The Alkaloids: Chemistry and Physiology, Volume VIII

THE ALKALOIDS Chemistry and Physiology

VOLUME VIII THE INDOLE ALKALOIDS

CONTRIBUTORS TO VOLUME Vm A. R. BATTERSBY E. COXWORTH B. GILBERT W. ASHLEYHARRISON A. HOFMANN H. F. HODSON

R.H. F. MANSKE J. E. SAXTON E. SCHLITTLER G. F. SMITH A. STOLL W. I. TAYLOR

THE ALKALOIDS Chemistry and Physiology Edited by

R. H. F. MANSKE Dominion Rubber Research Laboratory GueEph, Ontario,Canada

VOLUME V I I I THE INDOLE ALKALOIDS

1965 ACADEMIC PRESS

*

NEW YORK

*

LONDON

COPYRIGHT01965, BY ACADEMIC PRESSINC. ALL RIGHTS RESERVED. NO PART OF THIS BOOK MAY B E REPRODUCED I N A N Y FORM, B Y PHOTOSTAT, MICROFILM, OR A N Y OTHER MEANS, WITHOUT WRITTEN FERMISSION FROM THE PUBLISHERS.

ACADEMIC PRESS INC. 111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS INC. (LONDON) LTD. Berkeley Square House, London W.1

LIBRARY OF CONGRESS CATALOG CARD NUMBER : 50-5522

P R I N T E D IN T H E U N I T E D STATES OF AICERICA.

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

A. R. BATTERSBY, The Robert Robinson Laboratories, University of Liverpool, Liverpool, England (515) E. COXWORTH,Saskatchewan Research Council, Saskatoon, Saskatchewan, Canada (27) B. GILBERT, Centro de Pesquisas de Produtos Naturais, Faculdade Nacional de Farmhcia, Rio de Janeiro, Brazil (335)

W. ASHLEY HARRISON,Dominion Rubber Research Laboratory, Guelph, Ontario, Canada (679)

A. HOFMANN, Pharmaceutical-Chemical Research Laboratories, Sandoz Limited, Basel, Switzerland (725)

H. F. HODSON, The Wellcome Research Laboratories, Beckenham, Kent, England (515) R. H. F. MANSKE,Dominion Rubber Research Laboratory, Guelph, Ontario, Canada (47,55,581,679,693)

J. E. SAXTON, The University, Leeds, England (1, 59, 93, 119, 159, 673)

E. SCHLITTLER, Research Department, CIBA Pharmaceutical Company, Division of CIBA Corporation, Summit, New Jersey (287) G. F. SMITH,The University, Manchester, England (591)

A. STOLL,Pharmaceutical-Chemical Research Laboratories, Sandoz Limited, Basel, Switzerland (725)

W. I. TAYLOR,Research Department, CIBA Pharmaceutical Company, Division of CIBA Corporation, Summit, New Jersey (203,237,249,269,785)

V

This Page Intentionally Left Blank

PREFACE The explosive advance in the chemistry of the indole alkaloids in recent years has been occasioned not only by their intrinsic interest as problems in chemistry but by the possibility that some at least might have therapeutic value. The last review by J. E. Saxton in Volume V I I of this series was confined to two hundred pages. The present volume consists of nearly four times as many pages and it is pertinent to note that most of the content is new. Material reported previously in these volumes is only summarized to the extent that the present volume is self-consistent but not repetitive. Though there is not yet any evidence that the interest in indole alkaloids has declined there is sufficient new material to warrant the preparation of an up-to-date summary. The Editor is of the opinion that most of the structural types that plants elaborate have now been discovered: and it is likely that new alkaloids will largely fallinto presently known types. Modern methods, depending as they do upon mass- and NMR-spectra, are extremely powerful tools in this field of structural investigations, and new types, if and when they occur, can be readily recognized as such. The twenty-two chapters in the present volume are to some extent an arbitrary division of the subject matter. I n consequence there is occasional overlapping but such as there is appeared to be essential in the interest of continuity and clarity. Cross references and references to previous volumes are designed to expedite exhaustive study of a particular subject. Literature references are listed in the bibliography in the order in which they appear in each chapter. The entries in the subject index are limited t o the important topics for each substance or group ; substances mentioned only incidentally are not included. The Editor is again most grateful to the authors of this volume. All are thoroughly competent in their chosen topics and their conscientious labors are a pleasure to acknowledge. R. H. F. MANSRE January, 1965

vii

This Page Intentionally Left Blank

CONTENTS LIST O F CONTRIBUTORS ...................................................

PREFACE ............................................................... OF PREVIOUS VOLUMES ......................................... CONTENTS

V

vii xiv

Chapter 1. The Simple Bases

J. E. SAXTON I. I1. I11. IV. V. VI . VII.

Introduction ...................................................... Abrine and Hypaphorine ........................................... Gramine and Its Derivatives ........................................ Tryptamine and Its Derivatives ..................................... Psilocin and Psilocybin ............................................. 5-Hydroxytryptamine and Its Derivatives ............................ Cryptolepiqe ...................................................... References .......................................................

1 2 4 8 10 12 19 21

Chapter 2. Alkaloids of the Calabar Bean

E. COXWORTH I. I1. 111. IV. V. VI .

Source of the Alkaloids ............................................. Alkaloids Isolated ................................................. Physostigmine .................................................... Postulated Riosyntheses of the Physostigmine Ring System .............. Geneserine ....................................................... Pharmacology .................................................... References .......................................................

27 27 28 41 42 43 44

Chapter 3. The Carboline Alkaloids

.

R. H . F MANSKE I. I1. 111. IV.

Introduction ...................................................... Occurrence ....................................................... Properties ........................................................ Structure ......................................................... References .......................................................

47 47 49 49 52

Chapter 4. The Quinazolinocarbolines R . H . F. MANSKE I. Introduction ...................................................... I1. Occurrence ....................................................... 111. Structure ......................................................... References ........................................................

ix

55 55 56

58

CONTENTS

X

Chapter 5 . Alkaloids of Milragyna and Ouroupuria Species J. E. SAXTON I. I1. I11. IV . V. VI. VII . VIII .

Occurrence ....................................................... Mitragynine ...................................................... Mitraphylline ..................................................... Uncarine-A and Formosanine (Uncarine-B)............................ Rhynchophylline (Mitrinermine) ..................................... Adifoline ......................................................... The Mass Spectra of Mitraphylline and Rhynchophylline . . . . . . . . . . . . . . . . Rotundifoline. Isorotundifoline (Mitragynol). and Speciofoline . . . . . . . . . . . References ........................................................

59 62 64 70 75 80 82 85 89

Chapter 6. Alkaloids of Gelsemium Species

J. E . SAXTON I. I1. I11. IV . V. VI.

Occurrence ....................................................... Gelsemine ........................................................ Sempervirine ..................................................... Gelsemicine ....................................................... Gelsedine ......................................................... Gelseverine ....................................................... References ........................................................

93 95 107 110 112 115 115

Chapter 7. Alkaloids of Picrulima Nitido

J. E. SAXTON

....................... I . Occurrence ............................ Akuammigine ..................................................... Akuammicine ..................................................... Pseudoakuammicine ................................... ...... Aknammidine (Rhazine)............................................ VI. Pseudoakuammigine ............................ VII. Akuammine ...................................................... VIII . Picraline ......................................................... IX. Akuammiline ..................................................... x. Aknammenine .................................................... References ............................................. I1. 111. IV. V.

119 120 123 130 131 134 145 147 155 155 155

Chapter 8. Alkaloids of Alstonicc Species

J. E. SAXTON

I. Occurrence

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

I1. Alstonine and Tetrahydroalsto

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

e ...................................

111. Alstoniline ..................................................... IV . Alstonidine ....................................................... V . Echitamine ....................................................... VI. Echitamidine ..................................................... VII . Villalstonine . . . . .......................................... VIII . Macralstonine ....................................................

159 162 170 173 174 191 194 195

xi

CONTENTS

I X . Macralstonidine ................................................... X . AlkaloidC ........................................................ References ........................................................ Addendum:Venenatine ....................... . . . . . . . . . . . . . . . . . . . . .

196 197 199 202

Chapter 9. The Iboga and Voacanqn Alkaloids W. I. TAYLOR

I. The Iboga Alkaloids .............................................. I1. The VoacanguAlkaloids ............................................ I11. Miscellaneous ..................................................... References ........................................................

203 225 231 233

Chapter 10. The Chemistry of the 2.2 '.Indolylquinuclidine Alkaloids

W. I. TAYLOR

I. Determina.tion of the Structure of the Alkaloids ........................ I1. Synthesis of Cinchonamine .......................................... I11. Stereochemistry of Cinchonamine .................................... IV. Miscellaneous ..................................................... References ........................................................ Chapter 11.

238 243 244 246 240

The Pentacerns and the Eburnamine (Hunteria)-Vicamine Alkaloids

W . I. TAYLOR I. The Pentaceras Alkaloids (Canthin.6.ones) ............................. I T. The Eburnamine (Hunteria)-Vincamine Alkaloids ...................... TI1. The Hunteriu and Pleiocarpn Alkaloids ............................... References ........................................................

250 253 262 267

Chapter 12. The Vinca Alkaloids

W. I. TAYLOR

I. I1. 111. IV.

The Alkaloids of Vinca rosea L....................................... .The Alkaloids of Vinca minor L...................................... The Alkaloids of Vinca difformis Pourr. and V. major L.................. The Alkaloids of Vinca herbacea and V . lnncea .......................... References ........................................................

272 278 280 282 282

Chapter 13. Rauwolfia Alkaloids with Special Reference To the Chemistry of Reserpine

E. SCHLITTLER I. RauwoZJia Species and Their Alkaloids ................................ I T. The Chemistry of the Reserpine Group ................................ T I 1. Synthetic Work ................................................... References ........................................................

287 300 316 327

xii

CONTENTS

Chapter 14. The Alkaloids of Aspidosperm. Diplomhyncus. Kopsia. Ochrosk. Pkiocarpa. and Related Genera B GILBERT I. Introduction ...................................................... I1. The Aspidospermine Group ......................................... 1 x 1. The Aspidofractinine Group ......................................... I V . The Aspidoalbine Group ............................................ V. The Condylocarpine Group .......................................... VI. Alkaloids Related to Akuammicine ................................... VII . The Uleine Group ................................................. VIII . Tetrahydro j3-Carboline and Related Alkaloids ......................... IX . Alkaloids of Unknown Structure ..................................... References ........................................................

336 337 420 445 453 463 469 482 504 505

Chapter 15. Alkaloids of Calabash Curare and Strychnos Species A . R . BATTERSBY AND H. F. HODSON I. Introduction ...................................................... I1. The Czo-Alkaloids ................................................. I11. The Dimeric Alkaloids of Calabash Curare ............................. References ........................................................

515 522 537 576

.

Chapter 16. The Alkaloids of Calycanthaceae R . H . F. MANSKE

I. I1. I11. IV. V. VI.

Introduction ...................................................... Occurrence ....................................................... Calycanthine ...................................................... Calycanthidine .................................................... Folicanthine and Chimonanthine .................................... Hodgkinsine ...................................................... Addendum ....................................................... References ........................................................

.

Chapter 17

581 581 582 585 586 588 588 588

Strychnos Alkaloids

.

G. F SMITH I. Strychnine and Brucine: Historical Survey ............................ 11. The Reactions of Strychnine. Brucine. and Their Derivatives and Degradation Products ................................................ 111. a- and p.Colubrines ................................................ Iv . The Total Synthesis of Strychnine ................................... v. Vomicine: Historical Survey V I. The Reactions of Vomicine and Its Derivatives and Degradation Products . . VII . Minor Alkaloids ................................................... References

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

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

592 599 642 642 647 648 663 666

Chapter 18. Alkaloids of Haplophyton cimicidum J. E. SAXTON Text ............................................................ Addendum ....................................................... References

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

673 078

678

CONTENTS

...

Xlll

Chapter 19. The Alkaloids of Geissospermurn Species

.

R . H . F. MANSKE AND W ASHLEYHARRISON

I. I1. I11. IV. V. VI .

Introduction ...................................................... Geissoschizoline (Pereirine) ......................................... Geissoschizine..................................................... Geissospermine .................................................... Other Alkaloids ................................................... Flavopereirine .................................................... References .......................................................

679 681 683 685 687 688 690

Chapter 20. Alkaloids of Pseudocinchona and Yohimbe R . H . F. MANSEE

I. I1. I11. IV . V.

Introduction ...................................................... Yohimbane ....................................................... Heteroyohimbane ................................................. Corynane (17,18.Secoyohimbane) .................................... Corynoxane ....................................................... References ........................................................ Chapter 21.

694 695 707 716 720 721

The Ergot Alkaloids

A. STOLLAND A. HOFMANN

I. The Biology of Ergot and a Short History of I t s Active Principles up to the Discovery of Ergotamine ......................................... I1. Structural Types with Tables of the Natural Ergot Alkaloids ............ 111. Lysergic Acid and Isolysergic Acid ................................... IV. Simple Lysergic Acid Amides ....................................... V. Peptide Alkaloids .................................................. VI. The Alkaloids of the Clavine Series ................................... VII. Biogenesis of the Ergot Alkaloids .................................... VIII . Derivatives of Ergot Alkaloids ...................................... I X . The Pharmacology and Therapeutic Use of Ergot Alkaloids and Their Derivatives .......................................................... References ........................................................

726 729 734 746 748 760 766 768 772 779

Chapter 22. The Ajmaline-Sarpagine Alkaloids

W. I. TAYLOR I. I1. I11. IV.

The Ajmaline Group ............................................... TheSarpagineGroup .............................................. Mass Spectra of the Ajmaline-Sarpagine Alkaloids ...................... Pharmacological Notes ............................................. References ........................................................

789 804 808 811 812

AUTHORINDEX .......................................................... SUBJECTINDEX

815 851

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

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 NELSONJ . 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. . . . . .

8. 8. 9. 10. 11. 12. 13. 14. 15.

. . . . . . .

. . . . . . .

. . . . . . .

1 15 91 107 165 271 375

Contents of Volume 11 The Morphine Alkaloids I BY H . L . HOLMES . . . . . . . . 1 The Morphine Alkaloids I1 BY H.L.HOLMES AND (IN PART) GILBERT STORK 161 Sinomenine BY H . L . HOLMES . . . . . . . . . . . . 219 Colchicine BY J. W . COOKAND J . D. LOUDON . . . . . . . . 261 Alkaloids of the Amaryllidaceae BY J. 1%'.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 I1 BY H. L. HOLMES . . . . . . 513 Contents of Volume 111

16. 17. 18. 19. 20. 21. 22. 23. 24 .

The Chemistry of the Cinchona Alkaloids BY RICHARDB . TURNERAND R . B . WOODWARD . . . . . . . . . . . . . . . 1 Quinoline Alkaloids. Other than Those of Cinchona BY H . T. OPENSHAW 65 The Quinazoline Alkaloids BY H. T . OPENSHAW . . . . . . . 101 Lupin Alkaloids BY NELSONJ . LEONARD . . . . . . . . . 119 The Imidazole Alkaloids BY A. R . BATTERSBY AND H . T . OPENSHAW. 201 The Chemistry of Solanum and Veratrum Alkaloids B Y V . PRELOG AND 0. JEGER . . . . . . . . . . . . . . . . . . 247 P-Phenethylamines BY L . RETI . . . . . . . . . . . . 313 . . . . . . . . . . . . . 339 Ephreda Bases BY L . RETI JANOT . . . . . . . 363 TheIpecac Alkaloids BY MAURICE-MARIE

Contents of Volume I V 25. The Biosynthesis of Isoquinolines BY R . H . F. MANSKE . . . . . 26. Simple Isoquinoline Alkaloids BY L . RETI . . . . . . . . . 27. Cactus Alkaloids BY L . RETI . . . . . . . . . . . . . 28. The Benzylisoquinoline Alkaloids BY ALFREDBURGER . . . . . R . ASH29. The Protoberberine Alkaloids BY R . H . F. MANSKEAND WALTER FORD . . . . . . . . . . . . . . . . . . . 30. The Aporphine Alkaloids BY R . H . F. MANSKE . . . . . . . . 31. The Protopine Alkaloids BY R . H . F. MANSKE . . . . . . . . xiv

1 7 23 29 77 119 147

CONTENTS O F PREVIOUS VOLUMES

XV

CHAPTER 32. Phthalideisoquinoline Alkaloids

. .

33. 34. 35 . 36 . 37

BY JAROSLAV STANEKAND R. H F MANSKE . . . . . . . . . . . . . . . . . . Bisbenzylisoquinoline Alkaloids BY MARSHALL KULKA . . . . . The Cularine Alkaloids BY R . H F MANSKE a-Naphthaphenanthridine Alkaloids BY R . H . F. MANSKE . . . . The Erythrophleum Alkaloids BY G. DALMA . . . . . . . . The Aconitum and Delphinium Alkaloids BY E . S. STERN. . . . .

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 MARCEL J. DALLEMAGNE. . . Antimalarials BY 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 Alkdoids of Unknown Structure BY R . H. F. MANSKE.

.

. .

. . . . . . . .

167 199 249 253 265 275

Contents of Volume V . . . .

. . . .

. . . .

1 79 109 141 163 211 229 243 265 295 301

1 31 35 123 145 179 219 247 289

. . . . . . .

. . . . . . .

. . . . . . .

. . . . .

. . . . .

. . . . .

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

9.

Alkaloidsin theplant BY K . MOTHES . . . . . . The Pyrrolidine Alkaloids BY LEOMARION . . . . . Senecio Alkaloids BY NELSONJ. LEONARD. . . . . The Pyridine Alkaloids BY LEOMARION . . . . . . TheTropaneAlkaloidsBY G. FODOR . . . . . . The Strychnos Alkaloids BY J. B. HENDRICKSON. . . The Morphine Alkaloids BY GILBERTSTORK . . . . Colchicine and Related Compounds BY W . C. WILDMAN . Alkaloids of the Amaryllidaceae BY W. C. WILDMAN . .

. . . . .

. . . . . . . .

. . . . . . . .

Contents of Volume V I I 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20 . 21. 22 .

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 . . . . . . . Lupin Alkaloids BY NELSONJ. LEONARD . . . . . . . . . Steroid Alkaloids: The Holarrhena Group BY 0. JEGER AND v . PRELOG Steroid Alkaloids: The Solanurn Group BY v . PRELOG AND 0. JEGER . Steroid Alkaloids: Veratrum Group BY 0. JEGER AND V . PRELOG . . The Ipecac Alkaloids BY R . H . F. MANSKE . . . . . . . . . . . . . . . . . Isoquinoline Alkaloids BY R. H . F. MANSKE Phthalideisoquinoline Alkaloids BY JAROSLAV STAN~K Bisbenzylisoquinoline Alkaloids BY MARSHALLKULEA . . . . . The Diterpenoid Alkaloids from Aconitum, Delphinium, and Garrya SpeciesBYE. S. STERN. . . . . . . . . . . . . . 23. The Lycopodium Alkaloids BY R . H . F. MANSKE . . . . . . . 24. Minor Alkaloids of Unknown Structure BY R . H . F. MANSKE . . . .

. . . . .

1 201 229 247 253 319 343 363 419 423 433 439 473 505 509

This Page Intentionally Left Blank

-CHAPTER

1-

THE SIMPLE BASES

J. E. SAXTON The University, Leeds, England

I. Introduction .......................................................

1

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

2

11. Abrine and Hypaphorine..

111. Gramine and Its Derivatives.. ........................................ A. Gramine, 3-Aminomethylindole, and 3-Methylaminomethylindole. B. Donaxarine .....................................................

4 4

..

.................. Psilocin and Psilocybin. ........................................ 5-Hydroxytryptamine and I t s Derivatives.. .....................

IV. Tryptamine and Its Derivatives..

V. VI.

8

..... ...

A. 5-Hydroxytryptamine.. ................... . . . . . . . . . . . . . . . . . . . . . . . B. 5-Methoxy-N-methyltryptamine and 5-Methoxy-N,N-dimethyltrypt............ amine .............. C. Bufotenine ......................................................

....................... References .........................................................

VII. Cryptolepine.

8

10 12 12 16

16 19 21

I. Introduction I n view of the vital importance of indole-%acetic acid as a plant growth hormone, and the central position occupied by tryptophan both as a constituent of plant proteins and as the common biogenetic precursor of the complex indole alkaloids, it is not surprising to find that several simple derivatives of indole, which are presumably closely related to the roufes of biosynthesis and metabolism of indoleacetic acid or tryptophan, occur widely in the vegetable kingdom. Indole itself has been isolated from the flowers of many Jasminium and Citrus species ( l ) ,from Robinia pseudacacia L. (2), Cheirunthus cheiri L. (3), Narcissus jonquilla L. (4), and Chimonanthus fragrans Lindl. (5); it appears t o be an essential constituent of the perfumes of these flowers. Reports have also been made of the isolation of indole from nonfloral material, e.g., Celtis reticulosa Miq. (6) and Thlaspi arvense L. (7), but these have been criticized on the grounds that the indole was probably the result of bacterial action on a labile indole precursor (8). Numerous other claims 1

2

J. E. SAXTON

for the presence of indole in plants have been based solely on the color reactions given either by the plant extracts or by the volatile constituents of the oils obtained from the blossoms by enfleurage. The origin of indole in plants is not yet established; it was earlier suggested that it might be a degradation product of tryptophan, but this possibility does not appear to have been investigated. I n some plants, a t least, the reverse may be true, and indole may be converted into tryptophan by combination with serine. Whether this is the principal mode of biosynthesis of tryptophan remains to be determined; it is perhaps more likely that indole and tryptophan are products of alternative pathways of metabolism of indole-3-glycerol phosphate, and that conversion of the last-named into tryptophan does not proceed by way of indole. Indole-3glycerol phosphate may well be the vital intermediate between anthranilic acid and the naturally occurring derivatives of indole. Thus far, there is very little evidence for this route of biosynthesis in higher plants but it is well established in certain microorganisms, e.g., Escherichia coli and Neurospora (9). It would be dangerous to assume by analogy that the same route is used in the higher plants; nevertheless, it remains an attractive possibility in the absence of any evidence for an alternative. I n this connection, it is of interest that methyl anthranilate accompanies indole in the flowers of the jasmine and the bitter orange ( l ) ,and in Robinia pseudacacia (2). Although the natural occurrence of indole and the biosynthesis of the indole ring system are of importance and relevance to the wider question of the biosynthesis of the complex indole alkaloids, indole will not be discussed in detail here, as it is not an alkaloid. For a comprehensive and critical account of the occurrence of indole and its simple derivatives in plants, the reader is referred to the article by Stowe (8). t-Tryptophan is the ubiquitous indole derivative in plant proteins, and similarly need not be discussed a t length. However, it is satisfying to note in passing that the early assumption that tryptophan is the biogenetic precursor of all the indole alkaloids has been substantiated by the comparatively few radioactive tracer experiments that have so far been carried out, e.g., the formation of the ergot alkaloids in Claviceps purpurea (lo), and of ajmaline, reserpine, and serpentine in RauwolJia serpentina Benth. ex Kurz. (11).

11. Abrine and Hypaphorine

Abrine (I),the N,-methyl derivative of L-tryptophan, occurs in the seeds of the jequirity (Abrus precatorius L.) (12, 13); so far it has not

1.

3

THE SIMPLE BASES

been obtained from any other botanical source. The amino acid abrine must not be confused with abrin, the toxic protein mixture obtained from the same seeds, which was isolated and named as early as 1884 (14). COzH I

@

MeNH-C-H

C8, I

Me3N-C-H I

I Abrine

I1 Hypephorine

The constitution of abrine was proved by methylation with methyl iodide and methanolic sodium hydroxide, which gave the same methyl ester methiodide as did similar treatment of L-tryptophan (15). Since the product was almost completely racemized ( l 6 ) , this did not establish the configuration of the asymmetric center. The configurational identity of abrine and L-tryptophan was proved by Cahill and Jackson (16), who obtained the same optically pure methyl betaine (11) from both abrine and L-tryptophan by methylation with methyl iodide and methanolic sodium hydroxide. Under the appropriate conditions, the racemic methyl ester methiodide crystallized out, and was removed ; the methyl bktaine (11)which remained unesterified also escaped racemization, and was recovered from the mother liquors. The optically active methyl betaine (11)is identical with hypaphorine (16,17),which occurs widely in the seeds of Erythrina species. It was first isolated by Greshoff (18) from the seeds of E . subumbrans Merrill (Hypaphorus subumbrans Hassk.) ; much later, it was discovered in the seeds of E . variegata var. orientalis (L.) Merrill ( E . indica Lam.) (19, 20) and in E . cristagalli L. (21). More recently, the search for the major Erythrina alkaloids has revealed the presence of hypaphorine in the seeds of 23 other Erythrina species (22-28); according to Folkers et al. (25), it has been found in every Erythrina species so far examined. Occasionally, the amount of hypaphorine in the seeds is comparatively high; in E. ucanthocarpa E. Mey., for example, it is 5.8% (25), while in E . pallida Britton and Rose it is as high as 6.7% (26). Although many quaternary compounds exhibit curare-like activity, the physiological effects of Erythrina extracts are apparently not due to hypaphorine. The natural occurrence of hypaphorine is probably not confined to Erythrina species. Von Lippmann (29) isolated from beet shoots a substance with the appropriate physical and analytical properties which,

4

J . E. SAXTON

like hypaphorine, decomposed on being heated into indole and trimethylamine. Although final identification of this substance was not achieved, it Seems very probable that von Lippmann’s conclusion that it was hypaphorine is correct. The decomposition of hypaphorine into indole and trimethylamine occurs slowly in the rotting wood of Abrus precatorius L., and is responsible for its fecal odor. It also led van Romburgh (30) to propose the correct structure for hypaphorine, which was soon established by synthesis from L-tryptophan (17). The methyl ester of N,-dimethyl-L-tryptophan (IIa) also occurs naturally, and has recently been shown to be the major base of Pultenaea altissima F. Muell. ex Benth. (Leguminosae) (30a). COOMe

I

MeZN-C-H I

111. Gramine and Its Derivatives

A. GRAMINE,’&AMINOMETHYLINDOLE,AND 3-METHYLAMINOMETHYLINDOLE

H

I11 Gramine

The simplest well-authenticated indole alkaloid is gramine (111), which was originally isolated from chlorophyll-deficient barley mutants by von Euler and Hellstrom (31, 32). It was a t first believed that the presence of gramine in these mutants was genetically related to the chlorophyll deficiency (33), but this became untenable when gramine was shown to be a constituent of normal sprouting barley (Hordeum vulgare L.) (34,35).The Graminae have not been extensively investigated, so it is not yet known whether gramine occurs widely. However, the alkaloid donaxine, from the Asiatic sedge Arundo donax L. (36), has

1.

THE SIMPLE BASES

5

been shown to be identical with gramine (34, 37, 38); so far, this constitutes the only other recorded occurrence in the Graminae. I n other families, gramine has been isolated from the winged fruit of Acer rubrum L. (39), from the leaves of the silver maple, A , saccharinurn L. (40), and from Lupinus luteus L. (40a). Gramine is a monoacidic tertiary base which contains two methyl groups attached to nitrogen, and gives a typical indole UV-spectrum (32); it is optically inactive and possesses one active hydrogen atom (36). It was first formulated as 2-dimethylaminomethylindole, mainly on the basis of the close similarity of its UV-spectrum with that of 2-methylindole, but this was soon disproved by comparison of synthetic 2-dimethylaminomethylindole with gramine (41). Other formulations briefly 2-methyl-3-dimethylconsidered were 3-methyl-2-dimethylaminoindole, but none of these explained aminoindole, and 2-ethylmethylaminoindole, the absence of C-methyl groups (Kuhn-Roth) and, although the first accounted for the production of skatole by zinc dust distillation, this was not accepted as indicating the presence of a substituent a t position 3, owing to the drastic nature of the degradation and the poor yield of skatole obtained (38,41,42).The structure of gramine was then revealed fortuitously by Wieland and Hsing in an attempt to synthesize 3dimethylaminoacetylindole by reaction of indole magnesium iodide with dimethylaminoacetonitrile. Unexpectedly, the product, mp 134O, had the composition CllH14N2, and was identified as gramine (43). A second synthesis was later reported by Kiihn and Stein, who obtained a quantitative yield of gramine by the Mannich condensation of indole with formaldehyde and dimethylamine (44). As a Mannich base, gramine finds extensive application in preparative indole chemistry. Indeed, gramine is one of the few alkaloids which are more familiar as intermediates in organic synthesis, and which are readily accessible in high yield from cheap starting materials. I n alkaline media, gramine methiodide behaves as an effective alkylating agent, particularly in reactions with compounds containing an active methine or methylene group. For example, reaction of gramine methiodide with potassium cyanide gives indoleacetonitrile (45-47), which affords convenient preparations of indoleacetic acid (45, 47, 48, 49) and tryptamine (45). Condensation of gramine or its methiodide with the sodium derivative of acetamidomalonic ester yields ethyl a-acetamido-acarbethoxy-~-(3-indolyl)-propionate, which on hydrolysis and decarboxylation provides a valuable synthesis of ( )-tryptophan (50, 51). Alternatively, gramine will condense with ethyl nitroacetate a t 100" to give an intermediate which can also be converted readily into (k)tryptophan by appropriate transformations (52).

6

J. E. SAXTON

The biogenesis of gramine in barley has provided the subject for an interesting study using radioactive tracer techniques. Administration of 1-+ )-tryptophan-P-C14to sprouting barley led to the formation, in the leaves, of radioactive gramine, in which the activity resided specifically On the carbon atom of the methylene group, corresponding to the one originally labeled (53). When a mixture of ( & )-tryptophan-2-C14 and ( f)-trypto~han-P-C was ~ ~fed to sprouting barley, the gramine isolated contained radioactivity a t the methylene group and the 2-position only. Further, the ratio of these activities was the same as in the original tryptophan administered to the plant (54).These results establish beyond doubt that tryptophan is converted into gramine in barley by a process which does not involve fission of the indole-,&carbon linkage. Instead, fission must Occur between the CL and P carbon atoms of the tryptophan. The intermediates in this conversion have not yet been identified with certainty. Both 3-indolyl-/3-C14-pyruvicacid and 3-indolyl-P-C14-acrylic acid are converted in sprouting barley into radioactive gramine specifically labeled on the carbon atom attached t o the ring (55); however, the incorporation is very low, hence, these substances may not be direct intermediates. It may even be that they are first converted in the intact plant into tryptophan by enzymatic amination (56). This is consistent with the observation that the incorporation of 3-indolyl-8-acrylic acid and its conversion into gramine in excised barley shoots is much lower still (57), which suggests that the appropriate enzyme is not present in effective amounts except in the intact plant (56). Indole-3-glyoxylic acid, indole-3-aldehyde) and (perhaps surprisingly) indole-3-acetic acid are not converted by sprouting barley into gramine; however, the failure to incorporate indole-3-acetic acid was attributed to destruction of this compound before it reached the site of gramine synthesis (55). These studies have been taken a stage further by O’Donovan and Leete, who administered a mixture of ( & )-tryptophan-P-C14 and ( f )-tryptophan-P-H3 to intact barley seedlings (56). The radioactive gramine thus obtained was shown to contain the same ratio of C14 to tritium as the original tryptophan mixture, and it was further established that the radioactivity was present only in the methylene group of the side chain. These results prove very convincingly that the methylene group of tryptophan remains intact during its conversion into gramine under these conditions. Hence, 3-indolyl-/?-acrylic acid, indole-3aldehyde, and indole-3-glyoxylic acid cannot be precursors of gramine, since the intermediacy of these compounds would necessarily involve the loss of a part or the whole of the tritium attached to the P-carbon atom of the tryptophan side chain. Other conceivable intermediates, such as 3-indolyl-/3-pyruvic acid and indole-3-acetic acid, can also be eliminated,

1.

7

THE SIMPLE BASES

since the methylene hydrogen atoms in these compounds are located on carbon atoms attached to a carbonyl function, and would therefore be labile. All the above results are consistent with Wenkert’s recent suggestion (57a)that the biological conversion of tryptophan into gramine proceeds by condensation with pyridoxal phosphate (IIIa) with formation of a

IIIb

I11

Schiff’sbase (IIIb),which is then degraded by a reverse Michael reaction to the protonated 3-methyleneindolenine (IIIc). Addition of ammonia then yields 3-aminomethylindole, which on methylation affords gramine (111). This attractive hypothesis finds support in the recent isolation from barley seedlings of both 3-aminomethylindole and 3-methylaminomethylindole (57b).It is also supported by the demonstration that 3-aminomethylindole can be methylated t o 3-methylaminomethylindole

8

J . E. SAXTON

by ( -)-S-adenosyl-L-methionine in the presence of an enzyme preparation from barley shoots, and that the same system converts 3-methylaminomethylindole into gramine (57b). I n retrospect, the report that phenylalanine is not a precursor of tryptophan and therefore of gramine in barley is not surprising ; i t seems more probable that anthranilic acid is a precursor (5713).

B. DONAXARINE Donaxarine, C13HleN202, the minor alkaloid of Arundo donax, has been described on only one occasion, and little information is available concerning its structure. Apart from a positive pine splinter reaction and the fact that it occurs in the same plant as gramine, there seems little justification for its inclusion with the indole alkaloids. Donaxarine contains an N-methyl group and one active hydrogen atom, and is optically inactive. The function of the oxygen atoms is unknown, but they are not contained in phenolic hydroxyl groups or methoxyl groups (58, 59).

IV. Tryptamine and Its Derivatives

H IIId Tryptamine

The occurrence of tryptamine (IIId) in plants was first discovered by White, who isolated it from Acacia Jloribunda Sieb. and A . pruinosa Cunn. (60).Since that time it has been obtained from other Acacia species, namely, A . cultriformis Cunn., A . longifolia Willd., A . podalyriaefolia Cum. (61), A. acuminata Benth., A . cardiophylla Cunn., and A. vestita Ker. (62). Its presence has also been revealed in the inkcaF fungus, Coprinus micaceus Fr. (63), and in another fungus, Panaeolus foenisicii Pers. (63a).More recently, it has been discovered to be present in several edible fruits, namely, the tomato (64, 64a), plum, and eggplant, and also in traces in the orange (64a). There are also reports of its occurrence in mesquite (Prosopsis ,jul$/ora DC.) (65) and in lentils (Lens esculenta Moench., syn. L. culinare) (65a). Like gramine, tryptamine is more familiar as an intermediate in preparative indole chemistry than as an alkaloid. It was first synthesized

1. THE

SIMPLE BASES

0

by Ewins, who obtained it by Pischer cyclization of the phenylhydrazone of y-aminobutyraldehyde (66). It was later prepared by Majima and Hoshino by reaction of indole magnesium iodide with chloroacetonitrile, and reduction of the indoleacetonitrile so obtained (67). These preparations are now mainly of historical interest, having been superseded first by the preparative sequence involving gramine niethiodide (45), and more recently by Speeter and Anthony’s method, from indole via indole 3-glyoxylyl chloride and the corresponding amide (68, 69). This last method is of particular value, as with appropriate modifications it affords a convenient preparation of pure N,-substituted tryptamines using the same number of stages. A fifth method of synthesis of tryptamine involves the lithium aluminum hydride reduction of 3-P-nitroethylindole, prepared by the reaction of indole or indole magnesium bromide with nitroethylene (70).A related method utilizes the catalytic or electrolytic reduction of 3-P-nitrovinylindole, prepared by condensation of indole 3-aldehyde with nitromethane (70a). Finally, tryptamine may be obtained directlr from indole by reaction of indolyl magnesium bromide with ethyleneimine (70b). Dipterine, the N,-methyl derivative of tryptamine, occurs in two Asiatic members of the family Chenopodiaceae, Girgensohnia diptera Bge. (71, 72) and Arthrophytum leptocladum Popov (73), and also in the bark of Piptadenia peregrina Benth. (73a). Arthrophytum leptocladum also contains a closely related base, leptocladine (74), identified as Nbmethyltetrahydroharman by synthesis from dipterine and acetaldehyde (73). N,N-Dimethyltryptamine occurs more widely in nature, and is the simplest of several naturally occurring tryptamine derivatives which exhibit psychotomimetic activity. It was first identified as a constituent of the seeds and pods of P. peregrina and P . macrocarpa Benth. (Leguminosae) during an attempt to isolate the hallucinogenic principles present in the narcotic snuff prepared from these plants by certain American Indian tribes (75). The physiological activity of this snuff is only partly owing to dimethyltryptamine ; bufotenine is a second active constituent. Another plant which is used for a similar purpose is Prestonia amazonica (Benth.) Macbride (Haemadictyon amazonicum Benth.) (Apocyanaceae). A concoction from the leaves is consumed by some Colombian and Peruvian Indians for its hallucinogenic properties. Although the plant was earlier reported to contain two alkaloids (76), it seems probable that this was the result of botanical confusion with Banisteria caapi Spruce, which is used by the natives for the same purpose, often alone but sometimes mixed with Prestonia amazonica (77). I n later investigations using carefully identified P. amazonica, only

10

J. E. SAXTON

N,N-dimethyltryptamine was isolated ( 7 7 ) . N ,N-Dimethyltryptamine also occurs in the leaves of Lespedeza bicolor var. japonica (Leguminosae) (78), and in the roots of Mimosa hostilis Benth. (40, 79). The latter plant is also the source of an extract used by the local (Brazilian) Indians in their mysticoreligious ceremonies for its hallucinogenic properties. These rituals have been described by Gonqalves de Lima, who recorded the extraction of nigerine from Mimosa hostilis, but did not identify it as N,N-dimethyltryptamine (79). On account of the activity of these plant extracts and the isolation from them of N,N-dimethyltryptamine, the physiological activity of this base in humans is of interest. When injected intramuscularly, it causes hallucinations and illusions, which are characterized by their rapid appearance and brief duration (80). Apparently, dimethyltryptamine is rapidly metabolized and excreted mainly as indoleacetic acid, although the urine is enriched with 5-hydroxyindoleacetic acid ; whether this is the result of oxidation a t the 5-position or stimulation of the metabolism of serotonin in the brain is not yet known (80). The seeds of Piptadenia peregrina and P. macrocarpa also contain N,N-dimethyltryptamine oxide (75). Since N,N-dimethyltryptamine is readily oxidized on exposure to air, the oxide of this base may be an artifact.

V. Psilocin and Psilocybin OH

0%03H

A-,CHZCHpNhfe-zH

UJ

c3

QTJ,CIHzCHzNMez

H

IV Psilocybin

H V Psilocin

Psilocybin (IV) and psilocin (V) occur in several Mexican fungi, and, aside from mitragynine, represent the only derivatives of 4-hydroxyindole hitherto found in plants. Psilocybin was first isolated from Psilocybe mexicana Heim (81), and has since been obtained from P. caerulescens Mum. var. mazatecorum Heim, P. semperviva Heim et Cailleux, P. zapatecorum Heim, P. aztecorum Heim, and in Xtropharia cubensis Earle (82,83) and Panaeolus sphinctrinus (83a).All these fungi were of Mexican origin, but it is interesting that specimens of Stropharia cubensis procured from Thailand and Cambodia also contained psilocybin (82, 83). Psilocin occurs in very much smaller proportions, but has been detected in P. mexicana (81)) P. semperviva, P. aztecorum, and Stropharia cubensis

1.

11

THE SIMPLE BASES

(82). Other North American fungi which have recently been shown to contain psilocybin and psilocin are Psilocybe cyanescens and P . baeocystis Singer and Smith (83b, 83c) ; psilocybin also occurs in Gonocybe cyanopus (83b). Curiously, psilocin appears to be present in much larger amounts than psilocybin in P . baeocystis (83c). Both psilocybin, C12H17N204P, and psilocin are optically inactive, amphoteric substances, which exhibit UV-spectra closely similar to those of 4-hydroxyindole derivatives. Hydrolysis of psilocybin gives psilocin and one equivalent of phosphoric acid. Reaction of psilocybin with diazomethane gives dimethylpsilocybin, a neutral betaine which contains one saponifiable methoxyl group. The second methyl group introduced in the methylation is attached to nitrogen, since pyrolysis of dimethylpsilocybin gives trimethylamine ; psilocybin itself does not give trimethylamine on pyrolysis, and hence presumably contains a dimethylamino group (83, 84). The two carbon atoms which remain to be located were presumed to be present in an ethanamine side chain. Psilocin was (V), and therefore formulated as 4-hydroxy-AT,N-dimethyltryptarnine psilocybin as its 0-phosphoryl derivative (IV) ( 8 5 ); dimethylpsilocybin must accordingly be the betaine (VI). These formulations were substantiated by the synthesis of psilocin and psilocybin according to the illustrated sequence of reactions (83, 84, 85).

07SOQ OCH~CGH~

1. COCl

OCHzCsH5 I

I

\

N‘

H

H

-

COCONMez

LiAlHI

OCHzCsH5 I ~ ~ ~ C H z c H z N M e z

IH

HzlPd

(cs&CH~0)20PO

OH I

12

J. E. SAXTON

Since pre-Columbian times, many Mexican Indians have used narcotic and hallucinogenic drugs in their rituals ; in some remote parts of Mexico such drugs are apparently still used. I n recent years, it has been established that these drugs are prepared from various fungi, notably those belonging to the Psilocybe andstropharia genera (83, 86, 87,88,89). This discovery stimulated interest in the hallucinogenic constituents of these fungi, as a result of which psilocybin and psilocin were isolated. Ingestion of these fungi results in hallucinations and a state of intoxication (83,86,90); qualitatively, the effects are similar to those of mescaline and lysergic acid diethylamide (86, 91). The psychotomimetic activity of pure psilocybin is remarkably similar to that of Psilocybe mexicana extracts, and it is probable that the total activity of the Mexican drug prepared from this species can be ascribed to psilocybin (83, 92). This is not necessarily true of extracts of other fungi, however ; for example, P . yungensis is reported to be hallucinogenic, but it has been established that it does not contain psilocybin (93). I n P . sempervivu it has been demonstrated that tryptophan is a precursor of psilocybin (93a). It was simultaneously suggested that a similar oxidation of tryptophan or a tryptophan metabolite a t the 4-position constitutes an important intermediate stage in the biosynthesis of the ergot alkaloids from tryptophan.

VI. 5-Hydroxytryptamine and Its Derivatives

,!. 5 - H Y DROXYTRYPTAMINE

Since the discovery of 5-hydroxytryptamiiie (serotonin, enteramine, thrombocytin) (VII),and the demonstration of its physiological activity and its important function as a neurohormone, the possibility of its occurrence in plants has attracted much attention. It was first shown to be present in Mucunu pruriens DC. (cowhage), and is probably responsible for the intense irritation which results when cowhage comes into contact with the skin (94). This irritation could be a mechanical effect due t o the trichomes, but it is more likely to be the result of liberation of

1.

THE SIMPLE BASES

13

histamine. 5-Hydroxytryptamine is probably also the active irritant of Urtica dioicu L. (stinging nettle) (95), since it occurs to the extent of 0.02% in this species. It also occurs in Prosopsis juli$oru DC. (65), Gossypium hirsutum L., and Symplocarpus foetidus Nutt. (96), in the bark of Hippophae rhnmnoides L. (96a),and in several fungi belonging to the genus Panaeolus, namely, P . campanulatus (Fr.)Qu6let (P.linnaenus Imai) (83a, 97), P. acuminatus (Schff.ex. Fr.) Quklet, P . foenisicii Pers., P. semiovatus Fr., and P . subalteatus (Berk. et Br.) Sacc. (63a). Of much greater interest, however, are the reports of the presence of 5-hydroxytryptamine in several edible fruits, namely, the banana (64, 64a, 64b, 98, 99), tomato (64, 64a, 64b), pineapple (loo), plum, avocado, eggplant (64a), plantain (64a, 100a) and “Matoke” banana (both of which are varieties of M u s a paradisiaca L.) (100b), papaw (Carica papaya L.), passion fruit (Passijlorafoetida L.) (100a), and the walnut ( 1 0 0 ~ ) . The preparation of 5-hydroxytryptamine has been repeatedly investigated, and several convenient syntheses have been described ; in fact, virtually all the known routes to tryptamine derivatives have been employed. The first synthesis was an adaptation of the gramine route, starting from 5-benzyloxyindole and proceeding via 5-benzyloxygramine, 5-benzyloxyindoleacetonitrile, and 5-benzyloxytryptamine (10 1). I n common with all the other preparations involving 5-benzyloxytryptamine, the final stage, namely, debenzylation, was achieved by catalytic hydrogenation. Almost contemporaneously, a second synthesis of 5-hydroxytryptamine was reported via 5-benzyloxyiiidoleacetonitrile, prepared by the reaction of 5-benzyloxyindole magnesium iodide with chloroacetonitrile (102).An analogous route using 5-methoxyindole gave 5-methoxytryptamine, which was demethylated by means of aluminum chloride (103). The method of Speeter and Anthony (68) from 5-benzyloxyindole via 5-benzyloxyindoleglyoxylyl chloride and the related dibenzylamide affords a valuable preparation which proceeds in high over-all yield. A later synthesis involved the condensation of 5-benzyloxyindole-3-aldehyde, prepared by Vilsmeier-Haack formylation of 5-benzyloxyindole, with nitromethane ; reduction of the product with lithium aluminum hydride then afforded 5-benzyloxytryptamine ( 104, 105).A somewhat shorter method uses the reaction of 5-benzyloxyindole with nitroethylene a t loo”, which yields 3-(2-nitroethyl)-5-benzyloxyindole ; reduction with lithium aluminum hydride provides yet another route to 5-benzyloxytryptamine (106). Finally, in this series of preparations from 5-benzyloxyindole, a patented method describes the briefest synthesis hitherto reported, namely, the reaction of 5-benzyloxyindole magnesium bromide with ethyleneimine, which gives 5-benzyloxytryptamine directly (70a).

14

J. E. SAXTON

Several syntheses are on record which avoid the preparation of 5-benzyloxyindole ; in these procedures, the indole ring is usually formed after provision is made for the introduction of the ethanamine side chain. The first of these (107,108) was an adaptation of Ewins' original tryptamine synthesis. A subsequent route (109)started from ethyl a-cyano2,5-dimethoxycinnamate (VIII), which was prepared by condensation of 2,5-dimethoxybenzaldehydewith ethyl cyanoacetate. When this was boiled with potassium cyanide solution, addition of the elements of hydrogen cyanide was accompanied by hydrolysis of the ester function and decarboxylation, to give 2,5-dimethoxyphenylsuccinonitrile(IX). CN

IX

VIII CHzNH2

I

X

Hydrogenation of the latter gave the corresponding diprimary amine (X; R = M e ) , which on demethylation gave the phenol ( X ; R = H ) . Ferricyanide oxidation then gave 5-hydroxytryptamine in 25% over-all yield from 2,5-dimethoxybenzaldehyde(109). The synthesis developed by Sbramovitch and Shapiro (110) utilizes (XII) the formation of 6-methoxy- 1-keto-1,2,3,4-tetrahydr0-/3-carboline by the Fischer cyclization of the p-methoxyphenylhydrazone (XI) of 2,3-dioxopiperidine. Alkaline hydrolysis of XI1 and decarboxylation of the product (XIII) gave 5-methoxytryptamine (1l o ) , demethylation of which had previously been reported (103).

a:.NH 0 mqNH /

\OMe

MeO,/

H.COzH

H

XI

H XI1

O

1.

15

THE SIMPLE BASES

A different approach was adopted in the synthesis by Justoni and Pessina ( 111). The Japp-Klingemann reaction of p-benzyloxyphenylhydrazine with cyclopentanone carboxylic ester gave the p-benzyloxyphenylhydrazone of a-ketoadipic acid (XIV); Fischer cyclization of the corresponding dimethyl ester then yielded 5-benzyloxyindole-2-carboxylic-3-j3-propionic acid (XV). Decarboxylation of XV followed by Curtius degradation and debenzylation eventually afforded 5-hydroxytryptamine.

xv

XIV

Finally, mention may be made of two further syntheses, which employ as crucial stages the condensation of N-acetyl-5-benzyloxyindoxyl(112) or 5-methoxyisatin (113) with cyanacetic acid to give the intermediates XVI and XVII, respectively; these were then converted into 5-hydroxytryptamine by standard procedures.

XVIII

I n mammals 5-hydroxytryptamine is found in the brain, in the blood, and in the tissues of the stomach, intestines, and lungs; its function in all these sites has not yet been fully elucidated. However, there is little doubt that it plays an extremely important role in the central nervous system, and, in particillar, in the brain. The psychotomimetic activity of some drugs appears to be due to interference with the function of 5-hydroxytryptamine in the brain. It is also implicated in certain

16

J. E. SAXTON

abnormal pathological conditions ; some intestinal tumors contain appreciable amounts of 5-hydroxytryptamine, which is excreted as 5-hydroxyindoleacetic acid. The appearance of this acid in inordinate amounts in the urine is used in the diagnosis of such tumors. The biochemistry and pharmacology of 5-hydroxytryptamine have been repeatedly discussed (see, for example, Ref. 114 and 115). B.

5-METIIOXY -N-METHY LTRYPTAMINE AND

5-METHOXY -L~,I~-DIIvIETHYLTRYPTAMINE

A report that sheep fed on a perennial grass, Phalaris tuberosa L., developed a condition known as “staggers” led to the investigation of a related species, P . arundinacea L., and the subsequent isolation from it of 5-methoxy-N-methyltryptamine (XVIII ; R = H) (116). 5-MethoxyN,N-dimethyltryptamine (XVIII ; R = Me) also occurs naturally, and has been isolated from Dictyoloma incanescens DC. (40). Both bases have recently been shown to be present in the bark of Piptadenia peregrina Benth. (73a).

C. BUFOTENINE

XIX Bufotenine

Bufotenine, 5-hydroxy-N,N-dimethyltryptamine(XIX) occurs in the leguminous shrubs, Piptadenia peregrina (75, 117), P . macrocarpa Benth. (75),andP. colubrina Benth. (40).The seeds of the first two species have been used for centuries by certain Indian tribes of South America and the Caribbean islands as the source of a ceremonial, narcotic snuff, called cohoba, which is inhaled through a bifurcated tube. Some Brazilian Indians use the roasted seeds of P. colubrina for a similar purpose. Small doses of this snuff produce hallucinations and a kind of intoxication; excessive doses cause a violet temporary derangement. Whether bufotenine is the principal hallucinogen in these preparations has not yet been established, but it is certainly present in significant proportions ; P . peregrina seeds contain 0.94% and P . colubrina seeds as much as 2.1 yo of bufotenine. Intravenous injection of bufotenine is reported to cause hallucinations ( 1 18), but the possibility that inhalation of Piptadenia

1. THE SIMPLE

17

BASES

extracts can result in the absorption of sufficient bufotenine to cause these hallucinations has been refuted (119). It is therefore suggested that the hallucinogenic activity of the ancient Indian snuff was due to some more potent extraneous material introduced, or generated chemically, during preparation (119). The isolation of bufotenine from vegetable sources demonstrates its ubiquitous nature. It also occurs in the secretion of the parotid gland of the toad (Bufovulgaris Laur.) and several other Bufo species (120-125), in certain fungi [Amanita mappa Batsch., A . muscaria L., A . pantherina DC. (126),A . porphyria (126a),A . tomentella, and A . citrinapers. (126b)], and in human urine (127). Bufotenine was first isolated from Bufo vulgaris in 1893 by Phisalix and Bertrand (l20),but it was not fully characterized. Handovsky (128) later isolated the same oil, and obtained a crystalline oxalate, among other salts, which appeared to have the formula, C14H18N206, and from which he deduced that the base had the composition CsHgNO. Since the base gave a pine splinter color test, it was assigned a structure (XX) based on pyrrole (128).Wieland et al. reinvestigated these toad secretions, and from the basic fraction isolated two interconvertible, crystalline picrates, mp 17S0, which were formulated as derivatives of a base, C ~ ~ H I ~ O The ZN~ free . base was not obtained crystalline, but since a relationship with hypaphorine was suspected from its general properties,

xx

XXI

XXII

the structure X X I was tentatively proposed (129). However, this hypothesis was considerably weakened by comparison of bufotenine with ind-iV-methyltryptophan, and was completely invalidated by its eventual crystallization and purification, when the molecular formula C12H160N2 was established (121). Bufotenine was known to contain a 3-substituted indole nucleus and a tertiary amino group; the weakly acidic properties were now shown to be due to a phenolic hydroxyl group. A free imino group was also present, since the base contained two active hydrogens and yielded a diacetate. These data were combined in the formula XXII, in which the position of the phenolic group was unspecified ; however, positions 4 and 7 were provisionally eliminated, since a t that time no derivatives of 4- or 7-hydroxyindole had been found among natural products. The synthesis of the two remaining

18

J. E. SAXTON

isomers was therefore undertaken. Methylation of 6-methoxytryptamine, already known in connection with investigations in the harmine series, with methyl iodide and thallium hydroxide, gave a quaternary iodide which coincided in melting point (182"-183') with 0-methylbufotenine methiodide (mp 183"-184"), and corresponded closely in physical and chemical properties, but which gave a depression of almost 40" of melting mp 183", point on admixture. 5-Methoxy-N,N-dimethyltryptamine, was subsequently synthesized from 5-methoxyindole by condensation of the Grignard derivative with chloroacetonitrile, followed by reduction with sodium and alcohol and methylation of the 5-methoxytryptamine with methyl iodide and thallium ethoxide. The product was shown to be identical with 0-methylbufotenine methiodide in all respects (121). The synthesis of bufotenine itself followed closely upon the proof of its structure. Hoshino and Shimodaira reduced the ethyl ester of 5-ethoxyindole-3-acetic acid by the Bouveault-Blanc procedure to the corresponding primary alcohol, which was treated with phosphorus tribromide and then dimethylamine, to give the ethyl ether of bufotenine, which was demethylated with aluminum chloride (130). I n a later synthesis, 2,5-dimethoxybenzyl cyanide (XXIII)was alkylated by Eisleb's method with dimethylaminoethyl chloride in the presence of sodamide to give l-(2,5-dimethoxyphenyl)-3-dimethylaminopropylcyanide (XXIV), which was then hydrogenated over Raney nickel to yield 2-(2,5-dimethoxyphenyl)-4-dimethylaminobutylamine (XXV ; R = Me). Demethylation of this with hydrobromic acid, followed by oxidation of the product (XXV ; R = H) with potassium ferricyanide yielded bufotenine (XIX) via the related quinone (109). Two further syntheses of bufotenine have since been reported. I n the first of these, 5-benzyloxyindole was treated with oxalyl chloride to give 5-benzyloxy-3-indoleglyoxylyl chloride, which was converted by reaction with dimethylamine into 5-benzyloxy-N,N-dimethyl-3-indoleglyoxylamide. Reduction of this with an excess of lithium aluminum hydride yielded 0-benzylbufotenine, which was subsequently debenzylated (68). The fourth synthesis uses the gramine route (131). 5-Benzyloxyindole was converted into 5-benzyloxygramine, and thence into 5-benzyloxyindole-3-acetic acid, by standard procedures. 0-Benzylbufotenine was prepared from this by conversion into the related acid azide, reaction with dimethylamine, and reduction of the amide with lithium aluminum hydride. Catalytic debenzylation over a palladium catalyst gave bufotenine, identical with that from Amanita mappa in all respects except melting point. Whereas bufotenine has been reported in several instances to have mp 146"-147" (68, 117,121), Stoll et al. (131) found that

1.

19

THE SIMPLE BASES

their sample melted at 138"-140° in spite of the most diverse and careful methods of purification. This recalls the behavior of tryptamine, which has been reported to exist in two forms, of mp 118' (132) and 145' (66). CN MeO,A,CHzCN

MeO\/'.,,CH.

u L O M e

CHzCHzNMez

L)\OMe

XXIII

XXIV CHzNHz

xxv The seeds of Piptadenia peregrina and P. macrocarpa also contain bufotenine N-oxide (75). Since some tertiary derivatives of tryptamine, e.g., N,N-dimethyltryptamine, are readily converted by aerial oxidation into the N-oxide, it is possible that bufotenine oxide may be an artifact, generated during the extraction procedure or chromatographic separation. However, the formation of bufotenine oxide from bufotenine has never been observed in the absence of a specific oxidizing agent ; hence, it may be a genuine constituent of the seeds.

VII. Cryptolepine Extracts of the Cryptolepis genus (Asclepiadaceae), which are shrubs indigenous to tropical Africa, have found application as stomachics and in the dyeing of textiles and leather. The sap is extremely bitter, and is characterized by the rapidity with which it turns deep red on exposure to air. The alkaloid cryptolepine was first isolated from the roots of Cryptolepis triangularis N.E.Br., from the Belgian Congo, by Clinquart (133), and later by Delvaux (134),who obtained a base, mp 193"-194", analysis of which corresponded to the formula C17H160N2. Dry distillation of cryptolepine with quicklime gave a colorless sublimate, C13H10-12N2, mp 242'-243", and a yellow sublimate, mp 225", which were not further investigated. Cryptolepine was later isolated from Nigerian Cryptolepis sanguinolenta (Lindl.) Schlechter, and differs from all known alkaloids in that it forms deep violet needles, which give rise to solutions which are violet to red, according to the solvent (135). Analysis of the base (mp of

20

J. E. SAXTON

hydrate, 166"-169"), and several of its salts, which are yellow, indicated .the molecular formula, ClsH12N2. The earlier formula of Delvaux is understandable in view of the tendency of cryptolepine to form solvates, particularly with water and alcohols. The alcohol-free base, which can be obtained as a hemihydrate after being dried a t 120"/0.01 mm, does not contain any G-methyl groups, but has one methylimino group. Selenium dehydrogenation gives a colorless base, base A, C1,=,H10NZ,mp 249"-251", which forms a yellow methiodide, identical with cryptolepine hydriodide. Hydrogenation of base A affords a tetrahydro derivative which can be converted, via its methiodide, into the corresponding methonitrate, identical with tetrahydrocryptolepine nitrate. Distillation of the alkaloid over soda lime gives a pale yellow base, base B, C16H12N2, mp 264"-265", which possesses a C-methyl group, and which presumably arises from cryptolepine by migration of the methyl group from nitrogen to carbon. Hydrogenation of cryptolepine can proceed in three ways, according to the conditions employed. I n methanol solution, using a platinum oxide catalyst, a yellow dihydro derivative is rapidly formed. When the resulting methanol solution is shaken with air, the violet color of cryptolepine reappears in a few secmds, thus rendering impracticable the isolation of the reduction product. This reduction can also be accomplished using sodium hydrosulfite, and is assumed to involve the reduction of a pyridine ring to a dihydro derivative. More vigorous hydrogenation of the base in acetic acid using a platinum catalyst yields tetrahydrocryptolepine, whereas hydrogenation of the hydrochloride under the same conditions leads to the slow formation of octahydrocryptolepine. The absorption spectra of cryptolepine and the two hydrogenation products show, in common, a shift toward longer wavelengths in alkaline solution. It can therefore be inferred that the chromophore, which probably consists of the two nitrogen atoms linked by a system of conjugated double bonds, remains essentially unaffected by the hydrogenation, which involves the saturation of carbon rings. Crypt,olepine must be tetracyclic, and a consideration of its UV-spectrum suggests that these four rings are linearly arranged. One of the few ring systems capable of meeting all these requirements is contained in the known base, quindoline (XXVI). This was shown by direct comparison to be identical with base A ; tetrahydroquindoline and quindoline methiodide were identified with tetrahydrobase A (XXVII) and cryptolepine hydriodide (XXVIII), respectively ( 135). Cryptolepine (XXIX) is therefore the anhydronium base corresponding to quindoline methiodide, and it is of interest that it had been synthesized more than 20 years before its first isolation from Cryptolepis triangularis (136, 137, 138).

1.

xxx

21

THE SIMPLE BASES

XXIX Cryptolepine

XXVIII

lcOH

XXVI

XXVII

The structure of the isomer obtained on soda lime distillation is uncertain, but it is probably best formulated as XXX, since its spectrum is closely similar to that of quindoline. Cryptolepine has a significant hypotensive activity, and has been reported to cause a marked and prolonged fall in blood pressure in dogs (139, 140). REFERENCES 1. A. Hesse, Ber. 32, 2611 (1899); 33, 1585 (1900); A. Hesse and 0. Zeitschel, J . Prukt. Chem. 66, 481 (1902); R. Tsuchihashi and S. Tasaki, J . Chem. Ind. (Tokyo)21, 1117 ( 1 9 1 8 ) ; J . Soc. Chem. Ind. (London) 38A, 117 (1919). 2. F. Elze, Chemiker Ztg. 34, 814 (1910). 3. E. Kummert, Chemiker Ztg. 35, 667 (1911). 4. H. von Soden,J. Prakt. Chem. 110, 273 (1925). 5. G. Louveau, Rev. Marques Parfum. Savon. 9, 622 (1931); Chem. Zentr. 103 (Pt.11), 931 (1932). 6. C. A. Herter, J . Biol. Chem. 5, 489 (1909). 7. J. D. Ingle, Dairy Sci. Abstr. 2, 256 (1940). 8. B. B. Stowe, Fortschr. Chem. Org. Xaturstoffe 17, 248 (1959). 9. C. H. Doy, Rev. Pure AppZ. Chem. 10, 185 (1960). 10. D. Groger, H. J. Wenclt, K. Mothes, and F. Weygand, 2. Naturforsch. 14b, 355 (1959) ;W. A. Taber and L. C. Vining, Chem. Ind. (London)p. 1218 (1959);D. Groger, K. Mothes, H. Simon, H. G. Floss, and F. Weygand, 2. Naturforsch. 15b, 141 (1960) ; E. H. Taylor and E. Ra.mstad, Nature 188, 494 (1960).

22

J. E. SAXTON

11. E. Leete, Chem. Ind. (London) p. 692 (1960);J. Am. Chem. SOC.82, 6338 (1960). 12. N. Ghatak and R. Kaul, J. I n d i a n Chem. SOC.9, 383 (1932). 13. N. Ghatak, Bull. Acad. Sci. United Provinces Agra Oudh, India, 3, 295 (1934); Chem. Abstr. 29, 3344 (1935). 14. S. Martin, Proc. Roy. SOC. 42, 331 (1887). 15. T. Hoshino, Proc. I m p . Acad. (Tokyo) 11, 227 (1935); Chem. Abstr. 29, 6596 (1935); Ann. 520, 31 (1935). 16. W. M. Cahill and R. W. Jackson, J . Biol. Chem. 126, 29, 627 (1938). 17. P. van Romburgh and G. Barger, J. Chem. SOC.99, 2068 (1911). 18. M. Greshoff, Ber. 23, 3537 (1890); Mededeel. LandsPlantentuin 25, 54 (1898). 19. J. Maranon and J. K. Santos, Philippine J . Sci. 48, 563 (1932); Chem. Abstr. 26, 5609 (1932). 20. P. S. Rao, C . V. Rao, and T. R. Seshadri, Proc. I n d i a n Acad. Sci. Sect. A. 7, 179 (1938); Chem. Abstr. 32, 5240 (1938). 21. V. Deulofeu, E. Hug, and P. Mazzocco, J . Chem. SOC.p. 1841 (1939); V. Deulofeu, R. Labriola, E. Hug, M. Fondovila, and A. Kauffman,J. Org. Chem. 12, 486 (1947); V. Deulofeu, E. Hug, P. Mazzocco, and R. Labriola, Anales Asoc. Quim. Arg. 29, 121 (1941). 22. K. Folkers and F. Koniuszy, J . Am. Chem. SOC.61, 1232 (1939). 23. K. Folkers and F. Koniuszy, J . Am. Chern. SOC.62, 436, 1677 (1940). 24. R. T. Major and K. Folkers, U.S. Patent 2,407,713 (1946); Chem. Abstr. 41, 781 (1947). 25. K. Folkers, J. Shavel, and F. Koniuszy, J . Am. Chem. Soc. 63, 1544 (1941). 26. K. Folkers and J . Shavel, J. Am. Chem. SOC.64, 1894 (1942). 27. R.A. Gentile and R. Labriola, J . Org. Chem. 7, 136 (1942). 28. C. Lapiere, BuKSoc. Chim. Biol. 31, 862 (1949);J.Pharnz. Belg. 6, 71 (1951). 29. E. 0. von Lippmann, Ber. 49, 106 (1916). 30. P. van Romburgh, Koninkl. Ned. Akad. Wetenschap. 19, 1250 (1911); Chem. Zentr. 82 (Pt. I), 1548 (1911). 30% J. S. Fitzgerald, AustralianJ. Chem. 16, 246 (1963). 31. H. von Euler and H. Hellstrom, 2. Physiol. Chem. 208, 43 (1932). 32. H. von Euler and H. Hellstrom, 2. Physiol. Chern. 217, 23 (1933). 33. H. von Euler, H. Hellstrom and J. Hagen, Arkiv K e m i , Mineral. Geol. 11B,No. 36 (1934) ; Chem. Abstr. 28, 5502 (1934); H. von Euler, H. Hellstrom, and N. Lofgren, 2. Physiol. Chem. 234, 151 (1935). 34. K. Brandt, H. von Euler, H. Hellstrom, and N. Lofgren, 2. Physiol. Chem. 235, 37 (1935). 35. S. Kirkwood and L. Marion, J. Am. Chem. SOC.72, 2522 (1950). 36. A. P. Orekhov, S. S. Norkina, and T. Maximowa, Ber. 68, 436 (1935). 37. A. P. Orekhov and S. S. Norkina, Zh. Obshch. K h i m . 7, 673 (1937); Chem. Abstr. 31, 5801 (1937). 38. H. von Euler, H. Erdtman, and H. Hellstrom, Ber. 69, 743 [1936). 39. I. J. Pachter, J. Am. Pharm. Assoc. Sci. E d . 48, 670 (1959). 40. I. J. Pachter, D. E. Zacharias, and 0. Ribeiro, J. Org. Chem. 24, 1285 (1959). 40a. M. Wiewiorowski and H. Podkowinska, Bull. Acad. Polon. Sci. Ser. Sci. Biol. 10, 357 (1962). 41. H. von Euler and H. Erdtman, Ann. 520, 1 (1935). 42. H. Erdtman, Ber. 69,2482 (1936). 43. T. Wieland and C . Y. Hsing, Anit. 526, 188 (1936). 44. H. Kuhn and 0. Stein, Ber. 7 0 , 567 (1937).

1.

THE SIMPLE BASES

23

J . Thesing and F. Schulde, Chem. Ber. 85, 324 (1952). H. B. Henbest, E. R. H. Jones, and G. F. Smith, J . Chem. SOC.p. 3796 (1953). K. T. Potts and Sir Robert Robinson, J . Chem. SOC.p. 2675 (1955). H. R. Snyder, C. W. Smith, and J. M. Stewart, J . Am. Chem. SOC.66, 200 (1944). T. A. Geissman and A. Armen, J . Am. Chem. Soe. 74, 3916 (1952). H. R. Snyder and C. W. Smith, J. A m . Chem. Soc. 66, 350 (1944); E. E. Howe, A. J . Zambito, H. R. Snyder, and M. Tishler, J . Am. Chem. SOC.67, 38 (1945). 51. N. F. Albertson, S. Archer, and C. M. Suter, J . Am. Chem. Soe. 66, 500 (1944); 67, 36 (1945). 52. D. A. Lyttle and D. I. Weisblat, J . Am. Chem. SOC.69, 2118 (1947). 53. K. Bowden and L. Marion, Can. J . Chem. 29, 1037 (1951). 54. E. Leete and L. Marion, Can. J . C’hem. 31, 1195 (1953). 55. A. Breccia and L. Marion, Can. J . Chem. 37, 1066 (1959). 56. D. O’Donovan and E. Leete, J . Am. Chem. SOC.85, 461 (1963). 57. F. Wightman, M. D. Chisholm, and A. C. Neish, Phytochemistry 1, 30 (1961); Chem. Abstr. 56, 7722 (1962). 57a. E. Wenkert, J. Am. Chem. Soe. 84, 98 (1962). 57b. S. H. Mudd, Nature 189, 489 (1961). 57c. J. Massicot and L. Marion, Can. J . Chem. 35, 1 (1957). 58. J. Madinaveitia, Nature 139, 27 (1937). 59. J . Madinaveitia, J. Chem. SOC.p. 1927 (1937). 60. E. P. White, New Zealand J. Sci. TechnoZ. 25B, 157 (1944). 61. E. P. White, New ZeaZandJ. Sci. Technol. 33B, 54 (1951). 62. E. P. White, New Zealand J. Sci. Technol. 38B, 718 (1957). 63. P. H. List and H. Hetzel, Planta Med. 8, 105 (1960); Chem. Abstr. 54, 16556 (1960). 63a. V. E. Tyler, Jr. and A. H. Smith, Symposium, “Biochemie und Physiologie der Alkaloide.” Halle, 1960 (cited by H. G. Boit, “Ergebnisse der Alkaloid-Chemie bis 1960,” p. 986. Akademie Verlag, Berlin, 1961). 64. G. B. West, J . Pharm. Phurrnacol. 10, 589 (1958); 11,319 (1959); 11,Suppl. p. 275T (1959). 64s. S. Udenfriend, W. Lovenberg, and A. Sjoerdsma, Arch. Biochem. Biophys. 85, 487 (1959). 64b. R. K. Sanyal, P. K. Das, S. Sinha, and Y . K. Sinha, J . Phurm. P h a m c o l . 13, 318 (1961). 65. M. S. Fish, Unpublished work (1958), quoted in ref. 8. 65a. P. E. Pilet, Rev. Cen. Botan. 65, 605 (1958); Chem. Abstr. 53, 8330 (1959). 66. A. J. Ewins, J. Chem. SOC. 99, 270 (1911). 67. R. Majima and T. Hoshino, Ber. 58, 2042 (1925). 68. M. E. Speeter and W. C. Anthony, J . Am. Chem. SOC. 76, 6208 (1954). 69. F. V. Brutcher and W. D. Vanderwerff, J . Org. Chem. 23, 146 (1958). 70. W. E. Noland and P. J. Hartman, J . Am. Chem. SOC.76, 3227 (1954). 70a. T. Kametani and K. Fukumoto, Japan. J. Pharm. Chem. 33, 83 (1961); Chem. Abstr. 55, 19897 (1961). 70b. R. Bucourt, J. Valls, and R. Joly, U.S. Patent 2,920,080 (1960); Chem. Abstr. 54, 13018 (1960). 71. N. K. Yurashevskii and S. I. Stepanov, Zh. Obshch. Khim. 9, 2203 (1939); Chem. Abstr. 34, 4071 (1940). 72. N. K. Yurashevskii, Zh. Obshch. K h i m . 10, 1781 (1940); Chem. Abstr. 35,4016(1941). 73. N. K. Yurashevskii, Zh. Obshch. Khim. 11, 157 (1941); Chem. Abstr. 35, 5503 (1941). 73a. G. Legler and R. Tschesche, Nuturwissenschaften 50, 94 (1963).

45. 46. 47. 48. 49. 50.

24

J. E. SAXTON

74. N. K. Yurashevskii, Zh.Obshch. Khim. 9, 595 (1939); Chem. Abstr. 33, 7800 (1939). 75. M. S. Fish, N. M. Johnson, and E. C. Homing, J . Am. Chem. SOC.77, 5892 (1955). 76. A. M. Rarriga Villalba, J . Soc. Chem. I n d . (London)44, 205T (1925). 77. F. A. Hochstein and A. M. Paradies, J . Am. Ckem. SOC.79, 5735 (1957). 78. M. Goto, T. Noguchi, and T. TVatanabe,J . Phnrm. SOC. J a p a n 78, 464 (1958); Chem. Abstr. 52, 14082 (1958). 79. 0. Gongalves de Lima, Arquiw. I n s t . Pesquisas Agron. (Pernambucd) 4, 45 (1946). 80. St. SzBra, Ezperientia 12, 441 (1956). 81. A. Hofmann, R. Heim, A. Braek, and H. Kobel, Ezperientin 14, 107 (1955). 82. R. Heim and A. Hofmann, C'ompt. Rend. Acud. Sci. 247, 557 (1958). 83. A. Hofmann, R. Heim, A. Brack, H. Kobel, A. Frey, H. Ott, T. Petrzilka, and F. Troxler, Helu. Chim. Acta 42, 1557 (1959). 83a. V. E. Tyler, Jr. and M. H. Malone, J . Am. Phnrm. Assoc., Sci. E d . 49, 23 (1960). 83b. R. G. Benedict, L. R. Brady, A. H. Smith, and V. E. Tyler, Jr., Lloydia 25, 156 (1962); Chem. Abstr. 58, 5986 (1963). 83c. R. G . Benedict, L. R. Brady, and V. E. Tyler, Jr., J . Pharm. Sci. 51, 393 (1962). 84. d.Hofmann, A. Frey, H. Ott, T. Petrzilka, and F. Troxler, E'rperientia 14,397 (1958). 85. A. Hofmann and F. Troxler, Ezperientia 15, 101 (1959). 86. R. Heim, Compt. Rend. Acad. Sci.242, 965 (1956). 87. R. Heim, Compt. Rend. Acad. Sci. 242, 1389 (1956). 88. R. Heim, Compt. Rend. Acad. Sci. 244, 695 (1957). 89. R. Heim and R. Cailleux, Compt. Rend. Acad. Sci. 244, 3109 (1957). 90. R. Heim, A. Brack, H. Kobel, A. Hofmann, and R. Cailleux, Compt. Rend. Acad. Sci. 246, 1346 (1958). 91. J. Delay, J . Thuillier, H. Nakajima, and M. C. Durandin, Compt. Rend. SOC.Biol. 153, 244 (1959). 92. J. Delay, P. Pichot, T. LempBriBre, and P. Nicolas-Charles, Compt. Rend. Acad. Sci. 247,6235 (1958). 93. R. Heim, ActualitesPhurmacoZ. 12, 171 (1959); Chem. Abstr. 54, 15716 (1960). 93a. A. Brack, A. Hofmann, F. Kalberer, H. Kobel, and J . Rutschmann, Arch. Pharm. 294, 230 (1961). 94. K. Bowden, B. G. Brown, and J. E. Batty, Nature 174, 925 (1954). 95. H. 0. J . Collier and G. B. Chesher, Brit. J . Pharmacol. 11, 186 (1956); Chem. Abstr. 50, 14057 (1956). 96. C. Bulard and A. C. Lkopold, Compt. Rend. Acad. Sci. 247, 1382 (1958). 96a. G. P. Illenshikov, N. F. Petrova, and P. S. Masaagetov, U.S.S.R. Patent 137234 (1961); Chem. Abstr. 55, 26379 (1961). 97. V. E. Tyler, Jr. Science 128, 718 (1958). 98. P. Cartier, J. hforeau, and Y . Geffroy, Compt. Rend. SOC. Biol. 152, 902 (1958). 99. T. P. Waalkes, A. Sjoerdsma, C. R. Creveling, H. Weissbach, and S. Udenfriend, Science 127, 648 (1955). 100. D. U'. Bruce, Xature 188, 147 (1960). 100a. J. M. Foy and J . R. Parratt, J . Phurm. PharmacoZ. 12, 360 (1960). 100b. P. B. Marshall, J.Pharm. Pharmacol. 11, 639 (1959). 10Oc. E. Kirberger and L. Braun, Biochim. Biophys. Acta 49, 391 (1961). 101. K. E. Hamlin and F. E. Fischer, J . Am. Chem. Soc. 73, 5007 (1951). 102. M. E. Speeter, R. V. Heinzelmann, and D. I. Weisblat, J . Am. Chem. SOC. 73, 5514 (1951). 103. B. Asero, V. Colb, V. Erspamer, and A. Vercellone, Ann. 576, 69 (1952).

1.

THE SIMPLE BASES

25

E. H. P. Young, J . C'ltem. Soc. p. 3493 (1958). A. S. F. Ash and W. R. %Vragg,J . C'henz. Soc. p. 3887 (1958). W'. E. Noland and R. A. Hovden, J . Org. Chem. 24, 894 (1959). G. Bernini, Ann. Chim. ( R o m e )43, 559 (1953); Chenz. Abstr. 49, 2413 (1955). Z. J. Vejd&lekandL.Tuma,Cesk.Farm. 4,510(1955); Chem.Zentr. 128,10467 (1957). J. Harley-Mason and A. H. Jackson, J . Cltena. Soc. p. 1165 (1954). R. A. Ahramovitch and D. Shapiro, Chem. I n d . ( L o n d o n ) , p. 1255 (1955);J. Chem. Soc. p. 4589 (1956). 1 1 1 . R. Justoni and R. Pessina, Far?nnco (Puvia)E d . Sci. 10, 356 (1955); Chem. Abstr. 49, 13968 (1955). 112. C. E. Nenitzescu and D. Rlrileanu, Chem. Ber. 91, 1141 (1958). 113. S. Pietra, Fnrmuco (Puvia) E d . Sei. 13, 75 (1958); Chem. Abstr. 52, 13704 (1958). 114. G. P. Lewis, ed. " B-Hydroxytryptamine." Pergamon Press, New York, 1958; Symposiuni on 5-€Iydroxytryptamine, A7m. AT. Y . Acad. Sci.66, 592 (1957). 115. S. Udenfriend, P. A. Shore, D. F. Bogdanski, H. Weissbach, and B. B. Brodie, Recent Progr. Hormone Kes. 13, 1 (1957). 116. S. Wilkinson, J . Chem. SOC. p. 2079 (1958). 117. V. L. Stromberg,J. Am. Chem. Soc. 76, 1707 (1954). 118. H. D. Fabing and J. R. Hawkins, Science 123, 886 (1956). 119. IT'. J. Turner and S. Merlis, A . M . A . Arch. Neurol. Psychiat. 81, 121 (1959). 120. C. Phisalix and C. Bertrand, C'ompt. Rend. Acad. Sci. 116,1080 (1893); 135,46 (1902). 121. H. Wieland, W.Konz, and H. Mittasch, Ann. 513, 1 (1934). 122. H. Jensen and K. K. Chen, J . Biol. Chenz. 116, 87 (1936). 123. T7. Deulofeu and J. R. Mendive, Ann. 534, 288 (1938). 124. V. Ueulofeu and E. Duprat, J. Biol.Chem. 153, 459 (1944). 125. E. Titus and S. Udenfriend, Federation Proc. 13, 41 1 (1954). 126. T. Wieland, W. Motzel, and H. Merz, A n n . 581, 10 (1953). 126a. P. Catalfomo and V. E. Tyler, Jr., J. Pharm. S c i . 50, 689 (1961). 126b. V. E. Tyler, Jr., Lloydia 24, 71 (1961); C'hem. Abstr. 57, 15511 (1962). 127. F. M. Bumpus and I. H. Page, J . Biol. Chem. 212, 111 (1955). 128. H. Handovsky, Arch. Exptl. Pathol. Pharmakol. 86, 138 (1920); Chem. Abstr. 14, 2663 (1920). 129. H. Wieland, G. Hesse, and H. Mittasch, Ber. 64, 2099 (1931). 130. T. Hoshino and K. Shimodaira, Ann. 520, 19 (1935). 131. A. Stoll, F. Troxler, J . Peyer, and A. Hofmann, Helv. Chim. d c t a 38, 1452 (1955). 132. R. H. F. Manske, Cun. J . Bes. 5 , 592 (1931). 133. E. Clinquart, Bull. Acad. Roy. Med. Belg. [5] 9, 627 (1929); Chem. Abstr. 24, 1139 (1930). 134. E. Delvaux,J. Phnrm. Belg. 13, 955, 973 (1931). 135. E. Gellcirt, Raymond-Hamet, and E. Schlittler, Helv. Chim. Acta 34, 642 (1951). 136. F. Fichter and R. Boehringer, Ber. 39, 3932 (1906). 137. F. Fichter and H. Probst, Ber. 40, 3478 (1907). 138. F. Fichter and F. Rohner, Ber. 43, 3489 (1910). 139. Raymond-Hamet, Compt. Rend. SOC. Biol. 126, 768 (1937). 140. Raymond-Hamet, Compt. Rend. Acad. Sci. 207, 1016 (1938).

104. 105. 106. 107. 108. 109. 110.

This Page Intentionally Left Blank

-CHAPTER

2-

ALKALOIDS OF THE CALABAR BEAN E. COXWORTH Saskatchewan Research Council, Saskatoon, Saskatchewan, Canada

I. Source ofthe Alkaloids ...............................................

27

.................................................. 111. Physostigmine ...................................................... A. Elucidation of the Structure of Physostigmine. .......................

27

11. Alkaloids Isolated.

B. Some Comparisons with Other Alkaloids ............................ C. Syntheses of Physostigmine and Related Compounds . . . . . . . . . . . . . . . . . D. Other Attempted Syntheses of the Physostigmine Ring System . . . . . . . .

.............. Geneserine .......................................................... Pharmacology ...................................................... References ..........................................................

IV. Postulated Biosyntheses of the Physostigmine Ring System.

V. VI.

28 28 32 33 39

41 42 43 44

The chemistry of the alkaloids of the Calabar bean up to the year 1952 has been discussed fully in Chapter 13 (Section XI) of Volume I1 of this series. More recent work (up to 1960) has been described briefly in Chapter 10 of Volume VII of this series. The present chapter, then, will only briefly outline investigations reported prior to 1952, but will deal more extensively with work reported since that time.

I. Source of the Alkaloids The alkaloids are obtained from the seeds of the Calabar bean (Physostigma venenosum Balf.), which is a tropical woody vine. The beans were

used for many years as an ordeal poison in West Africa.

II. Alkaloids Isolated The alkaloids isolated from the Calabar bean are listed in Table I. The principal alkaloid found in the beans is physostigmine. 27

E. COXWORTH

28

TABLE I THE PHYSICAL CONSTANTSOF THE ALKALOIDS ISOLATED FROM CALABARBEAN

Compound

Formula

Physostigmine (alsoknownaseserine) Geneserine Eseramine" Isophysostigmine (unconfirmed) Physovenine" Eseridine [probably identical with geneserine (S)]

Melting point ("C)

C15H210aN3 Ci~H2103hi3 CitjH2503N4 C15H2102N3

105-106 128-129 245 200-202 (as the sulfate) C14H1803N2 123 C1sH2303N3 132

THE

[.ID -76" - 175"

Reference

1, 2 3

4 5

6 7

&Note added in proof: The structure of eseramine and reference t o the structure of physovenine can be found in Robinson and Spiteller (8a).

111. Physostigmine

I Physostigmine

Structure. The methyl carbamate of 5-hydroxy-1,2,3,3a,8,8a-hexahydro- 1,3a,8-trimethylpyrrolo[ 2,341indole. Derivatives. A table listing physical constants of derivaiives and transformation products of physostigmine is given in Volume I1 of this series (page 462 ff.). Pharmacological action. Widespread effects on the nervous system owing to the inhibition of the enzyme acetylcholinesterase.

A. ELUCIDATION OF THE STRUCTURE OF PHYSOSTIGMINE Physostigmine was first isolated by Jobst and Hesse ( 1 ) in 1864. The following year, Vee (2) obtained it in crystalline form for the first time and named it eserine (the other common name for the alkaloid). Straus (9) showed that physostigmine contained two N-methyl groups and behaved as a monoacidic tertiary base. Some of the early work on physostigmine indicated the unusual presence of a urethane grouping (carbamate, specifically -0CONHCH3)

2.

ALKALOIDS O F THE CALABAR BEAN

20

in the alkaloid. Thus, the basic hydrolysis of physostigmine yielded methylamine, carbon dioxide, and a phenolic, monoacidic tertiary base called eseroline (10). When sodium ethoxide was used, the products were eseroline and methyl urethane. Finally, physostigmine could be regenerated by the reaction of eseroline with methyl isocyanate (11). The methiodide of eseroline underwent an interesting cleavage reaction on heating to 200" (9). This reaction was of major importance in determining the structure of physostigmine. Thus, on heating the methiodide of eseroline to 200°, a new base, physostigmol (CloH11ON), was formed, the elements of C3H7N.CH31 being lost in the pr"3cess. The new base physostigmol showed the typical color reactions of an indole and contained an N-methyl grouping and a phenolic hydroxyl function. The ethyl ether of eseroline, eserethole, underwent a similar reaction (i.e., as the methiodide), the product in this case being the ethyl ether of physostigmol. The yield in this last reaction was 66% (12). Straus (9) proposed that the physostigmol was a hydroxy-1,3dimethylindole. Stedman (12 ) proved by syiiLhesis that physostigmol was in fact 5-hydroxy-1,3-dimethylindo!e; the route employed in this synthesis is outlined in structures I1 and 111.

I11

I1

The condensation of p-ethoxyphenylmethylhydrazine with a-ketoacid glutaric acid yielded 5-ethoxy-2-carboxy-l-methylindole-3-acetic (11),which was subsequently decarboxylated to physostigmol ethyl ether (111). This synthesis of physostigmol ethyl ether established the posjtion of the hydroxyl in eseroline and hence the position of the urethane group in physostigmine. i t also suggested that physostigmine was a substituted indoline. The ease with which physostigmol was formed from eseroline methiodide led Stedman and Barger (13) to believe that the P-methyl group in physostigmol was also present as a methyl group in physostigmine :

~zT"~ " 1z

CH3NHCOz

/

I

-CHzCHzN(TJH3)-

H3C Physostigmine

The C3H7N fragment must be joined to the indoline fragment as -CH2CT&N(GH3)since it was known thut physostigmine contained

30

E. COXWORTH

two tertiary N-methyl groups. Robinson suggested to Stedman and Barger (13) that formula I represented physostigmine ; this structure has since been confirmed by synthesis. Further evidence for this structure for physostigmine was obtained from other reactions on eserethole. It was found that eserethole (IV; R = CzH5) could be reduced (13, 14), one mole of hydrogen being absorbed. The reduction product, dihydroeserethole (V),was a secondary amine, indicating that a ring had been opened next to the nitrogen atom. When eserethole methiodide was treated with base, eseretholemethine was formed, a reaction which was reversible with hydriodic acid (15).

IV

V

This and other evidence (16) indicated that eseretholemethine (VI; R = CzH5) was a pseudo base or carbinol amine. Thus, on oxidation with silver nitrate an oxindole was obtained (dehydroeseretholemethine, VII; R = CzHs), which on subsequent Hofmann degradation and reduction of the Hofmann elimination product gave 5-ethoxy-l,3dimethyl-3-ethyloxindole (13). The methyl homolog ( V I I ; R = CHs),

VI

VII

i.e., dehydroesermetholemethine, was synthesized by King and Robinson (17), the route employed being outlined in structures V I I I to X. C H 3 0~ A ~ ~ c C ' & H ~ o c d %

+

H VIII

)

CH3 CH30 A dCHzCHzOCsH5 'JLgACH3

Jl

,

CH3 jCHzCHzOC.H5

C H 3 0 AV \ N / \ O

CH3 X

CH3

IX

+ VII; R

= CH3

IQ

2.

ALKALOIDS O P THE CALABAR BEAN

31

By heating 5-methoxy-2-methyl-3-/?-phenoxyethylindole (VIII) with methyl iodide under pressure, the indoleninium salt I X was obtained. Treatment of I X with sodium hydroxide yielded the corresponding 2-methyleneindoline which was oxidized with potassium permanganate to the oxindole X. The ether linkage on the ethyl side chain of X was thkn cleaved with hydrobromic acid to give the corresponding bromide, which on treatment with dimethylamine gave dl-esermetholemethine (VII). The dl-dehydroesermetholemethine thus obtained was resolved into the d- and 1-components and the 1-component proved to be identical with dehydroesermetholemethine obtained by degradation of physostigmine. The proof of structure of dehydroesermetholemethine gave considerable further support to the 3a-methylpyrrolo[2,3-b]indolestructure for physostigmine proposed by Robinson. The Robinson structure for the alkaloid also was derived easily from tryptophan, thus offering an obvious biogenetic route of synthesis by the plant (this will be discussed in more detail later). Although the investigations leading to the Robinson structure for physostigmine were completed in the 1930’s, it has only been in the last few years that the problem of the stereochemistry of the physostigmine ring system has been solved (18). I n the intervening years it was assumed that both cis and trans ring fusions were possible between pyrrole rings A and B of the physostigmine ring system (formulas Xa and Xb).

Xfb

Xb

Thus, certain differences between eseretholes (see, however, Section 111, C) prepared by two different routes were originally explained on the basis of one product having a cis ring fusion, the other having a trans ring fusion (19, see also Volume 11, Chapter 13, of this series). More recently, Witkop and Hill (20) concluded that physostigmine should have the more stable cis fusion structure Xa by analogy with the relative stabilities of cis and trans bicyclo[3,3,0]octane derivatives investigated by Linstead and Meade (21).Jackson (18) then pointed out that because of the fusion of the benzene ring to pyrrole ring A, the latter ring was planar, or nearly so, and, as a consequence of this planarity, only a cis ring fusion was possible between pyrrole rings A and B. Thus, although the physostigmine ring system has two asymmetric centers, only one pair of enantomers is possible (18, 22).

32

E. COXWORTH

I n agreement with this observation, all syntheses of the physostigmine ring system have given only one pair of enantiomorphs for each compound prepared (18). An X-ray crystallographic analysis of physostigmine to show this c i s ring fusion has not been carried out. However, X-ray analyses of two olher alkaloids which contain the same ring system, namely echitamine (23) and chimoiianthine (24), have shown this c i s ring fusion between the two pyrrole rings. The reactions of physostigmine and its derivatives in acid are of some interest. I n very dilute hydrochloric acid physostigmine and eserethole show a small hypsochromic shift in their UV-spectra. [This is in line with similar hypsochromic shifts displayed by other Ph-N,-Cz-N, alkaloids (22).] As the concentration of the acid is increased, the C2-Nt, bond is cleaved and an indoleniniam cation is formed (Xc). Confirmation of the

-.

CH3

xc

structure of Xc has been obtained from the NMR-spectra of physostigmine, esermethole, and deoxynoreseroline run in trifluoracetic acid ( 2 5 ) . This most recent work indicates that the acid media ring opening reaction is not restricted to Ph-N-C-N systems having an oxygen function on the benzene ring, as had been previously supposed ( 2 2 ) . As a corollary to these results, it has been suggested ( 2 5 ) that the hydrogenation of eserethole (13, 14), in which the Cz-N, bond is broken, proceeds via ring openirig t o the indoleninium (i.e., 3H-indolium) cation, which then undergoes hydrogenation. [Both reductions (13, 14) were carried out in acidic media.]

B. SOMECOMPARISONSWITH OTHER ALKALOIDS The decomposition of eserethole methiodide to yield physostigmol ethyl ether is analogous to the reaction of thebaine methiodide (XI)

+ . C H ~ C H Z N C HCH3I ~. fragment lost

XI

2.

33

ALKALOIDS O F THE CALABAR B E A N

when heated in the presence of acetic anhydride (26). I n both cases a --CHzCHzNCH3. CH3T fragment is lost and an aromatic system results. The reactions of physostigmine also show some similarities (e.g., facile reduction cleavage of the pyrrolo ring) to those of folicanthine, calycanthidine, and chimonanthine, which are now known to possess the structures of bis(pyrrolo[2,3,-B]indoles) ( 2 7 , 28,29, 30, 31, 32a, 32b; see also Chapter 16 of this volume).

C. SYNTHESESOF PHYSOSTIGMINE AND RELATED COMPOUNDS There have been various approaches to the synthesis of physostigmine ; these are outlined here, following the historical order in which they were developed. 1. The first satisfaztory approach to the synthesis of the physostigmine ring system was devised by King et al. (33). Their method is illustrated in structures XI1 to XIV. Synthesis of the indolenine XI1 was achieved

XI1

XIV

by condensation of p-ethoxyphenylhydrazine with y-phthalimidoa-methylbutyraldehyde, foIlowed by ring closure of the phenylhydrazone so obtained. Treatment of XI1 with methyl sulfate gave the N-methylated indoleniniuin salt XIII. The action of hydrazine hydrate, followed by the addition of hydrochloric acid, yielded the hydrochloride of dl-noreserethole (XIV). Methylation of XIV with methyl-p-toluenesulfonate gave a product of mp 79"-80". This was a t first believed t o be a

34

E. COXWORTH

diastereoisomer of eserethole (19), but later work (34) showed that the product of mp 79"-80" was a structural isomer of eserethole, namely, the

XIVa

indolenine XIVa. Several proposals have been advanced to account for the conversion of XIV into XIVa (18). 2 . Another method of building up the physostigmine ring system was devised by the same authors who devised the first (35). I n this method, the pyrrolo ring was closed by reaction of an appropriately aminosubstihted oxindole (XV) with phosphorus pentoxide to yield the cyclic amidine XVI, which was subsequently reduced to XVII and then niethylated to give a product claimed to be dl-esermethole (characterized as the methopicrate).

xv

XVI

XVII

At a later date, Kolosov et ul. (36) prepared dl-esermethole by another route (Julian and Pikl's procedure). Their material was not the same as that synthesized by King and Robinson, but did show some of the typical reactions of the physostigmine ring system. They therefore concluded that their material was indeed dl-esermethole, and that King and Robinson had not made dl-esermethole. They pointed out that the reduction of the cyclic amidine shown in structures XVI to XVII did not proceed under conditions which precluded reductive cleavage of the third ring. 3. The physostigmine ring system has also been synthesized by methylation of the Grignard compound prepared from suitably substituted tryptamines. This method, illustrated in structures XVIII and

2.

35

ALKALOIDS O F THE CALABAR BEAN

XIX, was developed by Hoshino and Kobayashi (37).The reaction of 5-ethoxy-N-methyltryptamine (XVIII) with ethyl magnesium iodide

uNj

CZH50-,/”,___

CH3

CHZCHZNHCH~

H XVIII

H

CH3

XIX

gave a Grignard compound, which when heated with methyl iodide yielded isonoreserethole (XIX).The hydrochloride of XIX, when heated with methyl iodide, was converted into dl-eserethole. Kobayashi (34) was able to resolve the dl-eserethole and proved that the ( -)eserethole was identical with material obtained from natural sources. 4. The first complete synthesis of physostigmine was carried out by Julian and Pikl(38).The pyrrolo third ring was formed in their procedure (see sequence XX to IV) by reductive ring closure of an appropriately amino-substituted oxindole. [The pyrrolo ring is not reductively cleaved under the conditions employed; the one reported synthesis to the contrary (39) is now believed to have actually given rise to the physostigmine ring system ( 1 8).] Compound X X was prepared by condensation

xx

XXI

XXII

of 5-ethoxy- 1,3-dimethyloxindolewith chloroacetonitrile in the presence of sodium ethoxide. The nitrile so obtained was catalytically reduced t o the primary amine XXT. Condensation of the amine function in XXI with benzaldehyde, followed by treatment with methyl iodide and hydrolysis of the methiodide so formed gave the N-methylated amine XXII. This compound (5-ethoxy-1,3-dimethyl-3-~-methylaminoethyloxindole) was easily separated into its optical antipodes. The 1-isomer was then ring closed smoothly by the addition of sodium to an alcoholic solution of the compound, the product being an I-esterethole (IV;

36

E. COXWORTH

R = C2H5) identical with the eserethole obtained from the alkaloid. By boiling a petroleum ether solution of I-eserethole (IV; R = CzH5), in which aluminum chloride was suspended, the phenolic base I-eserohe (IV; R = H) was obtained. For large-scale synthesis of physostigmine, Julian and Pikl (38) devised a simpler route to the key intermediate XXII. Condensation of the sodium salt of 5-ethoxy-l,3-dimethyloxindolewith 1,2-dibromoethane yielded 3-P-bromoethyl-5-ethoxy-1,3-dirnethyloxindole, which on heating with methylamine in methanol a t 100” gave the N-methylamine X X I I directly. Since it had already been shown by Polonovski and Nitzberg (11)that eseroline could be coilverted into physostigmine by the action of methylisocyanate, Julian and Pikl’s synthesis of I-eseroline constituted the first complete synthesis of physostigmine. There have been more recent syntheses of the physostigmine ring system which have employed the general method devised by Julian and Pikl. Kolosov et al. (36)prepared dl-esermethole by this procedure. By suitable modifications of the starting materials, these authors also succeeded in synthesizing dl-homoesermethole ( X X I I I; R = CH3) and dl-homoeseroline ( X X I I I ; R = H).

XXIII

Sugasawa and Murayama (40) have also prepared esermethole and homoesermethole by Julian and Pikl’s general method. It is interesting to note that when Sugasawa and Murayama attempted the reduction of 1-methyl-3-P-aminoethyloxindole (XXIV), the product was l-methyl-

XXVI

2.

ALKALOIDS O F THE CALABAR BEAN

37

3-6-aminoethylindole (XXV), and not the ring closed pyrr010[2,3-6]indole XXVI. The Japanese authors attributed this result t o the presence of an active hydrogen at position 3 of the oxindole. Hino and Ogawa (32b)recently found that lithium aluminum hydride can be used to effect the reductive cyclization of 3-aminoalkyloxindoles and have synthesized deoxyhomoeseroline by this method. 5. An entirely different approach to the synthesis of physostigmine was recently described by Harley-Mason and Jackson (41).This procedure was based on the earlier observation (42) that the ferricyanide oxidation of P-aminoethylhydroquinones(XXVIIa) led directly to the

XXVIIa

XXVIIIa

XXIXa

corresponding 5-hydroxyindoles (XXIXa). The reaction was presumed t o proceed through the quinone (XXVIIIa). This method of synthesizing 5-hydroxyindoles was successfully adapted to the synthesis of eseroline by the route outlined in structures XXVII-xxx-IV.

xxx

XXIX 1V;R = H

The acetophenone XXVII was condensed with ethylcyanoacetate (ammonium acetate-acetic acid catalyst) to yield the substituted acrylonitrile XXVIII. When XXVIII was heated in ethanol with

38

E. COXWORTH

potassium cyanide, addition of the elements of HCN (via Michael addition), saponification, and decarboxylation of the carbethoxy group all took place, the final product being the substituted succinonitrile XXIX. Catalytic reduction of XXIX gave the primary amine XXX (R = CH3, R‘ = H ) which was converted to the N-methylamine XXX (R = CH3, R’ = CH3) via condensation with benzaldehyde (2 moles), reaction of the di-Schiff base so formed with methyl iodide to form the dimethiodide, and hydrolysis of dimethiodide to give the desired N,N’-dimethyl-l,4-diaminobutaneXXX (R = CH3, R’ = CH3). On heating the last-named compound with hydrobromic acid, the free hydroquinone (XXX; R = H, R‘ = CH3) was obtained as the dihydrobromide. Treatment of this compound with potassium ferricyanide and sodium bicarbonate afforded dl-eseroline (IV; R = H ) directly, the yield in this double ring closure step being 30% (pure material). 6. In connection with a postulated route of biosynthesis of physostigmine (q.v.),Witkop and Hill (20) devised a new method of synthesis of the physostigmine ring system, outlined in structures XXXa to XXXIV. CH~CH(CHO)CHZC(NHCOCH~)(CO~CZH~)~ + XXXa

C~HSNHN=CHCH(CH~)CH~C(NHCOCH~)(CO + ~C~H~)~ XXXI

XXXII

XXXIII

XXXIV

The Michael addition of AcNHCH(C02C2H& to CHz=C(CH3)CHO in the presence of sodium ethoxide gave the aldehyde XXXa, which was condensed with phenylhydrazine to give the phenylhydrazone XXXI. Both rings were closed by heating the phenylhydrazone XXXI in refluxing acetic acid, and the product of the reaction was the bisnordeoxyeseroline XXXIII. Several hydrolysis and decarboxylation steps gave 2-carboxybisnordeoxyeseroline. Attempts to decarboxylate 2-

2.

ALKALOIDS O F THE CALABAR BEAN

39

carboxybisnordeoxyeseroline to bisnordeoxyeseroline were unsuccessful. Witkop and Hill assigned a cis configuration to the ring fusion of the two five-membered rings in the compounds they prepared. It is interesting to note that the nitrogen of the amide was sufficiently nucleophilic to be added to the indolenine in X X X I I or XXXII( H+),so that the product of the Fischer ring closure was the eseroline X X X I I I and not the intermediate indolenine XXXII. 7. Nakazaki (43) has recently described the synthesis of 1,2-dehydro2-ethoxy-9-methylbisnordeoxyeseroline (XXXVI), effected by the means illustrated in structures XXXV and XXXVI. The Grignard compound prepared from 2,3-dimethylindole and methyl magnesium iodide was treated with chloroacetonitrile to give the indolenine XXXV.

XXXV

XXXVI

Treatment of XXXV with hydrogen chloride in ethanol, followed by neutralization with sodium carbonate, gave the 172-dehydroeseroline XXXVI.

D. OTHERATTEMPTED SYNTHESES OF

THE PHYSOSTIGMINE RINGSYSTEM

1. I n 1956 Abramovitch (44, 45) reported a new method of synthesis of tryptamines, of which structures XXXVII t o XXXIX comprise an example. The phenylhydrazone XXXVII was synthesized via the

Japp-Klingemann reaction from 3-ethoxycarbonyl-2-oxopiperidine. Cyclization of the phenylhydrazone gave the tetrahydro-/3-carboline

40

E. COXWORTH

(XXXVIII), which on hydralysis gave the indole carboxylic acid XXXIX. Oecarboxylation of XXXIX then gave tryptamine. An attempt was made to extend this procedure to the synthesis of the physostigmine ring system, using the approach outlined in structures XL to XLIII (46).However, all attempts to cyclize the phenylhydrazone

H

I~ cI , + ( ~I 1

AZP,

C~H~NHN~\/NH

\/\Nfl\(NH

/I

0

-

0

XL

XLI

iz3c

CH3

A

A----LcH~cH~NH~

U L & rH / L NH J XLII

XLIII R = COOH, then H

XL were unsuccessful, and the indolenine XLI could not be obtained. [Robinson and Suginome have obtained 5-ethoxy-XLI by other means (47).1 2 . A further possible route to the physostigmine ring system (or a t least to the pyrroIo[2,3-b]indole ring system) would be ring closure of the appropriate amidrazone (XLIV) t o the a-aminoindole (XLV). (This

C,H,N-NH\N

U+W)

I R

\/

R’

N/\N

R

XLIV

R’

XLV

would then be analogous to the well-known Fischer indole synthesis employing ring closure of phenylhydrazones.) The synthesis of pyridino[2,3-b]indoles XLVI has recently been accomplished by this method (48) and the procedure has also been

XLVI

XLVIII n > 3

XLIX

2.

41

ALKALOIDS OF THE CALABAR BEAN

found useful in the synthesis of a-aminoindoles of types XLVIII and XLIX (49). However, attempts to extend this reaction to the synthesis of pyrrolo[2,3-b]indoles, and then to physostigmine, have not yet been successful.

IV. Postulated Biosyntheses of the Physostigmine Ring System I n the original paper which presented Robinson’s structure €or physostigmine Stedman and Barger (13) suggested a possible biosynthetic route t o the physostigmine ring system (see accompanying scheme). Tryptamine, produced by decarboxylation of tryptophan, was

Tryptophan

+

Tryptamine

---+

-

H3C ALCH2CH2NHCH3

d/LgJ

OH’

CH3

L

fl i

H3C CHzCHzNHCH3

W\N/
+

Deoxyeseroline (IV; R O replaced by H)

LI

conjectured t o be methylated to the quaternary hydroxide L. The isomer of L, i.e., the carbinol amine LI, was postulated to cyclize readily to the physostigmine ring system, specifically deoxyeseroline (XLIII ; R = CHs). Julian and Pikl (38) have also proposed a route of biosynthesis of physostigmine. They considered that the essential steps involved first oxidation and then methylation of tryptophan to an oxindole LII which was reductively ring closed to the physostigmine ring system. It will be H3C Tryptophan

--+

A-CH~CH~NHCH~ + Deoxyeseroline ( I V ; RO replaced by H)

UJ40 CH3 LII

noted that Julian and Pikl’s synthesis of physostigmine involved a laboratory imitation of the last step of their biosynthesis hypothesis. Most recently, Witkop and Hill (20) have proposed the biogenetic

42

E. COXWORTH

scheme outlined in structures L I I I and LIV. The 5hydroxytryptophan, obtained via enzymatic oxidation of tryptophan, was conjectured to undergo C-methylation in the presence of formaldehyde under acidic Tryptophan

--+

5-Hydroxytryptophan

+

LIV

LIII Eseroline (IV; R = H)

+

Physostigmine (I)

conditions. The indoleninium ion LIII, formed then, rapidly cyclized t o the amino acid LIV. Hydrogenolysis of the hydroxymethyl group in LIV, N-methylation, and decarboxylation would then give eseroline and, finally, physostigmine. Witkop and Hill noted that there was some analogy for the C-methylation step to be found in the recently reported C-methylation of dihydroberberine with formaldehyde (50).

V. Geneserine

LV Geneserine

Derivatives. A table listing derivatives and transformation products of geneserine is given in Volume I1 of this series (page 462 ff.). Elucidation of structure. Geneserine was found to behave in many ways similarly to physostigmine (3, 51). Thus, on heating t o 160" it yielded methyl isocyanate, indicating the presence of a urethane grouping. The alkaloid was decomposed by sodium ethoxide, yielding a new base called geneseroline, which behaved analogously to eseroline obtained in similar fashion from physostigmine. Significantly geneserine is reduced by zinc and acetic acid to physostigmine, which in turn can be oxidized by hydrogen peroxide to geneserine (51). Since geneserine differs in constitution from physostigmine only by the addition of one oxygen atom Polonovski concluded that geneserine was the N-oxide of physostigmine.

2.

ALKALOIDS O F THE CALABAR BEAN

43

VI. Pharmacology The main action of physostigmine is on the parasympathetic nervous system. It is now known (52, 5 3 ) that physostigmine produces its effect by inhibition of the enzyme acetylcholinesterase. The role played by acetylcholinesterase in the transmission of nerve impulses a t nerve endings (54,55) and in the conduction of impulses along nerve and muscle fibers (56, 55) has been described in detail elsewhere. Since the duty of acetylcholinesterase is to destroy (i.e., hydrolyze) acetylcholine, once that substance has served its function as chemical transmitter of a nerve impulse, the inhibition of the action of the enzyme will result in the accumulation of abnormally large quantities of acetylcholine. The effect of anticholinesterase drugs such as phystostigmine is in reality, then, the effect of abnormally high concentrations of acetylcholine a t those sites where it is normally released in the nervous system. The action of physostigmine (and other anticholinesterase drugs) is restricted to those sites of acetylcholine release to which it can readily penetrate (56), e.g., the nerve endings in the parasympathetic nervous system. In this connection it is interesting to note that physostigmine in particular has the ability to penetrate to many parts of the nervous system, and, as a consequence, affects many organs of the body (54). The widespread nature of the action of physostigmine has in fact limited its usefulness as a drug. Its application is now mainly confined to inducing myosis (constriction of the pupil of the eye). It has been shown (57) that the urethane group of physostigmine is responsible for its anticholinesterase action, although model compound studies have indicated that an amine function also appears to be necessary. A wide variety of urethanes of simple phenolic bases have been synthesized and evaluated as anticholinesterase drugs (see page 448 ff. of Chapter 13 of Volume I1 of this series). One of these substances, prostigmine (or neostigmine) (LVI) has now generally replaced physostigmine as an anticholinesterase drug (e.g., in the treatment of the disease myasthenia gravis).

The mechanism of the inhibition of acetylcholinesterase by physostigmine is related to the more general problem of the nature and mode of action of the active site (or sites) of the acetylcholinesterase molecule.

44

E. COXWORTH

The solution of this problem has been the object of a considerable number of investigations (54, 56, 58). It is generally considered that there is one especially reactive serine moiety at or near the esteratic site of action of the enzyme. For such powerful cholinesterase inhibitors as diisopropylfluorophosphate it has been shown that a stable covalent linkage is formed between the phosphate and this reacOive serine hydroxyl group (not necessarily present as a free hydroxyl in the enzyme) (59).A similar covalent linkage between urethane group and serine hydroxyl has been suggested for the physost,igmine-enzyme combination (59). Various partial structures have been suggested for the active esteratic site of the enzyme (60, 61, 62) but its actual nature remains unknown. REFERENCES 1. J. Jobst and 0. Hesse, A n n . 129, 115 (1864). 2. A. Vee, Jahresber. p. 456 (1865). 3. M. Polonovski and C. Nitzberg, Bull. SOC. Chim. France [4] 17, 244 (1915). 4. A. Ehrenberg, Verhandl. Vers. Deut. Ntf.Aerzte 11, 102 (1893); Chem. Zentr. 11, 439 (1894). 5. Ogui, Apotheker-Ztg. 19, 891 (1904). 6. A. H. Salway,J. Chem. Soc. 99, 2148 (1911). 7. C. F. Bohringer u. Sohne, Pharm. Post 21, 663 (1888); Chem. Zentr. 11, 1485 (1888); I&’. Eber, Pharm. Ztg. 37, 483 (1888); Chem. Zentr. 11, 1271 (1888). 8. F. Merck, Jahresber. 37, 39 (1924). 8a. B. Robinson and G. Spiteller, Chem. Iitd (Londorc), p. 459 (1964). 9. E’. Straus, A n n . 406, 332 (1914). 10. A. Petit and If. Polonovski, Bull. SOC.Chim. France [3] 9, 1008 (1893). Chim. France [4] 19, 27 (1916). 11. M. Polonovski and C. Nitzberg, Bull. SOC. 12. E. Stedman,J. Chem. SOC.125, 1373 (1924). 13. E. Stedman and G. Barger, J. Chem. SOC.127, 247 (1925). 14. M. Polonovski and M. Polonovski, Bull. Soc. Chim. France [4]23, 217 (1918). 15. G. Barger and E. Stedman, J . Chem. Soc. 123, 758 (1923). 16. M. Polonovski and M. Polonovski, Compt. Rend. Acad. Sci. 179, 57 (1924). 17. F. E. King and Sir Robert Robinson, 1.Chem. Soe. p. 326 (1932). 18. A. H. Jackson, Ph.D. Thesis, Cambridge Univ. (1954). 19. F. E. King and Sir Robert Robinson, J. Chem. SOC.p. 755 (1935). 20. B. Witkop and R. K. Hill, J . Am. Chem. SOC.77, 6592 (1955). 21. R. P. Linstead and E. M. Meade, J . Chem. Soc. p. 935 (1934). 22. B. Robinson, Chem. Ind. (London)p. 218 (1963). 23. J. A. Hamilton, T. A. Hamor, J. M. Robertson, and G. A. Sim, Proc. Chem. SOC. p. 63 (1961); H. Manohsr and S. Ramaseshan, Tetrahedron Letters p. 814 (1961). 24. I. J. Grant, T. A. Hamor, J. M. Robertson, and G. A. Sim, Proc. Chem. SOC.p. 148 (1962). 25. A. H. Jackson, Personal communication (1963). 26. K. W. Bentley, “The Chemistry of the Morphine Alkaloids,” p. 186. Oxford Univ. Press, London and New York, 1954.

2.

ALKALOIDS O F THE CALABAR BEAN

46

27. R. H. F. Manske and H. L. Holmes (eds.), “The Alkaloids,” Vol. 11, p. 98. Academic Press, New York, 1952. 28. H. F. Hodson and G. F. Smith, J . Chem. SOC.p. 1877 (1957). 29. H. F. Hodson, B. Robinson, and G. F. Smith, Proc. Chem. SOC.p. 465 (1961). 30. J. E. Saxton, W. G. Bardsley, and G. F. Smith, Proc. Chem. SOC. p. 148 (1962). 31. T. Hino, quoted in ref. 29. 32a. H. F. Hodson, G. F. Smith, and J. T. Wrobel, Chem. Ind. (London)p. 1551 (1958). 32b. T. Hino and K. Ogawa, Chem. Phurm. Bull. (Tokyo)9, 988, 991 (footnote) (1962). 33. F. E. King, N. Liquori, and Sir Robert Robinson, J . Chem. SOC.p. 1475 (1933); p. 1416 (1934). 34. T. Kobayashi, Ann. 539, 213 (1939); 536, 143 (1938). 35. F. E. King and Sir Robert Robinson, J . Chem. SOC. p. 1433 (1932). 36. & M.IKolosov, . L. I. Metreveli, and N. A. Preobrazhensky, J . Gen. Chem. USSR 23, 2143 (1953). 37. T. Hoshino and T. Kobayashi, Ann. 520, 11 (1935); and earlier references quoted therein. 38. P. L. Julian and J. Pik1,J. Am. Chem. SOC.57, 563, 755 (1935). 39. F. E. King and Sir Robert Robinson, J . Chem. SOC. p. 1433 (1932). 40. S. Sugasawa and M. Murayama, Chem. Pharm. Bull. (Tokyo) 6, 194, 200 (1958); Chem. Abstr. 53, 424, 425 (1959). p. 3651 (1954). 41. J . Harley-Mason and A. H. Jackson, J . Chem. SOC. 42. J. Harley-Mason and A. H. Jackson, J . Chem. SOC.p. 1165 (1954). 43. M. Nakazaki, Bull. Chem. SOC. Japan 32, 588 (1959); Chern. Abstr. 54, 7686 (1960). 44. R. A. Abramovitch and D. Shapiro, J . Chem. SOC.p. 4589 (1956). 45. R. A. Abramovitch, J . Chem. SOC.p. 4593 (1956). 46. R. A. Abramovitch, Can. J . Chem. 36, 354 (1958). 47. Sir Robert Robinson and .H. Suginome, J . Chem. SOC. p. 304 (1932). 48. H. Rapoport, D. S. Matteson, J. Gordon, andE. Coxworth, Unpublishedresults (1959). 49. H. Rapoport and E. Coxworth, Unpublished results (1959). 50. H. W. Bersch, Arch. Pharm. 283, 192 (1950); Chem. Abstr. 45, 1594 (1951). 51. M. Polonovski and C. Nitzberg, Bull. SOC.Chim. France [4] 17, 290 (1915); [4] 21, 191 (1917); [4] 23, 356 (1918). 52. E. Englhart and 0. Loewi, Arch. Exptl. Pathol. Phurmakol. 150, 1 (1930). 53. K. Mathes, J . Physiol. (London) 70, 338 (1930). 54. V. A. Drill (ed.), “Pharmacology in Medicine,” 2nd ed. McGraw-Hill, Toronto, 1958; esp. Chap. 26 by W. F. Riker, Jr. 55. D. Nachmansohn, “Chemical and Molecular Basis of Nerve Activity,” Academic Press, New York, 1959. 56. D. Nacbmansohn, Harvey Lectures, Ser. 49, 57 (1953-1954) (Pub. 1955). 57. E. Stedman, Am. J . Physiol. 90, 528 (1929). 58. Seefor example; (a) I?. D. Boyer, H. Lardy, and K. Myrbiick, “The Enzymes,” 2nd ed. Vol. IV, Chapters 28, 30, 31. Academic Press, New York, 1960; (b) M. L. Bender, Ghem. Rev. 60, 53 (1960); (c) G. E. Hein, reply by R. M. Krupka and K. J. Laidler, Nature 193, 1155 (1962) ; and earlier references quoted therein. 59. See f o r example: reference 58 (a),pp. 477, 518, 535. 60. G. R. Porter, H. N. Rydon, and J. A. Schofield, Nature 182, 927, 928 (1958). 61. S. A. Bernard, A. Berger, J. H. Carter, E. Katchalski, M. Sela, and Y. Shalitin, J . Amer. Chem. SOC.84, 2421 (1962). 62. E. Schatzle, M. Rottenberg, and M. Thurkauf, Helv. Chim. Acta 42, 1708 (1959).

This Page Intentionally Left Blank

-CHAPTER

3-

THE CARBOLINE ALKALOIDS R. H. F. MANSKE Dominion Rubber Research Laboratory, Guelph, Ontario, Canada

I. Introduction ........................................................

47

11. Occurrence .........................................................

47

111. Properties ..........................................................

49

IV. Structure ........................................................... A. Harmine ........................................................ B. Harmaline ....................................................... C. Harman ......................................................... D. Apoharmine .....................................................

49

References ..........................................................

49

50 51 51

52

I. Introduction The so-called harmala alkaloids, which are derivatives of carboline, were first found in Peganum harmala, and it was in a study of these that the nuclear structure in which an indole nucleus is fused with a pyridine nucleus was first recognized. Many of the polynuclear indole alkaloids contain the carboline (harman) nucleus, but it is often concealed in such a manner that it may be difficult to recognize. The following discussion will be limited to the derivatives of carboline in which no further nuclei occur. The numbering of the carboline ring system (I)adoptredhere is that of the Ring Index. It differs from some of the systems in use, but, since there has been considerable diversity and some confusion, it is hoped that the present system will be generally adopted. This chapter is a summary and extension of that dealing with the alkaloids of Peganum harmala, Volume 11,p. 393.

II. Occurrence Table I is a record of the occurrence of the harmala alkaloids. Their presence in plants of many families whose relation to each other is not 47

48

R. H. F. MANSKE TABLE I

PLANTS AND THEIRCONTAINEDALKALOIDS Plant and family

Peganum harmala L. Zygophyllaceae (Rutaceae)

Symplocos racemosa Roxb. Symplocaceae (Styracaceae) Sickingia rubra K. Schum. (Arariba rubra Mart.) Rubiaceae Eleagnus angustifolia L. Eleagnaceae Banisteria caapi Spruce Malpighiaceae Cabi paraensis Ducke Malpighiaceae Banisteriopsis inebrians Morton Malpighiaceae Leptactina densijora Hook. f. Rubiaceae Passijora incarnuta L. Passifloraceae

Alkaloid Harmine Harmaline Harmalol Harmidine Harman (Loturine) Harmine (Aribine) Tetrahydroharman Tetrahydroharmol N-Methyltetrahydroharmol Harmine Harmaline Tetrahydroharmine Harmine Harmine

Tetrahydroharmine Tetrahydroharman Harmine Harman Harmol Harmine Zygophyllum fabago L. Zygophyllaceae (first isolation of harmol) Harman Harmol 4-Methyl- and 3,4-dimethylArthrophytum leptocladum Popov 3,4,5,6-tetrahydro-4-carboline (Chenopodiaceae) Tetrahydroharman Petalostylis labicheoides R. Br. (Leguminosae) Strychnos melinoniana Baill. N,-Methylharman (Loganiaceae) Calligonum minimum Lipski Tetrahydroharman (Calligonine) (Polygonaceae) Base, ClzHloNz Base, CxzH140Nz

Reference 1

2 2, 3 4

5, 6

9, 10 11 11 12 13 13 14 15 16 16 17 17 17 18 18 18 19 20 21 21a

easily apparent indicates a biosynthesis that depends upon a ubiquitous' precursor and, as such, tryptophan is the more obvious. Indeed, harman and norharman have been identified in tobacco smoke. By tracer methods, it was shown that they arise a t least in part from tryptophan ( 2 2 ) . Norharman is often a product of the pyrolytic degradation of more

3.

49

THE CARBOLINE ALKALOIDS

complex indole alkaloids (23). Acid hydrolysis of proteins containing tryptophan also results in the formation of harman and some of its derivatives (24). 111. Properties The harmala alkaloids crystallize easily and have moderately high melting points. Even though the tetrahydro bases have an asymmetric carbon, only in one case was optical activity noted (tetrahydroharmine, mp 199"-200", [.ID + 32") (13). The bases in which the pyridine nucleus is a t least partly dehydrogenated exhibit a strong fluorescence in UV-light. Even the tetrahydro bases generally exhibit this property in consequence of air oxidation. The properties of the naturally occurring alkaloids, as well as those of several important artifacts, are recorded in Table 11. TABLE I1

PROPERTIES OF THE CARBOLINE ALKALOIDS Melting point Alkaloid

Formula

Harmaline Harmalol Harman Harmidine Harmidol Harmine Harmol N-Methyltetrahydroharmol Norharman Tetrahydroharman Tetrahydroharmine Tetrahydroharmol Tetrahydronorharman 2-Methyltetrahydro-2-carboline 1,2-Dirnethyltetrahydro-2-carboline

("C)

Reference

238 212 238 257 259 256 321 268 198 180 199 268 207 216 112

2 25 26 4 4 1 25 11 27 28 29 11 30 19 19

IV. Structure A. HARMINE Much of the early work on the structure of the harmala alkaloids was of necessity directed to the preparation and identification of degradative fragments, many of which were new. The following account will be

50

R. H. F. MANSKE

limited to the observations necessary to delineate unambiguously the structures of these alkaloids and, therefore, the subject will not be treated historically. Harmine is didehydroharmaline and is obtainable from the latter by mild oxidizing agents ; both bases are reduced by sodium in ethanol to the same tetrahydroharmine (29). Of the two nitrogens, only one is basic and takes part in salt formation. The oxygen is present as methoxyl and, when harmine is heated with fuming hydrochloric acid a t 140", there is formed methyl chloride and the phenolic base, harmol, which has recently been isolated from plants (17,18). Harmol is important because it can be made to react with ammonia in the presence of zinc chloride and ammonium chloride (anticipating Bucherer by several years) to yield the corresponding amino derivative from which the amino group was eliminated by the diazo reaction (31). The resulting demethoxy base named harman was later shown to be identical with a base obtainable from tryptophan by condensation with acetaldehyde in the presence of

I

I1

I11

oxidizin'g agents (26, 32). This observation made it virtually certain that harman had structure I1 (R = H ) . Confirmation was found in the fact that oxidation of harmine with strong nitric acid generated harminic acid and m-nitroanisic acid (33).Harminic acid (111;X = C02H), whose structural investigations played a prominent role in the early chemistry of harmine, on further oxidation gave pyridine-4-carboxylic acid (34). These observations clearly indicate that the benzene and pyridine rings are terminal and that the methoxyl is in the position shown for R in 11. Furthermore, the genesis of harman from tryptophan and acetaldehyde clearly fixes the position of the methyl and, therefore, harmine is I1 (R = OMe).

B. HARMALINE Harmaline is dihydroharmine and is easily oxidized to harmine. The position of the one double bond in the pyridine ring is fixed by its (IV) by synthesis from 2-acetyl-3-(~-aminoethyl)-6-methoxyindole warming in acid solution (35).Harmaline is therefore V. A number of later syntheses have been reported (28, 36, 37). A very recent one (38) achieved a synthesis in which the one double bond might have been

3.

51

THE CARBOLINE ALKALOIDS

expected to remain in the exocyclic position (VII). 6-Methoxytryptamine was condensed with glycolaldehyde to give VI in good yield

which, on dehydration with 90% phosphoric acid, yielded harmaline, presumably by way of the intermediate VII.

MeO-

H VII

VI

11

CHz

C. HARMAN The structure of harman (11; R = H) has been discussed already in connection with that of harmaline. Many additional syntheses of it and of its tetrahydro derivative have been recorded (39-44). Owing t o the extremely active 2-position in the indole nucleus, cyclizations which yield quinolines from P-phenethylamines proceed with greater ease from tryptamines, often yielding the tetrahydro bases under so-called physiological conditions (43,44).A novel synthesis of harman depends upon

n ‘J\N-N/VN H VIII

rl mcl I Me

N/

IX

I/N Me

rJ

\N

\/Me

I\/N 1 X

the Fischer indole ring closure of the reaction product (VIII) of 2-methyl3-hydrazinopyridine and cyclohexanone. The tetrahydro intermediate (IX) was smoothly dehydrogenated to harman in the presence of palladium (45).This would seem to be the first instance of an indole synthesis from a hydrazone derived from a hydrazinopyridine.

D. APOEMRMINE Though apoharmine (111; X = H ) played an important role in the early structural investigations of harmaline, it had not become easily

52

R. H. F. MANSKE

available by synthesis (46). A new and more convenient synthesis has been described, although the yield is still very poor (ca. 5 % ) . When 2-acetylpyrrole is condensed with aminodiethylacetal and ring closure is brought about by heating with polyphosphoric acid and a small amount of phosphorus oxychloride, there is formed a mixture of apoharmine (mp 183') and the liquid l-methylpyrrole[l,2-a]pyrazine (X) in 22 yo yield (47). REFERENCES J. Fritsche, Ann. 64, 360 (1848). F. Goebel, Ann, 38, 363 (1841). A. F. Ovejero, Farmacognosia (Madrid)6, 103 (1947); Chem. Abstr. 42, 5617 (1948). S. Siddiqui, Chem. Ind. (London)p. 356 (1962). 0. Hesse, Ber. 11, 1542 (1878). E. Spath, Monatsh. Chem. 41, 401 (1920). R. Rieth, Ann. 120, 247 (1861). E. Spgth, Monatsh. Chew. 40, 351 (1919). P. S. Massagetov, Zh. Obshch. Khim. 16, 139 (1946); Chem. Abstr. 40, 6754 (1946). G. P. Menshikov, E. L. Gurevich, and G. A. Samsonova, Zh. Obshch. Khim. 20, 1927 (1950); Chem. Abstr. 45, 2490 (1951). 11. T. F. Platonova, A. D. Kuzovkov, and P. S. Massagetov, Zh. Obshch. Khim. 26,3220 (1956); Chem. Abstr. 51, 8766 (1957). 12. A. L. Chen and K. K. Chen, Quart. J. Pharm. Phurmacol. 12, 30 (1939). 13. F. A. Hochstein and A. M. Paradies, J . A m . Chem. Soc. 79, 5735 (1957). 14. W. B. Mors and P. Zaltzman, BoZ. Inst. Quim. Agr. ( R i o deJaneiro) No. 34,17 (1954); Chem. Abstr. 49, 14906 (1955). 15. F. D. O'Connell and E. V. Lynn, J . Am. Phurm. Assoc. 42, 753 (1953); Chem. Abstr. 48, 2988 (1954). 16. R. R. Paris, F. Percheron, J. Mainil, and R. Goutarel, Bull. SOC.Chim. Prance p. 780 (1957); Chem. Abstr. 51, 16498 (1957). 17. J. Lutomski, Biul. Inst. Roslin Leczniczych 5 , 169 (1959); Chem. Abstr. 54, 16752 (1960). 18. B. Borkowski, Biul. Inst. Roslin Leczniczych 5 , 158 (1959); Chem. Abstr. 54, 15844 (1960). 19. T. F. Platonova, A. D. Kuzovkov, and P. S. Massagetov, Zh. Obshch. Khim. 28,3128 (1958); Chern. Abstr. 53, 7506 (1959). 20. G. M. Badger and A. F. Beecham, Nature 168, 517 (1951); Chem. Abstr. 46, 7574 (1952). 21. E. Bachli, C. Vamvacas, H. Schmid, and P. Karrer, HeZw. Chim. Acta 40, 1167 (1957). 21a. B. A. Abdusalamov and A. S. Sadykov, Uzbeksk. Khim. Zh. No. 6, 47 (1961); Chem. Abstr. 57, 9904 (1962). 22. E. H. Polindexter and R. D. Carpenter, Chem. Ind. (London)p. 176 (1962). 23. L. Marion and R. H. F. Manske, Can. J . Res. B16,432 (1938). 24. R. Tschesche, H. Jenssen, and P. N. Rangachari, Ber. 91, 1732 (1958). 25. 0. Fischer and E. Tauber, Ber. 18, 400 (1885). 26. W. H. Perkin and R. Robinson, J . Chem. Soc. 115, 967 (1919). 27. W. 0.Kermack, W. H. Perkin, and R. Robinson,J. C h m . Soc. 119, 1602 (1921). 28. S. Akabori and K. Saito, Ber. 63, 2245 (1930). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

3.

THE CARBOLINE ALKALOIDS

53

29. 0. Fischer, Ber. 22, 637 (1889). 30. E. Spath and E. Lederer, Ber. 63, 2102 (1930). 31. 0. Fischer, Chem. Zentr. (Festschr. 80. Geburtstug Regenten Luitpold. Erlanger I ) , 956 (1901). 32. F. G. Hopkins and S. W. Cole, J . Physiol. 29, 451 (1903). 33. 0. Fischer and W. Boesler, Ber. 45, 1930 (1912). 34. 0. Fischer, L. Angermann, and E. Diepolder, Ber. 47, 99 (1914). 35. R. H. F. Manske, W. H. Perkin, and R. Robinson, J . Chem. SOC. p. 1 (1927). 36. E. Spath and E. Lederer, Ber. 63, 120 (1930). 37. D. G. Harvey and W. Robson, J. Chem. SOC. p. 97 (1938). 38. I. D. Spenser, Can. J . Chem. 37, 1851 (1959). 39. Y. Asahina and S. Osada, J . Pharm. SOC. Japan No. 534, 63 (1926). 40. G. Hahn and H. F. Gudjons, Ber. 71, 2175 (1938). 41. G. Tatsui,J. Pharm. SOC. J a p a n 48, 453 (1928); Chem. Abstr. 22, 3415 (1928). 42. G. Hahn and H. Ludewig, Ber. 67,2031 (1934). 43. G. Hahn, L. BBrwald, 0. Schales, and H. Werner, Ann. 520, 107 (1935). 44. G. Hahn and A. Hansel, Ber. 71, 2163 (1938). 45. G. R. Clemo and R. J. W. Holt, J . Chem. SOC.p. 1313 (1953). 46. W. Lawson, W. H. Perkin, and R. Robinson, J . Chem. SOC.125, 626 (1924). 47. W. Herz and S. Tocker, J . Am. Chem. SOC.77, 6355 (1955).

This Page Intentionally Left Blank

-CHAPTER

4

THE QUNAZOLINOCARBOLINES R. H. F. MANSKE Dominion Rubber Research Laboratory, Guelph, Ontario, Canada

I. Introduction ........................................................

55

11. Occurrence .........................................................

55

111. Structure ........................................................... A. Rutaecarpine .................................................... B. Evodiamine ..................................................... C . Hortiamine and Hortiacine .................... D. Rhetsinine and Rhetsine ................... .................

56 66 57

~ f e r e n c e.......................................................... s

58

57

I. Introduction The quinazolinocarboline alkaloids, as the name indicates, contain both carboline and quinazoline nuclei. The first known representatives were rutaecarpine and evodiamine, and two more representatives have been encountered more recently. Rhetsinine, which was a t one time regarded as belonging here, is included in this chapter because of its affinity with rhetsine, though it does not contain the quinazoline ring system. This chapter is a summary and an extension of that dealing with these alkaloids in Volume 11, p. 402.

XI. Occurrence These alkaloids occur in the genera Evodia,Hortia, and Zanthoxylum, all members of the Rutaceae, and are restricted to only a few species. Most of the other species in these genera that have been examined yield furoquinoline and acridine alkaloids. Table I is a record of the occurrences, along with empirical formulas and melting points, of the quinazolinocarboline alkaloids. They are all optically inactive except evodiamine, [a]$ + 352' (acetone), which is easily racemized. Wuchuyine, , C10H1302NImp 237', [aID - 18', occurring in E. rutaecurpa ( l ) evidently 55

56

R. H. I?. MANSKE

does not belong to the group under discussion. Nothing is known of its structure. TABLE I

PLANTS A N D THEIR CONTAINED ALKALOIDS

Plant

Alkaloid

Evodia rutaecarpa Hook. f. & Thom. Rutaecarpine Evodiamine Hortia arborea Engl. Hortiacine Hortiamine Rutaecarpine Zanthoxylum rhetsa DC. (2.budrunga Rhetsime Wall.) Rhetsinine Hortia braziliana Vel. (Vand?) Hortiamine Base Nine other bases

Formula

Melting point ("C) Reference

CisH130N3 C19H170N3 CigHi50zN3 CzoHi70zN3

258 278 252 208

2, 3 2, 3

CigH170N3 CigHi70zNs

277 196

5 5

CzoHi70zNs

269

6

4 4

111. Structure

A. RUTAECARPINE The first-suggested structure of rutaecarpine was based upon an erroneous identification of one of the products of its fission with alkali ( 2 , 3). The fission fragment ultimately proved to be tryptamine ( 7 ) , and this led to the correct structure (I),since the second fragment had already been recognized as anthranilic acid. An early synthesis was achieved by with methylanthracondensing l-keto-l,2,3,4-tetrahydro-2-carboline

nilate in the presence of phosphorus trichloride (8). A number of other syntheses have been recorded (9-1 la), but that involving so-called physiological conditions is of special interest. An aqueous solution of

4.

THE QUINAZOLINOCARBOLINES

57

3,4-dihydro-2-carboline and o-aminobenzaldehyde as hydrochlorides was allowed to remain at 25" for 10 hours. The intermediate, presumably 11,was converted into rutaecarpine in 70 yo yield when it was oxidized with potassium ferricyanide in a phosphate buffer at pH 6.9 and 25" (J2).

B. EVODIAMINE This congener of rutaecarpine is transformed into a stable hydrate (isoevodiamine, mp 155') which, on dehydration with acetic anhydride, generates optically inactive evodiamine ( 2 , 3). Degradation with alkali generates dihydronorharman and N-methylanthranilic acid. More important still, evodiamine hydrochloride, upon heating, liberates methyl chloride and is converted into rutaecarpine (13). Consequently, evodiamine is I11;a synthesis involving the condensation of N-methylisatoic anhydride with tryptamine followed by heating with ethyl orthoformate confirmed the given structure (14).

C. HORTIAMINE AND HORTIACINE The structure of hortiamine (IV) follows from the fact that alkali and fission generates 6-methoxy-1-keto-1,2,3,4-tetrahydro-2-carboline N-methylanthranilic acid, and a recombination of these fragments regenerates the alkaloid (4). The structure of hortiacine (V) follows from the fact that it is formed from hortiamine hydrochloride by loss of methyl chloride on heating. It is the methoxy derivative of one of its congeners, namely, rutaecarpine (4). Hydrolysis of hortiamine gives rise to 6-methoxyrhetsinine (6).

D. RHETSININE AND RHETSINE The discoverers of these bases suggested certain structural formulas (5) which necessitated slight revisions. Rhetsinine was recognized as being identical with the product of the mild permanganate oxidation of evodiamine that was named hydroxyevodiamine (13). The IR-spectrum, however, showed no bands owing to hydroxyl, but well-defined bands ascribable to carbonyl and to NH. Consequently, rhetsinine is VI, and the older formulations of hydroxyevodiamine are no longer tenable. Furthermore, a synthesis of rhetsinine was easily achieved by condensing

58

R. H. F. MANSKE

l-keto-1,2,3,4-tetrahydro-2-carbolinewith methyl N-methylanthranilate in the presence of phosphorus oxychloride. Catalytic reduction of rhetsinine yielded rhetsine, which is identical with dl-evodiamine (15).

REFERENCES 1. A. L. Chen and K. K. Chen, J. Am. Pharm. Assoc. 22, 716 (1933); Chem. Abstr. 27, 5151 (1933). 2. Y. Asahina and K. Kashiwaki, J . Pharm. SOC.J a p a n No. 405, 1293 (1915); Chem. Abstr. 10, 607 (1916). 3. Y. Asahina and S. Mayeda, J . Pharm. SOC.J a p a n No. 416, 871 (1916); Chem. Abstr. 11, 332 (1917). 4. I. J. Pachter, R. F. Raffauf, G. E. Ullyot, and 0. Ribeiro, J . Am. Chem. SOC.82, 5187 (1960). 5. A. Chatterjee, S. Bose, and A. Ghosh, Tetrahedron 7 , 257 (1959). 6. I. J. Pachter, R. J. Mohrbacher, and D. E. Zacharias,J. Am. Chem. SOC.83,635 (1961). 7. Y. Asahina, J. Pharm. SOC. J a p a n No. 503, 1 (1924); Chem. Abstr. 18, 1667 (1924). 8. Y. Asahina, R. H. F. Manske, and R. Robinson, J . Chem. SOC. p. 1708 (1927). J a p a n No. 543, 541 (1927); Chem. 9. Y. Asahina, T. hie, and T. Ohta, J . Pharm. SOC. Abstr. 21, 3054 (1927). 10. T. Ohta, J . Pharm. Soc. Formosa 51, 2 pp. (in Ger.), 3 pp. (in Japan.) (1938); Chem. Abstr. 34, 5846 (1940). 11. Y. Asahina and T. Ohta, J . Pharm. SOC. J a p a n 48, 313 (1928); Chem. Abstr. 22, 3393 (1928). lla. A. G. Terzyan, R. R. Safrazbekyan, L. V. Khazhakyan, and G. T. Tatevosyan, Chem. Abstr. 57, 16532 (1962) [Izv.Akad. Nauk A r m . SSR, K h i m . N a u k i 14, 393 (196l)l. 12. C. Schopf and H. Steuer, A n n , 558, 124 (1947). 13. Y. Asahina and T. Ohta, J . P h a m . Soc. J a p a n No. 530, 293 (1926); Chem. Absh. 21, 2134 (1927). 14. Y. Ashina and T. Ohta, Ber. 61, 319 (1928). 15. I. J. Pachter and G. Suld, J. Org. Chem. 25, 1680 (1960).

-CHAPTER

5-

ALKALOIDS OF MITRAGYNA AND OUROUPARIA SPECIES J. E. SAXTON The University, Leech, Enghnd

I. Occurrence ........................................................

59

11. Mitragynine .......................................................

62

111. Mitraphylline ......................................................

64

IV. Uncarine-A and Formosanine (Uncarine-B)............................

70

V. Rhynchophylline (Mitrinermine) .....................................

....................................................... The Mass Spectra of Mitraphylline and Rhynchophylline . . . . . . . . . . . . . . . . Rotundifoline, Isorotundifoline (Mitragynol), and Speciofoline. ........... References.. ......................................................

VI. Adifoline.. VII. VIII.

75 80

82 85

89

I. Occurrence The genera Mitragyna and Ourouparia, ( Uncaria) (Rubiaceae) flourish in widely distributed tropical and subtropical regions. Several Mitragyna species that are large trees indigenous to West Africa, Southeastern Asia, and the Philippines find commercial use in the timber and paper industries. The leaves of M . speciosa Korth. are chewed as a narcotic in Siam, and, in common with leaves of M . parvifolia Korth., enjoy an undeserved reputation as a cure for opium addiction. In West Africa the bark of M . africana Korth. is used as a febrifuge. Species of the genus Ourouparia are large, woody climbing plants, also found in tropical regions, particularly in the Far East. The occurrence of alkaloids in Mitragyna species was first reported by Hooper, who isolated a crystalline alkaloid in 0.15% yield from the leaves of M . parvifoolia ( I ); this alkaloid was not named and has not since been reinvestigated. The leaves of M . speciosa contain the amorphous alkaloid mitragynine, mp 105°-1150 (2, 3), while the bark contains mitraspecine, C Z ~ H ~ ~ Nmp Z O244'-245', ~, [ C Z ] ~ ' - 59.15' (4, 5). A third base, giving an amorphous picrate, mp 123'-127", is also present, but this has not been studied further, since no crystalline derivatives could be obtained (6). The alkaloid mitraphylline, C21H24Nz04, mp 270°, 59

60

J. E. SAXTON

[.ID

- 9.84", was first isolatedfromthe bark of M . rubrostipulacea Havil. (Adina rubrostipulata K. Schumann) (7, 8 ) , and later from the bark of M . macrophylla Hiern ( 9 ) ; it also occurs in Uncaria kawakamii Hayata (10). Other reports of its extraction from M . rubrostipulacea ( A . rubrostipulata),under the name rubradinine, are available ( 1 1 , 12, 13). A point of interest is that, whereas a levorotatory form of mitraphylline occurs in the leaves of M . rubrostipulacea, a dextrorotatory form can be isolated from the bark (14) ; the dextrorotatory isomer is apparently also the one present in U . kawakamii (10).Under appropriate conditions, an optically inactive form of mitraphylline can also be obtained (9, 14).l Rhynchophylline (mitrinermine), C22HZgNz04, occurs in both Ourouparia and Mitragyna species ; it was originally discovered in the tendrils and stems of Ourouparia rhynchophylla Matsumura ( U . rhynchophylla Miq.) (15), and has since been shown to occur in M . africana ( M . inermis 0. Kuntze) (16), M . macrophylla Hiern ( M . stipulosa 0. Kuntze) ( 1 7 , 18), M . rotundifolia (Roxb.)0. Kuntze (19))M . ciliata AubrBv. et Pellegr. (14), and U . tomentosa DC. (0.guianensis Aubl. (20).The identity of rhynchophylline and mitrinermine has been proved repeatedly (14, 19, 21, 22)) although Millat maintains that they are different (23). The alkaloid crossoptine, obtained by Blaise (24)from the bark of the tropical African Crossopteryx kotschyana Fenzl. (Rubiaceae), which is used by the local natives as a febrifuge, is, according to Raymond-Hamet ( 2 5 ) , identical with rhynchophylline. However, it has been suggested (25) that the bark extracted, by Blaise was actually a Mitragyna species, and not C. kotschyana. Whether crossoptine (rhynchophylline) is identical with crossopterine, isolated much earlier by Hesse (26) from the same species, is not known.

It has occasionally been assumed that the dextro and lev0 forms of mitraphylline are true enantiomorphs, which, on the basis of the evidence available a t present, is a non sequitur. The formation in nature, i n the same plant, of two such enantiomorphs, which contain no less than five asymmetric centers, would be a remarkable circumstance indeed. A much more likely explanation is that these dextro and lev0 forms of mitraphylline ara simply mixtures of ( - )-mitraphylline and isomitraphylline [[.ID 18" (CHCIs)], which are readily interconvertible C7-epimers. This is also the view of Finch and Taylor ( 1 4 4 , who further point out that the discrepant rotation data may simply depend on the errors involved in the measurement of a small rotation using an inadequate concentration of alkaloid. I n a re-examination of Nozoye's mitraphylline, from U . kawakamii, for which [aID 3.8" (CHCls] had been recorded (lo), it was shown to be completely homogeneous (thin-layer chromatography] and identical in all respects (melting point, IR- and NMRspectra, and, more significantly, optical rotatory dispersion curve) with pure ( - )-mitraphylline, synthesized from ajmalicine (vide infra] (14a). Although other samples of mitraphylline with apparently anomalous rotations were not available, these data strongly suggest that they are also ( - )-mitraphylline,although they might conceivably be mixtures of ( - )-mitraphyllineand isomitraphylline.

+

+

5.

ALKALOIDS OP

Mitragyna AND Ourouparia

SPECIES

61

The leaves of M . rotundifolia (Roxb.) 0.Kuntze ( M .diversifolia Hook. f.) contain mitraversine, C22H26N204, mp 237' (2, 3), and rotundifoline, CzzH2,NzO5,mp 238"-240", [a]',"" + 124.7' (chloroform) (19). Rotundifoline is also a constituent of the leaves of M . ciliata (14, 26a), M . stipulosa, and M . inermis (26a). The alkaloid stipulatine, which occurs in the leaves of M . rubrostipulacea, and in the bark and leaves of M . speciosa together with rhynchophylline (26b), is identical with rotundifoline (26c). Isorotundifoline (mitragynol) was first extracted from M . rotundifolia (14), and has since been isolated from M . stipulosa and M . ciliata (26a). Speciofoline, a further isomer of rotundifoline, has recently been discovered in the leaves of M . speciosa (26c). Isorhynchophylline, C2~H28N204,[a]; + 8.3' (EtOH),a stereoisomer of rhynchophylline, occurs in 0. rhynchophylla (27) and in A . rubrostipulata (28). 0. kawakamii (Hayata) Hamet ( U . kawakamii Hayata) contains hanadamine, C21H24N204, mp 187') [a]h8"- 123.7' (29); uncarine-A, C21H24N204, mp 120'-130', [aID + 106.5' ; and uncarine-B, C21H24N204, mp 216-217") [aID +91.3' (30, 31). The last-named is identical with formosanine, mp 202'-218') [aID + 91.3', the principal alkaloid of 0.formosana Matsumura et Hayata (21, 32). The occurrence of formosanine in both 0. kawakamii and 0.formosana is not surprising, since there is little or no taxonomic distinction between these two species ; in fact, many botanist- do not consider them as separate species (33). A minor alkaloid, so far unnamed and uncharacterized, is also present in 0. formosana (34). This base shows some resemblance to gambirine, a n uncharacterized alkaloid isolated from 0. gambir Baillo; ( U . gambier Roxb.) (33, 35). The name gambirine has also been used to designate an alkaloid, C22H26N204, which occurs in the stems of U . gambier (36). Although these two alkaloids have been assumed to be identical (36), Raymond-Hamet states that they are different, and claims to have obtained both bases from U . gambier (33). Finally, the heartwood of the tree A . cordifolia Benth. et Hook. f. contains adifoline, C22H20N2Os (37); a substance having the same properties as adifoline had been isolated much earlier and named adinin (38), but in view of possible confusion with the purine derivative adenine the name adifoline is preferred (37). The pharmacology of these alkaloids has been discussed on numerous occasions (2, 3, 22, 23, 39-50). Mitragynine and, to a lesser extent, mitraphylline, exert a general depressant effect, and in some respects resemble cocaine (2, 3, 40, 44); mitraphylline is also a hypotensive (41). In connection with the use of Mitragyna extracts as a febrifuge, it is of interest that rhynchophylline (mitrinermine) exhibits a significant antipyretic action (45, 46). The hypotensive effects of the total extracts (39) are, in part, owing to rhynchophylline (23).

62

J. E. SAXTON

11. Mitragynine

The amorphous alkaloid mitragynine, initially formulated as is a monoacidic base and, in common with all the other alkaloids of this series, contains an ester group. The two remaining oxygen atoms are accounted for by the presence of two further methoxyl groups; the molecule also contains a t least one C-methyl group but is devoid of methylimino groups ( 2 , 6 ) .Saponification of mitragynine with methanolic potassium hydroxide yields a monobasic acid, apparently C21H28NZ04, containing two methoxyl groups, and an ether-soluble base, regarded as C23H34N205, which contains four methoxyl groups. The latter is presumably formed by addition of one molecule of methanol to the alkaloid; further heating with alkali converts it into the monobasic acid, from which mitragynine cannot be regenerated by methylation. The base does not respond t o hydrogenation procedures using palladium or platinum catalysts, and probably does not contain isolated double bonds (6). The UV-spectrum of mitragynine differs notably from the spectra of the other Mitragyna alkaloids. Whereas the absorption of the latter indicate the presence of oxindole nuclei, the spectrum of mitragynine shows a greater resemblance to that of the ajmalicine group of alkaloids (5). The presence of an indole nucleus is also suspected from its color reactions (2) and confirmed by the isolation of indole derivatives (so far unidentified) and 5-methoxy-9-methylharman (I)from the products of zinc dust distillation (6). The identification by synthesis (51) of this degradation product is of some interest, since the alkaloid itself does not apparently contain an N-methyl group. Moreover, this was the first demonstration of the occurrence of a 4-hydroxyindole derivative in nature. Recently, Hendrickson (52) has speculated on the structure of mitragynine on the basis of the preceding evidence, and has proposed the constitution 11. This pentacyclic formulation requires a revision of the hitherto accepted molecular formula from C22H30N204to C22H28N204 ; the latter, however, is equally consistent yith the available analytical data. The structure (11)can readily be derived from 4-methoxytryptamine and prephenic acid; the single-carbon unit, which normally appears as (3-21, is here omitted. The zinc dust degradation to 5-methoxy9-methylharman (I)was explained by a sequence of one reverse Mannich and two reverse Michael reactions. The methanol addition product obtained on reaction of mitragynine with methanolic alkali was presumed to be formed from I1 by reverse Michael reaction, which gives the unsaturated ester I I a (R = Me), followed by addition of methanol to the

C22H30N204,

5. ALKALOIDS OF Mitragyna AND Ouroupuria SPECIES

63

double bond. The reconversion of this product into mitragynine was explained by reversal of these reactions. The amino acid also obtained in the reaction of mitragynine with alkali was formulated as I I a ( R = H), which on esterification with diazomethane can give an ester different from mitragynine (52).

OMe

OMe

OMe

I

I

CHZ=C-COOR

COOMe

I

I1

Ira,

The structure of mitragynine (9-methoxycorynantheidine ; 111) was eventually established by comparison of its mass spectrum and NMRspectrum with those of its demethoxy relatives, corynantheidine (IIIa) and dihydrocorynantheine (IIIb) (53). As already mentioned, the UVspectrum of mitragynine resembles those of the heteroyohimbine alkaloids, and would appear to be owing to 4-methoxyindole and j3-alkoxyacrylic ester chromophores. The presence of this second chromophore is also suggested by the addition of the elements of methanol in the presence of alkali (cf. uncarine-A and formosanine) and confirmed by the IR-absorption bands a t 1690 and 1640 cm-1. The IR-spectrum also cQnfirms the presence of an imino group (3365 cm-1). Hence, mitragynine may be formulated as 9-methoxycorynantheidine or stereoisomer, in which case the correct molecular formula must be C23H30N204. This agrees with the analytical data for the methanolsolvated and unsolvated picrate, and is finally confirmed by the molecular weight (398) deduced from the mass spectrum. The gross structure 111, but not the stereochemistry, is also amply confirmed by the mass spectrum of mitragynine, which exhibits a fragmentation pattern closely similar t o those given by corynantheidine (IIIa) and dihydrocorynantheine (IIIb), if allowance is made for the aromatic methoxyl group in mitragynine. The stereochemistry of mitragynine is revealed by its NMR-spectrum, which, in the aliphatic region, is identical with that exhibited by corynantheidine ; in contrast, the NMR-spectrum of

64

J. E. SAXTON

dihydrocorynantheine is significantly different. Hence, mitragynine (111)is 9-methoxycorynantheidine (53). OMe

Me1OOC/"%HOMe I11 Mitragynine

MeO&/%HOMe

111. Mitraphylline

Mitraphylline, C21H~dN204,contains one methoxyl group, in the form of a carbomethoxy group, and a double bond, since it gives a yellow color with tetranitromethane and forms a dibromide (14). Its UV-spectrum is identical with those of formosanine and rhynchophylline, and very similar to that of gelsemine; hence, mitraphylline is also based on oxindole (10, 54). This conclusion is supported by the IR-spectrum of mitraphylline, which is consistent with its formulation as an oxindole; possessing an unsubstituted imino group (55). A further analogy with gelsemine is provided by its inability to couple with diazonium salts and its reaction with acetic anhydride and ferric chloride to form acetylmitraphylline, mp 164'-165" (55). However, the UV-spectrum of mitraphylline is not completely coincident with that of gelsemine; in the 250 mp region, the maximum occurs a t a significantly higher intensity, indicating the presence of a second chromophore. This could be an unsaturated enol-ester grouping of the type common to ajmalicine and its stereoisomers. The IR-spectrum is consistent with the occurrence in mitraphylline of such a grouping, since it exhibits absorption bands a t 1725 and 1626 cm-1 (56, 57). The presence of an oxindole nucleus in mitraphylline is established by the results of zinc dust distillation. The basic products include isoquinoline and 3,4-diethylpyridine, while the neutral fraction affords 3-spirocyclopropano-oxindole(IV),mp 179"-181" (55, 56). This degradation product is of the greatest importance in this series of alkaloids ;it was first obtained from the calcium oxide distillation of rhynchophyllic acid (19), and has since been obtained by zinc dust distillation of uncarine-A (58), by hydrogenation and pyrolysis of uncarine-A methiodide (59), and by zinc dust distillation of formosanine (uncarine-B) (60). The structure of this neutral degradation product was first proposed by Wenkert and Reid (61), who pointed out that its properties were very similar to those

5. ALKALOIDS

OF

Mitragyna AND Ourouparia

SPECIES

65

reported for 3-spirocyclopropano-oxindole(IV), which had been synthesized earlier from 1-phenyl-1-cyanocyclopropane (62). Independently, Kondo et al. (60) reached the same conclusion, which was finally

confirmed by comparison of a synthetic sample of IV with that obtained by degradation of the uncarines (63). The foregoing degradations, when considered in conjunction with the spectrographic properties of the molecule and the proved presence of one C-methyl group, resulted in the provisional formulation of mitraphylline as V, i.e., as the tetracyclic oxindole analog of ajmalicine (56). This expression was based on the molecular formula CzlHzaN204, and on the assumption that the molecule contains a carbomethoxy group. Later work established that the molecular formula of mitraphylline is C21H24N204, and provided the first conclusive proof of the presence of an alkoxyacrylic ester function. The structure for mitraphylline was accordingly modified to VI, i.e., the true oxindole analog of the heteroyohimbine alkaloids (57). Saponification of mitraphylline gives the corresponding acid, mitraphyllic acid, CzoH22N204, which can be reconverted into mitraphylline by esterification, although not without loss owing to side reactions of an undetermined nature. Reduction of mitraphylline with lithium aluminum hydride a t room temperature gives mitraphyllol, C20H24N203 (VI ; COOMe +CHzOH), which still contains

OR VI

VII

the oxindole grouping and the enol ether function, since its IR-spectrum contains bands a t 1730-1708 cm-1 and 1654 cm-l; the shift of the band

66

J. E. SAXTON

from 1626 cm-1 in the alkaloid to 1654 cm-1 in mitraphyllol is an expected consequence of the reduction of the conjugated ester group. Another result of the reduction of this grouping is the simple oxindole UV-spectrum exhibited by mitraphyllol. More vigorous reduction of mitraphylline with lithium aluminum hydride gives dihydrodeoxy--f mitraphyllol, CzoH,6N,02(VI; COOMe + CH,OH and -NHCO-NHCH,), which exhibits a typical dihydroindole UV-spectrum, and which, in contrast to mitraphylline, couples with diazotized sulfanilic acid. The formation of a dihydroindole derivat'ive, and not an indole derivative, in this reduction is an indication, though not perhaps a conclusive one, that the molecule is a 3,3-disubstituted oxindole ; hence, structure V is invalidated (57). I n accordance with the structure VI, mitraphylline on acid hydrolysis also suffers decarboxylation. The amorphous product can readily be converted into a well-defined crystalline derivative, containing one methoxyl group, with methanolic hydrogen chloride. This derivative possesses the properties of an acetal and exhibits a simple oxindole UV-spectrum, and is consequently formulated as V I I (R = Me), i.e., methylmitraphyllal. Acid hydrolysis of methylmitraphyllal gives an amorphous, methoxyl-free base, which exhibits a hydroxyl band but no aldehyde absorption in the IR-spectrum ;since it slowly reduces Tollens' reagent, it is presumed to be mitraphyllal (VII; R = H), and is probably identical with the amorphous hydrolysis product of mitraphylline. Further evidence for the presence of a hemiacetal grouping in mitraphyllal is provided by the Wolff-Kishner reduction, which gives mitraphyllane (VIII), C19H26N202 (57).

I

Et VIII

Independently, Nozoye (10) also proposed structure VI for mitraphylline, on the basis of the identity of its UV-spectrum and the close similarity of its IR-spectrum (except in the fingerprint region) with those of formosanine (uncarine-B), for which the constitution V I had already been proposed. The stereochemistry of mitraphylline has been discussed, and divergent views have been expressed (10, 64, 65, 66) ; nevertheless, there is sufficient experimental evidence available now t o allow configurational

5.

ALKALOIDS OF

Mitragyna AND Ourouparia

SPECIES

67

assignments to be made with some confidence. Mitraphylline, like its stereoisomer formosanine, undergoes a facile isomerization when heated with a variety of acidic and basic reagents to give an equilibrium mixture which contains (in pyridine) approximately 80% of isomitraphylline ; the same mixture is obtained when isomitraphylline is similarly treated. When equilibrated in aqueous acetic acid solution, mitraphylline predominates in the equilibrium mixture. Isomitraphylline can also be converted into a mixture of the two bases by oxidation with mercuric acetate followed by hydrogenation of the amorphous C-3-Nb dehydro derivative so obtained. It is noteworthy that mitraphylline is not oxidized by mercuric acetate. The site of the isomerism of these bases is clearly the p-amino lactam system, which allows isomerization by cleavage and re-formation of the C-3 to C-7 bond (28, 67) :

Mitraphylline and isomitraphylline must accordingly differ in the configuration a t C-3 and/or a t C-7. The NMR-spectra of mitraphylline and isomitraphylline show that they are stereochemically related to ajmalicine (IX) (64). In particular, the one-proton octets a t low fields in both spectra are almost identical in position and spin-spin coupling constants with the octets attributed to the C-19 hydrogen atom in ajmalicine and 3-isoajmalicine. The values for the C-19 methyl group in all four bases are also closely similar. Hence, it can be inferred that mitraphylline and isomitraphylline also have a trans D/E ring junction, and a trans disposition of the C-19 methyl group with respect to the C-20 hydrogen atom (64). The configuration of C-3 is less certain, owing to the facile isomerization of mitraphylline and isomitraphylline, a process which necessarily involves the destruction and reconstitution of the asymmetry a t C-3 and C-7. However, Wenkert et al., noting that these two bases have comparable thermodynamic stabilities, formulate them both with axial hydrogen a t C-3 ; consequently, mitraphylline and isomitraphylline differ only in the configuration a t C-7. It was argued that if one of the isomers had possessed equatorial hydrogen a t C-3, the presence of a bulky quaternary axial substituent (C-7) a t this position would have resulted in nonbonded interactions of such magnitude that the stability of this isomer would

68

J . E. SAXTON

necessarily be much smaller than that of the isomer with axial hydrogen at (3-3. Consequently, equilibration would be expected to result in almost complete conversion into the latter; this, clearly, is contrary to the experimental evidence. Hence, mitraphylline and isomitraphylline are presumed to have the same stereochemistry as ajmalicine at positions 3, 15, 19, and 20 (64).

XI OMe XI1

0

x 111 Mitraphylline

XIV Isomitraphylline

These conclusions have received convincing confirmation in a partial synthesis of mitraphylline and isomitraphylline from ajmalicine (IX)

5.

ALKALOIDS OF

Mitragyna

AND

Ourouparia

SPECIES

69

(66, 68). Reaction of ajmalicine with tertiary butyl hypochlorite gave which on methanolysis gave the the 7-chloroindolenine derivative (X), imidoether (XII), presumably via the intermediate XI. Hydrolysis of XI1 in weak acid solution then gave a mixture of the corresponding oxindoles (XI11 and XIV), which were identified as mitraphylline and isomitraphylline, not necessarily respectively. I n this sequence of reactions the rearrangement step (XI-+ XII) would be expected to proceed with retention of configuration of the migrating group, i.e., (3-3. Unfortunately, the force of this argument is lost since the conditions employed in the hydrolysis of the imidoether XI1 were sufficiently vigorous to allow the interconversion of mitraphylline and isomitraphylline. However, pseudoyohimbine, which has equatorial hydrogen a t C-3, gave only a minute yield (4%) of the same imidoether (XIIa) as did its C-3 stereoisomer yohimbine (IXa); consequently, it can be inferred that rearrangement is accompanied by almost complete retention of configuration at C-3. The subsequent stage, i.e., the hydrolysis of

IXa

KIIa

H

0 XIIIa

XIVa

the imidoether to the corresponding oxindole, was also shown to proceed with retention of configuration at C-3 by examination of the behavior of yohimbine derivatives in the same sequence of reactions. The products from yohimbine are two isomeric oxindoles, designated yohimbine oxindoles A and B. The methiodide of the intermediate imidoether

70

J. E. SAXTOK

(XIIa), however, hydrolyzed to give exclusively the methiodide of yohimbine oxindole A, which must therefore have the stereochemistry shown in XIVa, since quaternization of N, prohibits the equilibration of the isomeric oxindoles (cf. mechanism already outlined). Further evidence relating t o the structures of the yohimbine oxindoles A and B was derived from their NMR-spectra and pK, values, from which it was concluded that yohimbine oxindoles A and B are correctly formulated as XIVa and XIIIa, respectively (66). Returning t o the conversion of ajmalicine into mitraphylline and isomitraphylline, it was established by thin-layer chromatographic examination ofthe hydrolysis products of XI1 after appropriate intervals of time that the primary product of acid hydrolysis is isomitraphylline, which must therefore have the stereochemistry shown in XIV. The NMR-spectra of isomitraphylline and mitraphylline are analogous to those of the yohimbine oxindoles A and B, and hence these alkaloids must be XIV and XIII, respectively. The base-strengthening influence present in mitraphylline is attributed to the proximity of the lactam carbonyl group t o Nb; the ammonium ion obtained on protonation of Nbcan then be stabilized by hydrogen bonding with the carbonyl group. The stereochemistry shown in XI11 is the only one which permits a base-strengthening hydrogen bonding of this type ; accordingly, mitraphylline must be XI11 (66).

IV. Uncarine-A and Formosanine (Uncarine-B) The stem, wood, and bark of Uncaria kawakamii contain the two isomeric alkaloids, uncarine-A and uncarine-B, Cz1H24N204 (30, 31), the highest proportion of alkaloids (1.480/,) occurring in the bark (31). Uncarine-B has been identified with formosanine (28, 33)) which was isolated earlier from Ourouparia formosana (21, 32). The chemical behavior of uncarine-A and formosanine is identical in all respects, and their relationship as stereoisomers is shown by their ready interconversion and equilibration, which frequently hinders the separation and purification of individual isomers. Thus, uncarine-A in hot dilute acetic acid yields formosanine, and the reverse transformation can be achieved by heating forinosanine with ether, until it has all been isoinerized into the more soluble uncarine-A. Some formosanine is even obtained when uncarine-A is recovered from its hydrochloride, or from the hydrochloride of the related amino acid, by methylation with diazomethane (59).

Uncarine-A and formosanine are monoacidic bases, which contain one

5.

ALKALOIDS OF

Mitragyna AND Ourouparia

SPECIES

71

carbomethoxy group ; saponification yields the corresponding acid, from which the alkaloids can be regenerated with diazomethane. They are resistant to catalytic reduction, and do not form carbonyl derivatives. Both uncarine-A and formosanine add one mole of alcohol on treatment with anhydrous methanolic or ethanolic hydrogen chloride ;the products, in common with the alkaloids, are hydrolyzed and decarboxylated by aqueous acid to products which have been designated decarboxyuncarinol-A and -B (30). Palladium dehydrogenation of uncarine-A a t 250"-290" gives 3-ethyl-4-methylpyridine, 3-ethylpyridine, 3,4-diethylpyridine, 3-ethyloxindole, and 3-methyloxindole (59, 69). Under more vigorous conditions (300"-310"), isoquinoliiie is also obtained (69)) but this is regarded as a secondary product, derived from 3,4-diethylpyridine and not from uncarine-A itself (59). Zinc dust distillation gives the foregoing characteristic bases and, in addition, 3-spirocyclopropanooxindole (58); the last is also the product of pyrolysis of uncarine-A methiodide (59).Potash fusion of uncarine-A furnishes an acid, suspected to be o-hydroxyhydratropic acid, mp 166"-167" (70). The chemical evidence is thus overwhelming that uncarine-A and formosanine are based on oxindole and 3,4-disubstituted piperidine units. I n order to account for the degradation t o 3-spirocyclopropanooxindole, it was a t first assumed that the connecting linkage between these two halves consisted of a preformed cyclopropane ring (71). Further investigations with model substances, however, rendered this assumption unnecessary. N-Methyloxindole was alkylated with diethylaminoethyl chloride in the presence of sodamide to give l-methyl-3-(/3diethylaminoethy1)oxindole. When the methiodide of this base was heated a t 180'-200°, N-methyl-3-spirocyclopropano-oxindolewas formed, identical with the product of methylation of I V ( 7 1). Hence, the formation of 3-spirocyclopropano-oxindole in the degradation of uncarine does not necessarily imply the presence of a preformed cyclopropane ring, but can be explained equally well by a 3-(P-piperidylethyl)oxindole constitution. The structure of the piperidine fragment in uncarine-A and forinosanine can be inferred from their spectrographic properties. The UV-spectra are identical with the summation spectra of 3-ethyloxindole with either XV (R = H, R' = E t ) or XV (R = R' = Me) (72, 73, 74), and the IR-spectra in the carbonyl region are also entirely consistent with the presence of these chromophores (74). I n contrast, the UV-spectrum of " methyluncarinol-B methyl ether," prepared by addition of methanol to formosanine, is identical with that of 3-ethyloxindole (73). On the basis of these results, and the assumption that the molecular formula was C21H26N204, uncarine-A and formosanine were formulated as stereoisomers of mitraphylline (V) (73, 74).

72

J. E. SAXTON

As in the case of mitraphylline, a re-examination of the available evidence, together with some new experimental data, led t o the adoption of the constitution V I for uncarine-A and formosanine. Reduction of

IT

ROOC'\/O

I

R

xv

XVI Mayumbine

uncarine-A with lithium aluminum hydride gave a dihydroindole base, indicating that the alkaloid is a 3,3-disubstituted oxindole derivative. The same conclusion was drawn from the results of the methylation of uncarine-A and formosanine with an excess of sodium methoxide and methyl iodide, which gave only monomethyl derivatives. The stereochemistry of uncarine-A and formosanine was briefly discussed, and it was tentatively proposed that these two bases are epimeric a t C-3 (75). More recent investigations on formosanine include a comparison of its NMR-spectrum with that of the ajmalicine stereoisomer mayumbine, for which the epiallo stereochemistry shown in XVI was deduced (64). The similarity of these spectra indicates that in all probability formosanine is the oxindole analog of mayumbine ; hence, formosanine must also contain a c i s DIE ring junction and a c i s relationship of the C-19 and C-20 hydrogen atoms. Further, the one-proton octet due to the C-19 proton is centered at a somewhat higher field (6.247) in bothformosanine and mayumbine than the corresponding proton in ajmalicine and mitraphylline ( 5.6 T); this is interpreted as indicating that the C- 19hydrogen is oriented more favorably for intramolecular shielding by the C-16 to C-17 double bond in formosanine and mayumbine than in ajmalicine and mitraphylline. Consequently, in the former pair, a quasiaxial configuration is preferred for the C-19 hydrogen atom. Thus, the possible conformations of formosanine are given by XVII and XVIII, and its stereoisomer, uncarine-A, must be epimeric a t C-3 and/or C-7. Now formosanine and uncarine-A possess comparable stabilities, since,

-

5.

ALKALOIDS OF

Mitragyna

AND

Ourouparia

SPECIES

73

for example, the equilibrium mixture in pyridine contains 20% formosanine (28). Wenkert et al. argue that a compound of conformation XVII H

f

N

3

H

XVIII

would be expected t o be considerably less stable than one of conformation XVIII, regardless of the configuration of C-7, owing to the large nonbonded interactions present in both XVIIa and XVIIb, and would be expected to be converted almost entirely into XVIII a t equilibrium. Consequently, both formosanine and uncarine-A are formulated as the epiallo isomers XVIII (64). Since formosanine (pK, 5 . 5 ) is a somewhat stronger base than uncarine-A ( p K , 4.2), it may be assigned the conformation shown in X I X ; uncarine-A is then its C-7 epimer (XX). Hendrickson (65) expressed a different view regarding the stereochemistry of formosanine. I n the all0 series, providing the C-19 methyl group is trans with respect to the C-20 hydrogen atom, the favored conformation is probably that given by XVIIb, since the alternative XVIIa is subject to large nonbonded interactions, of which those owing to the C-5 and C-14 hydrogen atoms and the C-19 methyl group are probably the most important. Of the two possibilities based on XVIIb, one (XXI) contains the carbomethoxy group in close proximity to the

74

J. E. SAXTON

lactam function, a situation which is invoked t o account for the apparent hydrogen bonding of the imino group in formosanine (cf. IR-absorption a t 3226 em-1, compared with 3333 em-1 in uncarine-A) and its alkali

cr; \

COOMe

I

P

N

xx

XIX Formosanine (uncarine-B)

Uncarine-A

insolubility, in contrast t o the normal behavior of oxindoles. In none of the other cis D/E isomers does the carbomethoxy group approach the oxindole function. I n contrast to Wenkert et al., Hendrickson regards the

XXI

stereoisomers XX and X X I as explaining satisfactorily the comparable thermodynamic stabilities of uncarine-A and formosanine. Although the structures X X and X X I for uncarine-A and formo-

5.

ALKALOIDS OF

Mitragyna

AND

Ourouparia

SPECIES

75

sanine, respectively, explain satisfactorily their behavior towards alkali and the IR-absorption in the 3300 cm-1 region, they do not provide a convincing explanation of the different basic strengths of these alkaloids, since the ammonium ions derived from X X and X X I can neither be stabilized nor destabilized by other structural features within the molecule. I n the absence of a definitive proof of the stereochemistry of these alkaloids, the most acceptable conformations at present are X I X for formosanine and XX for uncarine-A.

V. Rhynchophylline (Mitrinermine) Rhynchophylline, C22H28N204, mp 212"-213", [a]k5'- 14.5O (chloro+ 5.9" (EtOH), form a form) and isorhynchophylline, mp 150°, [a]t8;"" third pair of interconvertible stereoisomers and both occur naturally. Their properties are very clearly reminiscent of the other members of this group. For example, the UV-spectrum of rhynchophylline is superimposable on that of mitraphylline (54), and it is also degraded by zinc dust distillation to 3-spirocyclopropano-oxindole (19). Rhynchophylline contains two methoxyl groups (76, 77), one of which is contained in a carbomethoxy group, since saponification gives the amorphous, amphoteric, rhynchophyllic acid (19, 27). Hydrolysis using acid conditions is accompanied by decarboxylation (19). Rhynchophylline decolorizes bromine in chloroform solution without liberation of hydrogen bromide, and absorbs one mole of hydrogen in the presence of a platinum catalyst; the double bond, however, is comparatively unreactive, since it does not suffer hydrogenation over palladium catalysts, nor give a yellow color with tetranitromethane (14). These properties, combined with the fact that the second methoxyl group is not attached to the benzene ring and must, therefore, be derived from the dihydroxyphenylalanine (or prephenic acid) precursor, suggest the presence in rhynchophylline of a /3-methoxyacrylic ester residue. Convincing support for this conclusion is provided by a comparison of the UV- and IR-spectra of rhynchophylline with those of appropriate model compounds. Thus, the UV-spectrum of rhynchophylline is identical with the summation spectrum of 3-ethyloxindole and ethyl /3-ethoxy-a-methylacrylate (74), and it is also closely similar to that of formosanine. The IR-absorption of rhynchophylline in the carbonyl region also resembles that of formosanine, except that rhynchophylline exhibits an adhtional band, of medium intensity, at 1645 cm-1; this band, however, is also present in the spectrum of ethyl /3-ethoxy-amethylacrylate.

76

J. E. SAXTON

It is thus apparent that the absorption spectra of rhynchophylline can be explained entirely satisfactorily by the assumption of oxindole and p-methoxyacrylate chromophores (74). Hence, the alkaloid was provisionally formulated as the oxindole analog of corynantheine or dihydrocorynantheine (56, 74) ; the resistance to hydrogenation and the KuhnRoth determination (0.6 C-Me) are clearly in favor of the latter. Abundant evidence has recently been accumulated, mainly owing to the work of Marion and his collaborators, that the structure of rhynchophylline is XXII, i.e., the tetracyclic oxindole analog of dihydrocorynantheine (28, 78). The presence of the P-methoxyacrylic ester grouping has been firmly established by a series of transformations similar to those carried out using corynantheine. The double bond in this grouping is extremely resistant t o hydrogenation, and in the presence of a platinum catalyst in acetic acid solution, rhynchophylline gives a hexahydro derivative by saturation of the benzene ring, but the double bond remains intact. Hydrolysis of rhynchophylline with dilute acid gives an byO Z , aldehyde base, rhynchophyllal (XXIII ; R = CHO), C ~ ~ H Z ~ N Z

XXII

XXIII

hydrolysis of the ester and enol ether functions, followed by decarboxylation of the ,f&aldehydo-acid so obtained. The absorption spectra of rhynchophyllal are consistent with this conversion, since its UV-spectrum is that of a simple oxindole derivative, and its IR-spectrum does not contain a band at 1650 cm-1 characteristic of the enol ether double bond. Reduction of rhynchophyllal with sodium borohydride gives rhynchophyllol ( X X I I I ; R = CHzOH), which can be further reduced by lithium aluminum hydride to dihydrodeoxyrhynchophyllol (XXIV). The latter can also be obtained by direct reduction of rhynchophyllal

-

I

CHzCH20H XXIV

5 . ALKALOIDS

OF

Mitragyna

AND

Ourouparia

SPECIES

77

with lithium aluminum hydride ; it behaves in all respects as a dihydroindole base (e.g., UV-spectrum in neutral and acid solution, and ability to couple with diazotized sulfanilic acid), indicating that its precursor is a 3,3-disubstituted oxindole derivative (78). This feature of the molecule is also established by the reduction of rhynchophylline with lithium aluminum hydride, which also yields a dihydroindole derivative (79). The points of attachment of the ring D substituents are proved by the palladium charcoal dehydrogenation of rhynchophyllane (XXIII ; R = Me ; later renamed isorhynchophyllane), the Wolff-Kishner reduction product of rhynchophyllal. The oily product, isolated as its crystalline picrate, was identified as 3,4-diethylpyridine. Similar dehydrogenation of rhynchophyllal gives 3-ethyloxindole and 3-ethyl-4-methylpyridine. Hence, the /3-methoxyacrylic ester grouping must be attached to the 4-position in a piperidine nucleus, and the ethyl group must be situated on an adjacent carbon atom (78). The structure XXII is clearly consistent with all these experimental results, and is also biogenetically satisfactory. The above evidence does not indicate, however, the point of attachment of the piperidine ring D to the ,&position of the oxindole nucleus, which was derived purely from biogenetic arguments. Later experiments furnished the required evidence relating to the size and nature of ring C. The equilibration of rhynchophylline or isorhynchophyllinein pyridine gives amixture containing 70~oofisorhynchophylliue. This base behaves in all respects as a stereoisomer of rhynchophylline, and can similarly be converted, by acid hydrolysis, into anl (amorphous) aldehyde. The latter, when reduced by the Wolff-Kishner procedure, gives a product which is identical with rhynchophyllane (XXIII ; R = Me); since the is0 bases appear in general to be the more stable, the name isorhynchophyllane is preferred for this product. Having regard t o the recent demonstration that the position of equilibrium in the rhynchophylline-isorhynchophylline system is dependent on the solvent employed (66), this view is no longer valid, and the stereochemistry of this base awaits elucidation. Isorhynchophyllane, in contrast to rhynchophyllane, reacts readily with mercuric acetate to give a dilactam which, although it has not been obtained crystalline, appears to contain oxindole (IR-absorption a t 1715 cm-1) and six-membered lactam (IR-absorption at 1625 cm-1) groupings. The formation of a second lactam function instead of an immonium salt in this oxidation can be rationalized by the sequence of reactions shown next page, which necessarily implies attachment of the /?-position of the oxindole nucleus to a carbon atom adjacent to N,. Ring C must therefore be five-membered, and the fact that the C-3to C-7 bond is severed in this oxidation is supported by the lithium

78

J. E. SAXTON

aluminum hydride reduction of the amorpho’usdilactam (XXV?), which yields a noncrystalline base having the UV-absorption of an indole derivative (28).

Et

The structure XXII for rhynchophylline has been confirmed, and the relative stereochemistry at C-15 and C-20 has been elucidated, by a total synthesis (80) of ( )-N-methylrhynchophyllane (XXVI) (Marion’s N methylisorhynchophyllane), which had been prepared earlier by methylation of (iso)rhynchophyllanewith sodium methoxide and methyl iodide (28). The lactone (XXVII) of threo-3,4-diethyl-5-hydroxyvaleric acid was converted by reaction with phosphorus pentachloride into the corresponding 8-chloroacid chloride (XXVIII),which on treatment with methylaniline gave the anilide XXIX. Reduction of XXIX with lithium aluminum hydride gave the aldehyde XXX, which slowly reacted with CIOCfl

__f

-Et

it

Et

XXVII

XXVIII

0

__f

NMeCO

CHzCl

I

u . - E t

At XXIX

XXVI

5. ALKALOIDS OF Mitragyna

AND

Ourouparia

79

SPECIES

2-hydroxy-Na-methyltryptamine (XXXI) at room temperature to give a mixture of two racemic (2-15, C-20-trans bases, which were separated by chromatography on alumina. One of these racemates, designated trans-A, exhibited an IR-spectrum identical with that of authentic N,-methyl(iso)rhynchophyllane (XXVI).The second racemate (trans-B), and the two corresponding racemates (&-A and cis-B) which were synthesized by the same route from erythro-3,4-diethyl-5-hydroxyvaleric acid, exhibited IR-absorption which was very similar to, but not identical with, that of XXVI (80).

".ciH=QT+ H H'

,

'-Et

'-Et H

I'

H'

MeOOC'

%HOMO

MeOOC'

H'

\CHOMe

XXXIII

XXXII

Me

-

-Et H

M~OOC/~\CHOM~

H

COOMe

xxxv

XXXIV

Rhynchophyllinr

XXXVI Isorh ynchophyllinr

The hydrogen atoms at C-15 and C-20 in rhynchophylline are thus oriented trans with respect to each other, a conclusion which is amply confirmed by the partial synthesis of rhynchophylline from dihydrocorynantheine (XXXII) (66). Reaction of XXXII with tertiary butyl hypochlorite gave the chloroindolenine derivative XXXIII, which on methanolysis was converted into the imidoether XXXIV. Hydrolysis of XXXIV with refluxing aqueous acetic acid then gave rhynchophylline.

80

J. E. SAXTON

This conversion not only demonstrates the trans disposition of the substituents at C-15 and C-20 in rhynchophylline, and therefore in isorhynchophylline, but also establishes that the absolute configuration of (2-15 is the same as in dihydrocorynantheine, i.e., the hydrogen atom has the a-configuration. Since rhynchophylline and isorhynchophylline are completely analogous in behavior to the oxindoles of known structure derived from yohimbine, they can be assigned the stereochemistry shown in XXXV and XXXVI. Rhynchophylline, pK, 6.32 (XXXV)is the stronger base [cf. isorhynchophylline, pK, 5.20 (XXXVI)] (79), and hence would be expected to have its lactam carbonyl group so situated with respect t o N, that its conjugate acid can be stabilized by hydrogen bonding (66). The position of the C-17 proton signal in the NMR-spectrum of rhynchophylline ( - 2.777) compared with the position of the comparable signal in the heteroyohimbine alkaloids (2.46 T ) is interpreted in terms of a trans relationship of the carbomethoxy group and the C-17 hydrogen atom in rhynchophylline and isorhynchophylline, as shown in XXXV and XXXVI (64).

VI. Adifoline The heartwood of the southeast Asiatic tree, Adina cordifolia, contains, in addition to umbelliferone and /3-sitosterol,a highly oxygenated, bright yellow alkaloid, adifoline, CzzHzoNzOg, which gives intensely green fluorescent solutions in ethanol (37). This base appears t o be identical with adinin, the yellow constituent of the same species, which was isolated earlier, and reputed to have the molecular formula C16H1407 (38).

Adifoline is an amphoteric substance, which contains one carboxylic acid group; one other acidic grouping is present, probably as a phenolic hydroxyl group, since adifoline gives a (gummy) dimethyl derivative with diazomethane. Attempts to characterize adifoline as a salt also failed, since no crystalline derivatives could be obtained (37). Adifoline contains one methoxyl and two C-methyl groups, one of which appears to be present as an acetyl group, since reaction with aqueous alkali liberates acetic acid. However, the IR-spectrum of adifoline does not contain a band attributable to an acetoxy group, and since the presence of an N-acetyl group is also unlikely, the origin of this acetic acid is as yet unknown. The IR-spectrum of the alkaloid exhibits complex absorption in the region 1710-1550 cm-1; the precise origin of most of these bands is unknown a t present, although a /3-alkoxyacrylic ester function might account for two of them. Evidence relating

5.

ALKALOIDS OF

Mitragyna

AND

Ourouparia

SPECIES

81

to this point could not be obkained by lithium aluminum hydride reduction, owing to the insolubility of adifoline in tetrahydrofuran. Attempted reduction with metal-acid combinations, and hydrogenation using Adams’ catalyst, also failed. The UV-spectrum of adifoline in neutral sohtion closely resembles that of 1-ethyl-8-carboline, while in acid solution the spectrum is very similar to that of p-carbolinium derivatives, e.g., serpentine nitrate. I n both spectra the only significant difference compared with the spectra of the model substances is that the longest wavelength maximum shows a bathochromic shift. The presence of a 8-carboline system in adifoline is confirmed by oxidation with hot concentrated nitric acid, which gives a tetracarboxylic acid, apoadifolinetetracarboxylic acid. The latter is regarded as one of the alternatives XXXVII, mainly on the basis of the

HOOC. CH2-

H

bOOH XXXVII

XXXVIII

Hoy;y-.$yooH

+\$

y

N ClOH1305

XXXIX Adifoline

NMR-spectrum of its tetramethyl ester, and its conversion, by total decarboxylation with zinc or copper, into apoadifoline, a methyl derivative of 1,6-diazaindene (XXXVIII). The destruction of ring A in adifoline by oxidation clearly indicates the presence of an oxygenated substituent, possibly a hydroxyl group. Such a substituent would account for the second, weakly acidic group in adifoline ; it is also consistent with the production of two p-carboline derivatives, one of which is phenolic, in

XL

the selenium degradation of adifoline. However, the UV-spectrum of adifoline does not give a precise indication of the position of the oxygenated substituent. On the basis of the above evidence, adifoline can

82

J. E. SAXTON

tentatively be formulated as XXXIX (37). The recent isolation (81) of a glycoside of the /3-carbolineacid (XL) from Aspidosperma polyneuron, together with an obvious biogenetic origin from tryptophan, render the p-carboline fragment of this structure readily acceptable.

VII. The Mass Spectra of Mitraphylline and Rhynchophylline Mitraphylline and rhynchophylline undergo characteristic fragmentation processes on electron impact, hence their mass spectra are markedly different from those of the tetrahydro-/3-carboline alkaloids in general, and of the heteroyohimbine alkaloids (e.g., ajmalicine) in particular. Consequently, the mass spectra of the oxindole alkaloids promise to be very valuable in the structural elucidation of new members of this subgroup; the structure of one such base, carapanaubine, a constituent of Aspidosperma carapanauba M. Pichon, has already been elucidated by this method, and the structure deduced has subsequently been confirmed by partial synthesis from isoreserpiline (82). The mass spectra of ajmalicine (IXa) and its deuterated derivatives IXb-IXd exhibit a prominent ion at M-1 (IXa) or M-2 (IXb-d), which is presumably due to formation of the corresponding 3-dehydro derivative XLI (Rz,Rs = H or D, as appropriate) by loss of the C-3 hydrogen atom. Other, but much less intense, ions at m/e 337 and 321 are formed by loss of a methyl or methoxyl group from ring E.

XLI

IXa; R1 = Rz = R3 = H IXb; R i = Rz = D, R3 = H IXC; R i = D, Rz = R3 = H IXd; R1 = R3 = D, Ra = H

The ions at lower m/e values reveal extensive fragmentation of the molecule, at the dotted lines indicated in IX, with formation of the ions XLIIa-d, which still contain the intact @-carbolinering system, and the ion XLIIe at m/e 156. This last ion is the most intense in the spectrum at lower m/e values and is formed by the comparatively facile rupture of three allylically activated bonds. There are no intense ions in this

5.

ALKALOIDS OF

Mitragyna AND Ourouparia SPECIES

83

region of the spectrum which correspond to fragments containing an intact D/E ring system (82, 83).

XLIIa (m/e 184)

XLIIb (169)

XLIIc (170)

I1

CHz XLIIe (156)

\/Me

XLIId (225)

In contrast, the mass spectra of mitraphylline (XIIIa)and its deuterated analogs (XIIIb-XIIId) reveal a totally different fragmentation pattern. There is no ion at m/e 367 (M-l), owing presumably to the

,-Me

H

H' XIIla; R1 = Rz = R3 = H XIIIb; R1 = R2 = D, R3 = H XIIIC; R1 = D,Rz = R3 = H XIIld; R1 = R3 = D, Rz = H

dHz

CH2

,*Me

H' Me00

\/

XLIIIa

XLIIIb

XLIIIC

J. E. SAXTON

84

impossibility of obtaining a conjugated ammonium ion by loss of hydrogen from c-3. However, an ion of low intensity a t m/e 351 (M-17)is attributed to the 3-dehydro ion XLI, formed by loss of a hydroxyl group from the molecular ion, and simultaneous rearrangement. The loss of OD2 from all three deuterated species XIIIb-d confirms that the C-3 hydrogen (or deuterium) is lost, but the mechanism of rearrangement to XLI is not known. As in the spectrum of ajmalicine minor ions are observed a t M-15 and M-31, which correspond respectively to loss of methyl and methoxyl groups from the parent molecule. The most intense ion in the mitraphylline spectrum is found a t m/e 223, which is attributed to the intact D/E fragment (XLIIIa); a less intense ion a t m/e 208 is presumably the result of loss of the ring E methyl group from XLIIIa. Decomposition of XLIIIa gives the fragment XLIIIb, which is responsible for an ion of moderate intensity a t m/e 69. Associated with the ion XLIIIa is one a t m/e 145, which can be represented by XLIIIc ;these two ions are probably the result of fragmentation of mitraphylline a t the C-3 t o C-7 and the C-5 to C-6 bonds. Other ions in the "indole" region of the spectrum are observed a t m/e 130 (XLIIId; R = H ) , 144 (XLIIIe), 146 (XLIIId; R = OH), and 159 (XLIIIf). 8

63

i"7l IfCHZ ?7l

~ \ N / \ R rH XLIIId

V \ N / H

-

XLIIIe

XLIIIf

From a diagnostic point of view, the most pertinent features of the mitraphylline spectrum are the formation of the D/E fragment XLIIIa a t m/e 223, and the absence of ions a t M-1 and in the ((p-carboline" region, i.e., a t m/e 169, 170, and 184 (82).

XLIV

The mass spectrum of rhynchophylline (XXII) shows an expected similarity to that of mitraphylline. The high m/e region exhibits peaks corresponding to loss of hydroxyl, methyl, or methoxyl groups from the 2

OD represents a deuteroxyl group, i.e. one oxygen and one deuterium atom.

5.

ALKALOIDS OF

Mitragyna

AND

Ourouparia

SPECIES

85

parent ion, and there is also one at m/e 355, which corresponds to loss of the ethyl group ;no ion is observed a t m/e 384 (M-1). The most intense ion in the spectrum occurs at m/e 239, due to cleavage of the spiran ring C, with formation of the fragment XLIV. This is accompanied by ions a t m/e 224, 210, and 208, which correspond respectively to loss of methyl, ethyl, or methoxyl groups from XLIV. At lower m/e values the mass spectrum of rhynchophylline also contains the ions XLIIIb-XLIIIf (82).

VIII. Rotundifoline, Isorotundifoline (Mitragynol), and Speciofoline

Rotundifoline was initially assigned the molecular formula CZ2H,,N2O5,and was shown to contain two methoxyl groups (14, 19). One methoxyl group was presumed to be situated in position 6 of the oxindole nucleus, since the UV-spectra of rotundifoline and mitraphylline bear approximately the same relationship t o each other as do those of a-colubrine and strychnine (5). The presence of the second methoxyl group of rotundifoline in a P-alkoxyacrylic ester function was indicated by its UV- and IR-spectra, and by its behavior on acid hydrolysis, which resulted in decarboxylation of the initially formed acid (19). The presence of a corynantheine- or ajmalicine-type structure was also suggested by the selenium degradation of rotundifoline, which gave a pyridine base, C9H,,N (191, later (14) identified as 3,4-diethylpyridine. No isolated double bonds are present, since the molecul~is resistant to hydrogenation under mild conditions. These data resulted in the provisional formulation of rotundifoline as a methoxymitraphylline (or stereoisomer) (56). Recent analyses of rotundifoline and several of its salts, together with the molecular weight obtained by mass spectrometry, have shown that

H

Et

rotundifoline has the molecular formula CzZH,,N2O5, and the data accumulated in several laboratories have established its gross structure as XLV (26a, 26b, 26c, 84). The NMR-spectrum of rotundifoline exhibits

86

J. E. SAXTON

multiplet signals a t low fields (four protons) owing to the C-17 hydrogen atom and three adjacent aromatic protons. Detailed analysis of the signals owing to these aromatic protons, and comparison with the NMRspectra of model substances (e.g., ar-methoxyoxindoles and aspidospermine), show that the oxygenated substituent is situated at position 9. The presence of two methoxyl groups and an ethyl group is also readily confirmed by the NMR-spectrum (26a-c, 84). The oxygenated substituent at C-9 in rotundifoline is a cryptophenolic hyclroxyl group, which is strongly hydrogen-bonded to the tertiary amino group; this results in a broad, weak band at 2450 cm-1 in the IR-spectrum, which is absent in the spectra of the salts (26a). Rotundifoline is insoluble in dilute alkali, and its UV-spectrum shows no significant change on addition of alkali; the molecule is not methylated by diazomethane nor 0-acetylated by acetic anhydride and sodium acetate. However, it responds to the ferric chloride test and the spot test for hindered phenols, and couples with diazotized sulfanilic acid. Reaction with acetic anhydride and sodium acetate, or with dimethyl sulfate, leads t o an N,-acetyl or N,-methyl derivative, respectively; the latter still affords a positive ferric chloride reaction and shows IR-absorption at 3570 cm-I (26b). Acid hydrolysis of rotundifoline yields a crystalline aldehyde, rotundimp 240"-241" (XLVI; R = CHO),which folal (stipulatal),C19H24N203, contains no methoxyl groups (26a, 84). Wolff-Kishner reduction of this product leads t o rotundifolane (XLVI; R = Me) (84).

I

CH,R XLVI

The mass spectrum of rotundifoline shows a fragmentation pattern entirely in accord with the constitution XLV (26b). Comparison with the spectrum of rhynchophylline (XXXV)reveals that ring D in rotundifoline furnishes an identical series of fragments ; however, the fragments derived from rings A and B in rotundifoline show a displacement towards higher m/e values by 16 units. Thus, the principal peaks in the spectrum of rotundifoline occur at m/e 239 (XLIV), and at m/e 224,210, and 208, which correspond to the loss of methyl, ethyl, or methoxyl groups, respectively, from XLIV. The principal fragment derived from the

5.

ALKALOIDS OF Mitragyna AND

Ourouparia

SPECIES

87

hydroxyoxindole grouping is observed at m/e 146, owing to the 4-hydroxy derivative of XLIIId (R = H). The spectrum of N,-methylrotundifoline (N,-methylstipulatine) exhibits an identical series of peaks owing to the ring D fragments; at lower m/e values the principal peaks are at m/e 160 and 174-176, presumably owing to the N-methyl-4hydroxy derivatives of XLIIId (R = H), XLIIIe, and XLIIId (R = OH) (26b). The phenolic base, mitragynol (isorotundifoline), CzzH,,NzOB,mp 131"-132" or 199"-201", [a]",""- 7.4" (chloroform), is readily soluble in dilute alkali, in contrast to rotundifoline (26a, 84). Its IR- and NMRspectra closely resemble those of rotundifoline ; one notable difference, however, is the absence of the weak, broad band at 2450 cm-1 in the IR-spectrum (26a). The absence of any association between the phenolic hydroxyl group and Nb which this implies is reflected in the greater basicity (pK, 7.3) of isorotundifoline, compared with rotundifoline (pK, 5.4) (84),and the much more rapid quaternization with methyl iodide which is observed with the is0 base (26a). That the nature of the isomerism between rotundifoline and isorotundifoline is in principle the same as that between rhynchophylline and isorhynchophylline is shown by the interconversion of these isomers. Isorotundifoline can be converted almost quantitatively into rotundifoline by thermal means, e.g., by heating at 220"-250" for 10 minutes (84). In pyridine solution the equilibrium is also in favor of rotundifoline, though not so markedly, whereas in acid solution (e.g., aqueous acetic acid) the reverse is true, presumably owing to stabilization by N,-protonation of that isomer (isorotundifoline) in which Nb is not hydrogen bonded to the phenolic hydroxyl group (26c).The slow equilibration observed, and the complete isomerization of isorotundifoline by thermal means, can be contrasted with the behavior of the rhynchophylline, mitraphylline, and uncarine pairs of isomers, which undergo a much more facile interconversion, and in general yield equilibrium mixtures which contain significant amounts of both isomers (26c). These data are best explained by the structure XLVII for rotundifoline; this is the only arrangement of this tetracyclic system in which effective hydrogen bonding between N, and the phenolic hydroxyl group is permitted (26a, b). Comparison of the NMR-spectra of rotundifoline, isorotundifoline, and rhynchophylline indicates that the C-17 hydrogen atom in rotundifoline and isorotundifoline is oriented trans with respect to the carbomethoxyl group (26c, 84). At present no evidence is available concerning the stereochemistry at C-15 and C-20. The stereochemistry of isorotundifoline is less certain. Badger et al. (84)

S8

J. E. SAXTON

observed that short heating of isorotundifoline just above its melting point resulted in the formation of rotundifoline and an unknown substance of intermediate mobility on thin-layer chromatograms. When

0

x 1,VII

heated a t 220"-250", this mixture was converted entirely into rotundifoline. This was interpreted as indicating that rotundifoline and isorotundifoline may be isomeric a t both C-3 and C-7, in contrast t'o the other pairs of oxindole alkaloid isomers. Another view has been expressed by Beckett and Tackie, who obtained the same methiodide, designated isorotundifoline methiodide, from both rotundifoline and isorotundifoline. Since isorotundifoline quaternized much more rapidly than rotundifoline, isomerization about the C / D ring junction by inversion of Nb was postulated in the quaternization of rotundifoline. Isorotundifoline was therefore formulated as XLVIII (26a). However, the experimental evidence can be satisfactorily interpreted without invoking isomerization about the C / D ring junction. If

x LVIII it is assumed that the isomerization in rotundifoline and isorotundifoline is concerned with C-3 and/or C-7, then, whatever the stereochemistry of isorotundifoline, Nb is certain to be more reactive towards methyl iodide than the strongly hydrogen-bonded Nb in rotundifoline ; the possibility of a purely thermal isomerization of rotundifoline into

5.

ALKALOIDS OF

Mitragyna AND Ourouparia

SPECIES

89

isorotundifoline (even if i t only proceeds very slowly) can account for the formation of the same quaternary salt from both alkaloids. Speciofoline, CzzHZ,N,O5, mp 202"-204", pK, 6.3, [a]:" - 103" (chloroform), exhibits IR-, UV-, and NMR-spectra which show an extremely close similarity to those of rotundifoline and isorotundifoline, and it is clear that it is a third stereoisomer of the structure XLV. The insolubility of speciofoline in alkali, and the presence of a weak, broad band at 2500 cm-l in the IR-spectrum, indicate further that the stereochemistry a t C-3 and C-7 is the same as in rotundifoline (26c). The (2-17 hydrogen atom is provisionally assumed to be cis with respect to the carbomethoxyl group, although the evidence available at present appears not to be decisive. It seems more probable that rotundifoline arid speciofoline differ in the stereochemistry at C-20.

REFERENCES 1. D. Hooper, Pharm. J . 78, 453 (1907). 2. E. Field, J. Chem. Soc. 119, 887 (1921). 3. Raymond-Hamet and L. Millat, Bull. Sci. Pharmacol. 40, 593 (1933); Chem. Abstr. 28, 1041 (1934). 4. P. Denis, Bull. Classe Sci., Acad. Roy. Belg. 24, 653 (1938); Chem. Abstr. 33, 1741 (1939). 5. Raymond-Hamet, Ann. Pharm. Franc. 8, 482 (1950); Chem. Abslr. 45, 822 (1951). 6. H. R. Ing and C. G. Raison, J . Chern. Soc. p. 986 (1939). 7. L. Michiels and Leroux, Bull. Acad. Roy. Med. Belg. [5] 5, 403 (1925); Chem. Abstr. 20, 964 (1926). 8. L. Michiels,J. Pharm. Belg. 17, 1049 (1935); Chem. Abstr. 30, 7780 (1936). 9. L. Michiels, J . Pharm. Belg. 13, 719 (1931); Chem. Abstr. 23, 3070 (1932). 10. T. Nozoye, Chem. Pharm. Bull. (Tokyo) 6, 306 (1958); Chem. Abstr. 53, 2270 (1959). 11. P. Denis, Bull. Classe Sci., Acad. Roy. Belg. 23, 174 (1937); Chem. Abstr. 31, 3928 (1937). 12. P. Denis, Bull. Classe Sci., Acad. Roy. Belg. 25, 177 (1939); Chem. Abstr. 35, 2521 (1941). 13. Raymond-Hamet, Bull. Sei. Pharmacol. 46, 327 (1939); Chem. Abstr. 33, 7959 (1939). 14. G. M. Badger, J. W. Cook, and P. A. Ongley, J . Chcm. Soc. p. 867 (1950). 14a. N. Finch and W. I. Taylor, Tetrahedron Letters p. 167 (1963). 15. H. Kondo, T. Fukuda, and M. Tomita, J . Pharm. Soc. Japan 48, 321 (1928); Chem. Abstr. 22, 3166 (1928). 16. Raymond-Hamet and L. Millat, Compt. Rend. Acad. Sci. 199, 587 (1934). 17. Raymond-Hamet and L. Millat, J . Pharm. Chim. 20,577 (1934); Chem. Abstr. 29,4133 (1935). 18. Raymond-Hamet and L. Millat, Bull. Sci. Pharmacol. 42, 602 (1935); Ghem. Abstr. 30, 1379 (1936). 19. G. Barger, E. Dyer, and L. J. Sargeant, J . Org. Chem. 4, 418 (1939). 20. Raymond-Hamet, Compt. Rend. Acad. Sci.235, 547 (1952). 21. Raymond-Hamet, Compt. Rend. Acad. Sci.203, 1383 (1936). 22. Raymond-Hamet, Compt. Rend. SOC.Biol. 128, 777 (1938).

90

J. E. SAXTON

23, L. BIillat, Ann. Pharm. Franc. 4, 27 (1946); Chem. Abstr. 41, 1228 (1947). 24. H. Blake, J . Phczrm. Chim. 16, 413 (1932). 25. Raymond-Hamet, Bull. Sci. Pharmacol. 47, 194 (1940); Chem. Zentr. 112 11, 3074 (1941). 26. 0. Hesse, Ber. 11, 1546 (1878). 26a. A. H. Beckett and A. N. Tackie, Chem. I d . (London)p. 1122 (1963). 26b. J . B. Hendrickson and J. J. Sims, Tetrahedron Letters p. 929 (1963). 26c. A. H. Beckett, C. M. Lee, and A. N. Tackie, Tetrahedron Letters p. 1709 (1963). 27. H. Kondo and T. Ikeda, J . Pharm. Soc. Japan 57, 881 (1937); Chem. Abstr. 32, 1272 (1938). 28. J. C. Seaton, M. D. Nair, 0. E . Edwards, and L. Marion, Can. J . Chem. 38,1035 (1960). 29. H. Kondo and K. Oshima, J . Pharm. SOC. Japan 52,528 (1932); Chem. Abstr. 27,1345 (1933). 30. H. Kondo and T. Ikeda, J . Pharm. SOC.Japan 61, 416, 453 (1941); Chem. Abstr. 45, 2960 (1951). 31. H. Kondo, T. Nozoye, and M. Tobita, Ann. Rept. Itsuu Lab. 4, 77 (1953); Chem. Abstr. 49, 1077 (1955). 32. Raymond-Hamet, Bull. Soc. C'him. France 10, 129 (1943). 33. Raymond-Hamet, Compt. Rend. Acad. Sci.245, 1458 (1957). 34. Raymond-Hamet, Compt. Rend. Acad. Sci. 240, 1721 (1955). 35. Raymond-Hamet, Bull. Acad. Natl. Med. (Paris) 112, 513 (1934); Chem. Abstr. 29, 7493 (1935). 36. T. Pavolini, F. Gambarin, and G. Montecchio, Ann. Chim. (Rome) 40, 654 (1950); Chem. Abstr. 46, 4552 (1952). 37. A. D. Cross, F. E. King, and T. J. King, J . Chem. Soc. p. 2714 (1961). 38. J . B. La1 and S. Dutt, J . Indian Chem. Soc. 12,257 (1935). 39. E. Perrot, Raymond-Hamet, and P. Larrieu, Bull. Sci. Pharmacol. 37, 401 (1930); Chenz. Abstr. 24, 4856 (1930). 40. K. S. Grewa1,J. Pharmacol. Exptl. Therap. 46,251 (1932); Chem. Abstr. 27, 139 (1933). Biol. 114, 692 (1933). 41. Raymond-Hamet, Compt. Rend. SOC. 42. Raymond-Hamet, Compt. Rend. SOC.Biol. 115, 255 (1934); Chem. Abstr. 28, 2061 (1934). 43. Raymond-Hamet, Compt. Rend. SOC.Biol. 116, 1337 (1934); Chem. Abstr. 28, 7359 (1934). 44. L. Massion, Arch. Intern. Pharmacodyn. 48, 217 (1934); Chem. Abstr. 29, 884 (1935). 45. E. Perrot, Raymond-Hamet, and L. Millat, Bull. Acad. Natl. Med. (Paris) 116, 266 (1936); Chem. Abstr. 31, 5875 (1937). 46. E. Perrot, Raymond-Hamet, and L. Millat, Bull. Sci. Pharmacol. 43, 694 (1936); Chem Abstr. 31, 2290 (1937). 47. Raymond-Hamet, Arch. Intern. Pharmacodyn. 56, 303 (1937); Chem. Abstr. 32, 1327 (1938). 48. Raymond-Hamet, Arch. Intern. Pharmacodyn. 63, 336 (1939); Chem. Abstr. 34, 5179 ( 1940). 49. Raymond-Hamet, Compt. Rend. SOC.Biol. 134, 459 (1940). 50. Raymond-Hamet, Arch. Intern. Pharmacodyn. 66, 330 (1941); Chem. Abstr. 36, 2917 (1942). 51. J. W. Cook, J. D. Loudon, and P. McCloskey, J. Chem. SOC.p. 3904 (1952). 52. J. B. Hendrickson, Chem. I n d . (London)p. 713 (1961). 53. B, S. Joshi, Raymond-Hamet, and W. I. Taylor, Chem. Ind. (London)p. 573 (1963). 54. Raymond-Hamet, Compt. Rend. Acad. Sci.230, 1405 (1950).

5.

ALKALOIDS OF

Mitragyna

AND

Ozcrouparia SPECIES

91

55. J. W. Cook, R. M. Gailey, and J. D. Loudon, Chem. Ind. (London)p. 640 (1953). 56. J. D. Loudon, in “Recent Work on Naturally-Occurring Nitrogen Heterocyclic Compounds,” Report of Symposium, held at Exeter, July, 1955 (K. Schofield, ed.), Chem. SOC. (London),Spec. Publ. No. 3, 12 (1955). 57. J. C. Seaton, R. Tondeur, and L. Marion, Can. J . Chem. 36, 1031 (1958). 58. T. Ikeda, J . Pharm. SOC.Japan 63, 393 (1943); Chem. Abstr. 44, 7332 (1950). 59. H. Kondo and T. Nozoye, Ann. Rept. Itsuu Lab. 1, 71 (1950); Chem. Abstr. 47, 3856 (1953). 60. H. Kondo, T. Nozoye, and M. Tobita, Ann. Rept. Itsuu Lab. 5,80 (1954);Chem. Abstr. 49, 15920 (1955). 61. E. Wenkert and T. L. Reid, Chem. Ind. (London)p. 1390 (1953). 62. D. G. Markess and A. Burger, J . Am. Chem. SOC.71,2031 (1949). 63. H. Kondo, T. Nozoye, and H. Tsukamoto, Ann. Rept. Itsuu Lab. 6 , 5 3 (1955); Chem. Abslr. 50, 10110 (1956). 64. E. Wenkert, B. Wickberg, and C. Leicht, Tetrahedron Letters No. 22, 822 (1961). 65. J. B. Hendrickson, J . Am. Chem. SOC.84, 650 (1962). 66. N. Finch and W. I. Taylor,J. Am. Chem. SOC.84, 1318, 3871 (1962). 67. E. Wenkert, J. H. Udelhofen, and N. K. Bhattacharyya, J. Am. Chem. SOC.81, 3763 (1959). 84, 1320 (1962). 68. J. Shave1 and H. Zinnes, J . Am. Chem. SOC. 69. T. Ikeda, J. Pharrn. Soc. J a p a n 61, 460 (1941); Chem. Abstr. 45, 2960 (1951). 70. H. Kondo and T. Nozoye, Ann. Rept. Itsuu Lab. 2, 72 (195i); Chem. Abstr. 47, 3857 (1953). 71. H. Kondo and T. Nozoye, Ann. Rept. Itsuu Lab. 7, 44 (1956); Chem. Abstr. 51, 2825 (1957). 72. T. Ikeda, J. Pharm. SOC. Japan 62, 15, 38 (1942); Chem. Abstr. 45, 2961 (1951). 73. H. Kondo, T. Nozoye, and M. Tobita, Ann. Rept. Itsuu Lab. 5,84 (1954); Chem. Abstr. 49, 15921 (1955). 74. H. Kondo and T. Nozoye, Ann. Rept. Itsuu Lab. 7 , 49 (1956); Chem. Abstr. 51, 2825 (1957). 75. T. Nozoye, Chem. Pharm. Bull. (Tokyo)6, 300 (1958); Chem. Abstr. 53,2269 (1959). 76. Raymond-Hamet and L. Millat, Bull. Sci. Pharmacol. 41, 533 (1934); Chem. Abstr. 29, 799 (1935). 77. T. Nozoye, Ann. Rept. Itsuu Lab. 8 , 10 (1957); Chem. Abstr. 51, 16504 (1957). 78. J. C. Seaton and L. Marion, Can. J . Chem. 35, 1102 (1957). 79. T. Nozoye, Chem. Pharm. Bull. (Tokyo)6,309 (1958); Chem. Abstr. 53,2270 (1959). 80. Y. Ban and T. Oishi, Tetrahedron Letters p. 791 (1961); Chem. Pharm. Bull. (Tokyo) 11, 451 (1963). 81. L. D. Antonaccio and H. Budzikiewicz, Montash. Chem. 93, 962 (1962). 82. B. Gilbert, J. A. Brissolese, N. Finch, W. I. Taylor, H. Budzikiewicz, J. M. Wilson, and C. Djerassi, J . Am. Chem. SOC.85, 1523 (1963). 83. L. D. Antonaccio, N. A. Pereira, B. Gilbert, H. Vorbrueggen, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi, J. Am. Chem. SOC.84,2161 (1962). 84. G. M. Badger, L. M. Jackman, R. Sklar, and E. Wenkert, Proc. Chem. SOC.p. 206 (1963).

This Page Intentionally Left Blank

ALKALOIDS OF GELSEMIUM SPECIES J. E. SAXTON The University, Leeds, England I. Occurrence .........................................................

93

11. Gelsemine ..........................................................

95

1x1. Sempervirine ........................................................

107

IV. Gelsemicine.........................................................

110

V. Gelsedine...........................................................

112

VI. Gelseverine .........................................................

115

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

115

References

I. Occurrence The genus Gelsemium (Loganiaceae) consists of only three species, of which G. sempervirens Ait., the Carolina or yellow jessamine indigenous to the southeastern United States, has been extensively investigated. This plant has a history of medical use in the treatment of neuralgia and migraine, but it is now very rarely used, although it is still included in the recent (1959) British Pharmaceutical Codex. The presence of alkaloids in G. sempervirens was first established by Wormley ( I ) , who extracted an amorphous base from the rhizomes and roots. Subsequently, Sonnenschein ( 2 ) obtained an amorphous base, gelsemine, which was formulated, on the basis of analyses of the free base, the hydrochloride, and the platinichloricie, as C22H38N204. Gerrard (3)appears to have been the first to obtain crystalline, if impure, gelsemine, which was described as a brittle, transparent solid of very low melting point. Several crystalline salts were prepared, analysis of which indicated the composition C24H28Nz04. Later, Thompson (4)described the extraction of gelsemine, and claimed with some confidence that it possessed the molecular formula C54H69N4012 (sic); a second alkaloid, gelseminine, was isolated as a dark brown, resinous mass, yellow when powdered. I n the ensuing years the extraction of gelsemine and gelseminine was recorded by other investigatoi*s,but there were inconsistencies in the nomenclature used, and the substances isolated, like the earlier ones, were not necessarily homogeneous. Thus, Spiegel’s gelseminine, a white amorphous powder, 93

94

J. E. SAXTON

mp 120", which was formulated as C24H28N204 or C22H26N203, was probably impure gelsemine (5). Cushny isolated both gelsemine and gelseminine as white amorphous substances ( 6 ) ; he rejected Spiegel's formula, and proposed instead the formulas C49H63N5014 for gelsemine and C42H47N3014 for gelseminine. Goldner ( 7 ) later repeated and confirmed Spiegel's results, and also recorded the first attempts to degrade "gelseminine " (gelsemine). Analytically pure gelsemine was eventually obtained as its solvate from acetone, mp 178", by Moore, who firmly established its molecular formula as CzoHzzN202 (8). An amorphous, strongly basic fraction corresponding t o gelseminine was also obtained, together with evidence for a third, highly toxic, base. Somewhat later Sayre reported further extractions of Gelsemium roots (9-13) ; although he appears not to have obtained pure gelsemine, he succeeded (with Stevenson) in isolating a second alkaloid, sempervirine, a reddishbrown, strongly basic substance which was presumably responsible, a t least in part, for the color and intense fluorescence exhibited by alcoholic extracts of the roots ( 1 1, 12, 13). The remainder of the alkaloidal extract was isolated as two amorphous, ill-defined fractions, designated gelsemidine and gelsemoidine (9, 10, 13). After a lapse of several years, Chou (14) described the isolation of a third alkaloid, gelsemicine, which is the most highly toxic constituent of Gelsemium roots. The presence of sempervirine was confirmed, and a fourth alkaloid was obtained in an amorphous condition (15). Forsyth, Marrian, and Stevens (16) later recorded an improved extraction procedure, which afforded higher yields of gelsemine and sempervirine than had hitherto been obtained. From the "gelsemoidine " fraction, some of the rather unstable gelsemicine was recovered, together with two methiodides and a series of picrates of minor, unidentified bases. Schwarz and Marion (17, IS) subsequently isolated two further bases from 0. sempervirens roots, namely, gelsedine, C19H24N203, mp 172.5"-174", and the noncrystalline base, gelseverine, C21H24-26N203. The gelsemicine isolated by Janot et al. (19), ClgH24N203,mp 170°-171", is almost certainly identical with gelsedine; it is definitely not the gelsemicine described by Chou (18). The second Gelsemium species, G. elegans (Gardn.) Benth., has been intermittently studied, although in no case has botanically wellauthenticated material been used. The Chinese drug, Kou Wen, which is reputed to be derived from G. elegans, contains four alkaloids, koumine, C20H22N203, mp 170°, [a];;0-2650, kouminine, kouminicine, and kouminidine, mp 200" (20). Kouminine and kouminicine were not obtained crystalline and, in fact, it was not claimed that kouminicine, physiologically the most potent of the four alkaloids, was homogeneous.

6.

ALKALOIDS OF

Gelsemium SPECIES

95

Kouminine was later shown to be a mixture of gelsemine with other, unidentified bases, while the kouminidine fraction yielded a base, mp 299", to which the original name, kouminidine, was assigned (21). The Chinese Gelsemiurn Ta Cha Yeh may also be G. elegans; it is therefore noteworthy that the leaves and stems of this plant contain koumine and O ~315O, , was gelsemine (21, 22). A new base, kounidine, C Z I H ~ ~ N Zmp also shown t o be present (22). Twan Chan Tsao is probably another synonym for G. elegans; however, its alkaloidal content has not been thoroughly examined. The only extraction reported to date described the isolation of the hydrobromide of a base of unknown identity from the leaves, stems and roots (23). Another base of unknown identity is humanine, extracted from the rhizomes of G. elegans; this base has not been chemically studied, and there are at present only brief reports of its pharmacological activity (24, 25). The third Gelsemiurn species, G. rankinii Small, has only recently been added to the Index Kewensis, and has apparently not yet been investigated. Xempervirine also occurs (26) in Mostuea buchholzii Engl. (Loganiaceae). Two bases closely resembling, and in all probability identical with, gelsemine and sempervirine have also been extracted from the roots of Gabonese Mostuea stirnulans A. Chev., which, although highly toxic, is used by the local natives as a stimulant and aphrodisiac (27). Owing to the use of Gelsemium extracts in pharmacy in the late niheteenth and early twentieth centuries several processes for the chemical (28-32) and biological assay (31, 33-35) of Gelsemium extracts have been developed. Since the chemical assay simply gives the proportion of bases present by acid-base titration, whereas the biological assay depends largely on the proportion of minor alkaloids that exert a much more potent pharmacological activity than gelsemine, the two methods do not always give concordant results (31, 36). Hence, the biological assay is usually preferred. The pharmacological activity of the total Gelsemium extracts (37-40), and of the pure alkaloids gelsemine (41-46), sempervirine (47-49), and gelsemicine (50-55) has been investigated. Gelsemicine is by far the most toxic of these alkaloids, and appears to be mainly responsible for the characteristic effects of Gelsemium extracts. I n minute doses it stimulates respiration but in larger doses it causes death by respiratory paralysis. 11. Gelsemine

Gelsemine, C ~ ~ H Z Z N ~mp O Z178O, , [cY]~ + 15.9", is a strong, monoacidic base, ph', 9.37, which forms salts with only one equivalent of

96

J. E. SAXTON

acid; the other nitrogen is nonbasic (8, 16). The molecule contains one methylimino group and one active hydrogen atom, but no C-methyl or methoxyl groups (56, 57). A terminal double bond is also present, since catalytic reduction yields dihydrogelsemine (16, 58, 59, 60), which possesses one C-methyl group (60). The presence of a double bond is also shown by the reaction of gelsemine with bromine, which gives dibromogelsemine, C~oHz2NzOzBr2,mp 309" (61). It also accounts for the products obtained by the reaction of gelsemine with concentrated hydrochloric acid, which results in the formation of apogelsemine, isoand O ~chloroisoapogelsemine ), (CzoHz3NzOzCl) apogelsemine ( C Z O H ~ ~ N Z by addition of the elements of water or hydrogen chloride (56). Hydriodic acid similarly gives iodoisoapogelsemine, CZOHZ~NZOZI, which on reduction with zinc and acetic acid affords dihydrogelsemine (16); both halogen derivatives can be hydrolyzed to isoapogelsemine (16, 56). The chloro compound is converted by boiling with diethylaniline into an isomeride of gelsemine, mp 14Oo-i45O (56);this product was not claimed to be pure, and may have been a mixture of gelsemine with an isomericie, possibly the isogelsemine, mp 200"-202", prepared by treatment of gelsemine with zinc and hydrochloric acid (58). Potash fusion of geisemine gives a mixture of bases. together with an indole derivative (57), later identified as 3-ethylindole (62); since the nitrogen atom in this degradation product is unsubstituted, it is probable that in gelsemine Ni, carries the methyl group. Zinc dust distillation yields skatole and two bases, which are suspected to be isoquinoline derivatives (62). The UV-spectrum of gelsemine is remarkably similar t o those of strychnine (63) and 3,3-dimethyloxindole (64, 65), and its color reactions are also reminiscent of strychnine (21, 58). The IR-spectrum discloses the presence of a carbonyl group, which must be contained in a lactam function, since gelsemine is not an aldehyde, ketoile, or ester. This is confirmed by lithium aluminum hydride reduction, which yields dihydrodeoxygelsemine, CzoH24Nz0, a diacidic base (pK, 8.4 and 3.4), which, in contrast to gelsemine, exhibits typical dihydroindole color reactions. Gelsemine is therefore based on oxindole, and since no indolaceous products are formed in this reduction, it must be a 3,3disubstituted oxindole derivative (64, 65). This scanty evidence, together with the assumption that gelsemine is derived biogenetica.lly from tryptamine and dihydroxyphenylalanine, was used by Gibson and Robinson to deduce the constitution I for gelsemine (60). Aside from the oxindole fragment of the molecule, this formulation was largely speculative; the presence of an N-methyl group and a terminal double bond was well founded, but there was no incontrovert-

6.

ALKALOIDS OF

Gelsemium SPECIES

97

ible evidence in support of a hydroisoquinoline ring or an ether link. The nature of this second oxygen atom was inferred mainly from its

I

complete lack of reactivity. Moore's earlier preparation (56) of an acetyl derivative, which was regarded as possibly an 0-acetylgelsemine, could not be repeated by subsequent investigators (16, 58). More recently, however, the acetylation has been accomplished by acetic anhydride and ferric chloride, and the product identified as N,-acetylgelsemine, mp 11 1.5'-113", is in all probability identical with Moore's acetylgelsemine, mp 106'-108" (66). The presence of an oxindole nucleus and a terminal double bond has received further confirmation. Thus, zinc dust distillation of gelsemine under comparatively mild conditions yields 3-methyloxindole, while ozonolysis yields formaldehyde (66). Reaction of N,-methylgelsemine with sodium metaperiodate in the presence of a catalytic amount of osmium tetroxide also leads to fission of the terminal carbon atom. The product, C20H22N203, mp 192"-194', pE, 6.15, is an aldehyde, since it forms an oxime which can be dehydrated to a nitrile, C20H21N302, mp 240'-241', pK, 4.95. Further, the oxidation product contains no C-methyl groups, but Wolff-Kishner reduction leads to a base, CzoH24NzO2, containing one such group (67, 68). Hence, the double bond in gelsemine is present in a vinyl side chain, a deduction which immediately excludes from further consideration all the structures proposed for gelsemine prior to 1957, since these were all based on the assumption that the methylene group is exocyclic. The significant reduction in basicity accompanying the transformation from N,-methylgelsemine (pK, 7 . 2 5 ) to the aldehyde, and thence to the nitrile, implies a steric proximity of the basic nicrogen atom to the vinyl group (68). However, gelsemine is not a vinylamine, since it is a weaker base than dihydrogelsemine (69). I n general, cyclic vinylamines are stronger bases than their saturated analogues (70). Much attention has been directed to the possibility that gelsemine is an allylamine. The Emde reduction normally proceeds readily with allylammonium salts, but gelsemine methiodide, with sodium in liquid

98

J. E. SAXTON

ammonia, gives only impure gelsemine. N-Bromosuccinimide, which degrades allylamines to unsaturated aldehydes (e.g., C6H5. CH=CH-CHz-NMeZ

--+

CoHsCH=CH. CHO

+ HNMeZ),

yields a bromine-containing base, which is probably identical with the bromination product of gelsemine, namely, bromoallogelsemine (69). The Hofmann degradations of gelsemine and its derivatives are also anomalous. Moore had earlier reported (56) the failure of gelsemine methiodide to give a methine base on being heated with potassium hydroxide ; gelsemine was recovered unchanged. Subsequent attempts with gelsemine or dihydrogelsemine methiodide gave products designated as de-N-methylgelsemine and de-N-methyldihydrogelsemine, which did not respond to attempts to degrade them further by the same process (71). It has since been realized that the products of the first Hofmann stage are simply N,-methylgelsemine and N,-methyldihydrogelsemine, formed by intermolecular transmethylation (72, 73). Indeed, N,-methylgelsemine can also be prepared by heating gelsemine with tetramethylammonium hydroxide a t 150" (73). These results are not consistent with an allylamine formulation for gelsemine, since in this event, dihydrogelsemine would contain a t least one hydrogen atom /3 to the amino group, and should therefore be capable of degradation by the Hofmann procedure unless the steric requirements of a polycyclic bridged ring system exclude this possibility (69). It was considered probable, therefore, that the vinyl group and N,, are separated by two or more carbon atoms. The bromination of gelsemine provided evidence regarding the relative positions in the molecule of the double bond and the oxindole nucleus. The product, bromoallogelsemine, CzoHzlNzOzBr, in contrast to gelsemine, has a reactive para position in the benzene ring, since it couples with diazonium salts. Its UV-absorption, which approaches that of an indolenine, is also similar to that of gelsemine in alkaline solution, The in which it probably exists as the anion -C6H4-N=C-oD. formulation of bromoallogelsemine as I1 is supported by the I R spectrum, which has no carbonyl band, but has a conjugated C=N band a t 1600 em-'. Catalytic reduction of bromoallogelsemine produces bromodihydroallogelsemine, which exhibits dihydroindole absorption in the UV-region and possesses neither C=O nor C=N bands in the IR. Dilute acid effects the hydration of I1 to give bromohydroxydihydrogelsemine (111),which behaves as an oxindole derivative. Lithium aluminum hydride reduces bromoallogelsemine t o hydroxydeoxytetrahydrogelsemine (IV), by reduction to a dihydroindole nucleus and

6. ALKALOIDS

OF

Gelsemiurn SPECIES

99

hydrogenolysis of the bromine atom. Both I1 and I11 are reconverted by zinc and acid into gelsemine. Evidently, the oxindole oxygen atom and the vinyl side chain are so situated in the molecule that cyclization of

I1

I11

H

IV

the primary bromination product can occur. Hence, in the above partexpressions, n = 1, 2 , or 3, while X may be C, N, or 0 (66). The von Braun degradation of gelsemine has also been investigated. Reaction of the alkaloid with cyanogen bromide furnishes N-cyanoN-norgelsemine, which can be reduced to N-cyanodihydro-N-norgelsemine (74). The latter is also the product of reaction of dihydrogelsemine with cyanogen bromide (69, 74); acid hydrolysis leads to dihydro-N-norgelsemine, which can be reconverted by methylation into dihydrogelsemine methiodide (69). Attempts to degrade dihydro-Nnorgelsemine further have so far met with failure. BenzoyldihydroN-norgelsemine does not react normally with phosphorus pentachloride, but yields benzoylchlorodihydro-N-norgelsemine,which can be hydrolyzed t o chlorodihydro-N-norgelsemine. Since the chlorine is not readily removed by hydrogenation, it is presumed to be attached to the benzene ring. Methylation gives chlorodihydrogelsemine, which can also be prepared from dihydrogelsemine and phosphorus pentachloride (74). The environment of Nb in dihydrogelsemine has been investigated by means of its reaction with diethyl azodicarboxylate. Instead of the expected demethylation, the intermediate adduct, when warmed with dilute acid, gives an unstable base (not isolated), and diethyl hydrazodicarboxylate ; no formaldehyde can be detected. The new base may be characterized as its methyl ether, C21H26N203, which is readily formed in methanol in the presence of alkali, or as the perchlorate, CzoH23NzOz.Cl04. The latter is clearly the salt of an anhydro base. This behavior, coupled with the IR-evidence, suggests that the unstable product is a carbinolamine, C~oH24N203,formed by reaction of diethyl azodicarboxylate, not with Nb-Me, but with an adjacent methylene

100

J. E. SAXTON

group ( 7 5 ) .This is confirmed by the observation that the carbinolamine reduces ammoniacal silver nitrate and Fehling's solution, and gives a dinitrophenylhydrazone. Reduction of the methyl ether with sodium borohydride regenerates dihydrogelsemine, while oxidation with chromic acid gives a neutral product, C20H22N203, which is evidently a Isctam. Its IR-spectrum exhibits twin peaks at 1718 and 1693 cm-1; the former is owing to the oxindole carbonyl, and the latter to the second lactam group, which, from its frequency of absorption, must be contained in a five-membered ring. I n accordance with this formulation, the lactam is reduced by lithium aluminum hydride to tetrahydrodeoxygelsemine, and is hydrolyzed to an amino acid, which cannot be isolated because of the ease with which it recyclizes ( 7 5 ) . These transformations prove conclusively that N, must be attached t o a methylene group, and can be summarized in the partial formulas illustrated in Scheme 1. -NMe-CH-N-NHCOOEt

1

1

-NMe-CH-

I

COOEt

OH

I

OM?

%

-NTMe-CO-

kQ,,

c18,

SCHEMEI

Attempts have been made to obviate the experimental difficulties associated with 3,3-disubstituted oxindoles, by converting gelsemine into the corresponding indole derivative. This transformation should be feasible by reduction to the carbinolamine stage, followed by dehydration and simultaneous rearrangement ; analogous transformations have been accomplished with simple spirocyclic oxindoles. However, reduction of N,-methylgelsemine does not proceed smoothly, and although a carbinolamine can be obtained by means of sodium in liquid ammonia, it remains apparently unaffected by polyphosphoric acid a t 220" (76). The constitution (I)for gelsemine is clearly incapable of explaining all these experimental data, but those which were subsequently proposed were also inadequate in one or more respects. On the basis of the bromination experiments, and the probable proximity of the double bond to the oxindole group, the alternatives Va and Vb were advanced (66, 7 1 ) . I n both formulas, the position of the ether link was not specified, although

6.

ALKALOIDS OF

Gelsemium SPECIES

101

Va represented one stereochemically satisfactory example of the possibilities ( 7 1). The reaction of gelsemine with diethyl azodicarboxylate rendered the first of these untenable, but the structure VI (position of ether link again unspecified), advanced in lieu of Va or Vb (75) was in turn invalidated by the discovery that the molecule contains a vinyl side chain.

I

I 0 I

Va

VI

The conclusion of Habgood and Marion that the basic nitrogen atom in gelsemine is contained in a five-membered ring was rejected by other investigators, who obtained a dihydrogelsemine lactam, isomeric with Marion’s lactam, by very slow oxidation of dihydrogelsemine with potassium permanganate (77). This lactam was regarded as a &lactam, since it exhibited carbonyl absorption a t 1724 cm-1 (oxindole carbonyl) and 1661 cm-1 (a-lactam carbonyl). On the basis of this and the earlier evidence, three further formulas for gelsemine were proposed, of which two, namely V I I and VIII, may be quoted (the third proposal was

VII

VIII

sterically impossible). Although these structures accommodated many of the known facts of gelsemine chemistry and were also consistent with normal biogenetic requirements, their principal disadvantage was their inability to provide a convincing explanation of the formation of

102

J. E. SAXTON

bromoallogelsemine. Although the vinyl and oxindole groupings are much closer (Dreiding models) than is apparent from a cursory glance at V I I and VIII, they are still not sufficiently close to allow ring closure after addition of Br+ to the vinyl group. The degradations described thus far establish that gelsemine is a 3,3-disubstituted oxindole derivative ; the remainder of the molecule is probably a compact, highly stable, tetracyclic ring system which contains an ether link, a vinyl group separated from the oxindole carbony1 group by not more than three atoms, and an N-methyl group. This last grouping is flanked by at least one methylene group, is contained in a five- or six-membered ring, and is possibly also part of a hydroisoquinoline ring system. The failure of the attempted Hofmann degradations is probably a consequence of the steric impossibility of introducing a double bond into the ring system rather than of the absence of ,!Ihydrogen atoms. The attempts a t degradation thus having failed, the structure of gelsemine (IX) was eventually elucidated by the X-ray analysis of gelsemine hydrochloride and hydriodide (78). Independently and simultaneously, the same structure was proposed by Conroy and Chakrabarti (79) on the basis of biogenetic considerations, together with some new evidence, mainly derived from the NMR-spectra of gelsemine and its derivatives. 17

IX Gelsemine

X

The 60 mcjsec NMR-spectrum of gelsemine exhibits three symmetrical quartets a t 3.72, 4.95, and 5.10 T with spin-spin coupling constants appropriate to those expected for a vinyl grouping. The absence of further splitting in the quartet at 3.72 T , due to the C-19 proton, is firm evidence for the attachment of the vinyl group to a quaternary carbon atom. This conclusion is confirmed by the NMRspectrum of dihydrodeoxygelsemine ( I X ; CO -tCHz), which exhibits a similar group of three quartets, and by the spectrum of the aldehyde (X), in which the aldehydic proton appears as a sharp singlet a t 0.11 T ;

6.

ALKALOIDS OF

Gelsemiurn SPECIES

103

Qp

-NMe

A(,

0

Ac

XIV

XI11

the quaternary nature of the adjacent carbon atom is thus established. If it is assumed that the vinyl group in gelsemine is the biogenetical equivalent of that in corynantheine, the presence of the grouping I 18

19

20

21

CHz=CH-C-CHzN I C

/ b\

is to be expected; the presence of a quaternary center at C-20 implies that an additional bond to this carbon atom has been formed, a situation which has no previous parallel among the indole alkaloids. In such a grouping, the (3-21 methylene protons should give rise to two doublets in the NMR-spectrum, in agreement with observation (79). The NMR-spectra of gelsemine and dihydrodeoxygelsemine exhibit four more groups of peaks at low fields owing to other CH protons adjacent to nitrogen or oxygen. Two of these show the splitting pattern expected of the methylene protons in the grouping :

>H-cH~-oThe two remaining protons are probably separate tertiary ones, but the substitution on the adjacent carbon atoms could not be deduced, owing to the complexity of the spectra and the incomplete resolution obtained.

104

J. E. SAXTON

A new degradation of gelsemine afforded convincing proof of the presence of the part-structure CH2=CH--CCHaN,, I

/

I

and also established that C-20 is contained in a five-membered ring. Oxidation of N,-acetyldihydrodeoxygelsemine methohydroxide (XI) with alkaline potassium permanganate gave the behaine (XII), which on concerted decarboxylation and /3-elimination in dilute solution in boiling dimethylformamide gave the unsaturated base (XIII). The NMR-spectrum of this base shows a symmetrical pair of peaks consistent with the presence of an exocyclic methylene group, while hydrogenation furnishes a base which contains one additional C-methyl group. Finally, oxidation of XI11 with osmium tetroxide and periodate gave a ketone (XIV), which was evidently a cyclopentanone derivative, since its IR-carbonyl absorption was observed at 1748 cm-1.

These data, together with the following biogenetic considerations, allowed the structure I X for gelsemine t o be deduced (79). The proposed precursor XV bears a close structural relation to rhynchophylline (XVI), and can be derived from tryptamine and dihydroxyphenylalanine according t o well-accepted principles. Dehydrogenation of XV at the 5,6-position then gives an intermediate (XVII) which can undergo an COOH HOCH=C

I

= & ? F H

CH. CHO HOOC~NCHOH xv

H

CHO

6.

ALEALOIDS OF

Gelserniurn SPECIES

105

COOH

COOH

N M e

‘

N

H

[yH

O

CH2 I CHO

H XVIII

XVII

COOH

OHC

4

IX I

CH2 I CHO

internal Michael addition with formation of the vital C-6 to C-20 bond. The product (XVIII) can further cyclize by means of a normal Mannich condensation to give XIX which can be converted into gelsemine (IX) by obvious transformations. The formation of the C-5 to C-16 bond has its parallel in ajmaline and sarpagine, but so far there are no other alkaloids known which contain the C-6 to C-20 bond. An alternative proposal for the biogenesis of gelsemine postulates its derivation from indole-3-acetaldehyde and the carbohydrate metabolite prephenic acid (80),in accordance with the tenets of Wenkert’s theory of indole alkaloid biogenesis. The structure I X affords a satisfactory explanation of most of the reactions of gelsemine already discussed. The failure of the attempted Hofmann degradations is clearly owing to the impossibility of introducing a double bond in the cage-like ring system ; the conversion into an indole derivative as attempted by Witkop would also have introduced considerable strain. Marion’s conclusion that his lactam was five-membered is apparently sound, since, in gelsemine, N, is enclosed in both a five- and a six-membered ring; hence, the lactam must exhibit IR-absorption characteristic of a y-lactam. The structure of Teuber’s lactam is obscure; on the basis of IX, it is not readily apparent how a product exhibiting 6-lactam absorption can be obtained by simple oxidation. Witkop’s isolation of an isoquinoline derivative is consistent with IX, which

106

J . E. SAXTON

contains an intact 4,7-disubstituted hydroisoquinoline ring system. since However, Witkop's base is unlikely to be 4,7-dimethylisoquinoline, degradative loss of the oxindole grouping could hardly be expected to lead to a methyl group. Finally, the structure of bromoallogelsemine (XX, assuming the new ether ring t o be six-membered) shows it to be a highly compact, heptacyclic molecule, in which rings C and D are chairare boatshaped and rings E and F (N,-C~-CI~-C~~-CZO-CZI) shaped.

&+ \

xx

21

CHzBr

I n spite of all the attempts discussed above, and others which have subsequently been reported (81, 82, 83), no satisfactory means of degrading gelsemine has yet been discovered, with the possible exception of Conroy's conversion of gelsemine into the ketone (XIV).A by-product in the formation of XIV was a neutral substance exhibiting carbonyl absorption a t 1720 cm-1. This might well be the result of reverse Mannich decomposition of the ,!3-amino ketone system present in XIV; unfortunately, the structure of this product has not been established and the degradation has not been pursued beyond this point. Roe and Gates attempted to apply the Hofmann degradation to N,-methylnorgelsemine carbinol methohydroxide (XXI),but instead of the expected elimination of formaldehyde (dotted arrows in XXI), an intramolecular displacement occurred, and the only product isolated was the unstable trimethylene oxide derivative (XXII).The latter could not

6.

ALKALOIDS OF

Gelsemium SPECIES

107

be degraded further, owing to the ease with which it reverted to its progenitor (XXI)even at room temperature. The interconversion of X X I and X X I I provides an impressive demonstration of the rigid geometrical requirements of the gelsemine ring system (81).

HI. Sempervirine Sempervirine, C1gH16N2, mp 258"-260", is an optically inactive orange-colored base, which gives intensely fluorescent (blue-violet) solutions in ethanol, even in very high dilution (15, 84, 85). It is a monoacidic base, which forms well-defined salts with one equivalent of acid ; the nitrate is remarkable for its very low solubility ( < 1 in 20,000) in water (84, 85). The molecule contains one active hydrogen atom but no methylimino groups (86). When subjected to catalytic hydrogenation, sempervirine absorbs three moles (palladium catalyst) or five moles (platinum oxide catalyst) of hydrogen, but in neither experiment could a simple hydrogenation product be obtained crystalline, owing to the ease with which the products resinified or absorbed atmospheric oxygen. From the platinum-catalyzed hydrogenation, a base, C19H24N20, of unknown constitution was eventually obtained. Sempervirine is stable in acid solution, and resists vigorous treatment with phosphorus and hydriodic acid. The Hofmann and van Braun degradations failed to give any useful results, as also did attempts at oxidation using a variety of reagents ( 16). The nature of the ring system present in sempervirine was revealed by selenium or palladium-charcoal dehydrogenation, which gave an isomer yobyrine (XXIII),together with a small amount of yobyrone (XXIII ; with C=O a t position 14). When boiled with Raney nickel in xylene solution, tetrahydroisoyobyrine (XXIV) was obtained. These are

XXIII

XXIV

xxv

characteristic degradation products of the yohimbine ring system ;hence, sempervirine was believed to be the unsaturated pentacyclic base XXV. This formulation appeared to account satisfactorily for the color and the

108

J. E. SAXTON

complex UV-spectrum of sempervirine (86, 87). The high basicity observed (pK, 10.6) was explained as the result of proton addition at C-14, the sempervirinium ion thus being regarded as XXVI (87).

XXVI

This structure for sempervirine was soon disproved by two syntheses of the compound XXV, which was shown to possess properties quite different from those reported for sempervirine (88, 89). I n particular, the compound XXY was a weak base, pK, 5.0, which exhibited only a faint fluorescence in ethanol solution. Its UV-spectrum in acid solution was markedly different from that in neutral solution. This was in contrast to sempervirine, which exhibits virtually identical spectra in neutral and acid solutions, but a markedly different spectrum in alkaline solution (88, 89). Independently and simultaneously, the anhydronium base structure XXVII was proposed for sempervirine (90, 91). This formulation clearly provides a more satisfactory explanation of the color of the alkaloid, its high basicity, and the absence of absorption corresponding to an imino group in the IR-spectrum (90, 91). I n contrast, the synthetic base XXV exhibits a sharp NH band a t 3480 cm-1 (88). Support for the constitution XXVII was obtained from the alkylation of sempervirine, which was shown to proceed with formation of N,-alkyl derivatives (90, 91). Thus, sempervirine methochloride gave, on distillation with selenium, a base which was identified as N,-methylyobyrine (90, 92). The structure XXVII is also consistent with the high dipole moment (8D) of seinpervirine (91, 93). The hydrogen atom obtained in the Zerewitinow determination is now attributed to the presence in XXVII of a virtual y-substituted picolinium system (90). The structure XXVII for sempervirine was readily established by a neat synthesis of N,-methylsempervirinium salts (94). The lithium derivative of N,-methylharman (XXVIII ; R = Me) condensed with 2-isopropoxymethylenecyciohexanone(XXIX) to give a product which, when treated with aqueous acid, gave N,-methylsempervirinium salts (XXX). Sempervirine itself was also obtained by an analogous synthesis

6.

ALKALOIDS OF

Gelsemium SPECIES

from the lithium derivative of harman (XXVIII ; R oxymethylenecyclohexanone (95).

109

= H) and

2-isoprop-

CHOCHMez

XXVIII

XXIX

A second synthesis of sempervirine was later reported by Swan (96)) according to the following reaction sequence. 3-Cyano-5,6,7,8-tetrahydroisoquinoline reacted with 3-ethoxypropyl magnesium bromide t o

SXXII

XXXIII

\ I \

XXXIV

xxxv

110

J. E. SAXTON

give the keto base (XXXI), which was converted by treatment with hydrogen bromide in acetic acid into 1,2,3,4,7,8,9,10-octahydro- 1oxobenzo(b)-pyridocolinium bromide (XXXII). Application of the Fischer synthesis to the phenylhydrazone of XXXII then gave the pentacyclic quaternary chloride XXXIII, which on dehydrogenation with tetrachloro-o-benzoquinonegave sempervirine, isolated as the insoluble nitrate (96). Recently, Potts and Liljegren (97) have synthesized 415920-yohimbene (XXXV) by condensation of 3-chloroacetylindole with 5,6,7,8-tetrahydroisoquinoline, and reduction of the tetracyclic quaternary chloride (XXXIV) so obtained with lithium aluminum hydride in tetrahydrofuran (THF) solution. Since Al5~20-yohimbene(XXXV) has previously been converted by palladium-maleic acid dehydrogenation into 5,6dihydrosempervirine (XXXIII) (98) and sempervirine (99), this constitutes a third formal synthesis of sempervirine.

IV. Gelsemicine Gelsemicine, C2oHzsN204, mp 171"-172", [aID-142", is a monoacidic base, which contains one C-methyl and two methoxyl groups (14,16,18) ; the presence of a methylimino group, reported earlier, has since been shown to be erroneous. The UV-spectrum of gelsemicine has a comparatively unfamiliar appearance and it is difficult to deduce the nature of the chromophore present. The spectrum shows some resemblance to those of 7-methoxyindole derivatives, although it does not possess a pronounced minimum at 250-260 mp in neutral solution (18, 100). I n alkaline solution, the curve is more like that expected for a substituted indole derivative. The IR-spectrum of gelsemicine in Nujol suspension exhibits twin maxima in the carbonyl region, but this does not necessarily imply the presence of two carbonyl groups, since dihydrogelsemine shows the same behavior. The spectrum also contains an imino band and a peak of high intensity at 1500 cm-1, which is characteristic of gelsemicine and all its derivatives. Gelsemicine forms acetyl and benzoyl derivatives which are neutral, show no imino band in the IR-spectrum, and have UV-spectra identical with that of gelsemicine. Hence, it can be inferred that one nitrogen atom is secondary and basic, while the other is tertiary (18). The structure of gelsemicine has been elucidated by the X-ray crystallographic investigations of Przybylska and Marion. Gelsemicine and N-methylgelsemicine form crystalline hydrobromide salts, but these are unsuitable for X-ray analysis since in both derivatives the asym-

6.

ALKALOIDS OF

Gelsemium SPECIES

111

metric unit contains two molecules which are, therefore, not crystallographically equivalent (101). One of the two crystalline forms of N-methylgelsemicine hydriodide exhibits the same phenomenon, which is probably related to the fact that in the solid phase gelsemicine (which has not been examined crystallographically) shows a split carbonyl band in the IR-spectrum. This is attributed to hydrogen bonding between the NH group and the carbonyl group in some, but not all, the molecules in the crystal. The second form of N-methylgelsemicine hydriodide crystallizes with only one molecule per asymmetric unit, and was therefore chosen for detailed X-ray examination (102, 103). From this, i t was established that the structure and absolute configuration of ( -)-N-methylgelsemicine are as given in XXXVI (R = Me) ; gelsemicine is therefore XXXVI; R =H.

0 Me XXXVI Gelsemicine (R=H)

The structure of gelsemicine thus deduced possesses two unusual features. The N-methoxyoxindole grnup is the first of its kind to be discovered among natural products. Secondly, the biosynthesis of gelsemicine appears to be a curiously aberrant process, which results in the omission of the C-21 one-carbon unit (the so-called “berberine bridge” carbon atom). However, a possible biogenetic route to gelsemicine is not hard to devise. The proposed precursor (XXXVII) differs essentially from Conroy’s precursor (XV) of gelsemine only in the omission of the berberine bridge carbon atom (C-21), and can readily be derived by accepted principles from 6-methoxytryptamine and dihydroxyphenylalanine (Robinson) or prephenic acid (Wenkert), with omission of the simple Mannich stage which normally serves to introduce the C-21 carbon atom. Addition of the primary amino group to the unsaturated aldehyde system in XXXVII gives XXXVIII, which by oxidation affords the intermediate imine XXXIX ; this can then cyclize by an internal Mannich reaction. Decarboxylation of the product (XL), adjustment of oxidation levels, formation of the ether ring, and intro-

112

J. E. SAXTON

duction of the N-methoxy group then lead to gelsemicine (XXXVI; R =H). COOH

H

H HOOC C H O H

XXXVII COOH

C'H2CHO

('HzCHO H XXXIX

XXXVIII

COOH

*

Me0 &C€12CH0

XXXVI;H.= H

H XL

V. Gelsedine Gelsedine, C I ~ H ~ ~ N mp Z O 172.5"-174", ~, [ c ~ ] ~ " - 1 5 9(chloroform), ~ can be isolated from the residual alkaloids of Gelsemium sempervirens, after removal of gelsemine and sempervirine, by benzoylation of the crude mixture and separation of the tertiary bases from the neutral benzoyl derivatives. Chromatography of these benzoyl derivatives yields a crystalline mixture, from which the secondary bases can be recovered by hydrolysis and separated by fractional crystallization of the perchlorates. By this process, the secondary basic fraction can be divided into its components, gelsemicine and gelsedine (18).The above physical constants for gelsedine, together with the data for benzoylgelsedine [mp 251 "-252" [cr]gO-116" (chloroform)},strongly suggest that it is identical with the gelsemicine (mp 171", [a]= -160"; benzoyl derivative, mp 262", [a],, -117") of Janot et al. (19).

6 . ALKALOIDS OF Gelsemium SPECIES

113

Gelsedine contains one methoxyl group and one C-methyl group, and is a strong, monoacidic base (pK, 9.3). Its UV-spectra in neutral (A,, 258 mp, Amin 235 mp, shoulder at 280 mp) and alkaline (A, 244 mp, shoulder at 280 mp) solutions are characteristic of 1,%disubstituted oxindoles (18, 19). Gelsedine is slowly hydrogenated, in acid solution using Adams’ catalyst a t slightly elevated temperatures, to a hexahydrodemethoxy derivative, by elimination of the methoxyl group and saturation of the benzene ring. The methoxyl group is also lost when the alkaloid is reduced with lithium aluminum hydride ; the product was not obtained in an analytically pure condition, but a second oxygen atom appeared to have been lost, presumably the oxindole oxygen (18). From these data it was inferred that gelsedine is a 1,3-disubstituted oxindole derivative, and a secondary base. The substituent on N , was presumed to be a methyl group, owing to an apparently satisfactory Herzig-Meyer determination for one methylimino group. The position of the methoxyl group was unknown, but in view of its ready elimination the most probable site of attachment appeared to be the benzylic position, i.e., CL to the lactam carbonyl group. However, the /3 position to this group could not be excluded with certainty, since an alkaline solution of gelsedine gradually developed a bright yellow color on being kept for a short while, and its UV-spectrum then resembled that of N-methylisatin. The third, inert oxygen, atom was assumed to be contained in an ether linkage.

I

OMe XLI Gelsed ine

xLII

I n a recent investigation Wenkert et al. (104) have shown that gelsedine (XLI) is 11-demethoxygelsemicine. The facile elimination of the N-methoxyl group has been confirmed, since it can also be removed by means of lithium in liquid ammonia or by sodium and butanol. The demethoxygelsedine so obtained gives a n N,-acetyl derivative (XLII), which exhibits IR-absorption a t 3413 cm-l. This absorption cannot be

114

J. E. SAXTON

owing to a hydroxyl group, since Nh-acetyldemethoxygelsedine resists acetylation even with ketene; it is much more likely to be due to an oxindole imino group, liberated by fission of an N,-methoxyl group. The presence of an N,-methoxyoxindole function in gelsedine excludes the presence of an N,-methyl group, reported earlier. This is confirmed by the NMR-spectrum of gelsedine, which contains no peak attributable t o an N-methyl group. Consequently, the apparently satisfactory Herzig-Meyer determination of one methylimino group must have been the result of interference by the N-methoxyl group. The gradual appearance in alkaline solution of a UV-spectrum resembling that of N-methylisatin was not observed by Wenkert et al., and it is not now clear t o what this earlier observation must be attributed. The NMR-spectrum of gelsedine confirms the presence of an o-disubstituted oxindole ring (four-proton multiplet a t 2.56-3.19 T), a methoxyl group (singlet a t 6.04 T), and an ethyl group (two-proton quartet centered at 8.28 T, and three-proton triplet centered at 9.0 T); the presence of this last grouping was also diagnosed by means of a modified Kuhn-Roth oxidation. There is no signal due to hydrogen adjacent to the oxindole carbonyl group, nor to the hydrogen of an oxindole imino group. I n contrast, the spectrum of XLII contains no peak due to a methoxyl group, but exhibits a concentration-dependent signal due to an oxindole imino hydrogen at 1.03 T. Comparison of the spectra of gelsedine and its N,-acetyl derivative reveals in the latter a downfield shift of three groups of protons, which can be attributed to the grouping 'CH-NH-CH-E

/

I

t

in gelsedine. The remaining downfield signals (two-proton multiplet a t 5.70-5.80 T and one-proton signal centered at 6.5 T ) are probably due to protons situated on carbon atoms adjacent to the ether oxygen atom. The great similarity of the IR-spectra of gelsedine and gelsemicine has already been noted (18); the above characteristics of the NMRspectrum of gelsedine also suggest a close structural affinity with gelsemicine. Comparison of the NMR-spectra of gelsedine and its Nhacetyl derivative with those of gelsemicine and its Nh-acetyl derivative confirms this relationship, since the two pairs of spectra are almost superposable. The only significant differences involve the aromatic methoxyl singlet (6.20 T ) and the three aromatic protons (multiplets in the region 2.74-3.52 T) in the spectra of gelsemicine and Nh-acetylgelsemicine ; in contrast, the spectra of gelsedine and N,-acetylgelsedine exhibit no peak due to an aromatic methoxyl group, but contain

6.

ALKALOIDS OF

Gelsemium SPECIES

115

multiplets in the aromatic region corresponding to four protons. Consequently, gelsedine (XLI) can be formulated as 1l-demethoxygelsemicine (104).

VI. Gelseverine Gelseverine, C21H24- 26N203, is the tertiary base isolated from the minor alkaloids of Gelsemium sempervirens. It has not yet been obtained crystalline but can be characterized as its perchlorate, mp 250"-252", or methiodide, mp 259"-261". The molecule contains one methylimino group and two methoxyl groups, but no C-methyl groups. The UV- and IR-spectra are consistent with its formulation as a 1,3,3-trisubstituted oxindole derivative (18). REFERENCES T. G. Wormley, Am. J.Phawn. 42, 1 (1870); Chem. Zentr. p. 678 (1870). F. L. Sonnenschein, Ber. 9, 1182 (1876). A. Gerrard, Pharm. J . 13, 641 (1883);Jahresber. p. 1353 (1883). F. A. Thompson, Pharm. J. 17, 805 (1887); Chem. Zentr. p. 802 (1887). L. Spiegel, Ber. 26, 1054 (1893). A. R. Cushny, Ber. 26, 1725 (1893); Arch. Ezptl. Pathol. Pharmakol. 31 (Pt. I), 49 (1893). 7. M. Goldner, Ber. Dent. Pharm. Ges. 5 , 330 (1895); J . Chem. SOC.70, 657 (1896). 8. C. W. Moore, J . Chem. SOC.97, 2223 (1910). 9. L. E. Sayre, Pharm. J . 86, 242 (1911); C h m . Abstr. 5 , 2146 (1911). 10. L. E. Sayre, J . Am. Pharm. Assoc. 1, 458 (1912); Chem. Abstr. 6, 1817 (1912). 11. A. E. Stevenson and L. E. Sayre, J . Am. Pharm. Assoc. 4, 60 (1915);J . Chem. SOC. 108, 85 (1915). 12. A. E. Stevenson and L. E. Sayre, J . Am. Pharm. Assoc. 4, 1458 (1915); Chem. Abstr. 10, 804 (1916). 13. L. E. Sayre,J. Am. Phurm. Assoc. 8, 708 (1919); Chem. Abstr. 13, 2972 (1919). 14. T. Q. Chou, Chinese J . Physiol. 5 , 131 (1931); Chem. Abstr. 25, 4085 (1931). 15. T. Q. Chou, ChineseJ. Physiol. 5 , 295 (1931); Chem. Abstr. 25, 5736 (1931). 16. W. G. C. Forsyth, S. F. Marrian, and T. S. Stevens, J . Chem. Soc. p. 579 (1945). 17. H. Schwarz and L. Marion,J. Am. Chem. Soc. 75, 4372 (1953). 18. H. Schwarz and L. Marion, Can. J . Chem. 31, 958 (1953). 19. M. M. Janot, R. Goutarel, and W. Friedrich, Ann. Pharm. Franc. 9, 305 (1951); Chem. Abstr. 46, 2553 (1952). 20. T. Q. Chou, ChineseJ. Physiol. 5, 345 (1931); Chem. Abstr. 26, 806 (1932). 21. Y. F. Chi, Y . S. Kao, and Y. T. Huang, J . Am. Chem. Soc. 60, 1723 (1938). 22. T. Q. Chou, C. H. Wang, and W. 0. Cheng, Chinese J . Physiol. 10, 79 (1936); Chem. Abstr. 30, 4270 (1936). 23. Y. F. Chi, S. T. Lee, and C. S. Lee, J . Chem. Eng. (China),5 , 40 (1938); Chem. Abstr. 33, 7045 (1939). 24. T. Mineshita, J a p a n J . Med. Sci. I V , Pharmaeol. 8, No. 1, Proc. J a p a n P h a ~ m a ~ o l . Soc. p. 49 (1934); Chem. Abstr. 29, 1512 (1935). 1. 2. 3. 4. 5. 6.

116

J. E. SAXTON

25. T. Mineshita, J. Orient. Med. 22, No. 3 (1935); Ber. ges. Physiol. exptl. Pharmakol. 87, 687 (1935); Chem. Abstr. 31, 7537 (1937). 26. E. Gellert and H. Schwarz, Helv. Chim. Acta34, 779 (1951). 27. R. Paris and H. Moyse-Mignon, Compt. Rend. Acad. Sci. 229, 86 (1949). 28. L. E. Sayre, Drug. Circ. 1, 55 (1911); Chem. Abstr. 5 , 2529 (1911). 29. L. E. Sayre, Am. J . Pharm. 84, 193 (1912); L'hem. Abstr. 6, 1811 (1912). 30. H. Neugebauer, Phurm. Ztg. 78, 1077 (1933); Chem. Abstr. 28, 1137 (1934). 31. E. E. Swanson and C. C. Hargreaves, J . Am. Phnrm. Assoc. 17, 23 (1928). 32. G. R. Lynch and N. Evers, Analyst 66, 108 (1941). 33. P. S. Pittenger, J . Am. Pharm. Assoc. 12, 1063 (1923). 34. H. Neugebauer, Apotheker-Ztg. 45, 80 (1930); Chem. Abstr. 24, 1703 (1930). 35. B. V. Christensen and L. G. Gramling, J . Am. Pharm. Assoc. 26, 32 (1937); 27, 1208 (1938). 36. L. S. Wu, Bull Natl. Formulary Comna. 15, 68 (1947); Chem. Abstr. 41, 5260 (1947). 37. T. Okanishi, Proc. J a p a n Pharm. Soc. p. 62 (1933); Gkem. Abstr. 29, 6951 (1935). 38. E. M. de Espanes, Cornpt. R e n d . Soc. Biol. 127, 128, 1088 (1938); 129, 386 (1938). 39. A. Risi, 2. Biol. 99, 446 (1939); Chem. Abstr. 34, 2071 (1940). 40. K. Uhlenbroock, M. Schweer, and L. Maschmann, Arzneimittel-Forsch. 9,419 (1959) 41. Raymond-Hamet, Compt. Rend. Acad. Sci. 205, 1449 (1937). 42. Raymond-Hamet, Compt. Rend. SOC.Biol. 151, 31 (1957). 43. E. M. de Espanes, Compt. Rend. SOC.Biol. 127, 1002, 1176, 1178, 1433 (1938); 129, 546 (1938). 44. R. Cahen and E. M. de Espanes, Compt. Bend. Acnd. Sci. 206, 280 (1938). 45. F. G. Henderson and K. K. Chen, J . Am. Pharm. Assoc. 32, 178 (1943). 46. 0. Eichler, F. Hertle, and I. Staib, Arzneimittel-Forsch. 7, 349 (1957). 47. Raymond-Hamet, Compt. Rend. Soc. Biol. 126, 690 (1937). 48. Raymond-Hamet and E. Rothlin, Compt. Rend. Soe. Biol. 117, 754, 859 (1934); 135, 478 (1941). 49. Rayhond-Hamet, Compt. Rend. SOC.Biol. 150, 1110 (1956). 50. H. C. Hou, ChineseJ. Physiol. 5 , 181,279 (1931); Chem. Abstr. 25,4062,5934 (1931). 51. H. C. Hou, ChineseJ. Physiol. 6, 41, 281 (1932); Chem. Abstr. 26, 57, 6018 (1932). 52. Raymond-Hamet, Compt. Rend. SOC.Biol.126, 1151 (1937). 53. K. K. Chen, R. C. Anderson, and E. B. Robbins, Quart. J . Phurm. 11, 84 (1938). 54. K. K. Chen and T. Q. Chou, Chinese J . Physiol. 14, 319 (1939); Chem. Abstr. 34, 4810 (1940). 55. K. K. Chen and H. M. Lee, ChineseJ. Physiol. 14, 489 (1939); Chem. Abstr. 34, 4811 (1940). 56. C. W. Moore,J. Chem. SOC.99, 1231 (1911). 57. L. Marion, Can. J . Res. B21, 247 (1943). 58. T. T. Chu and T. Q. Chou,J. Am. Chem. SOC.62, 1955 (1940). 59. M. Ferreiro, Ann. Sci. Nut. Botan. Biol. Vegetale 6, 148 (1945); Chem. Abstr. 41, 136 (1947). 60. M. S. Gibson and Sir Robert Robinson, Chem. I n d . (London)p. 93 (1951). 61. T. Q. Chou and T. T. Chu, J . Am. Chem. SOC.63, 827 (1941). 62. B. Witkop, J. Am. Chem. SOC.70, 1424 (1948). 63. M. M. Janot and A. Berton, Compt. Rend. Acad. Sci. 216, 564 (1943). 64. M. Kates and L. Marion, J . Am. Chem. SOC.72, 2308 (1950). 65. M. Kates and L. Marion, Can. J . Chem. 29, 37 (1951). 66. R. Goutarel, M. M. Janot, V. Prelog, R. P. A. Sneeden, and W. I. Taylor, Helw. Chim. Acta 34, 1139 (1951).

6. ALKALOIDS OF Gelsemiurn SPECIES

117

L. Marion and K. Sargeant, J . Am. Chem. SOC.78, 5127 (1956). L. Marion and K. Sargeant, Can. J . Chem. 35, 301 (1957). G. Jones and T. S. Stevens, J . Chem. SOC.p. 2344 (1953). R. Adams and J. E. Mahan, J . Am. Chem. SOC.64, 2588 (1942). R. Goutarel, M. M. Janot, V. Prelog, and R. P. A. Sneeden, Helv. Chim. Acta 34, 1962 (1951). 72. T. Habgood, L. Marion, and H. Schwarz, Helv. Chim. Acta 35, 638 (1952). 73. V. Prelog, J. B. Patrick, and B. Witkop, Helv. Chim. Acta 35, 640 (1952). 74. T. Habgood and L. Marion, Can. J . Chem. 32, 606 (1954). 75. T. Habgood and L. Marion, Can. J . Chem. 33, 604 (1955). 76. B. Witkop and J. B. Patrick, J . Am. Chem. SOC.75, 2572 (1953). 77. H. J. Teuber and S. Rosenberger, Chem. Ber. 93, 3100 (1960). 78. F. M. Lovell, R. Pepinsky, and A. J. C. Wilson, Tetrahedron Letters No. 4, 1 (1959). 79. H. Conroy and J. K. Chakrabarti, Tetrahedron Letters No. 4, 6 (1959). 80. A. Chatterjee and S. Ghosal, J . Sci. Ind. Res. ( I n d i a ) ,20B, 454 (1961). 81. A. M. Roe and M. Gates, Tetrahedron 11, 148 (1960). 82. E. Wenkert and J . H. Hansen, Iowa State J. Sci. 34, 163 (1959); Chem. Abstr. 54, 24837 (1960). 83. J. H. Hansen, Dissertation Abstr. 21, 1757 (1961). 84. V. Hasenfratz, Compt. Rend. Acad. Sci. 196, 1530 (1933). 85. V. Hasenfratz, Bull. SOC.Chim. France 53, 1084 (1933). 86. R. Goutarel, M. M. Janot, and V. Prelog, Ezperientia 4, 24 (1948). 87. V. Prelog, Helv. Chim. Acta 31, 588 (1948). 88. 0. E. Edwards and L. Marion, J . Am. Chem. SOC.71, 1694 (1949). 89. G. A. Swan,J. Chem. SOC.p. 1720 (1949). 90. R. B. Woodward and B. Witkop, J . Am. Chem. SOC.71, 379 (1949). 91. R. Bentley and T. S. Stevens, Nature 164, 141 (1949). 92. B. Witkop,J. Am. Chem. Soc. 75, 3361 (1953). 93. K. A. Jensen, Acta. Chem. Scand. 3, 1447 (1949). 94. R. B. Woodward and W. M. McLamore, J . Am. Chem. SOC.71, 379 (1949). 95. R. B. Woodward and W. M. McLamore, Unpublished work, quoted in ref. 92. 96. G. A. Swan, J . Chem. SOC.p. 2038 (1958). 97. K. T. Potts and D. R. Liljegren, J . Org. Chem. 28, 3066 (1963). 98. E. Wenkert and D. K. Roychaudhuri, J . Am. Chem. SOC.80, 1613 (1958). 99. C. F. Huebner, quoted in ref. 98. 100. Raymond-Hamet, Compt. Rend. Acad. Sci. 235,84 (1952). 101. M. Przybylska, Acta Cryat. 14, 694 (1961). 102. M. Przybylska and L. Marion, Can. J . Chem. 39, 2124 (1961). 103. M. Przybylska, Acta Cryst. 15, 301 (1962). 104. E. Wenkert, J. C. Orr, S. Garratt, J. H. Hansen, B. Wickberg, and C. L. Leicht, J . Org. Chem. 27, 4123 (1962). 67. 68. 69. 70. 71.

This Page Intentionally Left Blank

-CHAPTER

7-

ALKALOIDS OF PICRALIMA NITIDA J. E. SAXTON The University, Leeds, England

I. Occurrence ........................................................

119

11. Akuammigine ............................................ 111. Akuammicine ......................................................

....... Akuammidine (Rhazine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pseudoakuammigine.. . ................ Akuammine .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....................... Picraline. ............. ....................................... Akuammiline ...................................................... Akuammenine . . . . . . . . . .................................... References ........................................................

123

IV. Pseudoakuammicine.. . . . . . V. VI. VII. VIII. IX. X.

131 134

145 147

155 155 155

I. Occurrence The alkaloids discussed in the succeeding paragraphs occur in the seeds of Picralima nitida Th. and H. Durand (syn. P. klaineana Pierre), which is widely but sparsely distributed throughout tropical Africa. The seeds of this plant are used by the West African natives as a specific for malaria, and as an antipyretic (1, 2). The first chemical examination of Picralima was conducted by Clinquart, who obtained a crystalline alkaloid, mp 242"-243", and a second, amorphous base (3, 4). Various color reactions of these specimens were described, but beyond this no attempts were made t o characterize them completely. The alkaloids were shown to occur chiefly in the seeds, but they are also present in the leaves and bark (3, 4). A thorough investigation by Henry and Sharp ( l ) ,and later by Henry ( 5 ) ,of the constituents of Picralimaseeds,resulted in the isolation of eight alkaloids, of which the principal one, akuammine, appears to be identical with Clinquart's crystalline base. I n common with many other tropical and subtropical plant extracts used locally for their alleged antimalarial properties, the reputation of 119

120

J . E. SAXTON

the Picralima drug has not survived careful clinical and pharmacological testing. As a result of the demonstration that Picralima is inactive in avian, and therefore presumably in human, malaria (6), the chemical investigations were temporarily abandoned. More recent pharmacological inquiries have shown, however, that some of the alkaloids have pronounced activity (7-1 3). I n particular, akuammine augments the hypertensive effects of adrenaline, although, when administered individually without adrenaline, it exerts a hypotensive effect (7-1 1). Akuammine also shows a significant local anesthetic action, almost equal to that of cocaine (8), while a minor alkaloid, akuammidine, is approximately three times as effective (12). On the other hand, akuammigine appears to be almost devoid of physiological activity (14). I n recent years the chemical investigations on the Picralima bases have been resumed, and the structures of several of them have been elucidated. A ninth alkaloid, picraline, has been isolated from the seeds by fractional crystallization and chromatographic separation of the total alkaloid fraction (15, 15a). The occurrence of the typical Picralima alkaloids has occasionally been observed in other genera. Thus, akuammine (vincamajoridine) has been isolated from the leaves and branches of Vinca major L. (16 , 17, 18), and from the roots of Yinca rosea L. (syn. Lochnera rosea Reichb. and Catharanthus roseus G. Don.) (19). Rhazine, one of the constituent bases of the leaves and roots of Rhaxya stricta Decaisne (20), has recently been shown (21) to be identical with akuammidine. A further source of akuammidine has been found in the leaves of Vinca diflormis Pourr. (22). Finally, akuammicine N,-methochloride has recently been identified as one of the quaternary constituents of the root and stem bark of Hunteria eburnea Pichon (22a).

11. Akuammigine Akuammigine is obtained, as its hydrochloride, as the least soluble fraction during the extraction of the crude Picralima alkaloids with dilute hydrochloric acid. The free base, C Z ~ H Z ~ N HzO, Z O ~ . forms colorless, square plates from aqueous ethanol, mp 113", -42" (EtOH), pK,, 6.58. Attempted dehydration of the solvate leads to decomposition. The molecule contains one methoxyl group and one C-methyl group; Kuhn-Roth determination gives a value of 120% for one such group, which was initially interpreted as indicating the presence of two C-methyl groups ( 2 3 ) . However, a similar discrepancy occurs with a closely related alkaloid, tetrahydroalstonine, which is known to contain only one C-methyl group.

7. ALKALOIDS

OB

Picralima nitida

121

The UV-spectrum of akuammigine is almost identical with that of the indole alkaloids containing a heterocyclic ring E, and corynantheine (2, 23, 24). The presence of the /3-alkoxyacrylic ester group is confirmed by the absorption bands at 1706,1684,and 1629 cm-l in the IR-spectrum. The appearance of a second ester band must be owing to intramolecular hydrogen bonding, since it is not observed in the spectrum of a chloroform solution of akuammigine. These data indicate a close relationship with ajmalicine, which is supported by chemical evidence. Thus, selenium dehydrogenation gives alstyrine ; lithium aluminum hydride reduction gives a primary alcohol, akuammigol (I),which possesses a typical indole UV-spectrum, with a deep minimum a t 250 mp. The formulation of this product as an allylic alcohol, stereoisomeric with tetrahydroalstonol, is further shown by its

anomalous behavior with acids and the comparatively facile hydrogenolysis of the primary alcohol function. Since akuammigol does not form crystalline salts with acids, it may be characterized as its methiodide and 0-acetate. From this evidence, akuammigine is formulated as a stereoisomer of ajmalicine and tetrahydroalstonine (11)(2). I n analogy with these two alkaloids, the ester group of akuammigine is not readily saponified, nor is the double bond easily hydrogenated; in contrast to tetrahydroalstonine, however, dinitrophenylhydrazine has no effect on the molecule (2). The IR-spectrum of akuammigine in the 2800 cm-1 region was initially interpreted as indicating that it belongs to the pseudo or epiallo series, which possess /3 hydrogen at C-3 ( 2 5 , 2 6 ); however, the correct deduction to be drawn from the IR-spectrum is that in its preferred conformation akuammigine possesses equatorial hydrogen a t C-3 (27). Since akuammigine is dehydrogenated with palladium and maleic acid much more readily than 3-isoajmalicine, which was also assigned to the pseudo or epiallo series, it was tentatively suggested that akuammigine belongs to the pseudo series (28). The product of this dehydrogenation, isolated as its perchlorate, is identical with alstonine perchlorate ; hence, akuammigine is 3-isotetrahydroalstonine.

J . E. SAXTON

122

Recent work on the stereochemistry of the heteroyohimbine alkaloids has necessitated a revision of these conclusions. Thus, the conversion of tetrahydroalstonine (11) into 19-corynantheidone (111, with u hydrogen at C-20), and isomerization of the latter with sodium methoxide t o the (111,with ,fl hydrogen a t C-20), known 18,19-dihydro-19-corynantheone demonstrates that in tetrahydroalstonine, and therefore in akuammigine, the D/E ring junction must be cis (29). The hydrogenation of tetradehydroakuammigine with Raney nickel in alkaline solution furnishes

IV Akuemmigine

I11

tetrahydroalstonine ; since, in general, the cis D/E tetradehydro derivatives of the yohimbine and heteroyohimbine alkaloids afford predominantly the a110 stereoisomers in this reaction, tetrahydroalstonine must belong to the all0 series, and 3-isotetrahydroalstonine (akuammigine, IV) must belong to the epiallo series. The large spin-spin coupling constant observed in the one-proton octet owing to the C-19 proton in the NMRspectrum of tetrahydroalstonine indicates a trans diaxial orientation of

Akuammigine

the C-19 and C-20 hydrogen atoms (29). Therefore, these two hydrogen atoms must also be oriented trans in akuammigine, which can consequently be completely represented by the conformation V.

7 . ALKALOIDS OF Picralima nitida

123

111. Akuammicine

Akuammicine, C20H22N202, crystallizes from aqueous ethanol as colorless plates, mp 182', pK, 7 . 4 5 ,and ischaracterized by itsremarktibly high rotatory power, [a]:" -745' (5,23). The molecule contains C-methyl and methoxyl groups, and exhibits a UV-spectrum of approximately indoline type, readily distinguishable from the spectra of its congeners (23, 24). The IR-spectrum discloses the presence of an imino group and a 1,2-disubstituted benzene nucleus (747cm-l), and it has a pronounced band at 1658 cm-l, initially ascribed to an amide function (23, 30). The Otto reaction of akuammicine gives a deep blue color, reminiscent of benzylidenestrychnine, which also has a high rotatory power. Akuammicine was therefore regarded as having a similar chromophore, namely, Ph-NH-CO-C=C-

I

I

adjacent to an asymmetric center (30). However, vigorous reduction of the alkaloid with lithium aluminum hydride (it is resistant to this reagent under the usual conditions) gives a crystalline base, ClgH22N2, which behaves in all respects as an indoline derivative (31). The accompanying change of constitution is equivalent to the replacement of -COOMe by a methyl group, and this is supported by an increase in the C-methyl content. The presence of the ester group has been confirmed by saponification, although the intractable amino acid could not be characterized. Cautious acid hydrolysis is accompanied by decarboxylation, the product being an easily oxidized base, ClgH20N2. These data are explained by the assumption that the chromophore in akuammicine is I I Ph-NH--C=C-COOMe

which, as a vinylog of phenylurethane, should readily suffer decarboxylation after hydrolysis. It is significant that the analogous compound, .COOMe, also exhibits pronounced absorption a t Ph-NH-CMe=CH 1658 cm-I. Combining these results with biogenetic considerations, Aghoramurthy and Robinson arrived at the constitution VI for akuammicine. On the basis of this formulation, the aromatic base akuammicyrine, obtained in addition to skatole and 3-ethylindole on selenium dehydrogenation of akuammicine, was assigned the indololepidine structure VII. Synthesis of VII was achieved by the Fischer indole reaction followed by deusing cyclohexanone 4-methylquinolinyl-7-hydrazone, hydrogenation of the product. The identity of the synthetic base (VII)

124

J. E. SAXTON

with akuammicyrine was not established, but the absorption spectra are SO closely similar that they must contain the same basic ring system (31). Some further reactions of akuammicine consistent with the structure V I have been reported by Smith and Wrdbel(32). Thus, the isolation of acetaldehyde from the ozonolysis products of akuammicine formally establishes the presence of the double bond in an ethylidene grouping. The acid hydrolysis of akuammicine was earlier stated (31) to give a sensitive indolenine base, C18&&2, mp 147O-148O. I n evacuated sealed tubes, however, akuammicine reacts with hydrochloric acid a t 115' to give a stable, isomeric, indolenine base, mp 80"-84", in very high yield (32). The constitution of Robinson's indolenine base has not been elucidated, but Smith's base is almost certainly VIII, since this formulation accounts extremely satisfactorily for its behavior. Reduction of VIII with lithium aluminum hydride gives an indoline base, which is presumably the expected indoline analog of VIII. Reduction with methanolic potassium borohydride, however, gives an indole base, base A, C18H22N2 (X),which must be formed by reverse Mannich fission of the C-3 t o C-7 bond in VIII in the proton-donating solvent used (see

VI Akuammicine

VII

.

CH * Me VIII

CH Me

IX

1 X

7.

ALKALOIDS OF

Picralima nitida

125

arrows in VIII), followed by reduction of the immonium ion (IX). I n accordance with this suggestion the borohydride reduction of VIIT in acid solution, in which Nb is protonated, affords the corresponding

if*

x1

XI1

C-Fluorocurarine chloride

O d V '

XI11 Dihydrodeoxyisostrychnine

XIV

1

1. [OI 2. 0.4N HClllOOoll hr

I COOH XVI

XVII 19,20-Dihydroakuarnmicine

XVIII Strychene

126

J . E. SAXTON

indoline base. Further, the analogous degradation of 19,2O-dihydroakuammicine to the 19,20-dihydro derivative of X proves that the isolated double bond is not involved. Zinc dust distillation of the indole base A (X) gives comparatively high yields of 3-ethyl-2-methylindole and 3-ethylpyridine. Thus, it is possible to account for all the carbon atoms of X, and the structure of the latter is confirmed, with the exception of the point of attachment of the asterisked carbon atom (C-16)to the piperidine ring (32). Robinson’s structure for akuammicine (VI) has recently received ample confirmation by several independent routes. The first of these involved the acid hydrolysis of C-fluorocurarine chloride (XI), which gave a deformyl compound identified as XII. The same quaternary chloride was also obtained by hydrolysis and decarboxylation of akuammicine N,-methochloride with dilute hydrochloric acid (33). I n later experiments (33, 34) the permanganate oxidation of dihydrodeoxyisostrychnine (XIII) in neutral solution, followed by treatment with sulfur dioxide in methanol and then with hot, dilute methanolic hydrochloric acid, was shown to give strychene (XVIII), which is the hydrolysis and decarboxylation product of 19,20-dihydroakuammicine(XVII). This oxidation proceeds via the pyridone (XIV) and 19,2O-dihydronorfluorocurarine (XV). A second oxidation product of the pyridone (XIV), obtained in very small (3.5%) yield, is 19,20-dihydroakuammiciiie (XVII); this is presumably formed from the intermediate acid (XVI; not isolated) by esterification and hydrolysis of the amide function during the acid treatment. Thus, strychnine can be degraded in a few stages directly to 19,20-dihydroakuammicine (33, 34). The route adopted by Edwards and Smith (35) involved the conversion of strychnine and akuammicine into the common transformation product XXI. Reduction of akuammicine (VIa)with zinc and methanolic sulfuric acid gave 2,16-dihydroakuammicine (XIX), which, in contrast to akuammicine, exhibited a typical indoline UV-spectrum and normal ester carbonyl absorption in its IR-spectrum. Hydrogenation of XIX (XX), which, on equilibration gave 2,16,19,20-tetrahydroakuammicine with sodium methoxide and magnesium methoxide, was isomerized to isotetrahydroakuammicine (XXI), in which the ester grouping has the preferred equatorial configuration. The convergent route from strychnine proceeded via the Wieland-Gumlich aldehyde (XXII), which was converted first into its oxime and then into the related nitrile acetate (XXIII) by dehydration and concomitant acetylation with acetic anhydride in pyridine. Alkaline hydrolysis of XXIII, followed by esterification, gave the methyl ester (XXIV), the stereochemistry of which was established by its reduction with lithium aluminum hydride

7.

ALKALOIDS OF

Picralima nitida

127

COOMe XIX 2.1 6-Dihydroakuammicine

VIa Akuammicine

COOMe

xx 2,16,19,20-Tetrahydroakuammicine

dOOMe XXI Is0tetrahydroakuammicine (methyl 2~,16a,20a-curan-l’I-oate)

XXIII

IV“ I

H

1’-H COOMe

XXII Wieland-Gumlich aldehyde

I

CHzOH

XXIV

xxv

XXVI

bH2 Me XXVII Anhydro-Z,16-dihydroakuammicinol

I

128

J . E. SAXTON

to the Wieland-Gumlich diol, and by the configurational stability of its carbomethoxy grouping. Finally, hydrogenolysis of the allylic hydroxyl group in XXIV and reduction of the double bond gave a base, methyl 2/3,16a,20c(-curan-17-oate (XXI), which was identical with isotetrahydroakuammicine (35).

XXVIII

XXIX Geissoschizoline

A further reference compound in the strychnine-akuammicine series is the alcohol XXVI, which was prepared by Janot and his collaborators (36, 37) from 2,16-dihydroakuammicine (XIX) by saponification, followed by lithium aluminium hydride reduction of the 2,16-dihydroakuammicinic acid (XXV)so obtained. The alcohol XXVI was identical with one prepared earlier (38) from the Wieland-Gumlich aldehyde (XXII) by sodium borohydride reduction to the related diol and subsequent removal of the allylic hydroxyl group by zinc-acetic acid reduction of the corresponding allylic bromide. The lithium aluminum hydride reduction of 2,16-dihydroakuammicine (XIX) and the acid XXV gives rise to different diols; hence, it is clear that epimerization of C- 16 accompanies saponification of the ester XIX. The stereochemistry of the acid XXV follows from that of the alcohol XXVI, which was already known ; consequently, the stereochemistry of 2,16-dihydroakuammicine (XIX)is confirmed (37).I n accordance with this deduction the lithium aluminum hydride reduction of XIX, followed by catalytic hydrogenation of the product, affords geissoschizoline (XXIX) (37,37a). The lithium aluminum hydride reduction of akuammicine itself NZ 196", , furnishes an oxygen-free, secondary indoline base, C ~ ~ H Z ~mp which contains one exocyclic methylene group and one ethylidene group (36, 37). In contrast to an earlier report (31))this base contains only one C-methyl group, and is regarded as the pentacyclic base XXVII, i.e., anhydro-2,16-dihydroakuammicinol. This proposal is supported by the hydrogenation of XXVII to the tetrahydro derivative XXVIII, which can also be prepared by an independent route from geissoschizoline (XXIX) (37, 37a). There remains for discussion the very interesting decomposition

7.

ALKALOIDS OF

Picralima nitida

129

which occurs when akuammicine is heated in methanol solution (39). At 80" the product is a betaine, C1gHzoNzOz (XXX), which contains indole and pyridinium ion chromophores, since its UV-spectrum is almost identical with the summation spectrum of 2,3-dimethylindole

VIa

6OOMe

Me

bOOMe XXXI

/

CHMe

/

V ' N

Me COOMe XXXII

XXXIII

xxx

XXXIV

xxxv and 3-picoline ethobromide. The formation of XXX from akuammicine is explained by the degradation of a C-16-protonated species to XXXI (compare VIII --f IX), followed by proton removal a t C-14 to give an intermediate enamine (XXXII) that can break down, by an essentially irreversible retro-Michael reaction, to the ester (XXXIII). The subsequent stages to XXX are then unexceptional. This interpretation of the decomposition of akuammicine is consistent with the observations

130

J. E. SAXTON

that 2,16-dihydroakuammicine and akuammicine methiodide do not break down under comparable conditions; the former is simply epimerized a t c-16, and the latter remains unaffected. At 140" in methanol, a much more extensive decomposition of akuammicine occurs, and the products isolated are 3-ethylpyridine and 2-hydroxycarbazole (XXXV). The formation of 3-ethylpyridine is presumably the result of a normal Hofmann degradation of the pyridinium ester corresponding to XXXIII ; the other product should consequently have been methyl 3-vinyl-2-indolylacetate (XXXIV), but this was not obtained. It was accordingly suggested that the 2hydroxycarbazole obtained was formed from XXXIV by intramolecular nucleophilic attack of the vinylogous enamine methylene group on the carbomethoxy group (arrows in XXXIV), followed by elimination of methanol and aromatization (39).

IV. Pseudoakuammicine Pseudoakuammicine, C Z ~ H ~ Z N Z mp O ~187", , pK, 7.47, was first isolated from Picralima seeds by Henry ( 5 ) )who reported its rotation as [N]E" -83". Robinson and Thomas (30) noted that its I R - and UVspectra are very similar to those of akuammicine, but they also pointed out that pseudoakuammicine does not exhibit an unusually high negative rotation, which was also believed to be associated with the presence of the chromophore I I PhNH-C=C-COOMe

attached to an asymmetric center. I n more recent investigations, Edwards and Smith (40) have carried out further extractions of Picralima seeds, and have found pseudoakuammicine to be optically inactive; its 2,16- and 19,2O-dihydro derivatives are also optically inactive. Pseudoakuammicine and its two dihydro derivatives have IR-spectra identical with those of the corresponding derivatives of akuammicine ; hence, pseudoakuammicine is ( i )-akuammicine. This conclusion was confirmed by resolution of pseudoakuammicine by fractional crystallization of the ( + )-tartrate. The more soluble salt gave akuammicine, and the less soluble salt gave ( + )-akuammicine, mp 181"-182.5", [N]?' + 720" (MeOH). On being mixed with an equimolecular proportion of akuammicine, ( + )-akuammicine gave pseudoakuammicine.

7.

ALKALOIDS OF

PicraEima nitida

131

Pseudoakuammicine is the first racemic base to be discovered in the strychnine-yohimbine series of alkaloids, and the question of its origin naturally arises. The only stage in the extraction of Picralima seeds during which racemization of akuammicine might have occurred invoived prolonged percolation with hot methanol ; however, as already discussed, akuammicine is not racemized under these conditions but suffers a more extensive decomposition. I n any event, such a racemization would necessarily involve fission of the 3,7 and 15,16 bonds, followed by a nonspecific resynthesis, which is considered to be a very unlikely possibility. It was therefore suggested that, in the plant, pseudoakuammicine is produced by a nonspecific biosynthesis; this would accord with its formation from a tryptophan-phenylalanine precursor, but not from an optically pure prephenic acid derivative (40).

V. Akuammidine (Rhazine) Akuammidine, C21H24N203, mp 234", [c(]gO+ 24" (MeOH), is a monoacidic base which contains one methoxyl group and one C-methyl group, but no methylimino groups. The molecule also contains two active hydrogen atoms and gives rise to monoacetyl and monobenzoyl derivatives ( 5 , 41, 42). Its UV-spectrum is characteristic of the true indole alkaloids (14,41,42) ; hydrogenation gives a dihydroakuammidine by saturation of a double bond, but this is not conjugated with the indole nucleus, since the UV-spectrum of the product is also typically indolic (42). The IR-spectrum of akuammidine discloses the presence of an imino group, a strongly hydrogen-bonded hydroxyl group, and an ester group. The hydroxyl group accounts for the formation of monoacyl derivatives, while the ester group must be a carbomethoxy group, since hydrolysis of akuammidine gives akuammidinic acid, C20H,,Nz03, and reduction with lithium aluminum hydride gives akuammidinol, C20H24N202, the related primary alcohol. The position of the alcohol function in akuammidine relative to the carbomethoxy group is established by its reaction with potassium t-butoxide in benzene at SO", which results in the loss of formaldehyde by retroaldol cleavage. Thus, akuammidine contains a ,f?-hydroxypropionic ester residue, analogous to that in echitamine and dihydropseudoakuammigine. The product, dehydroxymethylakuammidine, on reduction with lithium aluminum hydride, furnishes a primary alcohol, dehydroxymethylakuammidinol, which is identical with normacusine-B (10-deoxysarpagine; XXXVI). Hence, akuammidine must have the structure and absolute configuration shown in XXXVII, and the only

132

J.

E. SAXTON MeOOC

XXXVI

XXXVII

XXXVIII Vincemejine

XLII Vincamedine

XXXIX

XLIII

I

1 HOC-Hz

OHQ

XL Vowhalotine

XLIV

/

/

XLV

/KB€?.

XLVI

7.

ALKALOIDS OF

XLV

T

133

Picrulima nitida

XLVI Polyneuridine LiAlH.

KOBU~

MeOOC __t

LiAlHd

I

Me

XXXVIIa Akuammidine

XLI

features of the molecule which remain to be established are the stereochemistry a t C-16 and the configuration at C-19 (42). The configuration of (3-16 can be deduced by correlation of akuammidine and its derivatives with vincamajine and congeneric alkaloids of Vincu species (43).Vincamajine has the structure XXXVIII, in which the configuration of C-16 is rigidly defined, owing to its presence in an ajmaline-like ring system. Oxidation of vincamajine with chromium trioxide in pyridine gives an indole aldehyde (XXXIX), which on reduction furnishes the hydroxy ester (XL), identical with the alkaloid voachalotine. Comparison of the molecular rotation differences between akuammidine (XXXVIIa) or its 0-acetyl derivative and akuammidinol (XLI) on the one hand with the molecular rotation differences between voachalotine or its 0-acetyl derivative and the related diol (AT,-methyl derivative of XLI) on the other, suggests that the configurations of C-16 in akuammidine and voachalotine, and therefore in vincamajine, are not the same. Hence, voachalotine must be epimeric a t C-16 with N,-methylakuammidine ; unfortunately, this conclusion could not be directly verified, owing to the failure to realize the N,-methyIation of akuammidine. Consequently, the converse route involving N,-demethylation of the vincamajine-voachalotine series was adopted. Vincamedine (XLII), which is the 0-acetyl derivative of vincamajine, on oxidation with chromium trioxide in pyridine, affords a mixture of products from which a small yield of the indolenine (XLIII) can be

134

J. E . SAXTON

isolated. I n alkaline solution the acetate grouping in XLIII is readily hydrolyzed with concomitant fission of the C-7 to (2-17 bond, but the expected aldehyde ester (XLIV) cannot be isolated, since it is readily deformylated to the ester (XLV), which is identical with dehydroxymethylakuammidine. However, reduction of the indolenine derivative (XLIII) with potassium borohydride gives the desired hydroxy ester (XLVI), which is the Aspidosperma alkaloid polyneuridine. Polyneuridine and akuammidine should be epimeric a t C- 16 ;this is confirmed by lithium aluminum hydride reduction of the two bases, which gives the same diol, akuammidinol (XLI) (43, 44). The configuration of C-16 in polyneuridine (XLVI) follows from its interrelation with vincamedine; hence, akuammidine has the configuration a t C-16 shown in XXXVIIa (43). These conclusions concerning the structure and stereochemistry of akuammidine have been verified by the X-ray analysis of akuammidine methiodide (21). This investigation has also established the configuration of the C-19 to C-20 double bond, which is as shown in XXXVIIa. The fragmentation pattern observed in the mass spectrum of akuammidine has also been discussed in relation to the structure XXXVIIa (44,45).

VI. Pseudoakuammigine Pseudoakuammigine, CzzHzsNzO3, crystallizes from ethanol as colorless, square plates, mp 165", [.]go -35" (EtOH), pK, 7.35, and contains methoxyl, methylimino, and C-methyl groups (5, 23, 46). Its UV-spectrum indicates that the molecule is based on an indoline nucleus (23, 46, 47); the IR-spectrum has bands corresponding to a 1,2-disubstituted benzene derivative (754 cm-') and a carbonyl group (1736 cm-l), but there is no absorption corresponding to imino or hydroxyl groups. I n many respects, the behavior of pseudoakuammigine is anomalous, on the assumption that it is an indoline derivative. Thus, its basic strength is closer t o that of strychnine (pK, 7.6) than to that of strychnidine (pK, 8.29). Surprisingly, it does not give all the color reactions exhibited by ajmaline or strychnidine. Nitric acid gives only a brownish-yellow color, instead of the customary deep red; the ferric chloride color is feeble, and appears only on warming; and the base does not couple with diazotized sulfanilic acid, except in dilute, buffered solution, and then very slowly. On the other hand, nitrozation occurs readily, to give a normal green p-nitroso derivative, which can be reduced to a p-aminopseudoakuammigine; this gives the deep red color with nitric acid characteristic of strychnidine (46).

7.

ALKALOIDS OF

Picralima nitida

135

The carbonyl function of pseudoakuammigine is contained in a methyl ester group, since reduction with lithium aluminum hydride gives a primary alcohol, pseudoakuammigol, which behaves in all respects as a typical indoline derivative, except that its ferric chloride reaction is still sluggish. It is noteworthy that the basicity of pseudoakuammigol (pK, 8.22) is almost the same as that of strychnidine. The ester group of the alkaloid is probably attached to quaternary carbon, since it can be recovered almost quantitatively after being refluxed for 7 hours with 10% ethanolic potassium hydroxide (46). I n the earlier work of Robinson and Thomas (46), it was found that neither pseudoakuammigine nor pseudoakuammigol could be smoothly hydrogenated, and the presence of a double bond could therefore not be rigorously established. A similar situation was encountered with the closely related alkaloid akuammine ; however, some (inconclusive) evidence for hydrogenation of this base was obtained, so it was assumed that the molecule contained a double bond. Owing to the close similarity of the IR-spectra of akuammine and pseudoakuammigine (except for the hydroxyl band in the spectrum of akuammine), the presence of a double bond in pseudoakuammigine was also assumed, and it was therefore tentatively formulated as a deoxyakuammine, i.e., XLVII, R = H (46). As expected from this formulation, pseudoakuarnmigine also yields 3-ethylpyridine on distillation with zinc dust (31). The unexpected deactivation of the aromatic position para t o N,, as testified by the anomalous color reactions and the reduced basicity of pseudoakuammigine, were explained by assuming that the proximity of the ester group to the indoline nitrogen atom resulted in deactivation of the potentially basic center across space by the carbonyl group (46). COOMe

Me Me00

XLVIII

XLVII R

=

H ; Pseudoakuammigine R = OH;Akuammine

The recent experimental work of Janot and Smith, and their respective collaborators, has demonstrated the inadequacy of structure XLVII (R = H) to explain the behavior of pseudoakuammigine; consequently, on the basis of their results, the French workers have proposed the constitution XLVIII (R = H ) (48). The reason for the apparently

136

J. E. SAXTON

anomalous color reactions of pseudoakuammigine becomes evident from a comparison of its UV-spectrum in neutral and strongly acid ( 5 N ) solution; in neutral solution, the spectrum is that of a n indoline base, but in acid solution the spectrum is that of the 3H-indolium ion. This behavior, together with the weaker basicity of pseudoakuammigine than is normal for an indoline base, are best explained by the participation of N, in a carbinolamine ether function. Since pseudoakuammigine can be recovered on basification of its solution in 5 N acid, the carbinolamine ether oxygen must be contained in a readily reformed ring, which is therefore likely to be five- or six-membered (48,49). In accordance with the structure XLVIII (R = H ) , reduction of pseudoakuammigine by zinc amalgam and sulfuric acid, or by Raney nickel in boiling dioxan, results in the addition of two hydrogen atoms, with liberation of an acetylatable hydroxyl group, which is strongly hydrogen bonded, presumably to the carbomethoxy group. Unexpectedly, however, the product, a dihydropseudoakuammigine, exhibits UV-spectra in neutral and in acid solution which are characteristic of the geminal dianiino system present in eserine-andfolicanthine. Accordingly, dihydropseudoakuammigine is formulated as LI, and is presumably formed from the pseudoakuammigine ammonium ion (XLIX) by hydrogenolysis of the C-3 to Nb bond, followed by closure of the fivemembered ring in the intermediate (L). Reduction of L I with lithium O~, aluminum hydride gives a dihydropseudoakuammiginol, C Z I H Z ~ N Zby conversion of the carbomethoxy group into a primary alcohol function. The product does not have the spectrographic properties of a base of the eserine type, but instead shows a small bathochromic shift in acid solution similar to, but less pronounced than, that shown by pseudoakuammigine.

XLIX

;i.-$ ._

CHMe

L

7.

ALKALOIDS OF

Picralima nitida

137

The postulated presence in dihydropseudoakuammigine (LI) of a /3-hydroxy-a,a-disubstituted propionic ester grouping, analogous to that in echitamine, finds confirmation in its reaction with potassium tbutoxide in benzene, which results in retroaldol loss of formaldehyde, with formation of dehydroxymethyldihydropseudoakuammigine(LII). The parent alkaloid and dihydropseudoakuammiginol, which do not contain the P-hydroxypropionic ester grouping, do not eliminate formaldehyde under comparable conditions. Reduction of dehydroxymethyldihydropseudoakuammigine with lithium aluminum hydride gives the

aT&

0 7 %

Me

Me

H

CHMc

‘CHM~

LIII

LII

q LIV

Et

related alcohol, which can be converted into the corresponding aldehyde by a modified Oppenauer oxidation, and thence by Wolff-Kishner reduction into dehydroxymethyldihydropseudoakuammiginane(LIII). All the bases in this series, from dihydropseudoakuammigine (LI) to the final product (LIII), exhibit UV-spectra consistent with their formulation as bases of the eserine type. The quantitative hydrogenation of LIII, and the results of Kuhn-Roth oxidation, confirm the presence in these transformation products, and therefore in pseudoakuammigine itself, of an ethylidene grouping (48). The preceding degradations and transformations of pseudoakuammigine are in consonance with, but do not prove, the structures proposed; it is therefore noteworthy that the mass spectrum of LIV, the hydrogenation product of LIII, is in complete accord with this geininal diamino structure (48). I n an independent investigation Smith and his collaborators have provided confirmation of the presence in pseudoakuammigine of some of the features contained in the structure XLVIII (R = H), and have also

138

J. E. SAXTON

contributed a fascinating study of the rearrangement of pseudoakuammigine under the influence of strong acid (49, 50, 51). The presence of an ethylidene grouping in pseudoakuammigine is confirmed by the production of acetaldehyde on ozonolysis, and by the NMR-spectrum; the N,-methyl group is also identifiable in the NMR-spectra of pseudoalcuammigine and all its derivatives (49). The presence of an N,-methyl tetrahydro-P-carboline system in pseudoakuammigine is proved by the mass spectrum of LIVa, the lithium aluminum hydride reduction product of Nb-methylpseudoakuammigine dihydromethine, since the second most abundant ion in the spectrum corresponds to the fragment LV (50).

Me

LV

LVI

The lithium aluminum hydride reduction of pseudoakuammigine furnishes an indoline base containing two hydroxyl groups; the carbinolamine ether function and the ester group have therefore been reduced, and the product, pseudoakuammigol (LVI; R = H), earlier formulated as C21H~GN202,is in fact C21H28N202. The two hydroxyl groups in this molecule are situated on alternate carbon atoms since pseudoakuammigol is unaffected by periodic acid but gives a crystalline isopropylidene derivative. Neopseudoakuammigol, the base in which the ester function in pseudoakuammigine is replaced by a primary alcohol group, can be obtained by reduction of the alkaloid with sodium and ethanol (49). The structure LVI (R = H) for pseudoakuammigol receives convincing confirmation from the fragmentation patterns observed in the mass spectra of pseudoakuammigol and its deuterated derivatives LVIa and LVIb ; the structure (XLVIII ; R = H ) for pseudoakuammigine itself can therefore be regarded as firmly established. Since these mass spectra,

7.

ALKALOIDS OF

Picralima nitida

139

when compared with the mass spectra of the alkaloid picraline and its lithium aluminum hydride (or deuteride) reduction products, also provide valuable evidence relating to the structures of the picraline series of bases, they will be discussed in detail later (see Section V I I I ) (50, 51, 51a, 5 3 , 54). HOHzC , ,CX20H

LVIa; X = D , Y = H LVIh; X = Y = D

The close structural relationship between akuammine and pseudoakuammigine, first suggested by Robinson and Thomas, has been confirmed by reduction of both 0-tosylakuammine and pseudoakuammigine by Raney nickel and hydrogen to dihydroisopseudoakuammigine, C22H~gN203,mp 264"-267". This base is almost certainly identical with Janot's dihydropseudoakuammigine (LI),mp 252", since it contains the same functional groups and exhibits the characteristic UV-spectra in neutral and acid solution of the eserine-like bases (49). The most important aspect of the chemistry of pseudoakuammigine which does not find a ready explanation in terms of the structure XLVIII (R = H) concerns the degradation to apopseudoakuammigine and the further transformations of this product. Apopseudoakuammigine itself is formed by prolonged treatment of pseudoakuammigine with hot mineral acid, and differs from its progenitor by the elements of a methylene group. Its IR-spectrum shows the presence of a y-lactone in which the carbonyl frequency (1746 cni-1) is lowered by hydrogen bonding, presumably with a hydroxyl group, since the 0-acetate exhibits normal carbonyl frequencies, a t 1773 and 1747 cm-1. Since the carbinolamine ether oxygen is situated /3 to the ester grouping, it cannot be involved in lactone formation, and must appear in apopseudoakuammigine as the free hydroxyl group ; apopseudoakuammigine therefore probably contains the part-structure LVIIa. The presence of this chromophore is confirmed by the UV-spectrum of apopseudoakuammigine, which is very similar to that of pseudoakuammigine, except that it is displaced slightly toward longer wavelengths ; however, the spectra of the corresponding ammonium ions (i.e. in concentrated acid) are identical. The formation of LVIIa from the proposed structure of pseudoakuammigine (XLVIII; R = H ) is clearly not a simple reaction, owing

140

J. E. SAXTON

to the steric impossibility of forming a lactone ring a t C-2 in this ring system. This, and other important features of the molecule, suggest very forcibly that a profound rearrangement occurs during the conversion of pseudoakuammigine into the apo base. Thus, apopseudoakuammigine gives no trace of pseudoakuammigol or stereoisomeric indoline base on reduction with lithium aluminum hydride ; instead, a mixture of indole derivatives is obtained. A further difficulty concerns the basicity of apopseudoakuammigine (pK, 5.6), which is a significantly weaker base than pseudoakuammigine. This is ascribed to the proximity of the carbon of the carbonyl group to N,, which results in a base-weakening field effect.

\. ./ 4

LVII~

LVIIb

LVIIId

LVIII~

LVIIIC

LVIIIb

7.

ALKALOIDS OF

Picralima nitida

141

I n contrast to pseudoakuammigine, which is not affected by alkali even under fairly vigorous conditions, apopseudoakuammigine is degraded a t room temperature, with the appearance of indole UVabsorption. With hot alkali, apopseudoakuammigine is converted in good yield into an indole base, C20H24N20 ; this change corresponds to the loss of carbon dioxide. The single oxygen atom in this product is contained in an aldehyde group (IR-absorption at 1726 and 2705 cm-I), since it gives an oxime which, on dehydration with acetic anhydride, gives a product exhibiting nitrile IR-absorption. These results can be explained by the reverse Mannich decomposition of a base (LVIIb) in which Nb is separated from the beta position of an indoline system by only one carbon atom. The product (LVIIIa) can then undergo an internal hydride shift with formation of a ,B-aldehydo acid (LVIIIb); decarboxylation to the indole aldehyde base (LVIIIc) then follows naturally. A related, but simpler, reaction, is the potassium borohydride reduction of apopseudoakuammigine to an indolic amino acid, which may be formulated as LVIIId. The unshared electrons on Nbare evidently involved in this reaction, since apopseudoakuammigine methiodide gives an indoline betaine, and not an indole derivative, on reduction with borohydride (49). CHz 'YHM~

LIXe

LIXC

The presence of an allylamine system in pseudoakuammigine, postulated in XLVIII, is confirmed by the platinum-catalyzed hydrogenolysis of pseudoakuammigine methiodide, which affords a dihydromethine base: C23H30N203, by fission of the C-21 to Nb bond; the apo base methiodide similarly yields a dihydromethine. Both bases still contain the ethylidene group, and contain one additional C-methyl group.

142

J . E. SAXTON

Hence, the part-structure LVIIb for apopseudoakuammigine may be expanded to LIXa. Whereas pseudoakuammigine dihydromethine, like the alkaloid itself, is unaffected by potassium borohydride, apopseudoakuammigine dihydromethine is reduced to an indolic amino acid. The latter was not isolated pure, but was converted, by reaction with methanolic hydrogen chloride, into a crystalline indolic hydroxy-y-lactone, which exhibits carbonyl absorption a t 1761 and 1726 cm-1. The formation of such a lactone can only occur if the double bond is involved; this, therefore, implies that the double bond must be situated in the P,y-or y,a-position with respect to the carbonyl group. The part-structure LIXa for apopsuedoakuammigine can therefore be expanded to the alternatives LIXb and LIXc. Since the NMR-spectrum of the lactone reveals that it contains a methyl and an ethyl group attached to quaternary carbon, the C-20 carbon atom must be the y-carbon atom with respect to the carbonyl function in both apopseudoakuammigine and the indolic y-lactone. The latter must therefore contain the part-structure LIXd and apopseudoakuammigine the part-structure LIXc (49, 50). One of the principal characteristics of apopseudoakuammigine is the ease with which it can be degraded to indolic products ; in this respect its behavior can be contrasted sharply with that of pseudoakuammigine. The part-structure LIXc explains this feature very neatly, and it is thus apparent that a complex rearrangement must have occurred during its formation from pseudoakuammigine. Although the complete structure of apopseudoakuammigine is not known with certainty, theconstitution LXII (R = H) has been tentatively proposed, and is consistent with all the experimental data a t present available (50). This structure can be obtained by acid treatment of pseudoakuammigine (XLVIII) by protonation to the corresponding ammonium ion (XLIX), followed by a double migration of the substituents at positions 2 and 7. Migration of the a substituent (C-6) at C-7 to C-2 yields the spirocyclic carbonium ion (LX); this is now capable of conversion by migration of the ,l3 substituent (C-3) at C-2 into an ammonium ion (LXI) with stereochemical inversion at C-7. Formation of the lactone ring at C-2 then yields apopseudoakuammigine (LXII; R =H). On the basis of this structure, most of the reactions of apopseudoakuammigine can be satisfactorily explained. The reduction with potassium borohydride can be accounted for, as postulated by Joule and Smith, by the reverse Mannich decomposition of LXII, followed by reduction of the intermediate (LXIII); the product of this reaction would then be the indolic amino acid LXIV, a structure which is consistent with the observation that this amino acid is extremely resistant to esteri-

7.

ALKALOIDS OF

Picralima nitida

143

fication. The mechanism by which LXIV is formed accords with the reported failure of apopseudoakuammigine methiodide to give an analogous product with potassium borohydride.

Pseudoakuammigine

COOMe

Me

Me

LX

LXI

l4

19

18

N ’

LXII R = H ; Apopseudoakuammigine R = OMe; 0-Methylapoakuammine

The degradation with alkali, following the reaction path outlined earlier (part-structures LVIIb --f LVIITa+ b +c) would result in the formation of the aldehyde LXVI. The production of this compound can be explained very convincingly by reverse Mannich decomposition of apopseudoakuammigine to the intermediate zwitter ion LXIII. The alternative conformation of this molecule is one (LXV) in which the (2-17 hydrogen is situated in close proximity to the trigonal C-3; hydride transfer, as postulated by Joule and Smith, would then be expected to

J. E. SAXTON

144

Mc

LXIII

LXV

& CH,OH

YHO

&HMe

Me

Me LXIV

LXVI

HOCHz

& C. Me

MeLXVIII

m

CHzCHO

CH=CHz Me LXVII

HOCHa HOCHz

Me LXTX

Mt?

LXX

proceed readily, to give the aldehyde (LXVI) (50, 51). This structure is consistent with the properties of this aldehyde, which affords, on zinc dust distillation, a mixture of 3-ethylpyridine and N-methylcarbazole ;

7.

ALKALOIDS OF

Picralima nitida

145

the latter is considered to be formed by decomposition of LXVI into 3-ethylpyridine and (possibly) the aldehyde LXVII, which then undergoes cyclization and aromatization (compare the thermal decomposition of akuammicine to XXXV, via the unsaturated ester XXXIV) (50, 51). Finally, the indolic y-lactone, obtained from apopseudoakuammigine dihydromethine (LXVIII) by borohydride reduction followed by treatment with methanolic hydrogen chloride, is probably LXX ; the intermediate noncrystalline acid is presumably LXIX.

VII. Akuammine Akuammine, C22H26N204, mp 254"-259" (dec.), pK, 7.5, [a]= -66.7' (ethanol), is the principal base in Picralima nitida, and occurs to the extent of 0.56% in the seeds. The molecule contains methoxyl, methylimino, C-methyl, and hydroxyl groups, but no aldehydic or ketonic carbonyl groups (1, 5, 23); it is a tertiary base, since its methiodide is a true quaternary salt (1).No evidence is available for the presence of an imino group ; benzoylation and acetylation produce 0-acyl derivatives (5). Akuanimine is soluble in alkali, which converts it into "akuammine hydrate," an alkali-soluble, microcrystalline substance which does not melt below 310", and brown, amorphous by-products, similarly soluble in alkali (1). Akuammine is also converted into intractable, resinous materials by boiling dilute hydrochloric acid (5))and it has even been known to decompose during attempted recrystallization from boiling methanol (23). The color reaction of akuammine with nitric acid is blood-red, and treatment with nitrous acid yields a scarlet, crystalline product, which does not give Liebermann's nitrosamine reaction and which is very probably a nitroakuammine hydrochloride (5). The UV-spectrum of the alkaloid is typical of indoline bases (23, 47)) which, combined with the evidence from color reactions, was initially interpreted as indicating that akuammine is a methoxylated indoline base. However, a detailed examination of its coIor reactions reveals that akuammine is a 5hydroxydihydroindole derivative, since it behaves in an exactly analogous manner to p-methylaminophenol ; its sensitivity to alkali is therefore readily understandable. The position of the hydroxyl group is confirmed by the IR-spectrum (band at 811 cm-', characteristic of 1,2,4-trisubstituted benzenes) (23), and by comparison of its UVspectrum with that of 6-methoxy-9,11-dimethylhexahydrocarbazole (LXXI). The spectra of the methoxyl isomers of LXXI are significantly different (52).

146

J. E. SAXTON

Zinc dust distillation of the amorphous material obtained when akuamrnine decomposed in methanol solution gave an indolaceous substance (probably skatole), a volatile base identified as 3-ethylpyridine (23), and carbazole (31). Hydrogenation studies were inconclusive, and although evidence for the formation of a dihydro derivative was obtained, this was not fully characterized (23).

MeooTl ""&p I

Me

'

N ' Me

Me

1

-O/

\-Me

MeOOC LXXI

LXXII

Of the three remaining oxygen atoms, two are present as an ester group (IR-band at 1736 cm-I), and the third, inert, oxygen is probably contained in an ether linkage. Taking cognizance of the fact that the oxidation color reactions and behavior of akuammine are characteristic of substances of type LXXI, and quite different from those of 2,3disubstituted indolines which readily suffer dehydrogenation to indole derivatives, akuammine must belong to the 8-series rather than the a-series of indole alkaloids. The constitution LXXII was proposed to explain all the available evidence; the double bond was provisionally located in the position adjacent to the methyl group, to account for the positive iodoform reaction (23). However, akuammine is not an enol ether, since its IR-spectrum does not exhibit an absorption band a t 1650 cm-1; the (highly strained) formula XLVII (R = OH) was therefore preferred (46). Much of the evidence obtained recently that has a bearing on the structure of akuammine has been derived from the degradation of the closely related alkaloid pseudoakuammigine, and has been discussed above. Robinson and Thomas (46) noted the very close similarity of the IR-spectra of akuammine and pseudoakuammigine, except for the hydroxyl band in the spectrum of akuammine ;they therefore formulated pseudoakuammigine as a deoxyakuammine. The correctness of this deduction has recently been demonstrated by Joule and Smith (49)' who converted 0-tosylakuammine and pseudoakuammigine into the same derivative, dihydroisopseudoakuammigine (LI ; Janot's dihydropseudoakuammigine), by boiling with Raney nickel in ethanol solution in an atmosphere of hydrogen. This close relationship is also confirmed by the similarity of the fragmentation patterns observed in the mass spectra of derivatives of 0-methylakuammine and pseudoakuammigine

7. ALKALOIDS

OF

Picralima nitida

147

(vide infra) (49). The structure proposed by Janot and his collaborators (48)for pseudoakuammigine is XLVIII (R = H) ; accordingly, akuammine is XLVIII (R = OH). The limited amount of experimental work that has been carried out recently on akuammine shows that its behavior is exactly analogous to that of pseudoakuammigine (49). Thus, the UV-spectrum of 0-methylakuammine in neutral and in dilute acid solution is typical of an indoline derivative ; a marked bathochromic shift is observed in concentrated hydrochloric acid, and the spectrum is now characteristic of the 3-Hindolium ion. The recovery of 0-methylakuammine from this solution shows that no rearrangement of the molecule has occurred, and suggests further that the readily reformed carbinolamine ether ring is five- or six-membered. I n accordance with the structure XLVIII (R = OMe), 0-methylakuammine is reduced by lithium aluminum hydride to 0-methylakuamminol (LVI; R = OMe), an indoline base which contains two alcoholic hydroxyl groups, and which, like pseudoakuammigol (LVI ; R = H), gives rise to an (amorphous) isopropylidene derivative. Although 0-methylakuammine (XLVIII; R = OMe) is stable to mineral acid under mild conditions, prolonged treatment with 3 N hydrochloric acid at 80" gives 0-methylapoakuammine by loss of the elements of a methylene group. The IR-spectrum of this product parallels that of apopseudoakuammigine in that it indicates the presence of a y-lactone grouping in which the carbonyl frequency (1756 cm-1) is lowered by hydrogen bonding with a hydroxyl group, since the carbonyl bands appear at the expected frequencies in the corresponding 0-acetate. By analogy with the structure proposed for apopseudoakuammigine (LXII ; R = H), 0-methylapoakuammine may be provisionally formulated as LXII, R = OMe (49, 50).

VIII. Picraline Picraline was first isolated by Thomas (15a) from the crude alkaloidal fraction of Picralima seeds by chromatography on alumina. Very recently its isolation has been reported by other investigators, and the molecule has been thoroughly characterized (15, 51a, 53, 54). Picraline, C~~HZ~N mp Z 180"-182", O~, [a]=-124" (MeOH),pK, 5.65 (50% aqueous ethanol), contains one methoxyl group, one active hydrogen, and two C-methyl groups. Its UV-spectrum exhibits maxima a t 231 and 289 mp and is unaffected by the addition of dilute acid or alkali ; in concentrated perchloric acid, however, a reversible change occurs, with the appearance

148

J . E. SAXTON

of maxima a t 241, 246, and 310 mp, characteristic of the protonated indolenine system. This behavior is clearly reminiscent of the carbinolamine ether grouping present in pseudoakuammigine. The IR-spectrum of picraline contains bands due to NH or OH (3400 cm-I), ester (1724 cm-I), and acetoxyl (1695 and 1250 cm-') groups, a double bond (1613 cm-l), and an ortho-disubstituted benzene nucleus ( - 750 cm-1); the ester group can be identified from the NMR-spectrum as a carbomethoxyl group. The presence of an acetoxyl group is also supported by the NMRspectrum, and confirmed by the formation of deacetylpicraline, C21H24N204, by acid hydrolysis of picraline. It is noteworthy that deacetylpicraline is also a constituent of Picralima seeds. The second C-methyl group and the double bond in picraline are present in an ethylidene group, since ozonolysis yields acetaldehyde ; the NMRspectrum also indicates clearly the presence of this group (15,51a, 53,54). Both picraline and deacetylpicraline react with aqueous alcoholic potassium hydroxide at 80" to give picrinine, CzoHzzNz03, which is formed by loss of the elements of formaldehyde from deacetylpicraline. Since picrinine still contains a carbomethoxyl group (IR- and NMRspectra) this reaction is very probably the retroaldol cleavage of formaldehyde from a 8-hydroxypropionic ester group, such as occurs in echitamine and akuammidine ; however, in the present instance, this reaction proceeds with unexpected ease. The third oxygen atom in picrinine must be present in an ether function, since the IR-spectrum affords no evidence for the presence of either a hydroxyl or a second carbonyl group. Since the UV-spectra of picrinine in neutral and in concentrated perchloric acid solution are very similar to those of picraline, this ether oxygen must be attached to the indoline a-position. The presence of the ,&hydroxypropionic ester unit in deacetylpicraline is established by oxidation with chromic acid in acetone, which yields an aldehyde base, picralinal, CzlHzzNz04;the latter is readily deformylated by short treatment with methanolic potassium hydroxide, which affords picrinine in quantitative yield. Reduction of picralinal with sodium borohydride regenerates deacetylpicraline. Vigorous treatment of deacetylpicraline with sodium borohydride gives a noncrystalline indoline base, which exhibits the UV-absorption of an anilinium ion in concentrated perchloric acid ; hence, the N,-carbinolamine ether function must have suffered reduction. Since acetylation of the noncrystalline base gives a product which exhibits acylaniline UV-absorption, picraline and its derivatives must contain an N,H group (53, 54).

'

1 Deacetylpicraline (burnamine) also occurs in the bark of Hunteria eburnea Pichon (56).

7.

ALKALOIDS OF

Picralima nitida

149

Reduction of picraline or deacetylpicraline with lithium aluminum hydride gives picralinol, C2,,H2,N202 (15, 51a), which, as expected from the reduction of a p-hydroxypropionic ester residue, behaves as a 1,3-diol since i t gives rise t o an isopropylidene derivative (53, 54). The UV-spectrum of picralinol is that of an indoline base, changed to anilinium ion in concentrated acid, which indicates that the Na-carbinolamine ether function has been reduced. The total loss of this oxygen atom suggests that its other point of attachment in picraline is also a carbon atom adjacent t o nitrogen, which can only be Nb;picraline must therefore be a biscarbinolamine ether. This deduction is supported by the comparatively weak basicity of picraline and the increased basicity (pK, 8.15) of picralinol (53, 54). The structural features present in picraline are, therefore, given by the part-structure LXXIII. Assuming a close structural relationship with pseudoakuammigine (XLVIII ; R = H), which is evident from the similarity of the IR-spectra of picralinol and pseudoakuammigol (LVI; R =H), this can be expanded to the complete structure (LXXIV) for picraline. The carbinolamine ether oxygen atom is attached to C-2 from the UV-evidence; the other point of attachment, adjacent to Nb, cannot be C-21, since this is sterically prohibited. Attachment at C-5 is favored, since the NMR-spectra of picraline, deacetylpicraline, and picrinine exhibit a signal (51, 53, 54) at -5.20 T , characteristic of a carbinolamine hydrogen : >N-CH-0-.

HOCH2,

OTJo>dN<

,COOMe C

/ \

H

LXXIII 19

I8

CHMe

a

LXXIV Pioralme

b

+ cn

TABLE I

0

MASSSPECTRA OF PICRALINOL AND RELATED COMPOUNDS;M/e VALUESOF PRINCIPAL PEAKS MQ- CHzOH

LXXIX - CHzO

or

Compound

Picralinol (LXXV)

M"

MQ- O H MQ-CDzOH LXXVIII" LXXIX

LXXIX - HzO

or

- CDzO LXXX" LXXXI' LXXXII'

326

309

295

25 1

196

178

166

144

143

130

330

313

297 299

253

199

181

167 169

146

145

131

Tetradeuteropicralinol

(LXXVII)

M

+ s G ra

Pseudoakuammigol (LVI; R = H)

340

323

309

265

196

178

166

158

157

144

Trideuteropseudoakuammigol (LVIa)

343

326

310 312

266

198

180

166 168

59

158

145

Tetradeuteropseudoakuammigol (LVIb)

344

327

311 313

267

199

181

167 169

60

159

145

0-Methylakuamminol (LVI; R = OMe)

370

353

295

196

178

166

188

a

174

Picralinol series: R = R' = H ; pseudoakuammigol series: R = Me,R' = H; 0-methylakuamminol series: R = Me, R' = OMe.

w

7.

ALKALOIDS OF

Picralima nitida

151

On the basis of this structure (LXXIV) for picraline, picralinol must be the diol LXXV. The latter differs from the proposed structure for pseudoakuammigol (LVI; R = H) only in that it possesses a hydrogen atom attached to N, instead of a methyl group ; accordingly, N,-methylation of picralinol should yield pseudoakuammigol. This methylation has been realized by the lithium aluminum hydride reduction of the triformyl derivative (LXXVI) of picralinol, which affords pseudoakuammigol directly (51a).

LXXV

LXXVII

LXXVI

The structures proposed for picralinol (LXXV) and pseudoakuammigol (LVI ; R = H),and also for 0-methylakuamminol (LVI; R = OMe), receive impressive support from a study of their mass spectra, together with those of tetradeuteropicralinol [the lithium aluminum deuteride reduction product (LXXVII) of picraline], trideuteropseudoakuammigol (LVIa), and tetradeuteropseudoakuammigol (LVIb). The fragmentation processes exhibited by these bases on electron impact are summarized in Table I, which also indicates possible structural assignments for the various fragments (50, 51, 51a, 53, 54). The deuterated positions in the fragments derived from LVIa, LVIb, and LXXVII are indicated by asterisks in the formulas LXXVIII-LXXXII. These data leave little room for doubt that picralinol has the constitution LXXV; the structure for picraline, however, is not so well established. Smith’s view that picraline is best represented by LXXIV is convincingly supported by the UV-evidence, but it is not accepted by other workers. The typically indoline spectrum of picraline undergoes a bathochromic shift in concentrated acid, and deacetylpicraline can be recovered on basification. This observation would seem to exclude the

152

J. E. SAXTON

possibility of the presence of a methoxyl group at C-2. Further, it may be deduced that the oxygen atom attached to C-2 must be so situated in the ring system that reformation of the C-2 to oxygen linkage is a very facile process. The structure LXXIV accords with this behavior, and is also consistent with the NMR-spectrum (51, 5 3 , 54).

LXXVIII

LXXIX

LXXXI

LXXX

LXXXII

A different view has been expressed by Janot, Djerassi, and their collaborators (51a),who reject the structure LXXIV on the grounds that an ether bridge between C-2 and C-5 is sterically impossible. However, the diagram LXXIVb was drawn from a Dreiding model of this structure, which would appear, as far as it is pessible to judge from the use of models, not to be severely strained. In lieu of LXXIV, Janot, Djerassi, and their co-workers consider the possibilities for picraline based on the structure LXXV for picralinol. The mass spectrographic evidence indicates that in picraline oxygenated substituents are attached to C-2, C-5, (2-17, and C-22, i.e., picraline has the partial formula LXXXIII. The various possibilities for picraline are consequently given by the structures LXXXIV-LXXXVI ; a fourth structure containing a ylactone function can be ignored, since it is eliminated by the IR-evidence. The first of these structures (LXXXIV) was eliminated by Janot and co-workers, who did not observe a bathochroniic shift of the UVspectrum of picraline in strong acid; a structure analogous to that of pseudoakuammigine thus appeared to be very improbable. The constitution LXXXV was similarly rejected since the UV-spectrum of deacetylpicraline was also reported not to exhibit a bathochromic shift

7.

ALKALOIDS OF

Picralima nitida

153

in acid solution; evidently, the presence of a hydroxyl group at C-2 would be expected to result in the ready formation of an indolenine derivative in acid solution. The remaining structure (LXXXVI) is supported, as the others are invalidated, by the mass spectrum of picraline, which shows a peak at M+ - 73, owing to loss of the acetoxymethyl group. In contrast, the mass spectrum of deacetylpicraline does not contain this peak, but instead exhibits one at M+ - 31, resulting from loss of the hydroxymethyl group. Hence, it was concluded that pieraline has the structure LXXXVI ; the methoxyl group at C-2 was tentatively assigned the cc-configuration,since this resulted in the least steric strain (51a).

* LXXXIII

H

LXXXIV

c~

0

H

co I kH3 LXXXV

CHMe

bfi’ Me

H

%HMe

LXXXVI

The two structures which have been firmly proposed for picraline are, therefore, LXXIV and LXXXVI. Both proposals account satisfactorily for the chemical reactions discussed above, and are consistent with its IR- and NMR-spectra; hence, a firm distinction between the two structures cannot be made on the basis of these data. The correction of the contradictory reports concerning the UV-absorption of picraline in acid solution would not materially assist the arguments. Both LXXIV and LXXXVI should exhibit indoline absorption in neutral and dilute acid solution, and both would be expected to show 3H-indolium ion absorption in the presence of concentrated acid. However, the recovery of deacetylpicraline from the solution of picraline or deacetylpicraline in

154

J. E. SAXTON

concentrated acid is a vital observation which is only consistent with the structure LXXIV, and effectively excludes the structure LXXXVI (55). Further evidence in support of LXXIV is provided by the mass spectrum of picraline, which, in addition to the peak at M+ - 73 (M/e 337), owing to loss of the acetoxymethyl group, contains two further prominent peaks, at M/e 351 (Mf - 59) and M/e 239 (M+ - 171). These are due, respectively, to loss of (a)the acetoxyl or carbomethoxyl group, and (b) C-16 and its substituent groupings (CH3COOCH,-C-COOMe) together with CO derived from C-5. There is no peak owing to loss of COz, and the peak owing to loss of the methoxyl group, at M+ - 31, is only very weak ;these characteristics of the mass spectrum would indeed be surprising if LXXXVI were correct. That the peak at M/e 351 is the result of loss of a carbomethoxyl group, and not an acetoxyl group, is evident from the mass spectrum of picrinine (LXXIV, with H in place of CHsCOOCHzat C-l6), which also exhibits a peak at M+ - 59, owing to loss of the carbomethoxyl group. Here the loss of an acetoxyl group does not come into consideration (55). The sodium borohydride reduction of picraline furnishes a (noncrystalline) base which exhibits typically indoline UV-absorption, and which must therefore arise by reduction of the N,-carbinolamine ether linkage. Since this product still contains amethoxyl group, observed in the NMR-spectrum as a singlet at 6.287, picraline does not possess a methoxyl group at C-2, and can therefore not have the structure LXXXVI. All these data, however, are consistent with the structure LXXIV (55) In addition to the formation of deacetylpicraline, the acid hydrolysis of picraline yields a yellow base, flavopicraline, CzoH,oNzOs,which exhibits UV-absorption at 245 and 390 m p , with a shoulder at 305-320 mp. In acid solution the long-wavelength maximum is shifted to 438 mp. The IR-spectrum of flavopicraline exhibits carbonyl absorption at 1761 cm-l, characteristic of a y-lactone, while the NMR spectrum discloses the presence of ethyl and CH=CH-N groups; there is no evidence for the presence of an ethylidene group. Sodium borohydride reduction affords a colorless indole base (UV-spectrum) which still contains the presumed y-lactone grouping (IR-absorption at 1760 cm-l) (15, 51, 53). The structure proposed for flavopicraline (LXXXVII) can be derived from deacetylpicraline by a mechanism which is formally analogous to that postulated for the conversion of pseudoakuammigine (XLVIII ; R = H ) into apopseudoakuammigine (LXII ; R = H) ; the closure of the lactone ring at C-20 implies that the stereochemistry at C-16 in picraline is the same as that in pseudoakuammigine (55) Finally the colorless indole base obtained on sodium borohydride

7.

ALKALOIDS OF

Picralima nitida

155

reduction of flavopicraline may provisionally be formulated as the indolic y-lactone (LXXXVIII). 19.18

Et

1

2 0 b '

LXXXVII

LXXXVIII

Flavopicreline

IX. Akuammiline Akuammiline, CzzH24Nz04, forms translucent prisms from ethanol, mp 160°, [u]$'" +47. 9 (EtOH), and contains methoxyl and two Cmethyl groups (5, 23). Its UV-spectrum is similar to that of 3,3-dimethylindolenine, but shows a small shift to longer wavelengths (46). The IR-absorption discloses the presence of hydroxyl or imino groups (3450 ern-I), an unconjugated ester (1736 cm-l), and possibly an orthodisubstituted benzene nucleus. The base gives a characteristic crimson Otto reaction, and a possible relationship to akuammigine has been suggested (23). (See note added in proof, p. 157.)

X. Akuammenine Akuammenine, CzoHzzNz04, is the least abundant alkaloid of this group, and is contained in the seeds to the extent of only 0.0006~0.As yet it has only been obtained as its scarlet picrate, mp 225", and no information regarding its constitution is available beyond the fact that it contains a methoxyl group ( 5 ) . REFERENCES 1. 2. 3. 4. 5. 6.

T. A. Henry and T. M. Sharp, J. Ghem. SOC.p. 1950 (1927). Sir Robert Robinson and A. F. Thomas, J. Chem. Soe. p. 3479 (1954). E. Clinquart, Bull. Acad. Roy. &fed. Belg. [5]6,492 (1926); Chem. dbstr. 21,151 (1927). E. Clinquart, J . Phann. Belg. 9, 187 (1927); Chem. Abstr. 22, 136 (1928). T. A. Henry, J . Chem. SOC.p. 2759 (1932). J. A. Goodson, T. A. Henry, and J. W. S. MacFie, Biochem. J . 24, 874 (1930).

156

J. E. SAXTON

7. Raymond-Hamet, Compt. Rend. Acad. Sci. 211, 125 (1940). 8. Raymond-Hamet, Arch. Exptl. Pathol. Phurmakol. 199, 399 (1942); Chem. Abstr. 37, 5782 (1943). 9. Raymond-Kamet, Compt. Rend. SOC.Biol. 137, 404 (1943). 10. Raymond-Hamet, Compt. Rend. Soc. Biol. 138, 199 (1944). 11. Raymond-Hamet, Rev. Intern. Botan. Appl. Agr. Trop. 31, 465 (1951). 12. Raymond-Hamet, Compt. Rend. SOC.Biol. 148, 458 (1964). 13. Raymond-Hamet, Compt. Rend. Scad. Sci. 255, 1482 (1962). 14. Raymond-Hamet, Compt. Rend. Acad. Sci.221, 699 (1945). 15. L. Olivier, J. LQvy,J. Le Men, and M.-M Janot, Ann. Pharm. Franc. 20, 361 (1962). 18a. A. F. Thomas, D. Phil. Thesis, Oxford Univ. 1954; and Personal communication (1963). 16. M.-M. Janot and J. Le Men, Compt. Rend. Acad. Sci. 238, 2550 (1954). 17. M.-M. Janot and J . Le Men, Compt. Rend. Acad. Sci. 240, 909 (1955). 18. M.-M. Janot, J. Le Men, K. Aghoramurthy, and Sir Robert Robinson, Ezperientia 11, 343 (1955). 19. M.-M. Janot and J . Le Ken, Contpt. Rend. Acad. Sci. 243, 1789 (1956). 20. A. Chatterjee, C. R. Ghosal, N. Adityachaudhury, and S. Ghosal, C'hem. I n d . (London) p. 1034 (1961). 21. S. Silvers and A. Tulinslry, Tetrahedron Letters p. 339 (1962); Acta Cryst. 16, 579 (1963). 22. J. Gosset, J . Le Men, and M.-M. Janot, Ann. Pharm. Franc. 20, 448 (1962). 22a. M. F. Bartlett, B. Korzun, R. Sklar, A. F. Smith, and W. I. Taylor, J . Org. Chem. 28, 1445 (1963). 23. M. F. Millson, Sir Robert Robinson, and A. F. Thomas, Experienlia 9, 89 (1953). 24. Raymond-Hamet, Compt. Rend. Acad. Sci.233,560 (1951). 25. E. Wenkert and D. K. Roychaudhuri, J . Am. Chem. SOC.78, 6417 (1956). 26. E. Wenkert and D. K. Roychaudhuri, J . Am. Chem. Soc. 80,1613 (1958). 27. W. E. Rosen, Tetrahedron Letters p. 481 (1961). 28. E. Wenkert and D. K. Roychaudhuri, J. Am. Chem. SOC.79, 1519 (1957). 29. E. Wenkert, R. Wickberg, and C. L. Leicht, J. Am. Chem. Soc. 83, 5037 (1961). 30. Sir Robert Robinson and A. F. Thomas, J . Chem. SOC.p. 2049 (1955). 31. K. Aghoramurthy and Sir Robert Robinson, Tetrahedron 1, 172 (1957). 32. G. F. Smith and J. T. Wr6be1, J . Chem. Soc p. 792 (1960). 33. K. Bernauer, W. Arnold, C. Weissmann, H. Schmid, and P. Karrer, Helv. Chim. Acta 43, 717 (1960). 34. C. Weissmann, H. Schmid, and P. Karrer, Helv. Chim. Acta 44, 1877 (1961). 35. P. N. Edwards and G. F. Smith, J . Chem. SOC.p. 152 (1961). 36. J. LBvy, J . Le Men, and M.-M. Janot, Bull. SOC.Chim. France p. 979 (1960). 37. M.-&I.Janot, J. Le Men, A. Le Hir, J. LBvy, and F. Puisieux, Compt. Rend. Acad. Sci. 250, 4383 (1960). 37a. M.-M. Janot, Pure Appl. Chern. 6,635 (1963). 38. K. Bernauer, F. Berlage, W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 2293 (1958). 39. P. N. Edwards and G. F. Smith, J . Chem. SOC.p. 1458 (1961). 40. P. N. Edwards and G. F. Smith, Proc. Chem. SOC.p. 215 (1960). 41. Raymond-Hamet, Compt. Rend. Acad. Sci. 236, 319 (1953); Bull. Soc. PhuTm. Bordeaux 90, 178 (1952); Chem. Abstr. 48, 8794 (1954). 42. J. LBvy, J. Le Men, and M.-M. Janot, Compt. Rend. Acad. Sci.253, 131 (1961). 43. M.-M. Janot, J. Le Men, J. Gosset, and J. LBvy, Bull. SOC.Chim.France p. 1079 (1962).

7.

ALKALOIDS OF

Picralima nitida

157

44. L. D. Antonaccio, N. a. Pereira, B. Gilbert, H. Vorbrueggen, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi,J. Am. Chem. Soc. 84, 2161 (1962). 45. E. Clayton, R. I. Reed, and J. M. Wilson, Tetrahedron 18, 1449 (1962); M. Ohashi, H. Budzikiewicz, J. M. Wilson, C. Djerassi, J. Levy, J . Gosset, J. Le Men, and M.-M. Janot, Tetrahedron 19, 2241 (1963). 46. Sir Robert Robinson and A. F. Thomas, J . Chem. SOC.p. 3522 (1954). 47. Raymond-Hamet, Compt. Rend. Acad. Sci. 230, 1183 (1950). 48. J . LBvy, J. Le Men, and M.-M. Janot, Bull. SOC.Chim. France p. 1658 (1961). 49. J. ,4. Joule and G. F. Smith, J. Chem. Soc. p. 312 (1962). 50. A. Z. Britten, P. N. Edwards, J. A. Joule, G. F. Smith, and G. Spiteller, Chem. I n d . (L ondon)p. 1120 (1963). 51. G. E’. Smith, Personal communication (1963). 51a. L. Olivier, J. LBvy, J. Le Men, If.-M. Janot, C. Djerassi, H. Budzikiewicz, J. M. IVilson, and L. J . Durham, Bull. Soc. Chim. Prance p. 646 (1963). 52. M. F. Millson and Sir Robert Robinson, J . Chem. Soc. p. 3362 (1955). 53. A. Z. Britten and G. F. Smith, J. Chem. Soc. p. 3850 (1963). 54. G. F. Smith, Lecture delivered a t Anniversary Meeting, Chem. SOC.,Cardiff, March 1963. 55. A . Z. Britten, G. F. Smith, and G. Spiteller, Chem. I n d . ( L o n d o n )p. 1492 (1963). 56. M. F. Bartlett, R. Sklar, A. F. Smith, and W. I. Taylor,J. Org. Chem. 28, 2197 (1963); W. I. Taylor, M. F. Bartlett, L. Olivier, J. Lplvy, and J. Le Men, Bull. Soc. Chim. France p. 392 (1964). 57. L. Olivier, J. LBvy, J. Le Men, M.-M. Janot, H. Budzikiewicz, and C. Djerassi, Ann. Phorm. Franc . 22, 35 (1964).

NOTEADDEDIN PROOF Section IX, Akuammiline (see p. 155): I n a recent study of this minor base, Olivier et al. (57) have deduced a close structural affinity with picraline, and have tentatively proposed the structure LXXXIX.

LXXXIX

This Page Intentionally Left Blank

-CHAPTER

-8

ALKALOIDS OF ALSTONIA SPECIES

J. E. SAXTON The University, L e d , England

I. Occurrence ........................................................

................................... 111. A l s t o n ~ i n e........................................................ IV. ALstonidine ........................................................ V. Echitamine ....................................................... VI. Echitamidine ...................................................... VII. Villalstonine ...... .................................................. VIII. Macralstonine ..................................................... IX. Macralstonidine.. .................................................. X. AlkaloidC ........................................................ 11. Alstonine and Tetrahydroalstonine

References ........................................................ Addendum : Veneratire

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

159 162 170 173 174 191 194 195 196 197 198 202

I. Occurrence The early chemistry of the alkaloids of Alstonia species is characterized by confusion and colored by an acrimonious dispute between the two principal protagonists concerning the constituents of A . scholaris. Most of the bases isolated from this and other species were a t best ill-defined and uncharacterized, and their identity and status as individual alkaloids are open to question; it is probable that several of the substances obtained were simply the same alkaloid in different states of purity. Investigations into the constituents of Alstonia species were stimulated by the knowledge that in the Far East extracts of Alstonia species were commonly used as a cure for malaria. For example, the Filipinos used an extract of A . scholaris, known as dita bark, and in Java and Batavia A. spectabilis was similarly used. This reputation has not survived careful pharmacological investigation, but it was widely believed by both physicians and natives, and two of these species, namely, A . scholaris and A. constricta, were formerly recognized in the British Pharmacopoeia. During an epidemic in Manila, dita bark was stated t o have superseded 159

160

J. E. SAXTON

quinine, but if the reports of the toxicity of this material are reliable this must have been a dangerous practice. The first extraction of an Alstonia species was reported by Palm, who isolated alstonine, an allegedly non-nitrogenous bitter principle from A. constricta F. Muell (1). The name alstonine was subsequently given to other preparations of doubtful purity from the same bark ( 2 , 3 ) ; a second base, alstonicine, was also obtained. I n 1865 Hesse isolated two bases, porphyrine and chlorogenine (4); later, he suggested that Palm’s alstonine was a mixture of these two bases, and that the alstonine of von Miiller and Rummel was impure chlorogenine (5). I n a more extensive investigation, Hesse obtained further details of porphyrine (probably identical with Oberlin’s alstonicine) and chlorogenine, and added a third alkaloid, alstonidine (6). He now admitted the identity of chlorogenine with Palm’s alstonine, and accepted the designation alstonine. Aside from alstonine, which was the only fully characterized base, these alkaloids were of doubtful homogeneity. In a careful examination of A . constricta bark, Sharp (7) confirmed the presence of alstonine and obtained it pure, and also isolated three further alkaloids, alkaloids A, B, and C, which could not be definitely identified with Hesse’s alkaloids. More recently, a second pure base, alstoniline, has been isolated (8).The difficulty of isolating pure crystalline bases from this material was owing to their susceptibility to atmospheric oxidation ;this was not appreciated by the earlier workers, but it frustrated their attempts to obtain satisfactory duplication of experimental results. The foregoing alkaloids were extracted from the trunk bark; the root bark contains reserpine (9, 10, 1l ) , alstonine, tetrahydroalstonine, a-yohimbine (rauwolscine), and a weak base which is probably alstonidine (9).Botanically, A . constricta differs markedly from other members of the same genus, and it is interesting that alstonine and its congeners do. not occur in other Alstonia species. However, alstonine does occur in some other genera, namely, in RauwolJin vomitoria Afzel., R. obscura K. Schum. (12), R. canescens L. (3. hirsuta Jacq.) (13, 14, 15), and in Vinca rosea L. (Lochnera rosea Reichb.) (16, 16a). Tetrahydroalstonine also occurs naturally, e.g., in the root bark of R. sellowii Muell. Argov. (17, 18),in the roots of V . rosea (16, 18a), and in V . Zancea Boj. ex A.DC. (L. lancea K. Schum.) (19); its N,-metho derivative is the quat,ernary alkaloid melinonine A, which occurs in the bark of Strychnos melinoniana Baillon (19a). The bark of A . scholaris R.Br. (Echites scholaris L.) was more widely used medicinally in the Far East than any other Alstonin species, and was intensively investigated in the years 1875-1880. Von Gorup-Besanez, working with a crude extract (“ ditaine ”) prepared from the bark by a

8.

ALKALOIDS OF

Alstonia

SPECIES

161

Manila apothecary, isolated a small amount of a crystalline base, but this was not further studied (20). Hesse later examined dita bark and obtained, in addition t o several non-nitrogenous constituents, the amorphous alkaloid ditamine, from the weakly basic fraction (21, 22, 23), and echitamine, from the strongly basic fraction (23). I n a more thorough investigation (24), he again described the isolation and characterization of ditamine and echitaniine and added a new alkaloid, echitenine. Meanwhile, Harnack reported the isolation of '' crystalline ditaine " (25), claiming later that this was the only alkaloid present (26). It was reputed to be identical with Hesse's ditamine, and was regarded as glycosidic in character. I n the controversy that fdlowed (23, 24, 26, 27, 28), it was eventually established that Harnack's ditaine was identical with echitamine, and that it was certainly not a'glycoside. Some years later, Bacon (29) repeated the extractions of dita bark and confirmed Hesse's work, but no new bases were isolated. It remained for Goodson t o discover the presence of one further alkaloid, echitamidine (30). The bark of A. spectabilis R.Br. was extracted by Hesse (23, 31), who obtained ditamine, echitamine, echitenine, and alstonamine, the lastnamed being possibly identical with Scharlke's alstonine, isolated some years earlier. Echitamine is the principal alkaloid of this group of Alstonia species; the other alkaloids are present in only very small amounts. Echitamine is also the principal base in A. congensis Engl., A. Gillettii De Wild. (30, 32), A . angustiloba Miq., A. spatulata Blume (30), A . verticillosa F. Muell. (33), and A . neriifolia D. Don. (34). The last-named bark also contains nerifoline, which is demethylechitamine (34) but, owing to the ease with which echitamine is known to be hydrolyzed, nerifoline may not be true constituent of this species. The bark of A . congensis also contains echitamidine in very small amounts (30). The third group of Alstonia species contains neither the alstonine nor the echitamine series of alkaloids; hence, the genus appears to be divided into three clearly defined sections as far as alkaloid content is concerned. A . macrophylla Wall. contains villalstonine (33, 35), macralstonine (33, 36), niacralstonidine, a base M which was obtained in minute amounts only (33), and macrophylline (37). A . viZZosa BIume contains villalstonine and base V, insufficiently characterized (33) ; A . somersetensis F. M. Bailey contains villalstonine and macralstonidine (33). Finally, a preliminary examination of A. muelleriana Domin. has yielded four alkaloids : alkaloid A, CzlH24-26N202, mp 322"-325" (dec.); alkaloid B, C40H50N404, mp 235"-270" (dec.); alkaloid C, C20H,2N,0,, mp 168"-169"; and alkaloidD, C20H,2N,0,, mp 172"-173" (38). Of these, only alkaloids B and C were obtained in adequate amounts for further

162

J. E. SAXTON

study, so it is not yet clear to which group of Alstonia species A. muelleriana belongs ; however, the recent elucidation of the structure of alkaloid C (videinfra) suggests that it may belong to none of these groups. Paper chromatographic examination of the total alkaloidal extracts of the bark indicated the presence of twenty alkaloids in addition to alkaloids A-D. Aside from the purely chemical aspects of these alkaloids, A . muelleriana warrants further study, as the total alkaloidal extracts possessed a hypotensive activity superior in some respects to that of pure reserpine. To which constituent of the bark this must be ascribed is unknown, but it is not owing to alkaloids B, C, or D (3%). The pharmacology of Alstonia extracts and of the pure alkaloids has been investigated on a number of occasions, but so far there is no indication of any effective antimalarial activity (8, 29, 32, 39-47).

II. Alstonine and Tetrahydroalstonine Alstonine, C21H20N203, is a yellow base which forms well-defined yellow salts. It can be obtained as crystalline hydrates, CzlHzoN203. 4&O and C2lH2oN203.1.25H20, but the anhydrous base cannot be obtained crystalline owing to decomposition. Alstonine is a monoacidic tertiary base which contains one methoxyl group but no methylimino groups ( 7 ) . Its salts are resistant to hydrogenation but the free base readily furnishes a colorless tetrahydro derivative which is also a monoacidic tertiary base. Tetrahydroalstonine can be saponified by prolonged boiling with strong alkali, resulting in the formation of tetrahydroalstoninic acid, C20H22N203, from which the alkaloid can be regenerated by esterification ; hence, the methoxyl group is contained in a carbomethoxy group, possibly attached to a quaternary center. The function of the remaining oxygen atom was unknown, but it appeared not to be present in a carbonyl or hydroxyl group (48). Early attempts to degrade alstonine met with only limited success. Oxidation with potassium permanganate gave a mixture of products, one of which was identified as N-oxalylanthranilic acid, a characteristic oxidation product of compounds containing the quinoline or indole ring system. Selenium dehydrogenation gave an oxygen-free base, alstyrine, which was formulated as ClsHzoNz or C19H22N2. Exhaustive methylation of alstyrine failed to give any identifiable nitrogen-free products, although an amorphous second methine base was obtained, which gave typical indole color reactions (48). The tentative conclusion that alstonine possesses an indole nucleus was amply confirmed by the results obtained by fusion of the alkaloid and its tetrahydro derivative

8. ALKALOIDS OF Alstonia SPECIES

163

with potassium hydroxide, and by the thermal decomposition of alstonine. The sole product identified from the alkali fusion of alstonine was harman (I;R = Me), but tetrahydroalstonine on similar treatment gave a complex mixture of products which contained harman, norharman (I;R = H), and three unidentified bases (possibly /3-carboline derivatives also) in the basic fraction, indole 2-carboxylic acid, and a neutral fraction suspected t o be a mixture of indole derivatives. Thermal decomposition of alstonine also gave a mixture of products ; although none of the bases was identified, the behavior of at least one of them suggested that it was a /3-carboline derivative (49). The UV-spectrum of tetrahydroalstonine closely resembles that of yohimbine, except that it contains an inflection a t 250 mp. Hence, tetrahydroalstonine probably contains an indole ring system and an additional chromophore. Reduction of tetrahydroalstonine with sodium O Zconversion , of and butanol gives a hexahydroalstonol, C Z ~ H Z ~ N Zby the ester function into a primary alcohol grouping and addition of two

Et

I

I1

I11

hydrogen atoms. The UV-spectrum of the product was exactly that expected for a 2,3-disubstituted indole. The addition of two extra hydrogen atoms was difficult to interpret ; it appeared not to be due to the saturation of carbonyl or imino functions, or to cleavage of an ally1 ether, since hexahydroalstonol contained only one acetylateable center, namely, the primary hydroxyl group. A preferable explanation was complete reduction of an c+unsaturated ester grouping, but in this case the resistance of this double bond to saturation during the preparation of tetrahydroalstonine was puzzling (49). From these observations no unequivocal conclusion could be reached concerning the identity of the chromophore which absorbs a t 250 mp, nor the function of the third oxygen atom. However, the presence of a /3-carboline system was accepted, in which N,, was substituted and N, was unsubstituted; the latter accounted for the presence of one active hydrogen in tetrahydroalstonine. Since alstonine hydrochloride absorbs a t higher wavelengths than 2-ethyl 8-carboline hydrochloride, it was assumed t o possess greater conjugation, and the partial structure I1 was therefore proposed for alstonine (49).

164

J. E. SAXTON

The constitution of alstyrine is clearly relevant to the elucidation of the structure of alstonine, since it contains all but two or three of the carbon atoms of the alkaloid. I n an investigatior of the structure of corynantheine, Karrer and Enslin (50) obtained a base, corynanthyrine, C19H22N2, by selenium dehydrogenation. The structure (111)of this base was established by ozonolysis and hydrolysis, which furnished o-aminopropiophenone and 4,5-diethylpyridine-2-carboxyiicacid. Corynanthyrine and alstyrine were suspected to be identical; this was confirmed by direct comparison (51). The structure thus deduced for alstyrine was acid was later confirmed by synthesis. 4,5-Diethylpyridine-2-carboxylic synthesized from 5-ethyl-2-methylpyridine and converted into the corresponding nitrile, which reacted with propyl magnesium bromide t o give 2-butyryl-4,5-diethylpyridine. Fischer cyclization of the phenylhydrazone of this ketone then gave 2-(4’,5’-diethy1-2’-pyridylf-3ethylindole (111))identical with alstyrine of natural origin (52). The isolation of ,B-carbolinederivatives and alstyrine from the degradation of alstonine suggests that the alkaloid contains the ring system IV ; to this must be added the chromophores, the ester group, and the third, unidentified, oxygen atom. Since the UV-spectrum of alstonine is very

IV

R

similar to those of tetradehydroyohimbine and serpentine, which were known to be anhydronium bases (53),it seemed more than probable that alstonine was also an anhydronium base (54).The nature of the second chromophore was deduced from the behavior of the products obtained by reduction of tetrahydroalstonine. Reduction with lithium aluminum hydride gives tetrahydroalstonol, C~oHz4NzO2,by reduction of the ester group ;the double bond which is reduced by sodium and butanol remains unaffected. Since the UV-spectra of tetra- and hexahydroaistonol are identical and charscteristic of 2,3-disubstituted indole derivatives, the double bond still presumed t o be present in tetrahydroalstonol is not conjugated with the indole nucleus. The IR-spectrum of tetrahydroalstonine exhibits an ester band a t 1715 em-’; hence, the double bond is conjugated with the ester grouping. The function of the remaining oxygen atom is more difficult t o establish. It is not likely to be present as a hydroxyl group, since hexahydroalstonol has only two active hydrogen

8. ALKALOIDS OF AZsstonia SPECIES

165

atoms (owing to N H and one OH), as compared with three in the isomeric yohimbol (i.e., four in yohimbol hemihydrate). Since tetrahydroalstonol also has two active hydrogen atoms and is not reduced further by lithium aluminum hydride, the remaining oxygen cannot be contained in a carbonyl group. On the other hand, tetrahydroalstonine is hardly affected by hydrobromic acid a t 140°, so if an ether link is p;esent, it must be highly unreactive (54). On the assumption that tetrahydroalstonine is an a$-unsaturated ester, tetrahydroalstonol must be an ally1 alcohol, which accounts for its anomalous behavior. This substance is unusually labile to acids, and cannot be recovered from strongly acid solutions. With methyl iodide and picric acid, salts are formed, but one molecule of water is simultaneously eliminated. With ethyl alcohol, tetrahydroalstonyl ethyl ether, C22H28N202, is formed, whereas on catalytic reduction, hydrogenolysis accompanies hydrogenation and the product is deoxyhexahydroalstonol, CzoHz~N20.I n contrast, hexahydroalstonol, in which the double bond has also been reduced, behaves normally. These data, together with the fact that alstonine contains a C-methyl group, were initially summarized in the constitution V, proposed for alstonine, and VI for tetrahydroalstonine (54).

V

VI

,/O

MeOOC/

I

Me VII

I

Me VIII

However, the UV-spectrum of tetrahydroalstonine (maxima a t 230 and 290 mp, inflection a t 250 mp) indicates that it is composed of an indole chromophore plus additional absorption in the 250 mp region. This spectrum shows quite different characteristics from the summation spectrum of 2,3-dimethylindole and 2,6-dimethyl-3-carbomethoxy-5,6dihydro- 1,2-pyran (VII); the summation spectrum exhibits a pronounced minimum a t 250 mp. Hence, formula VI for tetrahydroalstonine

166

J. E. SAXTON

is invalidated. On the other hand, the spectrum of 2,6-dimethyl-3-carbomethoxy-5,6-dihydro-1,4-pyran ( V I I I ) has a maximum at 250 mp, and a summation spectrum of this compound with 2,3-dimethylindole is superimposable on the spectrum of tetrahydroalstonine. Consequently, tetrahydroalstonine may be formulated as IX, i.e., as a stereoisomer of ajmalicine, and alstonine as X, i.e., as a stereoisomer of serpentine.

MeOOC/\/O

IX

A

Tetrahydroalstonine

Alstonine

/\ MeOOC CH

The IR-data are also consistent with these conclusions. Twin maxima a t 1695 and 1613 cm-1 are exhibited by tetrahydroalstonine, corynantheine, tetrahydroserpentine, and the model substance VIII, all of which possess the chromophore, ROOC-C=C-OR’ ; these are not observed in the spectra of VII and the saturated pyran derivatives ( 5 5 ) . The formula X for alstonine also explains all of its apparently anomalous properties. The resistance of the ester group t o saponification and of the ring E double bond t o hydrogenation are characteristic of V I I I , whereas the presumed molecular compound obtained earlier by reaction of alstonine with dinitrophenylhydrazine (54)is probably a derivative of its open-chain carbonyl equivalent (XI) ( 5 5 ) . With regard to the stereochemistry of tetrahydroalstonine, the IRspectrum in the 2800 cm-l region exhibits peaks of medium intensity on the high-wavelength side of the major C - H stretching vibration a t 2890 cm-l, characteristic of the yohimbine or heteroyohimbine stereo-

8.

ALKALOIDS OF

Alstonia

SPECIES

167

isomers containing CL hydrogen a t C-3, i.e., of the normal or all0 series (56, 57). This assignment is supported by the observation that reduction of alstonine catalytically or with sodium borohydride gives tetrahydroalstonine ; in reductions of this type the D/E trans derivatives give normal products, and those possessing a cis D/E ring junction give predominantly all0 products. Consequently, 3-isotetrahydroalstonine (akuammigine) belongs to the pseudo or epiallo series. By similar reasoning, it can be shown that 3-isoajmalicine also belongs to one of these series. Since 3-isotetrahydroalstonine is dehydrogenated to the corresponding tetradehydro derivative much more readily with palladium and maleic acid than is 3-isoajmalicine, it was inferred that 3-isotetrahydroalstonine belongs to the pseudo series ; hence, tetrahydroalstonine was presumed to belong to the normal series, and was formulated as XI1

MeOOC/\/O XI1

(57, 58). This deduction from the roughly determined rates of dehydrogenation in a heterogeneous reaction was based on the comparative rates of dehydrogenation similarly obtained with pseudo and epiallo stereoisomers of yohimbine and yohimban, for which the stereochemistry a t (2-3, C-15, and C-20 had been rigidly proved. Clearly, the extrapolation of such an argument to the heteroyohimbine series was largely empirical, and was not claimed to lead t o a definitive proof of the stereochemistry ; nevertheless, as a basis for further experimentation, it allowed useful interim proposals to be made. The same objection applies to a later argument, which supported the conclusions already reached. The IRspectrum of tetrahydroalstonine exhibits three ester bands (C-0 stretching vibration) a t 1227, 1202, and 1183 em-', whereas ajmalicine exhibits a comparatively simple peak a t 1183 em-'. The double-bond absorptions also differ; for tetrahydroalstonine, it is a t 1629 cm-l, and for ajmalicine, a t 1616 em-'. These differences were explained (from a study of appropriate molecular models) in terms of steric hindrance to rotation of the ester group by the hydrogen atoms at (2-14 in the D/E trans isomer (tetrahydroalstonine) with resultant nonplanarity of the unsaturated carbonyl system. I n the D/E cis isomer (ajmalicine), no such hindrance was observed in the models used (59).

168

J. E. SAXTON

On the basis of these conclusions it was perhaps surprising that a nonstereospecific synthesis of a heteroyohimbine should yield predominantly the presumed c i s 3 / E isomer, ( t )-ajmalicine, rather than the trans D/E isomer (60). However, later correlations by reliable chemical and physical methods proved beyond doubt that the earlier

1. Alcoholic

H,' ' Y O HO

XI11

XIV

Tetrahydroalstonine

XVI

XVII

XVIII Tetrahydroalstonine

conclusions were in error, and that tetrahydroalstonine does, in fact, possess a cis DIE ring junction. Treatment of tetrahydroalstonine (XIII) with alkali followed by aqueous acid resulted in hydrolysis and decarboxylation, and formation of the hemiacetal, tetrahydroalstonial

8. ALKALOIDS O F AbtOniU SPECIES

169

(XIV). Wolff-Kishner reduction of XIV gave 19-corynantheidol (XV), which was oxidized by the Oppenauer method to 19-corynantheidone (XVI). When this was equilibrated with sodium methoxide, it was converted into the more stable trans isomer, 18,lg-dihydro-19-corynantheone (XVII) (61). A similar sequence of reactions starting from ajmalicine also afforded the ketone XVII, but here the final epimerization stage was unnecessary (62). These results amply demonstrate that tetrahydroalstonine possesses a c i s D/E ring junction, and therefore belongs to the allo series. I n the NMR-spectrum of tetrahydroalstonine, the one-proton octet owing to the coupling of the C-19 hydrogen with the C-20 hydrogen and the hydrogens of the methyl group shows a large spin-spin coupling constant, which is explained in terms of a 19,20-trans diaxial arrangement of hydrogen atoms. The C-19 methyl group is consequently c i s with respect to the C-20 hydrogen atom, and the complete stereochemistry of tetrahydroalstonine is as given in XI11 ( = XVIII) (61, 63). The same conclusions concerning the stereochemistry of tetrahydroalstonine were reached from a comparison of the dissociation corstants and rate of methiodide formation of tetrahydroalstonine and its stereoisomers (64). Both the reduced basicity of tetrahydroalstonine, pK, 5.83 (cf. ajmalicine, pK, 6.31) and its reduced rate of methiodide formation when compared with ajmalicine are in consonance with its formulation as a c i s D/E isomer, in which Nbis to some extent sterically hindered

xx

0

AlStOnin6

XIX Ajmalicine

by the C-19 hydrogen atom. This point is evident from a comparison of the conformations of tetrahydroalstonine (XVIII) and ajmalicine (XIX) (64). On the basis of these arguments, therefore, alstonine is now formulated as XX.

170

J . E. SAXTON

111. Alstoniline

Alstoniline, C22H18N203, an optically inactive base, occurs to the extent of 0.02% in A . constricta bark, and is isolated therefrom as its bright-red hydrochloride. Fusion of the base with potassium hydroxide gives rise to 2-methylisophthalic acid, identified by synthesis. Catalytic reduction of alstoniline hydrochloride over platinum oxide yields tetrahydroalstoniline, which is unstable in the presence of reduced platinum and air, and is rapidly oxidized to alstoniline oxide, C22H18N204.2H20, which is also the product of aerial oxidation of alstoniline itself. However, pure tetrahydroalstoniline is stable, and can be crystallized from methanol as a solvate. It contains a tetrahydrocarboline nucleus (Adamkiewicz color reaction) and a carbomethoxy group, since lithium aluminum hydride reduction produces tetrahydroalstonilinol, C21H22N202, which contains only one methoxyl group, as compared with two in alstoniline. The IR-evidence is consistent with the replacement of an ester group by a primary alcohol function. I n contrast, similar reduction of alstoniline hydrochloride produces an unstable, paleyellow substance, which on catalytic reduction absorbs only one mole of hydrogen, the product being tetrahydroalstonilinol. Hence, of the two double bonds of alstoniline susceptible to catalytic reduction, one is also reducible by lithium aluminum hydride. These properties could be explained by the presence in alstoniline hydrochloride of a ~-carbolinium ion, but a comparison of its UV-spectrum with that of alstonine hydrochloride excludes this possibility. On the other hand, the spectrum bears some similarity to that of ketoyobyrine (65).The second methoxyl group is probably attached to the indole nucleus, since the spectrum of tetrahydroalstoniline hydrochloride has the same general shape as that of 6-methoxyindole, but shows a shift of 10 mp toward longer wavelengths (66). The presence of such a methoxyl group is supported by reaction of the alkaloid with hydrobromic acid, which gives a phenolic base, demethylalstoniline (65). Although this evidence was insufficient to deduce rigidly the constitution of alstoniline hydrochloride, formula X X I was tentatively proposed, since it explains satisfactorily the experimental data (65). The interconversions of the alkaloid can thus be summarized as illustrated o n p . 171. The methoxyl group was placed a t C-11 (yohimban numbering) by analogy with harmine, but C-10 could not be excluded with certainty. However, it has since been shown that the spectrum of alstoniline chloride is similar to that of synthetic 3-(6-methoxy-3-methyl-2indolyl)-2-methylisoquinoliniumiodide (XXII), whereas the spectrum

P i \A

0 2

+ \/\

I

I1

M e O O C / V XXIII

T

XXIV

Po

T

HzlPtOz

NaOHI CSH11OH

HzIPtOz

$-

COOH

F

m H

LiAlH4 t-

xxv XXI Alstoniline hydrochloride

b I; 0

N

m v M m

XXII

172

J. E. SAXTON

of tetrahydroalstoniline (XXIII) is almost identical with that of the tetrahydroisoquinoline base corresponding to (XXII) (66). The position of the methoxyl group is finally established by the alkaline degradation of alstoniline oxide (XXIV),which results in the formation of norharmine (XXV) and 2-methylisophthalic acid; thus, i t is possible to account for all the carbon atoms of alstoniline (67).

XXI

(as bromide)

XXVIII Ia/KOAc

xxx Alstonidine

XXIX

The structure of alstoniline was eventually confirmed by synthesis of alstonilinol and tetrahydroalstonilinol (68). 6-Methoxytryptophol, prepared from 6-methoxyindole via 6-methoxyindolylglyoxylyl chloride, was converted into the corresponding unstable bromide (XXVI), which was condensed with 5-carbomethoxyisoquinoline. The resulting quater-

8. ALKALOIDS OF Alstonia SPECIES

173

nary salt (XXVII) was then reduced with lithium aluminum hydride; the intermediate dihydroisoquinoline derivative cyclized spontaneously, and the product isolated was tetrahydroalstonilinol (XXVIII),identical with the product prepared from alstoniline. Finally, dehydrogenation of XXVIII with iodine and potassium acetate gave alstonilinol iodide (XXIX) (68). The synthesis of alstoniline itself was achieved in an extraordinarily direct manner, by prolonged heating of XXVI with 3-bromo-5-carbomethoxyisoquinoline a t 90"-95" ; the product was converted into the corresponding picrate, which was shown to be identical with authentic alstoniline picrate (68a).

IV.

Alstonidine

Alstonidine, C22H24N204, mp 186"-188", contains one C-methyl, one N-methyl, and one methoxyl group (9). Its molecular weight has been confirmed by the X-ray method (69) and by potentiometric titration; the latter also gives pK, 5.95 (66% dimethylformamide) (70). The UVspectrum of alstonidine clearly indicates that it is a P-carboline derivative ; since the spectrum resembles that of ind-N-methylharman more closely than that of harman, the indole nitrogen is presumed t o be methylated. The close similarity of these spectra also indicates that no oxygenated substituents are attached to the P-carboline system. Alstonidine is a methyl ester, since saponification gives alstonidinic acid, isolated as the trihydrate, mp 238"-240" ; esterification of the latter with diazomethane regenerates alstonidine. The IR-spectrum of alstonidine exhibits two intense peaks at 1698 and 1629 cm-l, indicating that the carbomethoxy group is present in a P-alkoxyacrylic ester residue. The remaining oxygen is contained in a hydroxyl group, since alstonidine gives an 0-acetate, mp (trihydrate) 92"-96". The position of the hydroxyl absorption in the IR-spectrum (3145cm-l, intensity independent of concentration) indicates that the hydroxyl group is involved in a strong intramolecular hydrogen bond. From these data, the structure X X X has been proposed for alstonidine (70).Although this structure is consistent with the known properties of this alkaloid, further work is desirable in order to establish the precise nature of the dihydropyran residue, which was formulated in this way partly by analogy with the structure of the heteroyohimbine bases. I n particular, a facile conversion of a structure such as XXX into a close relative of alstonine should be feasible. Unfortunately, lack of material has so far precluded all attempts at a structural correlation of this type (70).

174

J. E. SAXTON

V. Echitamine Echitamine was first described adequately by Hesse (24), who isolated it from A . scholaris as echitamine chloride, CzzHZ&zO&l, [.ID - 57", colorless needles from water. Hesse realized that this was the salt of a very strong base, echitammonium hydroxide, which could be precipitated from its salts by potassium hydroxide but not by ammonia ; the H zO0 ,~mp . 206", base thus obtained formed glassy prisms, C Z ~ H ~ O N3 Z [a]&5o - 28.8" from ethanol. When dried at SO", three molecules of water were lost, to give a product which was also a strong base. A fourth molecule of water was lost at 105" in vacuo ; the anhydrous base obtained, CzzHzsN204,was considerably weaker than the original, showing that a profound structural change had accompanied the removal of the last molecule of water. That this change was reversible was demonstrated by the reformation of echitamine chloride when the anhydrous base was treated with hydrochloric acid. These data were confirmed by Goodson and Henry (32), who established the presence in echitamine of one methoxyl group and one methylimino group. The methoxyl group is contained in a carbomethoxy group, since echitamine chloride is readily hydrolyzed by dilute alkalis to give demethylechitamine, C2lHzsN204.2Hz0, mp 268" (dec.), a neutral betaine-like substance which does not contain a methoxyl group and which gives rise to an acidic hydrochloride. The behavior of one of the two nitrogen atoms in echitamine chloride is thus characteristic of that of a quaternary ammonium grouping; the function of the other nitrogen atom is less obvious. Goodson and Henry reported the formation, in low yield, of a yellow nitroso derivative which gave a Liebermann's reaction, but the evidence for the formatian of an N-nitroso derivative was inconclusive. Consequently, the formation of a diacetyl derivative could be regarded as evidence for the presence of a t least one hydroxyl group but, owing to the undetermined nature of the second nitrogen atom, it was not clear whether the second acetyl group was attached to nitrogen or to oxygen (32). The presence of an indole nucleus in echitamine was suspected from the positive Hopkins-Cole color reaction, and was confirmed by the distillation of echitamine with alkalis (dry distillation with soda lime or distillation with 50% potassium hydroxide solution) which afforded a substance having a pronounced fecal odor and exhibiting a positive Ehrlich color reaction; although unidentified and of uncertain purity, this product was clearly an indole derivative. Methylamine was also isolated from the distillation with alkali. The UV-spectrum of echitamine chloride exhibits maxima a t 235 and

8. ALKALOIDS OF Alstonia SPECIES

175

295 mp, unaffected by the addition of strong acid (71, 72, 72a). This spectrum is characteristic of Nb-quaternary or N,-protonated eserine systems, in which the proximity of the positively charged nitrogen atom effectively prevents protonation of N,, even in the presence of an excess of strong acid (73). I n general, this type of spectrum is exhibited by dihydroindole derivatives in which N, and N, are separated by not more than two carbon atoms. I n echitamine chloride, the quaternary center was assumed t o be N,, which was also assumed to carry the methyl group (72). The IR-spectrum of echitamine chloride confirmed the presence of hydroxyl, imino, ester, and o-disubstituted benzene groups, group. The quaternary character of and the absence of an =N+Hechitamine chloride and the betaine character of demethylechitamine were confirmed by potentiometric titration (72). Hydrogenation experiments gave some information concerning the environment of Nb in echitamine chloride. I n aqueous solution in the presence of palladium charcoal, echitamine chloride gave a tertiary base whose UV-spectrum indicated that it was a dihydroindole derivative. The hypsochromic shift of 10 mp observed in the spectrum in acid solution (protonation of N, but not N,) suggested the presence of an eserine-like system. Since this product possessed a methylimino group and a C-methyl group, in contrast to echitamine chloride which had been presumed to contain no C-methyl groups, it was assumed that Emde fission of an allylamine system had occurred. It was not a t first clear whether the ally1 double bond had suffered hydrogenat’ion; hence, this base was first named tetrahydroechitamine (74) and, later, dihydroechitamine, when it was established that the hydrogenation product still contained the double bond (75). Since dihydroechitamine appeared to have the composition C22H28N203, it was suggested that echitamine chloride was actually a hydrate, C22H27N203Cl.H20, and the course of the hydrogenation was explained as follows :

I n fact, the true interpretation of the course of this reduction is more complex, and was only given later by Conroy et al. (82). Nevertheless, the mode of hydrogenation seemed to support the quaternary eserine formulation for echitamine chloride, and this was further confirmed by the properties of “ dihydroechitamine.” Although it did not undergo the expected reductive cleavage with zinc and hydrochloric acid, dihydroechitamine methiodide readily gave a methine base with cold dilute alkali, yielding dihydroechitamine methine. The UV-spectrum of this

176

J. E. SAXTON

product was typical of a dihydroindole derivative ; the hypsochromic shift observed in acid solution was explained by the recyclization of the initial carbinolamine base produced by Hofmann degradation of an eserine-like system :

Since this methine base was not oxidized by potassium ferricyanide, i t was assumed to contain a tertiary hydroxyl group at C-2, and since the carbinolamine was not readily dehydrated to an indole derivative, it was assumed t o be a p,P-disubstituted dihydroindole derivative (75). The fact that this methine base did not spontaneously dehydrate to give an indolenine was not regarded as a serious difficulty, as it was considered probable that some structural feature prevented such dehydration. The carbinolamine character of the methine base was supported by the result of reduction with zinc and hydrochloric acid, which yielded ‘‘ deoxyneonordihydroechitamine methine,” C22H30N203, a dihydroindole base which exhibited typical dihydroindole absorption in neutral and acid solution. The failure to protonate N, in acid solution was explained by the presumed spatial proximity of Nb even after cleavage of the C-2 to Nb linkage, this being a consequence of the ring system present in the molecule (75). The N,H group in echitamine chloride is also present in its acetyl derivative (IR-spectra) ; it is evident that the proximity of a quaternary B b prevents acetylation of N,. I n contrast, dihydroechitamine gives an hT,-acetyl derivative, identified by its ZR- and UV-spectra, and by the Positive Otto reaction. Hence, the presence of an imino group in echitamine is confirmed (72). Although echitamine chloride was readily hydrolyzed t o give demethylechitarnine, dihydroechitamine was found to be very resistant to saponification, thus suggesting that the carborhethoxy group was tertiary ( 7 5 ) . These results were extended (and to some extent anticipated) by the ihdependent work of Birch et al. (76), who established the presence in echitamine chloride of a C-methyl group by Kuhn-Roth determination, ahd also by ozonolysis, which gave a 25% yield of acetaldehyde. Hydrogenation of the amorphous base A, C22HzsNz04, obtained by basification Of echitamine chloride, gave a crystalline tertiary base B, CzzHzsNz03, which could be further hydrogenated t o a base C, C22H30N203. Both base B and base C exhibited eserine-like UV-spectra in neutral and acid solution. Base C gave no acetaldehyde on ozonolysis, but contained two

8. ALKALOIDS O F AktOnia SPECIES

177

C-methyl groups; a modified Kuhn-Roth oxidation gave a mixture of acetic and a-methylbutyric acids. These results allowed the following partial structure to be proposed for echitamine chloride and base C (76) :

Dehydrogenation experiments on derivatives of echitamine gave interesting, if unexpected, results. The dehydrogenation of “dihydroechitamine ” with selenium gave echitamyrine, ClzHloNZ, which was later identified as XXXI (R = H), identical with the product obtained

by oxidation of calycanthine with silver acetate ( 7 2 , 7 4 ) .A closely related base, XXXI (R = Me), was obtained by Birch et al. by distillation of base B with zinc dust (76). This was the first time that a pyrroloquinoline base had been isolated by degradation of an alkaloid presumed t o be derived from tryptamine and dihydroxyphenylalanine (or prephenic acid). This phase of the investigations was concluded by a proposal (XXXII) for the structure of echitamine (77-80). This was based mainly on the foregoing arguments, together with biogenetic considerations ; although it accounted for many of the reactions of echitamine, it was quite untenable for several reasons. Thus, a substance of structure XXXII would be expected t o exist in the aminoaldehyde form rather than the carbinolamine form, owing to the almost nonbasic character of N, and t o the fact that the carbinolamine ring would be eight-membered. The equivalent aminoaldehyde would be a formylacetic ester derivative, and as such would be expected to decarboxylate spontaneously following saponification ; echitamine itself should also give rise t o a dinitrophenylhydrazone derivative. Neither of these possibilities has been observed with echitamine chloride (81). Further, it had already been established by preparation of an N,-acetyl derivative from a hydrogenation product of echitamine base that, in echitamine, N, was secondary. The Indian workers (77) failed to obtain the nitroso derivative reported earlier by Goodson and Henry; instead, they obtained a higher

178

J. E. SAXTON

melting derivative which was formulated as a C-nitro derivative (79). This result was taken as an indication that, in echitamine, N, was tertiary (78, 79); however, the proximity of the positively charged Nb in a quaternary eserine-like system would be sufficient to account for the failure of N, to nitrosate (cf. acetylation).

HO-CH

C \H'

I COOMe

XXXII

The presence of an ethylidene group in echitamine chloride was also demonstrated by oxidation with periodic acid, which was reported to give acetaldehyde and indole-3-acetaldehyde (78). Alkali fusion and selenium dehydrogenation experiments gave inconclusive results, but the basic fractions were suspected to contain derivatives of /3-carboline (77, 78). Oxidation of echitamine with alkaline potassium permanganate afforded a low-melting base, which was considered to be Nb-methyltryptamine (80). Although the presence of a quaternary eserine-type system in echitamine chloride had been accepted by all the earlier investigators, other workers considered that this was not the only chromophore that could explain the UV-absorption spectra satisfactorily. The proposal of Conroy et al. (82) for echitamine chloride (XXXIII) was one example of

XXXIII

XXXIV

an alternative approach. This ingeniously derived structure was based on much new evidence which allowed the clarification of several obscure

8.

ALKALOIDS O F

Abtonia

SPECIES

179

and apparently anomalous reactions of echitamine chloride and its derivatives. The earlier tentative suggestion, based on the behavior of echitamine chloride on hydrogenation, that echitamine chloride is CzzH2703NzCl.H2O was dismissed since no "anhydrous " derivatives could be obtained. Further, the diacetyl derivative obtained by Goodson and Henry was clearly an 0,O-diacetyl derivative (IR-spectrum), which thus required the presence in echitamine chloride of two oxygen atoms in addition to the carbomethoxy group. This composition was confirmed when echitamine base, for which the structure XXXIV was favored, was obtained crystalline, as its solvate with benzene, CzzHzsNz04.C6H6, mp 139"-140". The NMR-spectrum of echitamine base (OMe, CH&H=C<, methyl attached to tertiary nitrogen) proved conclusively that it was a tertiary base, and not a quaternary ammonium hydroxide ; normal behavior for a tertiary base was demonstrated by the formation of a methiodide, which contained two N-methyl groups. That the interconversion of echitamine base and echitamine chloride was not a simple proton exchange was substantiated by the marked hysteresis observed in the reconversion of echitamine into echitamine chloride a t 25", a t which temperature the measured pK, was 7.8. I n contrast, titration of 11, appropriate for the echitamine chloride gave an apparent pK, titration of a quaternary chloride (82). The presence of a /3-hydroxypropionic ester residue in echitamine chloride was proved by treatment with potassium tert-butoxide in tertbutanol which gave alloechitamine, C21H26N203 (XXXV),by retroaldol loss of formaldehyde. The IR-peak a t 1689 cm-1 in alloechitamine was N

xxxv

XXXVI

attributed to a ketonic carbonyl group, whose carbonyl absorption frequency was reduced by transannular interaction with Nb (cf. methylpseudostrychnine); since this carbonyl band was not present in the spectrum of alloechitamine methiodide, the latter was assumed to contain the grouping

180

J. E. SAXTON

The course ofthe hydrogenation of echitamine and echitamine chloride, which had earlier been postulated to give base B (76) and dihydroechitamine (74, 75), both incorrectly formulated as CzzHzsNz03, was also elucidated by Conroy et al. Hydrogenation of echitamine in ethanol gave, after the absorption of one mole of hydrogen, a base named echitinolide, C21H26Nz03, mp 154"-157' (from benzene), which was identical with " dihydroechitamine," and probably also identical with base B. Echitinolide behaved as a a-lactone (IR-absorption at 1742 cm-1) and contained one N-methyl and two C-methyl groups (NMR-spectrum) but no methoxyl group ; it also contained one imino and one hydroxyl group. Acetylation gave an 0-monoacetate, which contained one imino group but no hydroxyl group; the low pK, value (5.4) of O-acetylechitinolide was consistent with the proximity of 0-acetyl and tertiary amino functions. Echitinolide was clearly formed by loss of methanol from the product of hydrogenation of echitamine. Since echitamine chloride shows no tendency to lactonize, it was inferred that this was prohibited for steric reasons; however, in one (C-3) epimer of the carbinalamine obtained by Emde fission of the Nb to '2-21 linkage in XXXIII there is evidently no barrier to lactonization, and echitinolide was accordingly formulated as XXXVI. The isomer, isoechitinolide, formed when echitinolide was heated with hydrochloric acid, contained neither a hydroxyl group nor a double bond, and was regarded as XXXVII. This requires epimerization at the carbon atom bearing the butenyl group prior to ring closure, a process which was evidently feasible in XXXVI, but sterically prohibited in echitamine chloride (XXXIII) which, in contrast, was stable to acid (82). The hitherto-accepted view that the UV-absorption of echitamine chloride was due to a quaternary eserine-type system was challenged by Conroy et al. on the grounds that echitamine base exhibited the same 235 and 295 mp) although neither nitrogen atom carried spectrum (A, a positive charge. Echitinolide and isoechitinolide exhibited typical dihydroindole spectra, with a small hypsochromic shift ( w 10 mp) in acid solution; this effect was attributed by other workers to protonation of Nb but not N, in a geminal diamino system. However, Conroy et a,!. explained these characteristics by postulating that, in those substances in which N, is axial to ring C, absorption occurs in the regions 245-250 mp and 305-310 mp, while those possessing an equatorial N, (echitamine, echitamine chloride, and methiodide) absorb at 235 and 295 mp. I n echitinolide (XXXVI = XXXVIa) and isoechitinolide, N, is axial to a boat-shaped ring C, but in acid solution the lactone ring is opened to give a conformation (XXXVIII) in which N, is equatorial to a chair-shaped ring C; this structure in acid solution is consistent with the appearance

8. ALKALOIDS OF A h t O n i U SPECIES

181

in the IR-spectrum of broad carboxylic hydroxyl absorption at 28603450 cm-1.

-0 XXXVII

w

XXXVIa

XXXVII I

The methine base obtained by Hofmann degradation of echitinolide methiodide, presumed earlier t o be C 2 3 H 3 2 N 2 0 4 but now shown t o be C 2 2 H 3 0 N 2 0 4 , was formulated as XXXIX, and its zinc-hydrochloric acid reduction product as XL. Since the lactone ring in XL was unaffected

XXXIX

XL

-0

on protonation, there was now no opportunity for conformational inversion in acid solution ; hence, the UV-spectrum of Govindachari's deoxyneonordihydroechitamine methine (75), now renamed deoxyneodihydroechitamine methine (XL),remained unchanged in acid solution.

182

J. E. SAXTON

I n all these derivatives of echitamine, protonation of N, was never observed, even in very strongly acid solutions; this was attributed to steric hindrance to solvation of the ion (82). A closely similar view concerning the structure of echitamine chloride was put forward independently by Robinson and his co-workers (83), who also regarded the arguments supporting the presence of an eserinelike system as inconclusive. The hydrogenation of echitamine chloride, which yielded a deoxyechitamine or a deoxydihydroechitamine according to the conditions employed, was explained by the hydrogenolysis of a quaternary carbinolamine function ; in the formation of the dihydro derivative, concomitant Emde fission of a quaternary allylamine system was also presumed t o occur, since the product contained an additional C-methyl group. The environment of N, was thus formulated as

This was consistent with an observed pK, of 9.1 for echitamine chloride, indicating that echitamine was intermediate in strength between a tertiary base and a quaternary ammonium compound. The absence of the usual dihydroindole color reactions was attributed, like the failure to protonate N, in strong acid, to transannular deactivation of N, by the carbomethoxy group, a situation analogous to that in pseudoakuammigine, for which the structure XLI had earlier been postulated. The first structure proposed by Robinson et al. (83) on the basis of these

\CH. Me MeOOh XLI

CHzOH XLII

arguments was XLII (R = OH, R' = H). This position of the carbinolamine hydroxyl group was preferred to that postulated by Conroy et al. since echitamine chloride exhibited stronger reducing properties than substances similar to XXXIII (e.g., pseudostrychnine) in the strychnine series. However, this structure had to be modified when it was proved that echitinolide is a lactone ; here, the presumed carbinolamine hydroxyl group is not removed on hydrogenolysis, since it is involved in lactone

8.

ALKALOIDS OF

Alstonia

SPECIES

183

formation. The new structure proposed was XLII (R = H, R’ = OH), and echitinolide was formulated as XLIII.

XLIII

With the accumulation of new experimental evidence, it became clear that the structures advanced by Conroy and Robinson for echitamine chloride were less than satisfactory. The inadequacies of formulas XXXIII and XLII were discussed by Birch et al. (Sl), but no alternative formulation was proposed. On the vexed question of the UV-absorption of echitamine and its derivatives, the earlier postulate concerning the failure of N, to protonate even in concentrated acid was rejected, since 2,16-dihydroakuammicine (XLIV), which contains a similar system to

bOOMe XLIV

H H & COOMe I OAc XLV

XXXIII and XLII, was completely protonated even in 0.5 N ethanolic hydrochloric acid. Further, the hysteresis reported by Conroy et al. in the reconversion of echitamine base into echitamine chloride was not observed; on the contrary, it was the opinion of Birch et al. (81) that the UV-absorption of echitamine base could not be measured in aqueous solution, since the p H rose very rapidly to a value indicating complete ionization as the quaternary hydroxide, whose UV-absorption was necessarily identical with that of echitamine chloride. When measured in petroleum solution, echitamine base did, in fact, exhibit different absorption, with maxima a t 227 and 283 mp. The validity of Conroy’s explanation of the hypsochromic shift of the UV-spectrum of echitinolide in acid solution was also questioned. Hydrogenation of 0,O-diacetylechitamine chloride gave a base,

184

J. E. SAXTON

C26H34N206,which still contained the carbomethoxy group, and which should be formulated as XLV on the basis of XXXIII for echitamine chloride. The UV-spectrum of this base also showed a hypsochromic effect in acid solution similar to that shown by echitinolide, although in this case conformational change of the type postulated by Conroy was not possible. Other experimental data which were difficult to explain were the resistance of echitinolide to reduction in acid solution, and the comparative reluctance of echitinolide and isoechitinolide t o undergo reduction with lithium aluminum hydride under mild conditions. The conformations of the substances involved, i.e.. XXXVIII, XXXVIa, and XXXVII, respectively, revealed no steric factors which could account for this behavior (81). Perhaps the most cogent argument advanced against the structure XXXIII for echitamine chloride was obtained from the Hofmann degradation of echitinolide methiodide. If echitinolide were correctly represented by XXXVI, Hofmann degradation should give rise to one and only one methine base, the structure of which would be XXXIX. However, in contrast to earlier workers, Birch et at. obtained two isomeric bases, C22H30N204 ; the isomerism involved was clearly not of the type exhibited by echitinolide and isoechitinolide, since the isomerism was also observed in the analogous degradation of /3-dihydroechitinolide, one of the two diastereoisomeric derivatives obtained by saturation of the double bond of echitinolide (81). These criticisms of the structure XXXIII for echitamine chloride apply with equal force to the proposals of Robinson et al., namely, XLII (R = OH, R’ = H) and XLII (R = H, R‘ = OH). I n addition, the presence of a potential aldehyde group was not easy to reconcile with the failure of echitamine to undergo reduction with borohydride in alkaline solution. Echitamine base would have to contain a free aldehyde function, or alternatively (on the basis of XLII; R = H, R’ = OH) a bridged carbinolamine structure in which the aldehyde group would be coupled with N,; echitamine would then be expected to exhibit benzenoid absorption in the UV-region owing to the inability of the p-electrons on N, to conjugate with the benzene ring (81). The final denouement in this challenging and intricate structural problem was provided by the X-ray crystal structure analysis of echitamine bromide methanol solvate, C22H29BrN204.MeOH (84),and echitamine iodide ( 8 5 ) , according to which the structure and absolute configuration of echitamine are as given in XLVI. The six-membered ring containing the primary alcohol and carbomethoxy groups is a distorted boat, with two cis-fused five-membered rings attached to it ; the secondary hydroxyl and the carbomethoxy groups occupy equatorial

8. ALKALOIDS

OF

Alstonia SPECIES

185

positions, while the primary alcohol group and C-20 are axially oriented (XLVIa). I n structure XLVI, the ring system is numbered in the usual way so that the constituent atoms can readily be identified with their biogenetical equivalents in the related indole alkaloids (cf., corynantheine, XLVII). As with the vast majority of indole alkaloids, the C-15 hydrogen has the o? configuration.

9

6

XLVII

The elucidation of the structure of echitamine terminated the discussions outlined briefly above, and the structures of many of its

XLVIII

transformation products became immediately apparent. However, there are still several unexplained points in the chemistry of the degradation products of echitamine chloride, and there are two proposals concerning

186

J. E. SAXTON

the structure of echitamine base. The important hydrogenation product, echitinolide, can be formulated as XLVIII ; it is formed on fission of the Nbto C-2 1 bond by conformational transformation of the six-membered ring into the alternative boat form, in which both the carbomethoxy and hydroxyl groups are axially oriented, and therefore appropriately placed for lactonization to occur (86, 87). It is thus evident that the primary hydroxyl group is conveniently situated to form an ether ring on protonation of the double bond. Hence isoechitinolide can be formulated as XLIX (86) or L (87). The intermediate formation of a tertiary, in

XLIX

L

preference to a secondary, carbonium ion prior to cyclization would be expected t o give an ether of structure L. I n accordance with this, the NMR-spectrum of isoechitinolide (L) shows a triplet centered at 9.38 r , and a quartet centred a t 8.69 7 , consistent with the presence of an ethyl group at.tached to a quaternary carbon atom. These characteristics are not present in the NMR-spectrum of echitinolide (87). The resistance of echitinolide and isoechitinolide to reduction by lithium aluminum hydride, mentioned above, is now attributed to steric hindrance in XLVIII and L. Hydrogenation of echitinolide gives a- and p-dihydroechitinolide, which are presumably the diastereoisomers formed on saturation of the 19,20 double bond. The same compounds can be obtained by hydrogenation of 0,O-diacetylechitnmine chloride, which gives an intermediate 0,O-diacetyldihydroechitamine, with carbomethoxy group intact ; further hydrogenation is then accompanied by lactonization, to give a mixture of a- and ,8-dihydroechitinolide (86). The Hofmann degradation of echitinolide methiodide is one aspect of the chemistry of echitamine which has not been explained entirely satisfactorily. The production of two methine bases, the amorphous Hofmann base A and the isomeric base B, mp 182"-187", was earlier used as an argument against Conroy's structure for echitamine. The structures of base A and base B, and the precise nature of the conversion

8. ALKALOIDS

OF

Alstonia

SPECIES

187

of base A into base B, are still obscure, but base B is tentatively formulated as L I (86). In accordance with its formulation as a carbinolamine, the Hofmann base B can be reduced to a deoxy base, pm 192'197', which gives a dihydroindole spectrum in neutral solution, but a benzenoid spectrum in the presence of strong acid, owing to protonation

LI

LII

of N, (86). The Hofmann base B is probably identical with Govindachari's echitinolide methine, mp 193' [originally named dihydroechitamine methine (75)], while the deoxy base, mp 192'-197', is probably the same as Govindachari's deoxyisoechitinolide methine, mp 200"-203' [originally named deoxyneonordihydroechitamine methine (75)and later deoxyneodihydroechitamine methine (82)l. As the final name implies, the hydrogenolysis in acid solution is accompanied by addition of the primary hydroxyl function to the double bond ; hence, deoxyisoechitinolide methine is probably L I I (87).

LIII

Hofmann degradation of base A or base B methiodide gives de-Nechitinolide-A, which contains both hydroxyl and imino groups (IRspectrum), and is formulated as LIII. De-N-echitinolide-A is readily converted by passage through alumina into an isomeric base, de-Nechitinolide-B, which can also be obtained by two Hofmann degradations starting from isoechitinolide (86). This alternative mode of preparation, together with the absence of hydroxyl absorption in the IR-region suggests that de-N-echitinolide-B is LIV.

188

J. E. SAXTON

Alloechitamine, obtained by reaction of echitamine chloride with potassium tert-butoxide, is produced by retroaldol loss of formaldehyde and generation of a second carbonyl group, and is formulated as LV

(86, 87). The low carbonyl stretching frequency (1689 om-1) observed in the IR-spectrum is the result of transannular interaction of the C-3 carbonyl group with Nb. Alloechitamine is presumably formed from echitamine chloride by a normal Hofmann elimination, which leads to the enol LVI; loss of formaldehyde then gives alloechitamine. The Hofmann stage would be expected to proceed smoothly, as the groups eliminated are situated trans diaxial to the boat-shaped ring.

J. COOMe

In aqueous solution in the presence of alkali, it is reasonable to expect the foregoing reaction to stop at the enol stage (LVI); the corresponding

8.

ALKALOIDS OF

Alstonia

SPECIES

189

keto form then contains a primary alcohol function suitably situated to allow hemiketal formation. This hemiketal (LVII) is one of the two structures attributed to echitamine base (87). The reason why ammonia does not precipitate echitamine base from echitamine chloride is thus

HI

---+

LVI

H"

attributed to the fact that ammonia is not a sufficiently strong base to effect the Hofmann elimination. The reconversion of LVII into echitamine chloride in the presence of acid may be envisaged as proceeding via the enol LVI ; isomerization of the latter to the related indolenine (LVIII)may also be involved, prior to closure of the five-membered ring. An alternative view has been expressed by Birch et al., who formulate echitamine base as the carbinolamine ether LIX (86). This could be formed by direct nucleophilic displacement of the positively charged

COOMe

a-@

HOH LIX

__f

LVIII

+ XLVI

Me

Nb by the primary alcohol function; alternatively, it could be formed in stages, via the indolenine LVIII or the related carbinolamine. Reformation of echitamine salts can also be explained as proceeding via the indolenine (LVIII), which is obtained by simple acid cleavage of LIX. Whichever structure of echitamine base (LVII or LIX) is correct, an

190

J. E. SAXTON

intermediate base can clearly be obtained in the reconversion of echitamine into its salts. This is in accordance with the observation that the hydrochloride of a base weaker than the echitamine quaternary hydroxide is the first product formed on titration of echitarnine with hydrochloric acid in 50% aqueous dioxane (82). Finally, the possible biogenesis of echitamine remains for discussion (88). This can readily be rationalized, using as an intermediate a base such as geissoschizine (LX), one of the hydrolytic fission products of the alkaloid geissospermine. Geissoschizine can clearly be obtained from tryptamine and dihydroxyphenylalanine or prephenic acid by a route which has many analogies in indole alkaloid chemistry. Dehydrogenation

MeOOC-4CHO H LX

161

4

M~OOC/C\CH-W-H LXI

I

XLVI

to give the C-3 to N, immonium ion, followed by hydrolysis and N b methylation, gives the keto base (LXI), in which C-7, normally susceptible to electrophilic attack, is now susceptible to attack by the nucleophilic C-16. The resulting cyclized base (LXII) can be converted by reduction and isomerization into the indolenine LXIII ( =LVIII), which can give the eserine system of echitamine (XLVI) by a wellestablished route (88).

8. ALKALOIDS OF A l S t O n i U

SPECIES

191

W. Echitarnidine Echitamidine, [a]F" -515") mp 244", was first isolated from A . scholaris and A . eongensis by Goodson (30), who assigned to it the molecular formula, CzoH2sN203. It behaved as a monoacidic base which apparently contained one methylimino group, but no methoxyl groups ; however, little reliance was placed on these results, as the alkaloid clearly behaved anomalously in these determinations. Many years later, its UV-spectrum (89) was observed to be virtually identical with that of akuammicine, indicating that the two alkaloids contain the same chromophore. This chromophore was identified by Aghoramurthy and Robinson in their proposal (LXIV) for the structure of akuammicine

=

& i C H .

booMe

LXIV Akmicine

Me

d

3

*

C COOMeH

. Me

LXV

(90). The presence of the same chromophore in echitamidine was also supported by the extremely high negative rotations of both alkaloids. In his study of akuammicine, Robinson commented that "if the molecular formula of echitamidine were C~oH24N203,it could be a hydroxydihydroakuammicine )' (90)) but no experimental data relating to echitamidine were reported. The anomalous behavior of echitamidine during methoxyl and methylimino determinations is probably the result of migration of a methyl group from oxygen to nitrogen; in fact, echitamidine contains one methoxyl group but no methylimino groups (91, 92). The apparent presence of one methylimino group determined by the Herzig-Meyer method has been confirmed, but its absence has been definitely proved from the NMR-evidence (92). The molecular formula for echitamidine was subsequently changed to C20H22N203, and the structure LXV was proposed. However, no analytical figures were given, hence the validity of structure LXV was clearly in doubt, particularly since it was derived from an erroneous structure (XXXII) for echitamine, on the assumption of a close biogenetical connection between the two alkaloids (91).

192

J. E. SAXTON

Very recently, the structure of echitamidine (LXVI) has been elucidated, mainly on the basis of the mass spectra of derivatives of echitamidine (92). Elementary analysis of echitamidine did not allow differentiation between the monohydrates of CzoHzzNz03, CzoHz4Nz03, or CzlHz&z03, and the analysis of the picrate was equally compatible with either of the last two formulas. The composition CzoHz&zO3 was

COOMe

LXVI Echitamidine

LXVII Condylocarpine

eventually determined from the mass spectrum which exhibited a molecular ion at m/e 340. Echitamidine contains one C-methyl group, but is inert to mild catalytic hydrogenation methods, and contains no olefinic protons (NMR-spectrum);structure LXV is therefore eliminated. The NMR-spectrum exhibited peaks at 6.11 7 (COOMe)and 1.32 7 (NH), a doublet centered at 8.84 T (three protons, due to CHrCH:), while the aromatic region resembled the same region in the spectrum of akuammicine ; no signals corresponding to a methylimino group were present. The peak of second highest m/e value in the mass spectrum of echitamidine occurs at m/e 322, which indicates loss of water from the parent molecule. Hence, echitamidine contains a hydroxyl group, a deduction which was confirmed by the preparation of 0-acetylechitamidine,whose mass spectrum indicated a molecular formula C22H2sN,04;its NMRspectrum resembled that of echitamidine, but it contained an additional peak at 7.91 T (three protons), owing to the acetate group. The evidence adduced thus far requires echitamidine to possess a pentacyclic ring system; no double bonds are present, other than those present in the chromophore Ar-N-C=C-COOMe. Assuming a close connection with akuammicine, echitamidine may thus be formulated as LXVI, in which the proved presence of CHs-CH< determines the position of the hydroxyl group. Alternatively, assuming a structural resemblance to condylocarpine (LXVII), it may be formulated as LXVIII. These two possibilities were distinguished, and the structure of echitamidine was established, by comparison of the mass spectra of

8. ALKALOIDS OF Alstonia SPECIES

193

*

tetrahydroakuammicine (LXIX ; R = H) and 2,16-dihydroechitamidine (LXIX; R = OH). The latter was prepared by the zinc-sulfuric acid reduction of echitamidine and, as expected, exhibited a normal

H

COOMe LXVIII

dihydroindole UV-spectrum, and had a molecular weight of 342 (mass spectrum). The fragmentation of tetrahydroakuammicine on electron impact proceeds with cleavage of ring C as indicated by the arrows in LXIX, followed by fission of the activated 5,6 bond to give two ions of m/e 130 and 196.An alternative mode of fission involves cleavage of the 20,21 bond in LXX to give a fragment of m/e 199. The usual /3-indolylethyl fragment at m/e 144 is also produced. The ions of m/e 130, 144, and 199, presumably due to identical fragments, are also

__f

H MeOOC

LXIX

Me

LXX

prominent in the mass spectrum of dihydroechitamidine ; however, the intense peak at m/e 196 in the spectrum of tetrahydroakuammicine is replaced by a similarly intense one at m/e 212,the difference of 16 mass units being due to the presence of an additional oxygen atom. Evidently this extra oxygen atom in dihydroechitamidine must be present in the non-nitrogenous fragment produced by cleavage of the 20,21 bond in LXX ; attachment at C-19 affords the most satisfactory explanation of the evidence. On the other hand, if echitamidine had the structure LXVIII, the mass spectrum of its dihydro derivative (LXXI) would have contained the peaks at m/e 130, 144, and 212 (fission of the 5,6 bond in LXXII), but instead of the ion at m/e 199,an ion of m/e 243

194

J. E. SAXTON

would have been produced following cleavage of the 20,21 bond, i.e., 44 mass units higher owing to the presence in hhe fragment of an additional hydroxyethyl group a t C-14. No such peak is observed in the

I

__f

\

N

H

\ COOMe

LXXI

N H

MeCH

I

OH MeOOC'

LXXII

mass spectrum of dihydroechitamidine, and hence formula LXVIII for echitamidine is eliminated. The available evidence is thus entirely consistent with the formulation of dihydroechitamidine as LXIX (R = OH), and with echitamidine as LXVI(92).

VII. Villalstonine Villalstonine, the major alkaloid of A . willosa and A . macrophylla, is amorphous, but when pure is stated to decompose at 235"-240" (33,35). Analysis of the base and several well-defined salts indicate that the molecular formula is C40H50N404 ;this is in agreement with the molecular weight determined cryosoopically in benzene (33). Villalstonine is a weak diacidic base, and forms a crystalline dimethiodide; since only one dissociation constant (pK, 5.8) has been recorded, it is probable that the molecule contains two centers of closely similar basic strength. The molecule also contains two methylimino groups and one methoxyl group. The latter is present in a carbomethoxy function, since alkaline hydrolysis gives an amphoteric substance, villalstonic acid, C39H48N404, and reduction with lithium aluminum hydride gives villalstoninol, CagH50N403, mp 274"-276", both of which are devoid of methoxyl groups (33, 35). One alcoholic hydroxyl group is present, since villalstonine can be characterized as an 0-acetate, C42H52N405, mp 228"-229". Oxidation with periodic acid is reported to give acetaldehyde ; hence, the molecule contains an ethylidene grouping. The UV-spectrum of villalstonine exhibits maxima at 231 and 286 mp, with a shoulder a t 250 mp and an inflection at 293 mp; this is consistent with the presence of indole and

8. ALKALOIDS OF Alstonia SPECIES

195

dihydroindole chromophores containing no substituents in the benzene rings. The only degradations that have so far been carried out with villalstonine are the drastic ones of potash fusion and selenium dehydrogenation. Alkali fusion yields products characteristic of the indole alkaloids, namely, 2-methylindole, indole-2-carboxylic acid, and a basic fraction which exhibits a typical p-carboline UV-spectrum ; a weak base also obtained in this degradation possesses a UV-spectrum closely resembling that of echitamidine. p-Carboline derivatives, so far unidentified, also appear to be the products of selenium dehydrogenation (35). At present, little can be said concerning the structure of villalstonine, except that it appears to be the first "dimeric" base in the Alstonia series. Its constituent chromophores are probably bz-unsubstituted indole and dihydroindole nuclei ; these may be contained in tetra- and hexahydro-p-carboline ring systems, but not even this can yet be regarded as established. The speculation that villalstonine may be regarded as composed of C19 and C21 units, the latter bearing the essential structural features of coronaridine (35), seems somewhat hasty on the basis of the published data. (See note added in proof, p. 201.)

VIII. Macralstonine Macralstonine, C44H54N405, crystallizes as colorless, rectangular rods, mp 293" (dec.), [ a ] D +27.5", and contains one methoxyl, a t least two C-methyl, and three, or possibly four, methylimino groups (33, 36). Its IR-spectrum discloses the presence of an imino or hydroxyl group, an alkoxyacrylic ester chromophore (peaks a t 1652 and 1628 em-I), and an o-disubstituted benzene nucleus. The formation of a derivative with dinitrophenylhydrazine indicates the presence of a potential carbonyl function. Macralstonine has also been reported to possess a marked hypotensive activity (36). The alkaloid macrophylline,l C45H54N405, mp 267"-268" (dec.), resembles macralstonine in both its mode of extraction and its physical properties (37),and was earlier assumed to be identical with it. Recently, however, the isolation of a base, mp 270"-272", reported to be identical with macralstonine, has been described from A . macrophylla; it is also Not to be confused with macrophylline, isolated from Senecio mcrophyllus (93), or with macrophylline-A and macrophyiline-B, isoiated, as their picrates, from Stryehnos nuzcrophylla (94).

196

J. E. SAXTON

stated that this base is not identical with macrophylline (36). Macrophylline (37) was reported to contain three methoxyl groups and three methylimino groups, and to be optically inactive (conditions not specified). The methoxyl analysis and the optical activity appear to be the only important differences in the properties of these two bases as far as is known a t present. However, the discrepancy in the reported rotations may not constitute an insuperable difficulty, since the rotation of macralstonine varies markedly with the solvent used ;in chloroform it is dextrorotatory, but in pyridine it is levorotatory (36).

IX. Macralstonidine

+

Macralstonidine, C41H50N403, mp 270" (dec.), [a]= 174.5O, pK, 6.6, (60% aqueous ethanol), the third dimeric alkaloid in this series, occurs to the extent of 0.04% in A . macrophylla and 0.06% in A . somersetemis (33). It is a diacidic tertiary base, which can be characterized as its dihydrochloride or dimethiodide. The molecular formula cannot yet be regarded as firmly established, although the NMR-proton integration (48.5 _+ 1) confirms the complexity of the molecule. Since macralstonidine exhibits a typical indole UV-spectrum both halves of the molecule must contain indole nuclei ; the hypsochromic shift observed in acid solution is reminiscent of alkaloids such as coronaridine, which contain the ibogaine ring system. Whether the benzene rings contain further substituents is uncertain; the presence of six aromatic protons, two of which are contained in an AB system (NMR) seems to be a clear indication for two such substituents situated in the 4,5 or 6,7 positions in one of the benzene rings. However, macralstonidine contains neither methoxyl nor phenolic hydroxyl groups, and although the UV-spectrum is consistent with the presence of methoxyl or hydroxyl groups para to the nitrogen atoms, it is not consistent with the presence of two adjacent oxygenated substituents attached to one of the benzene rings. Hence, if further substituents are present they are not the ones normally encountered in the indole series. Sharp (33) earlier suggested the presence of a methylenedioxy group on the basis of a positive Gaebel color reaction, but this has not yet been confirmed. An acetal function is probably present (NMR), but this may be involved in the union of the two halves of the molecule, since acid hydrolysis of macralstonidine gives a mixture of bases, which exhibit the same UV-spectrum as does the alkaloid. The NMR-spectrum and Kuhn-Roth determination disclose the presence of two C-methyl groups, one of which is contained in an ethylidene grouping. The NMR-spectrum also indicates that both

8. ALKALOIDS OB Alstonia SPECIES

197

indole nitrogen atoms, and one of the basic nitrogen atoms, bear methyl groups. The third oxygen atom is present in an alcoholic hydroxyl group, since macralstonidine gives an O-acetate (95). Selenium dehydrogenation of macralstonidine gives skatole, 2methylindole, and a basic fraction which is evidently a mixture. Its UV-spectrum closely resembles that of the mixture of bases (LXXIII and a base, C19H22N2, of unknown constitution) obtained by similar degradation of ibogamine (95). Hence, one (at least) of the indole nuclei in macralstonidine may be contained in a ring system of the ibogamine type. Et

Me

LXXIII

X. Alkaloid C The structure of alkaloid C, C20H22N203, mp 168"-169", [a]?' +200" (EtOH), from Alstonia muelleriann (38), has recently been established by X-ray crystallography. The UV-spectrum discloses the presence in alkaloid C of an oxindole chromophore ; the approximate coordinates in the unit cell of this large, rigid, planar grouping were deduced from the three-dimensional Patterson function, and the remainder of the atoms were then located by means of multiple superposition techniques (96). Refinement was carried out by the least-squares method in the usual way. The incorporation of a heavy atom into the molecule was thus rendered unnecessary. The structure deduced has the relative configuration shown in LXXIV; the absolute configuration has not been established, although it may provisionally be assumed to have the same configuration at C- 15 as the heteroyohimbine alkaloids, as depicted in LXXIV. Biogenetically, this interesting variant on the heteroyohimbine theme could originate from tryptamine and either dihydroxyphenylalanine (Robinson) or prephenic acid (LXXV) (Wenkert); as example, a possible route from prephenic acid via the seco-prephenate-formaldehydeunit (LXXVI) (97) is outlined below. A vital feature in this proposal (and

198

S . E. SAXTON

also in a possible biogenetic route from tryptamine and dihydroxyphenylalanine) is the fate of the one-carbon bridge (C-Zl), which is here postulated to provide one of the atoms of ring E, instead of appearing, as is usually the case, as the connecting linkage between C-20 and Nb. COOH I

co

COOH I CO

CHzOH

HOOC \

CHz

CHzOH

+ ‘cHz&

““.&OH

--+

HOOC

COOH LXXV

COOH

COOH

~;r””’”Hz

, ‘cH200

I

co

CO I‘CH2fyo C H z O H

CHzOH

lletroaldol

HOOC

OH

cleavage

HOOC

CH3 CHO

+

N ’ H

LXXVI

1. Oxidation a t CP 2. Mannich cycliration 3. -coz

LXXIV Alkaloid C

__f

8. ALKALOIDS OF Alstonia SPECIES

199

REFERENCES 1. C. Palm, Vierteljahresschrijt Prakt. Pharm. 12, 161 (1863). 2. F. von Miiller and L. Rummel, Pharm. J . Trans. 9, 441, 1059 (1879); J . Chem. SOC. 35, 31 (1879). 3. L. Oberlin and F. R. Schlagdenhauffen, Phurm. J . Trans. 9, 1059 (1879). 4. 0. Hesse, Ann. Szlppl. 4, p. 40 (1865). 5. 0. Hesse, Ber. 11, 2234 (1878). 6. 0. Hesse, Ann. 205, 360 (1880). 7. T. M. Sharp,J. Chem. SOC.p. 287 (1934). 8. W. L. Hawkins and R. C. Elderfield, J . Org. Chem. 7, 573 (1942). 9. G. H. Svoboda, J . Am. Pharm. Assoc. 46, 508 (1957). 10. W. D. Crow and Y. M. Greet, AustraZianJ. Chem. 8, 461 (1955). 11. R. G. Curtis, G. J. Handley, and T. C. Somers, Chem. Ind. (London), p. 1598 (1955), 12. E. Schlittler, H. Schwarz, and F. Bader, Helv. Chim. Acta 35, 271 (1952). 13. B. U. Vergara, J . A m . Chem. SOC.77, 1864 (1955). 14. K. Mezey and B. Uribe, Anales SOC.Biol. Bogota 6, 127 (1954); Chem. Abstr. 48, 12373 (1954). 15. K. Mezey and B. Uribe, Arch. Intern. Pharmacodyn. 98, 273 (1954); Chem. Abstr. 49, 1277 (1955). 16. M. Shimizu and F. Uchimaru, Chem. Pharm. BUZZ.(Tokyo)6,324 (1958); 7,713 (1959). 16a. P. P. Pillay and T. N. S. Kumari, J . Sci. Ind. Res. ( I n d i a )20B, 458 (1961). 17. S. C. Pakrashi, C. Djerassi, R. Wasicky, and N. Neuss, J . Am. Chem. SOC.77, 6687 (1955). 18. F. A. Hochstein, J . Am. Chem. SOC.77, 5744 (1955). 18a. I. Inagaki, M. Ogawa, H. Minamidani, and T. Kato, Nagoya Shiritsu Daigaku Yakugabuku Kiyo 8, 34 (1960); Chem. Abstr. 55, 10802 (1961). 19. M. M. Janot, J. Le Men, and Y. Gabhai, Ann. Pharm. Franc. 15, 474 (1957); Chem. Abstr. 52, 5745 (1958). 19a. E. Schlitter and J . Hohl, Hetv. Chim. Acta 35, 29 (1952). 20. E. von Gorup-Besanez, Ann. 176, 88 (1875). 21. 0. Hesse, Ann. 176, 326 (1875). 22. J. Jobst and 0. Hesse, Ann. 178, 49 (1875). 23. 0. Hesse, Ber. 11, 1546 (1878). 24. 0. Hesse, Ann. 203, 144 (1880). 25. E. Harnack, Arch. ExptZ. PathoZ. Pharmakol. 7, 126 (1877); Jahresber. p. 935 (1877). 26. E. Harnack, Ber. 11, 2004 (1878). 27. E. Harnack, Ber. 13, 1648 (1880). 28. 0. Hesse, Ber. 13, 1841 (1880). 29. R. F. Bacon, PhiZippineJ. Sci. 1, 1007 (1906); Chem. Abstr. 1, 1162 (1907). 30. J. A. Goodson, J. Chem. SOC.p. 2626 (1932). 31. 0. Hesse, Ann. 203, 170 (1880). 32. J. A. Goodson and T. A. Henry, J . Chem. SOC.127, 1640 (1925). 33. T. M. Sharp, J . Chem. SOC.p. 1227 (1934). 34. A. Chatterjee and S. Ghosal, Sci. Cult. (Calcutta) 26, 238 (1960); Chem. Abstr. 55, 11450 (1961). 35. A. Chatterjee, S. K. Talapatra, and N. Adityachaudhury, Chem. I d . ( L o n d o n ) p. 667 (1961). 36. S. K. Talapatra and N. A. Chaudhury, Sci. Cult. (Calcutta)24, 243 (1958); Chem. Abstr. 53, 11755 (1959).

200

J. E. SAXTON

37. F. Manas-Santos and A. C. Santos, Nut. Appl. Sci. Bull. (Univ. Philippines) 5, 133 (1936); Chem. Abstr. 31, 6243 (1937). 38. R. E. Gilman, Dissertation Abstr. 20, 1578 (1959). 3%. R. C. Elderfield, Am. Scientist 48, 193 (1960). 39. J. A. Goodson, T. A. Henry, and J. W. S. MacFie, Biochem. J . 24, 874 (1930). 40. Raymond-Hamet, Compt. Rend. SOC.Biol. 116, 1022 (1934); 135, 1565 (1941). 41. Rrtymond-Hamet, Compt. Rend. Acad. Sci. 209, 1013 (1939). 42. P. Keogh and F. H. Shaw, AustralianJ. Exptl. Biol. Med. Sci. 21, 183 (1943). 43. B. Mukerji, B. K. Ghosh, and L. B. Siddons, Indian Med. Gaz. 77, 723 (1942); Chem. Abstr. 37, 3877 (1943). 44. B. M. Das Gupta, L. B. Siddons, and Chakravarti, Indian Med. Baz. 79, 408 (1944). 45. B. Mukerji, Nature 158, 170 (1946). 46. K. G. Wakim and K. K. Chen, J . Pharmacol. Exptl. Therap. 90, 57 (1947). 47. S. T. Yang, J. Am. P h a m . Assoc. 37, 458 (1948). 48. T. M. Sharp, J. Chem. SOC.p. 1353 (1938). 49. N. J. Leonard and R. C. Elderfield, J . Org. Chem. 7 , 556 (1942). 50. P. Karrer and P. Enslin, Hetv. Chim. Act5 32, 1390 (1949). 51. P. Karrer and P. Enslin, Helv. Chim. Acta 33, 100 (1950). 52. T. B. Lee and G. A. Swan, J . Chem. Soc. p. 771 (1956);see also Robbins and Sir Robert Robinson, ibid., not0 on p. 772. 53. E. Schlittler and H. Schwarz, Helv. Chim. Acta 33, 1463 (1950). 54. R. C. Elderfield and A. P. Gray, J . Org. Chem. 16, 506 (1951). 55. F. E. Bader, Helv. Chim. Acta 36, 215 (1953). 56. E. Wenkert and D. K. Roychaudhuri, J. Am. Chem. SOC.78,6417 (1956). 57. E. Wenkert and D. K. Roychaudhuri, J . Am. Chem. SOC.80,1613 (1958). 58. E. Wenkert and D. K. Roychaudhuri, J . Am. Chem. SOC.79, 1519 (1957). 59. N. Neuss and H. E. Boaz, J. Org. Chem. 22, 1001 (1957). 60. E. E. van Tamelen and C. Placeway, J. Am. Chem. SOC.83, 2594 (1961). 61. E. Wenkert, B. Wickberg, and C. L. Leicht, J. Am. Chem. SOC.83, 5037 (1961). 62. E. Wenkert and N. V. Bringi,J. Am. Chem. SOC.81, 1474 (1959). 63. M. Shamma and J. B. Moss, J. Am. Chem. SOC.84, 1739 (1962). 64. M. Shamma and J. B. Moss, J. Am. Chem. SOC.83,5038 (1961). 65. R. C. Elderfield and S. L. Wythe, J . Org. Chem. 19, 683 (1954). 66. R. C. Elderfield and S. L. Wythe, J. Org. Chem. 19, 593 (1954). 67. R. C. Elderfield and 0. L. McCurdy, J . Org. Chem. 21, 295 (1956). 68. R. C. Elderfield and B. A. Fischer, J. Org. Chem. 23, 332, 949 (1958). 68a. Y. Ban and M. Seo, J . Org. Chem. 27,3380 (1962). 69. A. van Camp and H. A. Rose, J. Am. Pham. Assoc. 46, 509 (1957). 70. H. Boaz, R. C. Elderfield, and E. Schenker, J. Am. Pharm. Assoc. 46, 510 (1957). 71. X. Monseur and L. van Bever, J . Pharm. Belg. 10, 93 (1955). 72. T. R. Govindachari and S. Rajappa, PTOC. Chem. SOC.p. 134 (1959). 72a. T. R. Govindachari and S. Rajappa, Proc. Indian Acad. Sci.Sect. A . 51, 319 (1960). 73. H. F. Hodson and G. F. Smith,J. Chem. SOC.p. 1877 (1957). 74. T. R. Govindachari and S. Rajappa, Chem. I d . (London)p. 1154 (1959). 75. T. R. Govindachari and S. Rajappa, Chem. I d . (London)p. 1549 (1959). 76. A. J. Birch, H. F. Hodson, and G. F. Smith, Proc. Chem. SOC.p. 224 (1959). 77. S. Ghosal and S. G. Majumdar, Chem. Ind. (London)p. 19 (1960). 78. A. Chatterjee, S. Ghosal, and S. G. Majumdar, Chem. Ind. (London) p. 265 (1960). 79. A. Chatterjee and S. Ghosal, Natururissenschaften 47, 234 (1960). 80. A. Chatterjee and S. Ghosal, Chem. I d . (London)p. 176 (1961).

8. ALKALOIDS OF Alstonia SPECIES

201

81. A. J. Birch, H. F. Hodson, B. Moore, H. Potts, and G. F. Smith, Tetrahedron Letters No. 19, 36 (1960). 82. H. Conroy, R. Bernasconi, P. R. Brook, R. Ikan, R. Kurtz, and K. W. Robinson, Tetrahedron Letters No. 6, 1 (1960). 83. D. Chakravarti, R. N. Chakravarti, R. Chose, and Sir Robert Robinson, Tetrahedron Letters No. 10, 10; No. 11, 25; No. 12, 33 (1960). 84. J. A. Hamilton, T. A. Hamor, J. M, Robertson, and G. A. Sim, Proc. Chem. Soc. p. 63 (1961);J. Chem. Soc. p. 5061 (1962). 85. H. Manohar and S. Ramaseshan, Current Sci. (India)30, 5 (1961); Tetrahedron Letters p. 814 (1961); 2. Krist. 117, 273 (1962). Chem. Soc. p. 62 (1961). 86. A. J. Birch, H. F. Hodson, B. Moore, and G. F. Smith, PTOC. 87. T. R. Govindachari and S. Rajappa, Tetrahedron 15, 132 (1961). 88. G. F. Smith, Chem. I n d . (London)p. 1120 (1961). 89. Raymond-Hamet, Compt. Rend. Acad. Sci. 233, 560 (1951). 90. K. Aghoramurthy and Sir Robert Robinson, Tetrahedron 1, 172 (1958). 91. A. Chatterjee and S. Ghosal, Naturwissenschaften 48, 219 (1961). 92. C. Djerassi, Y. Nakagawa, H. Budzikiewicz, J. M. Wilson, J. Le Men, J. Poisson, and M. M. Janot, Tetrahedron Letters p. 653 (1962). 92a. B. S. Joshi and W. I. Taylor, unpublished work, mentioned in footnote 8 in ref. 92d. 92b. M. Hesse, W. von Philipsborn, D. Schumann, G. Spiteller, 11. Spiteller-Friedmann, W. I. Taylor, H. Schmid, and P. Karrer, Helv. Chirn. Acta 47, 878 (1964). 92c. W. G. Kump and H. Schmid, Helw. Chim. Acta 44, 1503 (1961). 92d. M. F. Bartlett, R. Sklar, A. F. Smith, and W. I. Taylor,J. Org. Chem. 28,2197 (1963). 93. A. Danilova, L. Utkin, and P. Massagetov, Zh. Obshch. Khim. 25, 831 (1955);J . Gen. Chem. C'SSR (Eng. Transl.) 25, 797 (1955);A. V. Danilova and L. Utkin, Zh. Obshch. Khim. 30, 345 (1960); Chem. Abstr. 54, 22698 (1960). 94. M. A. Iorio, 0. Corvillon, H. Magalhzes Alves, and G. B. Marini-Bettolo, Gazz. Chim. Ital. 86, 923 (1956). 95. S. Ghosal, G. Ganguli, and A. Chatterjee, Sci. Cult. (Calcutta)27, 406 (1961); Chem. Abstr. 56, 3521 (1962). 96. C. E. Nordman and K. Nakatsu, J . Am. Chem. SOC.85, 353 (1963). 97. E. Wenkert, J . Am. Chem. Soc. 84, 98 (1962). 98. T. R. Govindachari, N. Viswanathan, B. R. Pai, and T. S. Savitri, Tetrahedron Letters p. 901 (1964). 40, 1043 (1963). 99. A. B. Ray and A. Chatterjee, J . Indian Chem. SOC.

NOTEADDEDIN PROOF Section VII, Villalstonine (see p. 195):

According to a recent report on some unpublished work by Joshi and Taylor (92a), the acid hydrolysis of villalstonine affords pleiocarpamine (LXXIIa) (92b), one of the constituent bases of Pleiocarpa mutica Benth. (92c) and Hunteria eburnea Pichon (92d).

LXXIIa

J. E. SAXTON

202

Addendum. Venenatine Govindachari et al. have recently recorded the extraction of reserpine, venenatine (LXXVII), and 3-isovenenatine from the bark of Alstonia venenata R. Br. (98). 3-Isovenenatine may be identical with alstovenine, isolated earlier from the same species by Ray and Chatterjee (99), together with a second, unnamed base, m.p. 192". OMe

M e O O C V OH LXXVII Venenatine

----CHAPTER

9-

THE IBOGA AND VOACANGA ALKALOIDS W. I. TAYLOR Research Department, C I B A Pharmaceutical C m p a n y , Division of C l B A Corporation, Summit, New Jersey

I. The Iboga Alkaloids.. ............................................... A. The Structures of Ibogaine and Iboxygaine.. ......................... B. Ibogamine and Tabernanthine. ..................................... C. 18-Carbomethoxy Alkaloids ....................................... D. Voacryptine ..................................................... E. Catharanthine, Cleavamine, and Velbanamine ........................ F. Mass Spectra of Iboga Alkaloids. ................................... G. Other Alkaloids.. ................................................ 11. The Voacanga Alkaloids.. ............................................ A. Voachalotine .................................................... B. Vobasine, Dregamine, Tabernaemontanine, and Callichiline............. C. Bisindoles.. .....................................................

203 206 213 213 217 218 221 223 225 225 228 229

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

231

References ..........................................................

233

111. Miscellaneous.

I. The Iboga Alkaloids The iboga alkaloids (Table I) presently number twelve, if their oxidation products are excluded, all from apocynaceous plants of the genera Conopharyngia (Plumeria), Ervatamia, Gabunea, Stemmadenia, Tabernaemontana, Voacanga, Vinca (Lochnera,Catharanthus),and Tabernanthe. It was from the last genus that the parent pentacyclic heterocycle, ibogamine, was first obtained. The structures of these compounds depend entirely upon their interrelationships with ibogaine, whose structure was derived by degradation (illustrated schematically in Charts I to IV) and X-ray analysis. The absolute stereochemistry has not been rigorously determined, and none of the bases, at the time of writing, had been synthesized. The alkaloids can be conveniently grouped as shown in Table I, and it should be noted that the trivial names currently used obscure their similarities. Many of the alkaloids suffer facile autoxidation to yield hydroperoxyand hydroxyindolenines, whose further degradation products are 4hydroxyquinolines and pseudoindoxyls (Table I). Therefore, the 203

204

-4

W. I. TAYLOR

x x

x

x

x

x

3

X

0, x

x

B x

B. Hydroxyethyl side chain; R3 = OH Iboxygaine 234

B. Hydroxyethyl side chain; R3 = OH Me0

H n(5)

- 5" (CHCls)

112-114 or 166-167

(Kimvuline) 231-232

+ 4" (CHC13)

D . Oxidation and rearrangement products of parent bases

Voacristine (voacangarine) M e 0 - 25' (CHC13)

H o(15, 16)

C. Acetyl side chain; R3 = 0 --

9-Hydroxy-9H-ibogam ins + 82" (alc.)

H

168-172

Demethoxyiboluteine (ibogamine-ybindoxyl)

H

141

9-Hydroxy-9N-ibogaine + 74" (EtOH)

Me0

H n(6)

123-124

Me0

H n(7)

142

Iboluteine (ibogaine-$-indoxyl) - 114' (CHC13) Iboquine (ibogaine-4-quinolinol)

Me0

H

284-285

H n(6) 175-176

H

Voacryptine 25" (CHC13)

+

Me0

H

o(17)

n(6)

n(7)

"Sources: a, Callichilia barteri Stapf; b, C. sterwsepala Stapf; c, C. subsessilis Stapf; d, Corwpharyngia duriasima Stapf; e, Ervatamh coronaria Stapf; f, E . divarkata Burkill; g, Gabunia eglandulosa Stapf; h, Stemmadenia donnell-smithii R. E. Woodson; i, S. galeottiana Mierg; j, Tabernaemontana australis Muell. tlrg.; k, T . coronaria Wad.; 1, T . oppositifolia Urb.; m, T . psychot&jolia H.B. and K.; n, Tabevna,n$h iboga Baill.; 0, Voacunga ajricuna Stapf ex S. Elliot; p, V . bracteata Stapf; q, V . chalotiana Pierre ex Stapf; r, V . dregei E. Mey.; s, v. echweinfurthii Stapf; t, V . thouarsii Roem. et Schult. var. obtusa Pichon; u, Vinca rosea Lhn. Parenthetical numbers refer to reference list.

206

W. I. TAYLOR

isolation of these products from the plant cannot by itself be taken as proof of their natural occurrence. This situation is similar to that which exists for several of the tertiary bases obtained from Hunteria eb,urnea Pichon and possibly for some of the dimeric curare alkaloids derived from the Wieland-Gumlich aldehyde. A. THE STRUCTURES OF IBOGAINE AND IBOXYGAINE Although the isolation of the principal alkaloid, ibogaine, of Tabernanth,e iboga was described a t the turn of the present century ( l ) ,it was not until the early 1950’s that serious work on its structure was begun. It had been shown to contain a methoxy group and, by means of color reactions (19) and by measurement of its UV-spectrum, to be an indole (SO), but it was not recognized t o be a methoxyindole until permanganate oxidation was found to afford 5-methoxy-N-oxalylanthranilic acid (21).

I1

IV

/

VI Alloibogaine

V

CHARTI. KOH arid S e degradation of ibogaine.

9.

207

THE IBOGA AND vOUCU?lgU ALKALOIDS

A complete structure (22) for ibogaine (I)was derivable from a consideration of its potash fusion products (23, 24), 1,2-dimethyl-3-ethyl-5(111)on the one hand, hydroxyindole (11)and 3-ethyl-5-methylpyridine and its selenium dehydrogenation products (22, 25), the cyclic 2,2'aminophenylindole (IV), mp 208", and the indolo[3,2-c]quinoline (V), mp 17S0, on the other hand. I n each case all the carbons, both nitrogens, and the oxygen were accounted for. Alloibogaine (VI), amorphous, oxalate, mp 2003 [readily prepared from ibogaine by a more conventional route (as)] was an intermediate in the potash fusion (23, 24). The structures of the degradation products were confirmed by synthetic studies, and the routes which led to successful synthesis of the selenium degradation products (IV and V; Me0 = H) of ibogamine are given in Chart V. A third product (25) of the selenium dehydrogenation of ibogaine (ibogamine gave a similar product), characterized as its picrate, mp 165-167", may be V I I (R=MeO), but it has not been further examined (cf. the analogous dehydrogenation product of cinchonamine, p. 239).

HOOC F

/

\ -

%

'

R

VII

L

w N

HOOC

VIII

I n an attempt to confirm the formula for ibogaine derived from the foregoing results, rings A and B of ibogaine lactam (XI), mp 221°, [a],, - 16" (EtOH), were removed oxidatively to furnish a dibasic acid, believed to be VIII, in an amount unsuitable for further degradation (25). The chemistry of the (auto)oxidation products of ibogaine can now be considered (Chart 11).Ibogaine in air with or without catalytic assistance ( 2 5 , 26) was readily convertible into the indolenines, 9-hydroperoxy-9Hibogaine1 (IX), mp 218O-22Oo, and 9-hydroxy-9H-ibogaine1 (X). This is a well-understood process which requires no comment except that in 1 The names are derived systematically from the hypothetical tautomer of ibogaine, 9H-ibogaine:

208

-7

X

d ___*

L

0

W. I. TAYLOR

\

p

0

(:::: \

H

9.

THE IBOGA AND

Voacanga ALKALOIDS

209

this case it is unusually facile as compared with other indoles.The derivation of iboquine (XII) and iboluteine (XIII)from these indolenines was in agreement with the work on model compounds. With other oxidizing agents, especially chromic oxide in pyridine, ibogaine gave ibogaine lactam (XI) along with 8,19-dioxoibogaino (XIV), mp 318"320") [.IU - 49" (EtOH) (25).Thefacileformationof a lactam, in contrast to the behavior of other indole alkaloids, e.g., yohimbine, was a consequence of a suitable stereochemistry about the nitrogen. The formation of the 8-0x0 compound may be a result of an initial attack of the reagent a t C-9 followed by rearrangement to the C-8 hydroxy derivative which undergoes further oxidation. In the case of the chromic acid oxidation of yohimbine, an alternate pathway is preferred (2?), which leads to 433 14-yohimbine (28). The chromic acid oxidation of iboquine and iboluteine gave the analogous lactams XV and XVI, respectively ( 2 5 ) . I n none of the work were any ?-ox0 compounds detected. Lithium aluminum hydride reduction of ibogaine lactam regenerated ibogaine as expected (25). The same reduction of iboluteine, which also was claimed (26) to give back ibogaine, in actual fact produced dihydrodeoxyiboluteine, mp ?8"-?9" (29, 30). If iboluteine was reduced by sodium borohydride, two dihydroiboluteines (XVII) were obtained : A, mp 150"-152" [.IU + 28" (EtOH); and B, mp 184'-186", [.ID - 63" (EtOH). Both A and B upon treatment with acid afforded the inverted ibogaine derivative (XVIII), picrate, mp 201'-202" (30). This was the expected result, since in a Wagner-Meerwein rearrangement the more nucleophilic group generally migrates. The availability of iboluteine (XIII) made possible work which provided a second proof of the structure of ibogaine (Chart 111), along with its unequivocal correlation with ibogamine and tabernanthine (25). When 0-tosyliboluteine oxime (XIX) was refluxed in pyridine, it underwent an abnormal Beckmann rearrangement t o provide 4-methoxyanthranilonitrile (XX) and the ring C contracted ketone (XXI). The same ketone and the corresponding anthranilonitriles were obtained analogously from the pseudoindoxyls of ibogamine and tabernanthine. The amino ketone (XXI) subjected to a von Braun cyanogen bromide reaction (25) gave the N-cyano ketone (XXII) which, after reduction (LiAlH4) to X X I I I and dehydrogenation (Se), gave 6-methyl-8-ethylquinoline (XXIV). The rotatory dispersion of the ketone (XXII) was compared with a number of decalones and found to resemble most closely that of l-cis-9-methyl-4-decalone, whose absolute configuration is known. Assuming that a decahydroquinolone can be equated with a

210

7

-\

C

Y

W. I. TAYLOR

5+ x

x P.

N

x x

I4 Y

3

l-l

V

9.

THE IBOGA AND

Voacanga ALKALOIDS

21 1

decalone and that the angular methyl group does not make much difference, then the foregoing comparison is valid. This means that the stereochemistry of ibogaine is that pictured in I, with the exception of the configuration of the ethyl group. The configuration of the ethyl group, as well as an independent proof of the thus developed structure of ibogaine, came from a three-dimensional X-ray analysis of its hydrobromide (31). On the chemical side, the stereochemistry of the ethyl group of ibogaine was a consequence of the properties of 20-hydroxyibogaine [iboxygaine, kimvuline (XXV)]. Iboxygaine gave a positive iodoform reaction, formed an amorphous ketone, and furnished acetic acid upon chromic acid oxidation (5). Borohydride reduction of the ketone gave an amorphous alcohol, also obtainable by Wolff-Kishner reduction, but the relationship of this product to the starting alcohol is not known (32).

xxv Iboxygaine

XXVI

\

J XXVII

Ibogaine (I)

XXVIII d 2o-Ibogaine

Upon treatment with tosyl chloride in pyridine, a quaternary salt (XXVI) was produced ( 5 ) . The quaternary base was reconverted into iboxygaine by aqueous sodium hydroxide (33). With sodium ethylate, on the other hand, 20-ethoxyibogaice [XXVII, iboxygaine ethyl ether, y-isomer of Stauffacher and Seebeck (IS)], mp 194"-196", [aID -16" (EtOH), was the major product (33). In both the preceding reactions, Azo-ibogaine (XXVIII), mp 155"-156", [aID - 80" (EtOH), was also

212

W. I. TAYLOR

isolated, hydrogenation of which gave ibogaine. If the crude Hofmann products were reduced before work-up, ibogaine along with an unidentified isomer, mp 119"-120', [a],) + 95' (EtOH),was obtained (33). Treatment of XXVI under Emde conditions (Na/E:tOH) was reported to yield ibogaine, a p-isomer, mp 185'-187', [aID - 114' (EtOH), and 20-ethoxyibogaine (16), but a repetition of this experiment by another group gave only the last two substances (33). By far the most efficient

BrO

I Ibogaine

H

K31nO4 t-

HOOC,/

I

Et

HOOC,

xxx

I Et

XXIX N - Cyanoapoibogaine

HOOC,)

XXXI CHARTIV. Von Braun degradation of ibogaine.

I Et

agent for the conversion of XXVI t o ibogaine was lithium aluminum hydride (34). A third degradation (Chart IV) of ibogaine, not carried as far as a known compound, was complementary to the above results and is of

9. THE

IBOGA AND

Voacanga ALKALOIDS

213

interest because it is a conversion of the indoloazepine rings A, B, C into a derivative of a /3-carboline (25). The first step leading t o N-cyanoapoibogairie (XXIX) may be a Hofmann-type elimination. The subsequent reactions require little comment, except to point out that either XXX or XXXI may be suitable for correlation with substances of known absolute stereochemistry. B. IBOGAMINE AND TABERNANTHINE Both ibogamine and tabernanthine (formulas, Table I) have been related to ibogaine via fission of their respective pseudoindoxyls in the manner indicated in Chart I11 (XIX X X + XXI). In general, their chemical reactions were very similar to those of ibogaine, although tabernanthine, during the preparation of its pseudoindoxyl, gave rise t o an oxindole (XXXII), mp 191°-1970, a class of compound which was not picked up in the more exhaustive study of ibogaine (25). Both ibogamine lactam, mp 329"-331", and tabernanthine lactam, mp 312"-315", were also prepared (25).

M

d XXXII

XXXIII

Selenium dehydrogenation (25) of ibogainine gave products completely analogous t o ibogaine, viz., IV, V, and VII (Me0 = H in all three), the two major products being synthesized according to the procedures outlined in Chart V (35). Zinc dust distillation (36) of ibogamine yielded 3-methyl-5-ethylpyridinealong with an unexpected compound, 3methylcarbazole (XXXIII). The formation of the latter, although it was of no value in the elucidation of the structure of ibogamine, may be a characteristic pyrolytic product for this heterocyclic system under such reaction conditions. c . 18-CARBOMETHOXP ALKALOIDS The structures of these compounds (see Table I)have been established almost entirely by decarbomethoxylation to the parent heterocycle.

HooCuNO2

Me

I

NO2

oA(yt Me

I

Me I

Me

Me

I

Me

NH2

__f

I

Me

I

Me

CHsCHO

Me

I

Me

I

Me

CHARTV. Synthesis of the selenium degradation products of ibogaminc.

I Me

9.

215

THE IBOGA AND VOaCUn gU ALKALOIDS

The properties of voacangine (XXXIV, 18-carbomethoxy-12methoxyibogamine) are characteristic of this group. Voacangine is about 2 pK, units less basic than ibogaine. Voacangic acid was readily decarboxylated thermally or by reflux in mineral acid (37), and resembled indole-2-acetic acid (38)in this property. The ester also suffered decarbomethoxylation when refluxed with suitable amines such a8 hydrazine or

Heat or H e

COOR XXXIV Voacangine

1

RNHa

Heat

Ibogaine (I)

__f

I

I ’0,

LiAlH4

w H

Heat 4

Meo CHzOH

’

H \ q

‘

0-CHZ

xxxv Voacanginol; R = H

ethanolamine ( 14).Voacanginol (LiAlH4 reduction product of voacangine), mp 203”, [a]= + 38O, (CHC13))eliminated formaldehyde above its melting point to furnish ibogaine (39). The ease with which these eliminations proceed is facilitated by the almost planar geometry of the

216

W. I. TAYLOR

aromatic rings, c18, and its substituent. The reactions are believed to proceed by the illustrated mechanisms, which invoIve the intermediacy of the 9H-tautomer, the indolenine ( 2 5 , 14, 39). Voacanginol could be tosylated without quaternizatioii occurring (cf. iboxygaine, Section I, A), and the resulting sulfonate with lithium aluminum hydride afforded 18-methylibogaine, mp 18Y-19Oo (32). The site of the carboxyl in voacangine was confirmed by treating 20-hydroxyvoacanginol (XXXV; R = OH) with acetone containing hydrogen chloride to yield the acetonyl derivative, which was characterized as its 0-acetate (XXXVI) (16). A similar reaction, starting from d3-coronaridine (see Table I),led analogously to XXXVII (40).

With palladium charcoal, voacangine gave the expected 3-methyl-5ethylpyridine along with an as yet unidentified 5-methoxyindole (or indolenine), CllH13N02 ( Z ) , mp 80'-81'. I n the same paper, isovoacangine (see Table I) was shown to furnish, along with the pyridine, a product, mp 81°-820, which was assumed to be 6-methoxy-2-methyl-3ethylindole (11). When selenium was used, ibogaine was the only isolable product. Ozonolysis of voacangine has given a yellow compound,

I

COOMe

XXXVIII

I

COOMe XXXIX

mp 186O-187', [ Q ] ~+ 136' (CHC13),for which the structure XXXVIII was suggested (11).This formula might be in agreement with the recorded UV-spectrum (A,, 270, 386 mp), but its stability to base is more difficult to understand, since it furnished an acid reconvertible into the ester with

9.

THE IBOCA AND

Voacanga ALKALOIDS

217

diazomethane rather than a carbomethoxyiboquine. Voacanginol gave an analogous product (XXXVIII ; COOMe = CH20H). Voacangine has been subjected to the action of cyanogen bromide (11) and has yielded three compounds: the major product, an indole, C23H28N&Br, mp 203"-204', [.ID - 92' ; minor products, an indole, C23H28N303Br, mp 238"-240', [U]D +46", and an indolenine (?), C23H27N303, mp 175'-176'. [a]D -34'. The first two compounds were probably normal von Braun products, one of which may be convertible into N-cyanoapoibogaine (XXIX).Theformation of the indolenine would be a consequence of the nucleophilic reactivity of the indole a t C-9 and another example (quebrachamine and cyanogen bromide) is reported in the same paper. The catalytic oxidation of voacangine has been studied (32). After reduction of the "hydroperoxide," a compound, mp 249', [.ID -45' (CHC13),was isolated that was suggested to be 18-carbomethoxyibogaine lactam but, in the absence of proof of its nonbasic character, UVabsorption data, or decomposition into ibogaine lactam, its reformulation as an oxindole [cf. formation of tabernanthine oxindole (XXIII)] is a possibility.

D. VOACRYPTINE Voacryptine (see Table I) was recognized to be an oxovoacangine on the basis of its impirical formula and physical properties (41). The carbonyl group was placed on the side chain, since only acetic acid was produced after chromic acid oxidation and by its positive iodoform reaction. Voacryptine formed an oxime, mp 114°-1160, and under WolffKishner conditions it generated ibogaine. Because of the basic conditions of the last experiment, it was not possible to deduce the configuration of the acetyl side chain. However, reduction of voacryptine with potassium borohydride gave a mixture of diastereoisomeric dihydro compounds which was acetylated and resulted in the isolation of voacristine 0-acetate (XXXIX), mp 191'-193", ["ID -27" (CHC13), and a second compound, mp 180', which, it was suggested, may have been slightly impure 20-epivoacristine-0-acetate (41). Therefore, unless it can be shown that an epimerization a t C-4 preceded the borohydride reduction of voacryptine, its stereochemistry can be considered as established. It can also be deduced that the /3-configuration of the ethyl is preferred over the a-configuration; some support for this can be adduced by a conformational argument, based on the skewed nature of the isoquinuclidine moiety (cf. catharanthine and the related quinuclidine system in ajmaline, Chapter 2 2 ) .

218

W . I. TAYLOR

E. CATHARANTHINE, CLEAVAMINE, AND VELBANAMINE The occurrence of catharanthine (XL) is presently confined to Vinca L., and i t is the only known iboga alkaloid to contain an olefinic double bond (40). A derivative of this compound makes up the indolic portion of the clinically useful antileukemic drug, vincaleukoblastine rosea

LOOMe XL

(XL ; R = CH3). HyL-ogenation (12) of catharanthine led to only one isomer, dihydrocatharanthine (XLII, 18-carbomethoxy-4-epi-ibogamine), mp 63'-65', [aID + 35' (CHC13),the hydrogen coming in on the less hindered side of the isoquinuclidine residue (40). The alkaloid behaved similarly to voacangine, since on the one hand its dihydro derivative was readily decarbomethoxylated to 4-epi+ 86" (hydrochloride in MeOH), ibogamine (XLIII),mp 162'-164", and on the other its lithium aluminum hydride reduction product, catharanthinol, afforded an acetonide (XXXVII), mp 188'-191'. To account for the difficulty with which catharanthine eliminated the carbomethoxy group, it has been suggested that an intermediate in this reaction (XLIV) is highly strained (40). Since XLIV can be readily constructed from Dreiding Atomic Models, this explanation may not be correct. I n actual fact, it is probable that the acid-catalyzed decomposition of catharanthine takes a different course. It has more recently

XLII; H. = COOMe XLIIL; R = H

XLIV

been shown that catharanthine is converted in concentrated hydrochloric acid under reflux into A3-ibogamine (hydrochloride, mp 150'-154', [aID +go') and cleavamine (XLVII), the low yield being higher if a

9. THE

IBOGA AND

Voacanga ALKALOIDS

219

reducing agent (e.g., stannous chloride) was present (42,42a). This point is discussed further at the end of this section. The structure of catharanthine was completely established by the isolation of the indoloquinoline (Chart 1, V ; Me0 = H) from the selenium dehydrogenation products of XLIII (40). Whereas XLII required a temperature of 230"-250° in the presence of palladium on carbon to generate 3-methyl-5-ethylpyridine, catharanthine with the same catalyst at 150'-160" gave a good yield of 3-ethylpyridine. The latter reaction may proceed via the retro DielsAlder product, XLV, or, even better, via XLVI, whose fragmentation would be even more facile than that of XLII.

bOOMe XLV

XLVI

The structures of cleavamine and velbanamine are considered here since they are probably derived from precursors with the iboga skeleton. Cleavamine (XLVII), mp 109°-1130, ["ID + 56' (CHCIs),was obtained along with deacetylvindoline when leurosine (structure unknown but closely related to XLI) was refluxed with concentrated hydrochloric acid, stannous chloride, and tin (42). Velbanamine (XLVIII), mp 139"141", ["ID +56O (CHCIS), was the indolic product when vincaleukoblastine (XLI; R = Me) or leurocristine (XLI; R = CHO) were treated in the same way (42). Aside from the question of the mode of fission of the dimers (XLI), the production of the new tetracyclic systems in XLVII and XLVIII can be regarded as proceeding via a reverse Mannich

XLVII

XLVIII

reaction, reduction of the resulting iminium salt, and decarbomethoxylation. If the pentacyclic iboga system does not pre-exist in the dimer, a reverse Mannich reaction would not have to be invoked. Reduction of cleavamine gave a dihydro derivative, mp 1360-13S0, which is a C-ethyl isomer of quebrachamine, and, in fact, their IR-spectra

220

W . I. TAYLOR

are very similar. Of particular interest was a comparison of the mass spectra of dihydrocleavamine (XLVIX) and quebrachamine (L),which showed the expected similarities, viz., aromatic residues mie 156 and 143, hydroaromatic residue mie 124 (42a, 42b). This developed structure

CHz

,

CHI

C

282

i 56

124

XLVIX Dihydrocleavamine R

and I,

CHz

H 143

r

138

1

CHz a and c d

282

L Quebrachamine

\

156

+

I24

for cleavamine has been confirmed in all respects by the determination of its structure by X-ray crystallographic analysis (42c). I n a recent paper, the formation of cleavamine and A3-ibogamine from catharanthine (XL) under strongly acidic reducing conditions has been discussed in greater detail (42a). It is regarded as proceeding via the

9. THE IBOGA

221

AND ~ O ~ C G W L gALKALOIDS a

retro Mannich or equivalent product (LI) which a.fter decarboxylation could either be reduced t o furnish cleavamine or ring closed again to generate A3-ibogamine. An analogous intermediate (LI ; A 3 double bond reduced) has t o be invoked if the production of ibogamine along with

0:qy 3

OyJ 3

MeOOC

MeOOC

LI

LIi LIII (AS-reduced)

4-epi-ibogamine by the prolonged reflux of either 18-carbomethoxyibogamine or its C-4 epimer in concentrated hydrochloric acid is to be understood (42a). The carbomethoxy group is essential for the ring opening since, under identical acidic conditions, the C-4 epimeric ibogamines were recovered unchanged. LI as its equivalent indole has also t o be invoked to explain the formation of two new bases, pseudocatharanthine (LII), mp 114°-1160, [aID & 0" (solvent unspecified), and pseudodihydrocantharanthine (LIII), by refluxing catharanthine for 16 hours in acetic acid. The absence of water must be an important contrib!xtor to the success of this reaction because, in aqueous wid, it is the free acid which decarboxylates to drive the reaction in the direction of cleavamine (42a).

F. MASS SPZCTRA OF IBOGA ALKALOIDS Just as UV-, IR-, and NMR-spectra are used t o identify structural elements, to fingerprint, and t o elucidate the structures of molecules, so very recently has mass spectroscopy been applied to the problems of the organic chemist with very useful results in the case of indole alkaloids, of which the iboga group is but one class. When the mass spectra (34) of ibogamine, ibogaine, tabernanthine, and ibogaline are compared, they show a group of peaks m/e 122,124,135,136,and 149 owing to fragments of the molecule originating from a part which does not contain the additional substituents in the benzene ring. A second group of peaks in the case of ibogamine, a t m/e 156, 175, 251, 265, and 280, appear with almost the same intensity but 30 mass units higher for ibogaine and tabernanthine and 60 mass units higher for ibogaline. These results

Ibogaine 310

ib

f

cx f2

H

I

122

'-\\-\\

149

136

225

The astarisked structure becomes a positive ion if the alternate fission takes place; then. M = 122 is a rdical. CHARTVI. Principal slectmn.impact fragmentatlon products of ibogaine.

9.

T H E IBOGA AND

Voacanga ALKALOIDS

223

alone, accumulated on a few micrograms of material, show (stereochemistry excepted) that they differ only in the nature of the methoxyl substituent(s). In this way, the structure of ibogaline was proved to be 12,13-dimethoxyibogamine,which is in agreement with an earlier suggestion. The course of the fragmentation has been deduced (Chart VI) and was supported by the examination of deuterio compounds, e.g., ibogaine-20-d from XXVI and lithium aluminum deuteride ; ibogaine18-d from the decarbomethoxylation of voacangine by deuterated hydrazine ; and ibogaine-19-dz by reduction of ibogaine lactam with lithium aluminum deuteride. In the case of the carbomethoxy alkaloids, the fragmentation peaks which contain the aromatic nucleus are 58 mass units higher, and the typical iboga pattern in the range 120-150 mass units is retained without alteration. The pattern changes for d3-ibogamine (XL; COOMe = H ; decarbomethoxycatharanthine) and fissions " d" and "e (Chart VI) are negligible, since the ethyl group is no longer close to N,, which stabilizes the charge in the saturated bases. The presence of the double bond facilitates a retro Diels-Alder cleavage, the product of which directly yields m/e 122 or, by hydrogen rearrangement, mje 135, along with m/e 136, 143, and 156 analogous to those of dihydrocleavamine (XLVIX) (42a, 42b). ))

r

-I@

L

2

122

135

G. OTHER ~ K A L O I D S Under this heading are collected the alkaloids from the plants given in Table I, with the exception of the iboga bases already considered and those dealt with under Voacanga alkaloids. There were two bases isolated from Tabernanthe iboga, gabonine and kisantine, by only one group of workers (6). Gabonine, CZ1Hz8N2O4, mp 223"-226") [a]= + 65" (CHCl,), had bands in the carbonyl region at 1672 em-' (medium) and 1620 cm-I (strong); its UV-spectrum with maxima at 253,287, and 355 mp is indicative of extended conjugation. Kisantine, C21H28N203, mp 236"-23So, ["ID - 15" (CHCl,), had a medium-strength

224

W. I. TAYLOR

band a t 1670 cm-I and an indole-like UV-spectrum. It has recently been recognized that kisantipe is an oxindole, i.e., the 12-methoxy derivative of XXXII, and that gabonine may be the 13-methoxy equivalent of XXXVIII (COOMe=H). Both these alkaloids niay be artifacts obtained instead of ibogaline in the original isolation procedure (6). From Stemmadenia donnell-smithii, besides iboga alkaloids and voacamine, the indole ( + )-quebrachamine (L), mp 147"-149', [.ID + 1 1 1' (CHC13), and the indole stemmadenine (LIV, c(+x or a - t y bond), mp 199'-200' (dec.), [aID + 324' (pyridine),were isolated (11).The latter alkaloid also occurs in Diplorrhynchus condylocarpon (43) along with condylocarpine, shown (43a),to be LV (ratherthan LVI), and into which it was converted by potassium permanganate oxidation (44). Stem-

L'?

LVI

madenine is therefore LIV (a-y bond). These results were derived largely from a comparison of'the physical properties (UV-, IR-, NMRspectra) and mass spectra o f a number of derivatives, among which were the palladium dehydrogenation products of stemmadenine, the dimer (LVII),and 3-ethylpyridine, which accounted for all the carbons of the original molecule (44). i n the NMR-spectrum, the ethyl signals in dihydrocondylocarpine were a t abnormally high field which was

LVII

LVITI

9, THE

IBOGA AND

225

Voacanga ALKALOIDS

compatible with either structure, since the ethyl group would lie either above the aromatic ring [this is true for the C-ethyl of vindoline (42)] or above the acrylic ester system (44). From Ervntamia, besides coronaridine, the 2-acylindoles tabernaemontanine and dregamine (vide infra) were obtained ; from various Tabernaemontana species, iboga bases, voacamine, and olivacine (LVIII), mp 318O, were identified (10). Conopharyngia durissima has afforded iboga bases, two dimeric alkaloids discussed cnder Voacanga alkaloids, and a trace of a base, alkaloid E, mp 191°-1930, pK‘, 7.26, UV-maxima a t 210 and 305 m p , which differed from the other isolates in having no carbonyl absorption in the IR-spectrum (14). Conopharyngia pachysiphon, in contrast to C. durissima, has yielded only steroidal bases (45).

11. The Voacanga Alkaloids Plants of the Voacanga genus have given rise so far to four groups of bases, apart from the iboga type represented by voacangine, voacristine, and voacryptine (Table I); these are the sarpagine, 2-acylindole secosarpagine (derivatives), carbornethoxymethyleneindoline, and dimer types (Table 11).The terms “bisindoles” or ‘(dimers”were used to indicate the belief that the last group is probably derived by a doubling-up of monomeric systems. The reported production of voacangine from voacamine supports this view. The genus Callichilia is included here, since vobtusine is a constituent of the three species examined.

A. VOACHALOTINE This alkaloid is a member of the sarpagine group, and its structure was readily derived by simple transformations (58) which, among others, enabled it to be correlated with N,-methyldeoxysarpagine (LXII). Its CHzOH

It LXII

LXIII

TABLE I1 Voaoanga AND RELATED ALKALOIDS"

LX; R = CHCH3 LXI: R = Et

s

H

Name

Melting point (" C)

fl [aID(CHCls)

Observations

Sourceb L'

0

A. Sarpagine type Voachalotine (LIX)

B. 2-Acylindoles Vobasine (LX) Tabernaemontanine (LXI) Dregamine (LXI) Voacafrine (cZzHZS~Z04) Voacafricine (C22H24-26N204)

223-224

- 3"

111-113 208-210 217-219 106-109 186-205 135-137

- 158' - 57O

Dihydrovobasine

417) k(47)

- 93O

Dihydrovobasine

r(481, e(l0)

Structure unknown

o(49)

Structure unknown

o(49)

- 107' (B.HC1 in MeOH)

196-198

-

W. I. TAYLOR

LIX

C. Carbomethoxymethyleneindoline type Callichiline 235 (CzzH24Nz03)

-460'

D, Bisindoles : structures unknown (tentative formulas) Voacamine (voacanginine) 223 - 52" (C45H56N406) Voacorine [voacaline ? (54)l C45-46Hw-56N.tO7

UV NN &anilinoacrylate

2 COOMe, OMe, NMe; UV z 5 MeO-indole; acid yields voacangine (61) 2 COOMe, OMe, NMe; U V z 5 MeO-indole; acid yields no voacangine (61)

273

- 42'

Vobtusine (C4zH48N206)

305

- 321"

Voacamidine (C45HmN406)

128

- 174'

2 COOMe, OMe, NMe; UV z voacamine

Voacaminine

242

- 450

Mixgure of voacamine and voacarine (32)

Conodurine 222-225 (C~I-~ZH~O-~ZN~O~) Conoduramine (C41-42H50-52N406) a

2 15-217 (foaming)

- 101" - 770

c(50)

M

i;; 0 4

2 COOMe, 1 NMe; UV NN j-anilinoacrylate

2 COOMe, OMe, no NMe; UV z voacamine

2 COOMe, OMe, no NMe; UV z voacorine

Co-occurring iboga alkaloids and plant key are given in Table I, page 204. Parenthetical numbers refer to reference list.

u

rA

2 23

W. I. TAYLOR

chemistry and relationship to other menbers of the ajmaline-sarpagine group is discussed in Chapter 2 2 . The presence of a methyl substituent on N, is for the present of rare occurrence among indole alkaloids.

AND B. VOBASINE,DREGAMINX, TABERNAEMONTANINE, CALLICHILINE

Dregamine and tabernaemontanine were recognized (10) to possess a 2-acylindole chromophore, and vobasine has been correlated to them by showing that the first two alkaloids were diasteroisomeric dihydrovobasines (59). On the basis of degradation work, which has not been reported in full, the structures L X and LXI (see Table 11) have been deduced for these alkaloids, although the experimental results do no uniquely establish the heterocyclic system shown. Eiogenetic considerations may have influenced the authors since, among others, L X I I I also fits the published data. Voloasine, by either the action of strong base or hydrolysis followed by re-esterification, gave isovobasine (LX ; (2-16 ef?imer), mp 175”-178”, [a]=- 191’ (CHCI3) (17). Vobasine methiodide, subjected to a Hofmann degradation m d e r mild conditions, furnished vobasine methine, [a]= - 103” (UV-spectrum = 3-vinyl-2-acylindole). From isovobasine, an analogous methine ( [aID + 45” ; UV-spectrum = 3-vinyl-2-acylindole) was obtained. Both methines, upon treatment with sodium methoxide, formed the same optically inactive vobasineisomethine (no change in the UV-spectra). It should be noted that two optically active centers are ifivolved in this “racemization.” Vobasiae methine subjected to a second Hofmann degradation eliminated trimethylamine to yield deazavobasine, which retained the 3-vinyl-2-acylindole moiety and had, in addition, an isolated 1,3-diene function. ilexahydrodeazavobasine, MeOOC

CHzOH

H

LXIV

I

LXV

I

upon oxidation with chromic acid, formed a-methylbutyric acid (60). If the structure LX €or vobasine is correct, then LXlV may exist in acidic solution. Such an observation could be made the basis of a trivial solution

9.

THE IBOGA AND

Voacanga ALKALOIDS

229

for the structure of this class of compound, viz., the quaternary salt macusine R (LXV) should be the product of the acid treatment of the lithium aluminum hydride reduction product of either vobasine or isovobasine. Borohydride reduction (59) of vobasine gave vobasinol, mp 100°-102" (solvate) (0-acetate, mp 160°-162") upon which the action of acid could also be tried. The structure of callichiline, apart from recognition of its chromophoric moiety which must include the carbonyl and methoxy groups, is practically unknown; it has no N-methyl and is apparently not identical with any of the other indole alkaloids (about a dozen) with the same UV-spectra. If the functional group analyses are correct, biogenetic considerations would suggest that the formula be revised to C21H22-24N203.

C. BISINDOLES The known members of this group can be divided into three classes on the basis of their UV-spectra and base strengths. First, vobtusine with a UV- M two P-anilinoacrylates and pK, 6.95 may be built up of two identical monomeric units (two callichilines !). Second, voacorine with pK, 6.40 and a chromophoric moiety approximately equal to two isolated 5-oxyindoles may also be constructed from identical units. Third, voacamine and conduraniine, also with 5-oxyindole chromophores, with 5.4 (note similarity to voacangine, two dissociation constants pK', 7.0 may be derived from two dissimilar units. VoacpK', 5.6) and amidine and condurine remain unplaced, since their dissociation constants are unknown. I n all cases where a C-methyl determination has been made and where the resulting acids have been analyzed, they have been found to have formed acetic acid which, although excluding methyls attached to methylene, in these alkaloids cannot distinguish between an ethylidene residue (R un(resistant to hydrogenation '1) and/or a -CHOR-CHs specified). The nature of the fusion in these bisindoles is still unknown, and it does not appear t o involve an aldehyde function, as it does in certain curare and Geissosperma alkaloids. Among other possibilities are ether bonds, involvement, of N, (vide infra),or a linkage similar to that in the indole-indoline alkaloids represented by vincaleukoblastine (XLI). The isolation (61) of voacangine from voacamine (but not from voacorine), after reflux of the latter in 3 N hydrochloric acid or alcoholic hydrochloric acid, may indicate that voacangine is present as such in the dimer and is linked via, N, to the second unit. This would be in

-

N

230

W. I. TAYLOR

agreement with the PK', data, the isolation of 3-methyl-5-ethylpyridine after potash fusion of voacamine (other products were trimethylamine, acetic acid, propionic acid, and perhaps isobutyric acid), and those properties of the alkaloid which showed that one of its carbomethoxyl groups could be removed under the same conditions as were successful with voacangine (32).There are, however, other possibilities to consider; thus, in the foregoing fission with hydrochloric acid, an intermediate such as LXVI, LXVII, or LXVIII could have been formed, all of which could collapse into the isolated monomer, voacangine. The suggestion ( 1 7 ) that vobasine is a precursor of voacamine and that the latter compound undergoes an inversion of one of its carbomethoxyl functions upon hydrolysis (cf. vobasine, Section 11,B) cannot be decided on the

COOMe

LXVI

I COOMe

LXVII

I COOMe

LXVIII

basis of published work. Saponification of voacamine gave a dipotassium salt which, upon esterification in methanolic hydrogen chloride, resulted in a decarbomethoxyvoacamine (32). The same product was also produced upon lithium aluminum hydride reduction of voacamine in refluxing tetrahydrofuran. Selenium dehydrogenation of voacamine produced 4-methyl-3-ethylpyridine (p-collidine) and 13-carboline (32). This result could be interpreted as indicating that a voacangine nucleus does not pre-exist in voacamine, since voacangine with selenium has given the expected 3-methyl-5-ethylpyridine (13). Voacorine, like voacamine, furnished 3-methyl-5-ethylpyridine (56) upon potash fission, and vobtusine with selenium formed quinoline (13). A final property of significance is concerned with the behavior of both voacamine (62) and voacorine (56) upon pyrolysis in vacwo when trimethylamine and carbon dioxide (one mole equivalent of each) were given off. This has been

9.

231

THE IBOGA AND VOUCUngU ALKALOIDS

interpreted as occurring via a double betainization of the N-methyl and elimination via a retro Michael or fragmentation of a j3-amino ester and/or extrusion of nitrogen from a y or 8-dialkyl amino acid ester in which the corresponding lactone could be formed (56). Structure LXIX for voacamine has been put forward as a working hypothesis (32), although on the basis of the published information LXX is equally as attractive.

LXIX

LXX

I

III. Miscellaneous The mode of biosynthesis of none of these alkaloids is known but, in the case of the iboga group, some guesses have been made (39, 63, 64), all of which start from the amino acids, tryptophan and dihydroxyphenylalanine, and involve a fission of the latter's aromatic ring. A more sophisticated approach (65), starting from precursors of the aromatic amino acids, namely shikimic and prephenic acids, is apparently not in agreement with recent work on other indole alkaloids (66). The genesis of most indole alkaloids appears to stem from tryptophan and three

232

W. I. TAYLOR

formal fragments, 6C+ 1C+3C (67), whose exact nature remains t o be elucidated but are now thought tic involve 3-acetates +formic acid + malonic acid. Thus, the origin of the iboga and related bases is depicted formally as L X X I I I t LXXI +LXXII --f LXXIV, in which LXXII and LXXIV may be proximate precursors of the Voacanga alkaloids.

I R

LXXIII

LXXIV

From a pharmacological point of view, the Voacanga alkaloids are relatively nontoxic, rapidly eliminated, and of no great interest (68). The reported potent cardiotonic properties (9, 54, 69) of some of these alkaloids have apparently not withstood the test of time (52). The physiological actions of Tabernanthe iboga are very interesting. Its roots have been reported to exhibit sleep-combating and alerting effects (70). When chewed during stress, they are also said to prevent fatigue and hunger and have been used by natives in the Congo for this purpose (71). The roots have also been used in larger quantity by the same natives in fetishism. Modern research has traced this activity to the alkaloid portion and, in particular, to ibogaine (72). This alkaloid was shown to have distinct stimulating properties with weak anticonvulsant effects accompanied by reactions of apprehension and fear. Insertion of the carbomethoxyl group, i.e., voacangine, gave a product with only weak central nervous stimulating properties (73). Other papers have been published on pharmacological aspects which are beyond the scope of this article (74). A particularly good summary of the early botany and pharmacology

9.

THE IBOGA AND

Voacanga ALKALOIDS

233

of Tabernanthe iboga has been written (3), and a more recent paper has dealt with the former aspect in relation to the genus Daturicarpa (75). A paper has appeared on some aspects of the physiological properties of the alkaloids of Stemmadenia donnell-smithii (76). REFERENCES 1. J. Dybowski and E. Landrin, Compt. Rend. Acad. Sci. 133, 748 (1901); A. Haller and

E. Heckel, ibid. 133, 850 (1901). 2. C. A. Burckhardt, R. Goutarel, M.-Y. Janot, and E. Schlittler, Helv. Chim. Acta 35, 624 (1952). 3. J. Delourme-HoudB, Ann. Phamn. Franc. 4, 30 (1946). 4. N. Neuss, J. Org. Chem. 24, 2047 (1959). 5. R. Goutarel, F. Percheron, and M.-M. Janot, Compt. Rend. Acad.Sci. 246,279 (1958). 6. D. F. Dickel, C. L. Holden, R. C. Maxfield, L. E. Paszek, and W. I. Taylor, J . Am. Chem. SOC.80, 123 (1958). 7. R. Goutarel and &I.-M. Janot, Ann. Phurm. Franc. 11, 272 (1953). 8. M.-M. Janot and R. Goutarel, Compt. Rend. Acad. Sei. 240, 1800 (1955). 9. J. La Barre and L. Gillo, Bull. Acad. Roy. Med. Belg. 20, 194 (1955). 10. M. Gorman, N. Neuss, N. J. Cone, and J. A. Deyrup,J. Am. Chem. Soc. 82,1142 (1960). 11. F. Walls, 0. Collera, and A. Sandoval, Tetrahedron 2, 173 (1958). 12. M. Gorman, N. Neuss, G. H. Svoboda, A. J. Barnes, and N. J. Cone, J . Am. Pharm. Assoc. Sci. E d . 48, 256 (1959). 13. B. 0. G. Schuler, A. A. Verbeek, and F. L. Warren, J . Chem. Soc. p. 4776 (1958). 14. U. Renner, D. A. Prins, and W. G. Stoll, Helv. Chim. Acta 42, 1572 (1959). 15. U. Renner, Experientia 13, 468 (1957). 16. D. Stauffacher and E. Seebeck, Helv. Chim. Acta 41, 169 (1958). 17. U. Renner, Experientia 15, 185 (1959). 18. Olctta S. A., Belg. Patent 555,059 (1957). 19. Raymond-Hamet, Bull. Sci.Pharmacol. 33, 447, 518 (1926); Bull. Soc. Chim. Biol. 24, 10 (1942). 20. Raymond-Hamet, Bull. SOC.Chim. Biol. 25, 205 (1943); Compt. Rend. Acad. Sci. 229, 1359 (1949). 21. M.:M. Janot, R. Goutarel, and R. P. A. Sneeden, Helv. Chim. Acta 34, 1205 (1951). 22. W. I. Taylor, J . Am. Chem. SOC.79,3298 (1957). 23. E. Schlittler, C. A. Burckhardt, and E. Gellert, Helv. Chim. Acta 36, 1337 (1953). 24. R . Goutarel, M.-M. Janot, F. Mathys, and V. Prelog, Compt. Rend. Acad. Sci. 237, 1718 (1953). 25. M. F. Bartlett, D. F. Dickel, and W. I. Taylor,J. Am. Chem. SOC.80, i26 (1958). 26. R. Goutarel, M.-M. Janot, F. Mathys, and V. Prelog, Helv. Chim. Acta 39, 742 (1956). 27. W. I. Taylor, Proc. Chem. Soc. p. 247 (1962). 28. M. M. Robison, R. A. Lucas, H. B. MacPhillamy, R. L. Dziemian, I. Hsu, and R. J. Kiesel, J . Am. Chem. SOC. 83,2694 (1961). 29. R. Goutarel, Thesis, University of Paris, Paris (1954). 30. M. F. Bartlett, D. F. Dickel, R. C. Maxfield, L. E. Paszek, and A. F. Smith, J. Am. Chem. Soc. 81, 1932 (1959). 31. G. Arai, J. Coppola, and G. A. Jeffery, Acta Cryst. 13, 553 (1960). 32. F. Percheron, Ann. Chim. (Paris)4, 303 (1959). 33. U. Renner and D. A. Prins, Experientia 15, 456 (1959).

234

W. I. TAYLOR

34. K. Biemann and M. Friedmann-Spiteller, Tetrahedron Letters p. 68 (1961); J . Am. Chem. SOC. 83,4805 (1962). 35. H. B. MacPhillamy, R. L. Dziemian, R. A. Lucas, and M. E. Kuehne, J . Am. Chem. SOC. 80, 2172 (1958). 36. R. Goutarel, F. Percheron, J. Wohlfahrt, and M.-M. Janot, Ann. Pharm. Franc. 15, 353 (1957). 37. M.-M. Janot and R. Goutarel, Compt. Rend. Acad. Sci. 241,986 (1955). 38. W. Schindler, Helv. Chim. Acta 41, 1441 (1958). 39. F. Percheron, A. Le Hir, R. Goutarel, and M.-M. Janot, Compt. Rend. Acad. Sci. 245, 1141 (1957). 40. N. Neuss and M. Gorman, Tetrahedron Letters p. 206 (1961). 41. U. Ronner and D. A. Prins, Ezperientia 17, 106 (1961). 42. G. H. Svoboda, I. S. Johnson, M. Gorman, and N. Neuss,J. Plmrm.Sci. 51,707 (1962). 42a. M. Gorman and N. Neuss, Am. Chem. SOC., 144th Meeting, Los Angeles, 1963 p. 38M. 42b. M. Gorman and N. Neuss, Ann. Chim. (Rome)53, 43 (1963). 42c. J. P. Kutney, J. Trotter, T. Tabata, A. Kerigan, and N. Camerman, Chern. I n d . (London)p. 648 (1963). 42d. W. I. Taylor, J . Org. Chem. I n press. 43. D. Stauffachor, Helv. Chim. Acta 44, 2006 (1961). 43a. K. Biemann, A. L. Burlingame, and D. Stauffacher, Tetrahedron Letters p. 527 (1962). 44. A. Sandoval, F. Walls, J. N. Shoolery, J. M. Wilson, H. Budzikiewicz, and C. Djerassi, Tetrahedron Letters p. 409 (1962). 45. D. F. Dickel, R. A. Lucas, a n d H . B. MacPhillamy, J . Am. Chem. SOC. 81,3154 (1959). 46. J. Pecher, N. Defay, M. Gauthier, J. Peeters, R. H. Martin, and A. Vandermeers, Chem. Ind. (London)p. 1481 (1960). 47. A. N. Ratnagiriswaran and K. Venkatochalam, Quart. J.Pham. P h a m c o l . 12, 174 (1939). 48. N. Neuss and N. J. Cone, Ezperientia 15,414 (1959). 49. K. V. Rao,J. Org. Chem. 23, 1455 (1958). 50. R. Goutarel, A. Rassat, M. Plat, and J. Poisson, Bull. SOC. Chim. France p. 893 (1959). 51. M.-M. Janot and R. Goutarel, Compt. Rend. Acad. Sci. 240, 1719 (1955). 52. F. Fish, F. Newcombe, and J. Poisson, J. Pharm. Pharmacol. 12, Suppl. 41T (1960). 53. R. Goutarel and M.-M. Janot, Compt. Rend. Acad. Sci. 242,2981 (1956). 54. J. La Barre and L. Gillo, Compt. Rend. SOC.Biol. 150, 1628 (1956). 55. M. D. Patel, J. M. Rowson, and D. A. H. Taylor, J . Chem. SOC. p. 2587 (1961). 56. M.-M. Janot, F. Percheron, M. Chaigneau, and R. Goutarel, Compt. Rend. A d . Sci. 244, 1955 (1957). 57. J. La Barre, J. Lequime, and J. van Heerswynghels, Bull. Acad. Roy. Med. Belg. 20, 415 (1955). 58. J. Pecher, R. H. Martin, N. Defay, M. Kaisin, J. Peeters, and G. van Binst, Tetrahedron Letters p. 270 (1961); N. Defay, M. Kaisin, J. Pecher. and R. H. Martin, Bull. SOC. Chim. Belges 70,475 (1961);M.-M. Janot, J. Le Men, J. Gouset, and J. Levy, Bull. SOC.Chim. France p. 1079 (1962). 59. U. Renner and D. A. Prins, Ezperientia 17,209 (1961). 60. U. Renner and D. A. Prins, Chimia ( A a r a u ) 15, 321 (1961). 61. W. Winkler, Naturwissenschaften 48, 694 (1961). 62. R. Goutarel, F. Percheron, and M.-M. Janot, Compt. Rend. Acad.Scz. 243,1670 (1956). 63. W. I. Taylor, Ezperientia 13, 454 (1957). 64. D. Stauffacher, Chimiu ( A a r u u )12, 88 (1958).

9. THE

IBOGA AND VOUCUngU ALKALOIDS

235

E. Wenkert, J . Am. C h m . SOC.84, 98 (1962). E. Leete, S. Ghosal, and P. N. Edwards, J . Am. Chem. SOC.84, 1608 (1962). E. Schlittler and W. I. Taylor, Experientia 16, 244 (1960). A. Quevauviller, R. Goutarel, and M.-M. Janot, Ann. Phurm. Franc. 13, 423 (1955); A. Quevauviller and 0. Blanpin, Therapie 12, 635 (1957); J. La Barre, Compt. Rend. SOC. Bid. 150, 1493, 1494(1956); Compt. Rend.Soc. Biol. 151,615 (1957); G.lfogeland H. Uebel, Arzneimittel-Forsch. 11, 787 (1961). 69. J. La Barre, Compt. Rend. SOC. Biol. 149, 2263 (1955); J. La Barre, J.Physiol. (Paris) 48, 588 (1956); Compt. Rend. Soc. B i d . 155, 1017 (1956); A. Quevauviller and 0. Blanpin, ibid. 150, 1113 (1956); 151, 1864 (1957). Linne (Paris)1, 782 (1869). 70. H. Baillon, Bull. Mens. SOC. 71. A. Landrin, Bull. Sci. Phurmacol. 11, 319 (1905);Aubry-Lecompte, Arch. Med. Navale

65. 66. 67. 68.

2, 264 (1864). 72. J. A. Schneider and E. B. Sigg, Ann. N . Y . Acad. Sci, 66, 765 (1957). 73. G. Zetler and K. R. Unna, Arch. Exptl. Pathol. Phurmakol. 236, 122 (1959). 74. R. Paris and C. Vairel, Compt. Rend. Acad. Sci.228, 436 (1949); J. A. Schneider and M. McArthur, Experientin 12, 323 (1956); R. K. Rinehart and J . A. Schneider, J. Phrmacot. Ezptl. Therap. 119, 179 (1957); J. A. Schneider and R. K. Rinehart, Arch. Intern. Pharmacodyn. 60, 92 (1957); Raymond-Hamet and D. Vincent, Compt. Rend. SOC. Biol. 154, 2223 (1960). 75. H. Hiirlimann, Ber. Schweiz. Botan. ges. 67, 487 (1957). 76. Raymond-Hamet, Compt. Rend. Acad. Sci. 251, 2098 (1960).

This Page Intentionally Left Blank

---CHAPTER

10-

THE CHEMISTRY OF THE

2,2’-INDOLYLQUlNUCLIDINEALKALOIDS W. I. TAYLOR Research Department, CZBA Pharmaceutical Company, Division of CZBA Corporation, Summit, New Jersey

I. Determination of the Structure of the Alkaloids. .........................

238

11. Synthesis of Cinchonamine............................................

243

111. Stereochemistryof Cinchonamine. .....................................

244

IV. Miscellaneous. ......................................................

246

References..........................................................

246

There are recognized a t present three naturally occurring members of this group, cinchonamine, quinamine, and conquinamine, all minor alkaloids of certain Cinchona and Remijia species. The elucidation of their structures led to the suggestion that the quinoline moiety of the major bases, e.g., cinchonine and quinine, of these plants was probably derived from tryptophan via an indolic precursor. It has since been demonstrated from the results of feeding experiments with isotopically labeled tryptophan that this amino acid really can serve as a precursor of various indole alkaloids (1) as well as of quinine ( 2 ) . The details of these processes are not yet known but probably involve an intermediate@) related t o cinchonamine (2, 3, 6).

OMe Corynantheine

Cinchonamine 237

Cinchonine (quinine)

238

W. I. TAYLOR

I. Determination of the Structure of the Alkaloids1 Cinchonamine, mp 194", [a],, +121.1" (EtOH) was first isolated in 1881 (4) and shown to possess the formula C1gHz40Nz (4,5);its structure was determined in 1950 (6). Early work had shown cinchonamine to give color reactions typical of indole alkaloids (7), and this was also evident from its UV-spectrum (8). The base differs from the major cinchona alkaloids in yielding, upon oxidation with chromic acid (6), 3-vinylquinuclidine-8-carboxylic acid (111), mp 209", [a],, -29" (CHC13), which was first obtained from quinamine (9). The nature of the remainder of the molecule followed from the conversion of cinchonamine into O,N,-diacetylallocinchonamine (I),mp 159", [a],, -7" (CHClS), by refluxing acetic anhydride and its subsequent oxidation to 3-,6-acetoxyethylindole-2-aldehyde (11)(6). Dehydrogenation of cinchonamine with selenium or palladium-charcoal resulted in fission of the quinuclidine ring and the isolation of dehydrocinchonamine (V), mp 203", whose spectral characteristics and plZ, were in agreement with its formulation. From the selenium dehydrogenation products, a trace of a crystalline compound was also obtained which, from a single ultimate analysis and UV-spectrum, is considered t o be IV (10). Cinchonamine has been converted in two steps (11)into the quaternary chloride (VI), mp 320" [tosylate, mp 313', [.ID - 69.5" (90% MeOH)], identical with the same product derived from dihydrocorynantheol (VII) of known relative configuration. Also, cinchonine of known absolute configuration (12) has been converted into dihydrocinchonamine, mp 161", [aID 122' (EtOH) (13). These experiments serve to establish the absolute stereochemistry of cinchonamine (Chart I). The epimeric bases, quinamine, mp 185", [aID +104" (EtOH), and conquinamine (3-epiquinamine), C l g H 2 4 0 ~ N 2 ,mp 121°, [mID 204.6'

+

+

The numbering system for the alkaloidsin this chapter is based on biogenetic relationships, i.e., an atom is assigned the same number as its supposed equivalent in yohimban, viz. :

10.

/=

=J( “9 -

THE 2,2’-INDOLYLQUINUCLIDINEALKALOIDS

239

240

W. I. TAYLOR

(EtOH), were first isolated in 1872 and 1877, respectively (14). Quinamine was observed to give indole color reactions (7, 15), and 2,3dimethylindole was a result of zinc dust distillation of the alkaloid (15). With chromic acid (9), the quinuclidine carboxylic acid (111) was was obtained, and with nitric acid 3,6,8-trinitro-4-hydroxyquinoline isolated (15, 16). This quinoline is a consequence of fission of the indole and recyclization, with nitration preceding and following these steps [cf. ozonolysis of yohimbine to furnish a 2,3-disubstituted 4-hydroxyquinoline (17)]. Final clarification of the structure of quinamine was accomplished on the one hand by the production of cinchonamine upon lithium aluminum hydride reduction of the base (6),and on the other hand by the reverse reaction brought about by peracid (18) (Chart I). The reduction is explicable via the ring chain tautomeric hydroxyindolenine (VIII), which is also the primary product of the peracid oxidation of cinchonamine [cf. oxidation of tetrahydrocarbazole (19)]. The oxidation is stereospecific, but the stereochemistry of the introduced C-7 hydroxyl is still unknown.

11

IX

VIII

XI1

With the structure of quinamine firmly established, its other reactions can now be considered (Chart 11),all of which proceed via VIII. When the alkaloid was refluxed with acetic anhydride or acetyl chloride. acetylapoquinamine (0-acetyl-d 14-cinchonamine,XII), picrate, mp 143' was obtained. The change may be formulated, q u i n a m i n e ~ V I I I ~ I X + X I I , with the last, step being a special case of a general reaction of hydroxyindolenines (20). The analogous transformation can be realized very easily when 7-hydroxy-7H-yohimbine methiodide (X) is boiled in methanol (21). 37

10. THE 2,2'-INDOLYLQUINTJCLIDINE ALKALOIDS

241

Prolonged reflux of either quinamine or apoquinamine, mp 115') [.ID H2S04),in dilute acetic acid gave an amorphous material called ('quinamicine" (14, 15), to which structure XIV was given (16). - 32.9" (0.1 N

MeOOC/'\/

I OH

I OH

X

It was shown later (23) that ((quinamicine" was in reality crude quinamidine (XIV), mp 93" (ethanolate), which Hesse (14) had prepared in a pure state either by allowing quinamine to stand a t room temperature for several days in 13% hydrochloric acid or by heating the alkaloid a t 130" in aqueous tartaric acid. When quinamine was refluxed with sodium ethoxide (22, 23)) it slowly epimerized a t (2-3, quinamine 7 t V I I I SIX, to furnish 3-epiquinamine [conquinamine (14)], mp 120°, [a]D + 197" (EtOH) (Chart 11). While this sequence was reversible, an irreversible rearrangement via VIII (and 3-epi-VIII) gradually gave the two possible,pseudoindoxyls, isoquinamine (XIII),mp 211") [tr]D - 424" (0.1 N HzS04),and its epimer (XI), mp 183') [aID - 269" (EtOH) (22-24). Reduction of isoquinamine gave the dihydro derivative, allodihydroisoquinamine (XVI)which, in acid, unlike model dihydropseudoindoxyls, ring closed to the anhydro compound (XV) rather than rearranging to the indole (19). This tendency to cyclize was also evidenced when isoquinamine was heated with acids, the quaternary salt being produced (16, 22). Lithium aluminum hydride reduction of epiquinamine, as expected, afforded 3-epicinchonamine, mp 168", [.ID 48" (EtOH), also obtainable along with cinchonamine by the sodium-ethanol reduction of apoquinamine (23). Finally, it was shown in 1945 that, on heating quinamine or dihydroquinamine above its melting point, formaldehyde was evolved, and this was taken as evidence for the presence of a 2-hydroxymethyl on an indole a-carbon (15). I n the light of the true structure, the writer would like to suggest that this aldehyde is formed by pyrolysis of a 1,3-glycol, that is, a retro Prins reaction (25).

+

242

/===

/=

W. I. TAYLOR

/=

/====

10.

HzO f

243

THE 2,2’-INDOLYLQUINUCLIDINEALKALOIDS

+ CHzO f +

The pyrolysis of isoquinamine (XIII)to yield acetaldehyde has also been recorded (22). This fission may also proceed via a cyclic intermediate. A A 60

-

CH~CHOt

+

u,g J

(P)

I/

11. Synthesis of Cinchonamine

An early synthesis of cinchonamine is that originating from the Oxford group (23, 24), but the full details have not yet been published. It consisted of the coupling of tryptophol with meroquinene to furnish quinamidine, which was subsequently converted to 439 14-cinchonamine (apoquinamine). A second synthesis was carried out by a Russian school, starting this acid (26), according to the time from 3-vinylquinuclidine-8-carboxylic scheme shown in Chart I11 (27). These workers have also prepared the pyridine analog of cinchonamine (28). By far the most interesting and prolonged piece of work was carried out in Japan and had as its starting point cinchonine or cinchonidine (13,

244

W. I. TAYLOR

H

HOOC

29, 30). The synthetic pathway (Chart IV) leading to dihydrocinchonamine is probably a reversal of the biogenetic route. The same authors have also synthesized 10-methoxydihydrocinchonamine(30,3l).In none of these syntheses were any of the 3-epi compounds isolated.

Cinchonamine

C ~ A R T111. Synthesis of cinchonamine from 3-vinylquinuclidine-8-carboxylic acid.

III. Stereochemistry of Cinchonamine The best experimental evidence for the stereochemistry of cinchonamine is that indicated in Chart I, where dihydrocinchonamine and

-\ a

i

-

g 0 _3

-\

10. THE 2,2’-INDOLYLQUINUCLIDINEALKALOIDS

0 -L

-(b

(=I$

245

246

W . I. TAYLOR

dihydrocorynantheol yield the same quaternary base (VI) by processes which probably do not affect (2-3. All the syntheses of cinchonamine have proceeded via A3714-dihydrocinchonamine or 8-acyl-3-ethylquinuclidine intermediates. I n the latter case, the configuration of the acyl derivatives is unknown, although the Japanese group have derived the same ketone, probably because it is the least-soluble epimer, starting from either cinchonine or cinchonidine. Even if the stereochemistry of the crystalline ketone had been determined, it would not necessarily retain its configuration in solution, since analogous ketones of the Cinchona bases (32) are known to mutarotate. This comment is also relevant to the Russian work, since it was known (6) that their starting material, 2-vinylquinuclidine-8-carboxylicacid, niutarotated in solution. In spite of and recognizing this accumulated evidence it has been suggested (33), on the basis of molecular rotation differences between deoxycinohonidine and deoxycinchonine on the one hand and cinchonamine and 3-epicinchonamine on the other, that cinchonamine is configurationally the same as deoxycinchonidine. This conclusion would have been worthy of serious consideration if it had been shown that the indolyl and the 4-methylenequinolyl moieties were electronically equivalent for such a comparison.

IV. Miscellaneous Little df value has come out of pharmacological studies on these alkaloids (34) ; local anesthetic properties (35) have been reported, and none of the alkaloids have shown antimalarial properties (36). Crystal data for a number of cinchonamine salts (37) and apoquinamine (38) have been recorded. An old proposal (5)for taking advantage of the insolubility of cinchonamine nitrate as an analytical method for nitrate has been re-examined (39)REFERENCES 1. A. R. Battersby, Quart. Rew. (London) 15, 259 (1961). 2. E. Leete and N. Kowanko, Am. Chem.Soc. 142nd Meeting, Atlantic City, 1962, p. 53Q; J. Am. Chem. SOC.84, 4919 (1962). 3. P. de Moerloose and R. Ruyssen, Z n d . Chim. Belge 20, Spec. No., p. 492 (1955). 4. M. Arnaud, Compt. Rend. Acad. Sci. 93, 593 (1881); 97, 174 (1883). 5. L. Tshugaeff, Ber. 34,1824 (1901);L. Boutroux andP. Genvresse, Compt. Rend. Acud. Sci. 125,467 (1897); B. F. Howard and F. Perry, J . SOC. Chem. Znd. 24, 1281 (1905); B. F. Howard and 0. Chick, ibid. 28,53 (1909). 6. R . Goutarel, M.-M. Janot, V. Prelog, and W. I. Taylor, Helv. Chim. Acta 33,150 (1950); W. I. Taylor, ibid. 33, 164 (1950).

10.

THE 2,2‘-INDOLYLQUINUCLIDINE ALKALOIDS

247

7. Raymond-Hamet, Compt. Rend. Acad. Sci. 212, 135 (1941); 220, 670 (1945); 221,307 (1945); 227, 1182 (1948). 8. M.-M. Janot and A. Berton, Compt. Rend. Acad. Sci. 216, 564 (1943); RaymondHamet, ibid. 230, 1183 (1950). 9. T. A. Henry, K. S. Kirby, and G. E. Shaw, J . Chem. SOC.p. 524 (1956). 10. R. Goutarel, M.-M. Janot, V. Prelog, and W. I . Taylor, Unpublished observations. 11. E. WenkertandN. V. Bringi,J. Am. Chem.Soc. 80,3484 (1958); 81,1474,6535 (1959). 12. V. Prelog and E. Zahn, Helu. Chim. Acta 27, 545 (1944). 13. E. Ochiai and M. Ishikawa, Chem. Phurm. BUZZ.(Tokyo) 6, 208 (1958). 14. 0. Hesse, Ber. 5 , 265 (1872); 10, 2152 (1877); Ann. 166, 217 (1873); 199, 333 (1879); 207, 288 (1881); ef. also J. E. de Vrij, P h a n n . J . 4, 609 (1874); B. F. Howard, ibid. 5 , 1 (1875); A. C. Oudemans, Ann. 197,48 (1879). 15. K. S. Kirby, J . Chem. Soc. p. 528 (1945). 16. G. Bendz, C. C. J. Culvenor, L. J. Goldsworthy, K. S. Kirby, and Sir Robert Robinson, J . Chem. SOC.p. 1130 (1950). 17. B. Witkop and S. Goodwin, J . Am. Chem. SOC.75, 3371 (1953). 18. B. Witkop,J. Am. Chem. SOC. 72, 2311 (1950). 19. B. Witkop, Bull. SOC.Chim. France p. 423 (1954). 20. W. I. Taylor, Proc. Chem. SOC. p. 247 (1962). 21. W. I. Taylor, Unpublished observations. 22. K. S. Kirby, J . Chem. SOC. p. 735 (1949). 23. C. C. J . Culvenor, L. J . Goldsworthy, K. S. Kirby, and Sir Robert Robinson, J . Chem. Soe. p. 1485 (1950); and Unpublished observations. 24. C. C. J. Culvenor, L. J. Goldsworthy, K. S. Kirby, and Sir Robert Robinson, Nature 166, 105 (1950); C. C. J. Culvenor, Thesis, Oxford Univ. (1950). 25. E. Arundde and L. A. Mikeska, Chem. Rev. 51, 505 (1952). 26. R. P. Evstigneeva, C . Chang-pai, and N. A. Preobrazhenskii, Zh. Obshch. Khim. 30, 473 (1960). 27. C. Chang-pai, R. P. Evstigneeva, and N. A. Preobrazhenskii, Dokl. Aknd. N a u k SSSR 123, 707 (1958); Zh. Obshch. Khim. 30, 2085 (1960). 28. C. Chang-pai, R. P. Evstigneeva, and N. A. Preobrazhenskii, Zh. Obshch. Khim. 28, 3085 (1958). 29. E. Ochiai, M. Ishikawa, and Y. Oka, Chem. Phurm. Bull. Tokyo 7, 744 (1959); Ann. Rept. I t s u u Lab., No. 12, 29 (1962); E . Ochiai and M. Ishikawa, Chem. Pharm. Bull. (Tokyo)5, 498 (1957). 30. E. Ochiai, H. Kataoka, T. Dodo, and M. Takahashi, Chem. Phurm. Bull. (Tokyo) 10, 76 (1962); Ann. Rept. I t s u u Lab. No. 12, 11 (1962). 31. M. Ishikawa, Chem. Pharm. Bull. (Tokyo) 6, 67, 71 (1958); 5, 497 (1957). 32. R. B. Turner and R. B. Woodward, in “The Alkaloids” (R. H. F. Manske, ed.), Vol. 111,p. 29, Academic Press, New York, 1953. 33. R. L. Angustine, Chem. I n d . (London)p. 1071 (1959). 34. Raymond-Hamet, Compt. Rend. Soc. Biol. 134, 510 (1940); 135, 263 (1941); J. C. David and C. Vareed, I n d . J . Med. Res. 20, 1005 (1933); D. K. de Jough and E. G. van Proosdij-Hartzema, Arch. Intern. Pharmacodyn. 98, 320 (1954). 35. Raymond-Hamet, Compt. Rend. SOC.Biol. 147, 394 (1953); R. N. Chopra and J. C. David, I n d . J . Med. aes. 15, 343 (1927). 36. Raymond-Haniet, Compt. Rend. SOC.Biol. 150, 1883 (1956). 37. P. J. F. Griffiths, Acta Cryst. 5 , 290 (1952); 12, 418 (1959). 38. J. G. Scane, Acta Cryst. 15, 512 (1962). 39. C. Duval and N. D. Zuong, Anal. Chim. Acta 5,494 (1951).

This Page Intentionally Left Blank

-CHAPTER

11-

THE P E N T A C E U S AND THE EBURNAMINE (HUNTERL4)-VINCAMINE ALKALOIDS W. I. TAYLOR Research Department, CIBA Pharmaceutical Company, Division of CIBA Corporation, Summit, New Jersey I. The Pentaceras Alkaloids (Canthin-6-ones). .............................. 11. The Eburnamine (Hunteria)-Vincamine Alkaloids ........................ A. Eburnctmonine.. ................................................. B. Eburnamine, Isoeburnamine, and Eburnamenine ..................... C. Vincamine and Related Compounds. ................................ D. Mass Spectra of the Eburnamine-Vicamine Alkaloids.. ............... 111. The Hunteria and Pleiocarpa Alkaloids. References.....

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

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

250 253 253 258 259 262 262 267

Under this heading, the chemistry of those alkaloids known to have the skeletons I and I1 are considered, although they have been obtained from different plant families. This is more than a matter of convenience,

I

I1

Canthin-6-one group

Eburnamine-vincamine group

250

W. I. TAYLOR

for the rapid development of the chemistry of type I1 compounds was initially dependent on knowledge gained previously about type I. From a biogenetic point of view, type I can be dissected into tryptamine and 4C or 5C (then loss of carboxyl) units, whereas the HunteriaVinca group (11) involves a tryptamine and the lC+6C+3C units common to all complex indole alkaloids (1). The pattern in 11, upon transposing two of the bonds to adjacent carbons of the indole moiety, generates the skeletons I11 (quebrachamine) and IV (Aspidosperma)of two other classes of alkaloids whose chemistries, being quite different, obscure this close biogenetic origin. It has become obvious only quite recently that, in many apocynaceous plants,,types 11, 111, and IV co-occur-a state of affairs quite familiar in the case of yohimbinoid alkaloids (2); but in contrast to the yohimbinoid group all the stereochemical possibilities, including the racemates, are being discovered.

I. The Pentaceras Alkaloids (Canthin-6-ones) From Pentaceras australis Hook f. (Rutaceae), three optically inactive colored alkaloids were isolated (3) whose very characteristic UV-spectra closely resembled the carbazole derivative V. The structure of the simplest member of this group, canthin-6-one (VI), mp 162O, methiodide, mp 27lo-273O, N-oxide, mp 238”, was readily derived when with potassium permanganate, it was found to give ,B-carboline-1-carboxylic acid (3). Canthin-6-one has also been found in Zanthoxylum suberosum C.T. White (Rubiaceae) (4)and Picrasma crenata (Vill.) Engl. (Simarubaceae) ( 5 ) . The alkaloid was soluble in boiling aqueous ethanolic alkali, and acid, which was after acidification furnished 2-(l’-,B-carbolyl)-cis-acrylic reconverted into the parent lactam under mild alkaline conditions. Prolonged heating in alkali transformed the alkaloid into the transacrylic acid which could not be recyclized. The 4,5-double bond was readily hydrogenated, either with hydrogen in the presence of Raney nickel or by short reflux w-ith zinc and acetic acid. Prolonged reflux of the alkaloid with the latter reagent formed the deoxy compound VII, also produced from the N-oxide with zinc and hydrochloric acid. 4,5Dihydrocanthin-6-one, like 1-methyl-,B-carboline,with benzaldehyde in base gave the expected benzylidene derivative. Surprisingly, 2-( l’-/Icarboly1)propionic acid, obtained either by hydrolysis of 4,5-dihydrocanthin-6-one or by hydrogenation of the acrylic acids, could not be cyclized. However, dehydrogenation of the propionic acid derivative with selenium dioxide formed a mixture of canthin-6-one and the transacrylic acid. Hexahydrocanthin-6-one (VIII) had previously been syn-

11. Pentacerus AND

EBURNAMINE (HZ&eriG)-YINCAMINE

251

thesized (6), and conditions have been found for its successful dehydrogenation t o VI (7).

V

VI Canthin-6-one

VIII

VII

I X ; R = Me X;R=H 5-Methoxycanthin6-one

A second alkaloid, 5-methoxycanthin-Gone (IX), mp 24lo-242O, was demethylated with hydrobromic acid in acetic acid to the phenol X, which formed water-insoluble alkali metal salts ( 7 , 8). Upon treatment of the phenol with diazomethane in methylene chloride ( 7 ) or of its 0-acetate with the same reagent in moist ether (8), the alkaloid was regenerated. Its structure was proved by oxidation to ,6-carboline-1carboxylic acid, conversion of the phenol to a hydroxyquinoxaline (XI), and reduction with zinc and acetic acid to a mixture of canthin-6-one and its 4,5-dihydro derivative (8). The phenol (X) with the last reagent gave the same products, but with Raney nickel (in the presence of base) it yielded the ,6-carbolyl lactic acid (XII). When I X was hydrolyzed, it formed only one acrylic acid which could be recyclized t o the original lactam. A synthesis of the phenol was accomplished ( 7 ) by condensing diethyl oxalate with the dilithium derivative of 1 -methyl-,6-carboline. A formal reversal of this synthesis was brought about by heating I X methiodide in base, whereupon XI11 (or tautomer) crystallized (8). A third alkaloid contained sulfur (a rare feature in higher plant nitrogen heterocycles) and was shown to be 4-methylthiocanthin-6-one (XIV) (9). Hydrolysis with amyl alcoholic alkali gave methyl mercaptan and the phenol XV which, upon permanganate oxidation, furnished ,6-carboline-1-carboxylic acid. The phenol did not form a quinoxaline with o-phenylenediamine, and its acetate in moist ethereal diazomethane

252

W. I. TAYLOR

afforded, along with the betaine XVI, an ether isomeric with IX. A brief hydrolysis of the alkaloid resulted in the methylthioacrylic acid which reclosed in dilute aqueous alkali. With zinc and hydrochloric acid, XIV

unexpectedly gave 4,5-dihydrocanthine (VII), and perhaps the comparative ease of reduction of the CS carbonyl function in these systems points to a considerable amount of keto-carbonyl character in that moiety. In benzene solution with excess Raney nickel, the alkaloid afforded 4,5-dihydrocanthin-6-onebut, with an (‘aged” catalyst, canthin-6-one itself was the product.

H

XIV 4-Methylthiocanthin6-one

xv

XVI

A synthesis of XIV was accomplished as follows (9): /3-carboline-lcarboxylic acid chloride was condensed with the magnesium ethoxy derivative of malonic ester to yield, after acid hydrolysis, 4-hydroxycanthin-6-one (XV). This compound, with phosphorous oxychloride followed by heating in a sealed tube with potassium methyl mercaptide, gave the alkaloid XIV. If 4-hydroxycanthin-6-one is condensed with phosphorous oxychloride-phosphorous pentachloride, 4,5-dichlorocanare formed. thin-6-one and 4-hydroxy-5-chlorocanthin-6-one 4,5-Dimethoxycanthin-6-one has been isolated from Picrasma ailanthoides Sieb et Zucc. (Simarubaceae), and its structure was established by recognition of the characteristic UV-spectrum, functional group analysis, and oxidation to /3-carboline-l-carboxylic acid (10).

11. Pentaceras

AND EBURNAMINE

(Hunteria)-VINCAMINE

253

11. The Eburnamine (Hunteria)-Vincamine Alkaloids

A. EBURNAMONINE Among the alkaloids of Hunteria eburnea Pichon (Table I) were four interrelated ones (11). Two of them, eburnamine and isoeburnamine, were diastereoisomeric pentacyclic indoles (XVII) convertible by acids into eburnamenine, an N-vinylindole (XVIII), on the one hand, and by chromic acid into eburnamonine, an N-acylindole (XIX), on the other. Reduction of eburnamonine with lithium aluminum hydride regenerated the alcohols, XVII ( 7 , l l ) .When eburnamonine was heated with selenium

XVIII

XVII

21

XIX Eburnamonine

at 360" for 5 minutes, it gave 4-ethyl-4-propyl-4,5-dihydrocanthin-6-one (XX), [a],, + 36" (CHC13) picrate, mp 199"-200°, in almost quantitative yield. Prolonged heating of X X with selenium eventually gave the two possible canthin-6-ones X X I (Rz = E t or Pr), which were synthesized by condensation of diethyl oxalate and the dilithium derivative of the appropriate l-alkyl-,+carboline (XXIII; Rz = E t or Pr) followed by reductive removal of the 5-hydroxyl group in the product (XXII; R 1 = OH, Rz = E t or Pr). When X X was subjected t o prolonged reflux with sodium hydride in toluene, it extruded the acetyl moiety t o form Z-l-(l-ethylbutyl)-/3-carboline(XXIV), the racemic modification of which was readily synthesized. It should be noted that the optically active form cannot be prepared from the resolved amide under BischlerNapieralski conditions. A synthesis of racemic X X has also been accomplished by ring closure of the amide, XXV (12). It was hoped that this

254

W. I. TAYLOR

TABLE I THE

ALKALOIDS OF Hunteria eburnea"

XLVI

XLV

bH XLVII

Name

Melting point ("C)

A. Eburnamine group (7, l l ) b Isoeburnamine 217-220 CigHz4NzO 181 Eburnamine CigHz4NzO Eburnamenine 196 CrgHzzNz (picrate) 183 Eburnamonine CigHzzNzO

[.ID

+ 111" (CHC13) - 93' (CHC13)

+ 183" (CHC13)

+ 89" (CHC13)

Observations

XVII, axial hydroxyl XVII, equatorial hydroxyl XVIII XIX

B. Secocorynantheol group (32, 40) Burnamicine CzoHzsNzOz

198-200

C. Tertiary bases of unknown structure (40) Hunteramine 206-208 Cz6&4NzOio?

a

- 280" (CHC13)

?

XLVII

Indolic chromophore ; no carbonyl group ; possibly a glucoside

See footnote a of Table I1 (page 266) for bases of the Kopsinine group. Numbers in parentheses indicate references (see list).

11. PentUCerUS

AND EBURNAMINE

(Hunteriu)-VINCAMINE

255

TABLE I-(continued)

Name

Melting point ("C)

[a],

Observations

264-265

- 205" (CHC13)

285-290

- 199" (CHC13)

Indoline + 5-oxyindole chromophores ; 1 NMe; 1 COOMe Indolic

197-198

- 131" (CHC13)

Desacetylpicraline (47)

264-265

-

Yohimbol in Chapter 20

296-297

+ 101" (MeOH)

271-272

-567" (3:1, MeOH : HzO)

334

-

XLV

307-308

+ 105" (27.5%

XLV

D. Quaternary bases (27, 28) Yohimbol methochloride CzoHz7NzOC1 Dihydrocorynantheol methochloride CzoHzsNzOCl Akuammicine methochloride CziHz~iNzOzC1 Hunterburnine a-methochloride CzoHz7NzOzC1 Hunterburnine j3-methochloride CzoHz7NzOzC1 Huntrabrine methochloride CzoHz7NzOzC1 Hunteracine methochloride CzoHzsNzOCl

Corynantheol, see Chapter 20 Akuammicine, see Chapter 17

H20-MeOH) 285-287

+ 54' (HzO)

XLVI(?)

343-344

would lead to a solution to the absolute stereochemistry of this class of alkaloid by replacing the racemic a-ethyl-a-propylsuccinic acid with one of its resolved forms, but so far neither this acid nor its derivatives has yielded to attempts a t resolution (12). These results by themselves did not unequivocally establish the structure of eburnamonine (XIX); this was secured not only by a degradation of eburnamine (see Section 11, B) in which the uncertain features (ring D and the ethyl group) were not affected but also by a total synthesis of the alkaloid (Chart I). Condensation of /3-ethyl-/3-formyladipic acid with tryptamine gave in one step dl-eburnamonine lactam (XXVI) which,

256

W. I. TAYLOR

8 1

XXI

XXII

t

1. LiBu

2 . (COOEt)?

/

xxv

XXIII

XXIV

upon reduction with lithium aluminum hydride followed by oxidation of the resultant dl-eburnamines (XVII), afforded dl-eburnamonine (XIX), mp 203'-204" (7, 13). The desired aldehydodicarboxylic acid was prepared in four steps from p-ethylphenol, as outlined in Chart I. A second synthesis has been realized via an enamine synthesis (14). The condensation of ethylbromoacetate with XXVII gave the A 3 compounds XXVIII and XXIX. Sodium borohydride or catalytic reduction

XXVII

XXVIII

XXIX

/ XIX

xxx

11. Pentaceras

AND EBURNAMINE

OH

(Hunteria)-vINcAMINE

257

0

I

II

Et

Et

A

n

LiAlHa

+ XVII +

n

/

XXVI XIX CHARTI. A total synthesis of dl-eburnamonine.

of the latter gave a mixture of the DE trans compound, dl-3-epieburnamonine (XXX) and dl-eburnamonine (XIX) identical with the product of the first synthesis. Reduction of XXVIII by borohydride furnished only XIX. Eburnamonine can be hydrolyzed to an amino acid which recyclizes with great ease (7). This was thought t o be in favor of atrans DE system, but the foregoing results would indicate that some conformational aspects were not properly considered when the steric interactions in the two possible amino acids derived from X I X and XXX were analyzed. The tendency to cyclize must be quite strong since lithium aluminum hydride reduction of methyl eburnamoninate gave only the ring-closed alcohols (XVII) (7). dl-Eburnamonine (vincanorine) has been isolated (15) from Vinea minor L. ; its structure was recognized when selenium dehydrogenation was found to form dl-4-ethyl-4-propyl-4,5-dihydrocanthin-6-one (16). dl-Eburnamonine has been resolved into its antipodes by use of dibenzoyl tartaric acid as the active agent (16).

258

W. I. TAYLOR

B. EBURNAMINE, ISOEBURNAMINE, AND EBURNAMENINE Eburnamine and isoeburnamine are diastereoisomeric alcohols (XVII) whose structures followed from their chromic acid oxidation t o eburnamonine (XIX) and by Wolff-Kishner reduction t o d- 1,l-diethyl1,2,3,4,6,7,12,12b-octahydroindolo[2,3-a]quinolizine (XXXI), mp 106", [uID +93" (CHC13) (7). The last compound was catalytically dehydrogenated and rereduced t o furnish the racemic compound, mp 132".It was synthesized by condensing tryptamine with 4-ethyl-4-formylhexanoic acid and by reducing the resulting lactam (XXXII) with lithium aluminum hydride. Neither eburnamine (XXXIV) nor iso-

XXXI; R = H2 XXXII; R = 0

eburnamine (XXXIII)showed any properties, with the exception of the Wolff-Kishner reaction, which would indicate their being in equilibrium with the theoretically tautomeric aldehyde ; however, the reactions to be described imply the intermediacy of another tautomer, the iminium ion. The reactions are summarized in Chart I1 (partial formulas). If either eburnamine or isoeburnamine was allowed to stand a t room temperature in 0.5 N sulfuric acid for a period, it gave rise to a mixture of about 90% of the former and 10% of the latter and, if the solution was warmed briefly, eburnamenine was obtained. The driving force in this reaction probably lies in the relief of strain (1,3-diaxial interactions) in going from isoeburnamine to eburnamine. I n ethanolic picric acid, both alcohols gave one and the same 0-ethyleburnamine (Chart 11),mp 147O-148", [.ID + 64' (CHC13),which, if the solution was refluxed for a short period or the picrate itself was heated above it5 melting point, furnished eburnamenine in quantitative yield. Water was also eliminated from (iso)eburnamine methiodide(s) when it was crystallized from water. The above stereochemical conclusions were compatible with the NMRspectra (17, 18). I n agreement with these results, the alcohols were unaffected by sodium borohydride and only very slowly reduced to dihydroeburnamenine by lithium aluminum hydride because the reaction conditions

11. Pentaceras

AND EBURNAMINE

(Hunteria)-VINCAMINE

259

are poor for the formation of the iminium ion. It is for this reason that dihydroeburnamenine was best prepared by catalytic reduction of the olefin (7). Eburnamenine reacted readily with osmium tetroxide to form an amorphous glycol that, upon oxidation with chromic oxide, gave 15-hydroxyeburnamonine, which resisted further attack by the oxidant (7). “ 1

HO

N ‘’ EtO

HO (e) XXXIV Eburnainine

XVIII E burnamenine

CHART11. Interrelationships of eburnamine, isoeburnamine, and eburnamenine.

I n view of the foregoing described acid-catalyzed transformations and the fact that inorganic acids were used in the isolation of the alkaloids from the plant material, it is quite possible that the ratios of isoeburnamine to eburnamine and eburnamenine were quite different originally, even to the extent of the exclusion of the last two. By the use of mass spectroscopy coupled with deuterium labeling, eburnamonine, (is0 ? )eburnamine, and eburnamenine have been shown to exist (stereochemistry not specified) in Rhazya stricta Decaisine, and eburnamenine itself in Aspidosperma quebracho blanco Schlecht (19).The obtention of eburnamenine from Pleiocarpa species is described in Section 111.

C. VINCAMINEAND RELATED COMPOUNDS Under this heading is discussed the chemistry of those alkaloids, isolated from Vinca species (principally V . minor), that are closely related to eburnamine and its congeners. A table of Vinca alkaloids is given elsewhere (see Chapter 12). The principal alkaloid of V . minor is vincamine (XXXV; R = H) whose structure became clear (20) when it was found that treatment of

260

W. I. TAYLOR

the alkaloid with acid furnished I-eburnamonine (XXXVI; R = H), mp 174’, [a]= - 105’ (CHC13),the optical antipode of eburnamonine from Hunteria eburnea. Dehydrogenation of 1-eburnamonine gave 1-4-ethyl4-propyl-4,5-dihydrocanthin-6-one (16).

XXXVIII Apovincemine; R = H

XL

eoH

A

T I

CIIzNz

L

n

- 01 11 N HCl

RT

xxxv

\

R/\

t

dilution

n

a-.-

\N eJI \ / \

XXXIX

Vincamine; R = H Vincine; R = OMe

XXXVI I-Eburnamonine; R = H

XXXVII

CHART111. The properties of vineamine and vineine.

11-Methoxyvincamine (vincine, XXXV; R = OMe), whose ring A methoxyl is so placed because of its UV-spectrum and a color reaction specific for that position, had properties (Chart 111)that paralleled those of vincamine (21). This conclusion was reached independently on the basis of mass spectral comparisons (18). The stereochemistry of both alkaloids is regarded as being the same (21). Tetradehydro compounds (XXXVII) were formed by lead tetraacetate oxidation, and XXXVII

11. Pentaceras

AND EBURNAMINE ( f f U r L t e r i U ) - V I N C A M I N E

261

(R = OMe) with potassium isoamylate is said t o generate 7-niethoxy-jlcarboline in unspecified yield (21 ) . The hydroxyl group in XXXV could not be acetylated; instead, the N-vinylindole, the apo compound, XXXVIII was obtained. In fact, simple heating or solution of the alkaloid in strong acid was all that was necessary. The latter experiment, which finds an analogy in eburnamine itself, required the intermediacy of the iminium form (see Chart 11).This iminium form was apparently stable in strong acid, since vincamine in 11 N hydrochloric acid had a UV-spectrum with a long wavelength maximum a t 360 mp (log E = 3.85) compatible with XXXIX which reverted irreversibly into apovincamine (XXXVIII; R = H) upon dilution (18). I-Eburnamonine (XXXVI) has been produced by other reactions of vincamine. Oxidation of vincaminic acid (XL; R = H) by means of ammoniacal silver nitrate was one way, and periodic acid fission of vincaminol was another (16). A different group of workers, who had probably attempted to prepare vincaminol by lithium aluminum hydride reduction of vincamine, obtained instead I-eburnamonine in excellent yield (18). This has been rationalized as illustrated (partial formulas) by analogy with the base-induced decomposition of formic esters to carbon monoxide and alkoxide ion :

On the basis of the chemistry and NMR-spectra, the hydroxyl group in vincamine and 11-methoxyvincamine is regarded as being axially substituted (18). Some alkaloids of as yet undetermined structure may

Vincaminine (vincareine); R = H Vincinine; R = OMe

form further examples of the eburnamine system. The physical data on vincaminine (vincareine) (15, 22, 23, 24) and vincinine (22) are compatible with structures XLI (R = H) and XLI (R = OMe), respectively

262

W. I. TAYLOR

(now proved; see note added in proof Ref. 24), and thus resemble very closely, from a biogenetic point of view, the Aspidosperma /3-anilinoacrylates, 20-0x0-1-vincadifformine (XLII) and its 16-methoxy derivative, also isolated from F’inca minor (25).

D. MASS SPECTRAOF

THE

EBURNAMINE-VINCAMINE ALKALOIDS

Upon volatilization into the mass spectrometer, three alkaloidseburnamenine, eburnamine, and isoeburnamine-gave the same spectrum because of the facile loss of water from the last two alkaloids (18, 19). Apovincamine gave results that were equivalent to those of eburnamenine, if allowance was made for the extra 59 mass units (18). Eburnamenine, for example (Chart IV), showed a weak peak loss of the C-3 proton [cf. yohimbine class (as)], and the major primary fission was a retro Diels-Alder-like cleavage of ring C, then loss of one of the two chains ( “ a ” or “ b y ’split) to yield the major peaks mje 208 and 249. Eburnamonine split likewise (Chart IV), and in addition exhibited a further peak m/e 237 which was the mass of the fragment M-29 from which carbon monoxide had been ejected. These conclusions were confirmed by running the mass spectra of suitably labeled compounds prepared as follows : with lithium aluminum deuteride, eburnamine gave dihydroeburnamenin- 144, and eburnamonine gave dihydroeburnamenin-la-d,, and the oxygen of eburnamonine was exchanged for 0 1 8 (19). Dihydroeburnamenine behaved analogously to eburnamonine, and its two most prominent peaks apart from M and M-1 were M-29 (lossof ethyl) and M-70 ( “ a ” split) (19). Vincamine showed peaks from M, M-1, M-CH3, M-CzHs, M-COOMe, M-H20, and M-HCOOCH3. Since the last two peaks were equivalent to apovincamine and eburnamonine, it was not surprising to observe their characteristic fragmentation peaks in the line-rich spectrum. 11Methoxyvincamine behaved similarly to vincamine, except that the aromatic peaks were displaced by + 30 mass units (18).

III. The Hunteria and Pbiocarpa Alkaloids The alkaloids of Hunteria and Pleiocarpa are included here because they were early sources of the eburnamine-type alkaloids. The bases so far described from Hunteria eburnea Pichon (Table I) are made up of a large number of quaternary bases (27), the structures of five of which have been elucidated (27, 28). Interestingly, none of these

263

11. Pentucerus AND EBURNAMINE (HUnteriu)-VINCAMINE

208

B

193

R/k+/&/-

294

224

206

i. 265

237

CHARTIV. Principal electron-impact products of eburnamenine, apovincamine, and eburnamonine. The mi13 is given for R = H (eburnamine or eburnamenine); for R = COOMe, add 59 mass units (apovincamine).

2 64

W. I. TAYLOR

quaternary bases was a quaternary derivative of the co-occurring tertiary bases. I n fact, the quaternary bases whose structures were determined were derived from the yohimbinoid precursor XLIII, whereas the tertiary bases (and the Aspidosperma group) originated from XLIV (1).Does this mean that the quaternary bases are not derived

'h

(HOOCA

XLIII

YJJ

(HOOC)

XLIV

from their immediate tertiary precursors Z This is an important question which remains to be answered. An interesting pair of alkaloids, whose structures were derived by X-ray crystallographic analysis, turned out' to be epimeric N,-methylated quaternary salts (XLV; see Table I)of the as yet to be isolated tertiary base hunterburnine (28). This is not only the first case of the isolation of this new yohimbinoid variant, but also of such Nbepimers. The natural occurrence of such isomers may turn out to be quite common. It is also possible that the biochemical methylation step, if there is one here, may only be as specific as that of the analogous laboratory operation. It is known, for example, that treatment of yohimban with methyl iodide gives both diastereoisomeric methiodides (29). The absolute stereochemistry indicated for XLV is based on that expect,ed (30)if the asterisked carbon is equivalent to C-15 of yohimbine. Huntrabrine methochloride has an ethylidene group, a quaternary N-methyl, a 5-hydroxyindole chromophore, and a primary hydroxyl group. It underwent a facile Emde degradation, and the resulting tertiary base upon tosylation gave a phenolic-0-tosylate quaternary tosylate. The latter compound, upon selenium dehydrogenation, afforded uncharacterized products, one with a sempervirine-like UV-spectrum and another with a 2-pyridylindole chromophore. All these results are interpreted as supporting the structure XLVI (see Table I) for huntrabrine methochloride (27). Six further alkaloids were isolated in quantities insufficient for complete or accurate characterization and were given the alphabetical designations F, H (a pseudoindoxyl), I, J, K, and N (27). The total crude alkaloids had hypotensive activity (31) traced largely to the quaternary bases (27), but the only pure active compound isolated was hunterburnine a-methochloride (28). The only tertiary base aside from the eburnamine group whose structure has been established is burnamicine (XLVII; see Table I). It is a 2-acylindole, which in acidic solution existed as an indole. Its formula

11. Pentaceras AND

EBURNAMINE

(Hunteria)-vINcAMINE

265

followed from an analysis of its mass spectrum (32), but the amount available was insufficient for a classical proof, such as its conversion to the tetracyclic quaternary salt corynantheol methochloride. Whether XLVII is a natural product is a question still to be answered but, up to the present time, the conversion of XLVI-type quaternary salts into XLVII-type tertiary bases has not been realized. From Hunteria corymbosa Roxb., a crystalline alkaloid has been reported (33), but no physical constants were given. More recently, a reinvestigation has resulted in the obtention of corymine, mp 189"-192', [aJD + 2 7 O (CHClS), for which the constitution XLVIII has been proposed (34). A continuing examination of Pleiocarpa mutica Benth. and P. tubicina Stapf has revealed the alkaloids summarized in Table 11. Although extracts of these plants have shown prolonged blood pressure-lowering properties (41)) the active component(s) has not yet been identified.

The principal bases are all derivatives of kopsinine (L; see Table 11) and are closely related t o the Aspidosperrna alkaloids, aspidofractine [ 1-formylkopsinine (42)], refractine [ 17-methoxy-1-formylkopsinine (42)], pyrifoline [6,16-dimethoxy-l-acetyldecarbomethoxykopsinine (43)], and refractidine [6-methoxy-l-formyl-decarbomethoxykopsinine (43)]. The structures of all these compounds have been derived from a consideration of the physical data and the mass spectra (most important single tool) of suitable derivatives. The group has not yet been completely interrelated and none have been degraded (in any classical sense) to known compounds. The isolation of the 10-0x0 compounds is of particular interest (36). In a very detailed piece of work, the absence of %ox0 compounds in the crude alkaloids has been proved and, since the parent bases were shown to be stable to light and air even in chloroform, the lo-0x0 compounds cannot be artifacts (36). That the 0x0 group should turn up at the 10 position is in complete agreement with the commonly accepted view that tryptophan is the source of the tryptamine residue in the complex indole

266

W. I. TAYLOR

TABLE I1 THE ALKALOIDS OF PkiOCUTpa ntUtiCU (m) AND P . tubicina (t)

Name A. Kopsinine group Kopsinine 10-Oxokopsinime (kopsinilam)" 1-Methylkopsininea [Pleiocarpinine ; pleiocinine (38)]

Melting [a]= point (" C) (CHC13)

105

Source

-69' m and

135-136

- 124'

10-0x0-1-methylkopsinine 249-250 (pleiocarpinilam )' l-Carbomethoxykopsininea 141-142 [pleiocarpine ; pleiocine (38)]

- 53"

COOMe

C. Other alkaloids Pleiomutine Amorph. [C~Z-~~H~Z-~ ?)]B N ~ O Z (

(L) Kopsinine

- 145"

B. Eburnamine group Eburnamenine

'

97"

Pleiomutinine " & o H ~ ~ - ~ s N?)I ~OZ( Pleiocarpamine (CzoHzzNzOzb)

> 220 159

+ 123"

Pleiocarpinidine (formula unknown) Flavocarpine (CI~HI~NZOZ) Caffeine

154-156

?

307-308

+O"

a

Observations

?

m and t(36) m(35, 37) t(36)

See Table I, p. 254

435)

Indoline +indole chromophore; 1 COOMe; 1 NMe UV = p-anilinoacrylate

m(35)

1,2,3-Trisubstituted indole; 1 COOMe; no NMe Indole

m(35)

435)

XLIX

439)

m(35)

These bases also occur in Hunteria eburnea (35, 36).

' Revised formula; this base also occurs in Hunteria eburnea (40).

alkaloids. If it is assumed that the carboxyl group is not lost during the initial condensations, then it can either be retained, eg., in harman-3carboxylic acid from Aspidosperma polyneuron (44)and in adifolinefrom Adina cordifolia (45), or oxidatively removed [by a process equivalent to the well-known conversion of -CH(NHz)COOH to -CHO]; then, the resulting iminium salt can be either reduced to yield familiar alkaloids

1 1. PentUCera8

AND EBURNAMINE (Hunteria)-VINCAMINE

267

or further oxidized to lactams. If this occurs in the biosynthesis of indoline alkaloids, the product could still have a weakly basic N, but, in cases such as the yohimbinoid alkaloids or pleiocarpine, the product would be neutral and therefore very difficult to isolate. In a concise and complete paper flavocarpine (XLIX), a yellow amino acid with a sempervirine-like UV-spectrum has been recognized in P. mutica, its structure deduced, and synthesized (39).

REFERENCES 1. E. Schlittler and W. I. Taylor, Ezperientia 16, 244 (1960). 2. Sir Robert Robinson, “Structural Relations in Natural Products,” Clarendon Press, Oxford, 1955. 3. H. F. Haynes, E. R. Nelson, and J. R. Price, AustralianJ. Sci. Res. 5, 387 (1952). 4. J. R. Cannon, G. K. Hughes, E. Ritchie, and W. C. Taylor, Australian J . Sci. RES.6 , 86 (1953). 5. H. B. MacPhillamy, Personal communication. 6. G. Hahn and A. Hansel, Ber. 71, 2163 (1938). 7. M. F. Bartlett and W. I. Taylor, J . Am. Chem. SOC.82, 5941 (1960). 8. E. R. Nelson and J. R. Price, Australian J . Sci. Res. 5, 563 (1952). 9. E. R. Nelson and J. R. Price, AustralianJ. Sci. Res. 5 , 7 6 8 (1952). 10. N. Inamoto, S.Masuda, 0. Shimamura, and T. Tsuyuki, Bull. Chem. SOC. Japan 34, 888 (1961). 11. M. F. Bartlett, W.‘I. Taylor, and Raymond-Hamet, Compt. Rend. Acad. Sci. 249, 1259 (1959). 12. W. I. Taylor, Unpublished results. 13. M. F. Bartlett and W. I. Taylor, Tetrahedron Letters No. 20, 20 (1959). 14. E. Wenkert and B. Wickberg, 17th Natl. OrgnnicSymp. Am. Chem. SOC.Bloomington, Indiana, June, 1961. 15. J. Mokrj., I. Kompii, 0. Bauerova, J. Tomko, and S. Bauer, Ezperientia 17,354 (1961). 16. J . Mokrj., I. KompiB, and P. SefEoviE, Tetrahedron Letters No. 10, 433 (1962). 17. J. N. Shoolery, Personal communication. 18. M. Plat, D. D. Manh, J. Le Men, M.-M. Janot, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi, Bull. Soc. Chim. France p. 1082 (1962). 19. H. B.Schnoes, A. L. Burlingame, and K. Bieman, Tetrahedron Letters p. 993 (1962). 20. J. Trojanek, 0. Qtrouf, J . Holubek, and Z. Cekan, Tetrahedron Letters p. 702 (1961). 21. 0. Strouf and J. TrojBnek, Chem. I d . ( L o d o n )p. 2037 (1962). 22. J. Trojanek, 0. Strouf, K. KavkovL, and 2. Cekan, Chem. I n d . (London)p. 790 (1961). 23. J. Mokrjr, L. DGbravkovQ, and P. SefEoviE, Ezperientia 18, 564 (1962). 24. J. TrojBnek, 0. Strouf, K. Kavkova, and Z. Cekan, Collection Czech. Chem. Commun. 27, 2801 (1962). 25. M. Plat, J. Le Men, M.-M. Janot, H. Budzikiewicz, J . M. Wilson, L. J. Durham, and C. Djerassi, BUZZ.SOC.Chim. France p. 2237 (1962). 26. L. D. Antonaccio, N. A. Pereira, B. Gilbert, H. Vorbrueggen, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi,J. Am. Chem. SOC.84, 2161 (1962). 27. M. F. Bartlett, B. Korzun, R. Sklar, A. F. Smith, and W. I. Taylor J . Org. Chem. 28, 1445 (1963).

268

W . I. TAYLOR

28. J. D. M. Asher, J. M. Robertson, G. A. Sim, M. F. Bartlett, R. Sklar, and W. I. Taylor, Proc. Chem. SOC.p. 72 (1962); C. C. Scott, G. A. Sim, and J. M. Robertson, PTOC.Chem. SOC. p. 355 (1962). 29. B. Witkop, J. Am. Chem. SOC.71, 2559 (1949); B. Witkop and S . Goodwin, ibid. 75, 3371 (1953). 30. E. Wenkert and N. V. Bringi, J. Am. Chem. SOC.80, 3484 (1958). 31. Raymond-Hamet, Compt. Rend. Acad. Sci. 240, 1470 (1955); A. Engelhardt and H. Gelbrecht, h’aturwissenschaften 45, 547 (1959); Arzneimittel-B’orsch. 11, 414 (1961). 32. M. F. Bartlett and W. I. Taylor, J. Am. Chem. Soc. 85, 1203 (1963). 33. M. Greshoff, Ber. 23, 3537 (1890). Chem. SOC.p. 298 (1962). 34. A. K. Kiang and G. F. Smith, PTOC. 35. W. G. Kump and H. Schmid Helw. Chim. Acta 44, 1503 (1961). 36. C. Kump and H. Sehmid, Helv. Chim. Acta 44, 1090 (1962). 37. A. R. Battersby and D. J. Le Count, J . Chem. Soc. p. 3245 (1962). 38. F. A. Hochstein and B. Korol, Am. Chern.Soc., 140thMeetin9, Chicago, 1961, p. 11-0. 39. G. Biichi, R. E. Manning, and F. A. Hochstein, J. Am. Chern. SOC. 84, 3393 (1962). 40. M. F. Bartlett, R. Sklar, A. F. Smith, and W. I. Taylor,J. Org. Chem. 28, 2197 (1963). 41. Raymond-Hamet, Compt. Rend. Acad. Sci. 244, 2991 (1957); D. P. N. Tsao, J. A. Rosecrans, J. J. De Feo, and H. W. Youngken, Jr., Econ. Botany 15,99 (1961). 42. C. Djerassi, T. George, N. Finch, H. F. Lodish, H. Budzikiewicz, and B. Gilbert, J . Am. Chem. Soc. 84, 1500 (1962). 43. B. Gilbert, J. M. Ferreira, R. J. Owellen, C. E. Swanholm, H. Budzikiewicz, L. J. Durham, and C. Djerassi, Tetrahedron Letters p. 59 (1962). 44. L. D. Antonaccio and H. Budzikiewicz, Monutsh. Chem. 93, 962 (1962). 45. A. D. Cross, F. E. King, and T. J. King, J . Chem. SOC.p. 2714 (1961). 46. N. Neuss and N. J. Cone, Ezperientia 16, 302 (1960). 47. W. I. Taylor, M. F. Bartlett, L. Olivier, J. LBvy, and J. LeMen, Bull. Soc. Chim. France p. 392 (1964).

---CHAPTER

12-

THE VINCA ALKALOIDS W. I. TAYLOR Resenrch Department, C I B A Pharmaceutical Company, Division of C I B A Corporation, Summit, New Jersey

....................................... ....................................... 111. The Alkaloids of Vinca difformis Pourr. and Ti. major L. . . . . . . . . . . . . . . . . . . IV. The Alkaloids of Vinca herbacea and V . lancea. .......................... References .......................................................... 1. The Alkaloids of Vinca rosea L..

272

11. The Alkaloids of Vinca minor L.

278 280 282 282

The synonyms for the genus Vinca are Lochnera, Pervinca, and Catharanthus; since the botany of the plants does not concern us here, the term Vinca will be used throughout, although Pichon (1)has shown that there are considerable morphological differences between the first two genera mentioned. There is no doubt that the very interesting results being accumulated on these alkaloids will be of use to the systematic botanist, and a cursory glance at Tables I (alkaloids of V . rosea), I1 (alkaloids of V . minor), and I11 (alkaloids of V . dijformis and V . major) will bear this out. The plants of the genus Vinca described here represent only a few of the hundreds of members of the family Apocynaceae examined during the renaissance in indole alkaloid chemistry sparked by the discovery of reserpine in 1952. Because of the interesting pharmacological activity of its alkaloidal fraction, V . rosea L. has been subjected to an examination every bit as intense as has been recorded for the analgesics of the opium poppy, the smooth-muscle stimulants of ergot, the muscle relaxants of curare, and the hypotensive-sedative agents of RauwolJia. It is because of the power of modern chemical research, with its refined isolation and separatory methods, its application of physics to structure analysis, and its support from industry that a lifetime of classical research has been compressed into 7 years. I n fact, it is almost no longer a challenge to isolate and determine the structure and stereochemistry of a relatively simple natural product. The real problems today lie in the total analysis of a plant extract, in the determination of the biochemical significance 269

TABLE I THEALKALOIDS OF Vinca roo8ea

Name (Isolation reference)

Formula

Melting point (" C)

Maximum wavelength [a], (CHC13)

PK', (DMF)

A. Yohimbinoid bases Ajmalicine (16) [vinceine (17), vincaine (IS)] Serpentine (16) Alstonine (19, 20) Tetrahydroalstonine (20) Reserpine ( 21)

Sitsirikinen (22) I _

(kn&X)

Observations

For structures, see appropriate chapters

Cz1H~eNz03. iHzS04

239-241 (dec.)

+23" (Base)

7.6 (66%)

CzoHz4Nz03 CZIHZ~NZOZ CzzHdz04

211-214 (dec.) 190-193 (dec.) 168-169 (dec.)

+608"

5.5 (66%) 4.2 (66%)

Not isolated by the major investigators (6) Yohimbine isomer ( ?)

-

B. Strychnoid bases Lochneridine (22; Lochnericinea (23, 24) Lochnerinine" (25) Akuammine (vincamajoridine) (26, 54)

- 432' -424'

_ _ ~ ___ C. Sarpagine-like bases Lochnerine (16, 26) Perivine" (4, 31)

__

CZOHIZNZO3*

180-181

- 121'

7.5 (66%)

126-128

+ 30"

6.8 (66%)

D. Iboga bases Catharanthine (23)

Structure I ( 3 2 ) ; all three have UV = 8anilinoacrylate Not isolated by the major investigators (6) See appropriate chapter UV = 2-acylindole Structure 11;see appropriate chapter

E. Aspidosperma-type Vindoline (8, 23)

CZ~H~ZNZO~

154-1 55

- 18

5.5 (66%)

Vindolinine (23, 27)

CzlH24Nz0z.2HC1

210-212 (dec.)

3.3, 7.1 (66%)

Vindolidine (25)

Cz4H3oNz05

167

-8" (HZO) - 31"

+42" +26O (CH3CHClz) -58' 72" 61" +56" 42" 61" -t6' - 54" 35"

5.4, 7.4 (HzO) 5.0, 7.4 (33%)

Structure 111 (28); see appropriate chapter Structure IV (29)

-

Note conflict of name under G

F. Aspidosperma-iboga dimers 211-216 (dec.) 218-220 (dec.)

Vincaleukoblastine (12) Leurocristine (5)

C46H56N409 C46Hd40io

Neoleurocristine (30) Leurosine (4) Isoleurosine (22) Leurosidine (5) Neoleurosidine (30) Pleurosine (30) Carosine (30) Catharine (22) Catharicine (30'

188-196 (dec.) 202-205 (dec.) 202-206 (dec.) 208-211 (dec.) C48H6zN40iio 219-225 (dec.) C46H56N40ioa 191-194 C46H56N40ioQ 2 14-2 18 C46H52N409.CH30H" 271-275 (dec.) C46HazN40ioa 231-234 (dec.)

C46H56N4012~ C46H58N409 C46HsoN40gS

+

4.68 (33%) 5.5, 7.5 (HzO) 4.8, 7.3 (66%) 5.0, 8.8 (33%) 5.1 (33%) 4,4, 5.5 (33%) 4.4, 5.5 (33%) 5.34 (66%) 5.3, 6.3 (33%)

- 48"

5.4 (66%)

- 113"

+ + + +

214, 259 220, 255, 296

Structure V Structure V

214, 259 214, 261, 287 214, 265 214,268 267, 308 255, 294 222,265,292 214, 268, 293, 315

StructureXIV(?)

G. Other dimers Vindolicine (22) Vindolidine (30) Vincamicine (22)

C ~ O H ~ ~ N ~ O I ? 248-251 265-276 (dec.) C48Hs4N4010 244-250 (dec.) Analytical data (22) 224-228 (dec.)

Vincarodine (30)

C44H5zN40io

253-256 (dec.)

- 197'

5.3 (33%) 4.80, 5.85 (66%) 214,264, 315, 341 5.8 (66%) 230, 272, 298

Carosidine (30)

-

263-278 and 283 (dec.)

- 89"

-

E14, 254,303

C&IzeNz04 (31)

257-262 (dec.)

-160"

5.85 (66%)

226,270

+418'

H. Unassigned Virosine (4)

Provisional formulas and class assignments.

CzoHa4Nz03 in all major reviews.

Possibly symmetrical dimers of vindoline (6)

Related ( ? ) to vincine (6)

272

W. I. TAYLOR

of the individual components, and in the facile stereospecific synthesis of new molecular combinations.

I. The Alkaloids of Vinca rosea L. The importance of the alkaloids of V . rosea (more correctly Catharanthus roseus G. Don) lies in the detection and isolation (2-5) of potent antileukemic drugs. One of these, vincaleukoblastine, is now being used clinically, and a brief summary of these results has been presented (6). Although the plant has enjoyed a reputation as an oral hypoglycemic agent (7), this has not survived scientific examination ( 2 , 8). The total alkaloids possess a limited antibiotic activity along with a sustained hypotensive action (8). These and other pharmacognostic aspects have been described in greater detail elsewhere (9). The alkaloids isolated from V . rosea are listed in Table I; the wellknown yohimbinoid bases, including lochneridine (I)and lochnericine, will not be discussed ; catharanthine (11)and lochnerine (O-methylsarpagine) belong to the iboga and ajmaline-sarpagine chapters, respectively. And vindoline (111)and vindolinine (IV) are special cases of

I

I1

Aspidosperma-type molecules whose structures were derived largely from a detailed physical examination of the bases and their derivatives (principally by mass spectra; see Chart 111), a technique discussed on a broader basis under Aspidosperma. Similarly, the structures of some of the strychnoid bases were obtained by physical methods. There is no way of knowing which of the alkaloids listed in the table could be identical with either the amorphous bases of Greshoff (10) or the base tartrates and sulfates reported by Australian workers (11). The most interesting and important bases are the dimers represented by vincaleukoblastine (V). By very careful classical analyses, coupled with the full use of UV-, IR-, and NMR-spectroscopy, it was concluded that vincaleukoblastine, as well as leurosine, were indole-indoline dimers, most probably catharanthine linked to vindoline ( 1 2 ) . Leurocristine

12.

THE

VinCU

273

ALKALOIDS

( V; R = CHO) was correlated to vincaleukoblastine by reducing them both to the same pentahydroxy derivative (13). That at least vindoline made up the indoline portion of the dimers investigated was proved as a

111

IV

\'indoline

\'indolinine

result of a very important study. When the dimers, vincaleukoblastine (13), Ieurosine (13), leurosidine ( 6 ) , or isoleurosine ( 6 ) , were refluxed in concentrated hydrochloric acid in the presence of a reducing agent (Sn/SnClz), each gave rise to an indoline [deacetylvindoline (VI ; R = CH3) in all cases examined] and an indole [velbanamine (VII) in the first case, cleavamine (VIII) from the second, and unidentified indoles from the remainder]. In agreement with its established relationship to vincaleukoblastine, leurocristine under the same acidic conditions gave demethyldeacetylvindoline (VI ; R = H) and velbanamine (VII). The structure of cleavamine has been proved (14) beyond doubt, and its properties and preparation (15) from catharanthine (IX)'via X (see Chart I) by concentrated hydrochloric acid and a reducing agent (Sn/SnClz) are discussed more fully in the iboga chapter. The structure of velbanamine has not been established, but it is believed to be a hydroxydihydrocleavamine (13). The attachment of the indole moiety t o vindoline in these dimers has not been completely solved. It is not an ether linkage (cf. voacanga dimers) nor does it involve an aldehyde (cf. geissospermine). In vincaleukoblastine (and therefore in leurocristine), one end of the linkage is a t (2-15 of vindoline, a conclusion arising from an NMR-comparison of the differences between the number and coupling of the aromatic protons in the dimer, 15-bromovindoline and vindoline itself (13). Thus, in vincaleukoblastine, besides the four coupled proton resonances of the indole, there are two weakly coupled protons on the indoline residue, the same as in the para-substituted pair of 15-bromovindoline (13). Vincaleukoblastine has two hydroxyls; the one at (2-3 is hydrogen bonded, and the other in the indole is probably tertiary, since after acetylation there was observed in the NMR-spectrum no down-field

274

W. I. TAYLOR

V Vincaleukoblestine; R = Me Leurocristine; R = CHO

VI

VII Velbanamine

tz conc. HCI f---

Sn/SnClr

Leurosine

conc. HCl Sn/SnCIa

Leurosidine Isoleurosidine

H VIII Cleavamine

I

COOMe X

IX Catharanthine

CHARTI. Interrelationships among vincaleukoblastine, leurocristine, leurosine, velban-

emine, cleavamine, and catharanthine.

12.

THE VinCU ALKALOIDS

275

shift of the proton chemical shifts that would be expected (6, 13) after the conversion of a >CHOH

+

\ /

CH.OAc

Because of these results, the current formulations (V, Chart, I) for vincaleukoblastine have either a C-15 to C-4’ linkage (site of hydroxyl is then a mystery) or a C-15 to C-3’ linkage with the hydroxyl on C-3’ or C-4’ (stereochemistry not specified). These conclusions are acceptable, provided that the C-ethyl of the indole residue pre-exists in the dimer, but the only proof cited is the statement that the NMR is in agreement with structure V (13). The NMR argument should have excluded the possibility that an 18-carbomethoxydihydrocleavamineresidue, rather than the 18-carbomethoxyibogamine skeleton, formed the indole moiety of the dimer. If V (Chart I) is to be accepted as true, then the most plausible linkage is

and the fission can be looked at in two ways, keeping in mind that the yields of the isolated monomers are small. I n the first, the fission could be regarded as a reversal of a Friedel-Crafts condensation, followed by reduction of the ketone, then retro Mannich (C-5’ to C-18’ split) and reduction (cf. catharanthine --f cleavamine, Chart I) to yield velbanamine (VII; C30H). A second pathway, however, fits better into the known chemistry and makes the intermediate X I (Chart 11)the common one for all the ring-opening and fission products that have been isolated. If X I is reduced (no such compound has yet been isolated, but it could be one of the unknown dimers), elimination of the C-18’ carboxyl could take place but not cleavage of C-3’ to C-15; however, XII, the tautomeric enamine, could be split to yield deacetylvindoline (VI) and XIII. XI11 by reduction could form velbanamine (VII, Chart 11).If a steric requirement is added to this idea, it may provide an explanation also for the formation of cleavamine and make the difference between vincaleukoblastine and leurosine a matter of the stereochemistry at C-3’; i.e., if leurosine is XIV, hydroxyl extrusion (XV --f XVI) could precede the cleavage of the C-3‘ to C-15 bond (Chart 11). In the preceding discussion, the stereochemistry of the C-4’ ethyl has not been mentioned, although it is an additional source of isomerism in the dimers. I n the absence of experimental data, it is arbitrarily assigned a /3-configuration.

KJ

TABLE I1

l

u a

TEE ALKALOIDS OF V'inca minor

Formula Name A. Vincamine group Vincamineb [minorine (33)]

Melting point (" C)

231-232

0bservat ions

[XI,

Reference"

(34) [351 R3 = H ; R1= OH; R Z= COOMe (36) ~ 4 0 , 4 1 1

CziHzeNz03 39" (PY)

+

1 1 -Methoxyvincamine (vincine) rae -Eburnamonine (vincanorine ) Vincaminine" [vincareine (42)]

(42) [431 R3 = H; RlRz = 0 (44, 45) R1 = O H ; Rz = COOMe; R3 = 0 (44) 11 = OMe; R1= OH; Rz = COOMe; R3=O

VincinineC

B. Other indoles Vincaminorine 130-131

Vincadine 70-85

Vincaminoreine ( N ,-methylvincadine)

126

CzzH30NzOz +46" (EtOH) Cz~HzsNzOz +92" (EtOH) CzzH3oNzOz

COOMe

?

AOOMe

(47)

C. 13-Anilinoacrylates,Aspidosperma class 1-Vincadifformine CziHz6NzOz rac-base (86) - 540' (EtOH) 20-0x0-1-vincadifformine CziHz4N~03 (minovincine) B.HC1, 192 -480" (EtOH) 20-Hydroxy-1-vincadifformine CziHz6Nz03 (minovincinine) -418" (EtOH) 16-Methoxy-20-0x0-1CzzHz6Na04 -414" (EtOH) vincadiff ormine rac- 1-Methylvincadifformine CzzHzsNzOz (minovined) 79-80 & 0" Vincorine (16-methoxyCZZHZSN203 - 142" (EtOH) 1-vincadifformine" 93-94

D. Other compounds Vincamidine

CzoHd203 78-80

Vincaminoridine

CZ~H~ZNZO~ 58" (CHC13) CziHd203 159-160 - 158" (CHC13)

99-100 Vincoridine Reserpine

a

-+

COOMe Vincadifformine

COOMe and - OH UV A,,, 260 m p (3.74) N-Methyl-5-methoxy indole; also contains a COOMe Indoline; also has a COOMe Detected only by paper chromatography

(48) (39) (39) (50)

Reference dealing with the isolation of the alkaloids are given in parentheses; those referring to structure appear in brackets. Perivincine (37) and isovincamine (38) were shown (36) to be mixtures of vincamine and vincine. Suggested structures based on the present summary (proved, see footnote in Reference 46). Structure also by J. Mokr9, I. Kompig, L. DfibravkovB, and P. SefEoviC, IUPAC Meeting, Prague, August, 1962, and Reference 86. c3 -4 -J

278

W. I. TAYLOR

QqAT$ H

I

MeOOC

9’-cl5

MeOOC

0

OH

MeOOC

OH

H’ V Vincaleukoblastine ( ? ) (leurocristine)

XI1

XI

I MeOOC

OH

XI11

VII Velbenemine ( ? )

OH MeOOC XIV Leurosine ( P )

CIS

MeOOC

C1s

xv

MeOOC

XVI

MeOOC

C1s

VIII Cleavamine

CHART 11. Suggested position of the dimer linkage and course of the cleavage in vincaleukoblestine and leurosine (partial formulas).

11. The Alkaloids of Vincu minor L.

The alkaloids so far isolated from V . minor (Table 11) are quite different from those of the two other well-investigated species, V . rosea

I-

N

\

OOMe

\N

R

COOMe

CHART111. Important electron-impactfragmentation products for Aspidospemna-type bases (R, R', and R" unspecified).

280

W . I. TAYLOR

and V . difformis. I n fact, with the possible exception of the detection of reserpine, no yohimbinoid bases have been found. All the known V . minor alkaloids seem to be derived from tryptophan and the same formal precursor, indicated by the thicker lines in XVII (R2= COOH) (see Table 11). The chemistry of group A compounds is discussed in Chapter 11, on Pentaceras and eburnamine-vincamine alkaloids. The structures of group C compounds have been solved largely as a result of a stillcontinuing mass spectrometric examination of indole alkaloids that had as its starting point the study of aspidospermine (51). The characteristic fragmentation for this type of pentacyclic system is illustrated in Chart 111. The group B indoles have hydrogen contents, which would indicate that they may be tetracyclic. This has been confirmed, and vincadine turned out to be a carbomethoxyquebrachamine ( 8 7 ) , since upon heating in hydrochloric acid, d-quebrachamine is obtained in 90% yield (cf. conversion of voacangine to ibogaine). Decarbomethoxylation of vincaminoreine gave an indole identical in all respects with the AT,-methylation product of d-quebrachamine (87). It should be noted that in the group C bases, just as for the eburnamine-vincamine group, compounds turn up with different stereochemistries ; thus, in 8. minor there is also found the levo form of racemic vincadifformine first isolated from V . difloormis (Table 111). S' ince racvincadifformine has recently been reported (86) from V . minor, it is just possible that it might be an artifact produced from the true natural product, the levo form, during its isolation. Crude extracts of V . minor have shown a hypotensive effect in experimental animals (76) which is believed to be owing almost entirely to its vincamine content (77), and success has been claimed for clinical studies in middle Europe of purified alkaloid extracts (78).Besides this property, vincamine also has a curare-like property (79) and is known t o induce strychnine-like convulsions in rabbits (80). The ability of vincamine to reduce blood sugar in an acute test in rats is apparently not because of the inhibition of hepatic glycogenolysis (81). The medicinal uses of V . minor have recently been reviewed (82).

111. The Alkaloids of Vinca digormis Pourr. and V. major L. These plants are simple varieties of Vinca major L., so that 8.difformis Pourr. i s more properly named V . major L. var. difformis (Pourr.)Pich., and V . major L. ( V . pubescens Urv.) as V . major L. var. major (1). A glance at Table I11 will show that these varieties do produce different

TABLE 111 THEALKALOIDS OF Vincn difformis (d)A N D V . mcrjor ( m ) Formula

~-

Name ~

__-

Melting point (" C ) ~~

- 1090

242

Akuamrnine [vincamajoridine ( 5 2 ) ]

2.58-260

Sarpagine

(Py)

CzzHz6NzO4 - 104" ( P y ) CiqHzzNzOz

z 360 Akuamm itline

- 55"

(Py)

241

CziHz4Nz03 24" (MeOH)

315

-

+

A base (tetraphyllicine?) Vincamajinc 226

0 -Acetylviiicamajinc (vincamedine) rue-Vincadifformine

186

124-125

Vincamine [vincamirine (56); pcrivincine ( 6 i ) l Vinine

.

~~~~~

~

CzzH26Nz04

CzzHz6NzO3 - 55" (alc.) CZ~HZENZO~ - 66" (CHCl3) CziHz6NzOz

Referencea

Observations

[UID ~~~

Itcserpininc

~~~~~~

~~~~

.

Yohimbinoid ring I3 oxygen heterocycle; see appropriate chapter i-Hydroxy-S-methylindolinr of strychnoid t y p e ; see appropriate chapter See ajmaline-sarpagine chaptnr

See ajmaline.sarpagint?ie chapter See ajmaline-sarpagine chapter See ajmaline-sarpagine chapter Scc ajmaline-sarpagine chapter See Aspirlosprr?nrc chapter

f0"

See Table 11, p. 276

.

m(52)

m(54) [55?]

d(56) [6i]

d(58) [59, 601 m(52) d(61), rn(62)

[59] d(63), m(62) [RS] d(58)

[641 d(56, 36),

m(6i) m(65)

Pubcscine

m(65)

A base

m(65)

Vinramarjoreine

m(B6)

Alkaloid X

d(58, 61)

" References dealing with the source of t h e alkaloids appear in parentheses; those referring t o structure

appear in brackets.

2 R

282

W. I. TAYLOR

alkaloids ; note in particular the occurrence of rac-vincadifformine in V . difformis alone. The structures of most of these alkaloids are discussed elsewhere under appropriate headings. The bases vinine, pubescine, and an alkaloid, mp 194"-195', have not yet been reisolated. If the data on vincamarjoreine are correct, its structure may be novel, since it shows no carbonyl band in the IR-spectrum and both its oxygens are accounted for in methoxyl groups. The crude alkaloids of V . major var. major produce, by intravenous injection, a marked fall in the blood pressure of dogs (83) which is apparently not owing t o their reserpinine or vincamine content (67). A phytochemical investigation has also been reported (84). IV. The Alkaloids of Vinca herbacea and V. lancea From Vinca herbacea Waldst et Kit. var. libanotica (Zucc.) Pich. ( V . erecta Regel et Schmalh.), reserpinine (68) and vincamine (69, 33) have been isolated along with a number of other bases of unknown structure. These are vincanine (69, 70), C~gH22N20,mp 188", [E]D - 992' (MeOH), whose IR-spectrum is said to indicate carbonyl and NH groups; vincanidine (68, 70), C20H24N203, mp 250'-280" ; and erectine (71), CzzHzsN202 (bisnitrate, mp 224O), which has an indoline nucleus, one methoxyl, and two active hydrogen groups. From a V . herbacea (variety unknown) growing in Bulgaria, still another base was obtained, herbaceine, C24H32N206, mp 144O, [a]= - 219' (Py). It is an indole, takes up under catalytic conditions one mole equivalent of hydrogen, and has four methoxyl groups (72). Vinca Zancea, more properly named Catharanthus Zanceus (Boj. ex A.DC.) Pich. and indigenous to Madagascar, has been shown to contain yohimbine (73), ajmalicine (74), tetrahydroalstonine, and a new base (75) with a /3-anilinoacrylate chromophore, lanceine, C20H26N203 or Cz4H30N204, mp 198", [o(]D + 64" (EtOH) and + 62" (CHC13). The latter compound, upon lithium aluminum hydride reduction, yielded an indoline that was readily acetylated. Preliminary pharmacological studies on V . herbacea have not yielded any very interesting results (85). REFERENCES 1. M . Pichon, Mem. Museum Natl. Hist. (Paris)23, 439 (1951). 2. R. L.Noble, C . T . Beer, and J.H. Cutts, Ann. N . Y . Acud.Sci. 76,882 (1958);Biochem. P h a m c o l . 1, 347 (1958).

12. THE VinCU ALKALOIDS

283

3. J. H. Cutts, C. T. Beer, and R. L. Noble, Cancer Res. 20, 1023 (1960). 4. G. H. Svoboda, J . Am. Phrtn. Assoc. Sci. Ed. 47, 834 (1958). 5. G. H. Svoboda, Lloydia 24, 173 (1961). 6. G. H. Svoboda, I. S. Johnson, M. Gorman, and N. Neuss,J. Pharm.Sci. 51,707 (1962). 7. C. T. White, Queensland Agr. J . 23, 143 (1925); J. M. Watt and Breyer-Brandwijk, “The Medicinal and Poisonous Plants of SouEhern Africa,” Livingstone, Edinburgh, 1932; R. N. Chopra, R. L. Badhwar, and S. Ghosh, “Poisonous Plants of India,” Govt. India Press, Calcutta, 1949; A. Petelot, “Les Plantesmedicinales du Cambodge, du LEOS,and du Vietnam,” Vol. 2, p. 117. Saigon, 1953. 8. V. N. Kamat, J. de Sa, A. Vaz, F. Fernandes, and S. S. Bhatnagar, Indian J . Med. Res. 46,588 (1958);D. H. K. Lee and W. R . M. Drew, Med. J . Australia 1,699 (1929). 9. N. R. Farnsworth, Lloydia 24, 105 (1961). 10. M. Greshbff, Ber. 23, 3543 (1890). 11. R. C. Cowley and F. C. Bennett, Australian J . Pharm. 9, 61 (1928). 12. N. Neuss, M. Gorman, G. H. Svoboda, G. Maciak, and C. T. Beer, J . Am. Chem. SOC. 81, 4754 (1959); M. Gorman, N. Neuss, and G. H. Svoboda, J . Am. Chem. SOC.81, 4745 (1959). 13. N. Neuss, M. Gorman, H. E. Boaz, and N. J. Cone, J . Am. Chem. SOC. 84, 1509 (1962). 14. J. P. Kutney, J. Trotter, T. Tabata, A. Kerigan, and N. Cameran, Chem.Ind. (London) p. 648 (1963). 15. M. Gorman and N. Neuss, Ann. Chim. (Rome) 53, 43 (1963); Am. Chem. Soc. 144th Meeting, Los Augeles, 1963, p. 38M. 16. W. B. Mors, P. Zaltzman, J. J. Beerebom, S. C. Pakrashi, and C. Djerassi, Chem. I d . (London)p. 173 (1956). 17. R. R. Paris and H. Moyse-Mignon, Compt. Rend. Acad. Sci. 236, 1993 (1953). 18. A. Chatterjee and S. K. Talapatra, Sci. Cult. (Calcutta)20, 568 (1955). 19. P. P. Pillay, T. N. Santhakumari, J . Sci. Ind. Res. (India)20B, 458 (1961). 20. M. Shimizu and F. Uchimaru, Chem. Pharm. Bull. (Tokyo)6, 324 (1958). 21. N. K. Basu and B. Sarkar, Nature 181, 552 (1958); B. N. Nazir and K. L. Manda, J . Sci. Ind. Res. (India)18B, 175 (1959). 22. G. H. Svoboda, M. Gorman, N. Neuss, and A. J. Barnes, Jr., J . Pharm. Sci. 50, 409 (1961). 23. M. Gorman, N. Neuss, G. H. Svoboda, and A. J. Barnes, J r . , J . Am. Pharm. Assoc. Sci. Ed. 48, 256 (1959). 24. C. P. N. Nair and P. P . Pillay, TetrahwEron 6, 89 (1959). 25. B. K. Mom and J . TrojBnek, Chem. Ind. (London)p. 1425 (1962). 26. M.-M. Janot and J. Le Men, Compt. Rend. Acad. Sci. 243, 1789 (1956). 27. M.-M. Janot, J. Le Men, and C. Fan, Bull. SOC. Chim. France p. 891 (1959). 28. M. Gorman, N. Neuss, and K. Biemann, J . Am. Chem. SOC.84, 1058 (1962). 20. C. Djerassi, S. E. Flores, H. Budzikiewicz, J. M. Wilson, L. J. Durham, J. Le Men, M.-M. Janot, M. Plat, M. Gorman, and N. Neuss, PTOC. Natl. Acad. Sci. U.S. 48, 113 (1962). 30. G. H. Svoboda, M. Gorman, A. J. Barnes, Jr., and A. T. Oliver, J . Phrtn.Sci. 51, 518 (1962). 31. G. H. Svoboda, N. Neuss, and M. Gorman, J . Am. Pharm.Sci. Ed. 48, 659 (1959). 32. Y. Nakagawa, 3. M. Wilson, H. Budzikiewicz, and C . Djerassi, Chem. Ind. (London) p. 1986 (1962). 33. Z. Cekan, J. T r o j h e k , and E. S. Zabolotnaja, Tetrahedron Letters No. 18, 11 (1959). 34. E. Schlittler and A. Furlenmeier, HeZv. Chim. Acta 36, 2017 (1953). 35. J. Trojanek, 0. Strouf, J. Holubek, and Z. Cekan, Tetrahedron Letters p. 702 (1961).

2 84

W. I. TAYLOR

36. J. TrojBnek, K. KavkovO, 0. Strouf, and Z. Cekan, Collection Czech. Chem. Commun. 26, 867 (1961). 37. S . Scheindlin and N. Rubin, J . Am. Pharm. Assoc. Sci. Ed. 44, 330 (1955). 38. M. Pailer and L. Belohlav, Monatsh. Chem. 85, 1055 (1954). 39. J. Mokrs and I. KompiB, Naturwissenschaften 50, 93 (1963). 40. M. Plat, M. D. D. Manh, J . Le Men, M.-M. Janot, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi, Bull. SOC.Chim. France p. 1082 (1962). 41. 0. Strouf and J. TrojBnek, Chem. I n d . (London)p. 2037 (1962). 42. J. Mokrj., I. KompiS, 0. BauerovB, J. Tomko, and S.Bauer, Ezperientia 17,354 (1961). 43. J. Mokrj., I. Kompii, and P. Sefitoviit, Tetrahedron Letters p, 433 (1962). 44. J. Trojanek, 0. Strouf, K. Kavkova, and Z. Cekan, Chem. Ind. (London)p. 790 (1961). 45. J. MokrL, L. Dfibravkova, and P. SefEoviE, Ezperientia 18, 564 (1962). 46. J. Trojanek, 0. Strouf, K. KavkovB, and Z. Cekan, Collection Czech. Chem. Commun. 27, 2801 (1962). 47. J . Trojbnek, J. Hoffmannova, 0. Strouf, and Z. Cekan, Collection Czech. Chem. Cornmun. 24, 526 (1959). 48. J . Trojanek, 0. Strouf, K. KavkovB, and 2 . Cekan, Collection Czech. Chem. Commun. 25, 2045 (1960); Pharm. Acta Helv. 35, 96 (1960). 49. M. Plat, J . Le Men, M.-M. Janot, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi, Bull. Soc. Chim. France p. 2237 (1962). 50. P. M. Lyapunova and Y. G. Borisyuk, Famatseut. Zh. (Kiev) 16, No. 2, 42 (1961). 51. K. Biemann, “Mass Spectrometry. Organic Chemical Applications,” Chap. 8. McGraw-Hill, New York, 1961. 52. M.-M. Janot and J. Le Men, Compt. Rend. Acad. Sci. 238, 2550 (1954). 53. M.-M. Janot and J. Le Men, Compt. Rend. Acad. S c i . 240, 909 (1955). 54. M.-M. Janot, J. Le Men, K. Aghoramurthy, and Sir Robert Robinson, Ezperientia 11, 343 (1955). 55. J . A. Joule and G. F. Smith, J . Chem. SOC.p. 312 (1962). 56. M.-M. Janot, J. Le Men, and C. Fan, Ann. Pharm. Franc. 15, 513 (1957). 57. M. F. Bartlett, R. Sklar, and W. I. Taylor, J . Am. Chem. SOC.82, 3790 (1960). 58. J. Gosset, J. Le Men, and M.-M. Janot, Ann. Pharm. Franc. 20, 448 (1962). 59. M.-M. Janot, J. Le Men, J. Gosset, and J. LBvy, Bull. SOC.Chim. France p. 1079 (1962). 60. S. Silvers and A. Tulinsky, Tetrahedron Letters p. 339 (1962). 61. M. Gabbai, Thesis, Univ. Pharm., Paris, 1958, Ser. U, No. 291. 62. M.-M. Janot and J. Le Men, Compt. Rend. Acad. Sci.241, 767 (1955). 63. M.-M. Janot, J. Le Men, and Y. Hammouda, Compt. Rend. Acad. Sci. 243, 85 (1956). 64. C. Djerassi, H. Budzikiewicz, J. M. Wilson, J. Gosset, J. Le Men, and M.-M. Jrtnot. Tetrahedron Letters p. 235 (1962). 65. A. Orechoff, H. Gurewitch, S. Norkina, and N. Prein, Arch. Phclrm. 272, 70 (1934). 66. M.-M. Janot and J . Le Men, Ann. Pharm. Franc. 13, 325 (1955). 67. N. R. Farnsworth, F. J. Draus, R. W. Sager, and J. A. Bianculli, J . Am. Pharm. Assoc. Sci. Ed. 49, 589 (1960). 68. S. Y. Yunusov and P. K. Yuldashev, Dokl. Akad. Nauk Uz. SSR No. 9, 23 (1956); J . Gen. Chem. USSR (Eng. Trawl.) 27, 2072 (1957). 69. S . Y. Yunusov, P. Yuldashev, and N. V. Plekhanova, Dokl. Akad. Nauk Uz. SSR No. 7, 13 (1956). 70. S. Y. Yunusov and P. K. Yuldashev, Zh. Obshch. Khim. 27, 2015 (1957). 71. P. K. Yuldashev, V. M. Malikov, and S . Y. Yunusov, Dok. Akad. Nauk Uz. SSR No. 1, 25 (1960).

12.

THE ViTbCCA ALKALOIDS

285

72. P. Panov, I. Ognyanov, N. Mollov, K. Rusinov, V. Georgiev, and D. M. Zhelyazkov, Compt. Rend. Acad. BulgareSci. 13, No. 1,39 (1961);N. Mollov, I. Mokri, I. Ognyanov, and P. Dalev, Compt. Rend. Acad. Bulgare S c i . 13, No. 1, 43 (1961). 73. M.-M. Janot, J. Le Men, and Y. Hammouda, Ann. Pharm. Franc. 14,341 (1956). 74. M.-M. Janot and J. Le Men, Compt. Rend. Acad. Sci.239, 1311 (1954). 75. M.-M. Janot, J . Le Men, and Y . Gabbai, Ann. Pharm. Franc. 15, 474 (1957). 76. A. Quevauviller, J . LeMen, andM.-M. Janot, Compt. Rend.Soc. Biol. 148,1791 (1954); Ann. Phmrm. Franc. 12, 799 (1954); G. Gazet du Chatelier and E. Strasky, Ann. Pharm. Franc. 14, 677 (1956); J. Hano and J. Maj, Acta Polon. Pharm. 15, 1 7 1 (1957). 77. Raymond-Hamet, Compt. Rend. Soc. Bid. 148,1082 (1954);L. Szporny and K. Szasz, Arch. Ezptl. Pathol. Pharmakol. 236, 296 (1959); 2 . Szab6 and Z . Kagy, ArzneimittelForsch. 10, 811 (1960). 78. E. Szczeklik, J. Hano, 13. Bogdanikowa, and J. Maj, Polski Tygod. Lekar. 12, 121 (1957); Z. Szab6 and Z. Nagy, Arzneimittel-Forsch. 10, 811 (1960). 79. D. K. Zheliazkov, Suwremenna M e d . 9, 16 (1958). 80. M. B. Sultanov, Izv. A k a d . N a u k U z . SSRSer. M e d . No. 3, 38 (1959). 81. A. Kaldor and Z. Szab6, Ezperientia 16, 547 (1960). 82. F. H. L. Van Os, Pharm. Weekblad 96, 966 (1961). 83. A. Quevauviller, J. Le Men, and M.-M. Janot, Ann. Pharm. Franc. 13, 328 (1955). 84. N. R. Farnsworth, H. H. S. Fong, R. N. Blomster, and F. J. Draus, J . Pharm. S c i . 51, 217 (1962). 85. M. B. Sultanov, I z v . A k a d . N a u k . Uz. SSRSer. Med. No. 1, 29 (1959), [Chem. Abstr. 53, 20, 543 (1959)l; D. A. Bucharova, Aptechn. Delo 8, No. 2, 23 (1959), [Chem. Abstr. 53, 17,426 (1959)l; K. S . Roussinoff, D. K. Jelyazkoff, and V. P. Gueorguieff, Arch. Ital. Sci. FarmacoZ. 11, 83 (1961). 86. J. Mokrj., I. KompiB, L. Dubravkova, and P. SefEoviE, Ezperientia 19, 311 (1963). 87. J. Mokry, I. KompiB, L. Dhbravkovti, and P. SefEoviE, Tetrahedron Letters p. 1185 (1962).

This Page Intentionally Left Blank

-CHAPTER

13-

Rauwolfia ALKALOIDS WITH SPECIAL REFERENCE TO THE CHEMISTRY OF RESERPINE1 E . SCHLITTLER Research Department, CIBA Phumceutical Company, Diwision of CIRA Corporation, Summit, New Jersey

.

.

I. Rauwolfia Species and Their Alkaloids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Explanation of the Tables. . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . B. Crude Rauwolfca Extracts.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Extractive Methods.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Classification of RauwoZJiaAlkaloids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. ReserpineCongeners ..............................................

.

.

..

287 288 294 294 296 296

11. The Chemistry of the Reserpine Group. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Conformational Aspects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Modification of the Reserpine Skeleton. . . . . . . . . . . . . . , . . . . . . . . . . . . . . . C. Semisynthetic Reserpines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.. .

300 300 302 311

111. Synthetic Work ..................................................... A. TheTotal Synthesis ............................................... B. Syntheses of Simple Analogs. . . . . . . . . . . . . . . . . . . .. . . .. . . . . . . . . . . . .

316 316 325

..

.

. .

References....

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

327

I. Rauwoljia Species and Their Alkaloids Within tthe last 10 years, Rauwolfia. products have become important therapeutic agents, both as sedatives and antihypertensives. Although their production and use have fallen off since the peak years of 1955 and 1956, it is estimated that their total sales a t the consumers’ level in 1961 still amounted to $100 million in the United States alone. Since 1952, the year reserpine was first isolated, several thousand articles have been published on the isolation, chemistry, pharmacology, and clinical aspects of reserpine and other Rauwolfia alkaloids, and today these investigations are still being pursued. Botanists estimate the number of identified Rauwoljia species to be about 50, of which R. serpentina., For the most recent review of chemistry and pharmacology of the Rauwolfia alkaloids see R. A. Lucas in “Progress in Medicinal Chemistry” (G. P. Ellis and G. B. West, eds.), p. 146. Butterworths, Washington and London, 1963. 287

288

E. SCHLITTLER

R. canescens, R. vomitoria, and R. ligustrina have been investigated in detail. Only the first three species are important from the standpoint of supplies of therapeutically useful alkaloids ; the last species is not especially rich in reserpine, but has nevertheless been investigated in great detail (7). The Indian plant, R. serpentina, has lost much of its importance (except for local production), since its reserpine content (0.1%) is only about half that of R. vomitoria, which is a t present the most important species. It grows so plentifully in Central Africa, especially in the Congo, that cultivation is not necessary. It is considerably taller than R. serpentina (about 10 feet as compared with a shrub of 3 feet), and procedures have been developed by Congolese collectors by which the smaller side roots may be cut periodically (about every 3 years) without loss of the tree ( 1 10).

A. EXPLANATION OF THE TABLES Tables 1-111 list the Rauwolfia species investigated, the alkaloids so far isolated, and Rauwolfia alkaloid synonyms. I n Table 11, only the reference to the first isolation of a specific alkaloid is given, and only in case of simultaneous isolation is more than one reference quoted. Boxedin entries indicate that in this particular case the presence of the alkaloid was demonstrated by paper chromatography only. I n this table, the isolation of reserpine has not been specially reported because it was isolated from all RuuwolJia species, with the exception of b, de, f, ma, sf, and ve (see Table I for abbreviations). In addition t o these sources, reserpine has also been isolated from or recognized in other closely related apocynaceous plants, viz., Tonduzia longijolia Markgraf (111) (also contains rescinnaniine), Alstonia constricta F. Muell. (112), Vinca (Lochnercr) rosea L. (113), Vallesia dichotoma Ruiz et Pavon (114), Excavatia coccinea Markgraf (115), Vinca minor L. ( 1 l 6 ) , and Ochrosia poweri Baily ( 1 17). Table 111, which gives synonyms, is necessary because in the middle 1950’s many new alkaloids were being isolated, characterized, and named almost at the same time by widely dispersed groups. As a by-product of these investigations, the belief that many alkaloids were confined only to certain genera was invalidated. Thus, yohimbines and ring E heteroyohimbinoid alkaloids, which were thought to be confined to yohimbe and Pseudocincho?za,turned up in Rauwolfia species, and the sarpagine type first detected in the last genus occurs also in other members of apocynaceous and loganiaceous genera. During the

TABLE I Rauwolfia SPECIES' Code

Name

Code

Name ___ R. mauiensis Sherff* R.micrantha Hook. f . R.momhasiana Stapf 12. nana E. A. Bruce R. natulcnsis Sond. ( = caffru ?) R.nitida Jacq. R.obscura K. Sch. €2. paruensis Ducke R.pentuphylla Ducke R.perakensis King et Gamble R.rosea K. Sch. R.salicifolia Griseb. R. sandwicensis A.DC. R.sarapiquensis Woods. R. schueli Speg. R.sellowii Muell.-Arg. R. semperforens (Muell.-Arg.) Schltr. R. serpentina (L.) Benth. ex Kurz R. sprucei Muell. -Arg. R.sumutrana (Miq.) Jack R.ternifolia HBK. ( = ligustrina) R.tetruphylla L. R.verticillata (Lour.) B a a . R. viridis (Muell.-Arg.)Guillaumin R. vomitoria Afz. .-

af a bh b bo cf ca C

cb cu dc de di d f g

h ht i ie J

I 43 It mP mn

R. affhnis Muell.-Arg. ( = grandijiora) R. umsonicr.efolia (Miq.) A.DC. R. Bahiensis A.DC. R.beddomci Hook. f. R. boliuiana Mgf. ( = schueli) R. caffra Sond. R. cnmhodicma Pierre ex Pitard R.canescens L. ( = tetraphylla) R. czrhuna A.DC R. cumminsii Stapf R.decurun Hook. f. ( = densijiora ?) R. degeneri Sherff (= sadwicensis ?) R discolor* R densiftora Benth. ex Hook. f . R. fruticosa Burck. R. grandijloru Mart. ex A.DC. R. heterophylla Roem. et Schult. ( = tetruphylla) R. hirsuta Jacq. ( = tetruphylla) R. indecora Woods. ( = ligustrina) R. inebriaw K. Sch. ( = c a & z ?) R.,javanica Koord et, Val. R. kzmarckii A.DC. ( =uiridia Roem. e t Schult.) R.ligustrina Roem. et Schult. R. littoralis Rusby R.macrophylla Stapf R. mannii Stapf

'After R. E. Woodson (1lo), except for the asterisked plants.

ma m mo na n ni 0

Pa Pe P r

81 sd sa BC BW

Bf S

5P SU

tr t Ve vi V

TABLE I1

p.3

co 0

ALEALOIDSOF RauwoZjia SPECIES

Melting point (" C)

253-254

Rotation"

- 62"

Source* bl, ~

2 ~ f3 4 , , h5,

m

241-242 158-160 264-266 180-1 83 190-1 92

+ 141"

j6, lg7, ma, ni9, d o ,

p

swlz, ma78 bole, ~ 2 ~ 1 d19, 7,

de20,

s10, 8027, su4, sw12,

+ 72"

s29, v13

- 97"

S l O , sw30

-(?)132"

C2

~ ~ 1~1~, 1v13, 2 ,vd4,

5

hzl, izz, lg7, nZ3,oZ4,p25, naZ6, tzs, 4 3 , m 6

m,

M rn

8

p25,40

52

p25, 40

p25,40

> 300 190 220-223 228-229 240 230-231 231-232 228-232 240-242 32 1 228 130-131 183 155

- 59" (E)

+ 57" (E)

ht31, 0 3 2 , ~ 3 2 c34, hzl, lg7, sc27, sw12, m 3 3 ,

r;;;;l"

a35

c36 h37 S61

- 85" (P) - 137" + 184' (M) + 175" (P)

c2,s38 m3, c36, ma39 v67 024

- 121" (E)

+ 112" + 333O

P25 p40, v41 v79

I

lg7, af, cb, ht, j, It, sl, sp, t, t r p

8

Raugustine Raujemidine Raujemidine N-oxide Raumitorine Raunamine" Raunescine Isoraunescine Rauniticine Raunitidine Rauvanine Rauvomitine Rauwolfine" Rauwolfinine' Renoxidine (reserpine N-oxide) Rescidine Rescinnamine

160-170 144-150 2 15-2 17 138 206-207 160-170 241-242 233-235 276-278 129-135 115-117 235-236 235-236 238-241 183-186 237-238

-50" -88"

+ 60" $60" -74' -70" -35" -70" +32" -173"

- 35" (E) - 100" -63" -97"

1g7 c42 c79 v43

F

w

11145

c44, a

7

e44,(1g(7 ni9 ni9 v46 v47 cf18 548.. PI9 c49, s49, _

Jp -

v49,

v50

i, ie, 1, It, mo, ni, sp,su,trll5,1Ig7,

- 12"

Reserpiline

I

U

G153

0,pa,

pe, sl,

1

U

bole, c34, di54, lg7, m55, s56, 8027, v57, dc53, af, bh

cb, ht, ie, 1, It, mo, ni, pa, pe, r, sd, sl, sp,415 bo16, G4, ca58, dc53, lg7, ni9, pz5, 81237, v59, [i ,x - Z , V20,lgeo ~~

Isoreserpiline Isoreserpiline-#-indoxy1 Reserpine Isoreserpine $-Reserpine Reserpinine Isoreserpinine Sandwicensiue' Sandwicine Sarpagine Neosarpagine (sarpagine?)

211-212 251-254 264-265 152-156 257-258 243-244 225-226 260-262 r350 390

-82" (P' - 254" - 117" -164" -65" - 131" -18" 56" (M) 180" +53"(P)

+ +

sd39 m~39,sd39 bl, cz, d62, d v59, m45

,

l 1 5 I

v

H

~ h5,~ht31, ~ izz,, lg7,

11163.

p25, ~64,~

~ 6 5 ,

3 p.3

0 L

TABLE 11-continued

Name

Formula

Melting point (" C)

Rotation"

Sourceb

__ Semperflorine (tetraphyllicine?) Seredamine" Seredine Serpentine Serpentinine Tetrahydroalstonine Tetraphyllicine Tetraphylline Vomalidine Vomilenine Yohimbine a-Yohimbine /3-Yohimbine 3-epi-a-Yohimbine

#-Yohimbine a

295 297 291 158 265-266 228-230 320-322 220-223 242-243 207 242-243 238-239 246-249 125-1 28 181-183 222-223 268

+ 60" + 292" (M) +117" - 102" 61" - 78' +318" - 72" (P) 101" (P) - 12" - 54" (P) - 93"

+

+

-I-27" (P)

sf 6 6 v67 v43 b4, c68, h69, 197, m8, $0, sw30, E111, a6, dezo, j6, lg7, ma20, $0, sd39, t70, v13 ig7, sW12 de20, ma20, 024, sd39, sw12, t70 de20, sd39, t70, di54, H I 5 v71 v72 c68, h5, s75, v57, m7,m 4 c73, hzl, lg7, su4, v57, ht31 c74 s76

c77, t z 8

Rotations listed are for chloroform solution except as noted: E = ethanol, M = methanol, and P = pyridine. Letters refer to species (see Table I for code); superscript numbers indicate reference (see List) to first isolation of the alkaloid. Alkaloids of unknown structure.

13. RauwolJia ALKALOIDS

AND CHEMISTRY OF RESERPINE

293

TABLE I11

RuuwolJiu ALKALOID SYNONYMS

Synonym (Isolation)

Preferred name (identification) -

Neoajmalie (29) Alkaloid A (83) Alkaloid B (85) Alkaloid C (87) Alkaloid C (85) Alkaloid F (84) Alkaloid F-2 (71, 88) Alkaloid Su-3078 (76) Alkaloid 13,141 (90) Canescine (2) Chalcupine A (37) 11-Demethoxyreserpine (93) 3-Epirauwolscino (92) Heterophylline (21) Micranthine (8) Serpentidine (8) Perakenine (19) Raubasine (51) Raubasinine ( 51) Raugalline (97) Rauhimbine (38) Isorauhimbine (38) Raupine (64) ‘‘ Rauwolfine ” (85) Rauwolscine (73) Recanescine (104) “ Reserpinine ” (51) Reserpoxidine (49) Serpine (105) Serpinine (107) Substance I(108) Substance I1 (108) py-Tetrahydroserpentine (109) 6-Yohimbine (133)

Ajmaline (82) Reserpinine (84, 62) Serpentine (86) Reserpinine (87) Ajmalinine (86) Ajmalicine (84) Vomilenine (72) 3-Epi-a-yohimbine(89) Reserpinine (62) Deserpidine (91) or-Yohimbine (92) Deserpidine (91) 3-Epi-a-yohimbine(92) Aricine (21) Reserpiline (94) Serpentinine (94) Rauwolfinine (96) Ajmalicine (68) Reserpinine ( 5 1) Ajmaline (98) Corynanthine (87, 99) 3-Epi-or-yohimbine (100) Sarpagine (101) Ajmaline (102) or-Yohimbine (103) Deserpidine (91) Rescinnamine (51) Renoxidine (49) or-Yohinibine yohimbine (106) Tetraphyllicine (96) Reserpinine (68) Ajmalicine (51) Ajmalicine (109) Ajmalicine (109)

+

past few years, it has been clearly demonstrated that, although there are a large number of different structural types among indole alkaloids, they are all built out of the same building blocks-a tryptamine residue plus a Cs, a C1, and a Cs unit (118).

294

.

E SCHLITTLER

B. CRUDE RauwolJia EXTRACTS Although natural and synthetic RauwolJia alkaloids are available in high purity today, ground root in tablet form and partially separated Rauwolfia fractions are still used therapeutically. In this connection, the ‘‘ alseroxylon fraction ” should be mentioned, since it is defined as a “selected RauwolJia fraction from which the sympathicolytic and hypertensive alkaloids have been removed and which retains only the total antihypertensive, bradycardic, and sedative activity ” (119). Attention should also be drawn to recent publications on the isolation of a new hypotensive factor from the roots of R. serpentina (120). I n these publications, the separation of the crude alkaloids from fresh undried R. serpentina roots into four alkaloidal “complexes ” is reported. The first three fractions have been named resajmaline, ajmalexine, and serpajmaline, respectively. The serpajmaline fraction is said to be therapeutically active and also is claimed to contain serpentinine, serpentine, ajmaline, and two unknown substances, one of which is probably reserpiline. This fraction is said also to be free from reserpine and to be much more potent in its hypotensive activity than reserpine, but to lack reserpine’s sedative and central nervous depressant action. However, a later pharmacological investigation of an authentic water-soluble serpajmaline fraction (alkaloids present as salts) demonstrated that the type of antihypertensive activity observed closely resembles that of serpentine and serpentinine (121).This conclusion receives support from a chemical study of this fraction in which serpentinine, ajmaline, and tetraphyllicine were obtained in a pure state (121).

C. EXTRACTIVE METHODS For laboratory purposes, methanol is the usual solvent, and procedures based on its use have been described in detail (21, 39, 81). Of great practical usefulness for the separation of the weakly basic fraction is the solubility of certain alkaloidal acetates in chloroform, e.g., reserpine, ajmalicine, and aricine, whereas other acetates are insoluble in this solvent, e.g., ajmaline, yohimbine, and a-yohimbine. Since the anhydronium alkaloids are extremely strong bases, they can only be extracted into an organic solvent in their tertiary base form at p H 11.For industrial purposes, the best process extracts water-moistened RauwolJia root with hydrocarbons such as benzene, t,oluene, or xylene. I n this procedure, only the weak bases are extracted. No complicated separation processes itre involved, and reserpine is obtained in high yields (122).

13. Rauwolfia

ALKALOIDS AND CHEMISTRY OF RESERPINE

295

Extensive use of column chromatography has been necessary to separate the Rauwolja alkaloids, and in this connection attention is drawn t o a publication which concerned itself with the more refined technique of gradient elution chromatography ( 123). Paper chromatography has been used extensively for analytical, fractionation, and identification purposes (124, 15). This tool is not particularly useful for the assay of crude extracts, since certain alkaloids show up clearly whereas others cannot be resolved. More information as to specific alkaloidal composition is better obtained from more highly purified fractions.

$3qC00H'A

OHC

(COOH)

Lw4 I

I

Ajmaline type

Sarpagine type

I

HOOCA Yohimboid precursor

J

I

I

0 Yohimbine type (Yohimbinesand 18-hydroxyyohimbines)

Ring E heterocyclic type

CHARTI. Probable interrelationshipsamong Rauwolfia alkaloid types.

296

.

E SCHLITTLER

D. CLASSIFICATION OF Rauwolfia ALKALOIDS The arbitrary classification of Rauwolfia alkaloids (91) is simplified here, and it is slightly different from a recent arrangement (92). The Rauwo@a alkaloids can all be regarded as yohimbinoid derivatives as shown in Chart I, viz.: the yohimbines (all yohimbine isomers); 18hydroxyyohimbines (reserpine-type alkaloids) ; ring E heterocycles and their anhydronium analogs (ajmalicine, serpentine) ; ajmaline-type (which includes sarpagine) ; and compounds of unknown constitution.

E. RESERPINECONGENERS The occurrence of reserpine has been reported from all RauwolJia species, with the exception of about half a dozen in which it is probably present in minute amounts. Renoxidine, the N-oxide of reserpine, has been isolated from R. vomitoria, R. serpentina, and R. canescens, and it may riot be a natural product, since it could have been derived by autoxidation of the tertiary base which is abundant in these plants. If it was an artifact, the occurrence of other analogous N-oxides should have been noted, but so far the only other recognized case is raujemidine N-oxide, which is found along with the parent alkaloid, raujemidine (a minor base of R. canescens). I n contrast to reserpine, deserpidine and rescinnamine are of restricted distribution, each being recognized so far in about ten species only. Unlike the foregoing alkaloids, the remaining reserpine congeners listed in Table IV either have a greatly reduced tranquilizing and antihypertensive effect or are completely inactive. For activity, the methoxyl (R1=MeO) is not essential, but an ester function (R3= acyl) is indispensable, and the potency is increased if Rz = methyl. The isomeric esters (Rz = acyl) or the synthetic diesters (Rz = R3 = acyl) are inactive. Methyl reserpate (125), reserpic acid (125), and reserpic acid 18-0trimethoxybenzoate (126) have not been found in nature, and they are also pharmacologically uninteresting (126, 127). With the exception of raujemidine, the structures of the compounds in Table IV have been obtained by a combination of partial degradation, parallelism of properties with reserpine or deserpidine, and partial synthesis. Thus, the structure of methyl raunescate rests entirely upon those reactions which were authenticated in the case of methyl reserpate and deserpidate. Methyl pseudoreserpate was partially synthesized from reserpine (132), which established its structure except for the attachment of the acyl residue. Apart from the biological activity evinced by the

13. Rauwolfia

ALKALOIDS AND CHEMISTRY OF RESERPINE

297

lg-esters, there has been until recently no simple way of determining the site of the acyl group in the monosubstituted 17,18-dihydroxy compounds. It has been shown that, in the presence of aluminum isopropoxide, diazomethane is a practical methylating agent for yohimbinoid alkaloid alcoholic functions ( 128); thus, dihydrorescidine vtith this reagent gives dihydrorescinnamine (129). TABLE IV NATURALLY OCCURRINGRESERPINE-TWE ALKALOIDS

Me TMB‘ OMe - Reserpine N -oxide Me TMC~ OMe H TMB OMe H TMC OMe TMB H OMe H Me TMB H H TMB TMB H H

Reserpine Renoxidine Rescinnamine #-Reserpine Rescidine Raugustine Deserpidine Raunescine Isoraunescine Raujemidine Raujemidine N-oxide a

TMB = 3,4,5-trimethoxybenzoyl. TMC = 3,4,5-trimethoxycinnamoyl.

Very recently some light has been shed on the chemistry of raujemidine (42) which a t one time seemed t o be the only naturally occurring and pharmacologically active stereoisomer of reserpine (130). Raujemidine has now been shown to be an unsaturated derivative of reserpine, viz., the compound (133a). Upon comparing reserpine with raujemidine, a certain similarity in the space distribution pattern (viz., bulkiness to the right of the basic nitrogen and above the plane of the ~

l

~

~

3

~

~

c-3

/I]

Reserpic acid lactone

OR

[.-.

1

-

C-16

$MrOOC Me

yN PCoO-OH Y e = CHzOH

COOMe = COOH

a

L

COO,\IC Ale

C

% c2

Ib

tram -tru ns .cis

cis-trans-cis

cz

I\’

(Methyl-16-epireserpate)

m

h

-

3

13. Rauwoljia

/

3

ALKALOIDS AKD CHEMISTRY OF RESERPINE

/

299

H U

300

E. SCHLITTLER

DE rings) can be observed which might be responsible for the fact that this alkaloid exhibits about one-half of the tranquilizing property of reserpine as measured by the mouse ptosis assay (see Section VIII). The raujemidine content of R. vomitoria Afz. is very low (42), and for all practical purposes, raujemidine is of no importance.

11. The Chemistry of the Reserpine Group

A. CONFORMATIONALASPECTS Much of the chemistry of reserpine and its congeners becomes clear and compelling if the conformational mobility of the CDE rings of its 3-epialloyohimbane nucleus is remembered (134). It was the gradual realization of this property and how to take advantage of it that led to the very elegant experiments which established the complete relative stereochemistry of the reserpine molecule (135, 136). Not unexpectedly, it is found that reserpine under normal conditions prefers the cis-transcis CDE conformation (137, 138), but this does not prevent it from assuming the trans-trans-cis shape to make possible the formation of reserpic acid lactone (125) and the quaternary salt 111(136). Under appropriate reaction conditions, the reserpine molecule may epimerize to a thermodynamically more stable derivative, the driving force being the reorientation of the bulky axial substituent a t C-3 in Ia to an equatorial position. The most important of these isomerizations occurs at C-3 by one or more processes, e.g., in acidic media according to schemes A, B, or C (139), or under certain hydrogenation conditions (Section 11,B, 3). I n spite of the ease of this reaction, neither isoreserpine nor its congeners has yet been recognized as a natural product, although its 18-deoxy equivalent, a-yohimbine, is known (140). The stereochemical features of the isomolecule allow N-4 to ring-close a t C-18 to yield.VIII via the intermediacy of a skew-boat D ring (IIb) (135). Analogous quaternization is possible in the case of both reserpinol and isoreserpinol, viz., creation of an N-4-CH2-C-16 bond (136, 141). This stereochemical knowledge was used t o full effect in a stereospecific total synthesis of reserpine (139), since under suitable acidic conditions isoreserpic acid lactone can find considerable release of nonbonded interactions by inverting a t C-3 t o reserpic acid lactone (139, 142). Besides epimerization (C-3) of the axially oriented indole residue to give a more stable molecule via Ia, isomerization in ring E could also lead to a more thermodynamically stable compound if any one or more of the C-16, C-17, or C-18 substituents become equatorial in the trans-

13. Rauwolfia

ALKALOIDS AND CHEMISTRY OF RESERPINE

301

trans-cis conformer (Ib). Thus, under the conditions of the preparation of methyl reserpate by alkaline methanolysis of the trimethoxybenzoyl moiety, not unexpectedly methyl 16-epireserpate (IV) is also formed (79). This compound, subjected to more prolonged methanolysis conditions, sets up an equilibrium with some methyl reserpate (i.e., the energy difference is not too great between these two epimers), but finally both compounds are replaced by additional products which result from the intermediacy of an unsaturated acid (V) formed by a retro-Michael reaction. The sole characterized end product under certain conditions is methyl neoreserpate (VI), in which both the C-16 and C-17 substituents have become inverted ( 143). Since the intermediate a,P-unsaturated ester MeOOC' ' U O H OMe Methyl reserpate ( I a

I

MeOOG"\/\OH

+ t--

Me00W OMe

IV Methyl 16-epireserpate

I

I

OMe VI Methyl neoreserpate

V

VII

302

E. SCHLITTLER

could also isomerize into its p,y-equivalent (VII),other compounds could also be formed which would explain why the yield of methyl neoreserpate is never high. I n fact, under more vigorous methanolysis conditions, the keto ester corresponding to VII is the major product of the reaction (144). With the advent of NMR-spectra, the conformational analysis of some of these compounds has become somewhat easier, and in the reserpine class of compound, it can often distinguish between the cis-trans-cis form (equatorial C-3 proton) and the trans-trans-cis (axial C-3 proton). IR-spectroscopy has also proved useful in this regard, since the CD rings can be regarded as a quinolizidine, in which system the observation has been made that, when there are a t least two protons on the carbons a to the nitrogen trans-axially oriented with respect to its lone pair orbital, there will be a doublet between ca. 2700-2800 cm-1 (145). I n the case of the reserpine-type molecule, this would be true for the trans-trans-cis conformer (146))but in the cis-trans-cis case there is only one axial proton (at C-21) since the C-3 proton is equatorial and a t C-5 it is pseudoaxial [because the ring is the equivalent of a cyclohexene (143)j. Since isoreserpine also shows the expected multiplet, it is not possible from the IR-spectrum alone [as was once suggested (146)] to decide whether the C-3 proton in such compounds is u (below the plane of the rings, iso) or (above the plane of rings). As a final point, it should be remembered that the facile conversion of reserpine into isoreserpine is considerably favored by the presence of three equatorial substituents in ring E in the isomer Ia, but when the trimethoxybenzoyl moiety is hydrolyzed and one of the remaining groups inverted, it is no longer a priori possible to decide which one of the two conformers will predominate with certainty or even how easy the conversion to the corresponding C-3 is0 series will be (cf., in this regard, the inverted analog reserpine, vide infra). The absolute chemistry of reserpine has been derived directly (135),by making use of Klyne’s extension of the Hudson lactone rule (147) as applied to reserpic acid and its lactone, as well as by the application of Prelog’s asymmetric synthesis (148) to methyl reserpate (149). These results are in agreement with the conclusions obtained by more indirect but no less accurate means (91).

B. MODIFICATIONOF

THE

RESERPINESKELETON

1. Genera1 As has already become evident from the foregoing discussion, the reactions of reserpine that set it apart from the other yohimbinoid

13. RauwolJia ALKALOIDS

AND CHEMISTRY OF RESERPINE

303

alkaloids are largely owing to the extra substituent at C-18 in Ring E (Chart 11).Reactions which find exact parallels in other alkaloids are those that take place in rings A, B, and C; for example, the formation of A3 and tetradehydro compounds by reagents such as tertiary butyl hypochlorite (150), potassium dichromate (151), lead tetraacetate (91), and palladium-catalyzed hydrogen transfer to maleic acid (91). The indole system can be alkylated, and when the addendum is small, the N,-substituted product is obtained (152), but when it becomes large, e.g., benzyl, the C7-alkylindolenine is the important product (153). Ring A of reserpine and deserpidine has been substituted by bromine; substitution occurs in both cases in the 10 position (154).Acetylation of reserpine with acetic anhydride and perchloric acid (155) gave a mixture of the 10- and 12-acetyl reserpines in about equal yield. Interpretation of the IR-spectra excluded the possibility of N-acetylation and/or isomerization at C-3. Under the same conditions, isoreserpine was also acetylated in the identical positions of ring A. Introduction of additional hydroxyl groups into rings A and E of yohimbines by the use of microorganisms has been studied extensively. It is claimed that, by using a type of Streptomyces, a hydroxyl group is introduced into position 18 of both yohimbine and a-yohimbine (156). Hydroxylation of apoyohimbine, ,6-yohimbine methyl ether, and 3epiapoyohimbine with Cunninghamella blakesleana Lendner took place at position 10, and only for apoyohimbine (but not 3-epiapoyohimbine) is additional hydroxylation at 18 likely (157, 158).

2. Stereoisomersof Methyl Reserpate and StereoisomericReserpines A number of stereoisomers of methyl reserpate have been prepared by chemical manipulations. They have all retained the cis DE ring configuration, and so far no diastereoisomer of methyl reserpate with the equivalent trans-fused system has been prepared synthetically. Contrary to earlier assumptions that activity was present only in reserpic acid diesters, it was found that methyl 18-epireserpate itself possessed distinct sedative, but not antihypertensive, activity (159). Methyl reserpate and methyl isoreserpate were completely inactive, and the three other methyl reserpates have not been tested. The stereoisomeric methyl reserpates (see Table V) were prepared in the following ways. (a) Methyl isoreserpate can be obtained from methyl reserpate by refluxing in collidine containing p-toluene sulfonic acid (161) or by sodium borohydride reduction of methyl 3-dehydroreserpate. (b) Methyl 16-epireserpate is one of the products of the prolonged alkaline methanolysis of reserpine, and is isolated through its highly

304

E. SCHLITTLER

crystalline acetate hydrochloride. When methyl 16-epireserpate is further treated with alkali, it is partly reconverted into methyl reserpate and partly transformed into methyl neoreserpate (79, cf. Section 11,A). TABLE V ISOMERIC METHYLRESERPATES 18-0-3,4,5Trimethoxybenzoate Isomer (preparation)

Melting point (" C)

Methyl reserpate Methyl isoreserpate (161) Methyl 16-epireserpate(79) Methyl 16-epi-17-epireserpate(143) Methyl 18-epireserpate(151, 160) Methyl 18-epi-isoreserpate(160)

242-244 220-221 258-25ga 146-150" 222-223 2 10-21 3

[UID

(CHC13) -106' - 60" +49"6 53"

+

-8 1 O -

Melting point ("C)

(CHC13)

262-263 152-165 180 163-170 141-145 245-248

-117' - 164" 44" +29O +38" -14'

[UID

+

Acetate hydrochloride. In MeOH. ' Isopropanol of crystallization.

a

(c) Methyl 16-epi-17-epireserpate (methyl neoreserpate) is one of the readily isolatable epimerization products of the prolonged alkaline methanolysis of reserpine (143), and its formation has already been discussed under Section 11, A. Proof has been provided that the 18hydroxy group has not been epimerized, and as a consequence, neoreserpic acid does not form a lactone. Unlike methyl reserpate itself, methyl neoreserpate does not epimerize at C-3 when subjected to refluxing collidine in the presence of p-toluene sulfonic acid (see preceding discussion). This is because its conformation, which corresponds to that of the trans-trans-cisbackbone of Ib, is preferred over that of the C-3 isomeric series. (d) Methyl 18-epireserpate is prepared by heating the 18-p-bromobenzene sulfonate or m-nitrobenzene sulfonate of methyl reserpate with aqueous trimethylamine and dioxane (151, cf. Section II,C, 2). Another way of preparation is the reduction of methyl 18-ketoreserpate (see Section 11, B, 3) with sodium borohydride, with the 18-normal and the 18-epi compounds being obtained in roughly equal amounts (162). (e) Methyl 18-epi-isoreserpate was obtained by reduction of methyl 18-epi-3-dehydroreserpate with sodium borohydride ( 160).

13. RauwolJia

ALKALOIDS AND CHEMISTRY OF RESERPINE

305

Prom these intermediates, the stereoisomeric reserpines were prepared by trimethoxybenzoylation under the usual conditions. All these stereoisomers were pharmacologically inactive, even isoreserpine which differs from reserpine in the configuration of only one hydrogen atom. However, if the last two compounds are compared (Iversus 11,Chart 11), their shapes are found to be significantly different. There are, however, other apparently minor changes in the alkaloid, where the over-all shape is not involved, which also result in inactive compounds, e.g., N,-methylreserpine, N,-allylreserpine, and reserpamide (C-16 = CONH2) are inactive (152). I n fact, the first derivative acts as a reserpine antagonist. The fact that all reserpine-like activity is lost upon quaternization of Nb is less surprising in light of the resulting change in polarity.

3. 18-Ketones and Their Properties At an early date it was already recognized that the ketone (IX) derived fromanoxidationofthe C-18carbinolfunction ofmethylreserpate could be of considerable utility for further transformation of the reserpine pentacyclic ring system, but early attempts a t the preparation of the desired compound by conventional oxidation, e.g., by Oppenauer’s method, N-chlorosuccinimide, sodium dichromate, or chromic oxide in pyridine, were unsuccessful with both methyl reserpate and methyl 18-epireserpate. The ketone was finally obtained by heating methyl reserpate p-bromobenzene sulfonate with dimethyl sulfoxide in the presence of triethylamine (162), a method successfully used for simpler compounds (163). Subsequently, it was found that this oxidation could also be realized with other benzene sulfonate esters of methyl reserpate and 18-epireserpate. That the stereochemistry of the inolecule was unaffected was proved by sodium borohydride reduction of the ketone, which gave equal amounts of methyl reserpate and its 18-epimer. This and other simple reactions of the ketone are sketched in Chart 111, and additional observations will be given. Reductive amination of methyl ketoreserpate was studied in detail. The use of n-propylamine in the presence of a palladium-charcoal catalyst led not only to the expected mixture of a- and 13-aminoderivatives (X), but also to a methyl 17-demethoxy-18-deoxy-18-n-propylaminoreserpate (XII). I n the last case, no attempt was made to define the stereochemistry a t C-16 and C-18. Similar eliminations of the C-17 methoxyl under basic conditions were also encountered with other ketone derivativet. Reductive aminations were also effected with secondary amines. Piperidine yielded as the sole insolable product the

306

E. SCHLITTLER

17-demethoxy compound, whereas pyrrolidine produced the normal

18-a-amino derivative as well as the corresponding 17-demethoxy compound (XIII). The 18-a-amino derivative had a %is0 configuration as shown by its IR-spectrum.

\A MeOOC OMe mp 196"-201°, [a]?? +3S.5°

6Me mp 206"-207", [a]?," -56' I

\A

/

.-?*

MeOOC'

OU

\ / 7

\/\

MeOOC/'\,/iOSOzPhBr OMe

OMe mp 236"-237"

OMe mp 216", [ a ] o -65'

CHART111. Some transformations of methyl 18-ketoreserpate.

Whenever a hydrogenation step was required in the study of methyl ketoreserpate, there was always danger of inversion taking place at C-3 ;

13. RauwolJia ALKALOIDS

307

AND CHEMISTRY OF RESERPINE

thus, Adams catalyst was found to be a particularly potent isomerizing agent under hydrogenating conditions [this has also been noted for reserpine -+ isoreserpine and rescinnamine -+dihydroisorescinnamine (164)l. For this reason, palladium on charcoal, being somewhat less

z\

X

XI11

\

GC,-

$)mime ( 2 ) PClS

MeOOC \ XI

XI1

I OMe

Methyl 18-pyrrolidino-17-demethoxy18.deoxyreserpate HdR1 ‘RI

y), \i\i

MeOOC’VhO

MeOOC/\\/hO

+MeOOC/

bMe

/[ MeONa

OMe

>I?

NaOMe

IS

KeOOC/

+ PtO,

M e O O a N = O ]

\/NO

Methyl 18-keto1 5-epireserpate

-

MeOOC&IUOH

XIV

CHARTIV. Further transformations of m e t h y l 18-ketoreserpate.

active a C-3 isomerizing agent (except in the presence of triethylamine), was used for the reductive aminations of methyl 18-ketoreserpate. C-3 isomerization was certainly less rapid in the presence of diethylamine or propylamine, but turned out to be fast for pyrrolidine, which led to a pharmacologically inactive methyl 18-a-pyrrolidino-3-isoreserpate.

308

.

E SCHLITTLER

Elimination of the (2-17 methoxyl group was observed when either methyl ketoreserpate oxime (XIV) or semicarbazone was treated with methanolic sodium methoxide. The suggested course of the reaction is analogous to that given for the formation of a,P-unsaturated ketone phenylhydrazones from a-bronio ketones and phenylhydrazine ( 169). Although the double bond in methyl Al'-ketoreserpate is in the stable position, it can move into the P,y-position with respect to the carbonyl function, since under the proper experimental conditions (methanolic sodium niethoxide) methyl 18-keto-15-epireserpate [mirror image of yohimban, of which there are as yet no natural examples (165, 166)l is formed in yields up to 50%. (2-3 isomerization is observed in the form of a very low yield of methyl 18-keto-~l~~-3-isoreserpate when the semicarbazone is subjected to acidic cleavage. Treatment of the oxime with phosphorus pentachloride resulted in a second-order Beckmann rearrangement (161, 167, 168) generating XI, which offers obvious possibilities for the conversion of reserpine into its ring E seco equivalent. Pharmacological investigation of the many derivatives of methyl 18ketoreserpate gave few interesting results, since the majority were inactive. Methyl 18-ketoreserpate (IX) itself was not antihypertensive, but had about one-fifth the sedative activity of reserpine, lasting more than 24 hours, which a t first sight is surprising for a compound without the diester structure. Methyl 18-ketoisoreserpate,prepared by palladiumcharcoal reduction of methyl 3-dehydro- 18-ketoreserpate, was inactive. Methyl 18-epi-isoreserpate was neither sedative nor hypotensive. Methyl 18-pyrrolidino- and 18-piperidino-17-demethoxy-18-deoxyreserpates possess curare activity. The pyrrolidino derivative (XIII) has the same potency as Intocostrin upon intravenous administration to a dog, but it shows some interesting differences from the latter. For example, a t an intravenous dose of 150 pg/kg, the curare activity of the pyrrolidino compound comes on slowly and has an activity of several hours, whereas a t the same dose of Intocostrin, the activity appears within 5 minutes and has subsided within half an hour. An oral dose of 1 mg/kg of the pyrrolidino compound did not cause curare effects in the dog until the following day, but persisted for 5 days. I n comparison, Intocostrin did not have any effect after a dose of 8.0 mg/kg orally in the dog (159).

4 . Conversion of Reserpine into Inverted Reserpine and into Oxindoles It has recently been shown that the AB rings of the yohimbinoid alkaloids can be inverted with respect to the CDE rings, thus making available the y-carboline equivalents of the original ,6-carbolines. This conversion was made possible by the discovery of a route (170) t o the pseudoindoxyl derivatives of tetrahydro-/?-carbolines which have

13. RauwolJia

309

ALKALOIDS AND CHEMISTRY OF RESERPINE

preserved the stereochemistry and functionalities of the ring substituents. When reserpine, e.g., is reacted with one mole equivalent of a lead tetraacylate (e.g., acetyl, benzoyl, m-bromobenzoyl), the corresponding 7-acyloxy-7H-reserpine (XV) is formed, which upon alkaline methanolysis gives rise to a 7-hydroxy-7H derivative (XVI). The lastnamed compound, under more vigorous conditions, gives a mixture of methyl reserpate pseudoindoxyl and its 3-is0 derivative (XVII) (171).

!

OOMP

H NaBHI

XVIII Inverted compound

XVII

CHARTV . Invert reserpines and oxindole derivatives.

Esterification of the 18-hydroxyl followed by an acid-catalyzed equilibration (at (2-3, cf. isoreserpine -+reserpine) resulted in the formation and isolation of the inverted equivalent of reserpine (XVIII). The conversion into the reserpine configuration requires a comment, since it is probably facilitated in this case by an adverse interaction in the trans-trans-cis form between the C-9 proton and the methylene group at C-14. Congeners of reserpine were also subjected to this sequence of

310

E. SCHLITTLER

steps, but the new compounds lacked the hypotensive and sedative properties so characteristic of the parent indoles. When 7 -acyl-7H-reserpine derivatives (XV) are refluxed in dilute methanol with a few drops of acetic acid, conversion into the corresponding oxindole (XIX) with concomitant formation of the fivemembered spiro-ring C takes place. Under these conditions, also, carbon

OMe

xx

OMe

XXI rnp 242"-244",

-45" (CHCh)

i

H z + P d (5%) on charcoal

XXLI ( ? )

rnp 248"-250", [ a ] k

bMe

+ 139"

OMe

XXIII rnp 262"-265",

-

103"

CHARTV I . Hexacyclic derivatives of methyl reserpate.

atom 3 becomes inverted and its hydrogen is now in cr-position. Oxindole formation from yohimbinoid alkaloids has become a valuable tool for structure elucidation of the oxindole alkaloid group. Applicability of this method depends in part on the DE ring fusion, and it has been investigated in detail ( 172).

5. Hexacyclic Derivatives of Methyl Reserpate When methyl 3-dehydroreserpate (XX) is treated with ethylbromoacetate a t steam bath temperature, a new hexacyclic methyl reserpate derivative (XXI) is obtained (173). This reaction has also been carried

13. RauwoZJa

ALKALOIDS AND CHEMISTRY OF RESERPINE

31 1

out with the 3-dehydro derivative of the corresponding methyl 18epireserpate, the 18-alkylethers, and with reserpine itself. The substituents in position 18 influence to some extent the ease of this addition reaction, the lowest yields being obtained with reserpine. Catalytic reduction of the unsaturated addition product with 5% palladium on charcoal gave a single product (XXIII). When reduced with zinc and perchloric acid, the same compound was obtained only as a by-product, the major product being a derivative with an additional ring and only two methoxyl groups (XXII). Addition reactions to methyl 3-dehydroreserpate have been studied in some detail; unfortunately, none of the numerous compounds prepared possessed any pharmacologically interesting properties.

C. SEMISYNTHETIC RESERPINES 1. 18-Acyl Derivatives of Reserpic Acid and Methyl Reserpate Soon after the discovery of reserpine, work was initiated t o prepare derivatives with possibly higher and/or modified activities or with fewer side reactions. Alteration of the antihypertensive activity was more often aimed at in these investigations than tranquilizing activity. However, sedation could also be affected to some extent. The easiest approach to the problem of qualitative and quantitative changes was the variation of the ester groupings in positions 16 and 18 (126, 164). Among a very large number of diesters prepared, none proved to be quantitatively more active than reserpine and deserpidine. I n some compounds, the sedative activity was more reduced as compared with the antihypertensive activity, although a t the expense of total strength. With high doses of such compounds, the tranquilizing activity became noticeable again. Among the more than 100 diesters prepared from reserpic acid or methyl reserpate by partial synthesis was syrosingopine (XXIV) (174, 175), which is hypotensive in humans at an average daily dose of 3 mg (reserpine, 0.3 mg) with only minimal sedation. I n contrast, Su-51711 (XXV) proved to be a compound with practically no antihypertensive properties but with a tranquilizing activity (in animals) about half as active as reserpine. su-5171 was clinically investigated, but not introduced (176). Also, the pivalic ester of methyl reserpate (177) is claimed to possess sedative but no antihypertensive activity, but it 1 The designation “Su- ” is used for experimental compounds synthesized in the laboratories of CIBA Pharmaceutical Company, Summit, New Jersey.

312

E. SCHLITTLER

likewise has not been marketed. Thus, ester modifications in positions 16 and 18 have not yet produced spectacular results. Besides information given in the two papers already cited, additional details may be found in the patent literature (178).

XXIV Syrosingopine

MeOOC/j/'\O

I

OMe

xxv Su-5171

2. Methyl Reserpate Ethers The desire to prepare reserpine derivatives with higher and/or modified activities or with fewer side reactions also led to the preparation of methyl reserpate-18-ethers. When this work was started, the sedative activities of methyl 18-ketoreserpate and methyl 18-epireserpate (see preceding discussion) were not known and the problem was therefore approached entirely with the experience gained in the field of the methyl reserpate esters, viz., that again both functional groups in 16 and 18 had to be reacted. Since 18-lower fatty esters (acetate, propionate, and so forth) had only reduced activity, it was assumed that an ether group in position 18 had to have a certain bulkiness in order to give an active compound (151, 179). The first series of 18-alkoxy compounds (e.g., XXVI) was prepared by reaction of methyl reserpate with diazoalkanes in the presence of fluoboric acid (route A, Chart VII). Poor yields of this reaction cannot be increased by the use of a large excess of diazoalkanes which leads to the formation of Nb quaternary compounds (XXVII).

13. RauwolJia ALKALOIDS

AND CHEMISTRY OF RESERPINE

313

The use of aluminum isopropylate as a catalyst for this alkylation is apparently advantageous (128, 180). The resulting 18-ethers are formulated as 18-P-alkoxy compounds, since it has been demonstrated in the steroid series that this alkylation proceeds with retention of the configuration ( 181).

I MeOOC/bLOMe

MeOOC/'\/\OMe

OMe

6Me XXVI

XXVII

MeOOC'

MeOOC/ 6Me XXVIII

6Me

XXIX

CHARTVII. Ethers of methyl reserpate and methyl 18-epireserpate.

The unusually marked sedative activity of these ethers initiated an intensive search for alternative methods of synthesis. I n the course of this investigation, a method was discovered by which ethers epimeric a t C-18 (e.g., XXIX) could be prepared in high yield by alcoholysis of

314

E. SCIZLITTLER

methyl reserpate p-bromobenzenesulfonate (XXVIII) under mildly alkaline conditions a t 100" (route B, Chart VII). A number of structural studies demonstrated that the two methyl ethers differed exclusively in the stereochemistry in position 18. In route B, inversion occurs in the alcoholysis of the sulfonate ester, and in this case methyl 18-epireserpate is obtained when the corresponding brosylate (XXVIII) is refluxed with water, dioxane, and a small amount of triethylamine under nitrogen for 2 days. Methyl 18-epireserpate can be methylated with diazomethane-fluoboric acid (without inversion) to the 18-epi ether, although only in low yields. Pharmacological evaluation of the ethers and their water-soluble salts revealed a marked sedative action in dogs without any demonstrable antihypertensive effect. The tranquilizing activity differs from that of reserpine in that its onset occurs within minutes rather than hours and the duration of action is considerably shorter than that of reserpine. Cumulation is not evident on repeated administration; there is no effect on the gastrointestinal tract and no diarrhea occurs (182). Clinically, the normal, as well as the 18-epi, ethers were found to be effective when given to patients with mild or moderate anxiety. However, the tranquilizing activity of the normal ethers is considerably higher, and for all practical purposes the 18-epi ethers are not suitable for human therapy.

R

MeOOW

xxx

MeOOW

OMe

XXXI

,

-.R

OMe

C H ~R = -SCeHs (Su-8438);(Su-8463)(151) I - SCHzCeHs ; and

-o/

-O\o'/ SU-7064

SU-8055

SU-8344

\OMe Su-8607

Somewhat more difficult is the pharmacological evaluation of 18tetrahydropyranyl derivatives (XXX) and of 18-sulfides of methyl reserpate or methyl 18-epireserpate (XXXI). I n general, the tetrahydropyranyl ethers and their analogs show both sedative and antihypertensive activity, albeit in a reduced way. It seems that the

13. RauwolJia

ALKALOIDS AND CHEMISTRY OF RESERPINE

315

antihypertensive activity can be somewhat increased by preparing higher esters in position 16. A number of analogous sulfides (XXXI) were prcpared when p toluene sulfonate esters of methyl reserpate were treated with sodium salts of thiophenols or of methyl mercaptan. It can be assumed that, in this reaction, inversion takes place at position 18 as well. Some of the sulfides thus obtained show some antihypertensive activity when given intravenously to dogs, but they are not active orally. No sedative effects were observed with these compounds. If the alcoholysis of the foregoing p-bromobenzenesulfonate esters is carried out with higher alcohols under reflux or with any alcohol in a pressure vessel a t temperatures above 120°, partial isomerization at carbon atom 3 takes place, and mixtures of 3-normal- and 3-iso-18-epi ethers are obtained. Of special interest is the alcoholysis effected with glycol monomethyl ether where separation of the mixture proved to be possible, as it is in many other cases. The 18-epi compound (XXXII) was antihypertensive, whereas the 3-iso-18-epi ether (XXXIII) had a stimulant activity. This was actually the first compound from any of t,he 3-is0 series possessing any pharmacological activity.

MeOCHzCHzOOC/\/1'-OCH~CHzOMe i OMe XXXII Antihypertensive, nonsedative (hydrochlorideSu-9671)

OMe XXXIII Stimulant (maleate salt

=

Su-11.279)

Again, methylation of N, led t o inactive compounds, as in the case of reserpine itself (cf. 152).

316

.

E SCHLITTLER

III. Synthetic Work A. THE TOTALSYNTHESIS

1. Synthetic Outline Since its publication in 1956 and 1958, the Woodward synthesis of reserpine (183) has been modified by French, Czech, and Swiss chemists with an’eye toward its commercial exploitation, and it is to the credit of French workers (L. Velluz and colleagues) that today synthetic reserpine (and especially deserpidine) is competitive in price with reserpine extracted from plant material (184, 185). All these synthetic modifications of Woodward’s original approach either change the order of the steps and/or introduce simplifications into the highly complicated procedure. It is only natural that attempts to increase the yields of the individual reaction steps were intensely investigated. Since resolution into optically active intermediates a t the very end of a synthesis is wasteful, it is now done a t an early stage. Also, the introduction of the trimethoxybenzoyl radical into the potential 18 position is now effected earlier, which eliminates the need at some stage of a hydrolysis of the 18-acetyl group to the free hydroxyl and re-esterification with trimethoxybenzoyl chloride. Reduction of the A 3 compound directly to the normal series avoided the necessity t o prepare and isomerize the C-3 isolactone. The remarkable feature of Woodward’s procedure was the facile way in which the five adjacent stereochemical centers in ring E were built into the critical intermediate, the aldehyde acid (XLII) (Chart VIII). Parenthetically, the preparation of this compound has opened up the possibility of a general synthesis of all the yohimbines and their isomers, although this has not yet been done. Condensation of the aldehyde acid with 6-methoxytryptamine gave a Schiff base which was reduced in situ with sodium borohydride to furnish the amide (XLIII) which was ringclosed to the A 3 compound XLIV and subsequently transformed into the methyl dl-isoreserpate-0-acetate (XLVIII). I n the original synthesis, there still remained three problems : (a) conversion of the %is0 compound into the %normal; (b) replacement of the 18-0-acetate by a trimethoxybenzoyloxy group ; and ( c )the resolution of the racemic alkaloid. The last two problems require no detailed comments, but the solution provided for the fist led t o the reserpine configuration, resulting in an unequivocal proof of the stereochemistry a t C-3. The lactone of dl-isoreserpic acid was prepared, and this, upon refluxing in pivalic acid, was converted into the thermodynamically

13. Rauwolfia

317

ALKALOIDS AND CHEMISTRY OF RESERPINE

xxxv <-PIOH

cMeOH

0

,/

XXXVI

0

OM0

N.B.S.-HIO-H~~O~

Br“:Zn

Br

%n+AcOH ___f

AcOH

0

‘C

OH

HOOW

b

0

6M0

6M0

OM0

XLI (1) CHaYs (2) AcaO + Pyridinr

OH OSOI

(2) CH2N,

OAc

MeOOC/ OM0 XLII

OM0

CHARTVIII. Creation of ring E of reserpine.

OM0 XL

318

E . SCHLITTLER

more stable dl-reserpic acid lactone ; then, following known reactions, the lactone was converted by methanolysis into methyl dl-reserpate and to the corresponding trimethoxybenzoate. Resolution using the dcamphor-10-sulfonate was the last step.

2. Preparation of the Key Intermediate The Diels-Alder condensation of p-benzoquinone with vinylacrylic acid gives a cis decalin adduct (XXXIV),which is a prerequisite for the desired DE ring junction of the reserpine molecule. This cis configuration is preserved by proper choice of reagents, and in only one single instance has a transformation into a trans decalin been reported (186). In his “modified route,” Woodward has used methyl acrylate instead of the free acid, and others have used the corresponding ethyl or isopropyl esters (187, 186). The subsequent steps in the French and Czech modifications follow the Woodward synthesis in its broad lines, but it should be noted that Velluz already resolves Woodward’s XXXV into optical isomers with the help of either brucine or ephedrine (185). In an analogous study, the Czech investigators resolve a hydroxy acid (XXXVIII), obtained by hydrolysis of the Velluz compound (XXXVII) (185), with the help of brucine (187). Also, another dihydroxy acid (XXXIX), obtained from Woodward’s lactone acid (XXXVI),has been resolved, again by use of

I/

6

0 XXXVII

XXXVIII

XXXVI

XXXIX

brucine (187,188).Woodward had already pointed out that ring cleavage of XL could be done directly. However, in his own paper, the a$unsaturated ketone was first oxidized to a diol by means of osmium tetroxide and sodium chlorate, and ring cleavage was then effected by treatment with aqueous periodic acid (Chart VIII). I n the Velluz synthesis, XL is oxidized by means of ozone, whereas the Czech workers follow the original Woodward procedure (189). It should also be pointed out that the analogous cleavage process has been carried out with a compound containing a lactone ring that was un-

13. RauwolJia

ALKALOIDS AND CHEMISTRY OF RESERPINE

OMe

319

OMe XLVI

affected by the subsequent ozonization and condensation with tryptamine (190). I n another interesting modification (191), the trimethoxybenzoyl radical is already introduced into XLI and survives all further reaction steps in its potential 18 position.

-

MeO, XLI

+ MeO-

/rcocl

-

OMe

-\OMe

COOH I CHO CHs I Me00

,OMe

J+JLwe OMe XLV

-

\OMe

3. Ring Closure to the A 3 Compounds The condensation of XLII or one of its derivatives with tryptamine or a substituted tryptamine is a feature common to all modifications. Ring closure (XLIII +XLIV) is finally effected with boiling phosphorus oxychloride in excellent yield. Also the 18-trimethoxybenzoate (XLV) can undergo the same series of reactions as given for the acetate XLII. Ring closure with phosphorus oxychloride to an analogous quaternary compound (XLVII)can also be carried out with the racemic or optically active lactone XLVI (192).

320

E. SCHLITTLER

XLII

NaB&

1

H‘ Me00W XLIV

MeOOC’

OAc

OMe

I

OAc

OMe

XLIII

i-h-

POClS

M~OOC/\/\OH

--f

1,’

I

OMe

coo I

XLVI

OMe

4 . Reduction of the A3 Compounds Woodward’s reduction of the quaternary dehydro compound (XLIV) with sodium borohydride had led to the 3-is0 ( 3 ~ series ) (XL). As already mentioned, the corresponding 3-isolactone (XLIX) was then

13. Rauwoljia ALKALOIDS

321

AND CHEMISTRY OF RESERPINE

prepared by treatment of the free isoreserpic acid with N,N’-dicyclohexyl carbodiimide in pyridine. By treatment with pivalic acid, this 3-isolactone (XLIX) was transformed into the thermodynamically more stable lactone L. Hydrolysis with sodium methoxide and acylation with trimethoxybenzoyl chloride finally yielded dl-reserpine (LI),which was resolved by use of camphor sulfonic acid.

NaBHeMeO’

o&J?N

+

\

‘s_i

M~OOW XLVIII

,

OMe

OAc

OMe

XLIX Yivalic acid

MeO’

I

OMe

L

OMe

\ (2::; aoc-/

\-me

LI

The industrially preferred procedure is the reduction of the quaternary perchlorate with zinc and perchloric acid in a mixture of tetrahydrofuran and acetone (185),and it is claimed that reserpine is obtained exclusively. I n a later study (190),reduction of the quaternary dl-dehydro perchlorate with the same reducing agents gave a mixture of dl-methyl reserpate and dl-isoreserpate in a ratio of 311, readily separable by alumina chromatography. It has been pointed out by the same group of investigators that, in the synthesis of deserpidine, this reduction is more stereospecific and gives as the sole product the 3/3 form (193). I n another paper ( 191), quaternary dehydro compounds were reduced catalytically and with sodium borohydride, both in the presence of perchlorate ion. Only is0 compounds are obtained, and, therefore, the anion itself does not seem to possess a directing influence.

322

E. SCHLITTLER

5. Introduction of the TrinzethoxybenzoylRadical Trimethoxybenzoylation of XLI has already been mentioned under Section 111,A. The French synthesis introduces the trimethoxybenzoyl group immediately before ring closure, thus avoiding the use of methyl reserpate for the final step. The same process has also been reported later in another paper (189). I n many cases, the trimethoxybenzoyl group has been introduced into synthetic methyl reserpate as the last or second-to-last step. An optimal yield of 72% of reserpine is reported if a 50% excess of the acid chloride is used (190, 193),the yield diminishing as the amount of acylating agent is increased. In the French synthesis, acylation is carried out with trimethoxybenzoic anhydride in triethylamine and pyridine ( 185).

6 . Other Variants Here the synthesis of a dl-19-methyl homolog of deserpidine (LII)has t o be mentioned (194). For the initial Diels-Alder condensation, methyl sorbate was used instead of methyl vinylacrylate; for the rest, this synthesis followed well-trodden paths. French workers have also

OMe

LII

,OMe

LIII

reported the synthesis of 17a-cyano- and 17a-methyl acetamido-17demethoxyreserpine (195). A different approach to synthetic deserpidine is offered by Weisenborn (196). Although it did not reach its ultimate aim, it did yield dl-17-

13. RauwolJia ALKALOIDS

AND CHEMISTRY OF RESERPINE

323

demethoxydeserpidine (LIII), which possessed a typical but weak reserpine-like activity. Apart from these variations (listed in Table V I and requiring no further comment), one other variant of the reserpine molecule was TABLE VI

RESERPINE ANALOGS

Analog Ring A-substituted deserpidines" 10-Acetyl-11-methoxyc 12-Acetyl-l l-methoxy" 12-Aza 9,lO-Benzo 10-Benzyloxy 11-Benzyloxy 10-Butoxy 11-Butoxy 12-Butoxy 10-Chloro 11-Chloro 12-Chloro 9-Chloro-12-Methoxy 10-Chloro-11-methoxy 11-Chloro-lZ-methoxy 12-Chloro-11-methoxy 9,12-Dichloro 11,lZ-Dichloro 10,ll-Dimethoxy 11-Dimethylamino dl-10-Ethoxy 11-Ethoxy 11-Ethylmercapto dZ-10-Fluoro 11-Fluoro

Melting point ("C)

254-256 274-2 7 8 290 248 150-152 170 210 206 209 160-1 70 280 179

-

[.IDb

- 47 - 141 - 157 - 172 1 - 152 (P)

- 100 - 132 - 96 - 120 - 147 - 125 (E) - 116 - 132

300

- 120

183

- 129 - 106

230 172 222 264-266 150-1 60 230-232 282-283 250-252 230 261

- 119 - 97 - 157 (P) - 158 (P) - 160 (P) -

- 123

Reference

155 155 195 185, 202 199 200 200 200 200 185 185 185 185 185 185 201 202 202 199 195 203 199 199 204 205,202

E . SCHLITTLER

324

TABLE VI-continued

Analog 11-Hydroxy 9-Methoxy 10-Methoxy 12-Methoxy 9-Methyl dl-10-Methyl (hemihydrate) 11-Methyl 12-Methyl 12-Methyl-11-methoxy 10,ll-Methylenedioxy d-l0,ll-Methylenedioxy dl-10-Methylmercapto (hemihydrate) 11-Methylmercapto 11-Propoxy 11-Isopropoxy 9,10,11-Trimethoxy

Ring C-substituted deserpidines 3-Cyano-11-methoxy 6-P-Ethyl-10-chloro 6-a-Ethyl-10-methoxy 6-8-Ethyl-10-methoxy 6-Ethyl-11-methoxy C-Homo-11-methoxy 5-a-Methyl-11-methoxy 5-8-Methyl-1I-methoxy 6-a-Methyl-11-methoxy 6-P-Methyl-11-methoxy

Melting point ("C) 166-168 203-205 171 200 175-180 224-226 275 231 -

251-253 252-254 206 27 0-27 2 215-2 17 270 171-173

205 265 268 225 130 170,220,242 254 190-191 150, 223 220

[.IDb

- 165 (P) - 133 (P) - 142

- 120 - 137 (P)

-

- 132 - 124 -

- 164 (P)

+ 158 (P)

-

- 164 (P) - 160 (P) - 125 -111

- 136

- 80 - 129 - 105

- 82

- 125 (P)

- 165 (P) - 141 (P) -187 - 134

Reference 199 199 185 185 199 203 185 185 206 199 199 203 199 199 200,202 185

195 195 195 195 207,202 195, 208 199 199 202 202

Ring E-modified deserpidines dl-17-Demethoxy 17-Demethoxy-17-a-cyano-11-methoxy 17-Demethoxy-17-a-(N-acetyl-N-methylamino)-11-methoxy 17-Demethoxy-l7-ethoxy-ll-methoxy 17-Demethoxy-17-isopropoxy11-methoxy 17-Demethoxy-17-n-propoxy-l l-methoxy

247-248 297 300

- 110

- 127

196 195 195

240-242 248-250 215-217

- 149 (P) - 125 (P) - 157 (P)

199 199 199

a Compounds listed are the 1-enantiomers, corresponding to the naturally occurring I-reserpine, except as noted. * Rotations are given for chloroform solution except where noted: P = pyridine, E = ethanol. Compound partially synthetic.

13. Rauwolfia

ALKALOIDS AND CHEMISTRY OF RESERPINE

325

obtained by condensation of the aldehydo acid (XLII) with ,i3-phenylethylamines instead of tryptamine (195,197).Thus, hexahydroprotoberberine-like compounds (LIV) were obtained, for which the names of ) epialloberban (3P) have been proposed (197). No alloberban ( 3 ~ and detailed paper on the pharmacology of these compounds has yet appeared, although some reserpine-like activity has been claimed. These isoquinoline derivatives were synthesized along the lines already discussed and were racemic, except for the case described by Mueller and Allais (195).

LIV R = H (198) R = OMe (197)

B. SYNTHESES OF SIMPLEA N ~ L O G S One of the aims of pharmaceutical research is to profit by the knowledge of the structure of a useful naturally occurring active material by synthesizing simpler, effective compounds which would he more easily accessible and which would have more desirable or even different biological activities. During the last 10 years, a number of papers on syntheses of RauwolJia model compounds have been published. However, no compound with real therapeutically tranquilizing activity has been found whose structure could be connected in any way with the reserpine structure. I n some publications, a “reserpine-like ” activity is claimed, but such claims apply mostly t o an antihypertensive activity, which is rarely of the reserpine-type.” Another reason for the apparent failure to find active analogs is that there is no real incentive, since in the fields of both tranquilization and hypertension other unrelated synthetic compounds have been developed. With these reservations in mind, it is, however, interesting to point to a few contributions in this field. A number of papers have been published on the construction of the pentacyclic yohimbinoid ring system of reserpine (or yohimban) with an aromatic ring E, according to a procedure worked out more than 25 years ago (209, 210, 211, 212, 213, 214, 215). ((

326

E. SCHLITTLER

An attempt to synthesize a tetracyclic p-carboline system (LV) with rings A, B, C, and D analogous to reserpine is worth mentioning. It is claimed that by esterification with trimethoxybenzoyl chloride, physiological activity is obtained, but so far no pharmacological or clinical data have been published (216).

a-1 m)

A

Tryptamine 3-

1. Pd/Ha

__z

EtOOC( CH2)zCCOOEt II

CHOH

HNCH ---+ 2. A /I CCOOEt / EtOOC(CH2)Z

H

N

H

I

OC/”CHz

I

I

HZC\C/CHCOOEt Hz

Miller and Weinberg (217) have published data on reserpine-like activity of simple trimethoxybenzoates of amino alcohols and have claimed tranquilizing activity for compounds like the following :

Although these claims could not be confirmed (159), this publication has initiated a number of other investigations (218, 219, 220, 221, 222, 223, 224, 225, 226). Literature concerning model compounds is summarized in a recent publication (227), and subsequent papers by the same authors deal in greater detail with such work (228). From the material already presented in this chapter, it is obvious that structure-activity relationships in the reserpine field are still poorly understood, and so far the syntheses of model compounds, often prepared to test preconceived ideas, have not led to useful compounds. I n the present state of our knowledge, i t has been more profitable to discover substances whose clinical effects resemble reserpine as a result of careful pharmacological screening of

13. RauwolJia ALKALOIDS

AND CHEMISTRY OF RESERPINE

327

compounds, no matter what the reason for their synthesis. Analogous cases in other areas occur widely throughout medicinal chemistry. Tetrabenazine or 2-oxo-3-isobutyl-9,lO-dimethoxyhexahydrobenzo[u]quinolizine (LVI) (229))which has apparently emanated from work related to the synthesis of emetine, has obtained much attention recently. I n spite of the absence of an indole ring system, it seems to possess sedative reserpine-like activity. Tetrabenazine has been pharmacologically studied in great detail (230).

LVI

LVII

A similar compound is benzquinamide (LVII) (231), which has been experimentally employed as a psychotherapeutic agent in overt anxiety (232).Neither of the two Compounds is as yet commercially available in the United States, although tetrabenazine has been introduced in some European countries. REFERENCES 1. S. Bose, S. K. Talapatra, and A. Chatterjee, J . Indian Chem. Soc. 33, 379 (1956). 2. A. Stoll and A. Hofmann, SOC.Biol. chemists India Silver Jubilee Souvenir p. 248 (1955); J . Am. Chem. SOC.77, 820 (1955). 3. J. Keck, Naturwksemchaften42, 391 (1955). 4. A. Chatterjee, Proc. Symp. Phytochem. Kuala Lumpur, 1957 p. 142 (1957). 5. M. Ishidate, M. Okada, and K. Saito, Pharm. Bull. (Tokyo) 3, 319 (1955). 6. S. K. Talapatra, J . Sci. I d . Res. (India)21B, 198 (1962). 7. J. M. Mueller, Ezperientia 13, 479 (1957). 8. J. Shavel, G. Kane, and J. A. King, Am. Chem. SOC.,126th Meeting New York, 1954, p. 25N (1954). 9. R. Sakin, N. Hosansky, and R. Jaret, J . Pharm. Sci. 50, 1038 (1961). 10. S. Siddiqui and R. H. Siddiqui, J . Indian Chem. Soc. 8, 667 (1931). 11. N. A. Chaudhury and A. Chatterjee, J. Sci. I d . Res. (India)18B,130 (1959). 12. S. C. Pakrashi, C. Djerassi, R. Wasicky, and N. Neuss, J . Am. Chem. SOC.77, 6687 (1955). 13. R. Paris, Ann. Phrm. Franc. 1, 138 (1943). 14. H. R. Arthur, Chem. I d . (London)p. 85 (1956). 15. B. P. Korzun, A. F. St. Andr6, and P. R. ULshafer, J. Am. Pharm. Assoc. Sci. Ed. 46, 720 (1957); and Unpublished observations. 16. G. Iacobucci and V. Deulofeu, Anales Asoc. Quim.Arg. 46, 143 (1958). 17. B. P. Ghosh, Natumuissenschaften 45, 365 (1958). 18. J. B. Koepfli,J. Am. Chem. Soc. 54, 2412 (1932).

328 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51.

52. 53. 54.

55.

E . SCHLITTLER

A. Chatterjee and S. Talapatra, Naturwissenschaften 42, 182 (1955). N. Finch, P. R. Ulshafer, and W. I . Taylor, Ezperientia 19, 296 (1963). F. A. Hochstein, K. Murai, and W. H. Boegemann,J. Am. Chem. SOC.77,3551 (1955). M. Ishidate, M. Okada, and K. Saito, Pharm. Bull. (Tokyo) 3, 320 (1955). B. 0. G. Schuler and F. L. Warren, J . Chem. Soc. p. 215 (1956). M. Roland, J . Pharm. Belg. 14,347 (1959). A. K. Kiang and A. S. C. Wan, J. Chem. Soc. p. 1394 (1960). X. Monseur, J . Pharm. Be@. 12,39 (1957), G. Iacobucci and V. Deulofeu, J . Org. Chem. 22, 94 (1957). C. Djerassi, S. C. Pakrashi, and M. Gorman, cited by A. Chatterjee, S. C. Pakrashi, and G. Werner, Fortschr. Chem. Org. Naturstofle 13, 403 (1956). S. Siddiqui, J . Indian Chem. SOC. 16, 421 (1939). R. A. Seba, J. S. Campos, and J. G. Kuhlmann, Bol. Inst. Vital (Brazil)5, 175 (1954). B. U. Vergara, J . Am. Chem. Soc. 77, 1864 (1955); Ann. Soc. Biol. Bogota 6 , 277 (1956). E. Schlittler, H. Schwarz, and F. Bader, Helv. Ghim. Acta 35, 271 (1952). R. M. Bernal, A. Villegas-Castillo, and 0. P. Espejo, Ezperientia 16, 353 (1960). A. Stoll, A. Hofmann, and R. Brunner, Helv. Chim. Acta 38, 270 (1955). L. Gomez, P. da Silva, and P. Garria, J . Philippine Pharm. Assoc. 44, 127 (1957). S. Bhattacharji, M. M. Dhar, and M. L. Dhar, J . Sci. Ind. Res. ( I n d i a ) 15B, 506 (1956);J . Sci. Ind. Res. (India)16B, 97 (1957). E. C. Deger, Arch. Pharm. 275, 496 (1937); R. Paris and R. Mendoza, Bull. Sci. Pharmacol. 48, 146 (1941). A. Hofmann, Helv. Chim. Acta 37, 314 (1954). M. Gorman, N. Neuss, C. Djerassi, J. P. Kutneg, and P. J. Scheuer, Tetrahedron 1, 328 (1957). A. S. C. Wan and A. K. Kiang, Proc. Symp. Phytochem., Kuala Lumpur, 1957, p. 181 (1957). P. R. Ulshafer, M. F. Bartlett, L. Dorfman, M. A. Gillen, E. Schlittler, and E. Wenkert, Tetrahedron Letters No. 11, 363 (1961). P. R. Ulshafer, M. L. Pandow, and 9. H. Nugent, J . Org. Chem. 21, 923 (1956). J. Poisson, A. Le Hir, R. Goutarel, and M. M. Janot, Compt. Rend. Acad. Sci. 239, 302 (1954). N. Hosansky and E. Smith, J . Am. Pharm. Assoc. Sci. Ed. 44, 639 (1955). P. P. Pillay, D. S. Rao, and S. B. Rao, J . Sci. Ind. Res. ( I n d i a ) 19B, 135 (1960). R. Goutarel, M. Gut, and J. Parello, Compt. Rend. Acad. Sci.253, 2589 (1961). J. Poisson, R. Goutarel, and M. M. Janot, Compt. Rend. Acad. Sci.241, 1840 (1955). A. Chatterjee and S. Bose, Sci. Cult. (Calcutta)17, 139 (1951). P. R. Ulshafer, W. I. Taylor, and R. H. Nugent, Compt. Rend. Acad. Sci. 244, 2989 (1957). A. Popelak, E. Haack, G. Lettenbauer, and H. Spingler, Naturwissenschaften 48, 73 (1961). E. Haack, A. Popelak, H. Spingler, and F. Kaiser, Naturwissenschajten 41, 214 (1954); see also M. W. Klohs, M. D. Draper, and F. Keller, J . Am. Chem. SOC.76, 2843 (1954); and E. Haack, A. Popelak, and H. Spingler, Naturwissenschaften 42, 47 (1955). D. A. A. Kidd, Chem. Ind. (London) p. 1481 (1956). K. Atal, J . Am. Pharm. Assoc. Sci. Ed. 48, 37 (1959). R. Pernet, J. Philippe, and G. Combes, Ann. Pharm. Franc. 20, 527 (1962). D. S. Rao and S. B. Rao, IndianJ. Pharm. 18,202 (1956).

c.

13. RauwolJia

ALKALOIDS AND CHEMISTRY OF RESERPINE

329

56. M. W. Klohs, M. D. Draper, F. Keller, and W. Malesh, Chcm. I d . (London) p. 1264 (1956). 57. E. Haack, A. Popelak, and H. Spingler, Natur?uissensc?Laften43, 328 (1956). 58. D. A. A. Kidd, J. Chcm. Soc. p. 2432 (1958). 59. J. Poisson and R. Gout,arel, Bull. Soc. Chim. Prance p. 1703 (1956). 60. M. W. Klohs, F. Keller, R. E. Williams, and G. W. Kusserow, Chem. Ind. (London) p. 187 (1956). 61. B. Rakshit, Indian Pharmacist 10, 84 (1954). 62. E. Schlittler, H. Saner, and J. M. Mueller, Ezperientin 10, 133 (1954); see also F. L. Weisenborn, M. Moore, and P. A. Diassi, Chem. Ind. (London)p. 375 (1954). 63. S. K. Talapatra and A. Chatterjee, Sci. Cult. (Calcutta)22, 692 (1957). 64. K. Bodendorf and H. Eder, Naturwissenschaftcn 40, 342 (1953); see also A. Stoll and A. Hofmann, Helw. Chim. Acta 36, 1143 (1953). 65. N. A. Chaudhury and A. Chatterjee, J . Sci. Ind. Rcs. ( I u d i a ) 18B, 398 (1959). 66. E. Schlittler and A. Furlenmeier, Helw. Chim. Acta 36, 996 (1953). 67. J. Poisson, Thesis, Facult6 de Pharmacie, Paris (1959). 68. E. Haack, A. Popelak, H. Spingler, and F. Kaiser, Xnturwissenschnftcn 41,479 (1964). 69. M. M. Janot, R. Goutarel, and A. Le Hir, Compt. Rend. Acad. Sci. 238, 720 (1954). 70. C. Djerassi and J. Fishman, Chcm. Ind. (London) p. 627 (1955). 71. A. Hofmann and A. J. Frey, Helv. Chim. Acta 40, 1866 (1957). 72. W. I. Taylor, A. J. Frey, and A. Hofmann, Helw.Chim. Acta 45, 611 (1962). 73. A. Mookerjee, J. Indian Chem. Soc. 18, 33 (1941). 74. A. Hofmann, HeZw. Chim. Acta 38, 536 (1955). 75. F. E. Bader, D. F. Dickel, and E. Schlittler, J . Am. Chenz. Soc. 76, 1695 (1954). 76. F. E. Bader, D. F. Dickel, R. A. Lucas, and E. Schlittler, Ezperientia 10, 298 (1954). 77. A. Stoll and A. Hofmann, J . Am. Chem. Soc. 77, 820 (1955). 78. P . J. Scheuer, M. Y. Chang, and H. Fukami, J . Org. Chem. 28,2641 (1963). 79. P. R. Ulshafer, Unpublished observations. 80. J. M. Mueller, cited in Reference 20. 81. E. Schlittler (CIBA), U.S. Patent No. 2,752,351 (1956) [Chcm. Abstr. 50, 15,032 (1956)l; R. Schwyzer and J. Mueller (CIBA), U.S. Patent No. 2,833,771 (1958) [Chem. Abstr. 52, 20913 (1958)l; H. W. Ordway and P. A. Guercio (Pfizer), U.S. Patent No. 3,876,228 (1959) [Chem. Abstr. 53, 11,770 (1959)l; P. R. Ulshafer (CIBA), U.S. Patent No. 2,982,769 (1961) [Chem. Abstr. 55, 17,680 (1961)l. 82. F. A. L. Anet, D. Chakravarti, Sir Robert Robinson, and E. Schlittler, J. Chem. SOC.p. 1242 (1952). 83. N. Neuss, H. E. Boaz, and J. W. Forbes, J . Am. Chem. Soc. 76, 2463 (1954). 84. N. Neuss, H. E. Boaz, and J. W. Forbes, J . Am. Chem. Soc. 76, 3234 (1954). 85. L. van Itallie and A. J. Steenhauer, Arch. Phurnz. 270, 313 (1932). 86. S.Siddiqui and R. H. Siddiqui, J . Indian Chem. SOC.9, 539 (1932). 87. A. Hofmann, Helw. Chim. Acta 37, 849 (1954). 88. N. Neuss, “Physical Data of Indole and Indoline Alkaloids,” 1701. 11. Eli Lilly and Company, Indianapolis, 1962. 89. F. E. Bader, D. F. Dickel, C. F. Huebner, R. A. Lucas, and E. Schlittler, J . Am. Chem. Soc. 77, 3547 (1955). 90. E. Schlittler and H. Saner, Organic Symp. Am. Chem. Soc., A n n Arbor, Michigkn, June 15-19,1953 (unpublished). 91. E. Schlittler, in “Rauwolfia,” Chap. 3. Little, Brown, Boston, 1957. 92. J. E. Saxton, i n “The Alkaloids” (R. H. F. Manske and H. L. Holmes, eds.), Vol. V I I , p. 66. Academic Press, New York, 1960.

330

.

E SCHLITTLER

93. J. W. E. Harrisson, E. W. Packman, E. Smith, N. Hosansky, and R. Salkin, J . Am. Pharm. Assoc. Sei. E d . 44, 688 (1955). 94. J. Shavel, Personal communication. 95. A. Popelak, E. Haack, G. Lettenbauer, and H. Spingler, Naturwissenschuften 48, 73 (1961). 96. A. Chatterjee, S. C. Pakrashi, and G. Werner, Fortschr. Chem. Org. Naturstoffe 13, 393 (1956). 97. J. P. Le Gall, Ann. Phurm. Franc. 18, 817 (1960). 98. W. I. Taylor, P. Potier, and J. Le Men, Ann. Pharm. Franc. 21, 321 (1963). 99. A. Hofmann, Planta Med. 5 , 145 (1957). 100. R. Coutarel, A. Hofmann, M. M. Janot, A. Le Hir, and N. Neuss, HeZv. Chim. Acta 40, 156 (1957). 101. K. Bodendorf and H. Eder, Ber. 87, 818 (1954). 102. R. H. Siddiqui, Thesis, Aligarh. (1934). 103. A. Chatterjee and S. Pakrashi, Naturwissenschaften 41, 215 (1954). 104. N. Neuss, H. E. Boaz, and J. W. Forbes, J . Am. Chem. SOC.77, 4087 (1955). 105. A. Chatterjee and S. Bose, Experientia 10, 246 (1954). 106. F. A. Hochstein,J. Org. Chem. 21, 1516 (1956). 107. S. Bose, Naturwissenschaften 42, 71 (1955). 108. A. Popelak, H. Spingler, and F. Kaiser, Naturwissenschuften 40, 625 (1953). 109. M. W. Klohs, M. D. Draper, F. Keller, W. Malesh, and F. J. Petracek, J . Am. Chem. SOC.76, 1332 (1954). 110. R. E. Woodson, in “Rauwolfia,” Chap. 1. Little, Brown, Boston, 1957. 111. A. F. St. AndrB, B. Korzun, and F. Weinfeldt, J . Org. Chem. 21, 480 (1956). 112. R. G. Curtis, G. J. Handley, and T. C. Somers, Chem. I n d . (London) p. 1598 (1955); W. D. Crow and Y . Greet, Australian J . Chem. 8 , 461 (1955). 113. N. K. Basu and B. Sarkar, Nature 181, 552 (1958). 114. J . S. E. Holker, M. Cais, F. A. Hochstein, and C. Djerassi, J . OTg. Chem. 24, 314 (1959). 115. E,Macko and R. F. Raffauf, J . Med. Pharm. Chem. 2, 585 (1960). 116. P. M. Lyapunova and Y . G. Borisyuk, FuTmatsevt. Zh. (Kiev) 16, 42 (1961); Chem. Abstr. 56, 10,282. 117. F. A. Doy and B. P. Moore, AustraZianJ. Chem. 15, 548 (1962). 118. E. Schlittler and W. I. Taylor, Experientia 16,244 (1960). 119. W. Malamud, W. E. Barton, A. M. Fleming, P. M. Middleton, T. T. Friedman, and M. J . Schleifer, Am. J . Psychiat. 114, 193 (1957). 120. S. Siddiqui, Chem. Ind. (London) p. 1270 (1957); S. Siddiqui, S. A. Warsi, and H. Allauddin, Pakistan J . Sei. Ind. Res. 1, 3 (1958); Pakistan J . Sci. I d . Res. 2, 80 (1959); S. S. Siddiqui, U.S. Patent Nos. 3,047,562 (1963) and 3,047,563 (1963). 121. A. J. Plummer and D. F. Dickel, Personal communication. 122. H. Merz (C. H. Boehringer Sohn), Ger. Patent No. 967,469 (1957) [Chem. Abstr. 53, 18,402 (1959)l; A. Boehringer, C. Boehringer, I. Liebrecht, and J. Liebrecht (C. H. Boehringer Sohn), Brit. Patent No. 772,122 (1957) [Chem. Abstr. 51, 14,213 (1957)l. 123. E. E. van Tamelen and C. W. Taylor, J . Am. Chem. SOC.79, 5256 (1957). 124. F. Kaiser and A. Popelak, Ber. 92, 278 (1959). 125. L. Dorfman, A. Furlenmeier, C. F. Huebner, R. A. Lucas, H. B. MacPhillamy, J. M. Miiller, E. Schlittler, R. Schwyzer, and A. F. St. Andrb, Helv. Chim. Acta 37, 59 (1954). 126. R. A. Lucas, R. J. Kiesel, and M. J. Ceglowski, J . Am. Chem. SOC.82, 493 (1960). 127. J. A. Schneider, in “Rauwolfia,” Chap. 4. Little, Brown, Boston, 1957.

13. RauwolJa

ALKALOIDS AND CHEMISTRY OF RESERPINE

331

A. Popelak and G. Lettenbauer, Arch. Pluzrm. 295,427 (1962). A. Popelak and G. Lettenbauer, Arch. Pharm. 296, 261 (1963). M. Shamma and E. F. Walker, Chem. Ind. (London)p. 1866 (1962). C. F. Huebner and E. Schlittler, J . Am. Chem. SOC.79, 250 (1957). M. W. Klohs, F. Keller, R. E. Williams, and G. W. Kusserow, J . Am. Chem. Soc. 79, 3763 (1957). 133. H. Heinemann, Ber. 67, 15 (1934). 133a. M. Shamma, Personal communication (1964). 134. P. E. Aldrich, P. A. Diassi, D. F. Dickel, C. M. Dylion, P. D. Hance, C. F. Huebner, B. Korzun, M. E. Kuehne, L. H. Liu, H. B. MacPhillamy, E. W. Robb, D. K. Roychaudhuri, E. Schlittler, A. F. St. AndrP;, E. E. vanTamelen, F. L. Weisenborn, E. Wenkert, and 0. Wintersteiner, J. Am. Chem. SOC.81, 2481 (1959). 135. P. A. Diassi, F. L. Weisenborn, C. M. Dylion, and 0. Wintersteiner, J . Am. Chem. Soc. 77, 2028 (1955). 136. C. F. Huebner and E. Wenkert, J . Am. Chem. Soc. 77, 4180 (1955). 137. R. Pepinski, J. W. Turley, Y. Okaya, T. Doyne, V. Vand, A. Shimata, F. M. Lovell, and J. Y. Sayo, Acta Cryst. 10, 811 (1957). 138. W. E. Rosen and J. N. Shoolery, J . Am. Chem. SOC.83, 4816 (1961). 139. R. B. Woodward, F. E. Bader, H. Bickel, A. J. Frey, and R. W. Kierstead, Tetrahedron 2, 1 (1958). 140. T. A. Henry, “The Plant Alkaloids,” p. 500. Churchill, London, 1949. 141. P. A. Diassi, F. L. Weisenborn, C. M. Dylion, and 0. Wintersteiner, J . Am. Chem. Soc. 77, 4687 (1955); E. E. van Tamelen and P. D. Hance, J . Am. Chem. SOC.77, 4692 (1955). 142. C. F. Huebner, M. E. Kuehne, B. Korzun, and E. Schlittler, Ezperientia 12,249 (1956). 143. W. E. Rosen and J. M. O’Connor, J . Org. Chem. 26, 3051 (1961); W. E. Rosen and H. Sheppard, J . Am. Chem. SOC.83, 4240 (1961); W. E. Rosen, Tetrahedron Letters 14, 481 (1961). 144. L. A. Mitscher, J. K. Paul, and L. Goldman, Ezperientia 19, 195 (1963). 145. F. Bohlmann, Ber. 91, 2157 (1958). 146. E. Wenkert and D. K. Roychaudhuri, J . Am. Chem. SOC.78, 0417 (1956). 147. W. Klyne, Chem. Ind. (London) p. 1198 (1954). 148. V. Prelog and H. L. Meier, Hetv. Chim. Acta 36, 320 (1953). 149. Y. Ban and 0. Yonemitsu, Chem. Ind. (London)p. 948 (1961). 150. W. 0. Godtfredsen and S. Vangedal, Acta Chem. Scnnd. 10, 1414 (1956). 151. M. M. Robison, R. A. Lucas, H. B. MacPhillamy, R. L. Dziemian, I. Hsu, and R. J. Kiesel, J. Am. Chem. SOC.83, 2694 (1961). 152. C. F. Huebner, J. Am. Chem. SOC.76, 5792 (1954). 153. M. F. Bartlett, Unpublished observations. 154. F. L. Weisenborn (Olin Mathieson Chemical Corp.), U.S. Patent No. 2,912,436 (1959) [Chem. Abstr. 54,4652 (1960)l. 155. P. R. Ulshafer, R. H. Nugent, and L. Dorfman, Chem. Ind. ( L o d o n ) p. 1863 (1962). 156. S. C. Pan and F. L. Weisenborn,J. Am. Chem.Soc. 80,47 (1958);also (Olin Mathieson Chemical Corp.), U.S. Patent No. 2,964,451 (1960) [Chem. Abstr. 55, 6779 (196l)l. 157. W. 0. Godtfredsen, G. Korsby, H. Lorck, and S. Vangedal, Ezperientia 14, 88 (1958); W. 0. Godtfredsen (Lovens Kemiske Fabrik), Brit. Patent No. 824,496 (1959) [Chem. Abstr. 54, 7768 (1960)l. 158. E. S. Patterson, R. E. Hartman, W. W. Andres, and E. F. Krause, Chirn. Ind. BeZge 27, 556 (1962). 159. A. J. Plummer, Personal communication. 160. M. M. Robison, R. A. Lucas, H. B. MacPhillamy, R. L. Dziemian, I. Hsu, R. J. 128. 129. 130. 131. 132.

332

E. SCHLITTLER

Kiesel, and M. J. Morris, Am. Chem. SOC.139th Meeting, St. Louis, Missouri, 1961, p. 3N. 161. H. B. MacPhillamy, C. F. Huebner, E. Schlittler, A. F. St. AndrB, and P. R. Ulshafer, J . Am. Chem. SOC.77, 4335 (1955). 162. M. M. Robison, W. G. Pierson, R. A. Lucas, I. Hsu, and R. L. Dziemian, J . Org. Chem. 28, 768 (1963). 163. N. Kornblum, W. J. Jones, and G. J. Anderson, J. Am. Chem. Soc. 81, 4113 (1959). 164. R. A. Lucas, M. E. Kuehne, M. J. Ceglowski, R. L. Dziemian, and H. B. MacPhillamy, J . Am. Chem. SOC.81, 1928 (1959). 165. E. Wenkert and N. V. Bringi, J. Am. Chem. SOC.81, 1474 (1959). 166. A. K. Bose, B. G. Chatterjee, and R. S. Iyer, IndianJ. Phmrm. 18, 185 (1956). 167. C. Schoepf, Ann. 452, 211 (1927). 168. R. K. Hill, J . Org. Chem. 27, 29 (1962). 169. V. R. Mattox and E. C. Kendall, J. Am. Chem. SOC.78, 882 (1948). 170. W. I. Taylor, Intern. S y m p . Chem. N u t . Products, Brussels, 1962 (unpublished). 171. N. Finch and W. I. Taylor, In preparation. 172. N. Finch and W. I. Taylor, J . Am. Chem. SOC.84, 1318 (1962); J. Shave1 and H. Zinnes, J. Am. C'hern. Soc. 84, 1320 (1962); N. Finch and W. I. Taylor, J. Am. Chem. Soc. 84, 3871 (1962); N. Finch, C. W. Gemenden, I. Hsu, and W. I. Taylor, J. Am. Chem. SOC.85, 1520 (1963). 173. H. B. MacPhillamy, Personal communication. 174. a.J. Plummer, W. E. Barrett, R. A. Maxwell, D. Finocchio, R. A. Lucas, and R. Rutledge, Am. SOC. Pharmacol. Exptl. Therap. Fall Meeting, Ann Arbor, 1958, p. 27 (1958). 175. A. J. Plummer, W. E. Barrett, R. A. Maxwell, D. Finocchio, R. A. Lucas, and A. E. Earl, Arch. Intern. Phamnacodyn. 119, 245 (1959). 176. S. Garrattini, A. Mortari, A. Valsecchi, and L. ValzeKi, Nature 183, 1273 (1959). 177. T. Petrzilka, A. Frey, A. Hofmann, H. Ott, H. R. Schenk, and F. Troxler (Sandoz Ltd.), U.S. Patent No. 2,959,591 (1960) [Chem. Abstr. 55, 7444 (196l)I. 178. G. Szmuszkovicz (Upjohn Co.), U.S. Patent No. 2,819,271 (1958) [Chem. Abstr. 52, 14,717 (1958)l; G. Szmuszkovicz (Upjohn Co.), U.S. Patent No. 2,882,386 (1959) [Chem. Abstr. 5 3 , 16191 (1959)l; G. Szmuszkovicz and R. V. Heinzelman (Upjohn Co.), U.S. Patent No. 2,933,499 (1960) [Chem. Abstr. 54, 17,449 (1960)l; J . Mueller (CIBA),U S . Patent No. 2,960,506 (1960) [Chem. Abstr. 5 5 , 8452 (1961)l. 179. M. M. Robison, R. A. Lucas, H. B. MacPhillamy, W. E. Barrett, and A. J. Plummer, Ezperientia 17, 14 (1961). 180. A. Popelak and G. Lettenbauer, Ger. Patent Appl. 1,145,175 (1963). 181. M. Neeman,M. C. Caserio, J. D. Roberts, andW. S. Johnson, Tetrahedron 6 , 3 6 (1959). 182. W. E. Barrett, A. J. Plummer, R. A. Rutledge, and R. Weiss, J . Phurmacol. Exptl. Therap. 138, 78 (1962); W. E. Barrett, R. A. Rutledge, and A. J. Plummer, Pharmacologist 4, 167 (1962). 182a. W. E. Barrett, R. A. Rutledge, R. Weiss, and A. J . Plummer, Federation Proc. 23, No. 2 (1964). In press. 183. R. B. Woodward, F. E. Bader, H. Bickel, A. J. Frey, and R. W. Kierstead, J . Am. Chem. SOC. 78,2023,2657 (1956); Tetrahedron 2 , l(1958); U.S. Patent No. 2,883,384 [Chem. Abstr. 53, 20,118 (1959)l. Research Corp., Brit. Patent Nos. 799,269 (1958), 799,270 (1958), 799,271 (1958), and 799,272 (1958) [Chem. Abstr. 53, 16,186-90 (1959). 184. L. Velluz, B. Muller, R. Joly, G. NominB, A. Allais, J. Warnant, R. Bucourt, and J. Jolly, Bull. SOC.Chim. France p. 145 (1958).

13. Rauwol’a

ALKALOIDS AND CHEMISTRY OF RESERPINE

333

185. L. Velluz, G. Muller, R. Joly, G. NominB, J. Mathieu, A. Allais, J. Warnant, J. Valls, R. Bucourt, and J. Jolly, BUZZ.SOC.Chim. Prance p. 673 (1958). 186. J. 0. Jilek, B. Kakac, and M. Protiva, Collection Czech. C h m . Commun. 26, 2229 (1961). 187. L. Novak, J. 0. Jilek, B. Kakac, I. Ernest, and M. Protiva, Collection Czech. C k m . Commun. 25,2196 (1960). 188. L. Novak, J. 0. Jilek, B. Kakac, and M. Protiva, TetrahedronLettem No. 5, 10 (1959). 189. L. Blaha, J. Weichet, J. Zvacek, S. Smolik, and B. Kakac, Collection Czech. Chem. Commun. 25, 237 (1960). 190. M. Protiva, J. 0. Jilek, I. Ernest, and L. Novak, ~etrahedronLettersNo.11,12 (1959). 191. J. Weichet, K. Pelz, and L. Blaha, Collection Czech. Chem. C m u n . 26, 1529 (1961). 192. Sandoz, S. A., Belg. Patent 566,281 (1958). 193. J . 0. Jilek, I. Ernest, L. Novak, M. Rajsner, and M. Protiva, Collection Czech. Chem. Commun. 26, 687 (1961). 194. J. Weichet, B. Kakac, and L. Blaha, Collection Czech. Chem. C m m u n . 27,843 (1962); L. Blaha, B. Kakac, and J. Weichet, Collection Czech. Chem. Commun. 27,857 (1962). 195. G. Muller and A. Allais, Naturwissenschaften 47, 82 (1960). 196. F. L. Weisenborn and H. E. Applegate, J . Am. Chem. SOC.78, 2021 (1956); F. L. Weisenborn, J . Am. Chem. SOC.79, 4818 (1957). 197. K. Pelz, L. Blaha, and J. Weichet, Collection Czech. Chem. Cmnmun. 26, 1160 (1961). 198. J. 0. Jilek, J . Pomykacek, and M. Protiva, Collection Czech. C h m . Cmnmun. 26, 1145 (1961). 199. T. Petrzilka, A. Frey, A. Hofmann, H. Ott, H. R. Schenk, and F. Troxler (Sandoz), Union of S. Africa Patent Appl. 1,247,158 (1958). 200. L. Velluz, G. Muller, and A. Allais (Laboratoires francais de chimiotherapie), French Patent No. 1,238,756 (1960) [Chem. Abstr. 55, 24,807 (196l)l. 201. L. Vellue and G. Muller (Laboratoires francais de chimiothbrapie), French Patent No. 1,214,119 (1960) [Chem. Abstr. 55, 10,489 (1961)l. 202. L. Velluz, Ann. Pharm. Franc. 17, 15 (1959). 203. M. Protiva, M. Rajsner, and J. 0. Jilek, Monutsh. Chem. 91, 703 (1960). 204. L. Novak and M. Protiva, Naturwissenschaften 46, 579 (1959). 205. L. Velluz and G. Muller (Laboratoires francais de chimiothbrapie), French Patent No. 1,194,842 (1959) [Chem. Abstr. 56, 1491 (1962)l. 206. L. Velluz, G. Muller, and A. Allais (Laboratoires francais de chimiothbrapie), French Patent 1,214,122 (1959) [Chem. Abstr. 55, 9452 (1961)l. 207. L. SVelluz, G. Muller, and A. Allais, C m p t . Rend. Acad. Soi. 247, 1746 (1958). 208. A. Allais, G. Muller, and L. Velluz (Laboratoiresfrancais de chimiothbrapie), French Patent No. 1,238,738 (1960) [Chem. Abstr. 55, 23,575 (1961)l. 209. G. Hahn and H. Werner, Ann. 520, 123 (1935). 210. T. Nogradi, Monatsh. Chem. 88, 1087 (1957). 211. A. Buzas, C. Hoffmann, and G. Regnier, Bull. SOC. Chim. France p. 645 (1960). 212. W. Logemann, L. Almirante, L. Caprio, and A. Meli, Ber. 88, 1952 (1955); 89, 1043 (1956). 213. H. Plieninger and B. Kiefer, Ber. 90, 617 (1957). 214. C. Ribbens and W. T. Nauta, Rec. Traw. Chim.79, 854 (1960). 215. M. Onda and M. Kawanishi, J . Pharm. Soc. Japan 76,966 (1956) [Chem. Abstr. 51, 2824 (1957)l. 216. S. de Groot and J. Strating, Rec. Traw. Chim. 80, 121,944 (1961). 217. F. M. Miller and M. S. Weinberg, Chem. Eng. News 34,4760 (1956); Am. C h m . SOC., 130th Meeting, Atlantic City, 1956 p. 11N.

334 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228.

229. 230.

231. 232.

1. SCHLITTLER

Z. J. Vejdelek and V. Trcka, Collection Czech. Chem. C m u n . 24, 1860 (1959). G. di Paco and C. S. Tauro, P a m 0 (Pa&) Ed. Sci. 13, 64 (1958). B. V. R. Sastry and A. Laselo, J. Org. Chem. 23, 1677 (1958). F. A. Turner and J. E. Gearien, J . Org. Chem. 24, 1952 (1959). R. Ratotovis and G. Combes, BuU. SOC.C h k . France p. 576 (1959). G. Palazzo, L. Bizzi, and C. Pozzati, Ann. Chim. (Rome) 49, 853 (1959) [Chem. Abstr. 54, 24,510 (1960)l. P. B. Le Vine and J. E. Gearien, J. Org. Chem. 26,4060 (1961). M. S. Weinberg and F. M. Miller, Diwerfation Abstr. 20, 107 (1959). L. H. Schlager, Armeimittel-Forsch. 13, 226 (1963). T. Kralt, W. J. Asma, H. H. Haeck, and H. D. Moed, Rec. Trav. Chim. 80,313 (1961). T. Kralt, W. J. Asma, and H. D. Moed, Rec. Trav. Chirn. 80, 330, 431, 932 (1961); T. Kralt, H. D. Moed, V. Claassen, T. W. Hendriksen, A. Lindner, H. Selzer, F. Bruecke, G. Hertting, and G. Gogolak, Nature 188, 1108 (1960); A. Lindner, V. Classen, T. W. Hendrickson, and T. Kralt, J . Med. Phawn. Chem. 6, 97 (1963). A. Brossi, H. Lindlar, M. Walter, and 0. Schnider, Helv. Chim. Aeta 41, 119 (1958). A. Pletscher, Science 126, 507 (1957); A. Pletscher, H. Besendorf, and H. P. Baechtold, Arch. Exptl. Pathol. P h a m k o l . 232,499 (1958); G. P. Quinn, P. A. Shore, and B. B. Brodie, J. P h a m o l . Exptl. Therap. 127, 103 (1959); J. Reuse, Arch. Intern. P h a m o d y n . 126, 478 (1960); W. Stumpf, E. H. Grad, and H. Hundeshagen, Arznsimittel-Borsch. 11, 47 (1961), and others. J. R. Tretter (Chas. Pfizer Co.), U.S. Patent No. 3,053,845 (1962) [Chem. Abstr. 58, 3404 (1963)l. M. E. Smith, Am. J. Psych&. 118, 937 (1962); A. Scriabine, A. Weissman, K. F. Finger, C. S. Delahunt, G. W. Constantine, and J. A. Schneider, J. Am. Med. Assoc. 184, 276 (1963) where earlier literature is mentioned.

THE ALKALOIDS OF ASPIDOSPERMA. DIPL ORRHYNC US. KOPSIA. OCHROSIA. PLEIOCARPA. AND RELATED GENERA B . GILBERT Centro de Pesquisas de Produtos Naturais. Fuculdade Nucional de FarmCicia. Rio de Janeiro. Brazil

1. Introduction ......................................................

336

I1. The Aspidospermine Group ......................................... A. Introduction ................................................... B. Quebrachamine ................................................ C. Aspidospermine ................................................ D. NMR- and Mass Spectra of the Aspidospermine-Type Alkaloids . . . . . E. Some Minor Alkaloids of Aspidasperma quebrachoblancoand Rhazya strictu F. Demethoxyvallesine, Demethoxyaspidospermine, and Demethoxypalosine ....................................................... G. Demethylaspidospermine ........................................ H . Vallesine and Palosine ........................................... I. Aspidocarpine and Demethylaspidocarpine ......................... J. Aspidolimine ................................................... K . Pyrifolidme and Deacetylpyrifolidme .............................. L . Spegazzinine and Spegazzinidine .................................. M. Cylindrocaxpine and Cylindrocarpidine ............................ N . Limaspermine and Related Alkaloids .............................. 0. Tabersonine ................................................... P. Vinca Alkaloids of the Aspidospermine Group ......................

398 399 400 400 403 404 405 410 414 416 419

I11. The Aspidofractinine Group ......................................... A. Introduction ................................................... B . Intercorrelations and Skeletal Structure ........................... C. Aspidofractinine ................................................ D. Pyrifoline. Refractidine. and Refractalam .......................... E. Aspidofiline .................................................... F. Some Alkaloids of Aspidosperm populifolitm ...................... G. Kopsinine. Aspidofractine. Pleiocarpine. Pleiocarpinine. and Refractine H. Kopsinilam and Pleiocarpinilam .................................. I. Kopsiflorine. Kopsilongine. and Kopsamine ........................ J. Kopsine and Related Alkaloids ................................... K . K o p s k Alkaloids of Unknown Structure ...........................

420 420 421 429 429 432 433 434 439 439 441 444

IV. The Aspidoalbine Group ............................................ A. Aspidoalbine and Its N-Acetyl Analog ............................. B Aspidolimidine ................................................. 336

445 445 448

.

337 337 337 361 367 395

B . GILBERT

336 C. D. E. F. G.

Dichotamine and 1-Acetylaspidoalbidine . . . . . . . . . . . . . . . . . . . . . . . . . . . Haplocine and Haplocidine . . . . .... Cimicine and Cimicidine . . ................... Other Alkaloids of the Asp bine Group ......................... Obscurinervine and Related Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

449 450 451 451 452

453 V . The Condylocarpine Group .......................................... A. Iiitroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 B . Aspidospermatine, Aspidospermatidine, and Related Alkaloids . . . . . . . . 453 C. Condylocarpine and Stemmadenine . . . . . . . . . . . . 457 I). Tubotaiwine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462 E. 11.Methoxy.14,19.dihydroeondylocarpin e . . . . . . . . . . . . . . . . 462

VI . Alkaloids Related t o Akuammicine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . il. Introduction . . . . . . . ......................................... B. Mossambine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . Norfluorocurarine D . Compactinervine .

463 463 463 466 466

................. V I I . The Uleine Group . . . A. Introduction . . . . ................. B. Uleine and Relate C . Olivacine, Dihydroolivacine, and Guatambuine . . . . . . . . . . . . . . . . . . . . . D . Ellipticine, Dihydroellipticine, and N-Methyltetrahydroellipticine. . . . .

469 469 469 474 477

V I I I . Tetrahydro /3-Carbolineand Related Alkaloids . . . . . . . . . . . . . . . . . . . . . . . . . A . Yohimbine and Tetrahydroalstonine Derivatives . . . . . . . . . . B. Normacusine.B, Polyneuridine, and Akuammidine . . . . . . . . . . . . . . . . . . C. Quebrachidine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Harman-3-carboxylicAcid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E . Eburnamine and Related Alkaloids ............................... F. TuboAavine ............................... . . . . . . . . . . . . . . . . . . . . . G. Flavocarpine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H . Carapanaubine ...................... ........... ...... I. Isoreserpiline.+.indoxyl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J . Ochropamine and Ochropine ......................................

482 482 485 491 495 495 497 498 502 503 503

1X . Alkaloids of Unknown Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Alkaloids of Pleiocarpa Species ................ B. Alkaloids of Ochrosia Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . Alkaloids of Aspidosperma, Rhnzya, and Sternmadeniu . . . . . . . . . . . . . . .

504 504 504 505

..........

505

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

I Introduction

The great advance in our knowledge of the chemistry of this group during the past few years is indicated by the fact that a t the time of completion of the previous review (Volume VII. Chapter 10)no structure was known. whereas a t the time of the present writing. the structures of more than 100 alkaloids from these genera have been elucidated . A list

14. Aspidosperma AND RELATED ALKALOIDS

337

of the alkaloids at present known with their plant sources and physical properties is given in Table I. These bases may conveniently be divided into eight groups, each of these groups ranging over two or more of the seven genera under discussion.

11. The Aspidospermine Group

A. INTRODUCTION This group of alkaloids, whose basic carbon skeleton wtts not readily explained by the earlier biosynthetic hypotheses (1, a), appears to be relatively restricted in nature, occurring principally in the genera Aspidosperma, Kopsia, Pleiocarpa, and Stemmadenia of the family Apocynaceae. Similar alkaloids, for example, vindolinine (CVI, 3), vindoline (CIII, 4,5 ) , and vincadifformine (XCIII, 6), occur in Vinca or Catharanthus species and tabersonine (XCII, 7) is found in the genus Amsonia as well as in Stemmadenia (Section 11, 0 and P). The parent alkaloid of the group, quebrachamine (I),is also one of the most widely distributed. Although as an indole it is distinct from all the other members of the group, which are dihydroindoles, its obvious relation to the others justifies its inclusion here.

B. QUEBRACHAMINE The earlier literature on quebrachamine has been dealt with in previous volumes, and its occurrence in some new sources is recorded in Table I. Most important from the biogenetic point of view is the existence in nature of both optical enantiomers. The common levorotatory form corresponds in absolute configuration t o aspidospermine (11,vide infra), while the dextrorotatory form found in Stemmadenia donnell-smithii (8) is related to (+)-pyrifolidine (XLVI, 9, 10) and to (-)-tabersonine (XCII, 7). The Vinca minor alkaloids vincadine (11-A) and vincaminoreine (11-B)are the 3-carbomethoxy- and 3-carbomethoxy-Na-methyl derivatives, respectively, of ( + )-quebrachamine (10a). The same plant also contains ( k )-indolic-N-methylquebrachamine (lob). The determination of the structure of quebrachamine (I)followed on that of aspidospermine (11),and was indeed suggested at the time that this latter structure was published (11, 12, 13, 14). In work prior to this, Witkop and his co-workers were able to show that the two alkaloids were related, since both gave on zinc dust distillation a mixture from which

TABLE I

w w ca

PHYSICAL CONSTANTSAND PLANT SOURCES PART1. Quebrachamine and its derivatives

Compound'

Formula

mP ("C)

Quebrachamine, 1, a 38, 59; b 23, 63; c 34; d 5 2 ; e 51, 62; f Volume VII, 60, 61; E 38b

ClgHzeNz

146-147

Dextrorotatory, g 8 Racemic Salts (see Volume 11)

147-149 114-1 16

[uID2

-108 (c) -91 (e) - 110 (a) - 100 (d) +111 (c)

Reference 23, 34 55

8 6, 8

PART 2. Aspidospermidine and its derivatives Formula

Aspidospermidine(282A), XXXI, C ~ ~ H Z a ~ 28,51a; N Z , e,51 1,2-Dehydroaspidospermidine,IX, C ~ ~ H Z ~ e 51b NZ, Demethoxyvallesine, XXXII, CzoHz~Nz0,h 37 Demethoxyaspidospermine, XXXIII, C21HZsN20, h 40a; i 48; mm 48 Demethoxypalosine, XXXIV, Cz~H30Nz0,j 40; h 408 A',-methylaspidospermidine, XXXV, CzoHzsNz, a 28, 51a Aapidosine, XXXVI, C1gHzeNzO

w

0

16

17

Ne.

H

H

H

H H

H H

CHO

H H H

H H OH

mP (" C)

[elD

120-121 amorph.

+ 243(e)

+17(e)

Ac

amorph.

- 15 (c)

EtCO Me H

115-120

- 17 (c)

265.5-257.5

- 16 (e)

Reference

28, 51b 28, 51a, b 37 40a 40 28 38, 25, 26

Demethylaspidospermine, XXXVII, C21H2sN202, h 40e; i 48 Deacetylaspidospermine, VI, CzoHzsNzO, a 28, 51a; b 38a

H

OH

Ac

amorph.

H

OMe

H

109-111

11-Hydroxy-10-0xoaspidospermine(A),XII,

CzzHzsNz04. CsHs XV, C24H30Nz05 11-Acetoxy-10-oxoaspidospermine(A), 11-Hydroxy-10-oxoaspidosperrnine(B), XVI, CzzHzgN204 10,11-Dioxoaspidospermine, XIII, CzzHzsNz04. CeHa oxime, C22H27N304. CaHs 11-Chloro-10-oxoaspidospermine, XIV, CzzH27ClNz03 Vallesine, XXXVIII, Cz1H2sN202, k 41 ; 1 39 Aspidospermine, 11, CzzH30Nz02, a 38; b 23; m Volume 11; k 41,41a; 139 10-Oxoaspidospermine, XI, CzzHzsNz03. CeH6 8-Oxoaspidospermine, XX, CzzHzsNzOa Apoaspidospermine, LI, C15HzzNz02 Palosine, XXXIX, C23H32N202, b 23 N,-Isobutyryldeacetylaspidospermine, XL, Cz4H34Nz02 N,-Benzoyldeacetylaspidospermine, Cz7H32N202 N,-Carbethoxydeacetylaspidospermine,C23H3zN203 N,-Methyldeacetylaspidospermine,XLI, CzlH3oNzO N-Ethyldeacetylaspidospermine, XVII, CzzH32NzO N,-Benzyldeacetylaspidospermine,C27H34N20 0,N-Diacetylaspidosine, C23H3oN203 Demethylaspidocarpine, L, C21HzsNz03, d 42 Aspidocarpine, XLIV, C22H3oN203, n 30; j 31, 40; ii 48; p 113 i; d 42 Aspidolimine, LIII, C23H32N203, j 31, 40; q 47; p 113 i Deacetylpyrifolidine, XLVII, C21H30N202, r 37 Pyrifolidine, XLVI, C23H32N203, Dextrorotatory, r 10

H

H

OMe OMe

CHO Ac

40a, 25

228-230

25, 26, 28, 38, 38a 24

211 247 249 261 294-296 154-1 56 208-209

24 24 24 24 24 41, 39 23, 26, 38

115 174-176 224-225

H H H H H H H H OH OMe

OMe OMe OMe OMe OMe OMe OMe OAc OH OH

EtCO iPrC0 PhCO CO2Et CH3 Et PhCHz Ac Ac Ac

149-152 162-164 187-190 125.5-128 amorph. 111-112 119.5-120.5 158-1 58.5 156-158 168.5-169.5

OMe OMe

OH OMe

EtCO

OMe

OMe

Ac

150-1 51 153-154 148-150.5 152-154 147.5-150

H

+ 119 (c) +3

(0)

-92 (c) -93 (c) -99 ( 0 )

- 132 (c) - 128 (c) - 86 (0)

+ 125 (c) + 140 (c) + 133 (c) - 5 (c) + 7 (0) - 94 ( 0 ) 90 (c)

+

24 24 30 30 23,43 43 25,39 25 25, 28, 51s 25 25 25 30,42 30, 31 31 30 9 30,28 10

TABLE I-Continued W

Formula [elD

Compound

N,

16

17

N,-Ethyldeacetylpyrifolidine, LVII, Cz3H3sNz02

OMe

OMe

Et

0-Acetylaspidocarpine, XLV, Cz4H32N204 0-Acetylaspidolimine, LIV, C25H34Nz04 Picrate, B .C~H3N307 0-Propionylaspidolimine,L v , C26H36N204 Picrate, B. CsH3N307 0,O-Diacetyldemethylaspidocarpine, XLIX, C25H32N205

OMe OMe

OAc OAc

Ac EtCO

OMe

EtCOO

EtCO

OAc

OAc

Ac

-

-~

Reference

amorph. amorph.

-21

(c)

165-167

-8

(c)

+ 20 ( c )

semi-cryst.

193-195.5 179-181 144-145

+ 10 (c)

254-257 104.5-106

+ 63 ( c ) + 176 (c)

36 36 30,31 31 31 31 31 30

PART3. Aspidospermine-type alkaloids oxygenated a t C-3and their derivatives Formula Compound

Deacetylspegazzinine, LXII, C I ~ H Z ~ N Z O ~ Spegazzinine, LXIV, C21H28N203, c 34 0-Methyldeacetylspegazzinine, LXIII, CaoHzsNzOz 0-Methylspegazzinine, LXV, CzzH30NzOa 0-Methyl-0-acetylspegazzinine, LXVI, C24H32N204 0-Methyl-0-benzoylspegazzinine, LXVII, CzgH34N204 0,O-Diacetylspegazzinine,LXVIII, Cz5H32N205 Spegazzinidine, LXIX, C Z ~ H Z ~ N cZ 35, O ~36 , 0,O-Dimethylspegazzinidine, LXX, C23H32N204 0,O-Dimethyl-0-tosylspegazzinidine, LXXIV, C30H38Nz06S O,O-Dimethyl-3-dehydrospegazzinidine, LXXI, C23H3oN204

16

17

H

OH OH

H H H H H H OH OMe OM0 OMe

N,

3

H

OH OH OH OH OAc PhCOO OAc OH OH OTs

OMe

Ac H Ac Ac Ac Ac Ac Ac Ac

OMe

Ac

OMe OMe

OMe OMe OAc OH OMe

=O

amorph. amorph. amorph.

- 64 ( c ) - 136

-23 ( c )

amorph.

+ 97 + 123 ( c )

237-238 167-169 158-160

+ 186

- 156 (c) +42 (c)

185-187

- 53

192-194

or

155-160

34 34 34 34 34 34 34 35,36 35,36 36 35,36

b P

0

PART 4. Aspidospermine-type aIkaloids oxygenated at C-21 and their derivatives Formula mp ("C)

[c(ID

Reference

- 122 (c)

10 29, 10 10 10

COzH

11,s-118.5 168-169 146-147 236-238 and 245-248 265 (dec.)

DHC

CO2H

203-217

DHC Ac EtCO

H H

OMe OH OH OMe OMe OMe

H DHC Et

CHO CHaOH CH2OH CHzOH CHaOH CHzOH

H

OMe

CnHii

CHzOH

48-49

29, 10

H H

OMe OAc

Ac EtCO

CHzOAc CHzOAc

OMe

OH

Ac

CHzOH

148-150 amorph. 122-123 220

+ 131 (c)

52 33 33 120

OMe

OH

EtCO

CHzOH

174- 175

+118(c)

33,120

H

CHeOH

175-1 77

Compound

~

16

17

Cylindrocarpidine, LXXVI, C Z ~ H ~ O Ns 49, ~ O10 ~, Cylindrocarpine, LXXV, C30H34Nz04, s 49, 10 Dihydrocylindrocarpine, LXXVII, C30H36N~04 Cylindrocarpinic acid dihydrochloride, LXXX, CzoHz8ClzNz03. E t O H

H H H H

Cylindrocarpic acid hydrochloride, LXXVIII, CzsH33CINzO4 Dihydrocylindrocarpic acid hydrochloride, LXXIX, CzsH35ClNz04 Dihydrocylindrocarpal, L X X X I I I , CzgH34Nz03 Limapodine, L X X X I X , Cz~HzsNnOs,j 120 Limaspermine, LXXXVII, CzzH30N203, j 33 Decinnamoylcylindrocarpol, LXXXIV, CzoHz8NzOz Dihydrocylindrocarpol, LXXXII, CznH36Nz03 N,-Ethyldecinnamoylcylindroca.rpo1,LXXXVI, CzzH3zNzOz N,-y-phenylpropyldecinnamoylcylindrocarpol, LXXXI, CznH3sNzOz 21-Acetoxyaspidospermine, LXXXV, C z 4 H 3 ~ N ~ 0 4 0,O-Diacetyllimaspermine, LXXXVIII, Cz~H3.gN205 Picrate, B. C~H3N307 21-Hydroxyaspidocarpine( 16-methoxylimapodine), XC, C ~ ~ H ~ O Nj 120 ZO~, 21-Hydroxyaspidolimine( 16-methoxylimaspermine), XCI, C Z ~ H ~ Z N jZ120 O~,

N,

21

OMe OMe OMe OMe

Ac Cin DHC H

COzMe COzMe

H

OMe

Cin

H

OMe

H H H H

-

COzMe

COzH

amorph. 177-178 175-175.5 145-147 amorph. amorph.

- 181 (c) - 126 (c) - 3 (w)

10 10

+ 110 (c) f 108 (c)

+2 -98 (c)

10 120 33 52 10 52

.

N-Deacyl-0-methylaspidoalbinol, CXCIII, C2&€82N2O6

52

TABLE I-Continued PART 5. Aspidospermine-type alkaloids bearing a carbon substituent at C-3 and their derivatives -

~~~

~~

~~~~

Formula Compound Tabersonine, XCII, C Z I H Z ~ N Z Ot Z65; , g 66a; u 66a; v 66a; w 66a Hydrochloride, B .HCI Vincadifformine, ( = 6,7-dihydrotabersonine),XCIII, CziHzsNz0z Minovine (N,-methylvincadifformine), CVIII, CZZHZENZOZ Minovincinine, CVII, CziHzaNz03 Minovincine, CIX, CzlHzsNz03 16-Methoxyminovincine, CX, C2zHzeNz04 2,3-DihydrotabersonoI, XCVII, CzoHzsNzO N,,O-diacetate, C, Cz4H30Nz03

mP 2,3

6,7

3

20

(" C)

A

A

COzMe

H

amorph.

A

COzMe

H

A

COzMe

H

A A A

COzMe COzMe COzMe CHzOH CHzOAc

OH =O =O

A A

H H

196 (dec.) 124-125 amorph. 79-81 amorph. amorph. amorph. 186 196

[.ID

Reference 7, 64, 65

- 310 ( m ) rac. - 540 (e) 0 (e) -418 (e) - 504 (e) -414 (0) 82 ( e )

+

64, 65 6, 74 7, 32 74a 32, 76 32 32 32 7, 64 64

a,

8 W

M

3

PART6. Miscellaneous alkaloids with t h e aspidospermine skeleton a n d their derivatives mP -

__

Compound

Vindoline Deacetylvindoline Dih ydrovindoline

16-Methoxy-N,-methyl-4-oxoaspidospermidine Vindolinine CVI dihydrochloride, B .2HC1

Other derivatives (see Ref. 3)

Formula CIII CXIII CXI

CV CVI

CzsH3zNzOs Cz3Hm~Nz05 CzsHdzOs

CZIHZENZOZ CziHz4NzOz

Reference

(" C)

154-1 55 156-157 121-1 24 and 164-166 130-1 32 250-252 210-212 214-2 18

+42 (c) i-52 ( c )

+ 12 (c) - 72 (w) - 8 (w)

- 18 (w)

72 72 4

4 3, 72, 71 71 72 3

PART 7. Aspidofractinine-type alkaloids unsubstituted in position 3 Formula Compound

Aspidofractinine, CXVII, C I ~ H ~ x~ 102 N, Aspidofiline, CXXXVI, CzlHz6NzO~,r79 0-Methyldeacetylaspidofiline, CXXXIX, CzoHzaNzO, y 48

[.ID

mP 17

N,

6

(" C)

H OH

H Ac

OMe

H

H H H

amorph. 190-1 9 1 129-131

- 174 (c) -8 (c)

or N-Formyl-0-methyldeacetylaspidofiline, CXLI-B, CzlHz6NzOz, Y 48 0-Methylaspidofiline, CXXXVIII, CzzHzaNzOz 0-Acetylaspidofiline, CXXXVII, Cz3HzsNz03 16,17-Dimethoxyaspidofractinine,CXLI-A, CzlHzaNzOz, y 48 N-Formyl-16,17-dimethoxyaspidofractinine, CXLI-C, CzzHzaNz03, Y 48 Deformylrefractidine, CXXII, CeoHzaNzO Dihydrochloride, B .2HC1 Methiodide, B.MeI Refractidine, CXX, C Z ~ H Z ~ N ZxO81 Z, N-Methyldeformylrefractidine, CXXIII, CziHzaNzO 6-Demethyldeformylrefractidine, CXXVII, C19H~4N20 N-Aretyl-6-demethyldeformylrefractidine, CXXX, C z l H ~ 6 N z 0 ~ N,-Methyl-6-demethyldeformylrefractidine, CXXXI, CzoHz6NzO N,O-~iacetyl-6-demethyldeformylrefractidine, CXXIX,

Cz3HzsNz03 6-Dehydrodemethyldeformylrefractidine,CXXXV, C19HZzN20 Deacetylpyrifoline, CXXI, CzlHzsNzOz Perchlorate, B .HC104 Pyrifoline, CXIX, C Z ~ H ~ ~ Nr Z 81,O49 ~, 6-Demethyldeacetylpyrifoline, CXXV, CzoHz6NzOz 6-Acetyldemethylpyrifoline, CXXVIII, C Z ~ H ~ ~ N ~ O ~ 6-Dehydrodemethylpyrifoline, CXXXIII, CzoH24NzOz

OMe

CHO

H

amorph.

OMe OAc 16,17 (0Me)z 16,17(0Me)z H

Ac Ac

H H

H

H

amorph. 179-181 140- 142

CHO

H

amorph.

H

OMe

CHO Me

OMe OMe

H Me

OH OH OH

Ac

OAc

amorph. 189-193 258.5-259.5 158-160 104-106 163-164 2 18-2 19 160-161 193-194

H H

OMe

Ac

OMe

H

OH

Ac

OAc

H

=O

H

H H H H H

H OMe

OMe OMe OMe OMe

Ac

=O

132-136 amorph. 2 70-2 7 5 (dec.) 142-144 202-203 197-199 158-160

+ 3 (c)

Reference

102 80, 79 80 48

+ 53 (c)

- 141 (c)

80 80 48 48

- 53 (c) -31 (e)

- 140 (c) - 86 (c)

-24 (c) 4-39 (c) - 44 (c) +44 (c)

+ 10 (c) - 14 (c)

+ 102 (c) - 20 (c) + 170 (c) + 24 (c)

81 37 37 81 37 81 81 81,102 81 81,37 81 37 49, 81 81 81 81

TABLE I-Continued ~

~~

W rp

PART 8. Aspidofractinine-type alkaloids substituted in position 3

rp

Compound

[aID

Reference

- 62 t o - 77 (c)

91, 82,87

- 142 (c)

-116 ( m ) - 56 (c) -23 (c)

82, 91 96 91, 92 91 94 82 49, 82

+7 (c) + 3 (c) 34 (0) 76 (c) -67 (c) -18 (c)

37 82 82 82 87 87, 89

-48 (c)

87, 88, 85 89

~

Kopsinine, CXLII-A, CzlHz6NzOz, z 87, && 91; b b 96; o 48

H

H

COzM0

105

Aspidofractine, CXLIII-A, CzzHzeNz03, x 82 N-Acetylkopsinine, CXLIV-A, Cz3HzsNz03 Pleiocarpine, CXLV-A, Cz3HzsNz04, &a 91; b b 93, 96; cc 94 Pleiocarpinine, CXLVI-A, CzzHzsNzOz, tm 91; b b 96; cc 94

H

CHO Ac COzMe Me

COzMe COzMe COzMe COzMe

197 184-1 85 141-142 135-136

H CHO

COzMe COzMe

amorph. 157.5-159 and 191-192 amorph. 109-110 155-156 190-1 9 1 144-145 210-211

Doformylrefractine, CXLIX-A, CzzHzsNz03 Refractine, CL-A, Cz3HzsNz04, x 49; y 48

Deformylieoaspidofractine, CXLII-C, Cz1Hz6NzOz Isoaspidofractine, CXLIII-C, CzzHz6Nz03 Deformylisorofractine, CXLIX-C, CzzHzsNz03 Isorefractine, CL-C, Cz3HzsNz04 KopsiAorine, CLVI, Cz3HzsNz05, z 87 Kopsilongine, CLVII, Cz4H3oNz06, z 87

H

H H OM0

OM0

H

H OMe OMe

H OMe

CO~MO COzMe H COzM0 CHO COzMe COzMe OH,COzMe COzMe OH,COzMe H CHO

- 145 (c) - 124 (c)

W M

+

+

16.17

Kopsamine, CLVIII, Cz4HzsNz07, z 87 Kopsaminic acid, CLX, Cz3HzeNz07.+HzO

OzCHz

OzCHz

COzMe COzH

OH, C O ~ M O OH,COzMe

205-206 131

OH,COzMe OH,COzMe OH,COzMe

148-149 148-149 119-120

17

KopsiAoreine CLXI, CziHz6Nz03 Nitroso-, CLXIV, CziHz5N304 Kopsilongeine, CLXII, CzzHzsNz04

H H OMe

H NO

H

W

89 89 89

E

16,17

H NO

OH, COzMe OH, COzMe

145-146 185-186

89 89

H H

H H

COzH COzH

209-21 1 280-281 273-274

89 37 37

OM0

H

COzH

H

H H

OH, COzH OH, COzH

242-250 198 199

H

OH, COzH

220-221

H

H

COzMe

254-254.5

H H

COzMe

COzMe

CH3

COZMe

190 249-250

H H H H

H

COzMe COzMe

OzCHz OzCHz 17

Kopsinic acid, CXLII-B, CzoHz4NzOz Deformylisoaspidofractic acid, CXLII-B, CzoHz4NzOz Dihydrochloride, B. 2HC1 Deformylisorefractic acid, CXLIX-B, C Z I H Z ~ N Z O ~ Dihydrochloride, B .2HC1 Kopsifloric acid, CLXVII, CzoH34Nz03. HzO Kopsilongic acid, CLXVIII, CziHz6N~O4.3HzO

OMe

+20(m)

37 89 89

16,17

Kopsamic acid, C L X I X , C Z I H Z ~ N Z 3HzO O~.

OzCHz

89

17

Kopsinilam ( l o - o s o ) , CXLII-D, CziHz4NzO3, aa 96; bb 96; cc 96 Pleiocarpine lactam A( 10-oxo),CXLV-D, Cz3Hz+$"05 Pleiocarpinilam (lo-oxo), CXLVI-D, CzzHzsNzO3, aa 96; bb 96; cc 96 Kopsinine-8-lactait1, CXLII-E, CziHz4NzO3 Pleiocarpine lactam B (8-oxo),CXLV-E, Cz3HzfiNzOj N-Methylkopsinal, CLV, C Z ~ H Z ~ N ~ O Kopsinyl alcohol, CXLII-F, CzoHzfiNzO N-Methylkopsinyl alcohol, CXLVI-F, CZiHzsNzO 0-AcetaLe, CXLV1-K, Cz3HmNzOz Deformylrefractinol, C X L I X - F , CziHzsNzOz Deformylisorefractinol, CXLIX-T, CzlHzsNzOz .MeOH 0-Acetate, C~3H3oNzO3 Kopsinyl tosylate, C X L I I - H , C Z ~ H ~ ~ N Z O ~ S Deformylrefractinol tosylate, C X L I X - H , C Z E H ~ ~ N Z O ~ S Deformylisorefractinol tosylate, CXLIX-W, CzsH34Nz04S Kopsinyl iodide, C X L I I - J , CzoHzsINz

H H OMe OMe OMe

H OMe OMe

H

COzMe Me

H Me Me

H H H

H H H H

CHO CHzOH CHzOH CHzOAc CHzOH CHzOH CHzOAc

CHzOTs CHzOTs CHzOTs CHzI

205-207 201-202 131-132 162 136-137 124-125 153-154 101-103 amorph. amorph. 147-148 amorph. I62 (dec.)

- 13 ( c )

96

-53 (c)

95 96

- 82 (c)

-47 (c) -3 (c) + 6 (c) -57 (c) - 50 ( c ) - 17 (c)

96 95 95 82, 89, 91, 92 91, 92 91

+Z

U

*e

M

ti b

c

'F1

82 37 37 82 82 37 95

c3 @ 01

TABLE I-Continued

w

Formula Compound

mP

~

17

N,

H H H

CHzOCHz CO-Q-CHz H =CHz

H

Me

OM0

=CHz =CHz =C(CH3) CH3 CH3 CH3 =O =O =O

N,O-Methylenekopsinyl ether, CLIII, CzlHzeNzO Kopsinyl N,-carboxylate, CLIV, CzlH24NzOz Kopsinylene, CXLII-M, CzoHz4Nz Picrate, B. CeH3N307 N-Methylkopsinylene, CXLVI-M, CzlHzeNz Deformylrefract-3-ene, CXLIX-M, CzlHzeNzO N-Methylisokopsinylene, CXLVI-N, CzlHzeNz Kopsinane, CXLII-0, CzoHzeNz N-Methylkopsinane, CXLVI-0, CzlHzsNz Deformylrefractane, CXLIX-0, CzlHzsNzO Noraspidofractone, CXLIII-P, CzoHzzNzOz Deformylnorrefractone, CXLIX-P, CzoHz4NzOz

H H H OMe H OMe

H Me H Me H CHO H

Norrefractone, CL-P, C Z ~ H ~ ~ N Z O ~

OMe

CHO

[aID

Reference

(" C)

3

101-103 200-201 amorph. 183-187 68-70 amorph. 124-126 118 112-1 14 amorph. 140-142 202-209 (dec.) 184-189

-85 (c)

- 102 (c)

- 10 (0) - 240 (c)

95 95 82, 37 95 95 82 95 95 95 82 37 82

-51 (c)

37,102

[alD

Reference

PART 9. Kopsine and derivatives Formula Compound

Kopsine, CLXX, C2zHz4Nz04, dd 100, 103, 109

mP N,

3

3'

COzMe

OH

=O

(" C)

218

+ 16 (e)4 - 18 (c)

Dihydrokopsine-A (NaBH4), CLXXII, CzzHzeNz04

COzMe

OH

OH

256-257 (dec.)

104, 55 109, 112 100, 109, 112

W

8 W

M

E

0-Acetate, CzlHzsNzOs

COzMe

OH

OAc

COzMe

OH

OH

COzMe

OH

OAc

CLXXV, CzzHzzNz05 Kopsine Lactam A (10-OXO),

COzMe

OH

=O

CLXXVII, CzzHz4NzOs Dihydrokopsine Lactam R ( 10-OXO),

COzMe

OH

OH

170-1 7 1 (dec.) 223-225 (dec.) 214-215 (dec.) 234-235 (dec.) 222-224

Decarbomethoxykopsine, CLXXI, C~OHZZNZOZ, dd 113

H

OH

=O

240-242

Ethyl carbonate, Cz3HzeN204 Dihydrodecarbomethoxykopsine (NaBH4),CLXXIII, CzoHz4NzOz Dihydrodecarbomethoxykopsine (Hz/Pt),CzoHz4NzOz

H H

OCOzEt OH

=O OH

H

OH

OH

=O

217 266-268 and 276 178-180 and 190-1 9 1 190-192 160.5-162.5 238-240 248-250 and 260-261 278-280

H H H

252-254 244 (dec.) 163-1 64.5 162-163 155 154-155 174-1 75

Dihydrokopsine-B (Hz/Pt),CLXXIV, CzzHzsNzO4 0-Acetate, Cz4HzsNz05

Isokopsine, CLXXXVII, CzzHz4Nz04 Dihydroisokopsine, CLXXXVII-B, CzzHz6Nz04 Decarbomethoxyisokopsine, CLXXXVII-A, CzoHzzNzOz, dd 113 Dihydrodecarbomethoxyisokopsine,CLXXXVII-C, CzoHz4NzOz

Decarboinethoxykopsine lactam (10-oxo), CLXXVI, CzaHzoNz03 Kopsine methine, CLXXX, Cz3HzsNz04 Dihydro, Cz3HzsNz04 Tetrahydro, Cz3H30N204 Carbethoxytetrrthydro, C Z ~ H ~ Z N Z O ~ Kopsane, CLXXXIV, CzoH24Nz N-Acetyl, CLXXXV, C2zHzsNzO 10-Lactam, CLXXXVI, CzoHz2NzO

H

H Ac H

OH

H H H

100,112 100,101 100,112

+30 (e)

100, 101, 109,112 100, 101, 112 100, 109, 104 100 100,109 111

-82(~)

111 111 111 111

100,101 101,112 101,112 101,112 101,112 101,112 101 101

TABLE I-Continued W

rp

m

PART10. Compounds related t o aspidoalbine Formula Compound

mP 15, 16, 17

Fendleridine, CCI-L, C ~ ~ H ~ ~ G N 113f ZO, 1-Acetylaspidoalbidine, CCI-K, CzlHzsNzOz, 1~1130 Deacylhaplocine, CCI-F, ClgH24N202 Haplocidine, CCI-C, C21H26N203, F 113b, k 1138 Haplocine, CCI-D, CzzHzsN203, F 113b 0-Methylhaplocidine, CCI-B, CzzHzsNz03 0-Methylhaplocine, CCI-G, CzaH30Nz03 0-Acetylhaplocine, CCI-E, C24H3oN204 Dichotamine, CCI-A, C21Hz4Nz04, k 41, 113e Aspidolimidine, CCI, j 40, d 113h Fendlerine, CCI-0, G 113f N,-Acetyldepropionylaspidoalbine,CLXXXIX, CzaH30Nz05, d 42, 52; H 48 Aspidoalbine, CLXXXVIII, Cz4H32N205, d 42, 52 ; H 48 0-Methyldepropionylaspidoalbine, CXCII, CzzHaoN204 N,-Acetyl-0-methyldepropionylaspidoalbine, CXCI, C24H3zNz05 10-0x0, C24H30NzOs 0-Methylaspidoalbine, CXC, C25H34Nz05 CXCI-21-lactone, C24H3oNzOe CXC-21-lactone, Cz5H3zNzO6, CXCIX-A, ee 113g

N,

21

H

185-1 86 173-174 250 (dec.) 183-184 186-187 237-239 240-241 194-195 263-265 196- 199 185-186 175-179 or 194-1 95 174-177

Ac

H

17-OH 17-OH 17-OH 17-OMe 17-OMe 17-OAc 17-OMe 16-OMe,17-OH 16-OMe,17-OH 15, 16-(OMe)z,17-OH

Ac EtCO Ac EtCO EtCO CHO Ac EtCO Ac

15, 16-(OMe)2, 17-OH

EtCO

15, 16, 17-(OMe)3

15, 16, 17-(OMe)3 15, 16, 17-(OMe)3 15, 16, 17-(OMe)3

H

148-149

Ac

177-178

EtCO Ac EtCO

CH2

co

CO

[O L ] ~

Reference

(" C)

2 14-2 18 128-131 225-226 180-182

+ 239 (c) + 174 (c)

113f 1130 113b 113b 113b 113e 113c 113b 113e 40, 113h 113f 42, 48

+ 164 (c)

42

+ 46

(0)

- 116 (c)

-36 (m)

52 52

+22 (m) + 9 (c) - 114 (m)

52 42, 52 52 113g

ej W M

2

PART11. Compounds with t h e aspidospermatidine or strychane skeleton Formula

Compound 16

[.In

Reference

20

mP ("C)

=CHMe

H

184-1 86

28. 51a

=CHMe

€1

149-152

118

=CHMe

H

amorph.

28. 51a

=CHMe

H

amorph.

28, 51a

=CHMe

H

amorph.

28. 51a

=CHMe

H

157-159

Et

H

amorph.

Et

H

amorph.

+213 (c)

48

=CHMe

H

167-168

+900 (c) 870 (e) +584 (c)

116, 117, 75 120, 118a

~

N,

14 ~~

Aspidospermatidine, CCIV, ClsHzzNz, H H a 28, 51a 1,2-DihydrodecarbomethoxycondyloH H carpine, CCXXII, C18HzzNz N-Acetylaspidospermatidine, CCVIII, Ac H C Z O H Z ~ N ZaO28, , 51a H N-Methylaspidospermatidine, CCVI, Me C ~ ~ H Z ~a N28,Z 51a , Deacetylaspidospermatine, CCVII, H(12-OMe) H C ~ ~ H Z ~ NaZ28, O ,51a H Aspidospermatine, CCIX, CzlHz,jNzOz, Ac( 12-OMe) a 28, 51a Dihydroaspidospermatine, CCX, Ac( 12-OMe) H C21HZsNzOz, a 28, 51a 1l-Methoxy-14,19-dihydrocondylocarpine, H( 11-OMe) A , COzMe CCXXIV-A, CziHz1jNzO3, y 48 H Condylocarpine, CCXV, CzoHzzNzOz, ff 116 A, COzMe Tubotaiwine, CCXXIV, CzoHz4NzOz, bb 119; j 120 Picrate Tetrahydrocondylocarpine, CCXX, CzoHz~jNzOz Decarbomethoxyakuammicine, ClsHzoNz 19,2O-Dihydrodecarbomethoxyakuammicine, C C X X X I X , ClsHzzNz, bb 119 Tubifoline picrate5

H

A,COzMe

a-Et

H

amorph.

H

COzMe

Et

H

171-1 72 145-147

A A

H H

H H

=CHMe a-Et

80-84 amorph. 194-196

-73 (e)

28, 59, 51a 28, 51a

+

119,120 118

-361 (ea)

19 67, 118a 119

TABLE I-Continued Formula Compound

Na

16

14

20

mP (" C)

[aID

Reference

- 298 (c)

122

0 01

o

-

Decarbomethoxymossambine, CCXXXI, A CisHzoNzO 1,2-Dihydrodecarbornethoxyakuammicine, H CCIL CiBHzzNz Tetrahydrodecarbomethoxyakuammicine = H Tubifolidine, CCXXXIX-A, C I ~ H Z ~ N bbZ 119 , Akuammicine, CCXXV, CzoHzzNzOz H

H

OH

=CHMe

190

H

H

=CHMe

187-189

19

H

H

8-Et

176-177

119, 19

A , COzMe

H

=CHMe

181-182.5

H

A , COzMe

H

P-Et

173-175

H

A , COzMe

OH

=CHMe

240

- 470 (c) - 498 (c)

H H

A , COzMe A , COzMe

OAc OH

=CHMe Et

111

-523 (c)

H

A , COzMe

H

CHOHMe

H

A , COzMe

H

135 (hydr.) -515 244 (anhydr.) 220 211-214 607 (dec.)

H

A , COzMe

H

19-Dehydr0, CCXLI-G, CzoHzzNz04

H

A , COzMe

H

19,20-Diacetate, CCXLI-A, Cz4HzeNzOa

H

A , COzMe

H

19,20-Dihydroakuammicine,CCXXXVII, CzoHz4NzOz Mossambine, CCXXVI, CzoHzzNz03, ff 121, 116 0-Acetate, CCXXVII, CzzH24Nz04 19,20-Dihydromossambine,CCXXVIII, CzoHz4Nz03 Echitamidine, CCXLII, CzoHz4Nz03 0-Acetate, CzzHzaNzOa Lochneridine, CCXLIII, CzoH~4Nz03

-745 (0) - 727 (m) 720 (m) - 673 (m)

+

+

OH

19, 77, 69,131 125, 69, 19,1378 116, 122

130, 126 126 129, 127, 75

110-120 235 (dec.)

- 640 (PI

125a

224-226 (dec.)

- 607 (c)

125a

208

- 623 (c)

125a

C L M e

M e:{

1(;:

AcMe

p W

122 121

(Et Compactinervine, CCXLI, CzoHz4Nz04, gg 127

Pj 0 M

T:

Dihydro, CCXLI-B, CzoHzaNz04

H

COzMe

H

19-Epi, CCXLI-H, CzoHz4Nz04

H

A, COzMe

H

H

A, CHO

H

C:OHMe =CHMe

H

a-COzMe

H

=CHMe

265-270 (dec.)

-44 (c)

12th

222-224 (dec.)

-680 (p)

125a

184-186 270 (dec.) 140

-1230(~)

CzOHMe

Norfluorocurarine, CCXL, C19HzoNz0, ff 116 Methochloride, B HCl 2,16-Dihydroakuammicine,CCXXXVIII, CzoHz4NzOz Tetrilhydroakuammicine, CCXXXV, CzoHz6NzOz 2,16-Dihydromossambine, CCXXIX, CzoHz4Nz03 Tetrahydromossambine, CCXXX, CzoHzaNz03 2,16-Dihydrocompactinervine, CCXLI-B, CZOH26"204

.

H

H

m-COzMe

,5-Et

H

COzMe

OH

=CHMe

H

COzMe

OH

Et

H

COzMe

H

cr

- 18

135-137 197-198

+24(c)

116 133-1 37 68, 6% 137a, 75 137a 122 122

265-270 (dec.)

-44

(0)

~

+ b

5 8 2 01

48

OHMe

m

b

r-

PART12. Indoles related to stemmadenine

M U

Formula [a]=

Reference

20

mP (" C)

+329(p)

116

Compound

Stemmadenine, CCXIII, Cz1Hz6Nz03, g 8 ; u 66a; v 66a Degradation product (KBH4) ex akuammicine, CigHzeNz Degradation product (KBH4) ex mossambine, CCXXXII, CisHzsNzO Degradation product (KBH4) ex compactinervine

N,

16

14

H

CHzOH, COzMe

=CHMe

H

189-191

H

H

H

=CHMe

H

H

OH

=CHMe

125-150 and 160-1 62 2 15-2 16

H

H

m

/OH \CHOHMe

230-232

19

-62 (c)

122 48

w m c

W

TABLE I-Continued

01 E3

PART 13. Compounds related t o uleine and the pyridocarhazole bases

Compound

Formula

[aID

Reference

+ 18 (c)

138, 147

~

Uleine, CCXLV, hh 138; ii 140; j j 49; kk 141, 147; 11 143h; i 48; mm 48; nn 48; o 48; K 37 Dihydrouleine, CCXLVI Olivacine, CCLVIII, ii 140; kk 141, 147; 00 158; pp 145; nn 48; i 48; qq 148; J 48 1,2-Dihydroolivacine, CCLXXIV, hh 139 1,2,3,4-Tetrahydroolivacine N-Acetyltetrahydroolivacine Guatambuine, CCLX, dextrorotatory, hh 139; kk 141, 147; pp 145, 146;'ll 143b; J 48 Levorotatory, kk 147 Racemic, kk 147 Methiodide Ellipticine, CCLXXXIV, rr 156; ss 156; tt 161; uu 161; vv 161; 00 157, 158 Methonitrate, 00 158 1,2,-Dihydroellipticine,CCXCIII, hh 139, 00 158 Methonitrate, 00 158 1,2,3,4-Tetrahydroellipticine, CCXCIV N-Acetyltetrahydroelipticine N-Methyltetrahydroellipticine,CCLXXXV, h h 139; 00 158; C 163; I1 143b

ClsHZzNz, methanolate C18Hz4N2 Ci7Hi4N2

76-118 118-120 (cap.) 75-1 15 318-324 (dec.)

232-234 (dec.) 299-301 (dec.) 311-315 (dec.)

138 140, 141, 147 150 150 150 139, 141, 147, 146 147 147, 150 144, 147 156

293-304 (dec.) 296-300 301-303 (dec.) 160-165 (dec.) 272.5-273 224.5-225

158 158 158 158 158 139, 158

307-318 290-295 238-244 249-252

(dec.) (dec.) (dec.) (dec. )

-llO(c)

+112(p)

- 106 (PI

C18H2oNz

W

W

PART14. Indole and oxindole alkaloids related to /?-carboline and their derivatives mP Compound 3-Carbomethoxyharman, CCCXXXIX, b 180 11-Methoxyyohimbine stereoisomer, ww 162 Deacetylpoweridine, CCXCVIII Isodeacetylpoweridine /?-Lactone, CCCII Poweridine, CCXCVII, x x 161 LiAlH4 diol Dihydrocorynantheol, CCCVI, y y 163; zz 48

Formula

(" C)

252-253 148-149 222 (dec.) 197 (dec.) 277 (dec.) 226 (dec.) 215 (dec.) 181-183

[.ID

Reference 180 162 161 161 161 161 161 198, 163

-5(a)

- 19 (c)

- 37 (P) Methochloride, cc 164 10-Methoxydihydrocorynantheol, CCCVI-A, h 113d; ee 48; ww 164a 19,20-Dehydro, CCCVI-B, h 113d; ee 37; ww 48 Ajmalicine, CCCXVII, v 66a

272-273 165-166

Aricine, CCCIII, y y 163

186-187 (dec.)

Reserpinine, CCCIV, zz 48 Reserpiline, CCCLVI, h 113d

243-244 (dec.) amorph.

Isoreserpiline, CCCV, k 41; h 171; 113d; rr 156; tt 161; uu 161 ; vv 161 Methochloride, rr 182 Isoreserpiline-4-indoxyl, CCCLVII-A, h 113d Isoreserpiline oxindole (see Carapanaubine below)

210-212 (dec.)

184-185 253-254 (dec.)

283-985 (dec.) 250-253 (dec.)

+ 6 3 (w-m)

- 16 (p)

198 113d, 48

-65 (p) 113d, 48 - 61 (c) 66a, 55, -48 (p) 192b -89 (c) 55, 192b, -66(p) 192~ -59 (0) -131 (c) 65, 192b - 69 (m) 113d, 192b - 14 (P) -82 (p) 55, 192b -134(m)

182 113d, 192a

w

TABLE I-Continued

01 ip

Compound

Formula

Norrnacusine-B, CCCXX, ff 116, 121; b 165

Reference 245 and 275

172, 165, 116, 121

Methochloride 0-Acetate, CCCXXVIII

248-249 223

Dihydro, ZOor-Et

189-190 or 159-160 192 and 219-220 279

173 165, 121, 116, 177 172, 165

Dihydro-0-acetate, 2Oa-Et Methiodide (macusine-B iodide), b 173a

td

172 173, 121, 116 174, 176 165 174, 175

Dehydroxymethylpolyneuridine, CCCXXX t-Butyl ester, CCCXXXI Akuammidine, CCCXXII, k 113e; D (Volume V I I )

231 246-247 2 34-2 36

Polyneuridine, CCCXXI, b 165

245-247.5

165, 178, 176 173, 165 173 165, 176 174, 165

Methiodide = macusine-A iodide Methochloride 0-Acetate Akuainmidinol, CCCXXVI

B.Me1

274 (dec.) 252 (dec.) 278 260-265

0,O-Diacetate 17-Dehydropolyneuridine(aldehyde), CCCXXIX-A

C24H28NZO4 CziHzzNz03

224-227 285-286

165 165

z r ?:

Polyneuridinic acid, CCCXXIV Hydrochloride Quebrachidine, CCCXXXVIII-D, a 179b 0,N-Diacetate, CCCXXXVIII-E N-Methylquebrachidinol, CCCXXXVIII-G Hydrochloride Monoacetate, CCCXXXVIII-H N-Methylquebrachidinal, CCCXXXVIII-I Eburnamenine, CCCXLIII, cc 183; aa 91; a 28, 51a, 51; e 51 Picrate Eburnamine, CCCXL, cc 183; aa 53, 91; e 51; F 113b Isoeburnamine, CCCXLI, F 113b 0-Methyleburnamine, CCCXL-A, F 113b Eburnamonine, CCCXLII, cc 183; e 51 Tuboflavine, CCCXLVIII, bb 186 Carapanaubine, CCCLIV, A 171

CzoHz3ClNz03 CziHz4NzOa CzsHzsNz05 CziHz6NzOz Cz3HzsNz03 CziHz4NzOz C1gHzzNz CigHdzO CigHz4NzO CzoHzaNzO CigHzzNzO CiaHizNzO Cz3HzsNzOs

255-265 276-278 amorph. 255-260 199-201 212-215 amorph. 196 181 217 181 183 207-208 221-223

+54 (c)

+183 (c)

- 93 (0) +111 (c)

+ 89 (0) -101

(0)

165 179b 179b 179b 179b, 179c 179b Chapter11 Chapter 11 Chapter11 113b Chapter 11 186 171

3 %

i! b-

w

U

F

5

-

Pleiocarpamine, a a 91 Pleiomutine, aa 91 Dipicrate Distyphnate Pleiomutinine, a& 91 Lactam 3 (P.tubicina)

b

Q

z

PART15. Alkaloids of unknown structure and their derivatives

Compound

w rp

Formula ___ CzoHzzNzOz C4z-43Hsz-tieN402 B. 2C6H3N307 B. 2C~H3N308 C40H46-48N402

mP (" C)

159 amorph. 230 230 220 287-290 (dec.)

U b-

[@ID

Reference

+123(c) -97 (c)

91.53 91 91 91 91 96

k

s

0

u1

~

- u 1

TABLE I-Continued

0

01 Q,

mP Compound

Formula

(" C)

[a]=

Reference

108, 89, 56 .108, 89, 56 108, 89 109, 113, cf. 100 109, 113

Kopsingine, B 108, 89, 56

270-274 (dec.)

+75 ( c )

Kopsaporine, B 108, 89, 56 Kopsingarine, B 108, 89 Fruticosine, dd 53, 100, 109

224 (dec.) 230 (dec.) 225-226

+48 (c)'

0-Acetylfruticosine Decarbomethoxyfruticosine Isodecarbomethoxyfruticosine 0-Acetyldecarbomethox yfruticosine Fruticosamine, dd 109, 113 Elliptinine, rr 156 Unnamed, 0. sandwicensis Hydriodide Unnamed, 0. oppositifolin Methoxyellipticine, rr 156; ss 156; tt 161; uu 161; vv 161 Elliptamine Picrate Powerine LiAIH4 product, picrate Poweramine, xx 161 Ochropine, xx 192d Unnamed, A. australe

115-118 or 132-134 292-293 290-294 211-212 177-181 231-233 215 (dec.) 282-284 270-272 (dec.) 170 188-189 (dec.) 212 (dec.) 241-242 146 186-1 88

- 19 (c)

+43 (c) - 255 -I-27 0

-216 (a)

- 229 (a)

109 109 109 109, 113 156 156 194 156, 55 161 161 161 161 161 55, l92d 147

td

B

W M

E

Quebrachacidine, a 195 Aglycone Rhazinine, e 62, 197 Tosylate ~~

Cz6Hz8Nz011 Ci9Hz4NzO ~

~~~

234-238 315-32 1 115-1 16 280 (dec.)

- 250 + 4 (e)

195 195 197 197

~

Plant sources: a, Aspidosperma quebrachoblanco Schlecht; b, A. polyneuron Mull.-Arg. ; c, A. chakensis Speg. ; d, A. album (Vahl) R. Benth. ; e, R h z y a stricta Decaisne; f, G o n i o m k a m s s i E. May; g, Stemmadenia donnell-srnithii (Rose) Woodson; h, Aspidosperrna discolor A.DC. ;i, A . eburneum Fr. All. ;j, A. limae Woodson; k, Vallesia dichotoma Ruiz et Pav. ; 1, V . glabra (Cav.) Link; m, Aspidosperma quirandy Hassler (57, 58); n, A . megalocarpon Mull.-Arg.; 0,A . multijlorunz A.DC.; p, A . obscurinervium Azambuja; q, A . triternatum Rojas Acosta; r, A . pyrifolium Mart. ; s, A . cylindrocarpon Mull.-Arg.; t, Amsonia tabernaemontana Walt.; u, Stemrnadenia tomentosa Greenman var. palmeri; v, S. pubescens Benth. (S. obovata K. Schum.); w, Tabernaemontana citrifolia L. ( T. alba Mill. or Nicholson); x, Aspidosperma refracturn Mart. ; y . A. populifolium A.DC. ; z, Kopsia longifiora Merrill; aa, Pleiocarpa mutica Benth. ; bb, P. tubicina Stapf; cc, Hunteria eburnea Pichon ;dd, Kopsia fruticosa A.DC. ;ee, Adspidosperma spp. ;ff, Diplorrhyncus condylocarpon (Mull.-Arg.) Pichon spp. nzossambicensis (Benth.) Duvign. ( D . mossambicensis Benth.); gg, Aspidosperma compactinerviurn Kuhlm. ; hh, A. ulei Mgf. ; ii, A. olivaceum Mull.-Arg,; jj, A. pyricollum Mull.-Arg. ; kk, A. australe Mull.-Arg. ;11, A. dasycarpon A.DC. ; mm, A. gomezianum A.DC. ; nn, A. subincanum Mart. (Brazil)-see 0 0 3 ; 00, A. subincanum Mart. (Peru)-see nn3; pp, A. longipetiolatum Kuhlm. ; qq, Tabernaemontana psychotrzfolia H.B.K. ; rr, Oehrosia elliptica Labill. ; ss, 0. sandwicensis A.DC. ; tt, 0. moorei F. Muell. ; uu, 0. glomerata Valeton; vv, 0. coccinea Mgf. (Ezcavatia coccinea Mig. T. and B.); ww, Aspidosperma oblongum A.DC. ; xx, Ochrosia poweri Bail. ; yy, Aspidosperma marcgravianum Woodson; zz, A. auriculatum Mgf. ; A, A. carapanauba Pichon; B, Kopsia singapurensis Ridley; C, A. parvifolium A.DC. ; D, Picralima nitida (Stapf) Th. and H. Durand ( P . klaineanu Pierre); E, A . sandwithianurn Mgf.; F, Haplophyton cimicidum A.DC.; G, A . fendleri Woodson; H, A . spruceanum Benth.; J, A . nigricans Handro; K , A . hilurianurn Mull.-Arg. 2 Rotation in: c, chloroform; a, acetone; d, dioxane; e, ethanol; m, methanol; w, water; ea, ethyl acetate. 3 Two specimens of Aspidosperma subincanum Mart. have been studied, one collected in Peru and the other in Brazil. The alkaloids isolated, although chemically related, were different with the exception of olivacine, which occurs in both. 4 Whether this difference is due to solvent or to the existence of two enantiomeric forms of kopsine is not clear. 5 Cornpare strychene picrate, mp 150-153" and 178-184" (67). 1

? b

' 2

%

8

2U

2G

kL

Ei

z1

358

B. GILBERT

3,ij-diethyl- and 3-ethyl-5-methylpyridinecould be isolated as the mixed picrate (15, 16, 12). Other dehydrogenation products which were not identified consisted of /3-alkylindoles, methylcarbazoles, and a substituted ci- or /3-carboline. Oxidation of quebrachamine under a variety of

11-A; R = H 11-B; R = Me

XCII

conditions gives products which are probably indolenines. For example, ozone and peracids give a hydroxy base, C19HzsNzO (probably 111, 13); catalytic oxidation gives C19HzGNzOZ (probably IV) which is readily converted to I11 (15). A tribromide, C19HZ3NzBr3, probably of similar

111;R = O H I V ; R = OOH V; R = C N

structure, is obtained by treatment of quebrachamine with N-bromosuccinimide (13). The hydroxy base (111)is reconverted to quebrachamine (I)by lithium aluminum hydride, whereas alkali converts it to a mixture of an oxindole and an indoxyl. Yet another indolenine, of probable structure V, was obtained by the action of cyanogen bromide on ( + )-quebrachamine (8). Hydrolysis with alcoholic alkali gave back ( + )-quebrachamine.

14. Aspidosperrna

359

AND RELATED ALKALOIDS

An examination of the NMR-spectrum of quebrachamine (I) (17) showed the absence of an a-hydrogen on the indole nucleus and confirmed the absence of an N-methyl group. The real confirmation that quebrachamine had structure I came with the use of mass spectrometry by Biemann and Spiteller (18, 18a). First, the zinc dust distillation was examined and its products, separated by vapor phase chromatography (VPC), were shown to be 3-ethylpyridine ( l a % ) , 3-ethyl-4-methylpyridine( 5 % ) , (75y0),3-methyl-5-ethylpyridine and 3,5-diethylpyridine (5%) together with 3-methylindole, 2-ethylindole, 2,3-dimethylindole, 2,3-diethylindole, and a methylethylindole. It will be seen that the major pyridine fragment does not require rearrangement for its formation and, together with 2,3-diethylindole, contains all the carbon atoms of structure I . I n order to provide a model with identical aliphatic structure, aspidospermine (11)was converted to 17-methoxyquebrachamine (VIII). Deacetylaspidospermine (VI) was oxidized with iodine and alkali to the indolenine VII, which although contaminated by unchanged VI was recognized by its UV-spectrum and its conversion by lithium aluminum deuteride to monodeuterio-VI. Compound VII was susceptible t o a reverse Mannich condensation (19, see arrows in VII but note that reaction may not proceed in two stages as indicated) and, on reduction with sodium borohydride, it gave 17-methoxyquebrachamine (VIII) which was separated from the accompanying unchanged deacetylaspidospermine (VI) by alumina chromatography. A comparison of the mass spectra of I and VIII left no doubt as to the complete identity of the aliphatic portions of the two molecules as can be seen from Table I1 in which the main peaks are given. The fragments expected from the breakdown of I (20, 21) are in fact TABLE I1 MASSSPECTRA OF QUEBRACHAMINE (I),17-METHOXYQUEBRACHAMINE(VIII), AND ~ ~ - D E U T E R 17-METHOXYQUEBR IOACHAMINE (19d-VIII) Molecular weight of fragment/charge = m/e Indoles Alkaloid Quebrachamine VIII 19d-VIII

M+ M-Me M-Et 282 312 313

267 297

253 283 284

Piperidines 157 187 187

143 173 173

138 138 139

125 125 126

110 110 111

96 96 97

360

B. GILBERT

observed. In the spectrum of V I I I those that contain the indole nucleus are shifted to molecular weights 30 units higher, corresponding to the addition of the 17-methoxyl group, whereas those containing the aliphatic piperidine portion appear a t identical m/e values. Confirmation that the m/e 124 peak in fact represents the piperidinic ring was obtained by preparation of 19-deuterio-17-methoxy-quebrachamine(19d-VIII) by using sodium borodeuteride, and in its spectrum this peak was partly

I JICO

I

H

\'I

Meo

ii TI1

19d-VIII; R = D

i I

\ iiijc

1Ji

inje 110

injc 125

mie 96

i inje 124; R = H mje 125; R = D

14.

d S p ~ d O S p ~ ? W % AND Cl RELATED ALKALOIDS

361

shifted to m/e 125 (two-thirds), part remaining a t m/e 124 (note preferential loss of C-19 H over C-19 D). Both the transitions m/e 138t o m/e 110 and m/e 125 to m/e 96 were confirmed by observed metastable peaks (see Section 11,D). 17-Methoxyquebrachamine (VIII) has the rotation, - 103" in dioxane, compared with - 111" found for quebrachamine (I)in the same solvent, and the optical rotatory dispersion (ORD) curves are very similar. It may therefore be deduced that ( - )-quebrachamine has the same absolute configuration a t position 5 as does aspidospermine. The closure of a bond between positions 12 and 19, presumably involved in the natural synthesis of aspidospermine-type alkaloids from quebrachamine, has been shown (18) to occur during the zinc dust distillation of the latter when a compound I X was isolated whose mass spectrum was exactly comparable with that of the indolenine VII, except for the 30-unit shift in the indole-containing fragments (Table V). Substance I X has subsequently been found in nature (Section 11,E). The total synthesis of ( f )-quebrachamine has been achieved by a modification of the route used for the synthesis of aspidospermine (Section 11,C; ref. 27a). Condensation of the intermediate X X X - J with phenyl hydrazine gave dl-IX which on borohydride reduction yielded ( f )-quebrachamine (I).

C. ASPIDOSPERMINE The structure of aspidospermine (11) was practically solved by chemical degradative methods (see formula CCCCXXXIII, Volume VII, p. 131) but the final location of the ethyl side chain and the relative stereochemistry only became known after the single crystal X-ray study of its methiodide (22, 11) which was shown to have structure X.l Extensive degradative work, the results of which appeared side by side with those of the X-ray determination, had shown the nature of the structure around both nitrogen atoms and established the size of rings D and E (24, 12). Chromic acid oxidation of aspidospermine gave three lactams (XI, XII, and XIII, 24). The IR-carbonyl frequency (1680 cm-1) of the new amide group in lactam X I shows it to be in a five-membered ring, while the presence in the NMR-spectrum of a nonequivalence quartet a t 2.37 6 1 Stereochemical formulas show relative but not absolute configuration throughout Sections 11-IV.

362

B. GILBERT

(J = 16 cisec), attributable to the two protons at C-11, showed by the absence of further splitting that there is no hydrogen atom at position 12. This is supported by the fact that the single hydrogen atom at C-11 in the hydroxylactam XI1 shows only a singlet at 3.83 8 shifted to 5.10 8 in the acetate, XV, where it is well clear of other absorption. That no rearrangement had taken place during the formation of these two lactams was established by conversion of XI1 t o X I using phosphorus oxychloride to give XIV and then zinc dust followed by reduction of XI to the known N,-deacetyl-N,-ethylaspidospermine (XVII, 25). These results establish R3C. CHzCHzN in the five-membered ring E. Following this, other experiments elucidated the nature of ring D. Mild dehydrogenation of aspidospermine (11) with mercuric acetate gave 7,8-dehydroaspidospermine(XVIII),whose perchlorate (XIX)has the double bond in the expected 8,9 position showing C=N+ absorption at 1698 em-1. Sodium borohydride reduction of this salt to aspidospermine showed that no skeletal change had occurred, while silver oxide oxidation in aqueous dioxane gave the six-membered lactam (XX, vc0, 1625 em-1) in which the carbonyl group is in ring D. The characteristic 8-lactam absorption was clearer in the deacetyl derivative (XXI). IR-evidence and, particularly, coupling of 7,s-dehydroaspidospermine (XVIII) with phenyl diazonium chloride, which from its mechanism must take place at position 7 to give XXII, established the presence of a hydrogen atom in this position in XVIII and hence excluded structure CCCCXXXIII (Volume VII) for aspidospermine. The formation of 3,5during the dehydrogenation of diethyl- and 3-methyl-5-ethylpyridine aspidospermine under vigorous conditions (25) thus involves a rearrangement of the ethyl side chain. These results establish the series N-CHzCHZ in the six-membered ring D. The presence of a third adjacent methylene group was shown in another series of degradative steps. Aspidospermine N,-methiodide (X, 26, 2 7 ) which is reconverted t o aspidospermine under Hofmann degradation conditions, suffered the Emde degration (26, 24) to give the dihydromethine (XXIII) whose methiodide underwent Hofmann degradation to give the methine XXIV in which rings D and E have been opened and which has a terminal methylene group (IR-,910,995 em-1; NMR-, two one-proton quartets at 5.01 and 5.05 6 in the vinyl region). In addition, it has a vinyl hydrogen atom flanked by four vicinal protons (NMR-complexmultiplet at 4.24 6). As we know that position 12 is quaternary, this vinyl proton cannot be in position 11, and it must therefore be in position 7, the nitrogen having remained attached to carbon 10. From the aforementioned NMRabsorption, we may deduce that there is a methylene group in position 6.

XII; R = OH X I V ; R = C1 X V ; K = OAc X V I ; H. = O H (epimer of XII)

XX

I

i 0

J 363

PIIS*(’I

364

B. GILBERT

This result was confirmed by cleavage of the double bond in XXIV to give the aldehyde XXV, in which the aldehydic hydrogen atom shows a 1: 2 : 1 triplet a t 9.83 6 and therefore lies adjacent t o a methylene group in position 6. Furthermore, isomerization of the methine XXIV with hydrochloric acid moves the double bond to the 6,7 position (XXVI) as shown by the appearance of an allylic methyl group in the NMRspectrum (1.68 6, doublet, J = 2.6 cisec), and in this compound (XXVI) there are two vinyl protons (multiplet a t 5.47 6). The transformation of XXIV t o XXVI does not involve rearrangement because both compounds on reduction yield the same dihydro derivative, XXVII.

1

1. HCI-Hs0

z,3lcI

I

1. JleI 2. KOt-Uu

Me0

SSVIII

SSIS

XXIV

XSX

Ac XXVI

XXVII

No real evidence was adduced for the quaternary center a t position 5 in aspidospermine (11),but the formation of lactams and compounds containing C=N,, in which the nitrogen has necessarily the planar configuration excludes any bridging of the two rings D and E.

14. Aspidosperma

AND RELATED ALKALOIDS

365

Evidence for the six-membered ring C was lacking, and an attempt to clarify this portion of the molecule was made starting with the N,dimethiodide of deacetylaspidospermine (XXVIII, 12). Emde degradation gave X X I X in which the aryl-nitrogen bond has been cleaved. This underwent Hofmann elimination of N, on prolonged heating with methyl iodide to give a base containing only one nitrogen atom which could be XXX, which contains a hydrogenable double bond. Unfortunately, no quinoline derivative or other recognizable product could be isolated from the dehydrogenation of this compound, probably because of the two resistant quaternary centers in ring C. The synthesis of aspidospermine (11)has been achieved (27a).Starting with n-butyraldehyde in which the a-methylene group represents the eventual quaternary carbon atom a t position 5, acrylic ester was added by way of the enamine synthesis (27b) to give XXX-A. A second enamine synthesis using methyl vinyl ketone furnished the aldehyde XXX-B which by spontaneous condensation gave the isolable cyclohexenone, XXX-C, in which the final ring C is present and the propionic side chain provides the carbon atoms of ring D. A Mannich-type addition of ammonia to XXX-C led to the bicyclic amides XXX-D containing rings C and D. Reaction of XXX-D in the Fischer indole synthesis with o-methoxyphenylhydrazine led to the two isomeric indoles XXX-E and XXX-F, showing that enolization of XXX-D occurs away from, rather than toward, the ring junction. To induce enolization in the desired direction, it was necessary to build on ring E, which was achieved by lithium aluminum hydride reduction o f the amide carbonyl group of XXX-D (the ketonic carbonyl was protected as the ethylene ketal) to give the amine, XXX-G, which furnished the tricyclic amide, XXX-I in two steps. A slow acetic acid-catalyzed Fischer indole synthesis using this amide still gave only indolic material showing that enolization, even under these equilibrating conditions, proceeded away from the ring junction, due to the fact that enolization in the desired direction would have resulted in a strained five-membered ring containing three trigonal atoms. Reduction of the keto-amide, XXX-I to the keto-amine X x x - J removed two of these atoms, and the Fischer synthesis with this compound yielded an indolenine which was the dl-form of 1,2-dehydrodeacetylaspidosperniine (VII). The fact that this product, had the correct stereochemistry derives from the equilibration of the asymmetric centers a t positions 12 and 19 by the reversible conversion to dl-XXX-K (see Section 11, B) during the Fischer synthesis (note too that the intermediate phenylhydrazone could similarly equilibrate), VII being the most stable stereoisomer. Lithium aluminum hydride introduced hydrogen on the desired side of the molecule (see Section 11,B, Refs. 18,

366

B. GILBERT

xxx-I

XXX-H

XXX-J

\/+/U I H

I Me0

I Me0

dl-VII

Ac

dl-I1

14. Aspiclosperma

AND RELATED ALKALOIDS

367

18a) at position 2, and acetylation of the product furnished dl-aspidospermine whose IR- and mass spectra were identical with those of the natural alkaloid (27a).

D. NMR- AND MASS SPECTRA OF THE ASPIDOSPERMINE-TYPE ALKALOIDS Although NMR was only partly responsible for the structure determination of aspidospermine (11)and the mass spectrum of the alkaloid was only measured after its structure was known, these two physical methods have had immense application in subsequent investigations of related alkaloids. For this reason, the more important data have been collected in Tables I V and V. Considering first the NMR-spectra, the nature of the aromatic substitution pattern may often be deduced from the absorption between 7.3 and 6.6 6 [see pyrifolidine (XLVI), aspidocarpine (XLIV), and aspidoalbine (CLXXXVIII), for example], especially when the data are taken in conjunction with the UV-absorption spectrum (Table 111). Of the nonaromatic protons, the one found furthest downfield is the C-2 hydrogen atom, which appears as a quartet centered at 4.0-4.5 6 (29,45). The absence of this peak is indicative of substitution at this point and is an important means of recognizing alkaloids of the aspidofractinine group (Section 111) in which the sixth ring terminates at position 2 . Alkaloids which lack the C-2 proton but which bear a carbomethoxy group on C-3, e.g., refractine (CL-A),exhibit a quartet at about 3.8 8 due to the C-3 proton. The four protons next to nitrogen in positions 8 and 10 absorb in the region 3.3 to 2.9 6 and in all the alkaloids based on the aspidospermine skeleton form a characteristic pattern which is not obscured by other absorption (9, 29) and which is profoundly altered in the spectra of the related hexacyclic bases (Sections I11 and IV). The single proton on C-19, not having any neighbor, produces a singlet from 2.2-2.5 6 which is sometimes hidden by absorption due to the N-COCH3 where this group is present, though in these cases, its presence may be deduced from the integration curve. In addition to these absorptions, there are usually easily recognized three proton singlets due to aromatic methoxyl groups a t 3.75-3.90 8 and to the methyl group of an N-acetyl at 2.2 6. I n the case of N,-propionyl compounds a quartet may be observed at 2.3-2.8 6 due to the COCHz protons, though in a 60-mc spectrum, slight overlapping occurs with the N,-C( 19)H absorption in some cases. The methyl group of such a propionyl group exhibits a triplet centered a t approximately 1.25 8. The terminal methyl group of

TABLE 111

W

% --

UV-DATA Wavelength, mp

Chromophorel

Reference

log € 2 ~

Dihydroindoles (N,H) Unsubstituted, XCVI Unsubstituted, CXXII-Me1 16-Methoxy-, CXVI 17-Methoxy-, C X L I X - F

16,17-Dimethoxy-,X L V I I

n3

ac

n n

ac n aC

alk 15,16,17-Trimethoxy-, CXCII 17-Hydroxy-,X X X V I N-Alkyldihydroindoles N,-Methyl, CXLVI-A 16-Methoxy-,C I I I 17-Methoxy-, X L I

16,17-Dimethoxy-,L V I I N -Acyldihydroindoles N-Formyl, C X X N-Acetyl, CXLIV-A N-Carbomethoxy-, CXLV-A 17-Methoxy, -11 16,17-Dimethoxy-,X L V I

n

n

n

n n ac n

n

n n

n n

245, 2984 226, 260(sh),5 268(sh) 225, 250(sh), 300 212, 245, 288 215, 268, 275 215, 293 230, 279 as neutral 212, 240(sh), 305 245, 292

3.86, 3.48 3.58 3.98, 3.65, 3.52 4.56, 3.88, 3.41 4.36, 3.22, 3.23 4.40, 3.50 3.97, 3.30

206, 254, 211, 251, 220, 266, 218, 272, 218, 263,

4.42, 3.97, 3.52 4.52, 3.85, 3.71 4.42, 3.87, 3.40 3.47, 2.86, 2.86 4.36, 3.80, 3.48

300 304 305 278 305

208,253, 278,288 210, 257, 286, 294 206.5, 246, 282.5, 289.5 220, 257, 280-290(sh) 223, 252,286

~-

7 37 32 37

9, 10

4.48, 3.85, 3.63 3.66, 3.24

4.36,4.16, 4.33,4.15, 4.49, 4.20, 4.56, 4.09 4.55, 3.99,

3.72, 3.67 3.77, 3.74 3.51, 3.48 3.37

91 55, 72 25 36

81 96 91 25 10

n

16,17-Dimethoxy-,L X X 15,16,17-Trimethoxy-, CXC 17-Hydroxy-, L X X X V I I 16,17-Dihydroxy-,L X I X 16-Methoxy-17-hydroxy-,LIII 15,lR-Dimethoxy-17-hydroxy-, CLXXXVIII Indolenines Decarbomethoxymossambine, CCXXXI 1,2-Dehydrodeacetylaspidosperinine, VII

a-Methyleneindolines Vincadifformine, X C I I I Condylocarpine, CCXV Norfluorocurarine, CCXL 16-Methoxyrninovincine, CX 11-Methoxy-14,19-dihydrocondylocarpine, CCXXIV - A

n alk n alk n alk n alk

222, 250, 290(sh) 218,258, 295 221, 261, 292 230, 258(sh), 311 225, 260, 295(sh) 216, 237, 304 228, 264 224, 307 227, 267 308

4.48, 3.85, 3.27 4.51, 4.15, 3.58 4.40, 3.92, 3.55 4.39, 3.80, 3.74 4.30, 3.89 4.29, 4.24, 3.44 4.43, 3.94 4.42, 3.73 4.16, 3.88 3.64

n n

220, 262 228, 236(sh), 255, 307

4.32, 3.84 4.22,4.14, 3.69, 3.65

121

n

225, 300, 328 228, 295, 328 242, 299, 360 230, 250, 325 255, 286, 327

3.97, 4.03, 4.19 4.04, 4.01, 4.17 3.98, 3.57, 4.25 4.05, 4.03, 4.14 4.17, 4.04, 4.05

6 116 116 32

n n

n n n n

36 52 33 36 31 52

18

48

w t+

b

9. 3 % 3 -3

z U

P d

Fr3 M

tl P

t-

x

Extended indole Uleine, CCXLV

n

209, 309

4.38, 4.30

138

Py.ridociirbazoles Olivacine, CCLVIIl6

n

224, 239, 269, 277, 288, 293, 315, 330, 345, 377 242, 307, 351, 412

4.32,4.22, 4.54,4.71, 4.89,4.8S, 3.61, 3.73, 3.52, 3.58 4.45,4.92, 3.78, 3.59

160

ac

t. 5 F:

u

4 0

TABLE 111-Continued

Chromophorel

Ellipticine methonitrate Anhydronium base from olivacine methiodide, CCLIX 1,2-Dihydroellipticine, CCXCIII

1,2-Dihydroolivacine, CCLXXlV Guatambuine, CCLX7 Oxindole Carapanaubine, CCCLIV $-Indoxy1 Isoreserpiline-~-indoxyl,CCCLVII-A

Wavelength, mp

Reference

log €2

n n

241, 249, 307, 356, 423 270, 336, 373,400, 510

4.38, 4.36, 4.86, 3.72, 3.68 3.58,4.69, 3.88, 3.78,3.60

158 149

n

236, 244, 271(sh), 281, 302, 313,382 235, 270(sh), 279, 301, 312, 378 239, 248,275, 330, 344 235, 283, 314, 377 241, 250(sh), 263, 300, 330, 344

4.45,4.30,4.46,4.64, 4.18,4.30,4.43

158

ac n ac n

4.42, 4.45, 4.63, 4.48,4.46,4.59, 4.43, 4.63, 4.33, 4.64, 4.51, 4.38,

4.15, 4.28, 4.36 3.94, 3.63 4.32 4.30, 3.60, 3.52

W

139, 150

B z

W M

n ac

215, 244, 278(sh), 300(sh) 222, 246(sh), 278, 300(sh)

4.57, 4.23,3.80, 3.66 4.56, 3.79, 3.61,4.15

171

n

226, 251, 282, 402

4.36, 4.46,4.07, 3.74

113d

1 Many examples of the chromophores recorded are found in other alkaloid groups, especially that of the curare alkaloids (see Chapter 15). An excellent collection of spectra is found in refs. 55 and 56. 2 Log E rather than E values have been given for uniformity, as most literature values are so recorded. 3 Abbreviations: n = neutral, ac = acid, alk = alkaline. 4 Kopsinine shows peaks a t 205, 246, and 295 mp. 5 sh = shoulder. 6 Ellipticine (CCLXXXIV) has a similar spectrum (156, 55). 7 N-Methyltetrahydroellipticine(CCLXXXV) has a similar spectrum (158, 55).

TABLE IV

NMR- DATA^ PART1. Simple derivatives of aspidospermidine Position of protons; 6 , ppm; no. of peaks; J, cjsec -

Compound

Demethoxyaspidospermine, X X X I I I

Aromatic2

w

16,17

Na

19

8,lO

213

H

substitution

acyl

H

CHz

CH3

2.93-3.35

0.73

40a

2.3s

2.9-3.3

0.72

40

2.28s

2.9-3.3

0.72

40a

3 5U

2.9-3.3 2.8-3.3

0.67 0.69

29 30,45

$

2.23s

ca. 3.0

0.62

31

2.17s

2.9-3.3

0.67

9, 30

0.75

36

-4

29 29

8.13m( 17-H) 7.16m(3H) 8.13m(17-H) 7.07m(3H)

4.08q, J = 6,lO 4.08q, J = 6.5,lO

Demethylaspidospermine, XXXVII

6.9m

Aspidospermine, I1 Aspidocarpine, XLIV

7.0m 6.53t. J = 9

Aspidolimine, L I I I

6.65q, J = 8

4.07q, J = 6,lO 4.50q 3.95q, J = 5.5,lO 4.12q, J = 6.5, 10.5 4.50q, J = 6,lO 4.05d, J=8 4.5q 4.4% J = 6.5,lO 4.12q

Demethoxypalosine, XXXIV

Pyrifolidine, XLVI

6.6%

J = 8.2 Spegazzinidine, LXIX

6.57q

Cylindrocarpidine, LXXVI Cylindrocarpine, LXXV

7.0m 7.17m

Limaspermine, LXXXVII

6.92m

Reference

.”

2

b

6

&

2.27s

Y

10.83s 3.89s 3.80s 10.77s 3.88s 11.0s 3.77, 3.84 5.84m, 11.1s 3.88s 3.84s 10.88s

2.38q 1.24t J = 6.5 2.32s 2.20s 2.27s

Eti U

2.57q 1.25t J=7 2.22s 2.48s 2.22s -5

2.57q J = 7.5

2.45s 2.48s

2.9-3.35 2.9-3.3 ca. 3.0

-4

3.52t6, J=7

33

bw b 5 F1

0 4

w

TABLE IV-Continued

PART2. Aspidospermidine derivatives bearing a carbomethoxyl group at C-3

Compound

Aromatic

6,7 vinyl

19 H

COzMe

~-

~

8 CHz

21 CH3

Reference

~~

Tabersonine, XCII Vincadifformine, XCIII Minovine, CVIII Mmovmcine, CIX Vindolinine,* CVI

5.75s 7. O m 7.0m 6.95 6-7 lines 6 08d, 6 30% 6 91d, J = 2 5,s

Vindoline, CIII

3.64s 3.75s

-

5.6-6.3 12 lines 5.23, 5.88, ,,J = 10

3.78s 3.68s 3.80~9

8.98 8.88 3.25 8.8 4.5 approx. 2.68

0.77

2.65

3.4

1.85 0.95d 5=7 0.487

7,66a 6,66a 32 32

3 4

~

PART3. Aspidofractinine and kopsine-type alkaloids and their derivatives

Compound ~

~~

Aromatic .

Aspidofiline, CXXXVI Refractidine, CXX Pyrifolme, CXIX Deacetylpyrifoline, CXXI 6-Acety1-6-demethylpyrifoline, CXXVIII

Pleiocarpine, CXLV-A Refractine, CL-A

6 or 3 substitution

N* 17

substitution

-

19 H ~~

10.02s

7.0m 6.93m 6.75m 7.0m

7.72d( 17-H), 7.17m 6.9m

3.35s 3.33s 3.33s 2.02s, 4.63q, J = 5, 10.5 3.72s 3.68s

3.81s 3.78s 3.80s

2.32s 8.37s 2.12s

-

11 H

_____________~

3.0m

2.12s

-3.0m -3.15m

9.47s

-

3.82s 3.85s

8,lO CH2

Reference -~

80,37 81 81,37 37 81,37

3.0m

91

3.0m

82, 37

Isorefractine, CL-C Deformylrefractine, CXLIX -A Deformylisorefractine, CXLIX-C N-Methyldeformylrefractinol, CLI-F

6.9m G.73m 6.8m 6.8m

N-Methylkopsinylene, CXLVI-M N-Methylkopsinane, CXLVI-0

A'-Methylisokopsinylene, CXLVI-N

3.78s 3.75s 3.80s 3.95oct J = 2.5, 6.5,12 5.0 2 x 6 lines 1.3d J = 6.5 2.03d J = 1.5

3.83s 3.80s 3.82s 3.77s

--

9.20s

N

2.90s

3.0m 2.9m 3.1 3.0m

37 37 37 37

95 95 95 4.48q

X-Methylenekopsinyl ether, CLIII

95

J=7 Kopsine, CLXX

7.4m

7.2S'U

3.93s

Fruticosamine11 Fruticosinell

7.251~1 7.70(17-H) 7.2m

5.52br10 3.1710 4.79q'Z

3.90s 3.88s

100, 101, 109 109 109 3.75d J=1 3.5d J = -1 3.57s

Kopsine-l0-lactam,11CLXXV Dihydrokopsine- 10-lartam, CLXXVlI Kopsane- 10-lactam,ll CLXXXVl

2.82s

101

2.63d J = -1

101 101

PART4. Aspidoalbine-type alkaloids and their derivatives

Compound

Aromatic

21 CHz

l5,16,17 N, substltutlorl substitution

-~ ~

Aspidolimidine, CCI

Aspidoalbine, CLXXXVIII

6.73d 7.09d J=8 6.82s

0-Methyl-depropionylaspidoalbine, CXCII

6.89

3.88s 0.78s 3.86s 3.85s

2.32s

4.05m

-4.3m

8,10 CHz _ _ -2.85m

-

Reference 40

52

3. O m

4.02q _ _ _ _ ~

52 ~~

w 4 w

TABLE IV-Continued

W

4

ip

PART5. Alkaloids related to condylocarpine and mossambine and their derivatives ~~~

N,

Aromatic

Compound

16 substitution

3,15

H

19

18

Reference

~

3.78s

Condylocarpine, CCXV

2,16-Dihydromossambine, CCXXIX

6.8m

Decarbomethoxymossambine, CCXXXI Echitamidine, CCXLII

7.05m

8.68

3.89s

Lochneridine, CCXLIII Stemmadenine, CCXIII

7.35111

9.3

3.77s 3.79s 4.38

3.92m 4.12d J=2

5.32q

1.58d

117

5.55

1.66d J=6

122

5.33

5.4q

td

122 126

1.16d J=6 0.77m 1.7d

127 117

PART6. Indole and oxindole alkaloids and their derivatives ~~

~

Compound Uleine, CCXLV

Aromatic 7.34m

N,

COzMe

8.38s

17 __ 5.27s 4.98s (vinyl)

3-Carbomethoxy-1-methyl-P-carboline, CCCXXXIX

7.0-8.7m

9.40s

3.95s

18

19 -~

0.83m

Reference -

~

142

(Et) 180

Yohimbine, CCXCV 8-Yohimbine, CCXCVI Dihydrocorynantheol, CCCVI

7.25m 7.17m 7.2m

7.90s 7.87s 8.80s

0Acetyldihydrocorynantheol

7.15m

8.12s

Isoreserpiline, CCCV

6.77s, 6.90s 6.55s, 6.74s

7.95s

Carapanaubine, CCCLIV

0,O-Diacetylakuammidinol, CCCXXVII 0-Acetylnormacusine-B, CCCXXVIII

6.25m 6.25m

8.73s

8.43s 8.57s

3.77s 3.88s

3.72s 3.61s

4.05q (CH20Ac) 7.37s (vinyl) 7.44s (vinyl) 4.0m (CH20Ac) 3.93m (CH20Ac)

0.88 (Et) 0.87 (Et) 1.37d

(J = 6) 1.40d

(J = 6) 1.59d

(J = 6.5) 1.57d ( J =7)

37 37 63, 37 63, 37 4.44 oct 4.56 oct (J = 6, 5.7) 5.35q (J = 6.5) 5.38q ( J =7)

171

171

165 165

1 For model compounds, see refs. 45,17,101,66a. Abbreviations: s = singlet, d = doublet, t = triplet, q = quartet, oct = octet, m = multiplet, br = broad. 2 CHCl3 proton appears at 7.28 6. 3 Position of principal peak. 4 C02CH3; cylindrocarpidine, 3.58 6 ; cylindrocarpine, 3.56 6. 5 Trans-cinnamoyl vinyl protons show two doublets at 6.75 and 7.72 6 (J = 16). 6 Shifted to 4.0 6 in the acetate. 7 The high field position of the aliphatic proton absorption in tabersonine and vindoline may be due to increased shielding by ring currents from the aromatic ring (7, 51b). 8 c-2proton absorbs a t 3.85 6, doublet, J = 4.5 c/sec in N-methylvindolinol acetate (3). 9 Aromatic OCHs coincident. 10 Eliminated after replacement of active hydrogen by deuterium. 11 Absorption in 4.0-4.5 6 region attributed in the kopsine lactams to a C-8 proton. 12 Due to CH-CH(OH), converted to a doublet by treatment with D2O.

W

TABLE V

4 Q,

MASS SPECTRAL DATA PART1. Alkaloids which fragment as aspidospermine Mol. wt. of fragmentlcharge = m/el Compound

Reference

-

Mf2

a2

282 312 313 342 324 338 296 326 354 412

b3

c4

254 284 285 314 296 310 268 298 326 384

130 160 1615 190 130 130 144 174 160

-7

220 190 176 176

124 124 1246 124 124 124 124 124 124 182 182 124 124 124 124 124 124 1408 140 140 141 14P

Other peaks -

Aspidospermidine, X X X I Deacetylaspidospermine, VI 2-d Deacetylpyrifolidine, XLVII Demethoxyaspidospermine, X X X I I I Demethoxypalosine, X X X I V N-Methylaspidospermidine, X X X V N-Methyldeacetylaspidospermine, X L I Aspidospermine, I1 21-Acetoxyaspidospermine, LXXXV 0,N-Diacetyl-0-methyldeacylaspidoalbinol, CXCV Pyrifolidine, XLVI Aspidocarpine, XLIV Aspidolimine, L I I I Spegazzinine, LXIV Spegazzinidine, L X I X Demethylaspidospermine, XXXVII Limaspermine, L X X X V I I N-Decinnamoylcylindrocarpol, L X X X I V 0-Methyldeacylaspidoalbinol,CXCIII 19-d, CXCIV N-Et.hyldecinnamoylcylindrocarpo1, LXXXVI

384 370 3842 356 372 340 370 328 388

356 342 3562 312 328 3428 300

146

-7

220 220

356

328

.-

28, 21 9, 21 21

339, 311, 1099

9 40a 40 28,21 28 28,21 52 52 28, 37 40 40 36 36 40a 33 52 52 52 52

m

2 EM

N-Ethyl-0-methyldeacylaspidoalbinol, CXCVI N-CD3CH3, CXCVII 2,3-Dihydrotabersonine, XCVI Dihydrovincadifformine, CI 2,3-Dihydrotabersonol, XCVII Dihydrovincadifforminol, XCIX ( = tetrahydrotabersonol) 20-Hydroxydihydrovincadifforminol, CXIV 2,3-Dihydrotabersonane, XCVIII Dihydrovincadifforminane, CII 20-Hydroxydihydrovincadiff orminane, CXV Vindoline, CIII Dihydrovindoline, CXI 16-Methoxy-N,-methyl-4-oxoaspidospermidine, CV 20-Hydroxy- 16-methoxy-2,3-dihydrovincadifforminol, CXVI

416

7

338 340 310 312

252 254 252 254

122 124

328 294 296 312 45610 45810 340 358

270 252 254 270 296 298 298 300

140

248 250

140 121

124

174 174 174 160

124 140 122 124 124 140

309, 282, 135 311, 28410.11 166"

52 52 7 6 7 6, 7 32 7 32 32 21 4,21 4, 21 32

"

PART2. Aspidospermine-type compounds with a 2,3-double bond Mol. wt. of fragment/charge = m/e Compound

Reference

M+ Tabersonine, XCII Vincadifformine (6,7-dihydrotabersonine),XCIII Minovincinine, CVII 0-Acetylminovincinine Minovine, CVIII Minovincine, CIX 16-Methoxyminovincine, CX

336 338 3548 39610 352 35214 382

g12

c

Other peaks 135, 107, 9213

228 214 244

124 1408 18210 124 138 138

7

6, 7 32 32 32 32 32

w

4

4

W 41 06

TABLE V-Continued

PART3. Aspidospermine-type compounds with a l,2-double lbond (indolenines) Mol. wt. of fragmentlcharge = m/e Compound

Reference

~-

1,2-Dehydroaspidospermidine,I X 1,2-Dehydroaspidospermine, VII

M+

M-29

M-70

C

280 310

251 281

210 240

124

28 21

G)

PART4. Aspidofractinine-type compounds

E

b

~_

M

E

Mol. wt. of fragment/charge = m/e Compound

Reference M+

Aspidofractinine, CXVII 0-Methyldeacetylaspidofiline, CXXXIX 6,7-Dehydro-CXXXIX,CXLI N-Ethyldeacetylaspidofiline, CXL Aspidofiline, CXXXVI 0-Methylaspidofiline, CXXXVIII 0-Acetylaspidofiline, CXXXVII Deformylrefractidine, CXXII N-Methyldeformylrefractidine, CXXIII

280 310 308 324 3382 3522 3802 310 32419.20

h

252 282 280 296 3102 3242 3522 282 296

Indole

14415 17415 18816 18815 188 14415 15815

C

i

124 124

109 109 107 109 109 109 109 139 139

124 124 124 124 154 154

Other peaks

29517 130, 10918 10918

102 80 80 37 80 80 80 81 81,37

N-CD2H analog, CXXIV Deacetylpyrifoline, CXXI Refractidine, CXX Pyrifoline, CXIX Deformyldemethylrefractidine, CXXVII N -Acetyldeformyldemethylrefractidine,CXXX N-Methyldeformyldemethylrefractidine,CXXXI N-CDzH analog, CXXXII Deacetyl-6-demethylpyrifoline,CXXV 6-Deuterio analog, CXXVI N,O-Diacetyldeformyldemethylrefractidine, CXXIX 6-Acetyl-6-demethylpyrifoline, CXXVIII 6-Dehydrodeformyldemethylrefractidine,CXXXV 6-Dehydrodeacetyldemethylpyrifolhe,CXXXIII 1,7,7-Trideuterio analog, CXXXIV Kopsinine, CXLII-A Aspidofractine, CXLIII-A Pleiocarpine, CXLV-A Deformylrefraatine, CXLIX-A Refractine, CL-A Isorefractine, CL-C 3-Deuterio analog, CL-S Pleiocarpine lactam A, CXLV-D

326 340 33819 3822 2968 3388 3108 312 326 327 3802 410 294 324 327 338 36620 396 368 39620 396 397

298 312 31019 3542 2688 310 282 284 298 299 35229 10 382 266 296 299 310 338 368 340 368

16015

17415 174 14415 174 14415 17415 17515

157

154 154 154 154 140 140 140 140 140 141 182 182 138 138 140 12422 124 124 124 124 absent23 absent23

259 27212 215 22812

Pleiocarpinilam, CXLVI-D Kopsinyl alcohol, CXLII-F N-Methylkopsinyl alcohol, CXLVI-F N-Trideuteriomethyl-3’,3’, 10,lO-tetradeuteriokopsinyl alcohol, CXLVIII-G Deformylrefractinol, CXLIX-F

16015 17415 14415 17415 14415 14415 15815

12424

81 81 81 81 81 81 81 81 81 81 81 81 81 81 81 82, 37 82, 37 95 82,37 82,37 102 102 102

12424

102 82, 37 95 102

139 139 139 139 125 125 125 125 125 126 16710 167 123 123 125 10922 109 109 109 109

3108 324

282 296

124 124 126

10922 109 111

3408

312

124

109

109 10918 307,21 10918 130

130

154 154

37

TABLE V-Continued

W W

3

~

._

~

Mol. wt. of fragment/charge = m/e

Compound

Reference

~

h

M+

__

Indole

C

i

Other peaks

~-

0 -Acetyldeformylisorefractinol

38225

35425

109

3’,3’-Dideuterio analog, CXLIX-V

384

35625

109

Kopsinylene, CXLII-M

292

264

109

220,207,182, 168,129,12226

37,102

Aspidofract-3-ene, CXLIII-M

320

2922827

10922

213,164,130, 12226

37

Deformylrefract-3-ene, CXLIX-M

32228

109

250,16226

350

29428 3222.27

124

Refract-3-ene, CL-M

124

109

250, 243, 12226

37 37

Deformylnoraspidofractone, CXLII-P

294

266

124

10922

238,209,156, 130, 122,11526

37

Noraspidofractone, CXLIII-P

3222

2942

124

10922

238,209,156, 143,130,12226

37

Deformylnorrefractone, CXLIX-P

324

296

124

10922

268,239,201, 12226

37

Norrefractone, CL-P

352

3242927

124

281,268,12226

Deformylnorrefractol, CXLIX-Q

326

10922

96

37 37,102

1243

82 82

M

z *

H

62

M

z

PART5. Kopsine-like compounds Mol. wt. of fragment/charge = m/e Compound

-

-

Mf

-~

Reference

Other peaks

M-28

__

-

-

Kopsine, CLXX

38027

Decarbomethoxykopslne, CLXXI ,V,-Methyldecarbomethoxydlhydrokopslne, CLXXVIII 3'-Deuter1odecarbomethoxyd1hydrokopsine, CLXXLX Fruticosamine Fruticosine

35227

28225

254(w)

322 338

224 238

210(W)

325

224

196(w)

380 380

35227 35227

282 282

107,106, 109 107 107 107 243 321,230

109 109

PART6. Aspidoalbine-type compounds

Mol. wt. of fragment/charge = m/e Compound

~~~

M+

n'

M-44

b3

c'

310 326 340 356 400 414 400 358

294 310 324 3402 384 398 3842 342

130 146 160

138 138 138 138 138 138 138 138

Reference

Other peaks

~ _ _ _ _ _

1-Acetylaspidoalbidine, CCI-K Haplocidine, CCI-C 0-Methylhaplocidine, CCI-B Aspidolimidine, CCI Aspidoalbine, CLXXXVIII 0 -Methylaspidoalbine, CXC CXCI 0 .Methyl-N,-acetyldepropionylaspidoalbine, CXCII 0-Methyldepropionylaspidoalbine,

338 354 368 3842 428 442 42S2 386

176

220

160 149,310,311 110,160

110,160

113e 113e 113e 40 52 52 52 52

TABLE V-Continued

W 00 t.3

PART7. Aspidospermatidine- and strychane-like compounds Mol. wt. of fragmentlcharge = m/e Compound

Reference Mf

Dihydroaspidospermatidine, CCV Dihydroaspidospermatine, CCX Tetrahydrocondylocarpine,C C X X Aspidospermatidine, CCIV N-Acetylaspidospermatidine, CCVIII N-Methylaspidospermatidine, CCVI Deacetylaspidospermatine, CCVII Aspidospermatine, CCIX Tetrahydrodecarbomethoxyakuammicine,CCIII Tetrahydroakuammicine, C C X X X V Tetrahydromossambine, C C X X X 2,16-Dihydroechitamidine 16-Methyl-16-decarbomethoxytetrahydroakuammicine, C C X X X V I 16-Methyl-16-decarbomethoxytetrahydromossambine, C C X X X I V Spermostrychnine, C C X I Deacetylspermostrychnine, C C X I I 2,16-Dihydrodecarbomethoxyakuammicine, CCII 2,16-Dihydroakuammicine,C C X X X V I I I 2,16-Dihydromossambine, C C X X I X

r

8

b3

268 340 326 266 308 280 296 338 268 326 342 342 282

138 138 196 136 136 136 136 136 138 196 212 212 152

227

130 160 130 130 130 144 160 160 130 130 130 130

298

168

215

368 326 266 324 340

166 166 136 194 210

271 229

227

199 199 215 199 199

U

Other peaks 28,21 28,21 118 28,21 28,21 28,21 28,21 28,21 28,21 122 122 126 I22 122

160 160 130 130 130

139 139

25129 25129 26729

21 21 28, 21 122 122

td

8 W

2

PART 8. Yohimbine- and ajmalicine-like compounds

Mol. wt. of frctgment/charge = m/e Compound

Reference Mf

3548,25,27

V

W

2

Y

Other peaks

170

169

184

156

165, 162

17.Deoxy-a-yohimbine, CCCX

170

169

184

156

165

Alloyohimbone, CCCXI

170

169

184

156

165

16-Ketoyohimbane, CCCXII

170

169

184

156

165

Yohimbinone, CCCXIII

170

169

184

156

165

Seredone, CCCXIV

230

229

244

216

165

200

199

214

186

162

N-Methylyohimbane, CCCXV

184

183

198

170

165

3-Deuterioyohimbine, CCCVII

171

170

185

157

165 165

Yohimbine (and stereoisomers)

11-Methoxyyohimbine, CCXCVIII

38427.25

Tetrahydroalstonine, CCCXVIII

35227

170

169

184

156

Ajmalicine, CCCXVII

35227

170

169

184

156

225

165

3,5,6-Trideuterioajmalicine, CCCIX

35530

173

172, 171

187

158

228

165

3,14-Dideuterioajmalicine,CCCVIII

35430

171

170

185

158, 157

171

Tetraphylline, CCCXIX

38227

200

199

214

186

165

Aricine, CCCIII

38227

200

199

214

186

Tetramethylenetetrahydr0-/3-~arboline

22627

170

169

184(w)

156

165 197

165

w

TABLE V-Continued

PART9. Sarpagine-like compounds Mol. wt. of fragment/charge = m/e Compound

Reference

M+ Bisdeoxyajmalol-B, CCCXXXIV 0-Methyl-17-deoxydihydrosarpagine, CCCXXXV Dih ydronormacusine - B Normacusine-B, CCCXX 0-Acetylnormacusine-B, CCCXXVIII Polyneuridine, CCCXXI 0-Acetylpolyneuridine, CCCXXV Akuammidine, CCCXXII 0,O-Diacetylakuammidinol, CCCXXVII 17,17-Dideuterio-CCCXXVII

cc

29420,27,32 328 2968.27 29419,20,27 33620,25,27 3528,19.20,25,27 3941%2 0 , 2 5 , 2 7 . 3 3 35219.20.25.27 40810.25.27.33 41010,25,27,33

dd31

W

237 253 223 249 249 249 249 249 249(w) 249(w)

Other peaks __~ _ _ _ _ _

bb

167 167

1823 1983 16915 169 169 169 169 169 169 169

168 1683 1683 1683 1683 1683 1683 1683

259,267 275,207 275,261 275 275 282, 275, 208 282, 275, 208

165 165 37 165 165a 165, 165a 165, 37 165, 37

PART10. Eburnamine-like compounds

Mol. wt. of fragment/charge = m/e Compound

ff

ee

99

278

249

208

193

279

250

209

194

Mf E burnamenine, CCCXLII I (eburnamine, isoeburnamine) 14-Deuterioeburnamenine, CCCXLIV

hh

____Other peaks

206

Reference

51, 170 51

W

E

W M

Apovincamine( 14-carbomethoxyeburnamenine) 11-Methoxyapovincamine Dihydroeburnamenine, CCCXLV 14-Deuteriodihydroeburnamenine, CCXLVI Eburnamonine, CCCXLIl 14-018-Eburnamonine Vincamine( 14-carbomethoxyisoeburnamine) 11-Methoxyvincamine

336

307

266

251

264

170

366 28027 281

337 251 252

296 210 211

281 195 196

294

170 51 51

224 226 284 22434 314 25434

209(w)

265 267 3258(w) 26534 3558(w) 29534

29427 296 3548.20.25.27 29434 3848,20,25.27 32434

237 237

81, 170

20934

23734

23934

26334

267, 252

51 170

297, 282

170

PART11. Oxindoles __ Mol. wt. of fragmentlcharge = m/e

Compound

Reference 22

kk

11

mm

nn

M

___

C

Yohimbine oxindole A Yohimbine oxindole B Mitraphylline, CCCXLIX 3-Deuteriomitraphylline, CCCL 3,5,6-Trideuteriomitraphylline, CCCLI

37089 25 37089 25 3688.19,25 36930 37 1 3 O

2258 2258 22320

3,14-Dideuteriomitraphylline, CCCLII Aricine oxindole, CCCLIII Carapanaubine, CCCLIV Rhyncophylline

37030 39883 1 9 42882 19 38419.20.32

22520 22320 22320 23919.20.32

22520

1303 1303 1303 1303 132, 131 1303 1603 1903 1303

14522 14522 14522

37 171 171

b

14522927

159 159 159 159 161

171 171

2C

17522 20522 14522

159 189 219 159

171 171 171 171 ~

Fr m

__

W 00

u1

w

TABLE V-Continued

00 5 2

PART12. Quebrachidine-like compounds Mol. wt. of fragmentlcharge = m/e Compound

Reference M+

Quebrachidine, CCCXXXVIII-D 0,N-Diacetylquebrachidine, CCCXXXVIII-E Vincamedine, CCCXXXII

352 43625 408

b 13035 13035 14435

PP

QP

rr

122 264 264

222 222

190 190 190

179b 179b 179b

m

PART13. Miscellaneous compounds

B

_- __

W M

Mol. wt. of fragmentlcharge = m/e Compound

3-Carbomethoxyharman, CCCXXXIX Tetrahydroharman

Reference

M+

Other peaks

240 186

182 171

180 165

1 Only principal peaks are recorded. Blanks in the table do not signify that a peak is missing but rather that it is not recorded in the literature. 2 In compounds bearing a n acyl substituent, RCO, on N,, the molecular ion is accompanied by peaks at M-RCO and M-RCO +H. Similar peaks may often be detected accompanying a. An RCO peak may usually be found in the lower mass range. 3 Accompanied by a peak 14 units higher. 4 Accompanied by peaks 1 ( c f H), 14 (cfCH2, weak) and 28 (c+ 2CH2, strong) units higher, except in CV whero c+COCHz appeers a t m/e 166. 5 Homologs show identical shift.

Homologs show no shift. Weak or negligible M-28 peaks due perhaps to inverted stereochemistry at ‘2-19. 8 Accompanied by peaks 17 or 18 units lower due to loss of OH or water. 9 Due to loss of CHzOH (31 units) from Mf, a,and c. 10 Accompanied by peaks 60 units lower due to loss of acetic acid. 11 These peaks are due to fragment c plus atoms C-4 and C-3 with their substituents (see footnote 4). 1 2 Represents the indole fragment b carrying also atoms C-3 and C-4 with the carbomethoxyl [C(C02Me)=CHz]. 13 These fragments come from the piperidine or D ring and are seen in the spectrum of vindolinine, which also exhibits the following principal peaks: Mf, M-CH3, M-OCH3, m/e 277, 249, 230, 229, 216, 170, 156, 134, 122, 121, 120(3). 14 Accompanied by M-43 (loss of COCH3, side chain). 15 Accompanied by peaks 1 unit lower and 12, 13, and 14 units higher. 16 Base peak. 1 7 M-85 = M-(2CH3CO)+H. 18 Loss of CHzO from i. 10 Accompanied by a peak 31 units lower (loss of CH30 or CH20H). 20 Accompanied by a peak 15 units lower (loss of CH3). 21 M-75 = M-(CH3CO+MeOH). 22 Accompanied by a peak 1 unit higher. 23 Presence of m/e 124 has been shown to be due to refractine. 24i+H. 25 Accompanied by a peak 59 units lower (loss of CH3COO or COz CH3). 26 Cleaves abnormally due to extra double bond or ketone group. 27 Accompanied by a peak 1 unit lower. 28 Accompanied by a peak 30 units lower. 29 This fragment is due to an ion in which (2-16 aRd its substituent (where present) have been expelled from the molecular ion (115a). 30 Accompanied by a peak 2 units lower. 31 Structure dd (Section VIII, B formulas) has been ascribed to this fragment. 32 Accompanied by a peak 29 units lower (loss of CzH5). 33 Accompanied by a peak 119 units lower, M - ~ C H ~ C O Z H H or M-(CH3C02H+ C02CHa). 34 These are eburnamonine or methoxyeburnamonine peaks resulting from the loss of H .C02Me from vincamine or 11-methoxyvincamine, respectively. 35 Accompanied by peak 13 units higher. 6

7

F b P

b

*U

P

.

+

w

00 l

388

B. GILBERT

TABLE VI pK DATA Alkaloid

PK',

Solvent

Reference ~~

Aspidospermine, I1 Deacetylaspidospermine, VI N,-Benzoyldeacet ylaspidospermine Aspidocarpine, XLIV Pyrifolidine, XLVI Deacetylpyrifolidine, XLVII Spegazzinidine, LXIX 0,O-Dimethylspegazzinidine, LXX 3-Dehydro-O,O-dimethylspegazzinidine, LXXI Cylindrocarpine, LXXV Vindoline, C I I I Dihydrovindoline, CXI 16-Methoxy-A~~-methyl-4-oxoaspidospermidine, CV Vindolinine, CVI Refractidine, CXX Deformylrefractidine, CXXII Deformyldemethylrefractidine, CXXVII Pyrifoline, CXIX Deacetyldemethylpyrifoline, CXXV 6-Dehydrodeacetyldemethylpyrifoline, CXXXIII Kopsinine, CXLII-A Aspidofractine, CXLIII-A Pleiocarpine, CXLV-A Pleiocarpinine, CXLVI-A Refractine, CL-A Deformylrefractinol, CXLIX-F Kopsiflorine, CLVI Kopsilongine, CLVII Kopsamine, CLVIII Kopsine, CLXX Dihydrokopsine-A, CLXXII Decarbomethoxykopsine, CLXXI Dihydrodecarbomethoxykopsine-A, CLXXIII Fruticosamine Fruticosine 0-Methyldepropionylaspidoalbine,CXCIl

50% aq. EtOH 50% aq. EtOH 50% aq. EtOH

7.30 7.36 6.98 6.55 6.85 7.8 2.9,6.4,10.7 2.76,6.45 3.0,5.48

66% 50% 33% 33% 33%

DMFl aq. EtOH DMF DMF DMF

25 25 25 30 10,49 30 36 36 36

5.9 5.5 5.9 5.35

66% 66% 66% 66%

DMF DMF DMF DMF

10,49 72 4 4

66% DMF

72 92 81, 37 37 37 49 81 81

3.3,7.1 6.26 6.25 7.15 2.9,7.5 6.40 7.25 5.90

80% MCS1

DMF DMF DMF DMF aq. DMF aq. DMF

33% 33% 33% 66%

7.50 6.85 6.19 6.94 6.64 3.2,8.0 6.38 6.80 6.58 4.28 6.1 5.4 6.1

70% MeOH ttq. DMF aq. MCS aq. MCS aq. DMF aq. DMF 70% MeOH 70% MeOH 70% MeOH 800/, MCS

4.04 4.62 8.5 ~

DMF =dimethylformamide; MCS = methyl cellosolve.

87 37 91 91 49 37 87 87 87 109 100 100 100

80% MCS

109

807; MCS

109

33% DMF

52

14. Aspidosperma

389

AND RELATED ALKALOIDS

TABLE VI-Continued

pK DATA Alkaloid

-

PK:

~-

Solvent ~

~-

Uleine, CCXLV Dihydrouleine, CCXLVI Olivacine, CCLVIII

8.23 8.87 6.05-6.13

800/, MCS 80% MCS 80% MCS

Ellipticine, CCLXXXIV 1,2-Dihydro?livacine, CCLXXIV 1,2-Dihydroellipticine, CCXCIII Cuatambuine, CCLX N-Methyltetrahydroellipticirle, CCLXXXV 11-Methoxyyohimbine, CCXCVIII Polyneuridine, CCCXXI Akusmmidine, CCCXXII

5.78 8.09 7.53 7.87 7.49

80% MCS 80% MCS 80% MCS 80% MCY 800,6 MCS

7.1 6.60 6.30

66% U M F

Reference ~ _ _ _ _ 138 138 140, 149, 150 140 150 155 139 139 161 165 62

the C-ethyl side chain is easily distinguished from this, since its absorption is centered at 0.65 6 and contains two principal peaks separated by 5 4 clsec. The absence of this absorption shows that the side chain is substituted on its terminal carbon atom and this may either indicate the presence of a C-21 oxygen function as in cylindrocarpine (LXXV, 10, 29) or limaspermine (LXXXVII, 31), or that the carbon atoms 20 and 21 are involved in a sixth ring as in the aspidofractinine (Section 111) and aspidoalbine (Section IV) groups. Other functions which may be recognized from the NMR-spectrum include a phenolic hydroxyl whose presence is also seen by the change in the UV-spectrum in an alkaline medium (Table 111). Phenolic hydroxyls usually occur in this series in position 17 where hydrogen bonding with the carbonyl group of the N,-acyl is possible, and in these cases the OH proton peak is found a t 10.7-11.2 8. When the hydroxyl group is not in position 17 it is not found so far downfield, for example, the 16-OH of spegazzinidine absorbs at 5.84 6. Another downfield singlet found in the spectra of some alkaloids derives from the proton of an N,-formyl group which appears around 9.5 6. The three-proton singlet due t o the methyl group of a carbomethoxy function appears slightly upfield with respect to the aromatic methoxyl absorptions, occurring at 3.55-3.7 6. Other, less general deductions which may be made from NMR-absorption will be dealt with under specific cases in later sections. The use of NMR-spectroscopy mainly to determine the nature of peripheral groups has as its ideal complement mass spectrometry which

390

B. GILBERT

gives direct information about skeletal structure. I n some cases complete elucidation. of a structure has been possible by determining the mass spectrum by use of less than 1 mg of material ; it has therefore become possible to investigate successfully the minor Aspidosperma bases which were present in quantities too small to permit even elementary analysis (28, 51, 51a). I n the mass spectrometer, the alkaloid molecule is split by electron impact and those fragments which are sufficiently stable and which carry a positive charge are collected in order of their molecular weights. Ions carrying a double positive charge appear a t mi2 where m is their molecular weight. The information which may be derived from the pattern of relative intensity plotted against m/e (e representing the charge on the ion) can be divided into three parts (28a). First, the highest molecular weight peak is usually that of the molecular ion2, that is, the whole molecule singly charged. This peak is referred to as Mi- and is accompanied by minor peaks one and two units higher corresponding t o molecular ions containing heavier isotopes of nitrogen, hydrogen, carbon, and other elements (20, 21). From the molecular weight thus precisely determined, the molecular formula is derived, leaving no doubt as t o the correct number of hydrogen atoms. Although this was already possible by integration of the NMR-spectrum, the amount of material required for the mass spectral method is minute in comparison. Second, it is possible to classify an unknown alkaloid into a structural class provided that other alkaloids with the same carbon skeleton are known. The cleavage of the molecule is independent of many superficial substituents. I n the aspidospermine group, these have been found to include a hydroxyl or methoxyl substituent on the aromatic ring (18, lo), an oxygen function on the ethyl side chain, whether a t C-20 (32) or a t C-21 (33, 5 2 ) , a carbomethoxy or other one-carbon group or a hydroxyl group a t C-3 (6, 7 , 35,36), and a carbonyl group a t C-4 ( 3 6 ,4 ) . [Note that a 3-ketone fragments differently (361.1 Thus the characteristic pattern of aspidospermine (11)may be recognized in all related alkaloids and derivatives containing the same skeleton, the only difference being that those peaks which correspond t o fragments containing an extra substituent are shifted to higher molecular weight by a number of units equal to the molecular weight of the substituent, while alkaloids that do not contain the aromatic methoxyl group of aspidospermine exhibit a pattern due to the aromatic fragments a t correspondingly lower molecular weight. 2

For an exception see ref. 28b where peaks higher than the molecular ion were observed.

14. Aspidosperma

391

AND RELATED ALKALOIDS

Third, and most important, direct structural information may be obtained. It is often possible, especially when the method is used in conjunction with NMR and with deuteration, to determine the exact position of substituents and also to establish the nature of new skeletons without in many cases having to make any extensive degradations. The importance of this will be seen particularly in the sections on aspidofractinine-, aspidoalbine-, and condylocarpine-type alkaloids (Sections 111, IV, and V), whose structures do not lend themselves to easy degradation by classical methods, owing to the presence of quaternary centers.

b m/e 160

c m/e

124

The breakdown of deacetylaspidospermine (VI) is represented by the formulas a-c (9, 28, 21). That of aspidospermine itself (11) is similar, except that the group R which is acetyl in the parent alkaloid is either acetyl or hydrogen in fragment a, both peaks appearing in the spectrum, while in b the indole nitrogen atom bears a hydrogen atom. The evidence for the structures assigned t o fragments a , b, and c is threefold. First, these fragments may be predicted from the known breakdown patterns of simpler molecules (20, 21). It is known, for example, that bonds /3 to

392

B. GILBERT

nitrogen such as C-19 to C-12 and C-2 t o C-3 in V I are readily broken since the carbonium ion or radical produced, at C-19 or C-2, respectively, in the case under consideration, can be stabilized by the donation of electrons from the lone pair on the nitrogen atom. A similar stabilization, by delocalization, favors the production of allylic, benzylic, or similar radicals and ions, rendering the allylic and benzylic bonds subject t o facile cleavage. Stable ions are favored, especially those in which there are conjugated double bonds or enhanced aromatic character. I n addition, there is a tendency t o cleave bonds at a highly substituted carbon atom. In many cases a number of bonds are broken simultaneously by a concerted movement of electrons within a ring, and in deacetylaspidospermine (VI) the movement indicated by the arrows in the formula represents one of the principal manners of cleavage observed. The fact that the calculated molecular weights of the fragments a, b, and c so predicted are in accordance with the principal observed peaks is adduced as evidence in favor of these structures. It will be noted that C-3 and C-4 are expelled as ethylene producing the a peak at M-28, and this M-28 a fragment together with the c fragment at mje 124 are the characteristics of the aspidospermine group alkaloids (Table V). [Strychanone (CCXLIV) also shows these two peaks, although it does not have the aspidospermine skeleton (36).]It should be emphasized that the breakdown illustrated by the formulas is not the only one to take place. There is a strong peak at c + 28 t o which a structure such as d may be assigned, the C-4 to C-5 bond having resisted cleavage. Weaker peaks appear at c + 14 (e) and b + 14 (f).

CCXLIV

\

CI

U

m/e 254

I

COzMe

J. c m/e 124

Second, evidence for the structures of the fragments responsible for the peaks observed has been obtained by the detection of metastable peaks (20). Such a peak proves that a certain fragment, say x of mass m,, was formed by decomposition of another fragment, y, of mass my, the

14. Aspidosperma

AND RELATED ALKALOIDS

393

position of the metastable peak, m*, being given by the approximate equation :

As an example may be cited the case of dihydrovincadifformine (CI) in which the decomposition of the molecular ion (m/e 340) to fragment a (m/e 254) with expulsion of the elements of acrylic ester is documented not only by the presence in the spectrum of those two peaks but also by the appearance of a metastable peak a t 191. The further breakdown of a to give c (m/e 124) was recognized by a metastable peak a t 61 (6). Similarly, in deacetylaspidospermine the loss of ethylene from the molecular ion (m/e 312 to m/e 284) is accompanied by a metastable peak a t m/e 259 (21). Third, as mentioned previously, peaks a, b, and c have been found to occur in the spectra of alkaloids and derivatives substituted in various positions. I n many cases the location of these substituents has been established by independent evidence, and the substituted molecule has been converted into the unsubstituted by chemical methods establishing the identity of the skeleton. The variation of molecular weight found for fragments such as a, b, and c in such a series can be used to decide with certainty some of the constituent atoms of those fragments. For example, fragment a contains all the atoms in the original molecule with the exception of C-3 and C-4 and any substituents that these atoms carried. This is shown by the fact that no increase in the mass of fragment a is observed in the mass spectra of molecules substituted in these positions. Thus, for example, 2,3-dihydrovincadifformine(CI) and dihydrovincadifforminol (XCIX) both show fragment a a t m/e 254 (6) in the same position in which it is found in the spectrum of aspidospermidine (XXXI), the parent alkaloid of the aspidospermine series. I n aspidospermidine this peak lies a t M-28 (loss of CHzCHz), whereas in CI i t corresponds to M-86 (loss of CHz=CHCOzCH3) and in XCIX to M-58 (loss of CHz=CHCHzOH). The position of the carboxylic ester group in vincadifformine (XCIII) and hence of the corresponding substituents in C I and XCIX is established by the UV-spectrum as C-3 (6), so that the lost atoms are necessarily C-3 and C-4. A similar result is encountered in the mass spectrum of spegazzinidine (LXIX) where the a peak is found a t m/e 328 or M-44 (loss of CHOHCHz) and independent evidence has established that the hydroxyl group which is lost is located on (2-3. Independent chemical proof has also been obtained for the aspidospermine-type skeletons of vincadifformine and spegazzinidine.

394

B. GILBERT

A t the same time, substituents in positions 17 and 16 [series aspidospermidine (XXXI), deacetylaspidospermine (VI), and deacetylpyrifolidine (XLVII) in Table V] ; position 1 [series aspidospermidine (XXXI), demethoxyaspidospermine (XXXIII), demethoxypalosine (XXXIV), and N-methylaspidospermidine (XXXV)]; position 20 [pair dihydrovincadifforminane (tetrahydrotabersonane, CII) and 20hydroxydihydrovincadifforminane(CXV)]; and position 2 1 [pair aspidospermine (11) and 2 1-acetoxyaspidospermine (LXXXV)] result in corresponding changes in the molecular weight of a, demonstrating that a incorporates these carbon atoms.

Rz ~~~

XXXI VI XLVII XXXIII XXXIV XXXV

R3

Ri H H Me0

H H

H

1L

Rz K3 H H Me0 H Me0 H H Ac H EtCO H Me

Ji

HO

Ac

OH

CI; R = COzMe LXIX XCIX; R = CHzOH XCIII; R = COzMe 2,3-douhle bond

CH3

CII;

R =H CXV; R = O H

Me0

Ac

11; R = H LXXXV; R = OAc

The same type of evidence supports the structures assigned to fragments b and c. Variations in the aromatic substitution produce in all cases the corresponding variation in the molecular weight of b while c is not affected by these changes. The hydrogen atom on C-2 is retained in fragment b , for when it is replaced by deuterium (1S), the b peak, found at m/e 160 in deacetylaspidospermine (VI), is shifted to m/e 161 (21). On the other hand, alterations in ring D or in the ethyl side chain are reflected in the molecular weight of c which is altered appropriately. For example, limaspermine (LXXXVII), which has a primary alcoholic group at C-21, exhibits peaks at m/e 140 (strongest peak in the spectrum), 122, and 109 which may be interpreted as due t o c (16 mass units higher than in the aspidospermine spectrum due to the C-21 oxygen), c-HzO, and c-CH20H (33). In contrast, the highly substituted dihydrovindoline

14. Aspidosperma

AND RELATED ALKALOIDS

395

(CXI) has a normal c peak at m/e 124 because none of the substituents is located on ring D or the ethyl side chain (4). The molecular weight of fragment d , the structure assigned (21) to the c + 28 peak which occurs at mje 152 in VI, is shifted to m/e 166 in the spectrum of the ketone (CV); whereas in its 3,3-dideuterio derivative, a further shift to m/e 168 is observed. This is strong evidence that d retains atoms 3 and 4, since the shifts are those calculated for the substitution of CO for CH2 in

cv position 4 and of CD2 for CH2 in position 3. The position of the carbonyl at C-4 is independently based on NMR-evidence. Similarly, in dihydrovindoline (CXI) the d peak is shifted t o mje 284, the shift of 132 mass units corresponding to the substitution of hydrogen atoms on C-3 and C-4 by COzCHz(+ 58), OH( + 16), and OAc( + 58). Returning to the spectrum of the ketone (CV), fragment b and its homolog fragment f,at mje 174 and 188 respectively, do not change their positions when the ketone is deuterated (excluding structures which contain C-3, for these fragments).

E. SOMEMINORALKALOIDS OF Aspidosperma quebrachoblanco AND

Rhazya stricta

In a reinvestigation of the minor alkaloids of A . quebrachoblanco,whose presence had already been indicated by Kesse (59) in 1882 (Vol. 11), Biemann et al. (28, 51a) were able t o isolate by a combination of alumina and gas chromatography about twenty compounds. The identification or structure determination of fifteen of these by mass spectrometry was described. Six belonged t o the aspidospermine group and four of these were the known compounds, aspidospermine (11), deacetylaspidospermine (VI), N,-methyldeacetylaspidospermine (XLI),and ( - )-pyrifolidine (XLVI). The three last-named had not previously been encountered in nature, V I and XLI having been prepared from aspidospermine (11)and vallesine (XXXVIII) (38, 39, 25). (-)-Pyrifolidine is identical with 0-methylaspidocarpine (XLVI) which has been prepared

396

B . GILBERT

from aspidocarpine (XLIV) (30). 0-Methylaspidocarpine (XLVI) has been shown (9) t o be the enantiomer of the alkaloid (+)-pyrifolidine which occurs in A . pyrifolium.

XXXI XXXV VI XLI

rr

(-)-XLVI

Ri H

Rz H

H H H H OMe

H OMe OMe OMe OMe

R3 H Me H Me Ae Ac

The two previously unknown compounds in the series were designated “Alkaloid 2 8 2 8 ” [later (51) named aspidospermidine] and “Alkaloid 296A” (or N-methylaspidospermidine), the numbers referring to their mass spectrometrically determined molecular weights. The first has also been found in Rhazya stricta (51). The mass spectrum of aspidospermidine differed from that of aspidospermine only in the shift of the mie values of all those peaks containing the benzene nucleus and N,, corresponding to the absence of substituents in these positions. Thus the &I+ and fragment a peaks appear 7 2 units lower down on the mass scale, corresponding to the absence of methoxyl (30) and acetyl (42) (see Table V). I n a similar manner, the spectrum of N-methylaspidospermidine parallels exactly that of N,-methyldeacetylaspidospermine (XLI), showing only a lowering of the mie values for the same two ions by 30 mass units that corresponds t o the absence of the nuclear methoxyl group. The structures of aspidospermidine and N,-methylaspidospermidme were thus shown t o be XXXI and XXXV, respectively, and aspidospermidine may be recognized as the parent of the aspidospermine group. The indolenine corresponding t o aspidospermidine was also isolated (28,51a).It was denominated “Alkaloid 280A” [later named 1,2-dehydroaspidospermidine (51)] and has structure IX. This compound had previously been obtained by the zinc dust distillation of quebrachamine (18, Section 11, B) and was subsequently found in R. stricta (51). It is

14. Aspidosperma

AND RELATED ALKALOIDS

397

noteworthy that the 21-methyl group of I X absorbs a t unusually high field in the NMR-spectrum, due t o shielding by the benzene ring not observed in compounds lacking the 1,2-double bond (51b). A compound with apparently the same structure (IX)has also been obtained both in the racemic and the levorotatory form ([aJI, - 225" in ethanol) by the degradation respectively of racemic vincadifformine (XCIII, 6, see Chapter 12) and (-)-tabersonine (XCII, 7 , Section 11, 0). As the levorotatory form of I X is converted by alkaline borohydride to ( + )-quebrachamine (I),it is probably the optical antipode of the product isolated from 8.stricta. The latter gives ( - )-quebrachamine and the configurationally related ( + )-aspidospermidhe (XXXI)on borohydride reduction (51b).

XCIII XCII 6,7-double bond 2N HC1

I

105"

IX

I

1

Hz cat

XXXI

The structure (IX) of 1,2-dehydroaspidospermidinerests on comparison of its UV- (Table 111)and mass spectra (characteristic peaks a t mje 280, 251, and 210) with that of l,%dehydroaspidosperniine ( V I l , Section 11,B). Lithium aluminum hydride reduction also gave aspidospermidine (XXXI, 28, 18, 51a, b). Nine other minor alkaloids of A . quebrachoblanco are dealt with in Sections V and VIII.

398

B. GILBERT

F. DEMETHOXYVALLESINE, DEMETHOXYASPIDOSPERMINE, AND DEMETHOXYPALOSINE The three alkaloids named in the title (XXXII, XXXIII, and XXXIV) are respectively the N,-formyl, -acetyl, and -propionyl derivatives of aspidospermidine (Section 11, E). Demethoxypalosine (XXXIV) has been isolated from Aspidosperma limae (40) and A . discolor (40a) and was characterized as an N,-acyldihydroindole by its UV-spectrum (Table 111)and IR-absorption a t 5.89 p. A strong band in the IR-spectrum a t 13.1 p indicated an unsubstituted benzene ring. The foregoing information was confirmed and the substance was shown to belong to the aspidospermine group by NMR- and mass spectrometry. I n the NMR-spectrum (Table IV) the 17-proton absorption is found a t 8.13 6 well downfield from the three-proton multiplet due t o the other aromatic protons which is centered a t 7.07 6. This shift is due t o the proximity of the carbonyl group of the N,-propionyl group. In the aliphatic part of the spectrum, absorptions which are characteristic of the

(X)/" H

o=c,

COR

R

a M-28

b m/e 130

XXXII; R = H XXXIII; R = Me XXXIV; R = Et

c m/e 124

aspidospermine skeleton (see Section 11, D) were found a t 4.08 6 (C-2 proton), 3.3-2.9 6 (CH2-Nb-CH2), 2.60-2.17 6 (quartet) and 1.401.08 6 (triplet, propionyl CH2, and CH3, respectively), and 0.75-0.45 (triplet, CH3 of a C-ethyl group). The mass spectrum (Table V) shows the molecular ion peak a t 338, confirming the molecular formula, C22H30N20 ; the other principal peaks are those characteristic of alkaloids of the aspidospermine group, that is, M-28 (m/e 310, fragment a ) , m/e 130 (fragment b accompanied by a homolog a t m/e 144), and mje 124 (fragment c, base peak of much higher intensity than any other and

14. Aspidosperma AND RELATED ALKALOIDS

399

accompanied by homologs at m/e 138 and 152). A strong peak at m/e 57 is due to the propionyl group on N,. From A . discolor has also been isolated the N,-acetyl analog, demethoxyaspidospermine (XXXIII, 40a). The alkaloid, which was separated chromatographically from the accompanying demethylaspidospermine (XXXVII, Section 11,G), was not obtained crystalline but was identified as demethoxyaspidospermine (XXXIII) by examination of its NMR- and mass spectra. The former resembled in all respects that of demethoxypalosine (XXXIV) with the exception of a threeproton singlet at 2 . 2 2 6 due to the N,-acetyl group which replaced the propionyl absorptions observed in the spectrum of XXXIV. The mass spectrum (Table V ) shows the molecular ion and M-28 (fragment a ) peaks 14 mass units lower than in the spectrum of XXXIV, while the b and c fragments and their homologs do not suffer this shift. Demethoxyvallesine (XXXII)has also been detected in the same plant (37).

G. DEMETHYLASPIDOSPERMINE The simplest of the phenolic members of the aspidospermine group, demethylaspidospermine (XXXVII),has been found in A . discolor (40a). It was isolated as its crystalline perchlorate ( 2 5 ) from which the oily free

XXXVI XXXVII XLII

Ri Rz H H H Ac

Ac Ac

VI

XXXVIII; R I1; R XXXIX; R XL: R

=H = Me = Et = n-Pr

base was obtained. From the NMR- and mass spectra (Tables I V and V), the compound was recognized as the known aspidospermine derivative (XXXVII) which had previously been obtained by demethylation of aspidospermine (11) with aluminum chloride in nitrobenzene ( 2 5 ) . Confirmation of its identity was obtained by methylation with dimethyl sulfate and potassium carbonate in dry acetone t o give aspidospermine (11).It is noteworthy that the phenolic hydroxyl group in position 17 is strongly hydrogen-bonded to the carboayl group of the N,-acetyl (OH

400

B . GILBERT

shows NMR-singlet a t 10.83 6) and XXXVII does not suffer ready air oxidation as do phenolic bases in this series which lack the ilr-acyl group, e.g., aspidosine (XXXVI). As a result of the hydrogen bonding, the foregoing relatively vigorous conditions are necessary for inethylation but in spite of this the strong steric hindrance of N, (1 1, 12, 27) enables the direct isolation of the tertiary base aspidospermine, with little or no methylation of Nb having occurred (40a).

H. VALLESINE AND PALOSINE Vallesine (XXXVIII) and palosine (XXXIX) are respectively the N,-formyl and N,-propionyl analogs of aspidospermine (11).Vallesine which occurs in Vallesia glabra and V . dichotoma (Volume I1 and Ref. 41) has been related to aspidospermine (39, 2 5 ) and has therefore structure XXXVIII. Palosine (XXXIX) has been isolated from Aspidosperma polyneuron (23), where it occurs in admixture with aspidospermine and a ketonic base (43) in addition to other alkaloids (Table I). The separation of palosine from the mixture was very difficult but was finally achieved by chromatography on paper using formamide containing 10% of pure formic acid and 8% of ammonium formate as a buffered stationary phase with benzene: chloroform (4: 1) saturated with formamide as the mobile phase. Acid hydrolysis of palosine gave deacetylaspidospermine (VI) and propionic acid (43),the latter being identified by paper chromatography (44). Propionylation of deacetylaspidospermine gave palosine to which, therefore, the structure XXXIX may be given.

AND DEMETHYLASPIDOCARPINE I . ASPIDOCARPINE

Aspidocarpine (XLIV) was isolated by direct crystallization from the light petroleum extract of Aspidosperma megalocarpon (30), and was characterized as the perchlorate, hydrochloride, hydrobromide, and hydriodide. Its empirical formula is CZ2H30N203 and it contains one methoxyl group. Aspidocarpine represents an important chemical link between aspidospermine (11)and its relatives with one oxygen substituent in the benzene nucleus, on the one hand, and the related alkaloids with two oxygen substituents on the other, since it has been related directly with aspidospermine. The UV-spectrum of aspidocarpine, which is somewhat similar to that of aspidospermine, exhibits a large bathochromic shift on passing from

14. Aspidosperma

AND RELATED ALKALOIDS

40 1

neutral to alkaline solution (Table III),indicative of the presence of a phenolic hydroxyl group in the molecule. A carbonyl peak in the IRspectrum at 1632 em-1 indicated an amide group, and acid hydrolysis followed by steam distillation gave acetic acid, showing that this amide was present as an N-acetyl group. The low frequency of the amide carbonyl absorption and the absence of a hydroxyl peak in the IRspectrum, coupled with the olive-green ferric reaction of aspidocarpine and the fact that the phenolic hydroxyl group resists methylation with diazomethane, were indicative of strong hydrogen bonding between the O H and C=O groups. Evidence confirming these functional groups was obtained by acetylation to 0-acetylaspidocarpine (XLV) and methylation with dimethylsulfate in alkali (note that Nb resists methylation) to 0-methylaspidocarpine (XLVI) ([.Iu - 94" in chloroform, later shown t o be antipodal to ( + )-pyrifolidine, ["ID + 90" in chloroform; Refs. 9, 10) and in both XLV and XLVI the amide carbonyl absorption appeared a t higher frequencies (1666 and 1656 em-1, respectively) indicating that hydrogen bonding no longer occurred. 0-Methylaspidocarpine hydrolyzed with 10% hydrochloric acid gave 0-methyldeacetylaspidocarpine(XLVII) ( [aID- 4.9" antipodal to ( + )-deacetylpyrifolidine, [a]=+ 7" ; Refs. 9, lo), which could be reacetylated to XLVI. On these grounds, it was reasonable to propose the partial structure XLIII. The NMR-spectrum of aspidocarpine (XLIV) (30, 45) fully confirmed the foregoing, a hydrogen-bonded phenolic proton being found a t 10.83 6 and the N-acetyl methyl singlet a t 2.20 6. In addition, the methoxyl group exhibited a three-proton singlet at 3.73 6, and that it is an aromatic methoxyl is demonstrated by the presence of absorption due t o only two aromatic protons. These form a triplet with a large central peak a t 6.5 6, the small side peaks being separated by J = 9 cjsec which shows that these two protons are almost equivalent and are disposed ortho to one another. Most importantly, the aliphatic part of the spectrum closely resembles that of aspidospermine (II),showing the C-2 proton, CHz-N-CHz and C-ethyl absorptions (Table IV). The foregoing information really suffices t o place the methoxyl in position 16, but chemical evidence was also adduced to show that it was ortho to the phenolic hydroxyl. Demethylation of aspidocarpine (XLIV) with hydrobromic acid gave the unstable catechol XLVIII (ferric reaction, blood-red) which was acetylated to the triacetate (XLIX) in which the two 0-acetate absorptions in the IR-spectrum (1777 and 1767 em-1) showed them to be phenolic acetates. Alkaline hydrolysis of this triacetate gave demethylaspidocarpine (L), which was shown to be a catechol by the ferric reaction (blue-green passing t o deep-red in the

XLIII

THB*

XLVIII

XLIV

Ri

__ LII Me LIII Me X L V I I Me XLVI Me XLV Me LIV Me LV Me

It2

It3

H H

H EtCO H

Me

Me

Ac Ac EtCO

Ac Ac EtCO EtCO

14. Aspidosperma

AND RELATED ALKALOIDS

403

presence of sodium carbonate), and by the strong bathochromic shift of the UV-spectrum in the presence of boric acid buffered with sodium acetate (30, 46). Aspidocarpine derivatives were found to couple with diazotized sulfanilic acid when either the C-17 OH or N, was unsubstituted but not when they were both substituted; this was used as evidence that positions 14 and 15, para t o these groupings, were unsubstituted. Finally, proof was obtained that the aliphatic portion of the aspidocarpine molecule was identical with the correspondingportion of aspidospermine by oxidation of the two compounds with chromic acid in dilute sulfuric acid. In both cases, the aromatic ring was destroyed and there was isolated the a-keto amide (LI), mp 224"-225", [ C L ]-~ 132" (chloroform). The identity of this oxidation product shows not only that aspidocarpine (XLIV) and aspidospermine (11) have an identical skeleton but also that they have the same absolute configuration at all four centers. Aspidocarpine has also been found in other Aspidosperma species (Table I) while demethylaspidocarpine (L) has been isolated from A . album (42).

J. ASPIDOLIMINE Aspidolimine (LIII) is one of the main constituents of the bark of A . limae, where it occurs together with a number pf other alkaloids (Sections11, F, I and IV, B ;Refs. 31,40). It is also foundin A. triternatum (47). The IR- and UV-spectra are practically identical with those of aspidocarpine (XLIV), although elementary analysis and the mass spectrum (molecular ion peak at m/e 384) show it to contain one CH2 group more than that alkaloid. Examination of the NMR-spectrum (31) shows that there is an N,-propionyl group present (N-COCHzCH3 quartet at 2.57 6, J = 7 c/sec; N-COCHZCH~ triplet at 1.25 6 , J = 7 cjsec) while the N,-acetyl three-proton singlet shown by aspidocarpine (Section 11,I) is absent. It was reasonable to suppose that aspidolimine (LIII) was the N,-propionyl analog of aspidocarpine, and this was readily shown t o be true both chemically (31) and by examination of the mass spectrum (40). Vigorous acid hydrolysis of the alkaloid gave depropionylaspidolimine (LII), which without isolation was acetylated to give 0-acetylaspidocarpine (XLV). Similar hydrolysis of aspidocarpine (XLIV) and propionylation of the product gave 0-propionylaspidolimine (LV). The mass spectrum of aspidolimine shows all of the characteristic peaks of the aspidospermine group alkaloids (Section 11, D, Table V). In addition to M+ and M-28, peaks are found at M-56 (loss

404

B. GILBERT

of the propionyl group and pick-up of one hydrogen) and at M-84 [M-(56+28)]. The strongest peak of the spectrum is found at m/e 124 (fragment c ) ,and the peaks due to fragment b and homologs are found at m/e values 16 mass units higher than in the spectrum of aspidospermine, corresponding to the extra oxygen atom in the aromatic ring.

K. PYRIFOLIDINE AND DEACETYLPYRIFOLIDINE ( - )-Pyrifolidine (XLVI) has already been mentioned as a constituent of A. quebrachoblanco (Section 11,E) and as the product of methylation of aspidocarpine (Section 11, I). Dextrorotatory pyrifolidine and its ( + )-deacetyl derivative (XLVII) occur in Aspidosperma pyrifolium (49,56).At the time of isolation, the levorotatory compound (XLVI)was not known. The UV-spectrum (Table 111) is somewhat different from those of known N-acyldihydroindoles. The NMR-spectrum, however, not only confirmed the presence of two methoxyl groups and an N-acetyl function already detected by functional group analysis, but showed that the compound belonged to the aspidospermine group (9, 10). The C-2 proton quartet at 4.50 6, CHzN-CHz multiplet at 3.31-2.90 6, and the side chain terminal methyl absorption at 0.67 6 were all placed exactly as found in the spectrum of aspidospermine. Elementary analysis and the integrated proton count showed that the empirical formula was C23H32N203, differing from aspidospermine by CH20, equivalent to the addition of the extra methoxyl group known t o be present. The presence of only two aromatic protons occupying adjacent positions on the benzene ring was shown by a two-proton quartet with J = 8.2 c/sec, and this shows that both methoxyl groups are aromatic and must be in one of the three orientations, LVIII, LIX, or LX. Biogenetically, LVIII and LIX with a methoxyl at C-17 were attractive. Of these, LVIII with the other methoxyl at C-14 was the less likely, as this group would be expected t o alter the NMR aliphatic “fingerprint” region which is in fact almost superimposable upon that of aspidospermine (11). Conclusive evidence that the aliphatic portion of the pyrifolidine molecule was in fact identical with that of aspidospermine (11)was obtained from the mass spectra of their deacetyl derivatives (XLVII and VI, Table V). These proved to be identical, save only in the shift of 30 units observed in the peaks attributable to fragments containing the aromatic nucleus, this shift being due to the extra aromatic methoxyl group. Finally, direct comparison of ( + )-pyrifolidine with ( - )-@methylaspidocarpine (XLVI) and of their respective deacetyl derivatives

14. Aspidosperma

AND RELATED ALKALOIDS

405

(XLVII) showed complete identity by IR-spectra and by chromatography, while the rotatory dispersion curves of ( + )-XLVI and ( - )XLVI were mirror images of one another, demonstrating that the two compounds were enantiomeric at all four centers. Pyrifolidine is thus one of a small group of indole alkaloids which exist in both enantiomeric

+ )-XLVII;R = H (+)-XLVI; R = AC (

VI; R = H 11; H. = AC

forms ; others are quebrachamine (I),vincadifformine (XCIII, 6, 74), akuammicine (CCXXV, 7 7 ) , guatambuine (CCLX, 147), and vincanorine (CCCXLII, 78). It is noteworthy that these alkaloids either contain only one asymmetric center or could have been formed from a precursor with only one center by a concerted reaction in which the configuration of the remaining centers might depend on that of the original.

AND SPEGAZZINIDINE L. SPEGAZZININE

Evidence presented earlier (Volume VII, p. 132) led to the proposal of partial structure LXI (34) for spegazzinine and the suggestion was made that the molecule had the aspidospermine skeleton. This has subsequently been shown to be correct, but the evidence was obtained by a study of the related alkaloid spegazzinidine (LXIX). Both alkaloids occur in Aspidosperma chalensis, the proportion of each present having been found to vary with different specimens of the plant (35, 36). Spegazzinidine (LXIX) was shown by elementary analysis to be CzlHzsNz04, the number of hydrogens being confirmed by the mass spectral molecular weight determination. Its UV-spectrum resembles

406

B. GILBERT

that of spegazzinine (LXIV) and is of the aspidospermine type (Table 111).A bathochromic shift in alkaline solution similar to that shown by spegazzinine showed that it was also a phenolic base, and a carbonyl peak a t 6.13 p indicated the presence of the usual N-acyl function. The NMRspectrum of spegazzinidine (LXIX) confirmed the expected similarity to aspidospermine (11),exhibiting the characteristic CHz-N-CHz pattern, the N,-COCH3 methyl singlet, and the side chain terminal methyl absorption at 0.75 6. I n the aromatic region, a quartet due to two ortho protons, similar t o that of pyrifolidine (Section 11,K) was observed, while absorptions due t o two phenolic hydroxyl protons appeared, one a sharp singlet at 11.1 6 characteristic of a hydrogen-bonded hydroxyl in position 17, the other a broader band a t 5.84 6 due to a non-hydrogenbonded OH. Methylation of spegazzinidine (LXIX) with dimethyl sulfate (cf. aspidocarpine, Section II, I ; note that the alcoholic hydroxyl as well as N,, are unaffected) gave spegazzinidine dimethyl ether (LXX) whose NMR-spectrum now closely resembled that of pyrifolidine (XLVI). One striking difference existed between the NMR-spectra of spegazzinidine and its methyl ether on the one hand and aspidospermine and pyrifolidine on the other. This lay in the absorption due t o the C-2 proton which in the latter two alkaloids appears as a quartet, while with spegazzinidine and its methyl ether only a doublet is observed. This indicated that there was only one hydrogen atom a t C-3 and, as spegazzinidine dimethyl ether (LXX) still showed hydroxyl absorption in the IR-spectrum, it was reasonable to suppose that C-3 carried an alcoholic hydroxyl group. To confirm this, LXX was oxidized with chromium trioxide (50) to the ketone, 3-dehydrospegazzinidine (LXXI),in whose NMR-spectrum the doublet previously observed had disappeared and been replaced by a singlet a t 5.11 6. This observation is only consistent with a secondary alcoholic group having been present at C-3,and the further downfield shift of the C-2 proton peak in LXXI is caused by the adjacent carbonyl group (36). Independent evidence for the position of the carbonyl group in this ketone was obtained by exhaustive deuteration with sodium deuteroxide in DzO-CH~OD, when mass spectrometry showed that six deuterium atoms had entered the molecule (molecular ion raised from m/e 398 t o m/e 404). Three of these were in the acetyl group, since the m/e 4 3 peak (CH&O+) was shifted t o m/e 46 (CD&O+), and therefore three had entered a to the ketonic carbonyl (LXXII ; note that LXXIII is a by-product of the deuteration reaction). Only two positions in an aspidospermine skeleton have three a-hydrogen atoms-positions 3 and 20. The latter position is excluded by the NMRspectrum, and the carbonyl group is therefore at (2-3. Detailed study of the mass spectra of LXXI and its hexadeuterio derivative confirmed

14. Aspidosperma

AND RELATED ALKALOIDS

407

these findings (35, 36), although it should be noted that the 3-ketone does not, in contrast t o a 4-ketone (36, 4), fragment in the manner characteristic of aspidospermine (11) and its derivatives. Rather, it loses first the carbonyl group as CO, a behavior observed in other cyclic ketones (20, 21). The foregoing arguments have been based on the assumption that spegazzinidine has in fact the same skeleton as aspidospermine (11).That this is true was established in two ways. First, in the mass spectrum of spegazzinidine (LXIX) and of its dimethyl ether (LXX) the principal peaks are observed a t m/e values exactly coinciding with the calculated molecular weights of the fragments a, b, and c and their homologs that would be expected for a molecule with the aspidospermine skeleton but carrying two hydroxyl or, respectively methoxyl groups on the aromatic ring (Table V). It will be noted that the C-3 and C-4 atoms are not present in any of the aforementioned fragments. The nonappearance of the alcoholic oxygen, whose position has been proved, in these fragments has been cited (Section 11, D) as evidence that the C-3 and C-4 atoms are those initially lost in the breakdown of aspidospermine-type molecules under electron bombardment. This breakdown, which normally results in the loss of ethylene and the formation of the M-28 peak common to the spectra of most alkaloids of this group, results in the loss of vinyl alcohol, CHOH=CHz, or its equivalent, and the formation of an M-44 peak in the spectra of LXIX and LXX, which is raised t o M-45 when the alcoholic proton is replaced by deuterium. Second, direct chemical proof of the aspidospermine skeleton was obtained by conversion of spegazzinidine dimethyl ether (LXX) to its tosylate (LXXIV) which on reduction with lithium aluminum hydride gave ( - )-N,-deacetyl-N,-ethylpyrifolidine(LVII) identical, except for direction of rotation, with the product, ( + )-N,-deacetyl-N,-ethylpyrifolidine (LVII),obtained by a similar reduction of ( + )-pyrifolidine (XLVI). Since ( + )-pyrifolidine is antipodal to aspidospermine (Section 11, K), spegazzinidine has the same absolute configuration as aspidospermine (36). Finally, the coupling constant (J = 8 cjsec) observed €or the C-2 proton doublet in the NMR-spectrum of spegazzinidine shows that the C-2 and C-3 hydrogens are situated trans-diaxial with respect to one another. The C-3 hydroxyl group is therefore a-equatorial. A comparison of the NMR- and mass spectra of spegazzinine (LXIV) and it methyl ether (LXV) with those of spegazzinidine (LXIX) and its dimethyl ether (LXX)showed that the parent alkaloids differed only by the absence of the C-16 hydroxyl group in spegazzinine and of the C-16 methoxyl group in its methyl ether. In particular, the 5.84 6 signal due

!+

0 00

LXI

LXII LXIII LXIV LXV LXVI LXVII LXVIII

H H H Ac Me Ac Me Ac Me Ac Ac Ac

H Me

H

H H H Ac PhCO Ac

,/ RO LXIX

/

Ac a (M-44)

from LXIX; R = H fromLXX; R = Me

from LXIX; R = H from1,XX; R = Me

14. Aspidosperma AND RELATED ALKALOIDS

409

410

B. GILBERT

to the non-hydrogen-bonded phenolic grouping in LXIX is absent from the NMR-spectrum of spegazzinine, which shows absorption due to three aromatic protons, while in its mass spectrum the fragments a and b and relatives containing the aromatic ring appear a t m/e values 16 mass units lower (i.e., less one oxygen atom) than in the spectrum of spegazzinidine (LXIX). Similarly, a corresponding lowering of 30 mass units (CHzO) is observed when the spectra of spegazzinine methyl ether (LXV) and spegazzinidine dimethyl ether (LXX) are compared, all other peaks appearing at identical m/e values. I n accordance with previous experience (Section 11, D), therefore, spegazzinine may be allocated structure LXIV.

M. CYLINDROCARPINEA N D CYLINDROCARPIDINE Cylindrocarpine (LXXV) and cylindrocarpidine (LXXVI) occur in the trunk bark of Aspidosperma cylindrocarpon, and the former was first isolated by way of its crystalline perchlorate (49). The IR-spectrum of cylindrocarpine showed the presence of an ester (Aco, 5.79 p ) as well as an amide (A,,,, 6.06 p ) grouping, while the UV-spectrum showed the presence of a chromophore different from that of aspidospermine (11) and its relatives (Table 111).The integrated NMR-spectrum supported the empirical formula, C ~ O H ~ ~ Nand ZO confirmed ~, the presence of two methoxyl groups detected by functional group analysis. Preliminary chemical studies elucidated many features of the structure, leaving only the exact nature of the aliphatic portion of the molecule unknown. I n particular, hydrogenation gave dihydrocylindrocarpine (LXXVII) whose UV-spectrum was now superimposable on that of aspidospermine (11),showing that the molecule contained the N-acyl- 17-methoxydihydroindole nucleus. Alkaline hydrolysis of both cylindrocarpine (LXXV) and dihydrocylindrocarpine (LXXVII) cleaved only the ester grouping and gave the corresponding acids, cylindrocarpic and dihydrocylindrocarpic acid, as their hydrochlorides (LXXVIII and LXXIX, respectively). The latter was reesterified with diazomethane regenerating dihydrocylindrocarpine (LXXVII), proving that cylindrocarpine contains a carbomethoxyl group, and accounting for the second methoxyl group. I n order t o elucidate the nature of the AT,,-acylgrouping, both cylindrocarpine (LXXV) and its dihydro dcrivative (LXXVII) were hydrolyzcd with hydrochloric acid to cleave the amide as well as the ester grouping. I n both cases, cylindrocarpinic acid dihydrochloride (LXXX) was obtained, and from the nonbasic fraction were isolated,

14. Aspidosperma

AND RELATED ALKALOIDS

41 1

respectively, cinnamic and dihydrocinnamic acids. The unusual UVspectrum of cylindrocarpine is therefore simply a superimposition of the N-acyldihydroindole and cinnamoyl chromophores. The UV-spectra of cylindrocarpinic acid (LXXX, free amino acid) and deacetylaspidospermine (VI) in neutral and acidic media are practically identical, showing the characteristic reduction of wavelength and intensity on protonation of N, (Table Ill).This is in accordance with the presence of identical chromophores in both compounds. The preceding experiments cast no light on the nature of the aliphatic carbon skeleton, but examination of the NMR-spectrum showed that this was similar t o that of aspidospermine (11).In particular, the characteristic absorption due to the C-2 proton was found at 4.4S and those due to the CHz-N-CHz protons at 3.31-2.90 6, while in the absence of an N-acetyl peak the singlet due t o the isolated C-19 hydrogen atom is plainly visible. No side chain terminal methyl group absorption was observed and, assuming an aspidospermine skeleton, the carbomethoxy group could be provisionally placed in that position ((3-21). I n the aromatic portion of the spectrum, not only were the expected eight aromatic protons detected, but also two doublets with J = 16 cjsec due to the trans-cinnamic vinyl protons. It remained t o convert cylindrocarpine (LXXV) t o a known aspidospermine derivative. This was achieved by elimination of the carbomethoxy and cinnamoyl groups in the following manner. Cylindrocarpine (LXXV) was reduced with lithium aluminum hydride to a mixture of the y-phenylpropyl compound (LXXXI.) and dihydrocylindrocarpol (LXXXII). The latter, separated by chromatography, was oxidized with chromic acid in acetic acid t o the corresponding aldehyde, dihydrocylindrocarpal (LXXXIII),which by Wolff-Kishner reduction gave the known deacetylaspidospermine (VI).Cylindrocarpine can therefore be given the structure LXXV. Cylindrocarpidine, C23H30N204, the minor alkaloid of A . cylindrocarpon, was suspected from its NMR-spectrum (Table IV) t o be the N,-acetyl analog of cylindrocarpine (LXXV) since the absorption pattern was almost identical, except in that no peaks were seen corresponding to the cinnamoyl group, these being replaced by a threeproton singlet at 2.22 6 due t o an N-acetyl group. Confirmation was obtained by acid hydrolysis which gave cylindrocarpinic acid dihydrochloride (LXXX) and acetic acid (44),and the structure of cylindrocarpidine is therefore represented by the expression LXXVI (10, 29). The LXXX obtained had rotation identical with that from cylindrocarpine (LXXV) and consequently these two alkaloids have the same absolute configuration. That this is the same as that of aspidospermine

412 B . GILBERT

14. Aspidospermu AND RELATED ALKALOIDS

T

413

4 14

B. GILBERT

(11)was shown by the similarity in sign and shape of the ORD-curves of cylindrocarpidine and aspidospermine ( 10). Although the structures of these two alkaloids were proved without recourse to mass spectral measurements, spectra have subsequently been determined for the derivatives N-decinnamoylcylindrocarpol (LXXXIV), its O,N,-diacetyl derivative (LXXXV), and N-decinnamoyl-N-ethylcylindrocarpol (LXXXVI). All three show the expected pattern with peaks corresponding to a, b, and c fragments and their homologs as observed for other aspidospermine-type alkaloids (Section 11,D). It should be noted, however, that due to the extra oxygen atom in LXXXIV and LXXXVI and the acetoxy group in LXXXV the c peak appears a t m/e 140 in the first two cases and a t m/e 182 in the last instead of a t m/e 124 as in aspidospermine (52).

N. LIMASPERMINE AND RELATED ALKALOIDS Limaspermine, C~zH30N203,occurs in A . lirnae and was separated from the accompanying aspidolimine (LIII) and aspidocarpine (XLIV) by silica gel chromatography ( 33; see Section 11,J).I t s red-violet ceric sulfate reaction after heating was typical of an N-acylindoline and this was confirmed by the UV-spectrum (Table 111)which resembled that of demethylaspidospermine (XXXVII, 2 5 ) , showing the same bathochromic shift in alkaline solution (40a) which is attributable to the presence of a phenolic hydroxyl group. The presence of an amide group was confirmed by IR-absorption a t 1631 em-1, the low frequency of this peak indicating chelation with a hydroxyl group, which itself absorbed as a broad band, and it was therefore reasonable t o place the phenolic OH a t position 17. The nature of the amide group was elucidated by acid hydrolysis which gave propionic acid, and the third oxygen atom was accounted for by acetylation which gave 0,O-diacetyllimaspermine (LXXXVIII), which absorbs a t 1764 em-1 (phenolic acetate) and 1734 em-1 (alkyl acetate), thus indicating the presence of an alcoholic as well as a phenolic hydroxyl in limaspermine (LXXXVII). The foregoing information suggested a 17-hydroxy-N-propionylindoline structure, and this was fully confirmed by the NMR-spectrum (Table I V ) which showed the presence of three aromatic protons and a hydrogen-bonded phenolic hydroxyl group (singlet at 10.58 6). More important, the portion of the spectrum due to aliphatic protons was very similar to that of aspidolimine (LIII), except that no peaks were found that corresponded to a methoxyl group or a terminal methyl of an ethyl side chain. Other characteristic features of the aspidospermine (11)

14. Aspidosperma

415

AND RELATED ALKALOIDS

skeleton (29) such as the C-2 quartet a t 4.12 6 and the CH2--Nb-CH2 multiplet at about 3.0 6 were present, while the quartet and triplet due respectively to the methylene and methyl groups of N-propionyl were in the same positions as for aspidolimine (LIII). The absence of the side chain C-methyl and the presence of a triplet at 3.25 6 clear of other absorption, which is shifted downfield to 4.0 6 (overlapping the C-2 proton quartet) in the spectrum of the diacetate (LXXXVIII), show that the alcoholic hydroxyl group is primary and suggest that it is in

.

LXXXVII

LXXXVIII LXXXIX

I

xc

XCI

U

b

HH Me0 Me0

C

Ac H H

H

.

EtCO

Ac

Ac

H H H

Ac EtCO

c

position 2 1 of an aspidospermine-type molecule. The structure LXXXVII thus deduced for limaspermine is fully confirmed by its mass spectrum (Table V). Peaks are found corresponding t o the usual a, b, and c fragments, that due to c being, as is normal, the strongest peak of the spectrum. Its position (m/e 140) indicates that it contains the alcoholic oxygen atom which therefore must be in position 21. The a , or M-28, peak (m/e 342) is relatively weak, but is accompanied by a much stronger peak at m/e 324 due to loss of water, and by other smaller peaks a t m/e 267 (further loss of the propionyl group) and m/e 311 (a- CH2OH). Peaks due to loss of water and of CHzOH are also found accompanying c a t m/e 122 and 109, respectively, while a peak at m/e 110 is attributed

416

B. GILBERT

to loss of CH20 by a cyclic mechanism (c to c') (33, 54). Among the minor bases of A . limae there have also been isolated three other C-21 alcohols, the 16-methoxy derivative of limaspermine (XCI, 33), 1'imapodine (N-depropionyl-N-acetyllimaspermine,LXXXIX), and its 16-methoxy derivative, XC (120). The fact that other 17-hydroxy-N-acylindolinesof the ( - )-aspidospermine series have positive rotations of the order of 100" indicates that limaspermine ([ E]D + 108" in chloroform) and XCI ([ .IU + 118" in chloroform) also belong to this configurational group (54a). This has been confirmed by interrelation with haplocine (CCI-D) and ( - )-palosine (XXXIX, Section IV, D, 113c),and with cylindrocarpine (LXXV, 120).

0. TABERSONINE Tabersonine (XCII), a n amorphous base which occurs in the seeds of Amsonia tabernaemontana (65, Volume VII, p. 134), in some Xtemmadenia species (66,66a),and in Tabernaemontana alba (sea),is a member of a distinct group of aspidospermine-type alkaloids which bear an extra carbon atom in the form of a carbomethoxyl group linked t o position 3. The other members of this group are described briefly in Section 11, P and in detail in Chapter 12. The UV-spectrum (64, Table 111) and high negative rotation ([ED] - 310" in methanol) of tabersonine hydrochloride (64, 65) indicated at once that the alkaloid contained the same chromophore as akuammicine (CCXXV) and echitamidine (CCXLII, Chapter 8 ) , and this chromophore was subsequently shown (Volume VII, p. 127) to be Ph-NH-C=C-C02R. The NH (2.95 p ) and a$-unsaturated carboxylic ester bands (6.02 p and 6.2 p ) were also seen in the IR-spectrum. The NH group was not acetylated under usual conditions and the ester group was musually resistant t o hydrolysis (64). Examination of the mass spectrum (Table V, 7) showed that the correct molecuJar formula was C21H24N202, excluding a C20 akuammicine-like skeletm, and peaks a t m/e 92, 107, and 135 were similar to those encountered in vindolinine (CVI, 3) which has an aspidosperminelike skeleton with 6,7-doublebond. The presence of such a nonconjugated double bond had already been shown by catalytic hydrogenation t o 6,7-dihydrotabersonine (XCIII, an optically active form of vincadifformine which has identical IR-, NMR-, and mass spectra) in which the UV-spectrum remained similar to that of the parent alkaloid (64, 7 ) . Moreover, the NMR-spectrum of tabersonine (XCII) showed a peak due to two vinyl protons which disappeared in the dihydro derivative

14. Aspidosperma

417

AND EELATED ALKALOIDS

xcv

IX-A

T

4s HCI 110”

T

KBHI OH-

418

B. GILBERT

(XCIII), so that the double bond is disubstituted. The NMR-spectrum also showed the presence of four aromatic protons (consistent with a strong IR-band a t 13.5 p ) , and an NH and a C-ethyl group. The presence of a very strong mje 124 peak (fragment c ) in the mass spectrum of 6,7dihydrotabersonine (XCIII) indicated that the C-ethyl group probably formed part of an ethylpiperidine ring in an aspidospermine-type framework (7). The aspidospermine-type skeleton was then confirmed in two ways. First, tabersonine (XCII) was converted to ( + )-quebrachamine (I)by two independent routes. One involved reduction of the 6,7-double bond followed by strong acid hydrolysis, which removes the carbomethoxyl group, completely shifting the double bond t o the 1,2-position, a reaction known for the other vinylogous amides, akuamniicine (CCXXV, 70) and fluorocurarine (67). The resulting indolenine (IX-A) is the optically active form of decarboniethoxyvincadifformine (superimposable IRand mass spectra, see Chapter 12) and is possibly the antipode of 1,2dehydroaspidospermidine (IX). The former (IX-A) was converted by alkaline borohydride t o ( + )-quebrachamine, an example of the retrograde Mannich reaction and reduction already known in the akuammicine series (19, Section 11, A, and Chapter 7). The same series of reactions, carried out in a different order, gave successively decarbomethoxytabersonine (XCIV), 6,7-dehydro-(+ )-quebrachamine (XCV), and ( + )-quebrachamine (I). These conversions establish the stereochemistry of C-5 in (-)-tabersonine (XCII) as opposite t o that of ( - )-aspidospermine (11). A second proof of the skeleton of tabersonine as represented by XCII was obtained by examination of the mass spectra of the four reduction products XCVI, XCVII, XCVIII, and XCIX. I n all these compounds, the 2,3-double bond has been reduced, thus permitting the normal aspidospermine-type fragmentation in the mass spectrometer. The reduction of this double bond without affecting the isolated 6,7-double bond may be effected either with zinc and acid (cf. 68) which gives 2,3-dihydrotabersonine (XCVI), or with lithium aluminum hydride which gives chiefly 2,3-dihydrotabersonol (XCVII) (64) and its deoxy derivative [XCVIII, 7, contrast the reduction of akuammicine which gives the 3-methylene derivative (68, 69), a difference which, together with NMR-spectral changes, may be due to conformational or even stereochemical differences from aspidospermine]. Compounds XCVI and XCVII now showed typical indoline UV-spectra (64, 7, Table 111)and the N,O-diacetate (C) of 2,3-dihydrotabersonoI showed a typical AT,acylindoline spectrum (64). The mass spectra showed the expected peaks a t m/e 252 (a‘) and 122 (c’, accompanied by m/e 121 ; compare

14. Aspidosperma

AND RELATED ALKALOIDS

419

vindolinine, Ref. 3) in which the reduction of two mass units as compared with the spectrum of aspidospermidine (m/e 254 and 124,Ref. 28) results from the presence of the 6,7-double bond. I n all three cases the C-3 substituent had been expelled together with C-3 and (3-4. Catalytic reduction of 2,3-dihydrotabersonol (XCVII) gave the fully reduced tetrahydrotabersonol (XCIX, an optically active form of dihydrovincadifforminol) which showed mass spectral peaks a t m/e 254 and 124 corresponding to the expected fragments a and c. It remained only to locate the nonconjugated double bond. This must be in the D ring, for it appears in fragment c’ (above) and, as it is not

COzMe C V I ; R = COzMe CVI-A; R = H, 6,7-dihydro

CIII; R = Me CIV; R = CHO CXIII; deacetyl-CIII

attacked by potassium borohydride (XCIV to XCV) or by zinc and sulfuric acid (XCII to XCVI), it cannot be adjacent to nitrogen ; positions 6,7are thus the only ones left for a disubstituted double bond.

P. Vinca ALKALOIDS OF

THE

ASPIDOSPERMINE GROUP

A number of alkaloids with the aspidospermine skeleton occur in the genus Vinca and are dealt with in detail in Chapter 12.They include the very important base vindoline (CIII) which not only occurs as the free base in Vinca rosea ( = Lochnera rosea = Catharanthus roseus) but also as part of the dimeric alkaloids vinblastine (vincaleucoblastine) and leurosine, whereas the N,-formyl analog, CIV, forins part of the dimeric alkaloid leurocristine ( 5 , 72a). These dimeric alkaloids have been used successfully for the treatment of certain forms of cancer in man (5). Vindolinine (CVI) has also been isolated from V . rosea and its dihydrodecarbomethoxp derivative, tuboxenin (CVI-A), which is the parent member of the series, occurs in a Pleiocarpn species (53). ( f )-Vineadifformine (XCIII, 6, 74)has already been mentioned as the racemic form of ( - )-6,7-dihydrotabersonine(Section 11, 0). It has been found in V . difformis and in Rhazya stricta (51b)where the ( + ) form also occurs.

420

B. GILBERT

The ( k ) and ( - ) forms have been found in V . minor, where it is accompanied by four other alkaloids of similar skeleton, minovine (CVIII, 74a), minovincinine (CVII), minovincine (CIX), and 16-methoxyminovincine (CX) ( 3 2 , 7 3 , 74a, 76). The mass spectra of some of these alkaloids and of their derivatives CI, CII, and CXIV-CXVI have been included in Table V where useful for purposes of comparison ( 7 5 ) .

Ri Rz XCIII H H CVII H O H C V I I I Me H

Ri Rz CI H COzMe CII CH3 H CXIV H CHzOH CXV H CH3 C X V I OMe CHzOH

CIX; R = H CX; R = O M e

R3 H H OH OH OH

111. The Aspidofractinine Group

A. INTRODUCTION This group of mainly hexacyclic alkaloids is derived from the aspidospermine skeleton by closure of C-21 with C-2, the original ethyl side chain thus forming a new six-membered ring. Aspidofractinine has been chosen as the parent member as it bears no substituent in the aliphatic portion of the molecule. The oldest known member, kopsine, isolated in the last century, differs from the others in being heptacyclic, C-3 being linked to (2-11 by a one-carbon bridge. The extraordinary cage structure of this group of alkaloids with quaternary centers a t C-2, C-5, and C-12

14. Aspidosperma

42 I

AND RELATED ALKALOIDS

is not susceptible to ordinary chemical degradative methods, and the present evidence for the skeleton therefore rests mainly on physical evidence. The group is a t present limited to four genera of the family Apocynaceae, namely, Aspidosperma, Hunteria, Kopsia, and Pleiocarpa.

B. INTERCORRELATIONSAND SKELETAL STRUCTURE I n the case of the aspidospermine group (Section 11),the nature of the carbon skeleton was elucidated for aspidospermine only, the structures of the ‘other members of the group following by correlation with this alkaloid, either directly or mass spectrometrically. Although mass spectrometry alone suggested the structure CXVII for the skeleton of some Aspidosperma alkaloids, the final proof of CXVII rests on both

CXVII

CXIX

cxx

4 stages

CXXXVI

CXXXIX

physical and chemical evidence obtained in simultaneous work in a number of laboratories on diverse alkaloids. This evidence is brought together in the sequel. The validity of such evidence for alkaloids other than the one under study depends on intercorrelation and this has been largely achieved. Pyrifoline (CXIX) has been related t o refractidine (CXX) by the mass spectral comparison of the alkaloids and a series of derivatives, and to aspidofiline (CXXXVI) by mutual conversion to a common intermediate (CXXXIX). Aspidofractine (CXLIII-A), which has been mass spectrally related to refractine (CL-A),is readily converted

422

B. GILBERT

H6

COzMe CL-A

CLXXXIV

CLXX

i

3 stages

&zMe CXLIII-A

R

COzMe CXLII-A LiAlHd

COzMe CXLVI-A

CXLII-D +

CH~OH CXLII-F

~ H ~ O H CXLVI-F

.

I

4 stages

COzMe CXLV-A

14. Aspidosperma

AND RELATED ALKALOIDS

42 3

to its deformyl derivative kopsinine (CXLII-A) and the corresponding alcohol, kopsinyl alcohol (CXLII-F).These two compounds have in turn been related chemically to kopsinilam (CXLII-D), pleiocarpine (CXLV-A), and kopsine (CLXX). Pleiocarpine (CXLV-A) has further been interrelated with pleiocarpinine (CXLVI-A) and pleiocarpinilam (CXLVI-D) through the common intermediate N-methylkopsinyl alcohol (CXLVI-F). It remains to convert a member of the kopsinine group to a member of the aspidofiline group unsubstituted on C-3. 1. Number of Rings

The hexacyclic nature of the alkaloids, except kopsine, follows from accumulated analytical data, together with precise molecular weight determination by mass spectrometry, which indicate that apart from the dihydroindole moiety there are four other rings or double bonds. The absence of vinyl protons in the NMR-spectra of the bases and the fact that vigorous hydrogenation reduced only the benzene ring showed that only a hindered tetrasubstituted double bond could be present. Such a grouping was excluded by later evidence which demonstrated the presence of three quaternary centers ; therefore, the aliphatic part of the alkaloids contains four rings and the molecules are hexacyclic. Similar methods showed that kopsine is heptacyclic. 2 . General Evidence f r o m Mass Spectrometry

The mass spectra of many alkaloids and their derivatives in this group have been measured and the principal peaks are recorded in Table V (80, 81,82,95, 102). It is readily seen that a common skeleton is probable for all except kopsine and relatives, since the main differences observed are due only to peripheral substituents (RI-R~in structure CXVIII). A general similarity of the spectra to those of aspidospermine (11) and its derivatives is seen in the universal presence of an M-28 peak and usually of a peak a t m/e 124 which is modified only in the two alkaloids pyrifoline (CXIX) and refractidine (CXX) which bear a D-ring substituent, but which is otherwise invariable no matter what modifications are made in the indolic part of the molecule or to the C-3 carbomethoxyl group. The principal difference from the aspidospermine-type cleavage is seen in the fact that the main piperidine-containing peak is usually not c but a peak 15 mass units smaller, that is, normally a t m/e 109. Coupled with the NMR-spectral evidence that these alkaloids contain no G-ethyl side chain and no C-2 proton (absence of characteristic absorption a t 0.65-0.70 6 and 4.0-4.5 a), it is reasonable to postulate an aspidospermine skeleton in which the side chain is tied into a sixth ring terminating a t C-2, and t o propose the breakdown illustrated in the formulas, in which the cleavages

42 4

B. GILBERT

involved are favored ones (Section 11, D). Evidence has been obtained for the correctness of the structures allocated t o the various fragments h-p in a manner similar to that already described in Section 11, D. Metastable peaks have been observed corresponding, for example, to the decompositions, h t o i (at mje 41.5 for CXXXV and m/e 45.2 for CXLII-0) and M+ to c (at m/e 52.5 for CXLII-0). Fragment h always arises by loss of the unsubstituted bridge, C-21, C-20, in the cases where C-3 carries a substituent, and therefore it always appears at M-28. That h contains C-3 is shown by its molecular weight, which accompanies variations in the R3 group, and deuteration on (2-3. Fragment i contains carbon atoms 10,6 and 7 , for when these are substituted the position o f i in the mass spectrum moves appropriately. Thus, in CXVII and CXXXIX, i appears at mje 109, but in 3',3', 10,iO-tetradeuterio-N-trideuteriomethylkopsinyl alcohol (CXLVIII-G from LiAlD4 reduction of pleiocarpine lactam A) i is shifted to in/e 111 and that this shift is due to the C-10 deuterium atoms is shown by the invariability of i when R1, Rz, and R3 are modified in other molecules. When position 6 is substituted by OH, OMe, or =0, corresponding shifts are observed in the mass of i (series N,-rnethyl-(i-deformyl-6-demethylrefractidine,CXXXI; deformylrefractidine, CXXII; and deformyl-6-dehydro-6-demethylrefractidine, CXXXV; see Table 77) and deuteration of C-7 results in a shift of two units in the position of i (see pair, deacetyl-6-dehydro-6demethylpyrifoline, CXXXIII, and its 1,7,7-trideuterio derivative, CXXXIV in Table V). Evidence for the position of these substituents will be described in the sequel. These same carbon atoms are also found in fragment c which accompanies i in all the above changes. That the ethyl chain of c contains C-20 and C-21 and not C-4 and C-3 follows from its constant value, no matter what modifications are made to the substituent on C-3. The extra hydrogen atom that fragment c carries on C-21 does, however, come from C-3 as is shown by its increase by one mass unit when the C-3 hydrogen atom is replaced by deuterium (pair, N,-methyldeforniylnorrefractinol, CLI-Q and N,-dideuteriomethyldeformyl-3-deuterionorrefractinol, CLII-Q, obtained by the LiAlH4 and LiAID4 reduction, respectively, of the corresponding ketone). The transfer of hydrogen to C-21 by the mechanism pictured in the formula is only possible when the C-3 proton is and close t o C-21, as is the case in the naturally occurring alkaloids of the kopsinine-refractine type and their derivatives produced without epimerization. In the 3-is0 series of epimerized alkaloids and their derivatives, this proton is c( and its distance from C-21 precludes its transfer, in accordance with which is the absence of fragment c from their spectra. (Small c peaks sometimes observed have been attributed to contamination with nonepirnerized material, and to

14. Aspidosperma AND

425

RELATED ALKALOIDS

m

Ri

Rz R3

RI

R4 ~

CXVII CXXII CXXXI CXXXIII CXXXIV

cxxxv

CXXXVI CXXXIX

H H H H H OMe H Me H O H OMe H H =O OMe D H =0, i, Dz H H H = O OH Ac H H OMe H H H

H H

CXLII-F CXLII-G CXLII-0 CXLV-D CXLVI-D CXLVIII-G CLI-Q CLII-Q

M P

H H H H

H H OMe OMe

Rz H H H COzMe CH3 CDJ CDzH

R3

CHzOH CDzOH Me COzhlc CO&Ie CDiOH H,OH D,OH =CHz

=O

R4 H H H H , 10, =O H, 10, =O H, 10,Dz H H H H

426

B. GILBERT

check this, epimerization was conducted with NaOD in deuteriomethaiiol SO that the epimerized C-3 hydrogen would be replaced by deuterium, leaving the unepimerized C-3hydrogen unchanged. I n accord with expectation, no trace of deuterium transfer was observed.) The proximity which is necessary for this transfer is also illustrated by the lower relative intensity of the c peak in C-3 unsubstituted alkaloids such as aspidofiline (CXXXVI), refractidine (CXX), and pyrifoline (CXIX) where there is no sterically hindered bulky C-3 substituent to push C - 3 over toward the 21-20 bridge. Transfer of a hydrogen atom t o fragment i is also often observed, in molecules substituted on C-3, to produce a peakj a t m/e 110. I n cases where the C-3 substituent atom, C-3' or 0, carries an a-hydrogen atom, this is the proton transferred, as is shown by replacement by deuterium when j moves to m/e 111 (pair, kopsinyl alcohol, CXLII-F, and 3',3'-dideuteriokopsinyl alcohol, CXLII-G, by reduction of kopsinine with LiAlD4; see Table V). When the C-3 substituent does not carry an a-hydrogen atom, the mje 110 peak is due to the presence of natural isotopes in i , while in molecules having a double bond a t '2-3, for example, M and P, the main fragmentation paths are blocked and the m/e 110 peak cannot be ascribed a structure. A further peak a t m/e 81 which sometimes accompanies i results from its further decomposition (reversed Diels-Alder), as is established by a metastable peak a t m/e 60.5 (102). I n many spectra of alkaloids of this group, the indolic peaks are small as compared with the aspidospermine-type molecules. I n some cases they starid out, however, and peaks corresponding to fragments m and n occur in the spectra of O-methyldeacetylaspidofiline (CXXXIX)and deformylrefractidine (CXXII), while the lactams, pleiocarpinilam (CXLVI-D) and pleiocarpine lactam A (CXLV-D),exhibit intense peaks due to the fragments 1 and p (102). On the basis of the foregoing information, it has been possible to determine the complete structures of alkaloids in this group by use of reactions designed only to modify superficial substituents (80, 81, 82, 95). Other possible skeletal structures which might be compatible with the observed spectra are excluded by chemical evidence described subsequently. 3. N , I s Located in a Pive-Membered Ring ( E ) This is shown by the IR-absorption of the three lactams, pleiocarpine lactam A (CXLV-D), pleiocarpinilam (CXLVI-D), and kopsinilam (CXLII-D; Section 111, H), which is compatible only with their being y-lactams [compare, for example, 8- and 10-oxoaspidospermine (XX, XI), Section 11, C].

14. Aspidosperma

AND RELATED ALKALOIDS

427

4. NbI s Also Located in a Six-Membered Ring ( D ) Similar lactams, for example, refractalam (8-0x0-CXXII,Section Ill,

D), pleiocarpine lactam B (CXLV-E) and the related kopsinine lactam (CXLII-E,Section Ill, G)show IR-absorption due to an amide carbonyl group in a six-membered ring. Also, zinc dust distillation of pleiocarpine (CXLV-A, 95) and kopsinine (CXLII-A, 101) gives 3,5-diethylpyridine, an indication of structural similarity to aspidospermine which also gives this product derived from the piperidine ring D containing Nb. The ketones deacetyl-6-dehydrodemethylpyrifoline (CXXXIII) and deformyl-6-dehydrodemethylrefractidine(CXXXV, Section 111, D) are located in a six-membered ring as shown by their IR-absorption. Conclusive evidence has been obtained that the carbonyl groups in these two compounds are located in the same ring as Nb. 5. C-3 Lies in a Six-Membered Ring

The carbon atom, C-3, which in the kopsinine subgroup bears a carbomethoxyl substituent, has been shown by the mass spectral evidence summarized above not to be in the six-membered ring which contains Nb. That it lies in a strained six-membered ring is indicated by degradation of the carbomethoxyl groups of refractine (CL-A) and aspidofractine (CXLIII-A) to C-3 ketones (CXLIX-P and CXLII-P), whose IRabsorption occurs a t unusually low wavelengths ( 5 . 7 G 5 . 7 8 p) (Section 111,G). 6. Ring E Contains a Chain of T w o Methylene Qrouys Terminating in a

Quaternary Center, C-12, and the Tertiary Nb The vast majority of indole alkaloids contain a tryptamine unit in which Nb is linked to the P-position of the indole nucleus by an ethylene chain. On biogenetic grounds and also from the mass spectral similarity with aspidospermine (11),it is reasonable t o expect this feature in the aspidofractinine-type alkaloids. Furthermore, in kopsine lactam A (CLXXV), in which (2-11 is substituted by the C-3’ bridge and C-10 has been oxidized (five-membered lactam), the residual hydrogen atom on C-11 shows as a singlet (2.82 6) in the NMR-spectrum and C-12 is therefore quaternary. 7. Ring D Contains a Chain of Three Carbon Atoms between Nband (7-5

Evidence was presented under subsection 4 above tha.t ring D is six-membered and contains the ketonic carbonyl group of deacetyl6-dehydrodemethylpyrifoline (CXXXIII). When this ketone was exhaustively deuterated, only two deuterium atoms were introduced, which is consistent with the carbonyl’s lying between a quaternary

428

B. GILBERT

center, C-5, and a methylene group, C-7. Moreover, when the ketone was reduced under Clemmensen conditions, an olefin (CXLI) was obtained that contained a 6,7-double bond whose resistance to reduction by zinc and acid shows that this compound is not an enamine and there must be another carbon atom (C-8) between the double bond and Nb (Section 111,D, E). The preparation of the two lactams CXLII-E and CXLV-E shows that C-8 is a methylene group (Section 111,Q). 8. Ring F Contains a Pair of Carbon Atoms, C-3 and C-4, between Two Quaternary Centers, C-2 and C-5 As already mentioned, the NMR-spectra of the aspidofractinine-type alkaloids show that they do not have any hydrogen atom on C-2. The adjacent atom in ring F, C-3, bears a carbomethoxyl group in, for example, kopsinine (CXLII-A), the position of this group being established by the formation of a six-membered cyclic urethane (CLIV) involving N, from the derived alcohol, kopsinyl alcohol (CXLII-I?) (Section 111,G). The related N-methylkopsinyl alcohol (CXLVI-F) was converted by way of alkaline or thermal decomposition of its mesylata (CXLVI-I), to the olefins, N-methylkopsinylene (CXLVI-M) and N-methylisokopsinylene (CXLVI-N). N-Methylkopsinylene has an exocyclic methylene group whose vinyl protons show a 12-line ABXz pattern in the NMR-spectrum demonstrating the presence of two protons on C-4. I n isokopsinylene, the double bond has moved into the ring and the remaining C-4 proton, now absorbing in the vinyl region, clear of other absorption, shows weak coupling (J = 1.5 cjsec) only with the allylic C-3' methyl group. C-4 is therefme attached to a quaternary center (Section 111,G). 9. Positions 2 and 5 Are Linked by a n Unsubstituted Ethylene Chain

Only two carbon atoms have not been accounted for in the aliphatic portion of the molecule and these are clearly those expelled as ethylene in the formation of the ion h (M-28)in the mass spectrometer. In order to satisfy the conditions that C-2 and C-5 must be quaternary and that the C-3 ketone, e.g., CXLIX-P, is six-membered, these two carbon atoms must link positions 2 and 5 as in aspidospermine. 10. The Relation between C-3 and Ring El and the Stereochemistry of the Alkaloids The evidence already presented encompasses all the aliphatic carbon atoms of the alkaloids, the dihydroindole structure of rings A and B resting on UV-, NMR-, and mass spectral data as well as on the production of indole derivatives during the zinc dust distillation and alkali

14. Aspidosperma AND RELATED ALKALOIDS

429

fusion of pleiocarpine and kopsine (95, 103). The skeleton CXVII represents the only manner of linking the groupings described that is compatible with all the data mentioned. The demonstration of a C-3 to C-11 bridge through only one carbon atom in kopsine (CLXX) proves that the bonds 2-3 and 12-11 are disposed cis with respect to rings B and C in this alkaloid. From models, it is evident that ring E must be cis fused to ring C and the relative stereochemistry of C-19 is thus also established. If it is assumed that no inversion at C-12 and C-19 takes place during the pyrolysis of kopsinyl iodide (CXLII-J) to kopsane (CLXXXIV) which has been directly related to kopsine (Section 111, J),then the relative stereochemistry of all the ester alkaloids is established. The relative stereochemistry of the nonester alkaloids follows from the cis E/C fusion mentioned above. The correlation of the two groups with one another and with aspidospermine (11)has yet to be achieved.

C. ASPIDOFRACTININE From the residues remaining after the removal of the major alkaloids, refractidine, refractine, and aspidofractine (Sections 111, D, G) from Aspidosperma refractum, VPC and thin-layer chromatography enabled the separation and purification of a noncrystalline minor base, aspidofractinine, ClgH24N2. The IR-spectrum showed an NH band a t 3.10 p and the mass spectrum proved to be identical with that of O-methyldeacetylaspidofiline (CXXXIX, Section 111,E) with the sole exception that those peaks which correspond to indole-containing fragments were found at mass numbers 30 units lower. Aspidofractinine is thus the unsubstituted parent (CXVII) of the hexacyclic group which forms the subject of this section (102).

CXVII

D. PYRIFOLINE, REFRACTIDINE, AND REFRACTALAM Pyrifoline (CXIX)and refractidine (CXX) are abundant constituents, respectively, of the closely similar species, Aspidosperrna pyrifoliurn (49) and A. refracturn (81). Pyrifoline, shown by analysis and mass spectrometry to be C23H30N203, contains two methoxyl groups, while refractidine, CzlHzsN202, contains only one. The UV-spectrum of pyrifoline

430

B. GILBERT

is coincident with that of aspidospermine, and taken together with the fact that there is absorption due to three aromatic protons in the NMRspectrum, this shows that a 17-methoxy-N-acyldihydroindolechromophore is present. Refractidine showed UV-absorption characteristic of an N-acyldihydroindole unsubstituted in the aromatic ring (Table 111), confirmed by the appearance of absorption due to four aromatic protons in the NMR-spectrum. The NMR-spectra (Table IV) and acid hydrolysis of the two alkaloids to their deacyl derivatives, CXXI and CXXII,

Ri CXIX CXXI CXX CXXII CXXIII CXSlV

Olle

4c

OXe

H CHO H Me CD2H

H H

H H

CSXV CXXVI CXXVII CXXX CXXXI CXXXII

Rz

OMe H OMe H H H H Ac H Me H CDzH

R3

H D H

H

CXXXIII CXXXIV CXXXV

RI R2 Me0 H OMe D H

H

H

H

Ri

AC

CXXVIII; R = OMe CXXIX; R = H

respectively, showed that pyrifoline has an N,-acetyl and refractidine an N,-formyl group. The difference in molecular formula of the two alkaloids is thus explained by the 17- and N,-substituents. Functional group analysis and NMR-spectra show the absence of a C-ethyl side chain and a (2-2 hydrogen atom (absence of absorption a t 0.65-0.70 6 and 4.0-4.5 a), excluding the two alkaloids from the aspidospermine group, and the lack of evidence for a double bond indicates a hexacyclic structure (81). The mass spectra of the two bases and all their corresponding derivatives are identical except for appropriate shifts in the masses of the molecular ion and fragments h, m,and n,due t o the different substitution

14. Aspidosperma

AND RELATED ALKALOIDS

43 1

pattern in the indolic part of the molecule (Table V). Pyrifoline and refractidine therefore have the same aliphatic structure ( 2 l ) , and the evidence that this structure is based on the skeleton CXVII has already been presented (Section 111, B). It remained to locate the aliphatic methoxyl group present in both alkaloids. This group is selectively demethylated by boiling concentrated hydrochloric acid to give, respectively, the alcohols CXXV and CXXVII in which the c and i peaks are found 14 mass units lower than in the parent alkaloids (loss of CHZ). The secondary nature and equatorial configuration of the alcohol group was shown by acetylation to give CXXVIII and CXXIX, respectively, in whose NMR-spectra a clear one-proton quartet was observed a t 4.63 6 and 4.73 6, respectively (J = 5, 10.5 cjsec) due to the CH(0Ac) group flanked by two nonequivalent protons. Confirmation of the presence of only two vicinal protons was obtained by Oppenauer oxidation of the alcohol (CXXV) with cyclohexanone and aluminum phenoxide to the six-membered ketone (CXXXIII, IR, 5.89 p), which was equilibrated with excess sodium deuteroxide in deuteriomethanol to give the trisdeuterio derivative, CXXXIV, (accompanied by material not deuterated on Na). The carbonyl group thus has only two cr-hydrogen atoms and the mass spectra of the ketone and its deuterated derivative (Table V) showed that both the carbonyl group and the adjacent hydrogen, or, respectively, deuterium, atoms were retained in the c and i fragments. On the basis of skeleton CXVII and the evidence presented in Section 111, B, 2 for the structure of these fragments, only position 6 for the carbonyl group is compatible with these observations. Among other derivatives of the two alkaloids prepared in order to confirm the correctness of the interpretation placed upon the mass spectra, were the corresponding ketone in the refractidine series (CXXXV, I R , 5.9 p ) , the 6-deuterio derivative of deacetyl-6-demethylpyrifoline(CXXVI), and the compounds CXXIII, CXXIV, CXXXI, and CXXXII. I n all cases the mass spectral peaks were found a t the predicted m/e values (Table V). The methoxyl group of pyrifoline (CXIX) and refractidine (CXX) is therefore located securely in position 6. Of some interest is the ease of hydrolysis of the aliphatic methoxyl groups of these alkaloids which, in common with the sharp drop in pKi (33% dimethylformamide) which is observed on passing from the alcohol, CXXV (7.25), to the ketone, CXXXIII (5.90), would seem t o point to an interaction with Nb. Examination of models does not show proximity between these two positions, however. A similar change in basicity is seen in the pairs, ajmalidine (ketone, pKi 6.3) and sandwicine (alcohol, pKi 8.5) (98, 99), and kopsine (ketone, pKi 4.28) and dihydrokopsine A (alcohol, pKi 6.1) (109, 100). The two ketones, like CXXXIII,

432

B. GILBERT

are P-keto-amines in which the keto group is held rigidly away from the basic nitrogen atom. I n addition, the six-membered lactam 8-0x0-CXXII has been encountered in A . refracturn (37).

E. ASPIDOFILINE Aspidofiline (CXXXVI), one of the simplest members of the group, was the first alkaloid t o be isolated from A . pyrifoliurn (79). It was readily separated from the accompanying pyrifoline by virtue of its alkali solubility. The phenolic hydroxyl group so indicated is confirmed by the bathochromic shift of the UV-spectrum observed on addition of alkali. The spectrum is characteristic of a 17-hydroxy-N-acyldihydroindoleand the IR-spectrum shows a hydrogen-bonded amide carbonyl band a t

CXXXVI CXXXVII CXXXVIII CXXXIX CXL

Ri

Rz

H

Ac Ac Ac

Ac Me Me

H

CXLI

cxxxIII

H Et

6.14 p. Further confirmation is obtained from the NMR-spectrum which shows a singlet a t 10.13 6 due to the bonded 17-hydroxyl (compare demethoxyaspidospermine and aspidocarpine), while a three-proton singlet at 2.30 6 shows the N,-acyl group to be acetyl. Analysis and the mass spectrometrically determined molecular weight established the empirical formula as C21H~sN202.The oxygen atoms are already accounted for, so it remained to elucidate the nature of the hexacyclic carbon-nitrogen skeleton. Aspidofiline yields an 0-acetate (N,-tertiary) and is methylated with diazomethane in 24 days to 0-methylaspidofiline (CXXXVIII). The resistance to methylation of similarly hydrogenbonded phenolic hydroxyl groups has been encountered already with aspidocarpine and demethylaspidospermine (Section 11,G, I)and in those

14. Aspidosperma

AND RELATED ALKALOIDS

433

cases methylation was effected with dimethyl sulfate. This method was also used with aspidofiline, but as in this group N,-methylation is rapid, the O,N,-dimethyl quaternary sulfate resulted, from which CXXXVIII was regenerated by pyrolysis of the corresponding methohydroxide in high vacuum. Acid hydrolysis of the methyl ether yielded 0-methyldeacetylaspidofiline (CXXXIX). Mass spectral examination of aspidofiline and its three derivatives, CXXXVII, CXXXVIII, and CXXXIX, showed in each case an M-28 peak (h) and peaks a t m/e 124 and 109 (c and i ) . This similarity to other alkaloids of the group, considered together with the absence of NMR-absorption due to a C-ethyl side chain or a C-2 hydrogen atom, led t o the proposal of structure CXXXVI (see Section 111,R , 2). Proof that the proposed structure was correct was obtained by correlation with pyrifoline (CXIX).The ketone, CXXXIII (Section 111, D), was subjected to the Clemmensen reduction which unexpectedly gave the olefin, CXLI (i peak a t m/e 107) and th’is was hydrogenated t o 0-methyldeacetylaspidofiline (CXXXIX) (80).

F. SOMEALKALOIDS OF Aspidosperma populifoliurn Ten alkaloids have been isolated from A . populifolium, of which the greater part belong t o the aspidofractinine group (48). They include ( + )-0-methyldeacetylaspidofiline [CXXXIX, optical enantiomer of that derived from aspidofiline (Section 111,E)], its 16-methoxy analog, CXLI-A, and their respective N,-formyl derivatives, CXLI-B and CXLI-C. The structures of all four compounds could be established by comparison of their IR-, UV-, NMR-, and especially mass spectra with

?,I

‘*’ I

CXXXIX H CXLI-A Me0

H H

those of known aspidofiline derivatives and in the case of CXXXIX, by chromatographic comparison with its enantiomer. The UV-spectra of the dimet,hoxy compounds CXLI-A and CXLI-C, were closely similar to those of deacetylpyrifolidine (XLVII) and pyrifolidine (XLVI), respectively, while the presence of two ortho-hydrogens on the benzene

434

B. GILBERT

ring was established by NMR-spectroscopy (6.28 and 6.86 6 , doublets, J = 9 cjsec). On this basis, the methoxyl groups of these bases could be located in positions 16 and 17 (48).

G. KOPSININE, ASPIDOFRACTINE, PLEIOCARPINE, PLEIOCARPININE, AND REFRACTINE The five alkaloids mentioned in the title differ only in the substituent on N, or, in the case of refractine, by the presence of a methoxyl group a t C-17. It is convenient, therefore, t o consider their chemistry together. The natural sources of the alkaloids are recorded in Table I, and interconversions that have related them were described in Section 111,B. The bases are all levorotatory. Analyses of the bases and their salts and particularly mass spectral molecular weight determination established the empirical formulas, and examination of the UV-spectra and comparison with model compounds showed that all were dihydroindoles, kopsinine (CXLII-A)having a free N,-H, and aspidofractine (CXLIII-A),refractine (CL-A), and pleiocarpine (CXLV-A)being acylated. I n addition to the UV-, the IR- (N,-CHO, 6.0 p ; N,-COZMe, 5.87 p ) and NMR-data (N,-CHO, 9.5 6 ; N,-COZMe, 3.8 S), together with acid hydrolysis and reacylation (with formic-acetic anhydride or formic acid for aspidofractine and refractine and with methyl chlorocarbonate for pleiocarpine) established the nature of the acyl groups. The more strongly basic nature of pleiocarpinine (CXLVI-A) and its red ceric sulfate reaction, together with the absence of an NH band in the IR-spectrum suggested an N-alkyldihydroindole, and that the N, substituent was methyl was shown by the fact that pleiocarpine and pleiocarpinine give a common reduction product, N,-methylkopsinyl alcohol (CXLVI-F),with lithium aluminum hydride (87, 49, 91, 92).

The NMR-spectra (Table I V ) showed that pleiocarpine and aspidofractine were unsubstituted in the benzene ring (note the characteristic downfield doublet with fine structure due to the C-17 aromatic proton which lies close to the N,-acyl group), while refractine had a C-17 methoxyl (position from UV- and NMR-aromatic patterns which resemble those of aspidospermine). The remaining two oxygen atoms SO far unaccounted for in each of the alkaloids were located in a carbomethoxyl group (IR-, 5.7P5.80 p ; NMR-, ca. 3.7 6,methyl singlet) (82, 91, 92).

The principal evidence for the aliphatic structure of the alkaloids was presented in Section 111, B, where it was shown that they have the car-

14. Aspidosperrna

AND RELATED ALKALOIDS

435

bon-nitrogen skeleton represented in the formulas A-Q in the accompanying illustration. Some transformations of a peripheral nature are described below. 1. Rings D and E

Direct oxidation of pleiocarpine (CXLV-A) with permanganate gave a mixture of two Nb-lactams,the five-membered lactam A (CXLV-D, v, 1712 or 1683 cm-1) and the six-membered lactam B (CXLV-E, v, 1672 cm-1) whose UV-spectra were unaltered and which could be reduced back to a mixture of kopsinyl alcohol (CXLII-F) and its N-methyl derivative (CXLVI-F),thus demonstrating that no skeletal change had taken place. The formation of these neutral lactams has been cited (Section 111, B, 3, 4) as evidence that N, lies a t the junction of a fiveand a six-membered ring (rings E and D, respectively) (95). 2 . The Carbornethoxyl Group I s Linked to a Carbon Atom Which Also Bears a Xingle Hydrogen Atom

Mild alkaline hydrolysis of refractine and aspidofractine gave the free acids (CXLIX-B and CXLII-B, respectively) which on reformylation and reesterification did not give back the parent alkaloids (37; see, however, 89), but isomeric esters which were named isorefractine and isoaspidofractine (CL-C and CXLIII-C, respectively). Treatment of the deformyl derivatives CXLIX-A and CXLII-A with sodium methoxide in methanol effected the same isomerization directly to give the deformyl isoesters, CXLIX-C and CXLII-C, respectively. The carbomethoxyl group is thus attached in a sterically unfavorable configuration t o a carbon atom which also bears a hydrogen atom; the presence of this hydrogen atom may be confirmed by its replacement by deuterium to give CXLIX-S when the same epimerization is performed with sodium deuteroxide in deuteriomethanol (see Section 111, B, 2 and Table V). That only epimerization was involved was established by carrying deformylrefractine (CXLIX-A) and deformylisorefractine (CXLIX-C) CHzOH (For T) --+ through the series of transformations COzMe CHzOTs ( H or W) +=CH2 (M). The olefinic final product no longer has a hydrogen atom on C-3 and, in accordance with expectation, the same olefin (CXLIX-M), was obtained in both series. Further confirmation of the C-3 hydrogen is seen in the NMR-spectra of the hydrocarbons (0)in which the -C02Me of the original alkaloids has been reduced to CH3 [reduction of the tosylate (CXLIX-W) via desulfurization of the thiophenyl ether (CXLIX-Y) (82), Raney nickel in ethanol on the iodide (CXLII-J), or Wolff-Kishner reduction of the

-

436

B. GILBERT

14. Aspidosperma AND RELATED ALKALOIDS

PI

437

438

B. GILBERT

aldehyde (CLV, 95)]. I n all cases a new methyl doublet a t 1.1-1.3 6 (J = 6 c/sec) shows coupling with a single C-3 proton. 3. The Carbomethoxyl Group I s Located on C-3 Kopsinyl alcohol [CXLII-F, LiAIH4 reduction of kopsinine (deformylaspidofractine CXLII-A),or pleiocarpine (CXLV-A)which also gives the N-methyl derivative] forms two ring compounds which link the alcoholic group to N, through a one-carbon bridge. One is the carbinolamine ether, CLIII, formed by the action of formaldehyde, the other a cyclic urethane (CLIV) prepared with benzyl chlorocarbonate and alkali. The latter shows IR-carbonyl absorption (5.88 p) characteristic of a six-membered cyclic urethane (95; cf. 82, 89,91, 92) (note that CLIV can only maintain planarity of the amide group if formed from isokopsinyl alcohol), and the original CHzOH group, and hence the carboniethoxyl of the parent alkaloids is situated /3 to N,. Furthermore, the C-3 substituent, whether -C02Me or any derived group, is always found in the fragment h (M-28) in numerous mass spectra and therefore cannot be located on C-20 or C-21 which are expelled in the formation of this fragment. Fragment h decomposes to fragment i which has been shown (Section 111,B, 2) to incorporate the D ring, C-10 and C-4. The carbomethoxyl group is thus limited to C-3 and C-11, with the foregoing evidence excluding the latter position. The transformation of the carbomethoxyl group to the exocyclic olefin (M) has already been described. Further transformation of the N,-deformyl olefins (CXLII-M and CXLIX-M) via their N,-formyl derivatives and by ozonization to the N,-formyl norketones (CXLIII-P and CL-P, respectively) led to the deformyl norketones (CXLII-P and CXLIX-P). The IR-absorption of these (ca. 5.77 p) showed that the carbonyl group and hence the original carbomethoxyl group probably lay in a strained six-membered ring. 4.

The Stereochemistry of the Alkaloids

(See also Section 111,B, 10.) Only structure A for the five alkaloids is compatible with the foregoing facts. Moreover, as will be seen subsequently (Section 111,J),kopsinyl iodide (CXLII-J)may be transformed into kopFane (CLXXXIV)in which the C-3’ carbon atom forms a bridge between C-3 and C-11. This requires the C-3, C-4 bridge to be cis to ring E ; the stereochemistry of the remaining centers follows automatically. That the carbomethoxyl group has the a configuration is shown not only by this bridge formation but also by its instability with respect to the is0 series in which the group must possess the less hindered /3 orientation.

14. Aspidosperma

AND RELATED ALKALOIDS

439

H. KOPSINILAM AND PLEIOCARPINILAM The amides kopsinilam (CXLII-D) and pleiocarpinilam (CXLVI-D) occur together with the more strongly basic pleiocarpine, pleiocarpinine, and kopsinine (Table I). Their weak basicity derives from the dihydroindole nitrogen which, as is shown by the UV-spectra and ceric color reaction (N-CH3, red; NH, orange), is not acylated. The IR-spectra (v, COzMe, 1736-1740 cm-1; v , CO-Nb, 1684-1696 cm-1) suggest the presence of an ester function as well as a five-membered lactam. The nonreactivity of pleiocarpinilam toward methyl iodide confirmed that this involved Nb (note that steric hindrance prevents quaternization of Na). Lithium aluminum hydride reduction of pleiocarpinilam (CXLVI-D) and kopsinilam (CXLII-D) gave, respectively, N-methylkopsinyl alcohol (CXLVI-F) and kopsinyl alcohol (CXLII-F) and thus there was every likelihood that these alkaloids were the E-ring lactams derived from pleiocarpinine and kopsinine, respectively. The correctness of this view was shown by synthesis. Pleiocarpinilam was obtained by the direct permanganate oxidation of pleiocarpinine (CXLVI-A) in acetone, while kopsinilam could be prepared, in a similar manner, from N-acetylkopsinine (CXLIV-A) followed by acid hydrolysis of the acetyl group, or from pleiocarpine lactam A (CXLV-D) by hydrolysis and simultaneous decarboxylation of the N-carbomethoxyl followed by reesterification of the C-3 carboxylic acid. It was established that these two amides are not formed by the action of air and light on pleiocarpinine and kopsinine and that the amides are therefore not artifacts. No trace of pleiocarpine lactam A or any sixmembered lactam was encountered (96).

I. KOPSIFLORINE, KOPSILONGINE, AND KOPSAMINE The three alkaloids named in the title occur together with kopsinine in Kopsia longi$ora (86, 57, 90), and kopsamine was shown (88) to be identical with the “kopsine” isolated in 1920 from K . $avida (85, perhaps a wrong identification of K . pruniformis Reichb. f. et Zoll. ex Bakh. f.). The name kopsamine was retained, as “kopsine” now refers to another alkaloid. Analysis of the bases and their salts established the empirical formulas while the IR-spectrum indicated the presence of ester (v, 1740 cm-1) and amide (v, 1688 cm-1) groups in all three alkaloids. Mild alkaline hydrolysis of kopsamine, C24H28N207 (CLVIII),and of kopsiflorine, C23H~8N205(CLVI), resulted in cleavage of afi ester

440

B . GILBERT

group to give kopsaminic (CLX) and kopsiflorinic (CLIX) acids, respectively. These acids could be remethylated t o the parent bases, but on treatment with dilute acid, they decarboxylated to give kopseine (CLXIII) and kopsifloreine (CLXI), in which compounds a new weakly basic secondary amine had been generated, as was shown by formation of the nitroso derivatives, CLXVI and CLXIV. Kopseine and kopsifloreine retained the original ester group which could be hydrolyzed under more vigorous conditions to give, respectively, kopsamic (CLXIX) and kopsifloric (CLXVII)acids. A similar acid (CLXVIII) could also be NZO~ obtained by vigorous hydrolysis of kopsilongine, C Z ~ H ~ ~ (CLVII), and reesterification with diazomethane gave kopsilongeine (CLXII), a product corresponding in every way t o kopseine and kopsifloreine and

/

Ri H CLVI H CLVII C L V I I I 0-CHz-0 CLXI H CLXII H C L X I I I 0-CHz-0 CLXIV H CLXV H C L X V I 0-CHz-0

Rz H OMe

/

CO zMe

CbzMe R3 COzMe COzMe COzMe

Ri

COzQ

Rz

Ri

Rz

H H 0-CHz-0

H OMe

~

CLIX CLX

H 0-CHz-0

H

CLXVII CLXVIII CLXIX

H H OMe H H H NO OMe NO NO

forming a nitroso derivative, CLXV. These results indicated the presence of the groupings N-COZMe and C-C02Me in the three alkaloids; examination of the parent bases and their decarbomethoxy derivatives showed that they were dihydroindoles. The spectra and functional group analysis indicated that kopsilongine contained a 17-methoxy and kopsamine a 16,1i’-methylenedioxy grouping whereas kopsiflorine was unsubstituted in the benzene ring. The strong similarity in the chemical behavior of the three bases suggested that otherwise they were of identical structure, and this has subsequently been shown to be true (89, 97).

14. Aspiclosperma

AND RELATED ALKALOIDS

44 1

J. KOPSINEAND RELATED ALKALOIDS Although kopsine (CLXX) was probably isolated in the last century (83, 84), the true nature of its unusual cage-like structure has only become known recently (101). Preliminary investigations (103, 104, 105)

showed that it was a dihydroindole containing a carbomethoxyl group. In later work (100, log), the correct molecular formula, C~zH24N204,was established and by comparison of the UV-spectrum with model compounds and mild alkaline hydrolysis accompanied by decarboxylation to a dihydroindole (CLXXI) unsubstituted on N, (UV-, benzenoid in acid), it was shown that the carbomethoxyl group was attached to that nitrogen atom. From the IR-spectrum, a five-membered ketone (v, 1757 cm-1) and a hydroxyl group hydrogen-bonded to the N-carbomethoxyl (v, OH, 3268 cm-1; COZMe, 1679 cm-1) were recognized. The presence of the hydroxyl was confirmed by the NMR-spectrum which showed a singlet a t 7.2 6 that could be eliminated by previous treatment of the base with deuterium oxide (100, 101, 109). Furthermore, the hydroxyl and ketone groups were probably vicinal since kopsine reacted with periodate (107, 110). Both kopsine (CLXX) and decarbomethoxykopsine (CLXXI) resist acetylation, indicating that the hydroxyl group was probably tertiary and that N , was sterically hindered. The ketone could be reduced to two epimeric alcohols; dihydrokopsine-A (CLXXII)resulted with sodium borohydride reduction (100, 109))whereas catalytic hydrogenation gave dihydrokopsine-B (CLXXIV, 100, 104). Both formed only a monoacetate. Kopsine shows no vinyl absorption in the NMR-spectrum and could not be further reduced (except by catalytic hydrogenation of the benzene ring). The molecule therefore contains seven rings, since subsequent evidence excludes a tetrasubstituted double bond. Evidence was then accumulated to establish the relation of the ketonic carbonyl to Nb. This was first obtained by examination of two lactams. One, “lactam A” (CLXXV),was the final oxidation product of either of the epimeric alcohols dihydrokopsine-A or -B, in which the secondary hydroxyl had been reoxidized and a new carbonyl group had been introduced adjacent to Nb, as could be recognized both from its neutrality and from the IR-spectrum (v, 1675 cm-1) of its N,-decarbomethoxy derivative (CLXXVI). The same lactam A resulted from the direct oxidation of kopsine itself. The other, “lactam B ” (CLXXVII), was an intermediate oxidation product of dihydrokopsine-B in which the secondary hydroxyl had not suffered alteration. Comparison of the IR-absorption of both lactams with that of kopsinilam (CXLII-D) and

442

B. GILBERT 0

0

t-

CLXXV; R = COzMe CLXXVI: R = H

CLXXVII

CLXXII; R = COZMe CLXXIII; R = H

/

i

C,O, pyridine

t-

CLXXIV

CLXXXVI

CLXX; R = COzMe CLXXI; H = H

CLXXVIII; RI = CHs, RI = H CLXXIX; R I = H. Rz = D

CLXXXIV; R = H CLXXXV; R = Ac

CLXXX

I

J

/

CXL1I.J

CLXXXIII

CLXXXVII; R = COeMe CLXXXVII-A; R = H

CLXXXVII-B; R = COzMe CLXXXVII-C: R = H

RI CLXXXI

COzMe

Rt =O

CLXXXII-A COpMe

&H

CLXXXII-B COrEt


14. Aspidosperrna

AND RELATED ALKALOIDS

443

pleiocarpinilam (CXLVI-D) indicated that they were five-membered, and further comparison with 10,ll-dioxoaspidospermine(XIII, Section 11,C, Ref. 24) showed that “lactam A” could not be an a-keto lactam. However, in the NMR-spectrum of this compound (CLXXV) a singlet could be distinguished at 2.82 6 which became a slightly resolved doublet (2.63 6, J = 1 c/sec) in the hydroxy lactam B (CLXXVII). This could be attributed to the groupings CO-CH(CR2)-CONb in lactam A and CH(OH)-CH(CR2)-CONb in lactam B where R is not hydrogen and it was also possible t o say that the two hydrogen atoms in the latter were oriented at an angle of nearly 90” to one another, thus establishing the configuration of the C-3’ hydroxyl (101, see also 111). The same relation between the ketonic carbonyl group of kopsine and Nb was demonstrated by study of the Hofmann degradation of the methiodide. This yielded an a,/?-unsaturated ketone CLXXX whose double bond lay in a terminal methylene group (IR-, v, =CH2, 945, 920, 1629; Y , CO, 1748 cm-1; NMR-, two vinyl singlets at 5.07 and 6.3 6). Three reduction products (CLXXXI, CLXXXII-A, and CLXXXII-B) of this methine were prepared. In each case a new methyl doublet at 0.48-0.58 6, coupled (J = 7-7.5 clsec) to a single a-proton, could be recognized. This proton (on C-11) appeared in the spectrum of CLXXXI as a quartet at 3.7 6 with the same coupling constant (101). Reaction of the diol, CLXXXII-A, with periodate confirmed that it was an a-glycol. The foregoing experiments establish that kopsine contains two fivemembered rings apart from the dihydroindole moiety. One of these rings contains an a-hydroxy ketone in which the hydroxyl is tertiary. The other contains R&-CH-CH2-Nb, in which none of the groups R is hydrogen and the ketonic carbonyl is attached to the carbon atom to Nb.These two rings can only be accommodated in the partial structure CLXXXIII. To this may be added the fact that the hydroxyl group is close t o -N, and also that Nb is probably also involved in a six-membered ring since 3,Ei-diethylpyridineresults from the zinc dust distillation of kopsine (105). This and the occurrence of kopsine (CLXX) and pleiocarpine (CXLV-A)in plants of the same genus led to the belief that the two alkaloids might be structurally related. In fact, a direct correlation was achieved. Kopsinyl iodide (CXLII-J,Section 111,G), obtained from pleiocarpine (CXLV-A), was pyrolyzed to give in 60% yield a heptacyclic hydrocarbon kopsane (CLXXXIV) which forms an N,-acetate (CLXXXV) and a five-membered lactam (CLXXXVI) (NMR-spectrum similar to that of CLXXVII). The new ring which has been formed in kopsane does not therefore involve N, or C-10 and it is not a cyclopropane (NMR-spectrum clear from 0-0.9 6). Examination of models shows that, barring rearrangement, the new ring must involve closure of

444

B. GILBERT

C-3' with C-11. Prolonged treatment of kopsine with hydriodic acid and red phosphorus gave a small yield of the same hydrocarbon, kopsane (CLXXXIV), Taken in combination with the previously described evidence, this establishes the structure CLXX for kopsine. The mass spectrum of kopsine (Table V) is quite different from those of other alkaloids of this group which do not possess the extra ring, in t,hat the base peak, derived directly from the molecular ion (metastable peak at 209), occurs at m/e 282. That this involves the loss of the CO-C(0H) grouping and three more carbon atoms, but not of the N,substituent was shown by the spectra of the three derivatives CLXXI, CLXXVIII, and CLXXIX in which the base peak followed changes at N, but was not affected by alterations on C-3'. The results did not permit a unique structural interpretation (106, 107). Two other alkaloids, fruticosine and fruticosamine (Section 111, K , 109, 113), isomeric with kopsine, possess similar mass spectra. Kopsine (CLXX) undergoes reversible acyloin rearrangement by heating in tetralin t o give an isomer, isokopsine (CLXXXVII, 111).Similarly, bythe action of dilute alkali on kopsine or decarbomethoxykopsine (CLXXI),an equilibrium mixture of the latter with decarbomethoxyisokopsine (CLXXXVII-A) is obtained. I n isokopsine (CLXXXVII), hydrogen bonding is no longer observed between the tertiary hydroxyl group (v, 3509 cm-1, no downfield proton in the NMR-spectrum) and the N,-carbomethoxyl (v, 1706 cm-1). To confirm its structure, decarbomethoxyisokopsine (CLXXXVII-A) was directly related, via its sodium borohydride reduction product decarbomethoxydihydroisokopsine (CLXXXVII-C), to isokopsine, from which the same compound could be obtained by successive borohydride reduction to dihydroisokopsine (CLXXXVII-B) followed by hydrolysis of the N,-carbomethoxyl group (111). Both decarbomethoxykopsine (CLXXI) and its is0 derivative (CLXXXVII-A) occur in the leaves of K . fruticosa ( 113).

K. Kopsia ALKALOIDS OF UNKNOWN STRUCTURE Five alkaloids of unknown structure that may well belong to the aspidofractinine-kopsine group have been reported. Three, kopsaporine, kopsingine, and kopsingarine, have been isolated from K . singapurensis (108, 89, 56). The other two, fruticosine and fruticosamine, occur with kopsine in K. fruticosa (109, 100, 113). Both are isomers of kopsine and similarly contain an N-carbomethoxydihydroindole nucleus unsubstituted in the benzene ring, a five-membered ketone, and a hydroxyl group. Fruticosamine is converted into fruticosine by mild alkali or stronger acid treatment. Fruticosamine resists acetylation and does not

14. Aspidosperma

AND RELATED ALKALOIDS

445

form a methiodide, whereas fruticosine forms both acetate and methiodide. Although the similarity of the mass spectra of these alkaloids and kopsine suggests a similar skeleton, there are some notable differences, among which are the secondary nature of the hydroxyl group in fruticosine and the resistance of the five-membered ketone to reduction (109). IV. The Aspidoalbine Group A. ASPIDOALBINE AND ITS N-ACETYLANALOG The study of Aspidosperma album (see Volume VII, p. 129) in two laboratories resulted in the isolation of aspidoalbine and its N-acetyl analog [(CLXXXVIII+ CLXXXIX, 52) which have been separated (42,48)]as well as aspidocarpine and its demethyl derivative (Section 11, I, Refs. 42, 48). The UV-spectrum of aspidoalbine is similar to that of aspidocarpine (XLIV), showing the same bathochromic shift in alkali, and together with the IR-absorption a t 3.1-3.5 p (hydrogen-bonded hydroxyl) and 6.17p (amide carbonyl), suggesting that the alkaloid was Elementary analysis and mass also a 17-hydroxy-N-acyldihydroindole. spectrometry established the empirical formula as C24H32N205, although the latter showed the presence of a lower homolog, the nature of the mixture being resolved by acid hydrolysis which gave, after esterification and vapor phase chromatography of the volatile acid fraction, methyl propionate (mainly) and methyl acetate. The alkaloid mixture thus consisted principally of a 17-hydroxy-N-propionyldihydroindoletogether with a smaller proportion of the N-acetyl analog. Two methoxyl groups accounted for two of the remaining three oxygen atoms, and the NMR-spectrum (Table IV) showed that these were in the aromatic ring, since only a single aromatic proton singlet at 6.826 was observed. It was assumed that this proton occupied the rarely oxygenated 14 position. The fifth oxygen atom was not in a carbonyl group nor could it be acylated and it was therefore assumed to be ethereal. Methylation of the mixed alkaloid (CLXXXVIII + CLXXXIX) with dimethyl sulfate in acetone (note nonformation of an N,-metho salt, which is indicative of a steric hindrance around N, similar to that observed in the aspidospermine group) gave impure 0-methylaspidoalbine (CXC+ CXCI) from which the deacyl derivative, CXCII, was obtained by acid hydrolysis. Pure 0-methylaspidoalbine (CXC) and its lower homolog, CXCI, were prepared by acylation of CXCII, and in their NMR-spectra could be recognized the C-2 proton quartet, characteristic of the aspidospermine series, The mass spectra of aspidoalbine and its 0-methylated derivatives

5)

C X C I X ; R = Ac C X C I X - A ; R = COEt

CXCVIII

Series B

Serics A

_Ri _R2 JXXXIV LXXXV LXXXVI

H

H

Ac Et

Ac

~~

CXCIII CXCIV CXCV CXCVI CXCVII

H

Ri H H Ac Et CD2CD3

RZ R J H H Ac H H

H D H H H

C'

H2C\;/\

I,,I Me0

\CHzORz

R1 (H)

R C

Series A , Rz = H Series B. Re = OMe

a' 446

14. Aspidosperrna

AND RELATED ALKALOIDS

447

CXC, CXCI, and CXCII show the M-28 and weak indole peaks expected for an aspidospermine-like molecule. However, a strong M-44 peak (loss of CHzCHzO) also appears, and the base peak is found a t m/e 138 instead of a t m/e 124. Thus, aspidoalbine probably has an aspidospermine-type skeleton modified in the ring D area which contains the ethereal oxygen atom. Reduction of 0-methyldeacylaspidoalbine (CXCII) with lithium aluminum hydride gave an alcohol (CXCIII), and this showed the most intense peak a t m/e 140 which was shifted further to m/e 141 when the reduction was effected with lithium aluminum deuteride. The susceptibility to reduction under these conditions is characteristic of a carbinolamine ether, and since the indole peaks, 6 and homolog, remain unchanged during the foregoing transformations, this ether must terminate adjacent to Nb (52). The exact nature of the ether ring was demonstrated as follows. Oxidation of the 0-methyl-N,-acetyl derivative of aspidoalbine (CXCI) with chromium trioxide-pyridine yielded a five-membered lactam (CXCVIII) and a five-membered lactone (CXCIX, 52). The composition of the lactone shows that two hydrogen atoms in CXCI have been replaced by oxygen, and at the same time it was observed that a two-proton quartet which appears at 4.02 6 in the 100-mc NMR-spectrum of 0-methyldeacylaspidoalbine (CXCII) had disappeared. This quartet could therefore be ascribed to a methylene group (C-21 in formula CC) adjacent to the ethereal oxygen atom. It was shown by spin-decoupling to be coupled to two nonequivalent protons, respectively 2.05 and 2.72 ppm upfield (on C-20 in formula CC). When the coupling between these was also eliminated by double decoupling, then each in turn could be made to absorb as a singlet, thus demonstrating that the next adjacent carbon atom ((3-5in formula CC) bore no hydrogen atom (114). Confirmation that no hydrogen atom was attached to the carbon atom (C-19 in formula CC) on the other side of the ethereal oxygen atom was obtained by examination of the 100-mc spectra of 0-methyldeacylaspidoalbinol (CXCIII) and its deuterio derivative (CXCIV). The only difference between the spectra rests in the absence from the second of a sharp signal a t 2.17 6 [the position in which the C-19 proton absorbs in aspidospermine and pyrifolidine (Table IV)], showing that a lone proton is generated by the LiAlH4 reduction of the carbinolamine ether which therefore terminated at a quaternary center. The ether ring may thus be represented by partial formula CC (52). Both the aforementioned NMR-singlet exhibited by CXCIII and the existence of the five-membered lactam (CXCVIII) support the earliercited mass spectral evidence that aspidoalbine has an aspidosperminetype skeleton. Further confirmation of this was obtained by comparison

448

B . GILBERT

of the mass spectra of the three aspidoalbinol derivatives CXCIII, its

N,O-diacetate, CXCV, and its N-ethyl derivative, CXCVI, with those of three corresponding decinnamoylcylindrocarpol derivatives, LXXXIV, LXXXV, and LXXXVI, whose preparation was described in Section 11, M. All three pairs showed the expected identity of the piperidine-containing peaks (c and satellites) while the indole-containing peaks ( b and satellites) were shifted in the aspidoalbinol derivatives by 60 mass units due to the two extra methoxyl groups present. The M-28 species (fragment a in Section 11, D) which decomposes to give fragment c as is shown by the recognition of metastable peaks, was particularly weak in the spectra of the aspidoalbinol derivatives and this was attributed t o a probably opposite configuration at C-19 with respect to the cylindrocarpol derivatives. Subsequent work (1 13i)has shown, however, that the principal ring-opened reduction product of an aspidoalbine derivative has C-19, p-H (as in cylindrocarpol), the (2-19, a-epimer being a minor product. The accommodation of the ether ring as represented by CC in the aspidospermine skeleton can only be made in the manner represented by formula CLXXXVIII for aspidoalbine, and the mass spectral fragments Models permit the conobserved may be rationalized as a', b, and struction of two stereoisomers of this structure, one with the ether ring on the same side of the molecule as is the ethyl group of aspidospermine. The co-occurrence of aspidoalbine with alkaloids of the aspidosperminetype made it probable that this was the true configuration ( 5 2 ) ,and this has now been established by interrelation with obscurinervidine (CCI-R), neblinine (CCI-S),and aspidocarpine (XLIV, 113i). GI.

B. ASPIDOLIMIDINE I n addition to seven alkaloids of the aspidospermine group (Section 11, F, I, J, and N) isolated from Aspidosperma limae, this plant also yielded a base aspidolimidine, whose NMR- and mass spectra were not consistent with membership of that group. The UV-spectrum of aspidolimidine is very similar to those of aspidocarpine (XLIV) and aspidolimine (LIII), which accompany it in the plant. Indeed, the same aromatic substitution pattern is shown by NMR-spectra in which a hydrogen-bonded 17hydroxyl group may be recognized a t 10.78 6, an aromatic methoxyl group singlet at 3.88 6, and two ortho hydrogen atoms a t 6.73 6 and 7.09 6 (J = 8 clsec). An N,-acetyl group is also present (2.32 6). The difference from the aspidocarpine-type spectrum lies in the absence of C-ethyl absorption and the presence of absorption due to three protons

14. Aspidosperma AND

449

RELATED ALKALOIDS

instead of only one in the 4.0 6 region. A similar feature in the spectrum of aspidoalbine (CLXXXVIII) could be attributed to the absorption of the C-2 and the two C-21 protons. I n fact, a comparison of the mass spectra of aspidolimidine and aspidoalbine showed that the aliphatic portions of the two molecules were identical, each showing the base

CCI CCI-L CCI-K CCI-0

Ri Rz OMe OH H H H H OMe O H

Ri

R3 Ac H Ac COEt

CCI-B CCI-C CCI-D CCI-E CCI-F CCI-G

Me

H H Ac H Me

Rz Ac Ac COEt COEt H COEt

Ri CCI-A H CCI-M H CCI-N M e 0

Rz

R3

Me

CHO EtCO EtCO

H H

I Et CCI-I;

R

= OTs

R = SPh XXXIX; R = H CCI-J;

LXXXIV CCI-H LXXXVII

Ri

Rz

Me Me

H COEt COEt

H

peak (c) at m/e 138.The molecular ion appeared a t m/e 384 ( C Z Z H ~ ~ N Z O ~ ) , 30 mass units lower than in the case of aspidoalbine, and a similar shift was observed in the M-28 (a’),M-44 (loss of C-2O,C-21,0 bridge), and b peaks due to the absence in aspidolimidine of the C-15 methoxyl group. The alkaloid may therefore be assigned the structure CCI (40).

C. DICHOTAMINE AND 1-ACETYLASPIDOALBIDINE The alkaloid dichotamine, C Z I H ~ ~ Noccurs ~ O ~ in , Vallesia dichotoma together with vallesine, aspidospermine, reserpine, and akuammidine. The UV-spectrum was indicative of a 17-methoxy-Na-formyldihydroindole structure, and the IR-spectrum showed two carbonyl bands, one

450

B . GILBERT

due to the expected N-formyl function, the other to a five-membered lactone (5.67 p ) (41). Subsequently, the molecular formula was confirmed mass spectrometrically, while lithium aluminum hydride reduction of the deformyl compound gave decinnamoylcylindrocarpol (LXXXIV) and its 19-epimer in equal amounts, establishing the structure (CCI-A) for dichotamine (113e). Also found in V . dichotoma were haplocidine (CCI-C), whose structure and stereochemistry were proved independently by methylation with dimethyl sulfate to the highly crystalline 0-methyl derivative (CCI-B), removal of the N,-acetyl group by acid hydrolysis, and lithium aluminum hydride reduction to decinnamoylcylindrocarpol (LXXXIV) and its 19-epimer ; and the unsubstituted compound CCI-K, named 1-acetylaspidoalbidine (with the parent of the aspidoalbine series, CCI-L, being designated as aspidoalbidine to simplify nomenclature). The structure of CCI-K was determined by its analogous reactions and mass spectrometric fragmentation patterns of derivatives with those of haplocidine and its 0-methyl derivative (CCI-C and CCI-B) (113e).

D. HAPLOCINE AND HAPLOCIDINE Preliminary investigation of the plant Haplophyton cimicidum indicated the presence of an alkaloid haplophytine, possibly closely related to the Aspidosperma alkaloids (113a). A reinvestigation of the plant has resulted in the isolation of eburnamine (CCCXL), isoeburnamine (CCCXLI),0-methyleburnamine (CCCXL-A)(see Section VIII, E), and several alkaloids belonging to the aspidoalbine group (113b). Among these the structures of haplocine (CCI-D) and haplocidine (CCI-C) have 219, 258, 291-292 mp) of the been elucidated. The UV-absorption (A,, alkaloids was consistent with their being 17-hydroxy-Na-acyldihydroindoles. This was confirmed by the appearance of a hydrogen-bonded amide carbonyl peak a t 6.13 p in the IR-spectrum which shifted to 6.00 p in the spectrum of the 0-acetate, CCI-E. The phenolic nature of the hydroxyl group followed from the additional carbonyl absorption of the 0-acetate, CCI-E, at 5.69 p. Deacylation of both alkaloids gave the same product, depropionylhaplocine (CCI-F), and reacylation experiments showed that haplocidine (CCI-C) was its N-acetyl and haplocine (CCI-D), its N-propionyl derivative (113b). That the third oxygen atom of haplocine was present in a carbinolamine ether ring was shown by methylation to 0-methylhaplocine (CCI-G),followed by catalytic reduction which opened this ring stereospecifically to give a single product, CCI-H, which contained a new

14. Aspidosperma

AND RELATED ALKALOIDS

45 1

primary alcoholic group. The structure and stereochemistry of CCI-H were established by reduction of the -CHzOH group to -CH3 by way of the tosylate, CCI-I, and the crystalline thiophenyl ether, CCI-J, to palosine (XXXIX) identical with an authentic sample prepared from aspidospermine (11)(11310, 113c). Furthermore, catalytic reduction of haplocine led directly to the known alkaloid limaspermine (LXXXVII). The similarity of the observed behavior of haplocine with that of aspidoalbine led to the structure CCI-D for the alkaloid, and hence to CCI-C for haplocidine.

E. CIMICINE AND CIMICIDINE I n addition to the alkaloids haplocine and haplocidine described in Section IV, D, the plant Haplophyton cimicidum contains two hexacyclic alkaloids, cimicine, CzzHz6N204, (CCI-M) and cimicidine, Cz3HzsNz05, (CCI-N)which contain a y-lactone ring (IR,5.65 and 5.71 split carbonyl in nujol) (113j, k, 1, m). The NMR-spectra of these two alkaloids resembled one another, the principal difference resting in the presence of an aromatic methoxyl peak (3.876) in the spectrum of CCI-N absent from CCI-M. Furthermore, the NMR-spectrum of cimicine was similar to that of haplocine (CCI-D), leaving no doubt that it was a 17-hydroxy-Napropionyldihydroindole. The empirical formula suggested that cimicine (CCI-M)might be simply 21-oxohaplocine, and this was proved by direct oxidation of haplocine to cimicine in low yield using chromium trioxide in pyridine (compare the similar oxidation of 0-methylaspidoalbine, Section IV, A). The supposition that cimicidine is 16-methoxycimicine (CCI-N) is in accord with its NMR-(2 ortho aromatic H, 6.73, 7.026, J = 8 clsec) and IR-spectra. Both bases suffer cleavage of the carbinolamine lactone ring on catalytic hydrogenation to yield zwitterionic amino acids (compare reduction of haplocine, Section IV, D) (113j).

F. OTHERALKALOIDS OF

THE

ASPIDOALBINE GROUP

Recent studies have led t o the isolation of other alkaloids of this group. notably the parent member, CCI-L, which occurs in Aspidosperma fendleri Woodson, and has been named fendleridine. The principal base from this plant, fendlerine, is the Na-propionyl analog (CCI-0) of aspidolimidine (Section IV, B) with which it has been interconverted (113f).

A lactone, CXCIX-A, related to dichotamine, cimicine, and cimicidine

452

B. GILBERT

(Sections IV, C, D) but bearing three aromatic methoxyl groups, has been isolated from an Aspidosperma sp. Its structure follows from its interconversion with the known lactone CXCIX derived from aspidoalbine (Section IV, A) (113g).

G. OBSCURINERVINE AND RELATED ALKALOIDS I n addition t o aspidocarpine (XLIV) and aspidolimine (LIII), Aspidosperma obscurinervium Azambuja also contains three lactonic alkaloids : obscurinervine, mp 204"-205' (dec.), [a]L7 - 54" ; dihydroobscurinervine, mp 184°-1850 (dec.), [a]? - 61" ; and obscurinervidine, mp 206"-207" (dec.), [a]: -39". Recent studies have shown these to possess the structures CCI-P, CCI-Q, and CCI-R, respectively (113i). A dihydroindolic structure for these bases was suggested by their superimposable UV-spectra (A, 218, 253-255, 308-312 mp; E 50,000, 6,000, 3,000) while the presence of a y-lactone was indicated by strong IRabsorption a t 1755 cm-1. The first two exhibit strong M-Et and M-Me, and the last only M-Me peaks (loss of R2) in the mass spectrum, but the stability of the polycyclic skeleton results in only low intensity indolic peaks of which the most notable occurs a t mje 244 for all three molecules. The alkyl group Rz (ethyl or methyl),the lone aromatic proton, and two aromatic methoxyl groups may be recognized in the NMR-spectra of the bases, while with obscurinervine (CCI-P) and obscurinervidine (CCI-R) the two vinylic protons in positions 6 and 7 may be distinguished. Mild hydrogenation of obscurinervine gives the dihydro derivative (CCI-Q) which accompanies it in the plant. The alkaloid neblinine (CCI-S), mp 257"-258", from A . neblinae, is a demethoxy derivative of obscurinervidine (113i).3

CCI-P OMe CCI-Q OMe CCI-R OMe CCI-S H

Et Et Me Me

6,7-dihydro

3 Dihydroobscurinervidine also occurs in A . obscurinervium and the 22-ethyl homolog of neblinine (CCI-S, R1 = H, Rz = E t ) has been isolated from A . neblinae. Synthesis of a reduction product of neblinine from aspidocarpine established the relative stereochemistry a t four centers, that at the other two following from NMR data ( 1 13i).

14. Aspidosperma

AND RELATED ALKALOIDS

453

V. The Condylocarpine Group A. INTRODUCTION This small group, which at the time of writing comprises nine alkaloids occurring in the genera Aspidosperma, Diplorrhyncus, Pleiocarpa, and Stemmadenia of the family Apocynaceae, is important as its members represent close biogenetic relatives of the ulein group alkaloids (115). The common biogenetic origin of stemmadenine (CCXIII)and echitamine (Chapter 8) is also clear.

B. ASPIDOSPERMATINE, ASPIDOSPERMATIDINE, AND RELATED ALKALOIDS

Aspidosperma quebrachoblanco is a source of several aspidosperminetype alkaloids which have been discussed in Section 11, C and E. By use of alumina and vapor phase chromatography, six alkaloids were isolated whose mass spectra showed them to be related to one another but not to belong to the aspidospermine group. Five were characterized by a base peak a t m/e 136 (compare m/e 124 for the aspidospermine group) and contained a hydrogenatable double bond while the sixth had the base peak at m/e 138 and did not contain this double bond. The alkaloids were named by their molecular weights [later names (118) in parentheses] : 266-B (aspidospermatidine), 280-B (N,-methylaspidospermatidine), 296-B (deacetylaspidospermatine), 308-B (N,-acetylaspidospermatidine), 338-B (aspidospermatine), and 340-B (dihydroaspidospermatine), the last being the saturated member of the group. The UV-spectra of aspidospermatidine and aspidospermatine showed that they were, respectively, a dihydroindole unsubstituted in the aromatic ring and on N, and an N-acyldihydroindole bearing a methoxyl group ortho t o N, as in aspidospermine (11). The difference in molecular weight between the two corresponded to (Me0 + CH&O-2H) so that it was reasonable to assume that the N-acyl group in aspidospermatine was acetyl. In accordance with expectation, the indole peaks, b and homolog, differed by 30 mass units due to the methoxyl group alone (28, 51a). The isolation of 3-ethylpiperidine by the zinc dust distillation of aspidospermatidine and deacetylaspidospermatine made it probable that the m/e 136 peak derived from a piperidine ring incorporated in the aliphatic part of the molecule. An ideal model compound, dihydrodecarbomethoxyakuammicine (CCII, mol. wt. 266) was known (19). Its

454 B. GILBERT

R

---f

14. Aspidosperma

i

f

AND RELATED ALKALOIDS

455

456

B. GILBERT

mass spectrum was very similar to that of aspidospermatidine (266-B) but not quite identical. The strong m/e 136 peak was present and indicated that a molecule of this skeleton fragmented according to the scheme indicated in the formulas to give fragments b and r . It will be noted that there is no ethylene bridge to be lost in this molecule and this is in accord with the absence of an M-28 peak from the spectra. It was clear that the structure of aspidospermatidine represented some slight modification of the structure CCII and a clue was obtained from the mass spectra of the dihydro derivatives of both compounds (CCIII and CCV). I n both cases the base peak was now shifted to m/e 138 while the indole peaks remained unchanged. A new peak appeared in the spectrum of CCIII at m/e 199, which was absent from that of dihydroaspidospermatidine. An examination of the postulated initial breakdown product, q, of CCIII indicated that it would further decompose not only by cleavage a t x (to give r, m/e 138), but also at y which is both allylic and /3 to nitrogen (see Section 11, D and Refs. 20, 21) to give s, and indeed s has the required molecular weight of 199. The production of s involves loss of the three skeletal carbon atoms, 16, 15, and 20 as well as the ethyl side chain. If the side chain had not been at C-20, then the loss observed would have been 28 mass units less and the s ion would be found at m/e 227. Such a peak is in fact found in the spectrum of dihydroaspidospermatidine, and the ethyl group is therefore at 14, the only position which is consistent with the production of 3-ethylpyridine by dehydrogenation (28, 21, 51a). Structure CCIV is thus established for aspidospermatidine, some slight doubt remaining over the position of the double bond, although the absence of terminal methylene absorption in the IR-spectrum excludes the alternative C-CH-CHz formulation for the side chain. Since the only differences observed in the spectra of the accompanying alkaloids were in the indole peaks, it was possible to formulate them as CCVI-CCX. Confirmatory interconversions, CCVIII to CCIV, CCIX to CCVII, and CCIX to CCX, were made. Base 338-B (CCIX) was recognized by its mp, rotation, and composition as the aspidospermatine that Hesse isolated from the same plant in 1882 (59). The fragmentation pattern proposed for alkaloids of this skeleton has been confirmed by the observation that spermostrychnine (CCXI) and its deacetyl derivative (CCXII) (21), as well as numerous other derivatives of the same basic skeleton substituted in positions 14, 16, and 20 (see following sections and Ref. 115a), fragment in the same way. I n many cases the structures of these compounds have been proved independent of mass spectrometry. An exception to the normal breakdown pattern is seen in strychanone (CCXLIV)in which the 16-carbonyl group

14. Aspidospermu

AND RELATED ALKALOIDS

457

blocks the usual path (36, 128, Section 11,D, L). Breakdown patterns for various unsaturated derivatives of the aspidospermatidine skeleton have been worked out by examination of a series of compounds derived from akuammicine and other alkaloids (115a). An alternative and more probable cleavage of aspidospermatidine itself is illustrated by the production of q' and r ' . A minor alkaloid of Aspidosperma compactinervium has been identified as CCXII-A. Its methyl ether (CCXII-B) differs from aspidospermatine only in the location of the aromatic methoxyl group. As in the case of the foregoing alkaloids the structural determination was based on mass spectrometry of the alkaloid and its derivatives CCXII-B and CCXII-C. The latter, in which the ethylidene side chain has been reduced, shows the s peak at mje 299 (227 + Me0 + Ac - 2H),thus locating this grouping in position 14. The position of the aromatic hydroxyl group is based on theUV-spectrum of CCXII-A [A 218,252,260 (sh), 294, and 300 mp] (48).

A C

CCXII-A; R = H C C X I I - B ; R = Me CCXII-C; R = Me, 14,19-dihydro

C. CONDYLOCARPINEAND STEMMADENINE Stemmadenine (CCXIII) was first isolated from Stemmadenia donnellsmithii (8), where it occurs with (+)-quebrachamine (I) and some iboga-type bases. Subsequently, it has also been found in Diplorrhyncus condylocurpon ssp. mossumbicensis, where it occurs together with condylocarpine (CCXV), normacusine-B (Section VIII, B), two yohimbines, norfluorocurarine, and mossambine (Section VI, B and C, Ref. 116). Analysis and mass spectral molecular weight determination established the empirical formula, C21H26N203, for stemmadenine (8, 116, 117). Its UV-spectrum was characteristic of an indole (cf. ref. 5 5 ) , while the IRspectrum indicated the presence of a normal ester grouping (1718 cm-1) and the absence of any substituent in the indole aromatic ring (116). These findings were fully borne out by NMR-spectroscopy which showed the presence of an indole NK (9.3 6), four aromatic protons, and a carbomethoxyl methyl singlet (3.79 6). A single vinyl proton quartet (5.4 6)

1

CCXIII

CCXVI

/

CCXVIl

1 . KbfnOa, HCl, 5'

2. Heat

CHz

t CCXIV

CCXVIII

m/e 123

I

ccxv

ccxx

1

I

CCXIX

I

CCXXI CCXXIII-A 14,lY saturated l4p-H CCXXIII-B 14,19 saturated 14a-H

Q

S

I

LiAlHa

P

4

I

CH2

I A

I

COzMe

+ CCIV

CCXXII CCXXIII

14,19 saturated

ccxxIv

Ri

CCXXIV-A Me0 CCXXN-B Me0 CCXXIV-C Me0

Rz H H

T

2,16-dihydro

AO 2,16-dihydr0

tl

460

B. GILBERT

and a methyl doublet ( 1.7 6) were indicative of the grouping C=CH-CH3. This double bond could be hydrogenated (117). The mass spectrum of stemmadenine exhibited, besides the molecular ion, peaks a t M-30 (loss of CH20) and M-18 (loss of HzO). The latter suggested the presence of a hydroxyl group and the former, considered together with an NMR-spectroscopic signal a t 4.38 6 (CH2 of CHzOH) and the fact that formaldehyde can be isolated during the conversion of stemmadenine into condylocarpine (see below), further indicated that this hydroxyl group must be primary. Palladium dehydrogenation of stemmadenine gave 3-ethylpyridine and a compound, C25HzzNz02, showing carbonyl absorption a t 5.88 p and an extended indole chromophore in the UV spectrum. The structure of this compound, which is symmetrical, was solved entirely by NMRspectroscopy. I n the spectrum eight aromatic protons could be recognized, six normally placed and two further downfield which must be close to one of the two carbonyl groups. Four other protons made up an AzX2 pattern, of which two protons appear a t 4.43 6 (between CO and the indole double bond) and the other two a t 2.78 6 so that the grouping Ar-CH(C0)-CHZ-CH(CO)-Ar was present. Finally, the remaining hydrogen atoms could be located in two ethyl groups which absorbed as a quartet and a triplet. These features can only be assembled in the structure, CCXVI, which may be seen to be formed from two units of structure, CCXVII, with the loss of one carbon atom. The part structure, CCXVII, and 3-ethylpyridine together contain all the carbon atoms of stemmadenine, and the two fragments could be linked to give the part structure CCXVIII for the alkaloid [linkage of C-16 in CCXVII to N, is excluded as condylocarpine (CCXV, see below) is not an enamine] in which it remains to complete one more ring between (3-16 and some point on the piperidine ring. Such a part structure is fully in accord with the appearance of the strongest peak in the mass spectrum of stemmadenine a t m/e 123, a peak which can be rationalized as the piperidine fragment t (117). Proof of the point of attachment of C-16 came from the conversion of stemmadenine into condylocarpine by direct oxidation of the hydrochloride with aqueous permanganate followed by heating, which gave formaldehyde as a by-product (117). The condylocarpine obtained had the same physical properties, including high positive rotation, as the natural material ( 116). Condylocarpine has the characteristic a-methyleneindoline carboxylic ester chromophore found also, for example, in akuammicine (CCXXV) and tabersonine (XCII, Section 11, 0), and which gives rise to the high rotation and t o IR-spectroscopic bands a t 6.0 p (conjugated C02Me) and 6.25 p (conjugated double bond), and

14, Aspidosperma

AND RELATED ALKALOIDS

46 1

NMR-absorption at 8.72 6 (N,-H) and 3.78 6 (COzCH3). NMR-spectroscopy also showed that the ethylidene group of stemmadenine was still present. The absorption in the 4.0 6 region (which may be assigned to protons allylic to one double bond and next to nitrogen, or to protons allylic to two double bonds) showed two single proton absorptions. Position 3 (partial structure CCXIX) is therefore substituted and can only be linked to C-7, while C-16 may be linked to positions 15 or 21 (117).

Clemmensen reduction of condylocarpine (CCXV) gave the tetrahydro derivative (CCXX) [note the strange reduction of an isolated double bond for which a mechanism has been proposed (1IS)] in whose mass spectrum could be seen the now familiar breakdown pattern to give fragments r and s characteristic of the dihydroaspidospermatidine skeleton (Section V, B) but in which r now bears the C-16 carbomethoxyl group. In particular, the position of s a t m/e 227 [as in dihydroaspidospermatidine (CCV) and not as in tetrahydrodecarbomethoxyakuammicine (CCIII)] fixes the position of the ethyl group as (3-14 and not C-20. The structure of condylocarpine is thus CCXV and of stemmadenine, CCXIII, with the oxidative interconversion passing through the intermediate CCXIV. Further confirmation of structure CCXV was sought by decarbomethoxylation of condylocarpine in strong hydrochloric acid (compare tabersonine, Section 11, 0, and akuammicine, Ref. 19) to the indolenine CCXXI followed by reduction t o the indoline CCXXII. This compound was expected to be identical to aspidospermatidine (CCIV) but it was not. The identity of the mass spectra of CCXXII and CCIV, however, show that the former does in fact have the structure shown and this was confirmed by the identity of the mass spectra of the dihydro derivatives of the two compounds, CCXXIII and CCV. The difference must exist in the stereochemistry a t C-19 (or less probably C-2) (118). In this connection, it is interesting t o note that condylocarpine has a positive rotation in contrast to the negative rotations normally observed in a-methyleneindoline alkaloids. This a t first indicated an “unnatural ” configuration for both this alkaloid and stemmadenine, which itself occurs together with “ unnatural ” ( + )-quebrachamine. The full absolute stereochemistry of condylocarpine was finally established by correlation with akuammicine (CCXXV) by way of its dihydro-derivative, tubotaiwine (CCXXIV, Section V, D ; Ref. 118a). The configurations of both condylocarpine and stemmadenine a t position 15 are in fact “natural,” t h e positive rotation of the former being attributed t o its inversion at C-7 with respect t o the akuammicine group alkaloids.

462

B. GILBERT

D. TUBOTAIWINE From the leaves of Pleioearpa tubicina, in addition to bases described in Sections 111,G, and VI, C, there has been isolated an alkaloid, tubotaiwine, of structure CCXXIV (119). The same substance also occurs in the root bark of Aspidosperrna lirnae (120). Tubotaiwine is dihydrocondylocarpine and like the parent base is decarboxylated by hydrochloric acid. The product, condyfoline (CCXXIII-A, [aID+ 348"), undergoes heat-catalyzed transformation to a mixture of its 20-epimer (CCXXIII-B, [a]=+ 31aa) and tubifoline (CCXXXIX, [a]=- 361", Section 1'1, C), the latter of known absolute configuration. The transformation proceeds by way of an intermediate analogous to CCXIV in which the 3-7 bond has opened. The stereochemistry at position 15is unaffected and, as it controls the configurations a t positions 3 and 7, the absolute configurations of condyfoline (CCXXIII-A) and hence of tubotaiwine (CCXXIV) and condylocarpine (CCXV) at these centers follow (118a).

E.

1~-METHOXY-14,19-DIHYDROCONDYLOCARPINE

Although the alkaloids of Aspidosperma populifolium are preponderantly ofthe aspidofractinine type, one of them (CCXXlV-A)was clearly excluded from this class by its IR- (ACHC1- 5.98 and 6.25 p, MeOzC--C=C) and UV- (Table 111)spectra. The latter was somewhat similar to that of condylocarpine (CCXV),the differences being accountable to an aromatic methoxyl group whose presence was indicated by the NMR-spectrum (OCH3, 3.74 6, coincident with the COzMe peak; 3 aromatic protons only). A further difference from condylocarpine and akuammicine lay in the presence of an ethyl (0.70 6, triplet) rather than ethylidene side chain. Membership to the condylocarpine rather than to the akuammicine group was established by mass spectral examination of the dihydro derivative (CCXXIV-B).The spectrum paralleled that of condylocarpine except in that the s peak appeared a t rnle 257 (227 + 30 due to the aromatic methoxyl), while the indolic peaks were also shifted upward by 30 mass units. The UV-spectra of CCXXIV-B and its N-acetyl derivative, CCXXIV-C, corresponded t o those of la, 1,2,3,4,4a-hexahydro-7methoxycarbazole and its N-acetyl derivative, respectively (120a), thus placing the methoxyl group in the 1 1 position (condylocarpine numbering). The alkaloid CCXXIV-A may thus be formulated as 11-methoxy14,19-dihydrocondylocarpine(48).

14. Aspidosperma

AND RELATED ALKALOIDS

463

VI. Alkaloids Related to Akuammicine A. INTRODUCTION Alkaloids with the skeleton of akuammicine (CCXXV) are widely distributed in the family Apocynaceae and in the genus Strychnos of the family Loganiaceae. Comparatively few members of this group have been isolated so far, however, from plants of the genera under discussion. Together with the preceding isomeric group, they differ from the aspidospermine-type alkaloids in possessing a C19 skeleton if we include the C-16 carboxyl, in contrast to tabersonine (XCII),for example, which has a C20 skeleton including the carboxyl carbon atom.

B. MOSSAMBINE Mossambine occurs with a number of other alkaloids (Section V, C) in Diplorrhyncus mossambicensis (121, 122, 116). Its high negative rotation (-498' in chloroform), UV- (A 230, 300 and 330 mp), and IR-spectra (conjugated COZMe, v, 1653 cm-1; conjugated double bond, v, 1603 cm-1) marked it as a member of the a-methyleneindoline carboxylic ester group of alkaloids. Elementary analysis and mass spectral molecular weight determination established the empirical formula, C20H22N302. A modified Kuhn-Roth determination (44)showed the presence of one C-methyl group, whereas after catalytic hydrogenation, the same determination showed C-ethyl. The hydrogenation product, 19,20-dihydromossambine (CCXXVIII), had the same UV-absorption as mossambine itself and so the alkaloid must contain a grouping, C=CHCH3, unconjugated with the main chromophore. Although mossambine was too insoluble for NMR-determination, examination of the spectra of its 2,16-dihydro and decarbomethoxy derivatives (see below) showed the vinyl proton quartet (5.53-5.55 6) and coupled allylic methyl doublet (1.66 6, J = 6 cjsec) typical of such an ethylidene group. Acetylation of mossambine gave a mono-0-acetate (CCXXVII, v, 1740, 1355 cm-1) whose NMR-spectrum exhibited a new single proton triplet in which there are also two at 4.8 6 due to the group C-CH(0Ac)-C hydrogen atoms vicinal to the acetate group. There are three known types of alkaloid which contain the cr-methyleneindoline carboxylic ester chromophore, those which are typified by tabersonine (XCII), akuammicine (CCXXV), and condylocarpine (CCXV), respectively. The first is excluded since it has a C20 skeleton as against the C19 skeleton of mossambine. The molecular formula of

464

B. GILBERT

C C XXV; R = H C C XXVI; R = O H CCXXVII: R = OAC

U

C C X X X V I I I ; R = H ; 1Sg - H CCXXIX; R = OH

CCXXXV; R = H CCXXX; R = O H

P R' = COzMe or Me

CCX X V III; R = OH CCX X X V II; R = H

CCXXXIII; R = O H

CCXXXI

CCXXXVI; R = H CC XXXIV; J-t = OH

9

CCXXXII

r R' = COzMe or Me

For compounds with R = H, t h e following stereochemistry h a s been established: At C-5, (2-6, 8, C-3, a ;or-H at (3-3, (2-15, (2-20.

14. Aspidosperrna

AND RELATED ALKALOIDS

465

mossambine, however, was in accord with its being a hydroxy derivative of either of the other two compounds, both of which also possess an ethylidene group. Various other chemical similarities confirmed this supposition. For example, the conjugated double bond of mossambine could be selectively reduced with zinc and acid to give 2,16-dihydromossambine (CCXXIX), a dihydroindole (UV- and low rotation) unsubstituted in the aromatic ring (NMR-), while further catalytic reduction resulted in the uptake of two hydrogen atoms and the saturation of the ethylidene side chain. Vigorous acid hydrolysis of mossambine resulted in simultaneous decarboxylation with the production of an indolenine (CCXXXI, A, 220, 262 mp) which with alkaline borohydride underwent the reverse Mannich reaction and reduction to give an indole (CCXXXII, A, 227, 283, 289 mp). This series of reactions, typical of alkaloids of the types mentioned above (19, 123, 69), show that Nh is linked to the 8-indole position through one carbon atom, the sequence being pictured : 0

n c.

H+ N,=C--CCN,

__f

€IN,-C=C

C=Nt

+ HN*--C=C

CH-N,

Furthermore, lithium aluminum hydride reduction of mossambine (CCXXVI) followed the same course as with akuammicine (68), yielding a dihydroindole (A, 245, 301 mp) with an exocyclic methylene group (CCXXXIII,v, 1640 cm-1) in which the only remaining oxygen atom was that of the original hydroxyl group. Catalytic reduction of CCXXXIII gave the C-16 methyl derivative (CCXXXIV) (121, 122). The foregoing information does not distinguish between the akuammicine and condylocarpine skeletons. This distinction may be readily made as has been seen in the case of aspidospermatidine (Section V, B) by examination of the mass spectrum of a fully reduced derivative (115a). I n fact, comparison of the spectra of tetrahydromossambine (CCXXX) and tetrahydroakuammicine (CCXXXV) revealed a very similar pattern for both (Table V), the only important difference being that the r and s peaks were 16 units higher in the case of the mossambine derivative, showing that the alcoholic oxygen atom was present in both. If the ethyl side chain of tetrahydromossambine were in the 14 position as in tetrahydrocondylocarpine (CCXX), then the s fragment would contain it. The fact that, as in tetrahydroakuammicine, it does not, limits the ethyl group to position 20, the only carbon atom absent from s which could accommodate an ethylidene group in the original mossambine. The same conclusion may be reached from the spectra of the pair CCXXXIV and CCXXXVI. These results are also supported by the spectra of derivatives which retain the 2,16- and/or 19,20 double-bond (115a).

466

B . GILBERT

The secondary hydroxyl group present in mossambine can only be placed on C-5, (2-21, or C-14, since these are the only suitable atoms in common between the fragments r and s. Positions 5 and 21 are excluded because the hydroxyl is not eliminated during borohydride, zinc and acid, or lithium aluminum hydride reduction. Position 14 for this group is in full,accord with the NMR-absorption (above) and with other mass spectral data (absence in the spectra of CCXXX and CCXXXV of OH in the homolog of indolic fragment b which contains C-5, and in the spectra of CCXXIX and CCXXXVIII in fragment u which contains C-21). Mossambine thus has structure CCXXVI (122), the stereochemistry remaining a t present undefined. The fact that mossambine is, like akuammicine, levorotatory suggests, however, that the configuration at positions 7, 3, and 15 is the same as in the latter which has been related directly to the Wieland-Gumlich aldehyde (68, 77, 123, 124). It is of interest that Nb-metho salts of mossambine sometimes exhibit a carbonyl band in the IR-spectrum (1698 cm-I), pointing to a possible Hofmann cleavage (122, 121).

C. NORFLUOROCURARINE, DIHYDROAKUAMMICINE, TUBIFOLINE, AND TUBIFOLIDINE Among the seven alkaloids that were isolated from Diplorrhyncus mossumbicensis was a base C19HzoN20 with unusually high rotation ( - 12qO"in chloroform). Its carbonyl activity and UV-spectrum (A, 242, 299, 360 mp) were reminiscent of fluorocurarine (Chapter 15, Refs. 133-137) and comparison of the methochloride of the new base with fluorocurarine demonstrated complete identity. The alkaloid is therefore norfluorocurarine (CCXL, 116, 132). Three related alkaloids have been found in the leaves of Pleiocurpa tubicina. One is the known compound 19,20-dihydroakuammicine (CCXXXVII, 70, 69, 137a), while the other two, tubifoline and tubifolidine, are respectively its decarbomethoxy derivative, the indolenine, CCXXXIX, and the corresponding dihpdroindole, CCIII (119). A compound strychene of structure CCXXXIX has been prepared from dihydrodeoxyisostrychnine (67, 125).

D. COMPACTINERVINE The bark of Aspidosperma cornpnctinervium contains an alkaloid, compactinervine, whose high negative rotation, UV- (A 237, 297, 331) and IR-spectra (conjugated COZMe, 6.04 p ; conjugated double bond,

14. Aspidosperma

AND RELATED ALKALOIDS

467

6.30 p) showed that, like mossambine, it belongs to the akuammicine class. Although solvation precluded precise analyses, a molecular ion corresponding t o the empirical formula CzoH24N204 could be obtained when the alkaloid was introduced directly into the ion source of the mass spectrometer. Thus there are two oxygen atoms to account for, besides those present in the carbomethoxyl function; and both of these were shown to be alcoholic by smooth acetylation to a diacetate (CCXLI-A). The vicinal nature of the two hydroxyls was further shown by reaction with periodate which gave acetaldehyde (125a). Several characteristic reactions of akuammicine-type alkaloids are exhibited by compactinervine (CCXLI). For example, the conjugated double bond is reduced by zinc and acid to give dihydrocompactinervine (CCXLI-B), whose mass spectrum corresponds to that of tetrahydroakuammicine (CCXXXV), the s peak a t m/e 199 indicating a C-20 side chain (see Section V, B). Furthermore, treatment with strong acid results in decarboxylation to the indolenine CCXLI-C which undergoes the reverse Mannich condensation and reduction with alkaline borohydride to an indole (CCXLI-D). Lithium aluminum hydride reduces the m$-unsaturated ester grouping of compactinervine to an exocyclic methylene derivative (CCXLI-E) analogous to the corresponding reduction product of akuammicine (cf. Section VI, B). The mass spectra of both this product (CCXLI-E) and its trideuterio analog (CCXLI-F, obtained with LiAlD4) resembled those of the corresponding two akuammicine derivatives ( 125a). From these facts it was certain that compactinervine (CCXLI)had the dihydroakuammicine skeleton and it remained to place the two hydroxyl groups. On biogenetic grounds, positions 19 and 20 are attractive, since these are oxygenated in the related alkaloids echitamidine (CCXLII, 126) and lochneridine (CCXLIII, 127). That these positions are the true ones is shown by two observations. First, the r peak in the mass spectrum of dihydrocompactinervine (CCXLI-B) appears at m/e 228, 32 units higher than in the analogous tetrahydroakuammicine (CCXXXV), while the s peak is identical in the two spectra. The two oxygen atoms responsible for this shift are thus limited t o positions 15, 18, 19, 20, and 21, the only ones present in r but absent from s (see Section VI, B). Second, compactinervine (CCXLI) shows a methyl doublet at 1.15 in the NMR-spectrum analogous t o that of echitamidine (CCXLII) and due to the grouping CH(0H)CHs(125a). An attempt to synthesize compactinervine (CCXLI) by osmium tetroxide hydroxylation of akuammicine (CCXXV) failed since the product, CCXLI-H, differed stereochemically. However, both diols, CCXLI-H and compactinervine, gave the same 19-ketone CCXLI-G on oxidation with

468

B . GILBERT

the Jones reagent, and the only stereochemical difference was therefore at C-19 (125a). The complete absolute stereochemistry of compactinervine follows from this observation, for the vic-glycol, CCXLI-H, derived from akuammicine of known absolute configuration, must have both hydroxyl groups E . The 20-hydroxyl of compactinervine is therefore a , and the 19-hydroxyl/3. The equatorial tertiary hydroxyl in position 20 is note-

C‘CSLI; Ii = H C C X L I - A ; R = Ac

CCS1,I-c

CCXL-D

14. Aspidosperrna

AND RELATED ALKALOIDS

469

worthy in that it acetylates readily, but resists dehydration, in contrast to the 20 p-axial hydroxyl of lochneridine (CCXLIII) which dehydrates on attempted acetylation (125a, 127).

VII. The Uleine Group A. INTRODUCTION Alkaloids of this group have been discovered in the genera Aspidosperma, Tabernaernontana, Excavatia, and Ochrosia ; in the first they are well distributed, uleine itself being the major alkaloid of at least eight species. Uleine represents a link between the other more highly aromatic members of the group and the condylocarpine-type bases from which it differs in the lack of the 5,6 tryptamine bridge. Wenkert has suggested a biosynthetic route for the formation of this alkaloid (115).

B. ULEINEAND RELATED ALKALOIDS Uleine (CCXLV) was first isolated from Aspidosperrna ulei (138, Volume VII, p. 129) and has since been found in a number of other species (see Table I and Refs. 140, 141, 49, 48). It has the composition C18H22N2, and contains a double bond which is readily hydrogenated to give dihydrouleine (CCXLVI), a compound with the indole chromophore. The UV-spectrum of uleine (Table 111)showed that the double bond was conjugated with the indole nucleus; the similarity of the absorption to that of diacetylallocinchonamine (CCXLVII) suggested that one end of the double bond was linked (0 the a-indolic position. The NMR-spectrum showed that the double bond lay in an exocyclic methylene group (two vinyl singlets a t 5.27 and 4.98 6, absent in the spectrum of dihydrouleine) and established the presence of four aromatic protons and an indole NH grouping (Table IV). These results were in accord with the IR-absorption (v, 3440, N H ; 1635 and 877, =CH2; 740 em-1, 0-disubstituted benzene). An N-methyl (NMR-)and a C-ethyl group (modified Kuhn-Roth, Ref. 44, and NMR-) were also found to be present (138, 142). Uleine methiodide underwent a facile Hofmann degradation t o give an optically inactive compound (CCXLVIII), which retained the two nitrogen atoms and had carbazole UV-absorption. A further Hofinann degradation eliminated N1,and gave a product (CCXLIX) with somewhat extended carbazole absorption, which underwent ready catalytic reduction with saturation of one double bond to give a substituted

470

B. GILBERT

carbazole (CCL) which could also be obtained directly from uleine or dihydrouleine by selenium dehydrogenation (138). The second Hofmann product (CCXLIX) was recognized as a vinylcarbazole by oxidative cleavage of the double bond with osmium tetroxide and then periodate,

cCSL\-l

CCSLVlIl

I

CCLIV

2,3-saturated

CCXLVII

I

\

CCXLlS

CCL

CCLV 1. LIIS04 2. estcrification

CCLl

('CLII

CCLVI; R = H CCLVII; R = Ale

which gave a yellow aldehyde (CCLI) (v, CO, 1675 cm-1; NMR, CHO, 10.72 6). Comparison of the UV-spectrum of this aldehyde with those of 1-, 2 , 3-, and 4-formylcarbazole (143) left no doubt that it was a 2formylcarbazole. Furthermore, the NMR spectrum of CCLI showed the

14. Aspidosperma

AND RELATED ALKALOI1)S

47 1

presence of five aromatic protons, an aromatic methyl (2.75 S), and an aromatic ethyl group (quartet at 3.15 6, triplet a t 1.35 a), these two substituents occupying two of the remaining three positions ‘in the carbazole C ring. The location of the methyl group is clearly position 1 since it derives from the original exocyclic methylene group, and that of the ethyl substituent follows from the NMR spectrum of uleine itself which exhibits a one proton doublet a t 4.07 6 (J = 2 c/sec) which must be both “allylic ” and next to nitrogen (compare condylocarpine, Section v , C) and have only one neighboring proton, a requirement compatible only with strucure CCXLV for uleine and hence structure CCLI for the aldehyde (142). The structure of CCLI was confirmed by decarbonylation to l-methyl3-ethylcarbazole (CCLII) which was synthesized as follows. Ethyl 2-oxalylbutyrate and the Mannich base, ethyl-p-dimethylaminoethyl ketone condensed t o the cyclohexenone, CCLIII, which was reduced with zinc to CCLIV. The Fischer indole synthesis was then applied to this ketone and led via the phenylhydrazone, CCLV, to the tetrahydrocarbazole carboxylic acid, CCLVI, whose methyl ester, CCLVII, was dehydrogenated over palladium charcoal to 1-methyl-3-ethylcarbazole (CCLII) identical with the degradation product. The chemistry of uleine (CCXLV) parallels that of gramine, and the methiodide (CCLVII-D) suffers facile nucleophilic attack a t the benzylic 4-position with cleavage of the C-4-N bond. Thus with methoxide

m/e209 R = H m/e223 R = Me

m/e 180

m/e222 R = E t m/e 207 R = CH2’

m/e208 m/e 194

m/e 237

Principal mass spectral fragments from uleine (CCXLV)

R = Me R =H

CCLVII-D

CCLVII-A CCXLV; R CCXLVI; R CCLVII-P; R CCLVII-B; R CCLVII-C; R CCLVII-Q; R

= Me = Me, 1,l‘-dihydro = Me, 1,l’-dideuterio

=H = AC = Et

c R =H R = Me

td

J

t

m/e 209 m/e 223

Ri

CCXLVIII; R = Me CCLVII-J; R = AC CCLVII-K; R = H

CCLVII-E CCLVII-L CCLVII-F CCLVII-M CCLVII-G CCLVII-N CCLVII-H

H

Ri

Rz

OM0 OM0

Me

CN

Me Me, 1,l’-dihydro Me Me, 1,l’-dihydro

CN H H-

( -;>

Me,1,l’-dihydro

Ac

L/

CCLVII-I CCLVII-0

OMe OMe

Ac Ac, 1,l’-dihydro

14. Aspidosperma AND

RELATED ALKALOIDS

473

ion, cyanide ion, and lithium aluminum hydride the tetrahydrocarbazoles, CCLVII-E, CCLVII-F, and CCLVII-G, respectively, are obtained. The 4-methoxy-compound, CCLVII-E, readily loses methanol to give the same carbazole (CCXLVIII) as is obtained by Hofmann degradation of uleine. A similar nucleophilic attack a t position 4 results in the formation of the pyridinium salt, CCLVII-H, in which the pyridinium group can be readily displaced by methoxyl (warm methanol), the resulting 4-methoxytetrahydrocarbazole, CCLVII-I, being readily converted via CCLVII-J and -K to the above carbazole, CCXLVIII (143a). Lithium aluminum hydride reduction of uleine (CCXLV) gives dihydrouleine (CCXLVI), for which reaction a possible mechanism has been proposed ( 143a). Dihydrouleine methiodide behaves toward nucleophilic reagents similarly to uleine methiodide and the corresponding series of tetrahydrocarbazoles, CCLVII-L, CCLVII-MI, and CCLVII-N, have been prepared. The reaction of dihydrouleine with acetic anhydride and pyridine followed by methoxide also parallels that of uleine with the formation of CCLVII-0 (143a). Prom an examination of the mass spectra of uleine, dihydro- and dideuteriouleine (CCLVII-P, LiAlD4 on uleine), and the N-acetyl (CCLVII-C) and N-ethyl analogs (CCLVII-Q, LiAlH4 on CCLVII-C) of uleine much information has been accumulated on the probable breakdown path of these alkaloids in the mass spectrometer (see formulas). An important intermediate species is undoubtedly the carbazole, CCLVII-K, whose mass spectrum is practically identical with that of the isomeric uleine (143a).

CCLVIL-U

Among compounds related to uleine that have been isolated from A . dasycarpon are N-noruleine (CCLVII-B),dasycarpidone (CCLVII-A), the corresponding alcohol, dasycarpidol, N-nordasycarpidone, and 1 , l ’ dihydro- 1’-hydroxyuleine. Dasycarpidone may be obtained from uleine by ozonolysis, or from dasycarpidol by oxidation with chromium trioxide has been synthesized from in pyridine. 1,1’-Dihydro-1’-hydroxyuleine uleine by hydroboration (143b). Apparicine, an alkaloid of novel skeleton present in A . dasycarpon and several other species, has been shown t o have the structure CCLVII-U (37).

474

B. OILBEBT

C. OLIVACINE,DIHYDROOLIVACINE, AND GUATAMBUINE The yellow optically inactive alkaloid olivacine (CCLVIII) was first isolated from Aspidosperma olivaceum ( 140)and has subsequently been encountered in many Aspidosperma species (Table I,Refs. 141, 145,147, 48) as well as in Tubernuemontana psychotrifolia H.B.K. (148). Its empirical formula, C17H14N2, and complex UV-spectrum indicated a highly aromatic structure ; the alteration of UV-absorption in acid showed that the basic nitrogen atom was involved in the chromophore (140,149).Olivacine methiodide gave a red anhydronium base (CCLIX) with alkali (145).Meanwhile, another base, guatambuine (U-alkaloid-C, CCLX), C18H20N2, which had first been found in A. ulei (139)and which is present in several species often side by side with olivacine (Table I), was shown to be the N-methyltetrahydro derivative of olivacine both by dehydrogenation of the optically active base and by reduction of olivacine methiodide catalytically or with borohydride to racemic guatambuine (141,149).Guatambuine, remarkable in that both enantiomers have been encountered in the same species (147),is a much stronger base than olivacine (Table V I ) and has carbazole UV-absorption, the identity of the acid and neutral spectra showing that the reduced ring was that containing N1,which was now no longer linked to the chromophore. Analogy with ellipticine (Section VII, D) and comparison of the pKi and UV-spectroscopic data of a number of pyridocarbazoles then (151,152) demonstrated that olivacine was a lOH-pyrido-4,3-b-carbazole and led to the proposal of structure CCLVIII for this alkaloid and hence CCLX for guatambuine (149,153). The correctness of these proposals was readily shown by Hofmann degradation of guatambuine to CCLXI and thence to the same series, CCXLVIII, CCXLIX, and CCL, which had resulted from the degradation of uleine (Section VII, B). Alternative proof of the structure of olivacine was obtained by two syntheses. I n the first of these (150) the starting material, 2-amino-6cyanotoluene, represented ring C, and after transformation to the corresponding phenylhydrazine (CCLXII), rings A and B were built on by a Fischer (Borsche) indole synthesis to give the 2-cyano-1-methyl5,6,7,8-tetrahydrocarbazole(CCLXIII). Difficulty with the dehydrogenation of this compound led to its conversion to the corresponding ester (CCLXV) which smoothly dehydrogenated to the carbazole ester, CCLXVI, in which the carbomethoxyl group provides the starting point for building up ring D. The best method of homologation of the ester was found to be the Arndt-Eistert synthesis via the acid chloride, CCLXVIII, and diazoketone, CCLXIX, to the amide, CCLXX. This amide was then (CCLXXII) by transformed to 2-(~-aminoethyl)-~-methylcarbazole

CCLVIIl

CCLXXIV

I < CCL XXIII

CHzCHzNHCOCH3

T

IIofmann II

2.

CCLXI

CCLXX

I

I .1

CCLXIX

I

CCLXVIII

CHzCHzCOCH3

C C LX X X I

T

H2, I’d-C

CH=CHCOCH3

4

MepC10, KOH

CCLXXX

CHO

C C LX X I X

CH3

SOCh, DMF

I

CCLXVIl

COzH

CCLXVI

COZEt

I Me C CLX I I

CCLXIII; R = CN CCLXIV; R = COzH CCLXV: R = COpEt

/I

t-

COCl

T

CCL

C C LX X X l I

COCHNz

t

CCX LI X

II

CHzCHzC. CHI

CHzCONHz

1.

CCXLVIII

NOH CCLXXXlll

Ed,

CsHs

T

CCLXXV; R = Hz CCLXXVI; R = =CHOH CCLXXVII: R = =CHOPri

475

y

C C LX X V I I I

476

B. GILBERT

dehydration and hydrogenation of the intermediate cyanide, and thence into dihydroolivacine (CCLXXIV) by a Zischler-Napieralski ring closure of the N-acetyl derivative (CCLXXIII). Dehydrogenation of CCLXXIV gave olivacine (150).The second synthesis ( 154)began with the known l-keto-1,2,3,4-tetrahydrocarbazole (CCLXXV) in which rings A, B, and C are ready-formed. The initial atom of ring D was best introduced as a hydroxymethylene group a t position 2 by base-catalyzed condensation with ethyl formate. The hydroxymethylene group of CCLXXVI was then protected as its isopropyl ether (CCLXXVII) while a methyl group was introduced a t position 1 by reaction of the ketonic carbonyl with methyl lithium; loss of water and the protecting group gave the dihydrocarbazole aldehyde, CCLXXVIII. The aldehyde could be very readily dehydrogenated (with disproportionation and simultaneous production of CCLXXIX)to the carbazole aldehyde, CCLXXX. The building up of ring D was now achieved by a route quite different from that of the previous synthesis. Condensation of the aldehyde, CCLXXX, with acetone gave the a,P-unsaturated ketone, CCLXXXI, which was catalytically reduced and transformed to the oxime, CCLXXXIII. This oxime was subjected to a simultaneous Beckmann rearrangement and Bischler-Napieralski condensation to give dihydroolivacine which was dehydrogenated as before to olivacine itself (154). These two syntheses also constitute syntheses of racemic guatambuine. This alkaloid (CCLX) was also obtained by catalytic reduction of the methiodide of dihydroolivacine (CCLXXIV) (150). I n addition, the preparation of N-demethylguatambuine was described (150),as well as alternative routes to the aldehyde (CCLXXX) and corresponding acid (CCLXVII) (154).Dihydroolivacine occurs naturally in A . ulei and may be separated from the accompanying dihydroellipticine (Section VII, D) by thin layer chromatography (155,,139,149,150).Compounds with this chromophore may be recognized by the appearance of a new intense absorption peak in their acid UV-spectra a t approximately 377 nip as well as strong absorption in the IR-spectrum between 6 and 7 p (139, 149,150, 158). Recent investigation of Aspidosperma nigricans has resulted in the isolation of olivacine N-oxide (CCLXXXIII-A).This somewhat unstable compound may be reduced to olivacine (CCLVIII) by brief treatment with zinc and mineral acid. The N-oxide (CCLXXXIII-A) may be prepared from olivacine by perbenzoic oxidation. It is also the initial product of the action of hydrogen peroxide in warm acetic acid on olivacine, a reaction which yields as one of its final products the red isomeric amide, 2-oxo-2,3 dihydroolivacine (CCLXXXIII-B). The N oxide may be distinguished from the latter not only by its UV-spectrum

14. Aspidosperma

477

AND RELATED ALKALOIDS

(A,, 236,252,300,311,330,and 345 m p ; E, 16350, 14280,51490,65420, 5420, 5300) which differs little from that of olivacine, but also by the mass spectrum in which the base peak a t M-16 (m/e 246) corresponds to loss of one oxygen atom from the molecular ion to give a positively charged olivacine molecule-ion (48). Me

Me I A - A , H \ N H

Me

CCLXXXIII-A

CCLXXXIII-B

D. ELLIPTICINE, DIHYDROELLIPTIC~NE, AND N-METHYLTETRAHYDROELLIPTICINE Ellipticine (CCLXXXIV) was first isolated from Ochrosia elliptica and 0. sandwicensis A.DC. (156) and subsequently from Aspidosperrna subincanurn (see note 3, Table I) (157, 158) as well as from other plants (Table I, Ref. 161). I t s UV-spectrum is complex and very similar to that of olivacine (CCLVIII), and reduction of its methiodide with borohydrid2 gave A'-methyltetrahydroellipticine (CCLXXXV) ( 156) which had been previously isolated from A . ulei (U-alkaloid-B, 139) and also from A . subincanurn. Natural N-methyltetrahydroellipticinewas optically inactive and exhibited a carbazole UV-spectrum which, like that of the optically active guatambuine (CCLX), was unaffected by acid, thus showing that the basic nitrogen atom lay in a reduced ring and was insulated from tkie carbazole chromophore (139). The structure of ellipticine and hence of its tetrahydro derivative was established by a remarkable three-step synthesis from indole. Condensation with 3-acetylpyridine in the presence of zinc chloride gave the bisind olylpyridylethane (CCLXXXVI) in which i t remains only to form the C ring. Reductive acetylation of the pyridine ring with zinc and acetic anhydride yielded the N,C-diacetyldihydropyridine derivative (CCLXXXVII, A, CO, 5.80, 6.05 p ) in which the remaining two carbon atoms of ring C have been introduced. Pyrolysis of CCLXXXVII gave ellipticine (CCLXXXIV) in 2% yield (157). The result was confirmed by a second synthesis. The dimethylated C ring was built onto indole by condensation of hexane-2,5-dione with the reactive indolic 2,3 positions. Formylation of the resulting 1,4-dimethylcarbazole (CCLXXXVIII)with N-methylformanilide proceeded preferentially in the 3-position to give the aldehyde CCLXXXIX whose

47 8 B. GILBERT

0

\

J-Jx

G x

V v

1

Zn,AcaO

T

Ha, N :

Me I

CCLXXXVI

Me

ccxc

T

Me

I

CCLXXXVIII

CCLXXXIX

T structure was established by Wolff-Kishner reduction to 1,3,4-trimethylcarbazole. Condensation of CCLXXXIX with 2,2-diethoxyethylamine gave the Schiff’s base (CCXC) and, although this compound resisted attempts a t cyclization, its dihydro derivative (CCXCI)was successfully cyclized and dehydrogenated to ellipticine (CCLXXXIV) (159, 160). An independent synthesis of ellipticine follows, in its initial stages, the first olivacine synthesis reported in Section VTI, C. Instead of the monomethylcyanophenylhydrazine (CCLXII), a corresponding dimethyl (CCXCI-A), was emcompound, 3-cyano-2,5-dimethylphenylhydrazine ployed as starting material. As far as the ester, methyl 1,kdimethylcarbazole-2-carboxylate (CCXCI-B), the two syntheses are parailel. Ring D was then built up, however, by the series of reactions, COzMe

CHzOH

+CHO

___f

-CH=CH-NOz

--+

CHzCHzNHz

The final ring closure again paralleled the olivacine synthesis passing through the intermediate N-formyl rather than the N-acetyl, amine, to give 1,2-dihydroellipticine (CCXCIII) from which ellipticine was obtained by palladium dehydrogenation (16Oa). When the tertiary bases had been removed from the extract of Peruvian A . subincanum, the quaternary bases were extracted with butanol and crystallized as their nitrates. Two salts were separated by preparative paper chromatography, and one of these was recognized as ellipticine methonitrate. The second had a UV-spectrum similar to that of the 1,%dihydropyridocarbazole [e.g., 1,2-dihydroolivacine (CCLXXIV), Section VII, C] chromophore in acid solution. Furthermore, reduction with sodium borohydride gave N-methyltetrahydroellipticine (CCLXXXV). The second quaternary alkaloid was thus N-methyl-l,2-dihydroellipticine (CCXCII) in the form of its nitrate; this was confirmed by oxidation of the tetrahydro compound (CCLXXXV) with mercuric acetate and acidification with nitric acid to give the same salt. The corresponding tertiary base (CCXCIII) was also isolated and its structure confirmed both by conversion into the quaternary methochloride (CCXCII chloride) and by synthesis from ellipticine by reduction to the air-sensitive tetrahydroellipticine (CCXCIV) which gave 1,2-dihydroeIlipticine by dehydrogenation with 0,lO-phenanthraquinone (158). Dihydroellipticine (CCXCIII) was first encountered admixed with dihydroolivacine in A . uZei (U-alkaloid-D, 139, 155). Among other alkaloids of Ochrosia species there has been isolated a methoxyellipticine (156, 161).

Me

Me

CCXCI-A

Me

I

1. KOH/glycol 2. CHeNz

1. EtOCHO, 120'

i

2. POCIS, xylene

ccxcIII

CCXCI-B

482

33. GILBERT

VIII. Tetrahydro f3-Carboline and Related Alkaloids A. YOHIMBINE AND TETRAHYDROALSTONINE DERIVATIVES A number of these alkaloids, which are widely encountered in the family Rubiaceae and in the genus RauwolJa of the family Apocynaceae, have also been found in the genera under discussion. The occurrence of yohimbine (CCXCV),,L?-yohimbine(CCXCVI), and reserpine (CCXCVII) is recorded in Table I . I n addition, an 1l-methoxyyohimbine (CCXCVIII) has been isolated from Aspidosperma oblongum and its structure determined by mass and UV-spectrometry (see below, and Ref. 162). An alkaloid, poweridine, formulated as 17-0-acetyl-1l-methoxyyohimbine (CCXCIX), occurs in Ochrosia poweri and was shown to belong t o the yohimbine class by dehydrogenation to 7-hydroxyyobyrine (CCC)whose formation, together with the UV-spectrum of poweridine, establishes the position of the original methoxyl group. The preparation of a ,L?-lactone (CCCII, v, CO, 1803 cm-1) by dehydration of the 11-methoxyyohimbic acid (CCCI) from poweridine indicated that the hydroxyl and carbomethoxyl groups of this acid were vicinal and cis. The remainder of the stereochemistry remains a t present unknown (161). Among the yohimbine-like alkaloids with a heterocyclic ring E the most widespread is isoreserpiline (CCCV) whose methochloride has also been &countered. The occurrence of this alkaloid and of aricine (CCCIII), reserpinine (CCCIV),reserpiline (CCCLVI), and ajmalicine (CCCXVII), is recorded in Table I. A 5,6-dimethoxyindole related to isoreserpiline is elliptamine which occurs in four Ochrosia species as well as in Excavatia coccinea (161). A representative of the group in which ring E is open is dihydrocorynantheol (CCCVI), which occurs in two Aspidosperma species (163, 48). The methochloride of this compound has been isolated from Hunteria eburnea ( 164). 10-Methoxydihydrocorynantheol(CCCVIA) and a 19,20-dehydro derivative (CCCVI-B) have also been encountered in Aspidosperma discolor (113d) and other Aspidosperma species (48). Alkaloids which could be identical with CCCVI, CCCVI-A, and CCCVI-B as well as a dehydro derivative of CCCVI and a series of four similar bases bearing a 16-carbomethoxyl group have been found in A . oblongum (164a). The identification of alkaloids of this type has been greatly facilitated by the introduction of mass spectrometry (165,162,163).By examination of a large number of derivatives, the breakdown pattern has been established, the principal peaks being represented by the fragments v and w which are formed by a cyclic transfer of electrons in ring D

ai

OR3

Ri Rz R3 R4 Stereochemistry CCXCV H Me H H Y CCXCVI H Me H H P-y CCXCVII OMe Me Me OTMB 3-epi-a-Y CCXCVIII OMe Me H H unknown CCXCIX OMe Me Ac H unknown CCCI OMe H H H unknown Stereochemistry (substituent or angular H ) : Y = 3a,15a,l6a,17a,20P. /3-Y = 3~t,15a,16~(,17P,20/3. 3-epi-a-Y = 3/3,15a,16/3,17a,1813,ZO~t.

0

ccc

Q

CCCII

.U

bH20H

Ri CCCVI; R = H CCCVI-A; R = OMe* CCCVI-B; R = OMe, 19,20-dehydro*

* Stereochemistry unknown.

Rz

CCCIlI OMe H CCCIV H OMe CCCV OMe OMe

$

z1

484

B. GILBERT

(CCXCV,p. 486). These fragments are accompanied bv x and y (reverse Diels-Alder cleavage of ring C, formula CCCXVI) and all these four peaks occur in all yohimbine-type alkaloids whether ring E is homocyclic, heterocyclic, or open (165, 163). I n addition, there is always present in the mass spectra of this group a strong M-1 peak which is largely formed by the loss of the C-3 hydrogen atom. I n the heterocyclic ring E alkaloids, there is an additional indolic peak which has been assigned structure z (165).The described combination of peaks is not found in indole alkaloids based on the sarpagine, ibogamine, or eburnamine skeletons which each furnish a distinctive pattern (166, 167, 168, 169, 170, 51). The structures assigned to the fragments observed are based on the following evidence (165, 171). 1. The M - 1 Peak:

Fifty per cent of the M-1 peak is found a t M-2 in the spectra of 3-deuterioyohimbine (CCCVII, prepared by NaBD4 reduction of 3,4-dehydroyohimbine perchlorate) and of 3,5,6-trideuterioajmalicine (CCCIX, prepared by NaBD4 reduction of serpentine hydrochloride), so that onehslf of the hydrogen lost in the formation of this peak comes from C-3 or, in the case of ajmalicine, from C-3 and C-6. 2. Peaks v, w,x, and y Incorporate the Indolic Portion of the Molecule These four peaks remain invariable a t m/e 170, 169, 184, and 156 respectively when alterations are made in ring E as in the series, CCXCV, CCCX, CCCXI, CCCXII, and CCCXIII, and they do- not therefore contain atoms from this ring. The addition of substituents at positions 1, 3, 10, and 11 results in a corresponding increment in the mass of these fragments as is seen in the spectra of CCCXIV, CCCXV, and CCCVII (Table V). The importance of these peaks is explained by the ready cleavage of the allylically activated 3,14 bond.

3. The Structure of Peaks v and w Peak v, already shown to contain the indole residue, contains in addition carbon atoms 3,5, and 6, for in the spectrum of 3,5,6-trideuterioajmalicine (CCCIX) all three deuterium atoms are retained in this fragment. I n the spectrum of 3,14-dideuterioajmalicine (CCCVIII),prepared from serpentine by the successive action of NaBH4, HC1, and NaBD4, however, only one deuterium atom is retained in v which therefore does not contain C-14. Fragment w is derived from v by loss of one hydrogen from either C-5 or C-6 and this is shown by the fact that in the spectrum of 3,5,6-trideuterioajmalicine (CCCIX) w is split into two peaks, one in which one, and one in which two deuterium atoms have been retained.

14. Aspidosperma

AND RELATED ALKALOIDS

485

As the C-3 atom is always retained, the partial loss must occur from C-5 or C-6, which would seem t o indicate that there is no preference for the loss of deuterium over hydrogen in this case (compare 19-deuterio-17methoxyquebrachamine, Section 11, B). 4. The Xtructure of Fragment x I n the spectrum of 3,5,6-trideuterioajmalicine(CCCIX), all three deuterium atoms are retained, but in that of 3,14-dideuterioajmalicine (CCCVIII) only one deuterium atom appears in x. This fragment must therefore contain C-21 since it cannot contain C-14. 5. The Structure of Fragment y C-3 is present in this fragment (2 above) and from the spectrum of 3,14-dideuterioajmalicine (CCCVIII) it is seen that C-14 is also incorporated. Fragment y only retains two of the deuterium atoms of 3,5,6trideuterioajmalicine and therefore cannot contain (3-5. F'ragment y represents the strongest indolic peak in the spectra of the hetero ring E alkaloids because in these the 14,15 bond in the intermediate aa is doubly allylieally activated. The spectra may be used to distinguish the two most common stereochemical forms, those of ajmalicine (CCCXVII) and tetrahydroalstonine (CCCXVIII), since in the former the peak x is more intense than w or w, while in the latter and in yohimbine it is weaker. Examples are seen in the spectra of CCCIII-CCCV and CCCXVII-CCCXIX (Table V. Refs. 165, 37).

B.

NORMACUSINE-B,

POLYNEURIDINE, AND AKUAMMIDINE

Notmacusine-B was first obtained by the thermal decomposition of its quaternary methochloride, macusine-B, which was isolated from Strychnos toxifera." It was found to be identical with 10-deoxysarpagine (CCCXX) and its N,-methyldihydro derivative was identical with deoxyisoajmalol-B (CCCXXIII) (172, 173), thus establishing its structure and absolute stereochemistry (177, see Chapter 22). Subsequently, normacusine-B has been found in Diplorrhyncus mossambicensis (named tombozine, 116, 121, 122) and in Aspidosperma polyneuron (165). Another alkaloid of A. polyneuron, polyneuridine, C21H24N203, differs from normacusine-B by C02CH2. It contains a carbomethoxyl group as was shown by hydrolysis to the acid, CCCXXIV, which was

* Maciisine-R also occurs in A . polyneuron (173a).

I

Rz

Rz

Rz

2,

w

X

R4 Ri CCXCV H CCCX H CCCXI H CCCXII H CCCXIII H CCCXIV OMe CCCXV H

Rz Rs

R4

H COzMe OH H COzMe H H H =o H =O H H COzMe =O H H =O Me H H

Ri

Stereochemistry

Y a-Y allo-Y

Y

H R3

Y Y

R1 Rz R3 R4 CCCVII H CCCVIII H CCCIX D

H H D

D D D

H D H

Ri CCCXVII H CCCXVIII H CCCXIX H CCCIV H CCCIII OMe cccv OMe

Rz H H OMe OMe H OMe

Stereochemistry A T A T T T

Stereochemistry (substituent OT angular H ) : Y = 3a. 15a, 16e, 17a, 208. a-Y = 3a, 15a, 168, (17a), 20a. d o - Y = 3a, 15a, (168, 17/3), 20a. A z a 3a, 15a, 19a, 208. T = 3a, 15a, 19a, 20a.

488

2

P;

0,

\

/

U

P

Y

\

/

xEr

B . GILBERT

t

RI R2 R3 CCCXXIV H CHzOH CO2H CCCXXV H CH~OAC CO2Me CCCXXXVI Me CHzOH COsMe HOCH2,

CCCXX; R =H CCCXXVIII; R = Ac

CCCXXXII

,C02Me

R3,

COOMe

I CCCXXXVIII

I

cccxxxv11 CCCXXIII CCCXXIX CCCXXXIV CCUXXXV

Ri R2 H Me H H H Me OMe H

R3 H CHzOH H H

* 2O.m-Et;t 20,P-Et

bb

m/e 168

tu

m/e 169

cc

m/e249

I

dd rnle223

R4 CHzOH* C02Me Met Me

490

B. GILBERT

isolated as the hydrochloride and remethylated to the parent alkaloid. Polyneuridine also formed a mono-0-acetate (CCCXXV), demonstrating the presence of an alcoholic hydroxyl group. The possibility thus existed that polyneuridine was a carboinethoxy derivative of normacusine-B and this was supported by lithium aluminum hydride reduction to the diol, CCCXXVI, whose diacetate, CCCXXVII. was sufficiently soluble in deuteriochloroform for NMR-spectroscopic comparison with the monoacetate (CCCXXVIII) of normacusine-B. In fact a strong similarity was observed, the only major difference between the two spectra being the presence in that of CCCXXVII of absorption due to two acetate methyl groups (1.92 and 2.03 6) and to four protons corresponding to two CHzOAc groups while normacusine-B acetate showed absorption due to only one such group. The four aromatic protons, indole NH, and ethylidene group of normacusine-B were all also present in polyneuridine, the latter being confirmed by hydrogenation to dihydropolyneuridine (CCCXXIX). Meanwhile, comparison of the mass spectra of polyneuridine and normacusine-B had also revealed a strong similarity between the two alkaloids. Both exhibited M-31 peaks, and as this is only attributable to loss of the primary alcoholic function as CHzOH in normacusine-B, it was reasonable to assume the presence of such a grouping in polyneuridine. Moreover, both spectra contained intense peaks a t m/e 168 and 169 which remained invariable through a series of derivatives in which the carbomethoxyl group of polyneuridine and the alcoholic function and double bond of both alkaloids were modified. These P-carbolinic peaks compare with the v and w peaks of the yohimbine-type molecules which occur a t m/e 170 and 169, respectively, when the aromatic ring is unsubstituted. The lowering of one mass unit in each is attributable to the presence of the extra 5,16 bond which must be broken to produce the P-carboline fragments bb and w. These fragments are observed in sarpagine-type molecules (e.g., CCCXXXIV and CCCXXXV) substituted in the aromatic ring and on N, when the appropriate molecular weight shifts are observed (167), so that there is no doubt that they derive from the indolic part of the molecule (Table V). It was therefore assumed that polyneuridine (CCCXXI) possessed a sarpagine skeleton both the carbomethoxyl and primary alcoholic groupings being located on the C-16 atom, a supposition which was supported by the fact that the aldehyde (CCCXXIX-A) prepared by chromic acid oxidation of polyneuridine contained no a-hydrogen atom (failure to exchange with sodium deuteroxide in deuteromethanol), and by the recognition in the mass spectra of a fragment cc in which the two substituent)s and one skeletal carbon atom are missing (165, 165a, b).

14. Aspidosperrna

AND RELATED ALKALOIDS

491

Confirmation of the proposed structure was obtained by comparison with the isomeric alkaloid akuammidine (CCCXXII), which occurs in Picralima nitida (Volume VII, p. 122) and Rhazia stricta (named rhazine, 62), and whose structure has been established both chemically (174) and by X-ray diffraction (175). Both alkaloids undergo retroaldolization with loss of the primary alcohol group to an ester (CCCXXX or CCCXXXI) which, on lithium aluminum hydride reduction, furnishes normacusine-B (CCCXX) (174, 165). Also both alkaloids similarly reduced yield the same diol, akuammidinol (CCCXXVI).Thus, the structure CCCXXI for polyneuridine was established, including the absolute configuration of the skeleton (165). The orientation of the two C-16 substituents was established in two ways. First, the degradation of vincamedine (CCCXXXII) by chromic acid oxidation to the indolenine CCCXXXIII led, by way of the reverse Mannich condensation and reduction with alkaline borohydride (compare Sections 11, B, 0 and VI, B), to polyneuridine (CCCXXI) which cannot therefore have the alternative configuration (176). Second, the establishment of the C-16 orientation in akuammidine (175)required the structure CCCXXI for polyneuridine by difference. Polyneuridine is thus "normacusine-A," the tertiary base corresponding to the quaternary methochloride macusine-A which occurs in S . toxifera (173), and whose structure has been established by X-ray diffraction (178). Its N,-methyl derivative is the alkaloid voachalotine (CCCXXXVI) which occurs in Voacanga chalotiana (179, 176). The C-18 methyl group lies trans to the 20-21 bond (175, 178, see formulaCCCXXXVI1) as is also the case withechitamine,(CCCXXXVIII) and those alkaloids of the akuammicine type that have been related to the Wieland-Gumlich aldehyde (Chapter 7). It will be noted that polyneuridine is one member of a closely related group of alkaloids, which includes not only vincamedine and echitamine but also akuammine (CCCXXXVIII-A), #-akuammigine (CCCXXXVIII-B) (184a, b), picraline (179a, e), and quebrachidine (CCCXXXVIII-D, see following section). C.

QUEBRACHIDINE

A wide variety of alkaloids has been encountered in the bark of Aspidosperma puebrachoblanco (see Table I). An investigation of the leaf alkaloids yielded a new base, quebrachidine (CCCXXXVIII-D, 179b), whose UV- ( h 242, 290) and NMR-spectra were indicative of a dihydroindole unsubstituted in the benzene ring, while the IR-spectrum showed the presence also of the groupings NH, OH, and C02R. The NMRspectrum further indicated that the ester was a carbomethoxyl group

CCCXXXVIII-A; R = OH CCCXXXVJII-B; R = H

CCCXSXIX

-

8

-

;;:; IH N H' i ;

\

W M

%

H

Ri

Ri Rz CCCXXXVJII-E Ac Ac CCCXXXVIII-F CHO CHO CCCx xx I I Me Ac

oTqz tj

\

CCCXXXVTIJ-D

\

CCCXXI

14. Aspidosperma

T

X

Y

403

F-

d x 0

AND RELATED ALKALOIDS

/ II I1

X 4 SIX Y Y U

+x

55 xx xx xx

88 vv

494

B. GILBERT

(3.6 6) and that there was an ethylidine group present (three protons at 1.5 6, doublet; one proton a t 5.1 6, quartet, J = 6.5 clsec). A second basic nitrogen atom was indicated by the pKi 6.7. Quebrachidine formed an N,O-diacetate (OAc, v, 1745 cm-1, 3-proton singlet, 1.72 6 ; N-Ac, V , 1658 cm-I, 3-proton singlet, 2.45 6) whose UV-spectrum (A 250) showed that the dihydroindole nitrogen had been acetylated. A comparison of the NMR-spectra of the parent base and the diacetate showed that the former was a secondary alcohol since a single proton peak due to CHOH was shifted downfield by 2 ppm in the acetate. The nature of the skeletal structure of the alkaloid became clear from a comparison of the mass spectra of the base and particularly of its diacetate with that of vincamedine (CCCXXXII). Quebrachidine itself shows a molecular ion peak at m/e 352 establishing the molecular formula, C~lH24N203,while indolic peaks appear at m/e 130 and 143 confirmatory of the unsubstituted dihydroindole structure. Of special interest however is the parallel fragmentation of quebrachidine (CCCXXXVIII-D), quebrachidine diacetate (CCCXXXVIII-E), and vincamedine (CCCXXXII, Table V) which shows that all three have the same alicyclic skeleton. The fragmentation of this skeleton appears to involve as a principal process the rupture of the 2-3 and 5-6 bonds with the production of two fragments, either of which may bear the positive charge. Thus, for quebrachidine the indolic peak b at m/e 130 is accompanied by a peak, p p , at M-130. The diacetate also shows b at m/e 130 (N,-acetyl is lost as ketene) and p p a t M-(130+42). I n the case of vincamedine, the b peak is observed at m/e 144 due to the presence of an N,-methyl group, while p p appears at M-144. The peak p p thus represents that part of the molecule not present in fragment 6 , and in conformity with this is accompanied by satellites qq and w at mle values corresponding to the loss of acetyl (pp-42)and further loss of methanol (qq-32; or for quebrachidine, pp-32) (179b). Further confirmation of the structure (CCCXXXVIII-D) for quebrachidine was obtained by opening the five-membered ring by lead tetraacetate oxidation to an indolenine and reverse Mannich cleavage and reduction with borohydride. As in the case of vincamedine (Section VIII, B), the product was polyneuridine (CCCXXI). A similar oxidation of the product CCCXXXVIII-G (obtained from quebrachidine by successive formylation to CCCXXXVIII-F, and LiAlH4 reduction) gave by direct reverse Mannich cleavage the aldehyde CCCXXXVIII-I whose mass spectrum showed i t to have the same skeleton as deoxyajmalal-A (CCCXXXVIII-J). Morevoer, the diol CCCXXXVIII-G gave a monoacetate (CCCXXXVIII-H) identical with that obtained by reduction and acetylation of vincamedine (CCCXXXII, 179c).

14. Aspidosperma AND RELATED ALKALOIDS

495

Knowledge was still lacking of the stereochemistry at C-2 and C-17. A comparison of the mass spectra (cf. 179d) of ajmaline (CCCXXXVIII-K) and tetraphyllicine (CCCXXXVIII-L), which have @-H, with that of quebrachidine showed differences, although ajmaline shows peaks corresponding t o the expected b, b + 13, and p p fragments. A very close resemblance was observed between the spectra of quebrachidine and 2-epi-21-deoxyajmaline (CCCXXXVIII-M) and it may be assumed that the former has C-2, a-H (179b). The secondary hydroxyl group a t C-17 is cis to the carbomethoxyl function, as in ajmaline, since the diol, CCCXXXVIII-G, forms an isopropylidene derivative (CCCXXXVIII-N) impossible for the reverse configuration.

D. HARMAN-3-CARBOXYLIC ACID Among the water-soluble bases of Aspidosperma polyneuron there has been isolated an ester which after methanolysis yielded the known compound 3-carbomethoxyharman (CCCXXXIX) which was recognized by its NMR- and mass spectra (180, 181, Tables I V and V). The natural alkaloid is presumably a glycosidic ester of the corresponding acid.

E. EBURNAMINE AND RELATED ALKALOIDS The known alkaloids eburnamine (CCCXL),eburnamonine (CCCXLII), eburnamenine (CCCXLIII), and possibly isoeburnamine (CCCXLI) have been isolated from plants of the genera Pleiocarpa, Aspidosperma, and Rhaxia (Table I, Refs. 91, 53, 28, 51). These alkaloids were first discovered in Hunteria eburnea (Chapter 11)and their structures (183) and stereochemistry ( 183a) determined. Related alkaloids occur in the genus Vinca (184, 78, 170) and in both this genus and in Aspidosperma their identification in minute quantities has been made possible by mass spectrometry (51, 170) in which the breakdown path has been elucidated by isotopic replacement ( 5 1). The main fragments produced from eburnamenine (CCCXLIII) probably have the structures shown in the formulas ee, ff, and g g ; the fact that it is ring E that survives in these fragments is shown conclusively by the spectra of 14-deuterioeburnamenine (CCCXLIV), dihydroeburnamenine (CCCXLV), and 14-deuteriodihydroeburnamenine (CCCXLVI) in which each of the peaks suffers progressive increments of one mass unit. Peak gg is weak in the spectrum of eburnamonine (CCCXLII) in which another peak, hh, appears, derived from ff by the loss of 28 units. When the carbonyl

CCCXL

CCCXLVI

CCCXLI

ff M-29 CCCXLV

14,15-saturated

\

CCCXLII

ee‘

M-70

ff’ M-29

1

LiAlDl

CCCXLVII

CCCXLN 496

hh

14. Aspidosperma

AND RELATED ALKALOIDS

497

oxygen was replaced (partially) by Ol8, this peak, in contrast t o all the others, underwent no alteration. The peak hh therefore involves the loss of CO f r o m 8 and may be allocated the structure shown (51).It will be seen that the postulated cleavages 1 and 2 are highly favored ones, for they not only involve the breakage of allylically activated bonds but in most cases by a suitable shift of electrons the E ring in the resultant fragments can become fully aromatic (170,51).The reverse Diels-Alder reaction pictured as occurring in ring C is characteristic of the mass spectral breakdown of molecules containing a singly unsaturated sixEburnamine (CCCXL)gives the same spectrum membered ring (21,185). as eburnamenine (CCCXLIII) but may be distinguished by lithium aluminum deuteride reduction to 14-deuteriodihydroeburnamenine (CCCXLVI)(51). 0-Methyleburnamine (CCCXL-A),found in Haplophyton cimicidum, was identified by loss of methanol to eburnamenine (CCCXLIII) and chromic acid oxidation to eburnamonine (CCCXLII, 113b).

F. TUBOFLAVINE A novel canthine-type alkaloid, tuboflavine (CCCXLVIII), has been isolated from the bark of Pleiocarpa tubicina (186).Its highly aromatic nature was recognized from the UV-spectrum (A 215, 264,289,323, 401) which, although unchanged in alkali, underwent a large bathochromic shift in acid or on formation of the methiodide. The empirical formula,

CCCXLVIII

CCCXLVIII-A

CCCXLVIII-B

C ~ ~ H ~ Z N Zwas O , confirmed mass spectrometrically. Tuboflavine is reduced by lithium aluminum hydride to a mixture of two compounds, both of which have the UV-absorption of indolic-N-methylharman, indicating substitution of N,. By the successive action of dilute alkali and methanolic hydrochloric acid, tuboflavine was cleaved to l-carbomethoxy-/3-carboline (CCCXLVIII-A), a result which would exclude a true canthine-type structure based on the skeleton (CCCXLVIII-B).

498

B. GILBERT

The NMR-spectrum of tuboflavine exhibited absorption characteristic of an aromatic ethyl group (CH3, 1.32 6, triplet; CH2, 2.8 6, quartet, J = 7.5 cjsec) and seven aromatic protons, and was fully consistent with structure CCCXLVIII for the alkaloid (186).

G. FLAVOCARPINE Earlier work (186a, 186b, 186c) had shown the presence of physiologically active alkaloids in the plants Pleiocurpu tubicina and P. mutica and the chemistry of some of the tertiary bases from these plants has been described in Sections 111,G, H ; V, D ; VI, C ; and in the preceding section. After complete removal of the chloroform-soluble bases of P. mutica, it was found (186d)that the alkaline aqueous extract still gave a strong Mayer reaction and by way of butanol extraction, precipitation of the picrates, recovery of the free amino acid by ion exchange followed by countercurrent distribution, it was possible t o isolate the yellow zwitterionic alkaloid, flavocarpine (CCCXLVIII-C). The complex UVspectrum of the alkaloid was very similar to that of flavopereirine (CCCXLVIII-D, 186d, 186f, 186g) from which the principal difference lay in the presence in the former of an ionized carboxyl group (v, 1595 cm-1). The NMR-spectrum (trifluoroacetic acid solution) showed that an aromjttic ethyl group (1.55 6, triplet; 3.42 6, quartet) was also present in flavocarpine (186d). Methylation of flavocarpine (CCCXLVIII-C) with methanol and hydrochloric acid gave the methyl ester chloride (CCCXLVIII-E) which, with sodium carbonate, yielded the red anhydronium base, CCCXLVIIIF. The methyl ester chloride suffered reduction of the C ring with sodium borohydride, to give an indole (CCCXLVIII-G, cf. 186h), and this provided a safe distinction from a pyridocarbazole-type structure (cf. 152) for the ester which would have suffered reduction in ring D t o leave a carbazole chromophore in the product (149, 158). Mass spectral molecular weight determination established the presence of one residual double bond in CCCXLVIII-G, not conjugated with the indole chromophore (UV-spectrum), but conjugated with the carbomethoxyl group (IR, v, 1705 cm-1). The absence of vinyl proton absorption in the NMRspectrum showed that it was tetrasubstituted and this evidence, combined with biogenetical probability and the appearance of a v peak at m/e 170 in the mass spectrum (see Section VIII, A), suggests structure CCCXLVIII-G for the reduction product. Full confirmation of the skeleton and position of the ethyl substituent was obtained by direct

14. Aspidosperma

499

AND RELATED ALKALOIDS

decarboxylation of flavocarpine (CCCXLVIII-C) to flavopereirine (CCCXLVIII-D) (186d). Synthesis of flavocarpine was achieved by use of a modification of the method worked out by Ban and Seo (186i), in which a 2-chloropyridine is condensed with 3-( 2-bromoethyl)-indole to form th'e required ring system directly. Thus, to provide the eventual carboxyl group of flavocarpine, a cyano group was introduced into 3-ethylpyridine. This was

I

it CCCXLVIII-C; R CCCXLVIII-D; R CCCXLVIII-E; R CCCXLVIII-P; R CCCXLVIII-Q; R

= COz-

MeOzC CCCXLV1TI-G

v

in/e 170

H (salt) = COzMe = CONHz = COzH =

OQ

I

OMe I

I

R

RI

CCCXLVIII-H; R = H CCCXLVIII-K; R = CN

CCCXLVIII-I

Ri Rz CCCXLVIII-J C N H CCCXLVIII-L CN c1 CCCXLVIII-M CONHz C1

CCCXLVIII-D; R = H (anhydronium) CCCXLVIII-F; R = COzMe

CCCXLVIII-0

CCCXLVIII-N

effected by preparation of the N-oxide, CCCXLVIII-H, methylation to the iodide, CCCXLVIII-I, and direct cyanation (cf. 186j) followed by fractionation of the products (distinction of correct isomer, CCCXLVIIIJ, by NMR-spectroscopy and conversion to 3,4-diethylpyridine). The required chlorine atom was introduced by treatment of the N-oxide, CCCXLVIII-K, with phosphorus oxychloride in hot chloroform, the

500

XXXR

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

B. GILBERT

G L

N

$3

+

zh

14. Aspidosperma

(3 N

AND RELATED ALKALOIDS

V

5 v v

H H

55 v vv v vv

501

502

B. GILBERT

required isomer, CCCXLVIII-L, again being distinguished by NMRabsorption. For the condensation with the bromoethylindole, CCCXLVIII-N, the solid amide, CCCXLVIII-M, was preferred to the liquid cyanide, and the product, CCCXLVIII-0, was readily dehydrogenated with tetrachloroquinone t o the amide of flavocarpine (CCCXLVIII-P). Hydrolysis with aqueous hydrochloric acid furnished flavocarpine itself as the hydrochloride (CCCXLVIII-Q) from which the free base identical with natural flavocarpine (CCCXLVIII-C) was obtained by passage through an ion-exchange resin (186d).

H. CARAPANAUBINE The alkaloid carapanaubine, C23H28N206, which occurs in Aspidosperma carapanauba, was remarkable in that although it had a non-

indolic UV-spectrum its NMR-spectrum was very similar to that of isoreserpiline (CCCV, Table IV), the only significant difference being the position of the N, proton which was farther downfield (8.73 6) with carapanaubine than with isoreserpiline (7.95 6). The presence of an extra carbonyl band in the IR-spectrum and of an extra oxygen atom in the molecule coupled with the UV- and NMR-data led to the supposition that carapanaubine was an oxindole in which rings A, D, and E were identical with the corresponding rings in isoreserpiline (271). I n order to settle this question, a number of oxindoles (CCCXLIX-CCCLIII), prepared by the tertiary butyl hypochlorite (187, 188) or lead tetraacetate (189) oxidation and rearrangement of the corresponding indoles (CCCXVII, CCCL-A, CCCIX, CCCVIII, CCCIII), were examined mass spectrally. It was thus possible to establish the breakdown pattern represented by the formulas ii-nn in which the structures of the various fragments were elucidated by deuteration in the positions 3, 5 , 6, and 14 (CCCL-CCCLII) and by the presence of a methoxyl group in the aromatic ring (CCCLIII). I n the spectrum of carapanaubine the principal peaks were found in positions identical with those exhibited by mitraphylline (CCCXLIX) and aricine oxindole (CCCLIII), with the sole exception that those fragments which incorporated the benzene ring were shifted to higher mle values corresponding to the presence of two methoxyl groups in that ring. Carapanaubine thus has the structure CCCLIV and i t remained only to settle the stereochemistry. I n the NMR-spectrum of carapanaubine the coupling between the C-19 proton (octet a t 4.56 6) and the C-18 methyl protons (doublet a t 1.4 6, J = 6 c/sec) was eliminated by spin decoupling and the C-19 proton absorption then appeared as a doublet with J = 5.7 c/sec due to

14. Aspidosperma

AND RELATED ALKALOIDS

503

coupling with the lone C-20proton. Carapanaubine cannot therefore have the C-19 CC, DIE-trans configuration of mitraphylline (CCCXLIX) nor the DIE-cis configuration of formosanine (CCCLV) since these C-19 /I, alkaloids have been shown to have smaller 19,ZO-HHcoupling conshants (190).There was every possibility, therefore, that carapanaubine had the C-19 a , DIE-cis stereochemistry as in reserpiline (CCCLVI) and isoreserpiline (CCCV) (191, 192). This was fully confirmed by synthesis from reserpiline (CCCLVI)using the lead tetraacetate oxidation method applicable t o indolic alkaloids with cis D/E ring junction (189). The intermediate 7-acetoxy-7H-reserpiline (CCCLVII) was rearranged in methanol containing a little acetic acid and, as under these conditions oxindoles equilibrate to a mixture of the “A” (CONH, a ) and “ B ” (CONH, /3) stereoisomers, two products were obtained, one of them being carapanaubine (CCCLIV).It has been shown that the oxindoles resulting from this type of rearrangement have the 3-aH configuration irrespective of the starting material; the two oxindoles obtained were therefore isoreserpiline oxindoles A and B. The B configuration was allocated to carapanaubine, as it moved more slowly on paper chromatography-a property characteristic of the more strongly basic B isomers (189, 187).

I. ISORESERPILINE-+-INDOXYL The isolation of isoreserpiline-+-indoxy1 (CCCLVII-A) from Aspidosperma discolor has been reported (113d). The structure of this yellow alkaloid, which has also been encountered in Rauwolfia species, was established by synthesis from isoreserpiline (CCCV) (192a, 192g). The synthesis of +-indoxyls from indoles differs from that of oxindoles, described in the preceding section, only in the final treatment of the intermediate 7-acyloxy-7H-indolenine [in this case 7-m-bromobenzoyloxy-7H-isoreserpiline (CCCLVII-B)] with methanolic alkali instead of weak acid (192a).

J. OCHROPAMINE AND OCHROPINE The alkaloids ochropamine (CCCLVII-C)and ochropine (CCCLVII-D) from Ochrosia poweri are the only representatives of the growing 2acylindole class so far encountered in the genera under study (192d). The nature of the chromophore (A 243 and 315 mp, E 18,900, 17,700) present in CCCLVII-C was determined both by alkaline degradation to 2-acetyl-1,3-dimethylindole and by comparison with the known 1-keto1,2,3,4-tetrahydrocarbazole.The presence of carbomethoxyl (5.78 p,

504

13. GILBERT

2.58 8) and N-methyl groups suggested a relation t o vobasine (CCCLVIIE) whose structure and stereochemistry are known (192e, 192f). Comparison of the NMR spectra of vobasine, ochropamine, and ochropine left no doubt as t o the skeletal and relative stereochemical identity of the three alkaloids, and showed that ochropamine (CCCLVII-C)differed from vobasine only in possessing an N,-methyl group (NMR, 4.056), while ochropine (CCCLVII-D) contained in addition an aromatic methoxyl group. This was located by UV-comparison with model compounds (192d). MeOIC \

RI CCCLVII-C CCCLVII-D CCCLVII-E

Rz

H Me OMe Me H H

R,

Rz

IX. Alkaloids of Unknown Structure A. ALKALOIDSOF Pleiocarpa SPECIES Three alkaloids whose structure is unknown at the time of writing have been isolated from Pleiocarpa mutica (91). One of them, pleiocarpamine, C2oH22N202, is a pentacyclic indole in which N, is substituted. Nb is tertiary and the two oxygen atoms have been located in a carbomethoxyl group. The alkaloid bears an ethylidene side chain (91, 53). The other two, pleiomutine and pleiomutinine, are double alkaloids. Pleiomutine shows UV-absorption similar to that of leurosine and vinblastine and contains, therefore, an indole and a dihydroindole chromophore. The UV-spectrum of pleiomutinine, on the other hand, extends to longer wavelengths showing distinct similarity t o the spectra of vobtusine and callichiline (193). I n the species P. tubicina the occurrence of the two lactams pleiocarpinilam and kopsinilam has been described (Section 111, H). I n addition, a third lactam was isolated from this plant which exhibited IR-absorption bands a t 1763 and 1687 cm-1 (96).

B. ALKALOIDSOF Ochrosia SPECIES A number of species of the genus Ochrosia have been investigated, and among the alkaloids already discussed are ellipticine and methoxyellipticine from 0. elliptica (Section VII, D). Other alkaloids from this species include elliptinine for which structure CCCLVIII has been

14. Aspidosperma

AND RELATED ALKALOIDS

505

proposed (156). 0. sandwicensis yielded an unnamed base with UVabsorption a t 238 and 290 mp (156). An unnamed alkaloid from 0. oppositifoZia, C ~ ~ H ~ ~ Nhas Z Othe , same composition as methoxyellipticine. The UV-absorption (A 242, 275, 290, 335 mp; loge 4.35, 4.59,

CCCLVIII

4.66, 3.71) is also very similar (194). Australian and New Guinea species, notably 0. poweri, contain a number of alkaloids (161), including elliptamine, C24H3oNz05, powerine, C21H26N204, and poweramine, C23H30Nz04 (161). Elliptamine, the most widespread, is unstable in the form of the free base. It forms an orange-red picrate. Powerine and poweramine may be 5- and 6-methoxy indoles, respectively (161).

C. ALKALOIDS OF Aspidosperma, Rhazya, AND Stemmadenia I n addition to olivacine and guatambuine (Section VII, C), an unnamed alkaloid, mp 186"-188", resembling uleine in its color reactions, has been isolated from Aspidosperma australe (147)s. A glycosidic alkaloid, quebrachacidine, C26Hz8N2011, has been found in A. quebrachoblanco (195), and its aglycone has been prepared. Several alkaloids have been reported in A . oblongum and A . album (196), notably kromantine, mp 176", [aID+ 159" (in chloroform), from the latter species (196a). Among the many alkaloids of Rhazya stricta (seeTable I),the structure of the alkaloid rhazinine remains to be determined (197). This indolic base, C19H24N20, contains a primary alcoholic group but, unlike akuammidine (CCCXXII)which accompanies it in the plant, it contains no G-methyl (197). Other alkaloids of the same plant include rhazidine, CzoHzsN203, HzO, mp 278-279", [.ID - 21" (in ethanol) (199, 200, 62). Two unnamed alkaloids, mp 233"-235" (dec.) and 135"-139", respectively, were isolated from Stemmadenia donnell-smithii, but no further information is available on these a t the time of writing (8). REFERENCES 1. R.B. Woodward, Nature 162, 155 (1948). 2. Sir Robert Robinson, 'I The Structural Relationsof NaturalProducts." OxfordUniv. Press, London and New York, 1955.

3. C. Djerassi, S. E. Flores, H. Budzikiewicz, J. M. Wilson, L. J. Durham, J. Le Men, M. M. Janot, M. Plat, M. Gorman, and N. Neuss, Proc. NatE. Acad. Sci. U.S. 48, 113 (1962). 3

This alkaloid has been identified as apparioine [CCLVII-U p. 473 (37)].

506

B . GILBERT

4. M. Gorman, N. Neuss, and K. Biemann, J . Am. Ch,em. SOC.84, 1058 (1962). 5. G. H. Svoboda, I. S. Johnson, M. Gorman, and N. Neuss,J. Phrcrm. Sci. 51,707 (1962). 6. C. Djerassi, H. Budzikiewicz, J. M. Wilson, J. Gosset, and M. M. Janot, Tetrahedron Letters p. 235 (1962). 7. M. Plat, J. Le Men, M. M. Janot, J . M. Wilson, H. Budzikiewicz, L. J. Durham, Y . Nagakawa, and C. Djerassi, Tetralbedron Letters p. 271 (1962). 8. F. Walls, 0 . Collera, and A. Sandoval, Tetrahedron 2, 173 (1958). 9. C. Djerassi, R.Gilbert, J. N. Shoolery, L .F. Johnson, and K. Biemann, Ezperientia 17, 162 (1961). 10. C. Djerassi, A. A. P. G. Archer, T. George, B. Gilbert, and L. D. Antonaccio, Tetrahedron 16, 212 (1961). 10a. J . Mokrj., I. KompiB, L. Dubravkovb, and P. SepEoviE, Tetrahedron Letters p. 1185 (1962). lob. J. Mokry and I. KompiB, Chem. Zvesti 17,852 (1963). 11. J. F. D. Mills and S. C. Nyburg, J . Chem. SOC.p. 1458 (1960). 12. G. F. Smith and J . T. Wrobel, J . Chem. SOC.p. 1463 (1960). 13. H. Kny and B. Witkop, J . Org. Chem. 25, 635 (1960). 14. E. Wenkert, Experientia 15, 165 (1959) (see footnote 53). 15. B. U‘itkop,J. Am. Chem. SOC.79, 3193 (1957). 16. B. Witkop, J. Am. Chem. SOC.70, 3712 (1948). 17. L. A. Cohen, J. W. Ualy, H. Kny, and B. Witkop,J. Am. Chem. SOC.82,2184 (1960). 18. K. Biemann and G. Spiteller, Tetrahedron Letters p. 299 (1961). 18a. K. Biemann and G. Spiteller, J . Am. Chem. SOC.84, 4578 (1962). 19. G. F. Smith and J. T. Wrobe1,J. Chem. SOC.p. 792 (1960). 20. J. H. Benyon, “Mass Spectrometry and its Applications t o Organic Chemistry.” Elsevier, Amsterdam, 1960. 2 1. K. Biemann, “Mass Spectrometry, Organic Chemical Applications.” McGraw-Hill, New York, 1962. 22. J. F. D. Mills and S. C. Nyburg, Tetrahedron Letters No. 11, 1 (1959). 23. &. Schmutz and H. Lehner, Helv. Chim. Acta 42, 874 (1959). 24. H. Conroy, P. R. Brook, and Y . Amiel, Tetrahedron Letters No. 11, 4 (1959). 25. B. Witkop and J. B. Patrick, J . Am. Chem. SOC.76, 5603 (1954). 26. H. Conroy, P. R. Brook, M. K. Rout, and N. Silverman,J. Am. Chem. SOC.SO, 5178 (1958). 27. A. J . Everett, H. T. Openshaw, and G. F. Smith, J . Chem. SOC. p. 1120 (1957). 27a. G. Stork and J. E. Dolfini, J . Am. Chem. SOC.85, 2872 (1963). 27b. G. Stork, A. Brizzolara, H. Landesman, J. Szmuszkovicz, and R. Terrell, J . Am. Chem. SOC.85, 207 (1963). 28. K. Biemann, M. Friedmann-Spiteller, and G. Spiteller, Tetrahedron Letters p. 485 (1961). 28a. H. Budzikiewicz, C. Djerassi, and D. H. Williams, “Structure Elucidation of Natural Products by Mass Spectrometry,” Vol. I : Alkaloids. Holden-Day, Sen Francisco, 1964. 28b. P. Bommer, W. McMurray, and K. Biemann, J . Am. Chem. SOC.86, 1439 (1964). 29. C. Djerassi, A. A. P. G. Archer, T. George, B. Gilbert, J. N. Shoolery, and L. F. Johnson, Ezperientia 16, 532 (1960). 30. S. McLean, K. Palmer, and L. Marion, Can. J . Chem. 38, 1547 (1960). 31. M. Pinar and H. Schmid, Helv. Chim. Acta 45, 1283 (1962). 32. M. Plat, J. Le Men, M. M. Janot, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi, Bull. SOC.Chim. France p. 2237 (1962).

14. Aspidosperma AND RELATED ALKALOIDS

507

33. M. Pinar, W’.von Philipsborn, W. Vetter, and H. Schmid, HeZw. C’him.Acta 45,2260 (1962). 34. 0. 0. Orazi, R. A. Corral, J. S. E. Holker, and C. Djerassi, J . Org. Chem. 21, 979 (1956); Anales Asoc. Quim. Arg. 44, 177 (1956). 35. C. Djerassi, H. W. Brewer, H. Budzikiewicz, 0. 0. Orazi, and R. A. Corral, Experientia 18, 113 (1962). 36. C. Djerassi, H. W. Brewer, H. Budzikiewicz, 0. 0. Orazi, and R. A. Corral, J . Am. C‘hem. Soc. 84, 3480 (1962). 37. C. Cjerassi, J. M. Ferreira, S. E. Flores, T. George, B. Gilbert, H. J. Monteiro, H. Budzikiewicz, J. M. Wilson, L. J. Durham, R. J. Owellen, and E. Bianchi, Unpublished work (1963). 38. A. J. Ewins, J. Chem. SOC.105, 2738 (1914). 38a. A. C. Paladini, E. A. Ruveda, R. A. Corral, and 0. 0. Orazi, Anales Asoc. Quim. Arg. 50, 352 (1962). 38b. P. Relyveld, Pharm. Weekbhd 98, 175 (1963). 39. E. Schlittler and M. Rottenberg, Hell. Chim. Acta 31, 446 (1948). 40. J. Aguayo Brissolese, C. Djerassi, and B. Gilbert, Chem. Ind. (London)p. 1949 (1962). 40a. J. M. Ferreira, B. Gilbert, R. J. Owellen, and C. Djerassi, Experientia 19, 585 (1963). 41. J. S. E. Holker, M. Cais, F. A. Hochstein, and C. Djerassi, J. Org. Chem. 24, 314 (1959). 41a. V. Deulofeu, J . De Langhe, R. Labriola, and V. Carcamo, J . Chem. SOC.p. 1051 (1940). 42. C. Ferrari, S. McLean, L. Marion, and K. Palmer, Can. J . Chem. 41, 1531 (1963). 43. W. I. Taylor, N. Raab, H. Lehner, and J. Schmutz, Helw. Chim. Acta 42,2750 (1959). 44. C. F. Garbers, H. Schmid, and P. Karrer, Helw. Chim. Acta 37, 1336 (1954). 45. S. McLean, Can. J . Chem. 38, 2278 (1960). 46. L. Jurd, Arch. Biochem. Biophys. 63, 376 (1956). 47. V. Deulofeu, Personal communication (1962). 48. B. Gilbert, A. P. Duarte, Y. Nakagawe, J. A. Joule, S. E. Flores, J. A. Brissolese, J. Campello, E. P. Carrazzoni, R. J. Owellen, E. C. Blossey, hnd C. Djerassi. TObe published (1964). 49. B. Gilbert, L. D. Antonaccio, A. A. P. G. Archer, and C. Djerassi, Experientia 16, 61 (1960). 50. K. Bowden, I. M. Heilbron, E. R. H. Jones, and B. C. L. Weedon,J. Chem. SOC.p. 39 (1946). 51. H. K. Schnoes, A. L. Burlingame, and K. Biemann, TetrahedronLettem p. 993 (1962). 51a. K. Biemann, M. Spiteller-Friedmann, and G. Spiteller, J. Am. Chem. SOC.85, 631 (1963). 51b. G. F. Smith and M. A. Wahid, J . Chem. SOC.p. 4002 (1963). 52. C. Djerassi, L. D. Antonaccio, H. Budzikiewicz, J. M. Wilson, and B. Gilbert, Tetrahedron Letters p. 1001 (1962). 53. H. Schmid, Unpublished work (1963). 54. F. W. McLafferty, Anal. Chem. 31, 2072 (1959). 54a. S. McLean, Can.J. Chem. 42, 191 (1964). 55. N. Neuss, “Physical Data of Indole and Dihydroindole Alkaloids,” 4th rev. ed., Vol. I. Eli Lilly and Co., Indianapolis, Indiana, 1960. 56. N. Neuss, “Physical Data of Indole and Dihydroindole Alkaloids,” 4th rev. ed., Vol. 11. Lilly, Indianapolis, Indiana, 1962. 57. R. E. Woodson, Ann. Missouri Botan. Garden 38, 119 (1951).

508

B. GILBERT

J. Schmutz, Pham. Acta Helv. 36, 103 (1961). 0. Hesse, Ann. Ghem. 211, 249 (1882). E. Schlittler and E. GBllert, Helv. Chim. Acta 34, 920 (1951). E. GQllertand B. Witkop, Helv. Chim. Acta 35, 114 (1952). A. Chatterjee, C. R. Ghosal, N. Adityachaudhury, and S. Ghosal, Chem. Ind. (London)p. 1034 (1961). 63. L. D. Antonaccio, Rev. Quim.I d . ( R i o de Janeiro) 26, 149 (1957). 64. M. M. Janot, J. Le Men, and C. Fan, Compt. Rend. Acad. Sci. 248, 3005 (1959). 65. M. M. Janot, H. Pourrat, and J. Le Men, Bull. Soc. Chim. France p. 707 (1954). 66. A. Sandoval, Unpublished work (1958) cited in reference 7. 66a. 0. Collera, F. Walls, A. Sandoval, F. Garcia, J . Herran, and M. C. Perezamador, Bol. Inst. Quim. Univ. Nac. Auton. Mexico 14, 3 (1962). 67. K. Bernauer, W. Arnold, C. Weissmann, H. Schmid, and P. Karrer, Helv. Chim. Acta 43, 717 (1960). 68. M. M. Janot, J. LeMen, A. LeHir, J. Levy, andF. Puisieux, Compt. Rend. Acad. Sci, 250, 4383 (1960). 69. J. Levy, J. Le Men, and M. M. Janot, Bull. SOC.Chim. France p. 979 (1960). 70. K. Aghoramurthy and Sir Robert Robinson, Tetrahedron 1, 172 (1957). 71. M. M. Janot, J. Le Men, and C. Fan, Bull. Soc. Chim. France p. 891 (1959). 72. M. Gorman, N. Neuss, G. H. Svoboda, A. J. Barnes, and N. J. Cone, J . Am. Pharm. Assoc. Sci. Ed. 48, 256 (1959). 72a. N. Neuss, M. Gorman, W. Hargrove, N. J. Cone, K. Biemann, G. Buchi, and R. E. Manning, J . Am. Chem. Soc. 86, 1440 (1964). 73. J. Mokrj., I. KompiB, L. Dubravkova, and P. SevEoviE, I U P A C Congr. Nut. Prod., Prague, 1962; Experientia 19, 311 (1963). 74. J . Gosset, J. Le Men, and M. M. Janot, Ann. Pharm. Franp. 20,448 (1962). 74a. J. Mokrj., L. D6bravkov&, and P. SepEoviE, Experientia 18, 564 (1962). 75. M. M. Janot, in “The Chemistry of Natural Products,” Vol. 2, p. 635. Butterworths, London (1963). 76. M. Plat, E. Fellion, J. Le Men, and M. M. Janot, Ann. Pharm. Franp. 20, 899 (1962). 77. P. N. Edwards and G. F. Smith. Proc. Chem. Soc. p. 215 (1960).78. J. Mokry, 1. KompiB, and P. SefEoviE, Tetrahedron Letters p. 433 (1962). 79. L. D. Antonaccio, J . Org. Chem. 25, 1262 (1962). 80. C. Djerassi, R. J . Owellen, J. M. Ferreira, and L. D. Antonaccio, Experientia 18, 397 (1962). 81. B. Gilbert, J . M. Ferreira, R. J. Owellen, C. E. Swanholm, H. Budzikiewicz, L. J. Durham, and C. Djerassi, Tetrahedron Letters p. 59 (1962). 82. C. Djerassi, T. George, N. Finch, H. F. Lodish, H. Budzikiewicz, and B. Gilbert, J . Am. Chem. Soc. 84, 1499 (1962). 83. M. Greshoff, Ber. 23, 3537 (1890). 84. W. P. H. van den Driessen Mareeuw, Ned. Tijdschr. Pharm. Chem. Toxicol. 8, 199 (1896). 85. K. Gorter, Jaarb. Dept. Landb. Ned.-Ind. p. 240 (1920). 86. L. J. Webb, Australia, Commonwealth Sci. Ind. Res. Organization Bull. No. 268 (1952). 87. W. D. Crow and M. Michael, AustralianJ. Chem. 8, 129 (1955). 88. N. G. Bisset, W. D. Crow, and Y. M. Greet, AustralianJ. Chem. 11, 388 (1958). 89. W. D. Crow and M. Michael, Australian J . Chem. 15, 130 (1962). 90. W. D. Crow, Compt. Rend. Congr. Assoc. Sci.Pays Ocean Indien, 3“,Tananarive, 1957, Sect. G p. 39 (Pub. 1958); Chem. Abstr. 53, 17432e (1959).

58. 59. 60. 61. 62.

14. Aspidospermu

AND RELATED ALKALOIDS

509

W. G. Kump and H. Schmid, Helv. Chim. Acta 44, 1503 (1961). A. R. Battersby and D. J. Le Count, J . Chem. SOC.p. 3245 (1962). E. Seebeck and D. Stauffacher, Unpublished work (1961). F. A. Hochstein, Unpublished work cited by N. Neuss, reference 56. W. G. Knmp, D. J. Le Count, A. R. Battersby, and H. Schmid, Helv. Chim. Acta 45, 854 (1962). 96. C. Kump and H. Schmid, Helw. Chim. Acta 45, 1090 (1962). 97. W. D. Crow and K. Biemann, Unpublished work (1962). 98. S. C. Pakrashi, C. Djerassi, R. Wasicki, and N. Neuss, J . Am. Chem. SOC.7 7 , 6687 (1955). 99. M. Gorman, N. Neuss, C. Djerassi, J. P. Kutney, and P. J. Scheuer, Tetrahedron 1, 328 ( 1957). 100. T. R. Govindachari, S. Rajappa, and N. Viswanathan, J . Sci. Ind. Res. (India)20B, 557 (1961). 101. T. R. Govindachari, B. R. Pai, S. Rajappa, N. Viswanathan, W. G. Kump, K. Nagarajan, and H. Schmid, Helv. Chim. Acta 45, 1146 (1962). 102. C. Djerassi, H. Budzikiewicz, R. J. Owellen, J. M. Wilson, W. G. Kump, D. J. Le Count, A. R. Battersby, and H. Schmid, Helv. Chim. Acta 46, 742 (1963). 103. A. Bhattacharya, A. Chatterjee, and P. K. Bose, J . Am. Chem. Soc. 71, 3370 (1949). 104. A. Bhattacharya, J . Am. Chem. SOC.75,381 (1953). 105. A. K. Bhattacharya, Sci. Cult. (Calcutta)22, 120 (1956); Chem. Abstr. 51, 17951a (1957). 106. G. Spiteller, A. Chatterjee, A. Bhattacharya, and A. Deb, Naturwissenschaften 49, 279 (1962). 107. G. Spiteller, A. Chatterjee, A. Bbattacharya, and A. Deb, Monatsh. Chem. 93, 1220 (1962). 108. A. K. Kiang and R. D. Amarasingham, Proc. Symp. Phytochem. Kuakz Lumpur 1957, p. 165; Chem. Abstr. 53, 14131h (1959). 109. A. R. Battersby and H. Gregory, J. Chem. Soc. p. 22 (1963). 110. A. Chatterjee and A. Deb, Sci. Cult. (Calcutta)28, 195 (1962). 111. T. R. Govindachari, K. Nagarajan, and H. Schmid, Helw. Chim. Acta 46, 433 (1963). 112. T. R. Govindachari, B. R. Pai, S. Rajappa, N. Viswanathan, W. G. Kump, K. Nagarajan, and H. Schmid, Helv. Chim. Acta 4% 572 (1963). 113. A. Guggisberg, T. R. Govindachari, K. Nagarajan, and H. Schmid, Helw. Chim. Acta 46, 679 (1963). 113a.. H. R. Snyder, H. F. Strohmayer, and R. A. Mooney, J . Am. Chem. Soc., 80, 3708 (1958). 113b. M. P. Cava, S. K. Talapatra, K. Nomura, J. A. Weisbach, B. Douglas, and E. C. Shoop, Chem. Ind. (London) p. 1242 (1963). 113c. M. P. Cava, K. Nomura, and S. K. Talapatra, Tetrahedron 20,581 (1964). 113d. N. Dastoor and H. Schmid, Ezperientia 19, 297 (1963). 113e. K. S. Brown, Jr., H. Budzikiewicz, and C. Djerassi, Tetrahedron Letters p. 1731 (1963). 113f. R. H. Burnell, J. D. Medina, and W. A. Ayer, Chem. Ind. (London)pp. 33,235 (1964). 113g. J. M. Ferreira, A. Figueiredo, and K. S. Brown, private communication. 113h. L. Marion, private communication. 113i. K. S. Brown, Jr., andC. Djerassi,J. Am. Chem.Soc. 86, 2451 (1964). 113j. M. P. Cava, S. K. Talapatra, P. Yates, M. Rosenberger, A. G. Szabo, B. Douglas, R. F. Raffauf, E. C. Shoop, and J. A. Weisbach, Chem. Ind. (London)p. 1875 (1963). 113k. E. F. Rogers, H. R. Snyder, and R. F. Fischer, J . Am. Chem. Soc. 74, 1987 (1952). 91. 92. 93. 94. 95.

510

B . GILBERT

1131. H. R . Snyder, H.F. Fischer, J. F. Walker, H. E. Els, and G. A. Nussberger, J . Am. Chem. SOC. 76,2819 (1954). 113m. H. R. Snyder, R. F. Fischer, J. F. Walker, H. E . Els, and G. A. Nussberger,J. Am. Chem. Soc. 76,4601 (1954). 114. J. N. Shoolery, Discussions Paraday SOC.No. 34, 104 (1962). 115. E. Wenkert, J . A m . Chem. Soc. 84, 98 (1962). 115a. H. Budzikiewicz, J. M. Wilson, C. Djerassi, J. Levy, J . Le Men, and M. M. Janot, Tetrahedron 19, 1265 (1963). 116. D. Stauffacher, Helv. Chim. Acta 44, 2006 (1961). 117. A. Sandoval, F. Walls, J. N. Shoolery, J. M. Wilson, H. Budzikiewicz, and C. Djerassi, Tetrahedron Letters p. 409 (1962). 118. K. Riemann, A. L. Burlingame, and D. Stauffacher, Tetrahedron Letters p. 527 (1962). 11th. D. Schumann and H. Schmid, Helv. Chim. Acta 46, 1996 (1963). 119. W. G. Kump, M. B. Patel, and H. Schmid, Unpublished work (1963)cited in reference 118a. 120. M. Pinar and H. Schmid, Ann. Chem. 668, 97 (1963). p. 1115 (1957). 120a. J. R. Chalmers, H. T. Openshaw, and G. F. Smith, J . Chem. SOC. 121. R. Goutarel, X. Monseur, and J. Le Men, Compt. Rend. Acad. Sci. 253, 485 (1961). 122. X. Monsenr, R. Coutarel, J. Le Men, J. hf. Wilson, H. Budzikiewicz, and C. Djerassi, Bull. Soc. Chim. France p. 1088 (1962). 123. P. N. Edwards and G. F. Smith, J . Chem. Soc. p. 1458 (1961). 124. K. Bernauer, F. Berlage, W.von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 2293 (1958). 125. C. Weissmann, H. Schmid, and P. Karrer, Helv. Chim. Acta 44, 1877 (1961). 125a. C. Djerassi, Y . Nakagawa, J. M. Wilson, H. Budzikiewicz, B. Gilbert, and L. D. Antonaccio, Ezperientia 19, 467 (1963). 126. C. Djerassi, Y . Nakagawa, H. Budzikiewicz, J. M. Wilson, J. Le Men, J. Poisson, and M. M. Janot, Tetrahedron Letters p. 653 (1962). 127. Y . Nakagawa, J. M. Wilson, H. Budzikiewicz, and C. Djerassi, Chem. Ind. ( L o n d o n ) p. 1986 (1962). 128. C. Weissmann, H. Schmid, and P. Karrer, H e h . Chim. Actu 45, 62 (1962). 129. G. H. Svoboda, M. Gorman, N. Neuss, and A. J. Barnes,J. Pharm. Sci. 50,409 (1961). 130. J. A. Goodson, J . Chem. SOC.p. 2626 (1932). 131. Sir Robert Robinson and A. F. Thomas, J . Chem. SOC.p. 2049 (1955). 132. H. Fritz, E. Besch, a n d T . Wieland, Anyew. Chem. 71, 126 (1959). 133. H. Wieland, H. J. Pistor, and K. BBhr, Ann. Chem. 547, 140 (1941). 134. H. Fritz, E. Besch, and T. Wieland, Ann. Chem. 617, 166 (1958). 135. H. Fritz, Ber. 92, 1809 (1959). 136. W.von Philipsborn, K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 42, 461 (1959). 137. W. vori Philipsborn, H. Meyer, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 1257 (1958). 137a. P. N. Edwards and G. F. Smith,J. Chem. SOC.p. 152 (1961). 138. J. Schmutz, F. Hunzicker, and R. Hirt, Helv. Chim. Acta 40, 1189 (1957). 139. J. Schmutz and F. Hunzicker, Helv. Chim. Acta 41, 288 (1958). 140. J. Schmutz and F. Hunzicker, Pharnz. Acta Helo. 33, 344 (1958). 141. M. ,4. Ondetti and V. Deulofeu, Tetrahedron Letters No. 7, 1 (1959). 142. G. A. Biichi and E. W. Warnhoff, J . Am. Chem. SOC.81, 4433 (1959). 143. P. H. Carter, S. G. C. Plant, and M. Tomlinson,J. Chem. Soc. p. 2210 (1957). 143a. J. A. Joule and C. Djerassi, In press (1964).

14. Aspidosperrna

AND RELATED ALKALOIDS

511

143b. M. Ohashi, J. A. Joule, B. Gilbert, and C. Djerassi, Ezperientia I n press (1964). 144. M. A. Ondetti and V. Deulofeu, Tetrahedron Letters No. 1, 18 (1960). 145. G. B. Marini-Bettolo, Ann. Chim. (Rome)49, 869 (1959). 146. P. Carvalho-Ferreira, G. B. Marini-Bettolo, and J. Schmutz, Ezperientia 15, 179 (1959) 147. M. A. Ondetti and V. Deulofeu, Tetrahedron 15, 160 (1961). 148. M. Gorman, N. Neuss, N. Cone, and J. A. Deyrup,J. Am. Chem.SOC. 82,1142 (1960). 149. G. B. Marini-Bettolo and J. Schmutz, HeZv. Chim. Acta 42, 2146 (1959). 150. J. Schmutz and H. Wittwer, Helv. Chim. Acta 43, 793 (1960). 151. R. F. H. Manske and M. Kulka, Can. J. Res. 27B, 291 (1949). 152. R. F. H. Manske and M. Kulka, J. Am. Chem. SOC.72, 4997 (1950). 153. R. B. Woodward and G. A. Iacobucci, cited in reference 144. 154. E. Wenkert and K. G. Dave, J. Am. Chem. SOC.84, 94 (1962). 155. H. Lehner and J . Schmutz, Helv. Chim. Acta 44, 444 (1961). 156. S. Goodwin, A. F. Smith, and E. C. Homing, J . Am. Chem. SOC.81, 1903 (1959). 157. R. B. Woodward, G. A. Iacobucci, and F. A. Hochstein, J . Am. Chem. SOC.81, 4435 (1959). 158. G. Buchi, D. W. Mayo, and F. A. Hochstein, Tetrahedroia 15, 167 (1961). 159. P. A. Cranwell and J . E. Saxton, Chem. I n d . (London)p. 45 (1962). 160. P. A. Cranwell and J. E. Saxton, J. Chem. Soc. p. 3482 (1962). 16Oa. T. R. Govindachari, S. Rajappa, and V. Sudarsanam, IndianJ. Chem. 1,247 (1963). 161. F. A. Doy and B. P. Moore, Australian J . Chem. 15, 548 (1962). 162. G. M. Spiteller and M. Spiteller-Friedmann, Monatsh. Chem. 93, 795 (1962). 163. B. Gilbert, L. D. Antonaccio, and C. Djerassi, J . Org. Chem. 27, 4702 (1962). 164. J. D. M. Asher, J. M. Robertson, G. A. Sim, M. F. Bartlett, R. Sklar, and W. I. Taylor, Proc. Chem. SOC.p. 72 (1962). 164a. G. Spiteller and M. Friedmann-Spiteller, Monatsh. Chem. 94, 779 (1963). 165. L. D. Antonaccio, N. A. Pereira, B. Gilbert, H. Vorbrueggen, H. Budzikiewicz, J . M. Wilson, L. J. Durham, and C. Djerassi, J. Am. Chem. SOC.84, 2161 (1962). 165a. M. Ohashi, H . Budzikiewicz, J. M. Wilson, C . Djerassi, J. Levy, J. Gosset, J. LeMen, and M. M. Janot, Tetrahedron 19, 2241 (1963). 165b. E. Clayton, R . I. Reed, and J . M. Wilson, Tetrahedron 18, 1449 (1962). 166. K. Biemann, Tetrahedron Letters No. 15, 9 (1960). 167. K. Biemann,J. Am. Chem. SOC.83, 4801 (1961). 168. K. Biemann and M. Friedmann-Spiteller, Tetrahedron Letter8 p. 68 (1961). 169. K. Biemann and M. Friedmann-Spiteller,J . Am. Chem. SOC.83, 4805 (1961). 170. M. Plat, D. D. Manh, J. Le Men, M. M. Janot, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi, Bull. SOC.Chim. France p. 1082 (1962). 171. B. Gilbert, J . Aguayo Brissolese, N. Finch, W. I. Taylor, H. Budzikiewicz, J. M. Wilson, and C. Djerassi,J. Am. Chem.SOC.85, 1523 (1963). 172. A. R. Battersby and D. A. Yeowell, Proc. Chem. SOC.p. 17 (1961). 173. A. R. Battersby, It. Binks, H. F. Hodson, end D. A. Yeowell,J. Chem. SOC. p. 1848 (1960). 173a. F. Fish, M. Qaisuddin, and J. B. Stenlake, Chem. Ind. (London)p. 319 (1964). 174. J. Levy, J. LeMen, andM. M. Janot, Compt. Rend. Acad. Sci. 253, 131 (1961). 175. S. Silvers and A. Tulinsky, Tetrahedron Letters p. 339 (1962). 176. M. M. Janot, J. Le Men, J. Gosset, and J. Levy, Bull. SOC.Chim. France p. 1079 (1962). 177. M. F. Bartlett,, R. Sklar, W. I. Taylor, E. Schlittler, R. L. S. Amai, P. Beak, N. V. Bringi, and E. Wenkert, J. Am. Chem. SOC.84, 622 (1962).

512

R. GILBERT

178. A. T. McPhail, J. M. Robertson, G. A. Sim, A. R. Battersby, H. P. Hodson, and D. A. Yeowell, Proc. Chem. SOC.p. 223 (1961). 179. N. Defay, M. Kaisin, J. Pecher, and R. H. Martin, Bull. SOC.Chim. Belges 70, 475 (1961). 179a. L. Olivier, J. Levy, J. Le Men, andM. M. Janot, Ann. Pharm. Franp. 20,361 (1962). 179b. M. Gorman, A. L. Burlingame, and K. Biemann, Tetrahedron Letters p. 39 (1963). 179c. M. M. Janot, J. Le Men, and C. Fan, Compt. Rend. Acad. Sci. 247, 2375 (1958). 179d. G. Spiteller and M. Spiteller-Friedmann, Tetrahedron Letters p. 147 (1963). 179e. A. Z. Britten, G. F. Smith, and G. Spiteller, Chem. Ind. ( L o n d o n )p. 1492 (1963). 180. L. D. Antonaccio and H. Budzikiewicz, Monatsh. Chem. 93,962 (1962). 181. H. R. Snyder, C . H. Hansch, L. Katz, S. M. Parmerter, and E. C. Spaeth, J. Am. Chem. SOC.70, 219 (1948). 182. P. J. Scheur and J. T. H. Metzger, J . Org. Chem. 26, 3069 (1961). 183. M. F. Bartlett and W. I. Taylor, J . Am. Chem. SOC.82, 5941 (1960). 183a. E. Wenkert et al., Unpublished results (1963). 184. J . Trojanek, 0. Strouf, J. Holubek, and 2. cekan, Tetrahedron Letters p. 702 (1961). 184a. G. F. Smith and J . A. Joule, J . Chem. SOC.p. 312 (1962). 184b. J. Levy, J . Le Men, and M. M. Janot, Bull. SOC.Chim. France p. 1658 (1961). 185, K. Biemann, Angew. Chem. 74, 102 (1962); Angew. Chem. Intern. Ed. Engl. 1, 98 (1962). 186. C. Kump, J. Seibl, and H. Schmid, Helv. Chim. Acta 46,498 (1963). 186a. M. Raymond-Hamet, Compt. Rend. Acad. Sci. 244, 2991 (1957). 186b. A. K. Kiang and B. Douglas, Malayan Phurm. J . 6,138 (1957). 1860. D. P. N. Tsao, J. A. Rosecrans, J. J. de Feo, and H. W. Youngken, Econ. Botany 15 99 (1961). 186d. G. Biichi, R. E. Manning, and F. A. Hochstein,J. Am. Chern. SOC.84, 3393 (1962) 1860. 0. Bejar, R.'Goutarel, M. M. Janot, and A. Le Hir, Cmpt. Rend. Acad. Sci. 244, 2066 (1957). 186f. N. A. Hughes and H. Rapoport, J . Am. Chem. SOC.SO, 1604 (1958). 186g. K. B. Prasad and G. A. Swan, J . Chem. SOC.p. 2024 (1958). 186h. B. Witkop, J . A m . Chem. SOC.75, 3361 (1953). 186i. Y. Ban and M. Seo, Tetrahedron 16, 5 (1961); Chem. Ind. (London) p. 235 (1960). 186j. T. Okamoto and H. Tani, Chem. Pharm. Bull. (Tokyo)7, 925 (1959). 187. N. Finch and W. I. Taylor, J . Am. Chem. SOC.84, 1318, 3871 (1962). 188. J. Shave1 and H. Zinnes, J . Am. Chem. SOC.84, 1320 (1962). 189. N. Finch, C. W. Gemenden, I. H. Hsu, and W. I. Taylor, J . Am. Chem. SOC.85, 1520 (1963). 190. E. Wenkert, B. Wickberg, and C. Leicht, Tetrahedron Letters p. 822 (1961). 191. E. Wenkert, B. Wickberg, and C. Leicht,J. Am. Chem. SOC.83,505 (1961). 192. M. Shamma and J. B. Moss, J . Am. Chem. SOC.83, 5038 (1961). 192a. N. Finch, W. I. Taylor, and P. R. Ulshafer, Ezperientia 19, 296 (1963). 192b. R. F. Raffauf and M. B. Flagler, Econ. Botany 14, 37 (1960). 192c. F. A. Hochstein, K. Murai, and W. H. Boegemann, J . Am. Chem. SOC.77, 3551 (1955). 192d. B. Douglas, J. L. Kirkpatrick, B. P. Moore, and J. A. Weisbach, Australkn J . Chem. 17,246 (1964). 1920. U. Renner, D. A. Prins, A. L. Burlingame, and K. Biemann, Helv. Chim. Acta 46, 2186 (1963). 192f. M. P . Cava, S. K. Talapatra, J . A. Weisbach, B. Douglas, and G. 0. Dudek, Tetrahedron Letters p. 53 (1963).

14. Aspidosperma

AND RELATED ALKALOIDS

513

192g. N. Finch, I. H. Hsu, W. I. Taylor, H. Budzikiewicz, and C. Djerassi, J. Am. C'hern. SOC.8 6 , 2620 (1964). 193. R. Goutarel, A. Rassat, M. Plat, and J. Poisson, Bull. Soc. Chim. France p. 893 (1959). 194. A. Bums, M. Osowiecki, and 0. Schindler, Compt. Rend. Acad. Sci. 247, 1390 (1958). 195. P. Tunman and J. Rachor, Naturwissenschaften 47, 471 (1960). 196. K. H. Palmer, Thesis, Nottingham Univ., Nottingham (1954);cited in reference 58. 196a. P. Relyveld, Pharm. Weekblad 98, 47 (1963). 197. G. Ganguli, N. Adityachaudhury, V. P. Arya, and A. Chatterjee, Chem. Ind. (London) p. 1623 (1962). 198. C. Vamvacas, W. von Philipsborn, E. Schlittler, H. Schmid, and P. Karrer, Helw. Chim. Acta 40, 1793 (1957). 199. N. Aditya Chaudhury, G . Ganguli, A. Chatterjee, and G. Spiteller, Indian J . Chem. 1, 363 (1963). 200. N. Aditya Chaudhury, C. R . Ghosal, and G. Ganguli, Indian J . Chem. 1, 95 (1963).

This Page Intentionally Left Blank

-CHAPTER

1

6

ALKALOIDS OF CALABASH CURARE AND STRYCHNOS SPECIES* A. R. BATTERSBY The Robert Robinson Laboratories, University of Liverpool, Liverpool, England

and

H. F. HODSON T h e Wellcome Research Laboratories, Beckenham, Kent, England

I. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

515

11. The Czo-Alkaloids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 A. C-Mavacurine, C-Fluorocurine, and C-Alkaloid Y . . . . . . . . . . . . . . . . . . . . . . 522 527 B. ~ - F l u o r o c u r i n e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Alkaloids of Strychnos melinoniana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 D. Alkaloids Related to Sarpagine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534

.

111. The Dimeric Alkaloids of Calabash Curare. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......................... , . . . . . . . . . A. Introduction. . . . . . . . . . . . . B. Toxiferine-I, C-Dihydrotoxi e-I, and Related Alkaloids. . . . . . . . . . . . . C. Caracurine V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Synthetic Work.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. C-Fluorocurarine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. The “Families” of Alkaloids.. . . . . . .. . . .. . . . . .. . . . . . . , . . . . . .. .. . . .. G. C-Alkaloid D, Caracurine 11, and Caracurine I1 Dimethochloride . . . . . . . . H. C-Calebassine, C-Alkaloid A, and C-Alkaloid F.. . . . . . . . . . . . . . . . . . . . . . . I. C-Curarine, C-Alkaloid E, and C-Alkaloid G . . . . . . . . . . . . . . . . . . . .

537 537 539 546 547 548 553 555 560 567

Appendix. Alkaloids of Unknown Structure from Calabash Curare and Strychnos Species. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

574

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

576

.

References.

I. Introduction The South American Indian arrow and dart poisons known as curares are all concentrated aqueous extracts of plant materials, usually prepared according t o well-established ritual. All are powerful poisons which cause rapid paralysis of voluntary muscle. The fascination of this subject is well presented in a monograph by McIntyre (1) on the history, preparation, * The majority of Strychnos alkaloids which are based upon the strychnine skeleton are covered in Chapter 17. 515

516

A. R. BATTERSBY AND H. F. HODSON

and pharmacology of curare. A UNESCO symposium held in 1957 covered all aspects of curare and of synthetic compounds with curare-like activity ; the proceedings have been published ( 2 ) . Curare is a generic term which includes several types of arrow poisons prepared in South America. Two more or less distinct groups have been studied chemically ;tube-curare, so called because it is packed in bamboo tubes; and calabash curare, which is packed in calabashes or gourds. Within each group the composition of individual curares varies according to geographical origin and, no doubt, also according to the accessibility of plant materials ; further differences must be introduced by variation in alkaloid content of the plant materials. The active principles of curare were early recognized to be watersoluble quaternary alkaloids. Tube-curare was studied by King, who isolated (3) the crystalline quaternary bisbenzylisoquinoline alkaloid, d-tubocurarine (I),in 1935. The main plant constituent of tube-curare is the bark of menispermaceous plants, particularly of the genus Chondrodendron, and d-tubocurarine was later isolated (4)from C. tomentosum. Further work led to the isolation and structural elucidation of many more bisbenzylisoquinoline alkaloids from these sources (5).

I

v

Tuboeurarine

Calabash curare originates in the more northern parts of the subcontinent, particularly in the Amazon and Orinoco basins and surrounding regions. It is considerably more toxic than tube-curare and has presented much more formidable chemical problems with regard to both isolation and structural elucidation. Tubocurarine and synthetic compounds with essentially the same action are now extensively used in surgery. With their aid, i t is possible to achieve the muscular relaxation for successful surgery while using a lighter, and consequently safer, degree of anesthesia than would otherwise be required. Other clinical applications of curariform agents are constantly under investigation. The interest attaching to the highly active calabash curare alkaloids is thus obvious and has led to an intensive chemical study of this group of compounds.

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

517

It is generally agreed that the barks of several South American Strychnos species, particularly S . toxifera Rob. Schomb., are important constituents of calabash curare and that these are the plants which provide the toxic principles in the extracts. More than a century ago the explorer Robert Schomburgk saw S . toxifera bark being used in the preparation of curare, and the same observation has been made recently. As with all curares, other plant material is added, possibly to modify the physical character of the poison and perhaps also for purely ritualistic purposes. Various qualitative studies have shown that not all the South American Strychnos species known to be used for the preparation of calabash curare contain toxic alkaloids ; furthermore, quite high activity has been demonstrated in extracts of several Strychnos plants believed not t o be so used; these include S . trinervis (Vell.) Mart. and S. gardneri A.DC. ( 6 ) . However, many studies, including the extensive chromatographic studies of Marini-Bettolo and his collaborators ( 7 ) ,have shown that there is a general similarity in alkaloid content between calabash curare and extracts from the bark of many Strychnos species. Only a few of the quaternary alkaloids so far isolated have high curare-activity, and these are responsible for the toxicity of the curares and bark extracts ; the majority ofthe pure alkaloids are inactive or only weakly active. At present there is no evidence against the view that all the toxic alkaloids of calabash curare are derived from Strychnos species ; it is probable that the same also holds true for most of the nontoxic curare alkaloids. The first chemical work on calabash curare was carried out in 1897 by Boehm (8))who isolated a highly active amorphous material which was named “curarine.” This was soluble in water and insoluble in ether, so it is probable that Boehm was handling a mixture of crude quaternary alkaloids. Much later (1935), King described (9) the preparation of an equally active amorphous quaternary iodide from the bark of S. toxifera. However, the first isolation of well-characterized crystalline alkaloids was achieved by H. Wieland and his school (10-13). Calabash curares were extracted with methanol, and the water-soluble quaternary alkaloids in the extract were precipitated as the reineckate salts; this mixture was then fractionated by adsorption chromatography on alumina. The various reineckate fractions so obtained were converted into the corresponding chlorides by successive treatment with equivalent quantities of silver sulfate and barium chloride ; some of the quaternary alkaloids then crystallized as the chlorides or as the picrates. C-Curarinel 1 This alkaloid was originally (10, 13) called C-curarine-I,but the Roman one has not been used in the recent papers dealing with the chemistry of this substance.

518

A. R. BATTERSBY AND H. F. HODSON

chloride was the first crystalline calabash curare alkaloid thus to be isolated.2 Other well characterized alkaloids isolated in the early work were C-calebassine and C-dihydrotoxiferine from calabash curare and toxiferine I and toxiferine I1 from the bark of S. toxifera. King used essentially the same method (14) for a study of the alkaloids from S . toxifera, and he isolated toxiferine I and toxiferine I1 together with a series of new alkaloids, all in very small quantity, designated toxiferines 111-XII. Recently, it has been shown (15) that several of these salts are identical with well-characterized alkaloids described after King’s original paper; moreover, it was shown (15) that toxiferine V and toxiferine XI are identical with toxiferine I . Although chromatography of the alkaloidal reineckates gave the first crystalline alkaloids, the method has several disadvantages ; it has been shown, for example, that several well-separated fractions can all contain the same alkaloid (15). This sort of problem and the fact that these alkaloids are very difficult to characterize and identify by classical methods caused some confusion in the early work. The experimental difficulties can be understood when it is realized that most of the calabash curare and Strychnos alkaloids are chemically similar, that their salts often have high and uncharacteristic melting points, and that the amounts available were very small indeed. A major step forward in this respect came with the application (16, 1 7 ) of paper chromatography; two-dimensional paper chromatography generally allows resolution of most alkaloidal mixtures. This led naturally to the application of partition chromatography on cellulose columns for the preparative fractionation of alkaloid chlorides, and this method is now clearly established as the most satisfactory one. The calabash curare or the plant material is extracted with methanol, water, aqueous alcohol, or methanol containing 3% acetic acid. Separation of the quaternary chlorides from the concentrated extracts can then be achieved by precipitation as a mixture of reineckates and recovery of the chlorides as described in the preceding paragraphs. Alternatively, the alkaloids may be precipitated as a mixture of picrates ( 1 8), from which the chlorides may be obtained by ion exchange. A number of solvent systems have been described for partition chromatography on paper and in columns. Those most widely used are the systems “C” and “ D ” of Schmid et al. The quaternary alkaloids are commonly isolated and handled as the chlorides, and it is therefore convenient to use the name of the alkaloid to mean alkaloid chloride. Thus throughout this chapter, C-calebassine means C-calebassine chloride, and other alkaloids will be treated in the same way. Anions other than chloride will be named, e.g., toxiferine I picrate. The letter C- is used to indicate that the alkaloid is derived from a calabash.

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

519

(16), which were used in the early paper chromatographic studies; solvent “ C ” is water-saturated methyl ethyl ketone containing 1-3% methanol ; solvent “ D ” is ethyl acetate :pyridine :water in the proportions 7 . 5 : 2 . 3 : 1.65 by volume. Recently, Boekelheide and co-workers reported separations in this field in which the cheaper aqueous butanol is used in place of solvent system “ D ” (18). Repeated chromatography using different solvent systems, and often in combination with adsorption chromatography on alumina, is usually necessary before crystalline alkaloids are obtained. Almost all the many pure calabash curare alkaloids have been isolated in this way, and the power of the method may be judged from the way in which pure alkaloids could be separated (16) from a sample of calabash curare which contained a t least 41 quaternary salts. The method has also allowed the isolation of crystalline alkaloids from S. toxiifera; this plant material was shown to contain a t least 30 different alkaloids (15). Not only is partition chromatography important for the isolation of the quaternary alkaloids, but also for their identification and characterization. The mobilities of the alkaloids in the various solvent systems are often more valuable than melting points. Moreover, the spray reagent (19) which is used t o detect the alkaloids (1% solution of ceric sulfate in 2 N sulfuric acid) gives striking and often highly characteristic color reactions with most of the calabash curare alkaloids. As can be seen from Table I, there is a good correlation between the color reaction and the UV-chromophore of the alkaloid. Cinnamaldehyde is also a useful spray reagent ( 1 7 ) . For characterization purposes, the R, values of the alkaloids are given as R, values, which are displacements calculated with respect to C-curarine, the most abundant and one of the most stable of the calabash curare alkaloids. Thus : R,for toxiferine I =

Distance moved by toxiferine I Distance moved by C-curarine

R, values are usually quoted for solvent system “C” containing 1% methanol and for solvent system “D.” Clearly, the precise R, value found will depend upon the conditions (type of paper, temperature, etc.) used for the run. A worker in this field should therefore standardize his conditions by determining the R, values of several well-characterized alkaloids. Although partition chromatography is established as the most powerful single method for separating these alkaloids, it can often be used to advantage in combination with other methods. Countercurrent distribution (20, 21) and electrophoresis (7, 20, 22) have both been used in this

TABLE I ALKALOIDS OF CALABASH CURARE AND Strychnos SPECIES m =-

R,value Alkaloid C-Mavacurine Melinonine A Melinonine B C-Alkaloid T Lochneram Macusine A Macusine B Macusine C Caracurine I1 Caracurine I1 dimethochloride =Toxifarhe I X C-Akaloid D Caracurine V Caracurine VII Hemitoxiferine I C-Curarine

Formula of cation or base

(solvent ',C'')

Chromophore (formula number)

Color with ceric sulfate (Immediate/After 20 min) Carmine Nil Nil Very pale red

0.8

VI VI VI VI VI VI VI VI IV: ether

0.42 0.34 1.4 2.1 1.5 1.0

IV; ether IV;R=H IV; ether I1 I1 111; ether

Violet Red-violet/Yellowish Purple-red/Brown Stable orange Stable orange Blue/Chrome green

2.7 4.0 2.7 -

3.1 3.6 3.0

Isolation reference" 17, 23, 24, 25 26 26 27

28

Pale grey Pale grey Pale grey Purple/Brown

m

5 2M

6

* W

1

a x

15

r

15 29 18, 24, 30

0

14, 15 16, 30, 33, 35 23, 24, 30 23,24 15, 31 10-13, 16, 17, 18,32-34, Refs. in 7

x u 8z

C-Alkaloid E C-Alkaloid G C-Calebassine

0.36 0.65 0.8

111; ether 111; ether 111; R = H

Blue/Chrome green Blue/Chrome green Blue-violet/Carmine

C-Alkaloid A C-Alkaloid F C-Alkaloid Y Toxiferine I

0.23 0.49 1.6 0.42

111; R = H 111; R = H 111;R = H Vb

Blue-violet/Carmine Blue-violet/Carmine Red-violet/Olive green Red-violet/Colorless

C-Dihydrotoxiferine I

1.2

Vb

Blue-violet/Colorless

Nordihydrotoxiferine I C-Alkaloid H Caracurine VI Melinonine G Melinonine F C-Fluorocurine $-Fluorocurine C-Fluorocurarine EC-Curarine I11

1.2 0.71 1.6 3.2 2.0 2.1 2.1 2.2

Vb Vb Vb IX IX VIII VIII

Violet/Pale brown Red-violet/Colorless Purple/Brown Nil Nil Red-violet/Brownish

LXXX

Blue-green/Yellow.green

16,35 16,35 13, 16, 18, 32-34,36 1 P 1 6 , 32, 35 16, 35 23,24 13, 14, 16, 37, 38, and Refs. therein 13, 16-18, 34, 35,39 39 16,35 23,30 40 40 16, 17, 24, 41, 42 34 12, 16, 33-36, 38,43

~

The references given relate to the isolation of crystalline alkaloid. Many of these substances have been detected chromatographically in other plant materials (see Ref. 7).

522

A. R. BATTERSBY AND H. F. HODSON

way ;the main difficulty with the former is that the mixture of quaternary alkaloids tends to promote stable emulsions. Quite early in the chemical studies of Wieland and King, evidence accumulated that the calabash curare alkaloids are indole derivatives, and with present knowledge it is possible to correlate the UV-spectra of many of them with one or another of the following related chromophores formally derived from the indole nucleus by oxidation, reduction, and substitution, or combinations of these processes. They are the indoline (11),3-hydroxyindoline and the derived ethers (111),N-hydroxyalkylindoline and its ethers (IV),2-methyleneindoline or 1 -vinylindoline (Va or Vb, respectively), indole (VI), oxindole or 1-acylindoline (VIIa or VIIb, respectively), $-indoxy1 (VIII), and /3-carbolinium (IX) systems ; it is not possible to distinguish with certainty by spectroscopic methods between the chromophores I11 and IV, between Va and Vb, or between VIIa and VIIb.

I1

I11

IV

Va

Vb

VI

11. The C,,-Alkaloids A. C-MAVACURINE, C-FLUOROCURINE, AND C-ALKALOID Y C-Mavacurine was isolated by Wieland and Merz ( 1 7 ) from a calabash curare, and C-fluorocurine was crystallized, also from a calabash, by Schmid and Karrer (41) ; both alkaloids were subsequently found in

15.

ALKALOIDS O F CALABASH CURARE ; 8hyChnOs

523

Venezuelan S. toxifera bark (30), and they have been detected chromatographically in extracts from the barks of several Strychnos species (20, 25, 43-45). A study of these two alkaloids gave a first insight into the structures of alkaloids from calabash curare and commenced with a demonstration that they are interrelated. The quaternary C-fluorocurine, CzoH25N20: (picrate, mp 179O, [aID + 326" in methanol) has the UV-spectrum and fluorescence in UV-light typical of compounds with the #-indoxy1 chromophore (35). With sodium borohydride, it is reduced to the quaternary hydrofluorocurine C20H27N20; which has an indoline chromophore (42). Under the influence of warm dilute sulfuric acid, hydrofluorocurine undergoes dehydration and rearrangement to give a product CzoHz5N20+which is a quaternary 2,3-disubstituted indole as shown by the UV-spectrum. This change involves a Wagner-Meerwein type of 1,2 shift and is typical (46) of 2,2-disubstituted 3-hydroxyindolines; hydrofluorocurine thus contains this system (XI), and the presence of a 2,2-disubstituted #-indoxy1 system in fluorocurine (partial structure X ) is confirmed. It was further found that the indolic product XI1 is identical with natural C-mavacurine, C20H25NzOf (picrate mp 1790-180°) (42). The latter can also be obtained in low yield from C-fluorocurine by reduction with zinc and sulfuric acid (42). The reverse change from C-mavacurine to C-fluorocurine was demonstrated (47) by making use of the many investigations of Witkop and Patrick (48) on the oxidation and rearrangement of simple tetrahydrocarbazoles. Oxidation of C-mavacurine with oxygen and a platinum catalyst gives a product having an absorption spectrum of the indoline type which in alkaline solution undergoes a reversible bathochromic shift characteristic of 2-hydroxyindolines. The compound has the properties of a 1,2-diol and must have the partial structure XIV; its formation by way of the peroxide (partial structure XIII) is straightforward. With acid, the diol (XIV) undergoes dehydration and rearrangement as illustrated in XV to give C-fluorocurine (partial structure X). This completes the cyclic set of transformations, as shown, which led to a further important correlation. The diol (XIV) was found to be identical with C-alkaloid Y, previously isolated (23) in small quantity from a calabash. Thus, a triad of related alkaloids was established. C-Fluorocurine has one N-methyl group attached a t the quaternary nitrogen atom and one C-methyl group which must be located in an ethylidene side chain, since the alkaloid gives acetaldehyde on ozonolysis. One oxygen atom is involved in the +-indoxy1 chromophore, and the other in an alcoholic hydroxyl group; acetylation gives an 0-acetyl derivative which can be readily hydrolyzed to regenerate the parent

524

A. R. BATTERSBY AND H. I?. HODSON

alkaloid. Thermal decomposition of C-fluorocurine chloride proceeds smoothly a t 19O0-23o0 to give the corresponding tertiary base, norfluorocurine, C19H22N202, which gives C-fluorocurine iodide when methylated with methyl iodide; it is thus established that no skeletal change has occurred during the pyrolysis step. As in the quaternary series, norfluorocurine (partial structure X ) can be reduced with lithium aluminum hydride or sodium borohydride to norhydrofluorocurine (partial structure XI) which is converted by acid into normavacurine (partial structure XII). Acetylation of norfluorocurine gives O-acetylnorfluorocurine, and this in turn can be methylated to give O-acetylfluorocurine (42).

i XV Ha

XIV

XI11

Selenium dehydrogenation of normavacurine gives a small yield of a tertiary base shown t o be a P-carboline derivative (see XVI) by its UV-spectrum (49). The minute quantity available (ca. 0.5 mg) precluded complete identification of this product. However, the UV-spectrum of the derived N,-methiodide3 of this base shows the highly characteristic absorption of the N,-methocarbolinium cation (XVII) and is substantially unchanged by alkali. This indicates that N, is alkylated, because treatment of the carbolinium salt (XVII; R = H) with alkali generates the mesomeric anhydro base (XVIII), and a profound change in the absorption spectrum occurs. The degradation base can then confidently be assigned structure XVI (R = alkyl) ; it was shown not to be identical with N,-methylcarboline (XVI; R = Me). 3 The nitrogen atom of the indole or indoline system of these alkaloids and their parent heterocycles is designated N, and the second nitrogen atom is referred to as N,.

15.

ALKALOIDS OF CALABASH CURARE ; Xtrychnos

525

The double bond present in the side chain of the tertiary bases norfluorocurine and norhydrofluorocurine can be selectively hydrogenated over platinum catalyst ; the $-indoxy1 system of the former alkaloid is not affected. The product from the latter base, dihydronorhydrofluorocurine, gives both acetic acid and propionic acid on modified Kuhn-Roth oxidation, thus confirming the presence in norhydrofluorocurine of the ethylidene grouping. I n contrast, catalytic hydrogenation of the quaternary hydrofluorocurine involves an Emde degradation to give a tertiary

XVII

XVI

indoline base, hexahydrofluorocurine, C20H30N202. Hexahydrofluorocurine contains two C-methyl groups and gives a-methylbutyric acid (XXI) on modified Kuhn-Roth oxidation (49). Thus, hydrofluorocurine contains the quaternary allylamine system (XIX) which undergoes Emde degradation and simultaneous reduction, as illustrated, to give the tertiary base (partial structure XX). It follows that the system X I X must also be present in C-mavacurine, C-flwrocurine, and C-alkaloid Y, since it is clear that the changes shown in partial formulas X to XV involve only the chromophoric groups. Me

Me

XIX

XX

XXI

This evidence can now be combined in partial structure X X I I for C-mavacurine. The analytical evidence and the functional groups just established require that C-mavacurine be pentacyclic and, on the basis of partial structure XXII, two carbon atoms are available for the construction of two further rings which must be free from C-methyl groups. These requirements can be metin a number of ways, but best by

526

A. R. BATTERSBY AND H. F. HODSON

the biogenetically plausible structure XXIII (49). At present, the chemical evidence allows positions 6, 14, 15, and 16 as possible sites for the hydroxyl group ; the first two are somewhat unlikely on biogenetic grounds, and position 15 is favored. If this structure is correct for C-mavacurine, then C-fluorocurine must be XXIV and C-alkaloid Y is

xxv.

The location of the hydroxyl group has considerable biogenetic interest, and it is to be hoped that further work will be undertaken to establish its position and to confirm the skeleton of these alkaloids.

XXII

XXIII C-Mavacurine

0

XXIV C-Fluorocurine

OH

XXV C-Alkaloid Y

It remains to consider the interesting results obtained by catalytic hydrogenation of C-mavacurine (49). Reduction of the alkaloid in dilute sulfuric acid over a platinum catalyst results in selective hydrogenation (uptake of 1 mole) of the indolic double bond; the quaternary indoline thus obtained still contains the ethylidene side chain and gives acetaldehyde on ozonolysis. I n alkaline solution over a platinum catalyst, the hydrogenation product is a tertiary indolic base, CzoHzsN20, designated €2-dihydromavacurine.This base contains only one C-methyl group and, on modified Kuhn-Roth oxidation, gives acetic acid only; €2-dihydromavacurine must therefore retain the ethylidene side chain of C-mavacurine. It follows that €2-dihydromavacurine is formed by simple Emde fission of the Nb-C-3 bond and is formulated as XXVI. The Emde base exhibits a typical indolic UV-spectrum in neutral and in alkaline solution. I n acid solution, the absorption is that of an indoline; the change is reversible as illustrated (XXVI +XXVII). Protonation of

15.

527

ALKALOIDS O F CALABASH CURARE ; 8tTYJChnOS

€2-dihydromavacurine thus takes place a t C-7 concomitant with a transannular reaction to give the indoline cation XXVII. Methylation of €2-dihydromavacurine with methyl iodide takes a similar course, in that the product, after conversion into the chloride by ion exchange gives a crystalline quaternary chloride which must have structure XXVIII on the following evidence. It was established that the quaternization reaction involves C-methylation by preparation of the salt XXVIII using radioactive methyl iodide ; Kuhn-Roth oxidation of the product then gives radioactive acetic acid. The methochloride (XXVIII) has an indoline absorption spectrum in neutral solution, but in alcoholic potassium hydroxide solution the spectrum is completely changed to that of a 2-methyleneindoline, owing to proton abstraction from C-3 which gives the base XXIX ; acidification of the solution regenerates the ion XXVIII. COHO

CHMe

CHMe

HO

HO

XXVI

Me

XXVIII

XXVII

Me

XXIX

B. +-FLUOROCURINE This alkaloid, isolated from a calabash (34),has color reactions, R, values, a UV-spectrum, and a melting point of its picrate (179') identical with those of C-fluorocurine. However, unlike the latter alkaloid, it does not form a sparingly soluble iodide or a sparingly soluble p-nitrophenylhydrazone. Its relationship to C-fluorocurine is at present obscure.

528

A. R. BATTERSBY AND H. F. HODSON

C. ALKALOIDS OF Strychnos melinoniana Although it is reported that the bark of S . melinoniana Baillon is used in calabash curare preparation (1)) extracts of this species show no curare activity (50). The bark, however, has a relatively high content of quaternary alkaloids and was first examined by Schlittler and Hohl (26). Fractionation by chromatography of the mixed alkaloidal reineckates gave two alkaloids, melinonine A (0.61% of dried bark) and melinonine B (0.03%), both isolated as crystalline chlorides. A later investigation, by the Zurich group (40), employed partition chromatography and resulted in the isolation of melinonines A and B, together with C-fluorocurine and C-mavacurine and a series of new alkaloids, melinonines E, F, G, H, I, K, L, M ; all except melinonine L are quaternary salts. Most of these alkaloids have been studied chemically. Melinonine A, C ~ ~ H Z ~ N ~(chloride, O:, mp 260'-261', ["ID - 110' in water) contains one methoxyl group, one C-methyl group, and one N-methyl group attached at the quaternary Nb nitrogen atom (26). It gives no color reaction with ceric sulfate, but the Hopkins-Cole reaction is characteristic of tetrahydroharman derivatives. The IR-spectrum of melinonine A shows the absence of hydroxyl groups and, in keeping with this, neither acetyl nor benzoyl derivatives can be formed.

I

R

xxx

I I MeOzC-C=C-0XXXI

Pyrolysis of the alkaloid chloride a t 210" smoothly gives the corresponding tertiary base, normelinonine A, C21H24N203, which can be methylated at Nb to regenerate melinonine A. Normelinonine A is not

15.

ALKALOIDS OF CALABASH CURARE;

Strychnos

529

hydrogenated over Adams catalyst in acetic acid. Zinc dust distillation of normelinonine A gives a mixture of 3-methylindole and 3-ethylindole, whereas selenium dehydrogenation gives alstyrine (XXX; R = Et). This base is also formed by dehydrogenation of serpentine, corynantheine, and alstonine, all cr-indole alkaloids with a heterocyclic or cleaved ring E . The UV-spectrum of normelinonine A is very similar to that of those indole alkaloids which contain a 2,3-disubstituted indole chromophore together with the chromophore X X X I ; for example, the spectrum is identical with that of tetrahydroalstonine (XXXII). These results strongly suggested that normelinonine A is either tetrahydroalstonine (51) itself or some stereoisomer with this gross structure. I n fact, careful comparison showed the former possibility to be the correct one, and it therefore follows that melinonine A has the structure XXXIII (26). The stereochemistry and absolute configuration shown follow from subsequent work on tetrahydroalstonine (52, 53). Melinonine B is formulated as CzoHz7NzOf on the basis of analyses of the chloride, iodide, and perchlorate; the chloride, mp 311' (dec.), has [.ID - 14.8' (methanol-water) (54). The alkaloid has one N-methyl group but no methoxyl or C-methyl groups. Its UV-spectrum is that of a 2,3-disubstituted indole and its IR-spectrum exhibits both hydroxyl (3.05 p ) and >N-H (3.20 p ) absorption. Acetylation of melinonine B gives a crystalline 0,N-diacetyl derivative which results from attack at the indolic N,, since the UV-spectrum of this derivative is that of an N-acylindole and its IR-spectrum has both ester and amide bands. Unlike melinonine A, pyrolysis of melinonine B chloride does not give a good yield of the corresponding tertiary base. The only crystalline product from the reaction is a base, mp 196.5'-198.5' designated TI, obtained in low yield, which is not converted into the parent alkaloid by methylation; it appears to be an N,-methyl derivative of the desired nor base formed by a transmethylation reaction. Methylation and chromatography of the amorphous pyrolysis product show that some normelinonine B is present, but it was not possible to isolate it. The UV-spectrum of melinonine B is identical with the spectra of yohimbine and its derivatives, apart from a hypsochromic shift of 5-8 mp. This " short-wavelength indolic absorption is characteristic of yohimbine derivatives protonated or quaternized a t the Nb nitrogen atom; thus, the spectrum of melinonine B is identical with that of yohimbol N,-methochloride (N,-methoderivative of XXXIV). These indications from UV-spectroscopy are supported by dehydrogenation of melinonine B over palladium which gives yobyrine (XXXV),previously obtained under these conditions from a variety of yohimbine alkaloids with a carbocyclic ring E. Dehydrogenation of melinonine B with ))

530

A. R. BATTERSBY AND H. I?. HODSON

selenium gives a small amount of a base shown to be an a-pyridylindole derivative by the identity of its UV-spectrum with that of alstyrine ( X X X ; R = E t ) . This base is not identical with alstyrine itself nor with de-ethylalstyrine ( X X X ; R = H) and in all probability is the lower homologue ( X X X ; R = Me). However, its identification as an a-pyridylindole gives valuable structural information.

XXXIV

OH

xxxv

XXXVI

0

CHz

I

CHzOH

CHz. O H XXXVII Melinonine B

XXXVIII

I

XL

CHzOH

All the foregoing evidence supports the formulation of melinonine B as the N,-metho derivative of one of the eight possible stereoisomers of yohimbol (gross structure XXXIV) ; such a formulation is biogenetically acceptable and was tested as follows. Four stereoisomers of yohimbone (XXXVI) are known, corresponding to the various possible arrangements a t the centers 3 and 20, with the hydrogen atom a t position 15 in the a-configuration. These four ketones were reduced separately, and the resulting mixture of two epimeric alcohols in each case was quater-

15.

ALKALOIDS OF CALABASH CURARE ; Xtrychnos

531

nized with methyl iodide. Finally, fractionation by partition chromatography yielded all eight stereoisomeric N,-metho yohimbols (see gross structure XXXIV) ; the Zurich group (54) proved that none of these is identical with melinonine B. However, a t this stage further evidence was obtained (54) which is clearly inconsistent with a hydroxyyohimbane structure. Melinonine B gives formaldehyde on ozonolysis, and on catalytic hydrogenation yields dihydromelinonine B, C20H29N20+, and the IRevidence suggests that this reduction involves simple saturation of a double bond. Dihydromelinonine B contains one C-methyl group, and on modified Kuhn-Roth oxidation it gives both acetic and propionic acids. This establishes the presence of a C-ethyl side chain in dihydromelinonine B, and so melinonine B itself contains a vinyl side chain and only four rings. Since melinonine B contains no C-methyl group, the hydroxyl group must be placed in a 2-hydroxyethyl side chain which leads to structure XXXVII as a reasonable possibility for this alkaloid, although it is by no means firmly established. For example, the original workers pointed out that the evidence allows no distinction to be made between structure XXXVII and a structure with the side chains a t positions (3-15and C-20 interchanged, although the latter is less attractive biogenetically (54). If structure XXXVII does, in fact, represent melinonine B, then dihydromelinonine B must have the gross structure XXXVIII. Two stereoisomers of the structure XXXVIII were prepared by the Zurich group. Dihydrocorynantheal (55) was reduced with lithium aluminum hydride to give dihydrocorynantheol, which on methyllation with methyl iodide, followed by ion exchange, gave the crystalline N,-methodihydrocorynantheol chloride (XXXIX). Similarly, corynantheidal (56) was converted into corynantheidol and thence into N,-methocorynantheidol chloride (XL).Neither of the salts is identical with dihydromelinonine B chloride. Therefore, if melinonine B has gross structure XXXVII, it must belong to the pseudo or epiallo series (54). The formation of yobyrine by palladium dehydrogenation of a compound with structure XXXVII is surprising. I n every other case, yobyrine has only been produced by dehydrogenation of yohimbine-like alkaloids with a carbocyclic ring E, such as yohimbine (XLI) itself, and has been regarded as evidence for this feature. It is clear that the chemistry of melinonine B is sufficiently unusual to attract further efforts toward a complete elucidation of structure. Melinonine F, C13H13N2f, was isolated as the crystalline chloride, mp 288O, and it also affords a crystalline picrate (40).Its UV-spectrum in neutral and in alkaline solution is identical with that of /3-carbolinium

532

A . R. BATTERSBY AND H. F. HODSON

salts, which led to a direct comparison of this alkaloid with the N,metho derivative (XLII) of harman. The two were identical. Although harman itself has been isolated several times from natural sources, this is the first reported occurrence of the quaternary derivative. Melinonine G which gives a crystalline iodide and picrate, mp 229.5"230.5', is formulated as C17H15Na; it contains no N-methyl groups (40). I n keeping with its low hydrogen content, its UV-spectrum in neutral and in alkaline solution is closely similar to that of sempervirine salts

Me02C-*'V

OH XLI Yohimbine

XLIII Seinpervirine

XLII Melinonine F

XLIV Melinonine,G

HOCHz'\/ XLV

I

1

X LVI

(XLIII); the IR-spectrum of its iodide is also similar to that of sempervirine iodide and further shows the absence of a vinyl group. Catalytic hydrogenation of melinonine G over Adams catalyst in aqueous alkaline solution gives an indole which yields propionic and acetic acids on modified Kuhn-Roth oxidation and which, therefore, contains a C-ethyl group. All this evidence is accommodated in structure XLIV for melinonine G ; the indolic reduction product must then have structure

15.

ALKALOIDS OF CALABASH CURARE;

Strychnos

533

XLV. Structure XLIV is, in fact, the one established (57, 58) for flavopereirine, isolated from Geissospermum species (Apocynaceae), and subsequently confirmed by many syntheses (59-66). Surprisingly, no direct comparison of the two alkaloids has been reported. Melinonine G (flavopereirine) has considerable biogenetic interest because it lacks the two or three carbon atoms which in most other indole alkaloids are attached at the asterisked position (see XLIV). Thus, this alkaloid may be regarded as a partly degraded system or, on more recent views (67, 68), the three-carbon unit may never have been attached. Because of the small amounts isolated, much less is known about the remaining alkaloids of S. melinoniana. Melinonine E, isolated as the picrate, double mp 120.5"-122" and 2 16"-219", also gives a crystalline perchlorate and nitrate ; analyses of these salts allow no distinction between the two possible formulas, C20H23N20+and C2oH25N20+.The alkaloid has no N-methyl, 0-methyl, or C-methyl groups, and the Doeuvre test indicates the absence of vinyl groups. The UV-spectrum is of the type shown by /3carbolinium salts and is very similar to that of melinonine F. Acetylation with acetic anhydride in pyridine gives an 0-acetyl derivative isolated as the crystalline picrate. On this slender evidence, the biogenetically possible structure XLVI is suggested as a working hypothesis for melinonine E (40). Melinonine H, formulated as C ~ O H ~ I - ~ ~from N ~ analyses O+ of the picrate, mp 290"-292", and perchlorate, has one N-methyl and one C-methyl group. Its UV-spectrum shows that this alkaloid is the only quinoline derivative so far isolated from calabash curare or Strychnos species. It is not identical with the quaternary N,,-metho derivatives of cinchonine or cinchonidine (40). Melinonine I and melinonine K were isol.ated as crystalline picrates, mp 160"-170' and 196"-199", respectively, although in amounts insufficient to allow their empirical formulas to be established (40). They are both indoles, and their UV-spectra suggest that they may be 5- or 8-hydroxyindoles. The tertiary alkaloid melinonine L was isolated as the free base, C22H26N204, mp 248"-250" (40). It has one methoxyl group, one Cmethyl group, and one N-methyl group; two active hydrogen atoms were shown to be present by the Zerewitinov determination. With acetic anhydride in pyridine, it gives a crystalline monoacetyl derivative. The UV-spectrum of melinonine Lis very similar to that of melinonine Abut, unlike the latter, undergoes a marked bathochromic shift in alkaline solution. The alkaloid is not quaternized by methyl iodide under normal conditions, but i t reacts with dimethyl sulfate t o give a quaternary salt which can be isolated as the picrate.

534

A. R . BATTERSBY AND H. F. HODSON

Melinonine M, picrate mp 242'-245O, is the least abundant alkaloid of S . melinoniana. The only information a t present available is that it is an anhydronium base with a UV-spectrum similar to those of melinonine E and melinonine F (40).

D. ALKALOIDS RELATEDT O SARPAGINE Several alkaloids, which have structures closely related to that of sarpagine (XLVII; R = H), have been isolated from calabash curare and S. toxifera. Full details of the work leading to their structures and stereochemistry are given in Chapter 22 ; here only a short outline will be given. 1. C-Alkaloid T (0-Methylsarpagine)and Lochneram

Fractionation of the contents of a Brazilian calabash (27) yielded a tertiary base, C-Alkaloid T, C20H24N202, mp 202.5"-203.5", ["ID + 72" (ethanol). This was found t o be indistinguishable from O-methylsarpagine (69) (identical with lochnerine) (70) which is now proved ( 7 1 ) to have the structure XLVII (R = Me); evidence for the illustrated configuration of the ethylidene system will be presented subsequently. CAlkaloid T, however, may be a difficultly separable mixture of vinyl and ethylidene isomers as illustrated in formula XLVIII, since ozonolysis afforded a mixture of formaldehyde and acetaldehyde. It should be made clear, however, that this possibility of a mixture depends entirely on the results from ozonolysis. Sarpagine (XLVII; R = H) also gives formaldehyde and acetaldehyde on ozonolysis (27). This same Brazilian calabash curare also yielded (28) a quaternary alkaloid, CzlH27N20$, mp of iodide 235"-238", ["ID + 41' (96% ethanol). This was named lochneram, since it was proved (28) to be identical with N,,-metho lochnerine (XLIX). Moreover, lochneram gave only acetaldehyde when subjected to ozonolysis and is therefore regarded as the pure ethylidene isomer (XLIX). 2. Macusines A , B, and C

These alkaloids formed a difficultly separable mixture in the "fastrunning" alkaloidal fraction from S. toxifera (15). Macusine B chloride, mp 248"-249O, ["ID +16O (in water), has the molecular formula C20H25NzO+Cl-. It undergoes smooth pyrolytic decomposition to the corresponding tertiary base, normacusine B, CigH22N20, and most of the structural work was carried out on this material. The key stages in this involved (72) the demonstration that

15.

ALKALOIDS O F CALABASH CURARE ; 8tTyChnOS

535

normacusine B contains a 2,3-disubstituted indole system, a primary alcohol function, and ethylidene double bond (NMR),and a caged system around N, which prevents the introduction of a >C=N(b)- residue by any of the standard reagents for such dehydrogenations. When these results are considered in conjunction with the usual biosynthetic arguments, they lead t o structure L as the most probable one for normacusine B. Confirmation was obtained by direct correlation (72) of normacusine B with deoxyisoajmalol B (see Chapter 2 2 , Section I) of established structure and absolute stereochemistry (71). Structure L is thus a complete expression for normacusine B, including the absolute stereochemistry; only the configuration of the ethylidene system is left unknown by the foregoing work. The configuration has been proved t o be as shown in formula L by correlation of normacusine B with macusine A of known structure by X-ray analysis (see later discussion). Macusine B therefore has the structure and absolute stereochemistry shown in formula LI. Normacusine B has also been shown (72) to be identical with deoxysarpagine (XLVII; RO = H) (71), and so the earlier correlations allow the illustrated configuration to be written for the ethylidene systems in sarpagine (XLVII; R = H), 0-methylsarpagine (XLVII; R = Me), and lochneram (XLIX). More recently, many other alkaloids have been correlated with normacusine B, and the same ethylidene configuration is thereby established for all of them; these include tombozine (73), akuammidine ( 7 4 , 7 5 ) ,voachalotine (76), polyneuridine (77), vincamajine (74), vincamedine (74), vellosimine, and geissolosimine (78). It is interesting that the ethylidene group in echitamine (79) (Chapter 8) has the same configuration. The amount available of macusine A chloride, C22H27NzO$Cl-, mp 252', [.ID -58' (in water), was initially too small for a solution of its structure by chemical means. Structural analysis by the X-ray method (80) established the constitution L I I I with only the absolute configuration to be determined; this is considered later in this discussion. When the structure of polyneuridine was determined (77), it was found t o be the tertiary base corresponding t o macusine A. All the chemical work (80, 81) on macusine A is in accord with the given structure. Pyrolysis of macusine A chloride did not yield the expected nor-base but rather the ester (LII) formed by loss of formaldehyde in a retroaldol reaction. This was established by reducing the base from the pyrolysis with lithium aluminum hydride to afford normacusine B (L). With the absolute configuration of normacusine B firmly established, it follows that the absolute configuration shown in formula LIII correctly represents macusine A.

536

A. R . BATTERSBY AND H. F. HODSON

Macusine C chloride, C ~ Z H ~ ~ N ~ O mp $ -260"-261", , UI.[ - 61" (in water), is isomeric with macusine A ; that the two alkaloids differ only in the configuration a t position 16 is shown by the following results (81). Pyrolysis of the chloride gave in this case a mixture of two bases, one corresponding in molecular weight (mass spectrum) to normacusine C. The second base was shown to be the same retroaldol product (LII)

Me XLVII R = H ; Sarpagine R = Me; 0-Methylsarpagine

XLIX Lochnerarn

Me

XLVIII C-Alkaloid T

Me

L Normacusine B

Mc LIII Macusine A

Me LIV Macusine.C

15.

ALKALOIDS OF CALABASH

CURARE; Xtrychnos

537

produced earlier from macusine A by reducing it with lithium aluminum hydride to normacusine B (L). This result and the known configuration a t position 16 in macusine A allow the structure LIV to be written for macusine C, which also correctly represents the absolute configuration and the stereochemistry of the ethylidene system. Macusine C is thus the Y,-metho derivative of akuammidine (74). The production of the ester LII in the pyrolysis of macusine C chloride is owing to the inversion of the carbanion iiitermediate in the reverse-aldol step, as would be expected, since the carbomethoxy group occupies the less compressed position in this configuration ( 7 1). A determination of the structure of akuammidine by X-ray analysis ( 7 5 ) provides a second starting point for the collection of a set of interrelated structures having the same configuration a t the ethylidene group ; again, one must make use of the various chemical correlations already outlined in the chapter.

111. The Dimeric Alkaloids of Calabash Curare A. INTRODUCTION

As the work progressed on the iaolation of pure materials from calabash curare and Xtrychnos species, it became clear that there are two fairly well-defined groups of alkaloids. One group contains those which have comparatively high mobilities in the various solvent systems used for partition chromatography, and these are found to have little or no curarizing activity. All the alkaloids discussed so far fall into this group, and they are based upon C19 or C20 skeletons. The second group contains those alkaloids with high curarizing activity ; these are all found to move slowly on paper chromatograms and cellulose partition columns. All the early work on this second group of alkaloids was confused because there was no certain knowledge of their molecular formulas. Formulas based upon the presence of two nitrogen atoms in molecules of about twenty carbon atoms were assumed. A major step forward was taken when it was shown by the Zurich group (82) that the alkaloids of this group contain four nitrogen atoms in a unit of size around c40. Two of the four nitrogen atoms are weakly basic and correspond to the N, nitrogen atoms of the alkaloids considered so far ; the other two are the quaternized Nb atoms. The method used to determine the molecular size invoIves pyrolysis of the alkaloid chloride LV t o yield the corresponding nor base, that is,

538

A. R . BATTERSBY AND H. F. HODSON

the tertiary base LVI. Several of these large quaternary alkaloids can be made to give good yields of the nor base if a small quantity of the quaternary salt is spread as a very thin film over the surface of a glass bulb and this is then evacuated t o around 10-4 mm and plunged into a metal bath a t about 250O. The nor base, for example, norcurarine (LVI), is formed and distills rapidly onto the cooler parts of the bulb. By treating the nor base (LVI) with one-half equivalent of mineral acid, an I8

81

c1Q -N-N-

I

I

Cl0

Me

Me LV

J I

I8 -N-N-

7

I

H LVI

I8

A

ICE

-N-N-

I H

I

H LVIII

LVII

I -N-N-

I8

ICE

I

I

Me

Me

LIX

-N-N-

I8

18

I

I

H

Me

LX

ICE

I8

-N---NI

I

H

H LXI

equilibrium mixture of the species LVI, LVII, and LVIII is formed. On methylation of the total mixture of salts, LVI is converted into the starting material, C-curarine (LIX = LV), the diprotonated ion (LVIII) is unaffected, whereas the monoprotonated salt (LVII) yields mono-N,methonorcurarine (LX). These substances are readily separable by partition chromatography, and the mono-metho derivative (LX) can be further N,,-methylated t o afford C-curarine (LIX). This formation of a molecule is monoquaternary salt carrying one N-methyl group in a conclusive proof that the original C-curarine possesses two quaternary N,-methyl residues in a C40 system. The success of this method depends upon the fact that the two basic N, nitrogen atoms in the tertiary base (LVI) are sufficiently separated in space so that protonation or quaternization a t one basic nitrogen atom does not greatly affect either of these processes a t the second basic center. This same method, used initially for C-curarine, also showed C-calebassine to be C ~ O H ~ ~ N ~and O ; C-dihydrotoxiferine + I to be C40H46Nq++

15.

ALKALOIDS OF CALABASH CURARE

; Xtrychnos

539

(82). Indeed, all the calabash curare alkaloids with high curarizing activity are now known to possess C40 molecules. The converse, however, is not true, for caracurine V dimethochloride (Section 111, C) has the molecular formula C ~ ~ H ~ G N ~yet O ; +has , a surprisingly low physiological activity. Tubocurarine (I) and the various synthetic curarizing agents such as succinylcholine (LXII) have two quaternary nitrogen atoms set some distance apart. The foregoing experiments allow the calabash curare alkaloids to be placed in the same class of compounds.

B. TOXIFERINEI, C-DIHYDROTOXIFERINE I, AND RELATEDALKALOIDS In their early isolation work on X. toxifera grown in British Guiana, Wieland et al. (13) obtained the quaternary alkaloid toxiferine I, which has a remarkably high physiological activity. The same alkaloid was later isolated by Schmid and Karrer (37) from a Venezuelan calabash, and recently it has been shown (15) that two of the picrates isolated by King (14) from S . toxifera bark are, in fact, toxiferine I picrate. King’s plant material was the same as that used by the German workers. Also from a Venezuelan calabash, Wieland et al. (13) isolated an alkaloid with properties very similar to those of toxiferine I which they called C-dihydrotoxiferine I, although no formal relationship to toxiferine I was demonstrated. Indeed, the true relationship between the two alkaloids has only been established as a part of the recent elucidation of their structures. The two possess the same constitution, apart from the replacement of the two -CH20H groups present in toxiferine I by -CH3 groups in “C-dihydrotoxiferine I.” The latter name can be seen to be a misnomer, and C-bisdeoxytoxiferine I would be more appropriate. Nevertheless, the incorrect name is so firmly established in the literature that any attempt to change it a t this stage would only lead to confusion. The tertiary base corresponding to C-dihydrotoxiferine I, named nordihydrotoxiferine, has also been isolated (39) from the tertiary fraction of a Venezuelan sample of S. toxiferu bark. C-Dihydrotoxiferine I chloride, C40H46NtfC1r, [aID - 600’ (1 : 1 aqueous alcohol), has two N-methyl groups attached a t the quaternary Nb nitrogen atoms (39). Molecular distillation of the alkaloid chloride gives nordihydrotoxiferine with loss of methyl chloride ; this ditertiary base can be converted back into the bisquaternary alkaloid, as the diiodide, by methylation with methyl iodide (39). Dehydrogenation of C-dihydrotoxiferine I with sulfur or with zinc dust gives isoquinoline

540

A. R. R A T T E R S B Y AND H. F. HODSON

(83), whereas distillation with zinc dust gives a mixture of 3-methylindole and 3-ethylindole (83).Palladium dehydrogenation of nordihydrotoxiferine gives a trace of a P-carboline derivative (39). C-Dihydrotoxiferine I yields acetaldehyde on ozonolysis (39) and, as will be seen subsequently (Section 3, F),this alkaloid can be converted into C-calebassine and into C-curarine, both known to contain the grouping LXIII. Ii; is thus highly probable that C-dihydrotoxiferine I also contains this quaternary allylamine system (LXIII). I n common with toxiferine I, the UV-spectrum of C-dihydrotoxiferine I is almost identical with that of simple a-methyleneindolines (LXIV),and it was assumed a t this stage that this represented the chromophore of the two alkaloids (35). The amount of alkaloid isolated was in all cases very small, and the foregoing account covers all that was known about C-dihydrotoxiferine I a decade after its first isolation. e3

C H Z .CO .0 .C H Z.CHz. NMe3

I

e3

CHz. CO. 0 .CH2. CH2. NMe3 LXII

\m 1 --NzCHz. C=CHMe /

LXIII

LXIV

Still less was known about toxiferine I a t this time. Because of strong similarities between the properties of toxiferine I and those of curare alkaloids with established C49 molecular formulas, the earlier formulas were tentatively revised to C40H46-48NzOi+ (82).Toxiferine I dichloride has [“I,, - 540’ (in water) and the picrate has mp 278”-280’ (dec.). The first important contribution to the structural elucidation of C-dihydrotoxiferine I and toxiferine I came from investigations by the Karrer-Schmid group (24)of the acid-catalyzed transformations of some of the nine tertiary aikaloids, caracurines I-IX, which had earlier been isolated (30) from Venezuelan S. tozifera bark. Caracurine V is rapidly converted by dilute mineral acid into an unstable base, caracurine Va, with UV-spectrum and color reactions very similar to those of C-dihydrotoxiferine I and toxiferine I. A slower acid-catalyzed change then takes place to convert caracurine Va into a mixture of caracurine I1 and caracurine VII ; caracurine VII has an indoline UV-spectrum and shows a highly characteristic orange color reaction with ceric sulfate. The similar dilute acid-catalyzed transformation of toxiferine I dichloride was investigated by Battersby and Hodson (84, 31), who isolated two crystalline quaternary products from the reaction mixture. One was shown to be identical with a quaternary alkaloid, provisionally called AS, which they had previously isolated (15) from S. toxifera bark

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

541

from British Guiana. Alkaloid A8 chloride has indoline UV-absorption, a secondary N, nitrogen atom, and gives an orange color reaction with ceric sulfate ; it was proved to be identical with caracurine VII methochloride ( 3 1) . The two sets of transformations just described take place, therefore, on related tertiary and quaternary molecules. Caracurine Va must be nortoxiferine I, aiid the second product obtained by Battersby and Hodson must be caracurine I1 dimethochloride; these ideniities were confirmed by the necessary methylations and comparisons (31, 85). It was also shown that alkaloid A8 in hot acetic acid is converted to toxiferine I in moderate yield (31). These transformations are summarized in Scheme I, which includes the subsequent demonstration (Section 111, G) that the formation of caracurine I1 and its dimethochloride involves atmospheric oxidation. Caracurine V

Caracurine I1

HsO+/Oz f---

Caracurine Va

1

Methylation

Methylation HsO+IOa

Caracurine I1 methochloride

&O+

w

Caracurine VII Methylation

&O+

Toxiferine I

7-f

Alkaloid A8

AcOH

SCHEMEI

Karrer, Schmid, and others (24, 86) examined the degradation of Cdihydrotoxiferine I which, with dilute mineral acid, is slowly converted into C-alkaloid D, previously isolated from a calabash. The course of this reaction was followed by paper chromatography, and it was found that a fast-running component with an orange ceric sulfate reaction rapidly appears in the reaction mixture and then slowly disappears with the formation of C-alkaloid D. Under modified conditions, the fast-running component was isolated as the main product and named hemidihydrotoxiferice I (for reasons that will be seen later) (86);it was subsequently shown (Section 111, G) that C-alkaloid D arises in the above reaction mixture by atmospheric oxidation. Hemidihydrotoxiferine I in aqueous acetic acid is converted back into C-dihydrotoxiferine I (86). Thus can be written the fol!owing reaction scheme, essentially similar to that given previously for toxiferine I and caracurine V, but characterized by an equilibrium between C-dihydrotoxiferine I and hemidihydrotoxiferine I, which even in aqueous acid allows the presence of appreciable

542

A. R. BATTERSBY AND H. F. HODSON

amounts of the former. I n contrast, toxiferine I is converted almost quantitatively into alkaloid A8 under these aqueous conditions. C-Alkaloid D

HS0+

H30+/0z

Hemidihydrotoxiferine I

aC-dihydrotoxiferine I ACOH

Very shortly after these many relationships had been established came the highly important discovery by the Zurich group (87) that caracurine V l I is identical with the Wieland-Gumlich aldehyde (LXV), first obtained (88) 25 years earlier as a degradation product from strychnine (LXVI). The aldehyde group in the base LXV is “masked” as a cyclic hemiacetal (89). This identification is of further interest in showing the natural occurrence of an intermediate proposed by Woodward (90) in his biogenetii: scheme for strychnine. There is also set up in this way a structural link between the alkaloids of South American and North American Strychnos species. Alkaloid A8 chloride has already been shown t o be caracurine VII methochloride and is thus the Wieland-Gumlich aldehyde N,-methochloride LXVII (hemitoxiferine I), which was confirmed (31) by direct comparison.

OC\/’\O/’

HOr\O/

LXV Wieland-Gumlich Aldehyde Caracurine VII /\e

k\

LXVI Strychnine

/\ \/\NA\i/*-H ((1) H H

LXVII Hemitoxiferine I

c1”

/\e

ClO

H

/

I

CHO

LXVIII Hemidihydrotoxiferine I

Hemidihydrotoxiferine I chloride contains >N,-H and its I R spectrum indicates the presence of an aldehyde group. On the basis of this and the foregoing knowledge, the structure LXVIlI was proposed (86) for it, i.e., that of the 18-deoxy-Wieland-Gumlich aldehyde methochloride, and this structure was confirmed in the following manner

15. ALKALOIDS OF CALABASH CURARE ; Stryehnos

543

(85). The degradation of C-dihydrotoxiferine I by acid, discussed previously, can also be carried out on the related tertiary base nordihydrotoxiferine I to give norhemidihydrotoxiferine (tertiary base corresponding to LXVIII) which, with sodium borohydride, is reduced to the primary alcohol (LXIX). The alcohol is identical with that prepared from the Wieland-Gumlich aldehyde by reducing it with lithium aluminum hydride to give the diol (LXX ; R = OH), followed by selective bromination of the allylic hydroxyl group to yield the halide (LXX; R = Br). Reductive dehalogenation with zinc in acetic acid then gives the desired product (LXIX). The relative and absolute stereochemistry of strychnine is established (91) to be as shown in formula LXVI, and it follows that hemitoxiferine I (LXVII) and hemidihydrotoxiferine I (LXVIII) have the relative and absolute configurations illustrated.

LXIX

1,XX

All the foregoing information allows one to state that the formation of C-dihydrotoxiferine I from hemidihydrotoxiferine I (LXVIII) in acetic acid involves the condensation of two molecules of LXVIII, with the loss of two molecules of water and with the disappe,arance of the aldehyde and >N,-H functions. Evidence for such a condensation is provided by the fact that a mixture of equivalent amounts of norhemidihydrotoxiferine I and hemidihydrotoxiferine I in acetic acid gives a reaction product containing nordihydrotoxiferine I, its mono-N,-metho salt, and C-dihydrotoxiferine I (86). The formation of toxiferine I from hemitoxiferine I (LXVII) must be strictly analogous, and ' the only structures which can be written to accommodate the above evidence are LXXI (R = H) or LXXII (R = H) for C-dihydrotoxiferine I and LXXI (R = OH) or LXXII (R = OH) for toxiferine I. Structure LXXIII can be written as a formal representation of the intermediate in both condensation and fission reactions. Of these alternatives, the structures LXXI were a t first preferred (86, 31), mainly because of the striking similarity between the UV-spectra of the two alkaloids and those of simple cr-methyleneindolines. The above chemical evidence allows of no distinction between formulas LXXI and LXXII. Boekelheide and his collaborators sought to distinguish between these possible structures by NMR-spectroscopy (92). The olefinic region

544

A. R. BATTERSBY AND H. F. HODSON

of the NMR-spectrum of C-dihydrotoxiferine I shows a sharp peak, assigned to the C-17 and C-17' protons in structure LXXII (R = H), superimposed upon an apparent quartet obviously owing to the (2-19, (2-19' protons; the total area of the olefinic absorption was stated to be equal to half that of the total aromatic absorption. Assuming the correctness of these assignments, the spectrum is clearly in accord only

LXXII

LXXI

R = H; C-DihydrotoxiferineI R = OH; Toxiferine I

LXXIII

with structure LXXII (R = H ) for C-dihydrotoxiferine I ; the spectrum of a compound of structure LXXI (R = H ) should exhibit only a quartet in the olefinic region owing to the C-19, C-19' protons, the area being q u a 1 to one-quarter of the total aromatic absorption. Similar evidence was adduced from the spectrum of toxiferine I in support of structure LXXII (R = OH) for the alkaloid. A more comprehensive study of the NMR-spectra of C-dihydrotoxiferine I, toxiferine I, their derivatives, and model compounds was made by the Swiss workers who showed that the above assignments are not

15.

ALKALOIDS O F CALABASH CURARE ; f&yChnOs

545

completely coi.rect (94). The ratio of “aromatic” to “olefinic ” absorption in C-dihydrotoxiferine I is, in fact, 5 : 2; the “aromatic” region consists of absorption resulting from the eight aromatic protons upon which is superimposed a singlet now assigned t o the C-17, C-17’ protons of structure LXXII. The ‘iolefinic” region contains the quartet due t o the C-19, C-19’ protons together with the superimposed sharp peak which can only be assigned to the C-2, C-2’ methine protons. Similar assignments were made for the spectrum of toxiferine I . The spectra thus remain in agreement with structures LXXII and incompatible with structures LXXI. Chemical evidence for structure LXXII(R= H) for C-dihydrotoxiferine I was provided as follows (93).

..

H

I1

0

LXXIV

The ethylidene side chains of nordihydrotoxiferine (tertiary base corresponding to LXXII ; R = H) can be hydrogenated over platinum in ethanol to give two stereoisomeric tetrahydro derivatives which were separated by chromatography and fully characterized. When one of these isomers was subjected to ozonolysis in methanol, it gave strychanone unequivocally known to have the structure LXXIV ( R = H ) ; strychanone is prepared (94) Erom dihydrodeoxystrychnine (see p. 633). The other tetrahydro isomer gives only a poor yield of impure strychanone by ozonolysis. A similar result was obtained on ozonolysis of the two isomeric nortetrahydrotoxiferines I ; one isomer gave 18-hydroxystrychanone (LXXIV; R = OH). These results clearly support the (2-16-17,C-16’-17’ position of the double bonds for C-dihydrotoxiferine I (LXXII; R = H) and for toxiferine I (LXXII; R = OH). C-Alkaloid H, picrate mp 189O-l92O, was isolated (16) from a calabash and its structure has been shown to be the the “hybrid” one of 18hydroxydihydrotoxiferine I (LXXV). This foliows from its preparation

546

A. R. BATTERSBY AND H. F. HODSON

by the mixed condensation of hemitoxiferine I (LXVII) with norhemidihydrotoxiferine I (tertiary base corresponding to LXVIII) in acetic acid. The chloroform-soluble ditertiary and tertiary-quaternary products were extracted and converted into a mixture of diquaternary chlorides. Separation of the products by partition chromatography then gave C-dihydrotoxiferine I and the salt (LXXV) identical with the natural C-alkaloid H (95). C.

CARACURINEV

Caracurine V has an indoline chromophore and is readily converted into nortoxiferine I by dilute mineral acids and by dilute acetic acid; its IR-spectrum shows the absence of hydroxyl groups but exhibits strong

LXXVII

LXXVI Caracurine V

BrHzC

I

LXXVIII

LXXIX

ether absorption. Accordingly, it has been assigned (85) the aminohemiacetal structure (LXXVI) which has the same stable sevenmembered ring that is present in the Wieland-Gumlich aldehyde ; further support for this structure comes from the partial synthetic work

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

547

shown subsequently. A second amino-hemiacetal structure can be written for caracurine V in which the side chain oxygens are linked to the 2 and 2' positions as shown in structure LXXVII. However, the NMR-spectrum of caracurine V (96) firmly establishes the presence of two >N-CH--0residues, thereby confirming structure LXXVI. I

Caracurine V can be reduced catalytically over rhodium (93) to give two stereoisomeric bases, both of which are tetrahydro derivatives formed by saturation of the 19,20,and 19', 20' double bonds. These bases can be N-methylated, whereupon the hemiacetal rings open to afford the isomeric tetrahydrotoxiferine and isotetrahydrotoxiferine (LXXII; R = OH with 19,20 and 19',20' double bonds saturated).

D. SYNTHETIC WORK It was noted earlier that condensation of the Wieland-Gumlich aldehyde N,-methochloride, i.e., hemitoxiferine I (LXVII), in hot acetic acid gives only a low yield of toxiferine I . Caracurine V di-N,methochloride (di-N,-methochloride of LXXVI) can also be isolated from the reaction mixture, but the major reaction product (31, 97) is 0,O-diacetyltoxiferine I dichloride (LXXII ; R = OAc). Condensation of hemitoxiferine I (LXVII) in pivalic acid, with which there is strong steric hindrance to acylation reactions, gives a reaction mixture from which a t least 70% of pure toxiferine I can be isolated by direct crystallization (31). If the condensation is run in acetic acid in the presence of sodium acetate, again only a low yield of toxiferine I is obtained. I n this case, the yield can be raised (97) t o over 60% by (a) treatment of the reaction product with p-toluene sulfonic acid in acetic acid, which converts caracurine V di-N,-methochloride into toxiferine I ;and (b)treatment of the reaction mixture with aqueous ammonia which hydrolyzes the diacetate (LXXII ; R = OAc) to yield toxiferine 1(LXXII ;R = OH). I n contrast to these findings, condensation of the Wieland-Gumlich aldehyde itself (LXV), in pivalic acid (31) or in acetic acid (86), gives mainly caracurine V (LXXVI); only traces of nortoxiferine I (tertiary base corresponding to LXXII ; R = OH) are formed. The synthesis of C-dihydrotoxiferine I has also been achieved. When the Wieland-Gumlich aldehyde (LXV) was treated with hydrogen bromide, acetic acid, and phosphorus, it yielded the 18-bromo derivative (LXXVIII; R = Br), which without isolation was debrominated with zinc and acetic acid. The resulting amorphous aldehyde (LXXVIII ; R = H or stereoisomer) was converted into its N,-metho salt which was

548

A. R.

BATTERSBY AND H. F. HODSON

warmed in acetate buffer to effect self-condensation. The resulting product was shown t o have color reactions, a UV-spectrum, R, values, an IR-spectrum, and a rotation identical with those of C-dihydrotoxiferine I . Its picrate, however, melted some 60’ higher than the picrate of the natural alkaloid, and the two picrates could not be interconverted. There can therefore be no doubt that the two compounds are different, and by elimination one is forced to the conclusion that they must be isomeric about the 19,20 and 19’,20’ double bonds of structure LXXII. The new product was designated C-dihydrotoxiferine I* (85). A second approach, however, gave the desired compound. Caracurins V (LXXVI),prepared from the Wieland-Gumlich aldehyde, reacts with hydrogen bromide by ring-opening and bromination to give the allylic bromide (LXXIX).Reduction of this bromide with zinc and acetic acid, followed by N,-methylation of the product, gives material in all respects identical with C-dihydrotoxiferine I (98). If, instead of quaternizing the N, nitrogen atom of the WielandGumlich aldehyde (LXV) with methyl iodide, other alkyl or alkenyl halides are used, then analogs of hemitoxiferine I (LXVII) may be obtained. The preparation from these substances of analogs of toxiferine I (LXXII; R = OH) has been studied by several pharmaceutical houses (99) and one such analog, in which the N,-methyl groups of toxiferine I are replaced by ally1 residues, is a valuable curarizing agent (100). Simple model compounds containing the diazacyclooctadiene ring system which forms the heart of the toxiferine molecule have also been prepared (101).

E . C-FLUOROCURARIBE This pale-yellow quaternary alkaloid, picrate mp 189’, [.ID of chloride (in water), was first isolated from a calabash curare (12) ; it was subsequently isolated from other calabash curare preparations (34, 35) and has been identified chromatographically in extracts from the bark of S. mitscherlichii ( 3 3 )and other Strychnos species (7). The identification of C-fluorocurarine is greatly helped by its deep-blue fluorescence in UV-light. Analyses of the crystalline iodide and anthraquinone sulfonate which showed its molecular formula to be C ~ ~ H Z ~ (Nl a~) , O + is in agreement with all subsequent work. The properties of C-fluorocurarine, particularly its low toxicity and high R, values, suggest that it is a Czo alkaloid rather than a “double” molecule. This was confirmed (102) by application of the partial quaternization method to one of its derivatives (see subsequent discussion) ; the method cannot be applied to C-fluorocurarine itself, since on pyrolysis it is not smoothly demethylated to the - 930”

15.

ALKALOIDS OF CALABASH CURARE ; Xtrychnos

549

corresponding tertiary base. However, after the structure of C-fluorocurarine had been elucidated, the tertiary base, aorfluorocurarine, was isolated (73) from Diplorrhynchus condylocarpon Muell. Arg. Although C-fluorocurarine is a CZOalkaloid, it is discussed in this section because of its important relationships with the C40 alkaloids of the C-dihydrotoxiferine I “family” (Section 111, F). Thus, it has been shown by Boekelheide’s group (18)and by T. Wieland’s group (103) that C-fluorocurarine is produced by the action of concentrated hydrochloric acid on C-curarine, and it is also obtained (104) by treatment of C-calebassine with the mixed anhydride of formic and acetic acid. C-Fluorocurarine has one N-methyl group attached at the quaternary N,) nitrogen atom and one C-methyl group in an ethylidene side chain; ozonolysis gives acetaldehyde but no formaldehyde. The N, nitrogen atom is secondary ; N,-acetylfluorocurarine can be prepared and is readily hydrolyzed to the parent alkaloid. The most striking feature of the alkaloid is its characteristic and hitherto unique UV-spectrum, with a long wavelength peak a t 358 mp, which undergoes a reversible bathochromic shift in the presence of 0.01 N alkali (102). Basification of a concentrated aqueous solution of the alkaloid chloride gives the yellow “ C-fluorocurarine base chloride ” as a crystalline solid ; its UV-spectrum is identical with that of C-fluorocurarine in alkali, and, when this “base chloride ” is treated with acid, C-fluorocurarine is regenerated (105). With dimethyl sulfate, the alkaloid gives the corresponding monoquaternary N,-methyl derivative, which has a UV-absorption identical with that of C-fluorocurarine itself but which does not undergo the bathochromic shift in alkali. This shift must therefore involve the ready removal of a proton from the N, nitrogen atom of C-fluorocurarine (102). The nature of the chromophore was deduced by the Zurich group (102) by a careful study of the borohydride reduction of N,-methylfluorocurarine. Reduction of a very dilute solution of this quaternary salt in phosphate buffer at pH 8, with an excess of sodium borohydride, gives a substance (LXXXIII) with a pure a-methyleneindoline chromophore ; in acid solution, this product (LXXXIII) is rapidly and irreversibly converted into a second substance with a typical indoline UV-spectrum. The nature of these products and their relationship will be considered subsequently. When the same reduction is attempted on a preparative scale, with a higher concentration of reactants, the pH cannot be kept below 9, and the product contains both a-methyleneindoline and indoline components. Acidification of the reaction mixture gives a mixture of substances, all of which have the UV-spectrum of an indoline ; the mixture can be separated by partition chromatography into three crystalline quaternary compounds (LXXXIV, LXXXV, and LXXXVII).

550

A. R. BATTERSBY AND H. F. HODSON

The substance LXXXV, C21H27NzO+, has a typical indoline UVspectrum which in 0.1 N alkali undergoes the reversible bathochromic shift characteristic of 2-hydroxyindolines ; in 5 N hydrochloric acid, the spectrum is reversibly changed to that of an a-vinylindoleninium cation (LXXXVI). The IR-spectrum shows absorption corresponding to the presence of hydroxyl and vinyl groups. Ozonolysis gives acetaldehyde from the ethylidene side chain and also formaldehyde. Further, i t can be shown by chromatographic studies that the hydroxyl group forms a

LXXXI

LXXX

Y\rt I Me

LXXXII

+I CHzOH

LXXXIII

Me

CHzOH

LXXXIV

J readily hydrolyzable methyl ether when treated with methanolic hydrogen chloride ; this suggests that the hydroxyl group forms part of a carbinolamine system. The UV-spectrum of the second compound (LXXXIV),CZ1HzgNzO+, has an indoline UV-spectrum which shows no shift in alkali but which in 1 N hydrochloric acid is characteristic of the indolinium cation. The IR-spectrum shows the presence of a hydroxyl but no vinyl group. I n contrast to LXXXV, which cannot be acetylated, the alcohol LXXXIV with acetic anhydride in pyridine gives a crystalline 0-acetyl derivative.

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

551

The third product (LXXXVII), deoxyisodihydro-N,-methylfluorocurarine, C Z ~ H ~ ~ also N ; , has the UV-spectrum of an indoline, and the spectrum is unchanged by alkali. 7 N Hydrochloric acid, however, changes the spectrum to that of the indoleninium ion. The IR-spectrum shows absorption characteristic of the vinyl group. Pyrolysis of the chloride salt (LXXXVII)yields the corresponding N, tertiary base with loss of methyl chloride and, by the method of partial quaternization, this base was shown to have only one strongly basic nitrogen atom, i.e., to be a CZOcompound. A consideration of the properties (102) of the three indolinic products discussed and their possible modes of formation leads to the partial structures LXXXIV, LXXXV, and LXXXVII as the only satisfactory representation. On this basis, the partial structure LXXX follows for N,-methylfluorocurarine and the partial structure LXXXI to represent the chromophore of C-fluorocurarine. The a-methyleneindoline formed by borohydride reduction of N,-methylfluorocurarine in dilute solution must then be the allylic alcohol LXXXIII. It is reasonable to assume that the reduction in alkaline solution gives a mixture of the products LXXXIII and LXXXIV ; when the reaction mixture is acidified, the alcohol LXXXIII presumably undergoes allylic rearrangement to yield the isomer LXXXV, some of which is further reduced under the acidic conditions to yield the indoline LXXXVII. This can be envisaged as proceeding by way of the indoleninium cation LXXXVI. The partial structure LXXXI for C-fluorocurarine is supported by the following facts : C-fluorocurarine forms an unstablg oxime ; the IRspectrum of the alkaloid is consistent with the presence of an a,/3unsaturated /3-aminoaldehyde group ; and, by oxidation with hydrogen peroxide, C-fluorocurarine gives a substance with a UV-absorption characteristic of N,-unsubstituted oxindoles. From a study of C-fluorocurarine, its N,-methyl and N,-acetyl derivatives, and model cornpounds, Wieland and associates ( 105) arrived a t similar conclusions, although the aldehydic nature of the carbonyl group was not recognized. However, Fritz (106)subsequently reported the synthesis of the aldehyde LXXXVIII by C-formylation of N-formylhexahydro-11-methylcarbazole (LXXXIX); in this way, the first simple compound to contain the chromophore LXXXI was prepared. Its UV-absorption, in neutral and in alkaline solution, is closely similar to that of C-fluorocurarine, thus confirming the foregoing assignment. The bathochromic shift in alkali must then be due to formation of the mesomeric anion (LXXXII). From these considerations and on biogenetic grounds, the structure XC was proposed (102) as a hypothetical one for C-fluorocurarine. This was confirmed (107) as follows. Hydrogenolysis of the allylic hydroxyl

552

A. R. BATTERSBY AND H. F. HODSON

group in the Wieland-Gumlich glycol (LXX; R = OH) over platinum in ethanol gave the corresponding deoxy compound (LXIX), which was converted by Oppenauer oxidation into norhemidihydrotoxiferine (tertiary base corresponding to LXVIII) ; this underwent autoxidation in air to yield norfluorocurarine, which, on N,-methylation, gave C-fluorocurarine identical in all respects with the natural alkaloid. Independent proof was obtained by reducing C-fluorocurarine with zinc in sulfuric acid to give hemidihydrotoxiferine (LXVIII) which, without isolation, was self-condensed in buffered aqueous acetic acid to yield (108) C-dihydrotoxiferine I. The latter compound had already been chemically related to the Wieland-Gumlich aldehyde.

xc

LXXXVIII

C-Fluorocurarine

XCI

XCII

XCIII

Structure XC allows ready interpretation of the deformylation (109) of C-fluorocurarine under the influence of hot acid t o yield the indolenine XCI. This reaction is strictly analogous to that which occurs when akuammicine is heated with acid. C-Fluorocurarine chloride is reduced (108) by sodium Forohydride in alkaline solution to the well-characterized tetrahydrofluorocurarine, CzoH27N20+ (XCII). The alcohol XCII, however, proved t o be a

15.

ALKALOIDS OF CALABASH CURARE ; Xtrychnos

553

diastereoisomer of the N,-metho salt (XCIII) of the alcohol (LXIX) obtained from the Wieland-Gumlich aldehyde as described previously ; comparison of these substances did not, therefore, yield any structural information.

F. THE

“FAMILIES”

OF

ALKALOIDS

The formation of C-alkaloid D from C-dihydrotoxiferine I by aerial oxidation in dilute aqueous mineral acid has already been mentioned. This is one of several important reactions which show that many of the calabash curare and Strychnos alkaloids can be grouped together in socalled “families ” containing mutually related alkaloids. Thus, Cdihydrotoxiferine I is converted into C-curarine by irradiation of the solid alkaloid in the presence of oxygen (110); similar irradiation of a solution of the alkaloid in the presence of eosin gives C-calebassine (110). C-Calebassine is also formed from C-dihydrotoxiferine I by the action of hot acetic acid in pyridine in the presence of oxygen (111). The production of C-fluorocurarine from C-calebassine and from C-curarine has already been noted. These reactions are summarized in Scheme 2 . C-Alkaloid D

T

HzO+ 0

2

Oa h,

C-Dihydrotoxiferine I 02,

pyridine, AcOH

11

C-Curarine

C-Calebassine

.1

conc. HCI

02 h, HCO. OAc

A

C-Fluorocurarine

SCHEME 2. Interrelationships in the C-dihydrotoxiferine I “family.”

A similar scheme can be written for alkaloids related to toxiferine I, the corresponding fundamental changes involved being strictly analogous (see Scheme 3). C-Alkaloid E in the toxiferine I rcfamily” corresponds to C-curarine in the C-dihydrotoxiferine I (‘family;” similarly, C-alkaloid A corresponds to C-calebassine, and caracurine I1 metkochloride to C-alkaloid D. The interconversions of the tertiary alkaloids related to toxiferine I have already been discussed. A comparison of these two reaction schemes reveals two, possibly related, points of interest. Hemidihydrotoxiferine I arid its nor compound have not been isolated from natural sources; neither has the analog of C-fluorocurarine in the toxiferine I series, which would have structure

554

A. R. BATTERSBY AND H. F. HODSON

XCIV or, less probably, that of the corresponding cyclic hemiacetal. The Wieland-Gumlich aldehyde and its N,-metho derivative (hemitoxiferine I, alkaloid A8 ; LXVII) are stabilized as the cyclic hemiacetals, and both occur naturally. I n hemidihydrotoxiferine I (LXVIII), there is no such

4

Caracurine I1 methochloride Ha0+02

(refs. 31, 113)

Hemitoxiferine I

0 2

Toxiferine I

h,

(Ref. 112)

C-Alkaloid E

or AcOH (Ref. 112)

C-Alkaloid-A SCHEME 3. Interrelationships in the toxiferine-I “family”.

stabilization, and one can understand a greater tendency towards selfcondensation to C-dihydrotoxiferine I or oxidation to C-fluorocurarine. Attempts t o prepare the aldehyde XCIV by oxidation of the WielandGumlich aldehyde have so far been unsuccessful ( 1 14, 115).

XCIV

The toxiferine-like “hybrid,” C-alkaloid H (LXXV), is converted into C-alkaloid F by aerial oxidation in the presence of pivalic acid and pyridine; irradiation of solid C-alkaloid H in the presence of oxygen gives a product identical with natural C-alkaloid G. The properties of these transformation products confirm that C-alkaloids F and Q stand in the C-alkaloid H “family,” as do C-calebassine and C-curarine in the C-dihydrotoxiferine I “family.” The tertiary alkaloid caracurine VI is the nor base corresponding to C-alkaloid H. Many of the foregoing transformations take place under relatively mild conditions of aerial oxidation which might well obtain during the largely unknown processing of plant extracts to give calabash curare preparations. Other transformations take place under equally mild conditions of acid catalysis, which are often attained during isolation of the alkaloids. These two factors must be, a t least to some extent, responsible for the large number of components that can be detected by

15.

ALKALOIDS OF CALABASH

CURARE; Strychnos

555

paper chromatography of calabash curares and bark extracts (see Section I). The material of this section illustrates the key position held by the Wieland-Gumlich aldehyde (LXV) in this group of alkaloids. Of the calabash curare and Strychnos alkaloids of known structure, no fewer than nineteen4 can be derived from the aldehyde LXV, its 18-deoxy derivative, and the corresponding N,-metho derivatives.

G. C-ALKALOID D, CARACURINE11, AND CARACURINEI1 DIMETHO CHLORIDE The quaternary C-alkaloid D was first isolated (16) from a sample of calabash curare and was later encountered (24) as one of theproducts formed when C-dihydrotoxiferine I is treated with dilute acid in the presence of oxygen. Because of its low mobility on paper chromatograms, it was considered to be a "dimeric alkaloid," and the formula ( 1 16) on the basis of analyses of the several C ~ O H ~ ~ N ~was O ;proposed ' crystalline salts afforded by this alkaloid. Despite its bisquaternary nature, C-alkaloid D has a low curarizing activity, and in this respect it is similar to caracurine V dimethochloride mentioned earlier. C-Alkaloid D chloride shows [.ID - 51" (in 1 : 1 aqueous acetone). caracurine 11, C38H38N402, [.ID - 232" (in CHC13), mp 248"-249", is a tertiary base present in the bark of Venezuelan S. toxifera (30), and it is interesting that the bark of apparently the same species collected in British Guiana yielded (14, 15)the corresponding bisquaternary alkaloid, caracurine I1 dimethochloride. This alkaloid has [.ID - 106" (in 1 :1 aqueous acetone). The formation of caracurine I1 from caracurine Va (nortoxiferine I ) under the influence of acid and oxygen was noted earlier (24), together with the similar conversion of toxiferine I into caracurine I1 dimethochloride (84, 31). It was clearly demonstrated (116, 113) that light is not required, whereas oxygen is essential for the three acidcatalyzed changes described ; they thus involve nonphotochemical oxidation reactions. When toxiferine I and nortoxiferine I are treated with acid in the strict absence of oxygen, high yields of the simple fission products, hemitoxiferine I and the Wieland-Gumlich aldehyde, respectively, are obtained (1 13, 116). 4 Caracurine VII, hemitoxiferine I, C-fluorocurarine,diaboline, toxiferine I, C-dihydrotoxiferine I, nordihydrotoxiferine (caracurine IX), caracurine V, C-alkaloid A, C-alkaloid E, caracurine I1 methochloride, caracurine 11, C-calebassine, C-curarineI, C-alkaloid D, C-alkaloid H, C-alkaloid F, C-alkaloid G, caracurine VI.

556

A. R . BATTERSBY AND H. F. HODSON

The chemistry of caracurine I1 and its dimetho derivative and of C-alkaloid D has been studied by Swiss (116) and English (113) groups of workers, with the final solution of the structures coming as the result of a joint effort (96). The UV-spectrum of C-alkaloid D is that of an indoline, and it undergoes a bathochromic shift in the presence of alkali. This points to the chromophore XCVI, or the equivalent 2-hydroxyindoline system, as that responsible for the UV-absorption, two such units being necessary

p] =

HO-C-

2

SCI’

XCVI

XCVII

XCVIII

to account for the intensity of absorption. Further support comes from the very ready formation of C-alkaloid D dimethyl ether (partial formula XCVIII) when the alkaloid is treated with dry acidic methanol. This product, C40H46N4(0Me)z+,shows an indoline UV-spectrum which is unchanged in the presence of alkali, as would be expected; aqueous acid regenerates C-alkaloid D. Moreover, the UV-spectrum of C-alkaloid D measured in 12 N hydrochloric acid is completely changed and corresponds t o an indoleninium chromophore or its equivalent (XCVII); again, the intensity of absorption corresponds to t w o such units being present in the molecule. This change is a reversible one, for dilution of the acidic solution allows the quantitative recovery of the alkaloid. Carbinolamine systems (e.g.,XCVI) are normally reduced by zinc and acid to the deoxy derivative, and under these conditions C-alkaloid D affords bisdeoxy-C-alkaloid D, C40H48Nif (partial formula XCV), the indoline UV-spectrum of which is changed to a pure indolinium (benzenoid) spectrum (chromophore XCV with N protonated) when measured in

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

557

12 N hydrochloric acid. Both N, nitrogen atoms are thus protonated under these conditions. Pyrolysis of this bisdeoxy derivative yields the corresponding tertiary base, norbisdeoxy-C-alkaloid D, C38H42N4, [a]= - 36" (in CHC13). No skeletal change occurs in this step, for quaternization of the tertiary base with methyl iodide yields bisdeoxy-C-alkaloid D iodide. The IRspectrum of norbisdeoxy-C-alkaloid D shows no bands corresponding to 'OH or >NH residues, and this base is thus a convenient substance for perhydrogenation studies. Under vigorous conditions, it absorbs eight moles of hydrogen, corresponding to the saturation of two aromatic rings and two ethylidene systems. The presence of the latter may be inferred from the production of acetaldehyde by ozonolysis of C-alkaloid D and from the presence of two ethylidene residues in C-dihydrotoxiferine I, from which C-alkaloid D may be prepared. Modified Kuhn-Roth oxidation of C-alkaloid D yields acetic acid only. It follows from this perhydrogenation result that norbisdeoxy-Calkaloid D, and so the parent alkaloid also, contain twelve rings, one more than C-dihydrotoxiferine I. The conversion of the latter into C-alkaloid D must therefore involve the disappearance of the two central double bonds of C-dihydrotoxiferine I (LXXII ; R = H), the appearance of the two hydroxyl groups shown to be present in carbinolamine systems, and the formation of a new carbon-carbon bond. It was clear a t this stage that C-alkaloid D resembles the isomeric C-calebassine (see Section 111, H) in many ways, but the two differ completely in their behavior toward hot strong acid. The former can be recovered unchanged after vigorous treatment, whereas the latter is converted under relatively mild conditions into isocalebassine. The conversion of toxiferine I (LXXII ; R = OH) into caracurine I1 dimethochloride is best carried out in weakly acidic aqueous solution with oxygen and a platinum catalyst. Under these conditions, a yield of 86% may be obtained, and the product can then be pyrolyzed to afford the tertiary base, caracurine 11.This is converted back into caracurine I1 dimethiodide by treatment with methyl iodide. A simpler route to caracurine-11, however, involves the treatment of caracurine V(LXXV1) with p-toluene sulfonic acid and oxygen in isobutyric acid. There can be little doubt that the first step in this process is the fission of the cyclic ether systems in caracurine V to afford nortoxiferine I (caracurine Va, tertiary base corresponding t o LXXII, R = OH) and that the oxidative changes then take place on this molecule. The indoline UV-spectrum of caracurine I1 is virtually unchanged in the presence of 1 N hydrochloric acid and corresponds to the presence of two indoline systems in the C38H38N402 molecule. Mass-spectrometric

558

A. R. BATTERSBY AND H. F. HODSON

determination of the molecular weight of caracurine I1 firmly established the formula given above. When measured in concentrated sulfuric acid, the UV-spectrum of caracurine I1 corresponds to the presence of two

residues, and this change is reversible since the alkaloid is recovered unchanged from the diluted acid. Caracurine I1 is thus quite stable to strong acid. Measurements made in 12 N hydrochloric acid show the presence of indoliniuin and indoleninium chromophores. No bands corresponding to -OH or >NH groups appear in the IR-spectrum of caracurine If, but strong absorption corresponding t o ether residues appears. Bearing in mind the structure of toxiferine I (LXXII; R = OH), the preceding results can best be accommodated by partial structure C for caracurine 11. On this basis, the ion present in concentrated sulfuric acid is CI, and the indolinium component of the spectra measured in weaker acids corresponds to protonation a t N, in the partial structure C.

“71’< \/-N

I

HO.HzC HA II

CHz

b,’<

I

I/

HC /I A\,’

CC\

HO.H2C I HC

CH

,A<

I/

I

/C\,’ 2

XCIX

CI

C

BrHzC HO.hH I I/ HC
/I

A \ .

CII

Support for the partial structure C comes from the reduction of caracurine I1 with zinc and concentrated sulfuric acid in methanol, which affords a diol (XCIX) showing strong hydroxyl absorption in the IR-spectrum. The basic strength of the N, nitrogen atoms in this product

15.

ALKALOIDS OF CALABASH CURARE ; Xtrychnos

559

is considerably higher than obtains in the starting material, as shown by the ready protonation of one of the two N, atoms in 1 N hydrochloric acid. I n concentrated sulfuric acid, both N, nitrogen atoms of the diol (XCIX) are considerably protonated. This is as would be expected, now that the base-weakening effect ( 1 17) of the carbinolamine ether link to the carbon, a to the N, nitrogen, has been removed. Hot acetic acid readily converted the diol XCIX into its 0,O-diacetate in keeping with the allylic nature of the hydroxyl groups. Hydrogenation of caracurine I1 under carefully controlled acidic conditions yields tetrahydrocaracurine 11, which contains two indoline residues (UV-spectrum). The side-chain double bonds (see C) are saturated in this derivative, and rigorous proof comes from the NMR-study (see p. 560). It follows from this and the established molecular formula that caracurine I1 contains fourteen rings, three more than toxiferine I (LXXII; R = OH). The cyclic carbinolamine systems account for two of these, and the third is owing to a new carbon-carbon bond. As with C-alkaloid D, the two central double bonds of toxiferine I and nortoxiferine I have been lost in the conversion into the caracurine I1 series. The relationship of caracurine I1 to C-alkaloid D was established by converting the former, with hydrogen bromide, into the allylic dibromide (CII). Reduction of this product with zinc and acid then gave norbisdeoxy-C-alkaloid D. Information derived from one series can thus be applied in the other.

CIII

CIV

Caracurine I1

C-AlkaloidD

The location of the new carbon-carbon bond in these alkaloids was determined by NMR-spectroscopy (96), and this led to the complete structures C I I I and CIV for caracurine-I1 and C-alkaloid D, respectively. Caracurine I1 dimethochloride is the di-N,-metho derivative of structure CIII. The NMR-spectrum of caracurine I1 showed, in addition to signals

560

A.

R. BATTERSBY AND H. F. HODSON

corresponding to the aromatic protons and to those a t positions 19 and 19', a very sharp singlet a t 5.147 equivalent to two protons. This important signal was assigned to the protons a t positions 17 and 17' on the basis of NMR-studies with simpler carbinolamine ethers, and because of the absence of a signal in the 5.07 region in the NMR-spectrum of caracurine I1 diol (partial structure XCIX). Caracurine V (LXXVI) also shows this signal (at 5.31 T),but in this case it appeared as a doublet (J = 2-3 clsec, depending on conditions), owing to spin-spin coupling of the protons a t positions 17 and 17' with those a t 16 and 16'. It thus follows that positions 16 and 16' are fully substituted in caracurine 11. Similar results were obtained from NMR-measurements on C-alkaloid D and on derivatives of both alkaloids. The interlocking evidence leaves no doubt that the new bond in these alkaloids joins positions 16 and 16'. When this conclusion is considered with the chemical evidence, it leads to structures C I I I and CIV for caracurine I1 and C-alkaloid D, respectively. The relative and absolute stereochemistry shown follows from that which is established for the parent alkaloids toxiferine I (LXXII; R = OH) and C-dihydrotoxiferine I (LXXII ; R = H) and from the fact that the new rings can only be closed in the illustrated way unless impossible strain is introduced. It is highly satisfying that the full structure and stereochemistry of caracurine I1 have been determined by McPhail and Sim (118), using X-ray measurements on caracurine I1 dimethiodide, and that this completely independent method also leads to structure CIII in every detail for this alkaloid.

H. C-CALEBASSINE,C-ALKALOIDA,

AND

C-ALKALOID F

C-Calebassine is a colorless alkaloid, picrate mp 210"-212", [.IU + 72" (chloride, in water), is one of the more common constituents of calabash curares, and is comparatively readily isolated as its sparingly soluble picrate. When it was first obtained (13), the name C-toxiferine I1 was used, and it must be stressed that this alkaloid is not identical with toxiferine I1 from S . toxifera (13, 14). Later, C-toxiferine I1 was shown to be identical with C-calebassine isolated by the Zurich group (32), and the latter name has become established. The name C-strychnotoxine I was also proposed ( 1 7 , 119) for this alkaloid, but was later abandoned (120). Early analyses suggested a molecular formula of CzoHz3Nb or C20HzsN20+ (13, 119). However, the high physiological activity and low Rfvalues of C-calebassine together with its relationship to C-dihydrotoxi-

15.

ALKALOIDS O F CALABASH CURARE ; 8tryChnOs

561

ferine I (see discussion to follow) indicate a C40 formulation and the molecular formula was tentatively reassigned (82) as C40H48- 50N40$+. Direct proof of the “dimeric” formulation could not be obtained by the partial quaternization method because C-calebassine chloride is not smoothly pyrolyzed to the corresponding tertiary base ; instead, extensive decomposition occurs. C-calebassine has an indoline UV-spectrum (maxima at 250 and 300 mp) (35). It contains two double bonds similarly situated in ethylidene groups as shown by the following results (120, 121). Hydrogenation in aqueous solution over a platinum catalyst gives tetrahydrocalebassine with a UV-spectrum and color reactions identical with those of the parent alkaloid; the olefinic residues are therefore isolated from the chromophore. Ozonolysis of C-calebassine, oxidation with performic acid, and oxidation with osmium tetroxide and sodium chlorate, followed in the last two cases by periodate fission, all gave high yields of acetaldehyde by cleavage of the ethylidene residues. Modified Kuhn-Roth oxidation of the alkaloid gives acetic acid only. Similar oxidation of tetrahydrocalebassine, which gives no acetaldehyde on ozonolysis, yields a mixture of propionic and acetic acids, indicating the presence of a C-ethyl group in the reduction product. Under the influence of strong mineral acids, C-calebassine is converted into a yellow product which is probably isomeric; this substance is named isocalebassine and is readily isolated as a highly crystalline acid adduct ; e.g., with hydriodic acid, isocalebassine diiodide hydriodide C40H48-50N40212. HI is formed (121, 122). Tetrahydrocalebassine behaves in an analogous manner to give isotetrahydrocalebassine diiodide hydriodide (121, 122). Although C-calebassine and tetrahydrocalebassine are resistant to hydrogenation over a platinum catalyst in alkaline solution, isocalebassine and isotetrahydrocalebassine readily undergo simple Emde fission under these conditions. The Emde base from isotetrahydrocalebassine gives acetic acid, propionic acid, and a-methylbutyric acid on modified Kuhn-Roth oxidation ; since isocalebassine gives acetaldehyde on ozonolysis, this Emde base has partial structure CX and isotetrahydrocalebassine partial structure CIX. It was mentioned before that C-calebassine also gives acetaldehyde on ozonolysis, and it follows that the ethylidene systems are not changed in the production of the is0 compound; the same holds true for the equivalent parts of the molecule in the tetrahydro series. I n some way, at present not understood, Emde reduction is facilitated in the is0 compounds. The partial formulas CV, CVI, and CVII are assigned (121) to C-calebassine, isocalebassine, and tetrahydrocalebassine, respectively ; the Emde base from isocalebassine has partial structure CVIII.

562

A . R. BATTERSBY AND H. F. HODSON

C-Calebassine contains two hydroxyl groups ( 123) present as carbinolamine residues \ I ,N-%OH I

Reductive removal of the hydroxyl groups with zinc and acetic acid (123, 124) gives bisdeoxycalebassine

which may be reconverted into C-calebassine by photooxidation (1 12). As would be expected, C-calebassine readily forms a dimethyl ether by treatment with dry methanol in the presence of a trace of acid; this ether formation is readily reversed in dilute aqueous acid a t room temperature. Tetrahydrocalebassine forms an analogous dimethyl ether, and the changes involved ( 125) are :

A certain amount of confusion was caused in the early work on C-calebassine because this ether formation and hydrolysis under such mild conditions was considered to be some type of isomerization reaction. I n this respect, it is worth noting that C-calebassine dimethyl ether, or “C-calebassine 0. Rv ” (125) (ohne Rotverschiebung, see next paragraph) is identical with the C-calebassine in Ref. 122, the C-toxiferine I1 in Ref. 17, and the C-strychnotoxine I a in Ref. 119; it is probably identical also with the C-curarine I1 in Ref. 11. C-Calebassine itself, or “ C-calebassine m. Rv.” ( 125) (mit Rotverschiebung), is identical with the C-calebassine A in Ref. 122, the C-toxiferine I1 in Ref. 13, and the C-strychnotoxine I in Ref. 119. The indoline UV-spectra of C-calebassine and tetrahydrocalebassine exhibit a bathochromic shift of some 10 mp in alkaline solution (35, 125). This shift is not shown by the corresponding dimethyl ethers and is known to be characteristic of the 2-hydroxyindoline chromophore (47) (CXI) or, of course, its equivalent (CXII).The alkali-induced shift must be owing to the generation of the anion CXIII, or the equivalent anion from CXII, by the removal of the carbinolamine proton. This spectroscopic shift is to be compared with the hypsochromic shift (126) caused by protonation or quaternization of N, in the system CXIV which

15.

ALKALOIDS OF CALABASH CURARE ; Xtrychnos

563

generates the cation CXV. The chromophore of C-calebassine is thus CXI or CXII. Further support for the chromophore CXI or CXII comes from the demonstration that in strong mineral acids the indoline UV-spectrum of C-calebassine changes to that of the indoleninium cation (CXVI) or its equivalent (CXVII) (127). When the hydroxyl groups of C-calebassine

I

are removed to form the bisdeoxy derivative, the basicity of the N, nitrogen atoms increases so that the spectrum of bisdeoxycalebassine in 10 N hydrochloric acid is benzenoid. Complete protonation of both N, nitrogen atoms has thus occurred. With methyl iodide a t 70°, bisdeoxycalebassine gives a N,,N,,N,-trimethyl derivative, again no doubt because of the increased basic strength of N, in the bisdeoxy series. The isolation of this derivative, containing three N-methyl groups in a (341-molecule, gave the first direct proof that C-calebassine is a C40 alkaloid (124). The failure to form aN,,N,,N,,N,-tetramethyl derivative can be rationalized in terms of the field effect of the one quaternary N, atom preventing quaternization a t the other closely situated N, atom ; this is in contrast to the relatively large distance between the N, atoms. All the foregoing evidence must now be considered in the light of the relationship between C-calebassine and C-dihydrotoxiferine I . Analyses

564

A. R. BATTERSBY AND H. F. HODSON

of C-calebassine and its derivatives allow no clear distinction between the formulas C40H48N40:+ and C~OH~ON~O:+. C-Calebassine is formed from C-dihydrotoxiferine I, C40H46Nif, by photooxidation and by the action of pyridine and acetic acid in the presence of atmospheric oxygen (110, 111). Therefore, the formula C40H480;+ is preferred, since the H50 formulation corresponds to hydration rather than oxidation of C-dihydrotoxiferine I. Pyrolysis of bisdeoxycalebassine dichloride gives the

CXIV

oq CXVI

cxv

/\

CXVII

corresponding nor base, C38H42N4, by the loss of two moles of methyl chloride. This base undergoes perhydrogenation over platinum in dilute sulfuric acid, with the uptake of eight moles of hydrogen corresponding to the saturation of two benzene rings and two ethylidene residues. Under the same conditions, C-calebassine diperchlorate consumes ten moles of hydrogen, the extra two moles being used for reductive removal of the hydroxyl groups. This knowledge, when considered with the molecular formula C40H48N40;+, allows the deduction that C-calebassine contains twelve rings. One more ring has thus been formed in the conversion of C-toxiferine I into C-calebassine and corresponds to one extra carboncarbon bond, presumably in the central ring system (127). The central portion of the C-dihydrotoxiferine I molecule is shown in partial formula. CXVIII. On the basis of the foregoing information, one can state that the corresponding portion of the C-calebassine molecule is saturated, carries two hydroxyl groups cc to the N, nitrogen atoms, and contains one additional carbon-carbon bond. Models show that the Cz,C2’ position for this bond is sterically impossible. It is further found

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

565

that bisdeoxycalebassine is not identical (116) with the isomeric bisdeoxyalkaloid D which is known (Section 111, G) to have the partial structure CXIX. Therefore, the extra bond must be in the C17,C17' position, giving partial formulas CXX for bisdeoxycalebassine and the two possibilities, CXXI and CXXII, for C-calebassine itself. An unsymmetrical disposition of the extra bond can be ruled out, since the NMR-spectra of C-calebassine and its derivatives have all the signals derived from two identical protons or groups of protons (127).

CXVIII

cxx

CXIX

H

HO

H

H

CXXI

CXXII

Analysis of the NMR-spectrum of C-calebassine (127) gives support for the partial structures CXXI or CXXII. In addition to signals which can be assigned unambiguously to the periphery of the molecule, there is a two-proton singlet a t 291 cjsec corresponding to the C-2 and C-2' protons in CXXII or to the C-17 and C-17'protons in CXXI. I n addition, a two-proton singlet which appears at 179 cjsec is assigned to the c-16 and C-16' protons. Treatment of C-calebassine with DCI or DzS04 in deuterium oxide followed by isolation through the picrate to the chloride gives a dideuterocalebassine in which the signal at 179 cjsec is absent ; this change can be reversed by treatment with concentrated hydrochloric acid followed by water. This is owing to deuterium exchange a t the C-16 and C-16' positions, that is, at the /3-position of the potential enamine systems in CXXI or CXXII, and thus confirms the assignment of the 179 c/sec signal. It remains, therefore, to distinguish between the possible structures CXXIII (R1=Rz = H) and CXXIV for C-calebassine. The following

566

A . R . BATTERSBY AND H. F. HODSON

evidence establishes the former, with the hydroxyl group in the 2 and 2' positions (127). 1. I n 10 N hydrochloric acid, the UV-spectrum of calebassine is of the diindoleninium type to be expected of the partial structure CXXV derived from CXXI. Structure CXXII would be expected to give the bridged, and no doubt very strained, dianhydronium structure CXXVI. 2. More direct evidence comes from the NMR-spectrum of C-calebassine; the assignments of the signals were confirmed by NMR-studies of

CXXIII C-Calebassino; R1 = Rz = H C-Alkaloid A ; R1 = Rz = OH C-Alkaloid F; R1 = H, Rz = O H

cxxv

CXXIV

CXXVI

bisdeoxycalebassine and deutero derivatives of this substance. Dreiding models of structures CXXIII (R1 = R2 = H) and CXXIV indicate the stereochemistry shown ; only cis coupling of rings C/F, C'/F', and F/F' is allowed. If structure CXXIV were correct, there would undoubtedly be spin-spin coupling between the cis protons a t C-2 and C-16 and C-2' and C-16'. I n fact, the 291 and 179 clsec signals, which on the basis of structure CXXIV would have to be assigned respectively to the protons a t positions 2,2' and positions 16,16', show no splitting. It is important that the models show that the dihedral angle between the trans protons a t (2-16 and C-15 and C-16' and C-15' is such that no spin-spin coupling

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

567

would be expected. Structure CXXIII (R1=Rz = H), however, is fully consistent with these results, for there are no neighboring protons cis to the C-16 and C-16' protons ; again, the absence of coupling between the trans-situated C-16 and C-17 and the C-16' and C-17' protons has previous analogy in the very small or nonexistent spin-spin interaction between trans-situated neighboring protons in pyrrolidine and tetrahydrofuran systems. The combined evidence proves that C-calebassine has structure CXXIII (R1= R:!= H). What is now known as C-alkaloid A was first isolated by King (14) from S. toxifera bark and named toxiferine IV ;when it was again isolated from a calabash (l6), its identity (15) with King's alkaloid was not known, and the second name has become established. C-Alkaloid A, picrate, mp 228"-229", ["ID +64O (chloride in water), is formed from toxiferine I under the same conditions that produce C-calebassine from C-dihydrotoxiferine I . These include aerial oxidation in a mixture of pyridine, water, and isobutyric acid (127) or pivalic acid (114);when the acid used is acetic acid, C-alkaloid A is produced along with its 0 , O diacetyl derivative (114) as described earlier in the case of toxiferine I. Most of the properties of C-alkaloid A and its UV-spectrum in neutral, alkaline, and strongly acidic solution are identical with those of C-calebassine. Also, its NMR-spectrum corresponds with that of C-calebassine, with the exception of those signals owing to the hydroxylated side chains. C-Alkaloid A therefore has structure CXXIII (R1= R2 = OH) ; C-alkaloid F (see Section 111, F) is CXXIII (R1=H,

Rz = OH). I. C-CURARINE, C-ALKALOID E,

AND

C-ALKALOID G

C-Curarine, C ~ O H ~ ~ N ~ yields O + +a ,picrate of mp 306°-3070, and, like C-calebassine, is one of the more frequent and abundant constituents of calabash curare. It was the first crystalline calabash curare alkaloid to be isolated (10). As mentioned earlier, C-curarine is formed from C-dihydrotoxiferine I by photooxidation and is converted into C-fluorocurarine by the action of concentrated hydrochloric or hydrobromic acid. A considerable amount of evidence relating to the structure of C-curarine was obtained before its relationships to C-dihydrotoxiferine I and C-fluorocurarine were established. Pyrolysis of the alkaloid dichloride a t 300" gives the ditertiary base norcurarine, C S ~ H ~ ;~ N ~ O norcurarine can be methylated with methyl iodide to regenerate Ccurarine diiodide (32). Early information concerning the skeleton of

568

A. R . BATTERSBY AND H. F. HODSON

C-curarine came from zinc dust distillation of norcurarine, which yields 3-methylindole, 3 ethylindole, 3-ethylpyridine, carbazole, and methylcarbazole (128). C-Curarine gives acetaldehyde on ozonolysis or on cleavage with the osmium tetroxide-periodate reagent ( 129) and contains two quaternary allylamine systems (CXXVII). On heating with aqueous alkali, the dichloride undergoes Hofmann elimination to give the so-called “ditertiary ether base ” ( 1 1 , 1 2 , 1 3 0 ) ,C40H42N40 (partialformula CXXVIII) containing two diene residues. This base has four terminal methylene groups (Doeuvre estimation) and, relative to C-curarine and derivatives of this alkaloid which lack diene systems, shows additional strong UV-absorption at about 226 mp (c = 22,000 to 29,000, depending on the comparison made). This absorption corresponds to the diene chromophore and has an intensity corresponding to two diene units in the C40 molecule. The presence of diene systems is further supported by the formation by the “ditertiary ether-base” of an adduct with maleic anhydride. The “ether-base” is clearly formed as the result of a vinylogous Hofmann degradation as illustrated (CXXVII+CXXVIII). Reduction of the base (CXXVIII) with sodium in amyl alcohol gives a tetrahydro derivative, C40H46N40, which can be reduced catalytically to an octahydro derivative, C40H~oN40, also obtained directly by catalytic reduction of the base (CXXVIII). The reduction of C-curarine chloride with sodium in amyl alcohol gives a base C40H46N40, as the result of an Emde degradation, which is identical with the above tetrahydro derivative. This tetrahydro base appears to be a mixture of the isomers CXXIXa and CXXIXb. The octahydro base which has partial formula CXXX contains four C-methyl groups and, on modified KuhnRoth oxidation, gives m-methylbutyric acid in addition to acetic and propionic acids ; this was the first application in the alkaloid field of a technique which has since been widely used in structural determinations. All the IR-spectra and the results of ozonolysis and oxidation experiments on the foregoing compounds fully support the partial structures CXXVIII to CXXX (130). On catalytic hydrogenation in the presence of an excess of alkali, C-curarine chloride gives a homogeneous crystalline Emde base which, in contrast to the foregoing product from the reduction by sodium-amyl alcohol, is clearly the terminal methylene derivative CXXIXa and not a mixture of double-bond isomers (129). Emde reduction with allylic shift has thus occurred. A monoquaternary base, C40H43N40+, is produced when an equimolecular mixture of C-curarine chloride and C-curarine hydroxide is heated in ethanol. This base, descurarine, corresponds to a single vinylogous Hofmann degradation of C-curarine and has partial structure

15. ALKALOIDS OF

CALABASH CURARE ; Strychnos

569

CXXXI. Its properties are those to be expected of such a structure ; for example, the UV-spectrum shows absorption at 2 2 4 mp (c = 14000) corresponding to one diene chromophore in the C4o-molecule, and oxidation by osmium tetroxide-sodium chlorate followed by periodate fission gives both acetaldehyde and formaldehyde. The Hofmann degradation can be completed by treatment of descurarine with alkali to yield the “ditertiary ether base” CXXVIII (129).

c CXXVIII

CXXVII

_------_I

CHa-C=CH-CHa

Me c

CXXIXa

CXXIXb

/---. I

Me

cxxx

CXXXII

CXXXI

CXXXIII

The oxygen atom in C-curarine is present as an ether linkage ; thus, the IR-spectra of the alkaloid and the derivatives mentioned previously all have strong ether absorption in the region 1200-1170 cm-1 and no absorption corresponding to hydroxyl groups (13 1, 132). A most important clue to the structure of C-curarine comes from its degradation to the Czo-alkaloid, C-fluorocurarine (XC), by the action of

570

A. R. BATTERSBY AND H. F. HODSON

concentrated hydrochloric or hydrobromic acid. This reaction, which takes place in the absence of oxygen, has been shown by a very neat method (131) to be a simple hydrolysis according to the equation:

Partial quaternization (see Section 111, A) of norcurarine gives N b monomethonorcurarine which, by reaction with C14-methyl iodide followed by ion exchange, gives C-curarine chloride in which one of the N,,-methyl groups is labeled with C14. Degradation of this radioactive C-curarine chloride with hot, concentrated hydrobromic acid then gives a 7% yield of C-fluorocurarine which has exactly half the molar activity of the starting material. This shows clearly that the molecule of Ccurarine is composed of two identical halves, each with the skeleton of C-fluorocurarine. The NMR-spectrum of C-curarine further supports the identity of the two halves ; as with C-calebassine, it is evident that all the signals are derived from either two identical protons or two identical proton groups (132).The skeleton of C-curarine must thus be based upon that of C-dihydrotoxiferine I, and this is in keeping with the formation of C-curarine from the latter alkaloid under comparatively mild conditions. It is of interest that vigorous acid treatment of descurarine (partial structure CXXXI) also gives a small yield of C-fluorocurarine (129). A t one stage, unsymmetrical formulas (133) were considered for C-curarine because of the reported formation of hemidihydrotoxiferine I in the vigorous hydrolysis of C-curarine. However, a careful study of the hydrolysis showed (134) that none of this supposed fission product is, in fact, formed. The structure of the central ring system of C-curarine must now be considered, bearing in mind the close relationship of this alkaloid t o C-dihydrotoxiferine I . The UV-spectrum is unfortunately no help in this respect, as it is unlike that of any alkaloid so far discussed and unlike any simple indole derivative. The empirical formula, C40H44N40++,requires that the central ring system contain two double bonds but, although the analytical results do not exclude C ~ O H ~ ~ N ~the O +established +, molecular weight of norcurarine [mass spectrum (132)] leaves no doubt about the H44 formulation for C-curarine. The following evidence from the hydrogenation of norcurarine also indicates that the central ring system does indeed contain two double bonds, together with the ether linkage ; structures CXXXII and CXXXIII can thus be considered for this part of the C-curarine molecule (132). Hydrogenation of norcurarine over a platinum catalyst in acetic acid gives a crystalline tetrahydronorcurarine, C38H42N40,in which only the ethylidene double bonds have

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

57 1

been reduced; this reduction product has the same UV-spectrum and color reactions as the parent base but gives no acetaldehyde on ozonolysis. Hydrogenation in the presence of a large excess of catalyst results in a greater uptake of hydrogen and the formation of a mixture which can be separated by partition chromatography into hexahydronorcurarine, C38H44N40,and octahydronorcurarine, C38H46N40.It has been shown that no Emde degradation occurs during the formation of either of these bases and that they are produced by reduction of the ethylidene double bonds, together with reduction of one and two double bonds, respectively, of the central ring system. The UV-absorption of hexahydronorcurarine is a " mixed " spectrum corresponding to the summation of a N-vinylindoline and an indoline chromophore ; the spectrum of octahydronorcurarine is typical of N,-substituted indolines. Evidence (132) in favor of partial structure CXXXIII for C-curarine comes from the ozonolysis of tetrahydronorcurarine, in which only the ethylidene double bonds have been reduced. The UV-spectrum of this tetrahydro product is as that of C-curarine itself, and thus the central part of the molecule (CXXXII or CXXXIII) has been unaffected in the conversion into the tetrahydro derivative. After treatment of the ozonolysis products with active zinc, acid hydrolysis, and finally reduction with sulfur dioxide, the reaction mixture yields strychanone (LXXIV), indicating the 16,17 and 16',17' positions for the central double bonds of C-curarine as in CXXXIII. It is pointed out, however, that this evidence is not conclusive, and that the illustrated mechanism can be written which would account for the formation of 1 molecule of strychanone from a compound with partial structure CXXXII (132).

1) Reduction

0 1

(CXXXII)--+

I

Strychenone

I

A careful study of the NMR-spectra of C-curarine and norcurarine shows that they can best be interpreted in terms of partial structure CXXXIII, and the combined chemical and spectroscopic evidence is convincingly in favor of this system (132). Structure CXXXIV is thus proposed for C-curarine. Hexahydronorcurarine is then CXXXV, which accounts for its UV-spectrum described earlier. I n C-curarine, the central ring system is too rigid to allow full orbital overlap within the

572

A. R . BATTERSBY AND H. F. HODSON

two enamine systems ; the UV-spectrum of C-curarine is therefore not that of N-vinylindoline systems but is characteristic of the system CXXXIII alone. I n hexahydronorcurarine (CXXXV), with one central double bond saturated, the central ring system is less rigid, and full overlap within the one enamine system can occur ; the absorption is thus a combination of indoline and N-vinylindoline contributions.

cxxxv

CXXXIV C -Curmine

CXXXVI Violet form I

>?’Y HC CH OI / IP \

1

@T/\/C1I H HC .A& ps? I \

CXXXVII Yellow form

Finally must be mentioned the interesting halochrome reaction of C-curarine ; similar color reactions are given by those derivatives of C-curarine which retain the chromophore of the parent molecule. This reaction has been studied most closely with tetrahydronorcurarine. I n 40-807, aqueous sulfuric acid, the base dissolves to give a deep-violet solution. I n concentrated sulfuric acid, the solution is yellow but, on dilution with water to 70%, the color becomes deep-violet and finally disappears on further dilution. Tetrahydronorcurarine can be recovered unchanged from these diluted acid solutions. A plausible explanation of

15. ALKALOIDS OF CALABASH CURARE ; Strychnos

573

these color changes is that in 70% sulfuric acid reversible protonation of the chromophore could occur to generate the mesomeric cyanine system CXXXVI. I n concentrated sulfuric acid, a reversible ionization step could generate the 1,5-diazacyclooctatetraenesystem (CXXXVII), and this could be responsible for the yellow color in the strongly acidic medium. The chemistry of C-curarine has been studied probably more extensively than any other calabash curare alkaloid, and a large number of transformation products, many of unknown structure, have been obtained, e.g., the ultracurines A and B (135). Several further substances have been obtained (132) from some of the reduction products described earlier in the chapter. To avoid confusion, these have not been described in the foregoing account, which includes all the reactions and spectroscopic studies required to establish the structure of C-curarine.

CXXXVIII C-Alkaloid E; R1 = Rz = O H C-Alkaloid G; R1 = H; R t = O H

C-Alkaloid-E, C ~ O H ~ ~ N ~ was O ; +isolated , (16, 35) from calabash curare and characterized as the crystalline picrate, mp 272". Its UVspectrum (35)is virtually identical with that of C-curarine, so that the unusual chromophore established previously for the latter alkaloid is present also in C-alkaloid E. Since C-alkaloid E is formed from toxiferine I, just as C-curarine is produced from C-dihydrotoxiferine I, there can be no doubt that C-alkaloid E has the structure CXXXVIII (R1 = Rz = OH). This holds true also for C-alkaloid G, which is the C-curarine analog of C-alkaloid H. C-Alkaloid G thus confidently can be assigned the st'ructure CXXXVIII (R1 = H ; Rz = OH). This alkaloid, also from a calabash (16, 3 5 ) , affords a crystalline picrate, mp 285"-286O and its UV-spectrum is superimposable on that of C-curarine.

APPENDIX ALKALOIDS OF UNKNOWN STRUCTURE FROM CALABASH CURARE AND Strychnos SPECIES

Provisional formula

Alkaloid ~

~~

L-+

R, value (solvent “C”) ~~

Picrate mp (“C)

Color with ceric sulfate Immediate/After 20 min

Isolation reference

@ W

e$

~

M

Xanthocurine Fedamazine Melinonine E

CzoHziNzO+ CzoHziNzO+ C2 oHz3-2sNzO

Melinonine M C-Fluorocurinine Kryptocurine C-Alkaloid J Melinonine L Melinonine K Toxiferine 111 Toxiferine VIII Caracurine VIII methochloride Caracurine I X methochloride Alkaloid 1 Alkaloid M

-

1.9 +

CziHzgNzOz+ CzoHz7-zsNzOz CisHziNz CzzHz6Nz04

-

2.23 2.95 1.04

-

Over 275 233-235 120.5-122 and 2 16-2 19 242-245 213 -

Over 320 -

196-199

Blue-green Stable blue Nil

36 30

Nil Carmine/Wine-red Nil Red orange Weak blue Yellow

Purple-red/Brown

40 16,35 136 35 40 40 14 14, 15 39

40

-

-

-

-

1.12

Over 300 Dec. over 240 Dec. over 260

-

-

0.63 1.40

Red-violet/Brown

39

CzoHziNz+ (C~oHz3NzOz+)~

-

-

Deep-blue Nil/Yellow

17 23

CzoHwNzO+ (CzzHzsNz03+),

1.45

@ rn

m

P

zU

x

r

E 8

C-Alkaloid UB C-Alkaloid 0 C-Alkaloid P C-Alkaloid B

-

C-Alkaloid C

0.34

Caracurine I Caracurine I11 C-Alkaloid R C-Alkaloid S

0.7 0.8 0.68 0.51 as methochloride

C-Isodihydrotoxiferie I Toxiferine XI1 C-Guaianine C-Alkaloid I Caracurine IV Alkaloid 2 C-Alkaloid Q C-Alkaloid L Melinonine I

-

Melinonine H

-

Croceocurine Macrophylline A C-AlkaloidBL

1.6 0.75

3.95

0.34

238-240 237-238 224-232 Dec. over 270 Dec. over 270

Carmine/Brown Nil Blue Red-violetlBrown

-

Purple-red/Violet Purple-red/Brown-red Violet Red-violet/Colorless

-

250

2.3 1.12 0.89 1.o

242 Over 233 Over 305 194 Over 300

-

Blue-violet Blue -violet/Carmine Violet/Orange-brown Red-violet Nil Red Yellow-brown Nil

40

-

-

> 300

Violet Deep-blue/Blue-green

136 44 137

-

-

-

2.50

171 160-1 70 dec. 290-292 dec.

172-178

CI

01

Red-violet/Brown

13 14, 15 36 35 30 17 34 35 40

-

41, 16 36 36

-

a E ".

3

%

$T 2

2 ui

576

A. R. BATTERSBY AND H. F. HODSON

REFERENCES 1. A. R. McIntyre, “Curare, I t s Natural History and Clinical Use.” Univ. Chicago

Press, Chicago, Illinois, 1947. 2. D. Bovet, F. Bovet-Nitti, and G. B. Marini-Bettolo (eds.), “Curare and Curarelike Agents.” Elsevier, Amsterdam, 1959. 3. H. King, J. Chem. SOC. p. 1381 (1935). 4. 0. Wintersteiner and J. D. Dutcher, Science 97, 467 (1943). 5. M. Kulka, in “The Alkaloids” (R. H. F. Manske and H. L. Holmes, ecls.), Vol. IV, p. 199. Academic Press, New York, 1954. 6. D. F. Marsh, Ann. N . Y . Acad.Sci. 54, 307 (1951); P. Karrer,J. Phurm. Pharmacol. 8, 161 (1956). 7. G. B. Marini-Bettolo and G. C. Casinovi, J. Chromatog. 1, 411 (1958). 8. R. Boehm, Arch. Pharm. 235, 660 (1897). 9. H. King, Nature 135, 496 (1935). 10. H. Wieland, W. Konz, and R. Sonderhoff, Ann. 527, 160 (1937). 11. H. Wieland and H. J. Pistor, Ann. 536, 68 (1938). 12. H. Wieland, H. J. Pistor, and K. Bahr, Ann. 547, 140 (1941). 13. H. Wieland, K. Bahr, and B. Witkop, Ann. 547, 156 (1941). 14. H. King, J. Chem.Soc. p. 3263 (1949). 15. A. R. Battersby, R. Binks. H. F. Hodson, and D. A. Yeowell, J . Chem. Soc. p . 1848 (1960). 16. H. Schmid, J. Kebrle, and P. Karrer, Helv. Chim. Acta 35, 1846 (1952). 17. T. Wieland and H. Merz, Ber. 85, 731 (1952). 18. A. Zurcher, 0. Ceder, and V. Boekelheide, J . Am. Chem. Soc. 80, 1500 (1958). 19. H. Schmid and P. Karrer, Helv. Chim. Acta 33, 512 (1950). 20. G. C. Casinovi, Gazz. Chim. Ital. 87, 1174 (1957). 21. A. R. Battersby and R. Binks, Unpublished observations (1957). 22. G. B. Marini-Bettolo and M. A. Iorio, Gazz. Chim. Ital. 86, 1305 (1956); G. C. Casinovi, Gazz. Chim. Ital. 87, 1457 (1957). 23. H. Asmis, E. Bachli, E. Giesbrecht, J. Kebrle, H. Schmid, and P. Karrer, Helv. Chim. Acta 37, 1968 (1954). 24. H. Asmis, E. BBchii, H. Schmid, and P. Karrer, Helw. Chim. Acta 37, 1993 (1954). 25. G. B. Marini-Bettolo, M. A. Iorio, A. Pimenta, A. Ducke, and D. Bovet, Gazz. Chim. Ital. 84, 1161 (1954). 26. E. Schlittler and J. Hohl, Helv. Chim. Acta 35, 29 (1952). 27. W. Arnold, W. von Philipsborn, H. Schmid, and P. Karrer, H e h . Chim. Acta 40, 705 (1957). 28. W. Arnold, F. Berlage, K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 1505 (1958). 29. A. R. Battersby and D. A. Yeowell, J . Chem. SOC. (1964). In press. 30. H. Asmis, H. Schmid, and P. Karrer, Helv. Chim. Acta 37, 1983 (1954). 31. A. R.Battersby and H. F. Hodson, J . Chem. Soc. p. 736 (1960). 32. P. Karrer and H. Schmid, Helw. Chim. Acta 29, 1853 (1946). 33. J. Kebrle, H. Schmid, P. Waser, and P. Karrer, Helw. Chim. Acta 36, 345 (1953). 34. H. Meyer, H. Schmid, and P. Karrer, H e h . Chim. Acta 39, 1208 (1956). 35. J. Kebrle, H. Schmid, P. Waser, and P. Karrer, Helv. Chim. Acta 36, 102 (1953). 36. E. Giesbrecht, H. Meyer, E. Bachli, H. Schmid, and P. Karrer, Helv. Chim. Acta 37, 1974 (1954). 37. H. Schmid and P. Karrer, Helw. Chim. Acta 30, 1162 (1947).

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

577

38. G. B. Marini-Bettolo, M. Lederer, M. A. Iorio, and A. Pimenta, Gazz. Chim. Ital. 84, 1155 (1954). 39. H. Asmis, P. Waser, H. Schmid, and P. Karrer, Helv. Chim. Acta 38, 1661 (1955). 40. E. Biichli, C. Vamvacas, H. Schmid, and P. Karrer, Helv. Chim. Acta 40, 1167 (1957). 41. H. Schmid and P. Karrer, Helw. Chim. Acta 30, 2081 (1947). 42. H. Bickel, E. Giesbrecht, J. Kebrle, H. Schmid, and P. Karrer, Helv. Chzm. Acta 37, 553 (1954). 43. G. C. Casinovi, M. Lederer, and G. B. Marini-Bettolo, Gazz. Chim. Ital. 86,342 (1956). 44. M. A. Iorio, 0. Corvillon, H. Magalhaes Alves, and G. B. Marini-Bettolo, Gazz. Chim. Ital. 86, 923 (1956). 45. A. Penna, M. A. Iorio, S. Chiavarelli, and G. B. Marini-Bettolo, Gazz. Chim. Ital. 87, 1163 (1957). 46. B. Witkop, J. Am. Chem. SOC.72,614 (1950); B. Witkop and J. B. Patrick, i b a . 73, 713 (1951). 47. H. Fritz, T. Wieland, and E. Besch, Ann. 611, 268 (1958). 48. B. Witkop and J. B. Patrick, Ezperientia 6, 183 (1950); J . Am. Chem. SOC.73, 2188 (1951). 49. H. Bickel, H. Schmid, and P. Karrer, Helw. Chim. Acta 38, 649 (1955). 50. H. King, J. Chem. SOC. p. 955 (1949). 51. R. C. Elderfield and A. P. Gray, J . Org. Chem. 16, 506 (1951). 52. F. E. Bader, Helw. Chim. Acta 36, 215 (1953). 79, 1519 (1957). 53. E. Wenkert and D. K. Roychaudhuri, J. Am. Chem. SOC. 54. C. Vamvacas, W. von Philipsborn, E. Schlittler, H. Schmid, and P. Karrer, Helv. Chim. Acta 40, 1793 (1957). 55. M. M. Janot and R. Goutarel, Bull. SOC.Chim. France p. 588 (1951). 56. M. M. Janot, R. Goutarel, and J. Chabasse-Massonneau, Bull. SOC.Chim. Prance p. 1033 (1953). 57. 0. Bejar, R. Goutarel, M. M. Janot, and A. Le Hir, Compt. Rend. Acad.Sci. 244,2066 (1957). 58. N. A. Hughes and H. Rapoport, J. Am. Chem. SOC.80, 1604 (1958). 59. A. Le Hir, M. M. Janot, and D. Van Stolk, Bull. SOC.Chim. France p. 551 (1958). 60. K. B. Prasad and G. A. Swan, J. Chem. SOC.p. 2024 (1958). 61. J. Thesing and W. Festag, Eqerientia 15, 127*(1959). 62. H. Kaneko, J. Pharm. SOC. Japan 80, 1374 (1960). 63. Y. Ban and M. Seo, Tetrahedron 16, 5 (1961). 64.’ E. Wenkert and J. Kilzer, J . Org. Chem. 27, 2283 (1962). 65. E. Wenkert, R. A. Massey-Westropp, and R. G. Lewis, .J. Am. Chem. SOC.84, 3732 (1962). 66. E. Wenkert and B. Wickberg, J . Am. Chem. SOC. 84,4914 (1962). 84, 1068 (1962). 67. E. Leete, S. Ghosal, and P. N. Edwards, J. A m . Chem. SOC. 68. A. R. Battersby quoted in footnote 11 of ref. (67); see also E. Schlittler and W. I. Taylor, Ezperientia 16, 244 (1960). 69. A. Stoll and A. Hofmann. Helv. Chim. Acta 36, 1143 (1953); D. Stauffacher, A. Hofmann, and E . Seebeck, Helv. Chim. Acta 40, 508 (1957); J. Poisson, J. Le Men, and M. M. Janot, Bull. Soc. Chim. France p. 610 (1957). 70. W. B. Mom, P. Zaltzmann, J. J. Beereboom, S. C. Pakrashi, and C. Djerassi, Chem. Ind. (London)p. 173 (1956). 71. M. F. Bartlett, R. Sklar, W. I . Taylor, E. Schlittler, R. L. S. Amai, P. Beak, N. V. Bringi, and E . Wenkert, J. Am. Chem. SOC.84, 622 (1962); and refs. therein to preliminary communications.

578

A . R. BATTERSBY AND H. F. HODSON

72. A. R. Battersby and D. A. Yeowell, Proc. Chem. SOC. p. 17 (1961). 73. D. Stauffacher, Helv. Chim. Acta 44, 2006 (1961). 74. J. Levy, J . Le Men, and M. M. Janot, Compt. Rend. Acud. Sci. 253, 131 (1961); M. M. Janot, J . Le Men, J. Gosset, and J. Levy, Bull. SOC.Chim. France p. 1079 (1962). 75. S . Silvars and A. Tulinsky, Tetrahedron Letters X o . 8, 339 (1962). 76. N. Defay, M. Kaisin, J. Pechar, and R. H. Martin, Bull. SOC.Chim. Belges 70, 475 (1961). 77. L. D. Antonaccio, N. A. Pereira, B. Gilbert, H. Vorbrueggen, H. Budzikiewicz, J. M. Wilson, L. J . Durham, and C. Djerassi, J . Am. Chem. SOC. 84, 2161 (1962). 78. H. Rapoport and R. E. Moore, J . Org. Chem. 27, 2981 (1962). 79. J. A. Hamilton, T. A. Harmor, J. M. Robertson, and G. A. Sim, Proc. Chem. SOC. p. 63 (1961). 80. A. T. McPhail, J. M. Robertson, G. A. Sim, A. R. Battersby, H. F. Hodson, and D. A. Yeowell, Proc. Chem. SOC.p. 223 (1961). 81. A. R. Battersby and D. A. Yeowell, In press; D. A. Yeowell, Thesis, Univ. of Bristol, Bristol (1961). 82. W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 39, 913 (1956). 83. H. Wieland, B. Witkop, and K. Bahr, Ann. 558, 144 (1947). 84. A. R. Battersby and H. F. Hodson, Proc. Chem. SOC.p. 287 (1958). 85. K. Bernauer, F. Berlage, W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 2293 (1958). 86. K. Bernauer, H. Schmid, and P. Karrer, Helv.Chim. Actu 41, 1408 (1958). 87. K. Bernauer, S. K. Pavanaram, W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Actn 41, 1405 (1958). 88. H. Wieland and W. Gumlich, Ann. 494, 191 (1932); H. Wieland and K. Kaziro, Ann. 506, 60 (1933). 89. F. A. L. Anet and Sir Robert Robinson, J . Chem. SOC.p. 2253 (1955). 90. R. B. Woodward, Nature 162, 155 (1948). 91. A. F. Peerdeman, Acta Cryst. 9, 824 (1956). 92. V. Boekelheide, 0. Ceder, T. Crabb, Y. Kawazoe, and R. N. Knowles, Tetrahedron Letters No. 26, 1 (1960). 93. W. Arnold, M. Hesse, H. Hiltebrand, A. Melera, W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 44, 620 (1961). 94. C. Weissmann, 0. Heshmat, K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 43, 1165 (1960). 95. F. Berlage, K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 42,2650 (1959). 96. A. R. Battersby, D. A. Yeowell, L. M. Jackman, H. D. Schroeder, M. Hesse, H. Hiltebrand, W. von Philipsborn, H. Schmid, and P. Karrer, Proc. Chem. Soc. p. 413 (1961). 97. F. Berlage, K. Bernauer, W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 42, 394 (1959). 98. K. Bernauer, F. Berlage, W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 42, 201 (1959). 99. E.g. N.R.D.C., Belg. Patent 583,711 (1958); Hoffmann-LaRoche, Brit. Patent 874,560 (1958) and Brit. Patent 886,946 (1959). 100. A. Fiirst, A. Boller, and H. Els, Unpublished observations (1960). 101. K. Bernauer, Helv. Chim. Acta 46, 197 (1963). 102. W. von Philipsborn, H. Meyer, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 1257 (1958). 103. H. Fritz and T. Wieland, Ann. 611, 277 (1958).

15.

ALKALOIDS OF CALABASH CURARE ; Strychnos

579

104. H. Volz and T. Wieland, Ann. 640, 1 (1957). 105. H. Fritz, E. Besch, and T. Wieland, Ann. 617, 166 (1958). 106. H. Fritz, Ber. 92, 1809 (1959). 107. H. Fritz, E. Besch, and T. Wieland, Angew. Chem. 71, 126 (1959). 108. W. von Philipsborn, K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim.Acta 42. 461 (1959). 109. K. Bernauer, W. Arnold, C. Weissmann, H. Schmid, and P. Karrer, Helw. Chim. Acta 43, 717 (1960). 110. K. Bernauer, H. Schmid, and P. Karrer, Helv. Chirn.Acta 40, 1999 (1957). 111. H. Asmis, H. Schmid, and P. Karrer, Helw. Chirn. Acta 39, 440 (1956). 112. K. Bernauer, F. Berlage, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 1202 (1958). 113. A. R. Battersby, H. F. Hodson, G. V. Rao, and D. A. Yeowell, Proc. Chem. SOC. p. 412 (1961). 114. A. R. Battersby and H. F. Hodson, Unpublished observations (1959). 115. H. Schmid, Personal communication (1959). 116. H. D. Schroeder, H. Hiltebrand, H. Schmid, and P. Karrer, Helv. Chim.. Acta 44, 34 (1961). 117. S. M. Kupchan, W. S. Johnson, and S. Rajagopalan, Tetrahedron 7, 47 (1959). 118. A. T. McPhail and G. A. Sim, Proc. Chern. SOC. p. 416 (1961). 119. T. Wieland and H. Merz, Ann. 580, 204 (1953). 120. T. Wieland, H. Fritz, K. Hasspacher, and A, Bauer, A n n . 588, 1 (1954). 121. K. Bernauer, H. Schmid, and P. Karrer, Helw. Chim. Acta 40, 731 (1957). 122. K. Bernauer, E. Biichli, H. Schmid, and P. Karrer, Angew. Chem. 69, 59 (1957). 123. H. Volz and T. Wieland, Naturwissenschajten 44, 376 (1957). 124. K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta41,26 (1958). 125. K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 4 1 673 (1958). 126. H. F. Hodson and G . F. Smith, J . Chem. SOC. p. 1877 (1957). 127. M. Hesse, H. Hiltebrand, C. Weissmann, W. von Philipsborn, K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 44, 2211 (1961). 128. H. Schmid, A. Ebnother, and P. Karrer, Helv. Chim. Acta 33, 1486 (1950). 129. V. Boekelheide, 0. Ceder, M. Natsume, and A. Zurcher, J . Am. Chern. SOC.81, 2256 (1959). 130. W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Acta 38, 1067 (1955). 131. W. von Philipsborn, W. Arnold, J. Nagyvary, K. Bernauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 43, 141 (1960). 132. J. Nagyvary, W. Arnold, W. von Philipsborn, H. Schmid, and P. Karrer, Tetrahedron 14, 138 (1961). 133. T. Wieland, H. Remshagen, and H. Fritz, Naturwissenschajten 48, 50 (1961). 134. M. Hesse, H. Schmid, and P. Karrer, Helv. Chim. Acta 44, 1873 (1961). 135. H. Fritz and H. Meyer, Ann. 617, 162 (1958). 136. 1%.Meyer, H. Schmid, P. Waser, and P. Karrer, Helv. Chim. Acta 39, 1214 (1956). 137. I. Schmidt, P. Waser, H. Schmid, and P. Karrer, Helv. Chim. Acta 43, 1218 (1960).

This Page Intentionally Left Blank

-CHAPTER

16-

THE ALKALOIDS OF CALYCANTHACEAE] R. H. F. MANSKE Dominion Rubber Research Laboratory, Guelph, Ontario, Canade

...................................................... ....................................................... 111. Calycanthine. ....................................................... A. Properties.. ..................................................... B. Structure ........................................................ IV. Calycanthidine ...................................................... V. Folicanthine and Chimonanthine ....................................... VI. Hodgkinsine ........................................................ References.. ........................................................ I. Introduction..

11. Occurrence..

581 581 582 582 582 585 586 588 588

I. Introduction The family Calycanthaceae is represented by only two genera, Calycanthus L., with about four species in the United States, and Chimonanthus Lindl. (Meratia Loisel.) with two Asiatic species. Its relation to other plant families is uncertain, and the nature of its chemical constituents has not shed any light upon possible affinities. Calycanthine is the chief alkaloid in the seeds of these plants. It is not an indole alkaloid, but its facile conversion to indoles and the fact that most of its congeners are indoles is justification for treating it in this volume.

II. Occurrence Following earlier reports of the presence of toxic alkaloids in the seeds of Calycunthus gluucus Willd., the isolation and characterization of calycanthine was recorded by Gordin ( 1 ) . A later attempt by the same author ( 2 )to obtain more calycanthine from the same presumed botanical source led to the isolation of what was assumed to be an isomeric base, 1 Formerly reviewed by Leo Marion in Volume 11, p. 434, and by J. E. Sexton in Volume VII, p. 147. 581

582

R . H. F. MANSKE

isocalycanthine. Subsequent workers have failed to find isocalycanthine, and its existence is very much in doubt, particularly since the reviewer has on several occasions encountered an unstable lower-melting-point crystalline form of calycanthine. A reexamination of C. glaucus yielded calycanthine and a second well-characterized base, calycanthidine (3). Calycanthine has also been isolated from C..poridus L. (4),C . occidentalis Hook. & Am. ( 5 ) , and Chimonanthus fragrans Lindl. (Meratia praecox Rehd. & Wils.) (4).The last contains two alkaloids, as yet uncharacterized, while the leaves of Calycanthus $oridus (6) and Calycanthus occidentalis ( 7 ) contain folicanthine. Recently, the leaves of Chimonanthus fragrans have been examined and found to contain a new alkaloid, chimonanthine (8). The rather remarkable alkaloid hodgkinsine from Hodgkinsonia frutescens C. T. White is almost certainly related to the chimonanthinefolicanthine pair. This relation is the more remarkable because Hodgkinsonia F. Muell. belongs to the family Rubiaceae, which does not have obvious affinities with Calycanthaceae.

111. Calycanthine

A. PROPERTIES Calycanthine crystallizes from most solvents with one molecule of water of crystallization which is tenaciously retained. It then melts a t 216'~218'. It may be obtained in anhydrous form by dissoIving in chloroform, boiling to remove the water, and allowing to crystallize. It then melts a t 245'. Neither form is readily soluble in organic solvents. It is optically active : [ a ] g + 684' (EtOH) (9), [a]? + 698O (acetone) (1). Many of the salts are crystalline, and a nitroso derivative of uncertain structure has been described. Though it soon became evident that the original C11H14Nz formula had to be doubled, there remained some doubt as to the number of hydrogens, and this was only determined in favor of CzzHzsN4 when the structure of the alkaloid was finally settled. When calycanthine is boiled in ethanol with a trace of p-dimethylaminobenzaldehyde (Ehrlich's reagent) and hydrochloric acid, it gives a clear pink color diagnostic of a monosubstituted indole.

B. STRUCTURE There are few examples in the history of organic chemistry where deductions of structure based on identified fragments have been as wide of the mark as in the work on calycanthine. Reactions, which because of

16.

583

THE ALKALOIDS OF CALYCANTHACEAE

their mildness could be presumed to proceed without molecular rearrangements, yielded products that only served to confuse the real structure. Conversely, pyrolyses a t high temperatures which yielded calycanine and could be presumed to involve rearrangements, were the first to give real insight into the structure of this enigmatic alkaloid. The belief in the presence of an indole nucleus in calycanthine was so firmly rooted that the first structures (10) which bear some semblance t o the true one still retained the indole nucleus, even though they were largely based on the then-known structure of calycanine. The first insight into the structure of calycanthine was gained when i t was shown that its benzoylation followed by mild oxidation gave rise to N-benzoyl-N-methyltryptamine (I).Barger and co-workers (12) also I

CO. Ph H

I

I1

obtained N-methyltryptamine. Furthermore, pyrolytic experiments under a variety of conditions gave rise to 3-carboline, skatole, 3-ethylindole, lepidine, quinoline, and, lastly, calycanine (12, 13). This evidence for the presence of a carboline skeleton seemed overwhelming, particularly when it was shown that the product of heating calycanthine with phthalic anhydride was I1 (13). The s€ructure of calycanine was correctly understood to be I11 (lo), and this led Robinson and Teuber to write their epochal and biogenetically feasible structures for calycanthine (a preferred one was IV),

U W H I11

Me

IV

V

which, though still erroneous, served to show the untenability of hithertosuggested structures. These need not be reproduced here, but their correctness was brought into serious doubt by the results of oxidizing

584

R. H. F. MANSKE

calycanthine with silver acetate. The main product was a base (V) (14, 15) shown by a synthesis to have the structure shown (16). It has proved t o be another artifact whose structure does not mirror that of

VI

VII

its parent source. The earlier synthesis of calycanine received only semi-officialstatus a t first by being reported in “The Alkaloids ” (Volume VII) via a private communication. It was again synthesized by Govindachari and co-workers ( 1 7 ) as follows: The intermediate obtained by condensing isatin with o-nitrophenacetyl chloride, on treatment with alkali and subsequently with acid, generated VI. This, on catalytic

VIII

IX

reduction, lactonized spontaneously to VII, which on zinc dust distillation gave calycanine (111).A more recent synthesis of calycanine utilized the condensation of equimolecular quantities of o-aminoacetophenone and o-nitrophenylpyruvic acid in an inert solvent (toluene) in the presence of zinc chloride. The postulated intermediates (VIII and IX) were isolated and IX, on heating with zinc chloride, also generated calycanine (18). It would perhaps be more correct to state that Robinson and Teuber arrived at the correct structure for calycanine on the basis of their calycanthine structure, which was derived from a presumed biogenesis from two molecules of N-methyltryptamine. It had the correct number of functional groups (two each of =NMe and =NH) and explained the readiness with which it could be coupled with diazonium compounds. Furthermore, its conversion to calycanine involved only a minor readjustment of bonds in addition to the usual dehydrogenation and loss of N-methyl. The true structure (X) of calycanthine was revealed by Hamor and

16.

585

THE ALKALOIDS OF CALYCANTHACEAE

co-workers as the result of an X-ray analysis of the dihydrobromide and showed that the alkaloid is a compact three-dimensional unit in which the planes of the two benzene rings are inclined a t 61’ to each other (19, 2 0 ) . The remaining four six-membered rings are in the “chair” form and fused cis to each other. H

Me

\

N ‘’

H

X

Me

H

N Me

H

XI

Me

XI1

Simultaneously, Woodward and co-workers ( 21 ) confirmed structure X for the alkaloid by chemical methods. They first recorded the calycanine synthesis already mentioned. The intermediate tetraaminodialdehyde, first postulated by Robinson and Teuber ( l o ) , was also the basis for their structural speculations. The mercuric acetate oxidation product (dehydrocalycanthine) of Marion and Manske (13), formed by loss of two hydrogens, was smoothly converted by the action of alcoholic alkali into methylamine, and the resulting amide alcohol was written as XI. It was concluded that dehydrocalycanthine is an ene-imine, the relevant portion of the molecule being shown as XII. Other structures derivable from the tetraaminodialdehyde would more probably generate on amidine, and this would be impossible with X because of steric strain a t a bridgehead double bond. Further evidence for structure X was spught in comparing NMRspectra of calycanthine and physostigmine. This comparison indicated that five-membered N rings are not present in calycanthine. Firially, the absolute configuration of calycanthine was arrived a t by measuring the circular dichroism in the region of the long-wavelength absorption band a t 3100 A ( 2 2 ) .

IV. Calycanthidiue

It is rather remarkable that the correct molecular formula [C23H28N4, mp 142O, - 285O (MeOH)] of this alkaloid should have eluded the scrutiny of competent chemists for so long. Xaxton et al. (23) have now shown that the solvate, C23H28N4. BMeZCO, has a molecular weight of 391 t 2, in excellent agreement with theory and with the value of 360

[.IF

586

R . H. F. MANSKE

obtained by mass spectrographic methods. UV-spectra, electrometric titration, IR-spectra, and specific rotation data show that calycanthidine clearly belongs to the folicanthine-chimonanthine group. That it is a once N-methylated chimonanthine follows from the fact that i t is smoothly reduced by zinc and hydrochloric acid t o a mixture of N methyl-2,3-dihydrotryptamine(46% ) and 1,N-dimethyl-2,3-dihydrotryptamine (35%) (7). Calycanthidine is therefore XIV, with one R being Me and the other R being H. The X-ray analysis of chimonanthine dihydrobromide has confirmed the molecular structure XIV for this trio of chimonanthine, calycanthidine, and folicanthine, where R = R = H, R = Me, R = H ; and R = R = Me, respectively.

V. Folicanthine and Chimonanthine

Folicanthine melts a t 118'-119' with [a]$.5 - 364.4' (MeOH). Its earlier empirical and structural formulas (6, 24) have been abandoned in view of the work of Hodson and Smith (7, 25), who showed that it is C24H30N4 and therefore isomeric with N,N-dimethylcalycanthine. It seemed inescapable that folicanthine should be derived from l,Ndimethyltryptamine (C12H&2, picrate, mp 160"-163' and 174'-176'), whose identity was only ascertained by Hodson and Smith (7) although it is easily obtained from folicanthine by zinc dust distillation (26). Moreover, dehydrofolicanthine (a product of silver acetate oxidation) was recognized as 9-methylnorharman. That two dimethyltryptamine units are symmetrically disposed in folicanthine is confirmed by the reductive fission with zinc and hydrochloric acid which gives an 86% yield of N,N-dimethyl-2,3-dihydrotryptamine. The foregoing evidence, as well as that ascertainable from spectra, prompted the latter authors to suggest a structural formula which they amended in a later publication (8).

Chimonanthine (C22H26N4) melts a t 188'-189' with [aID - 329' (EtOH)and is therefore isomeric with calycanthine (8).Like folicanthine, i t readily suffers reductive fission and gives rise to N-methyl-2,3-dihydrotryptamine. The IR-spectra of the two bases are much alike, and it is reasonable to assume that folicanthine is dimethylchimonanthine. The indoline structure was retained for these alkaloids because they either do not yield calycanine a t all or only in insignificant amounts upon dehydrogenation. Furthermore, folicanthine dimethiodide, on Hofmann degradation, generated a crystalline base whose suggested structure (XIII) (27) has been confirmed by a synthesis of the racemate (28). The

16. THE ALKALOIDS OF CALYCANTHACEAE

587

relative stereochemistry of XI11 foUows from the fact that it is opticaily active. In contact with dilute sulfuric acid, it suffers a remarkable cleavage into 1 ,N,N-trimethyltryptamine and l-methyl-3-(/3-dimethylaminoethy1)oxindole in equimolecular proportions and in virtually quantitative yield. Consequently, folicanthine (R = Me) and chimonanthine (R = H) have been given either structure XIV or XV; evidence from mass spectrographic data is in slight favor of XIV, which has been shown to be correct by X-ray methods (29).

XV

A remarkable synthesis of meso- and racemic chimonanthine has recently been published (30).The sodium salt of the urethane (XVI) was dimerized by oxidation with iodine in tetrahydrofuran. Two diastereoisomeric dimers (XVII; R = CHzCHz. C02Et) were obtained; XVIIa, mp 243'-245", and XVIIb, mp 214"-216'. Reduction of the former with LiAlH4 in tetrahydrofuran yielded an isomer of calycanthine (mp 240'243') which was regarded as retaining one of the indole nuclei. Similar reduction of the second isomer (XVIIb) yielded optically inactive chimonanthine (mp 202"-203") whose IR- and mass-spectra were identical with those of the natural base.

XVI

XVII

588

R. H. F. MANSKE

VI. Hodgkinsine Hodgkinsine was isolated from Hodgkinsonia frutescens by Anet et al. from the leaves of this Queensland plant (31). It crystallizes with benzene of crystallization and then melts at 128". The solvent-free base was amorphous and had [a]: + 60". The formula of the authors (C22H26N4) has been revised by Smith (32) to C21H24N4 and is looked upon as possibly a monodemethylchimonanthine. I n view, however, of the reported presence of two N-methyl groups and because of the [a]=value differing greatly from that of chimonanthine, the reviewer tentatively favors the discoverers' formula and considers that it may be a stereoisomer of chimonanthine. Addendum

Hino (33) had already synthesized a number of 3,3'-disubstituted 3,3-bisoxindoles among which was XVIII (R = R' = H). When the Schiff base of this (R,R'=PhCH=) is heated with methyl iodide, the corresponding bismethiodide is formed and hydrolysis with dilute hydrochloric acid generates the base XVIII (R = H, R' = Me). Reduction of the last with lithium aluminum hydride in dioxane afforded ( )-folkanthine (mp 168"-169") identical in the IR-spectrum with that of natural folicanthine (34). NILR'

NRK'

XVIII R=R'=H R = H , R'=Me

REFERENCES 1. H. M. Gordin, J . Am. Chem. SOC. 27, 144, 1418 (1905). 2. H. M. Gordin, J . Am. Chem. SOC.31, 1305 (1909). 3. G. Barger, A. Jacob, and J. Madinaveitia, Rec. Traw. Chim. 57, 548 (1938). 4. R. H. F. Manske, J . Am. Chem. SOC.5 1 , 1836 (1929). 5. R. H. F. Manske and L. Marion, Can. J . Res. B17, 293 (1939). 6. K. Eiter and 0. Svierak, Monatsh. Chem. 82, 186 (1951); Chem. Abstr. 45,7576 (1951).

16. THE ALKALOIDS OF 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

CALYCANTHACEAE

589

H. F. Hodson and G. F. Smith, J . Chem. Soc. p. 1877 (1957). H. F. Hodson, Sir Robert Robinson, and G. F. Smith, Proc. Chem. SOC. p. 465 (1961). E. Spath and W. Stroh, Ber. 58, 2131 (1925). Sir Robert Robinson and H. J. Teuber, Chem. I n d . (London)p. 783 (1954). R. H. F. Manske, Can.J. Res. 4,275 (1931). G. Barger, J. Madinaveitia, and P. Streuli, J . Chem. SOC. p. 510 (1939). L. Marion and R. H. F. Manske, Can. J . Res. B16,432 (1938). E . Spath, W. Stroh, E . Lederer, and K. Eiter, Monatsh. Chern. 79, 11 (1948). K. Eiter, Monatsh. Chem. 79, 17 (1948). K. Eiter and M. Nagy, Monatsh. Chem. 80, 607 (1949). K. W. Gopinath, T. R. Govindachari, and S. Rajappa, Tetrahedron 8, 291 (1960). T. Kobayashi and R. Kikumoto, Tetrahedron 18, 813 (1962). T. A. Hamor, J. M. Robertson, H. N. Shrivastava, and J. V. Silverton, Proc. Chem. SOC. p. 78 (1960). T. A. Hamor and J. M. Robertson, J . Chem. Soc. p. 194 (1962). R. B. Woodward, N. C. Yang, T. J. Katz, V. M. Clark, J. Harley-Mason, R. F. J. Ingleby, and N. Sheppard, Proc. Chem. SOC.p. 76 (1960). S. F. Mason, Proc. Chem. Soc. p. 362 (1962). J. E . Saxton, W. G. Bardsley, and G. F. Smith, Proc. Chern. SOC. p. 148 (1962). P. R. Levy and Sir Robert Robinson, Feschr. Paul Karrer p. 39 (1949). H. F. Hodson and G. F. Smith, Chem. Ind. (London)p. 740 (1956). K. Eiter and 0. Svierak, Monatsh. Chem. 83, 1453 (1952); Chem. Abstr. 48, 691 (1954). H. F. Hodson, G. F. Smith, and J. T. Wrobel, Chem. Ind. ( L o n d o n ) p. 1551 (1958). T. Hino, Unpublished observations (1963). I. J. Grant, T. A. Hamor, J. M. Robertson, and G. A. Sim, Proc. Chem. SOC.p. 148 (1962). J. B. Hendrickson, R. Rees, and R. Goschke, Proc. Chem. SOC.p. 383 (1962). E. F. L. J. Anet, G. K. Hughes, and E. Ritchie, Australian J. Chem. 14, 173 (1961). G. F. Smith, Personal communication (1963). T. Hino, Chern. Pharm. Bull. ( T o k y o ) 9, 979, 988 (1961). T. Hino and S. Yamada, Tetrahedron Letters p. 1757 (1963).

This Page Intentionally Left Blank

17-

+"AFTER

STRYCHNOS ALKALOIDS G. F. SMITH The University. Manchester. England I . Strychnine and Brucine: Historical Survey .............................

I1. The Reactions of Strychnine. Brucine. and Their Derivatives and Degradation Products .......................................................... A . Electrophilic Substitution on the Aromatic Ring ..................... B . Reactions Around Rings I11 and VII (Not Involving Oxidation or Reduction) ..................................................... C . The Quaternary Salts and Their Reactions with Nucleophilic Reagents : The Hofmann and von Braun Degradations ......................... D . Reductions ..................................................... E . Oxidations ...................................................... F . Pseudostrychnine (XXV) and N-Methyl-sec-pseudostrychnine(CCVI. CCVII) ......................................................... G . High-Temperature Degradation of Strychnine. Brucine. and Derivatives H . Neostrychnine (LXVII-A)........................................

592 599 599 600 607 612 621 636 64 1 641

I11. a-and /3.Colubrines .................................................

642

IV. The Total Synthesis of Strychnine .....................................

642

V . Vomicine: Historical Survey .......................................... VI . The Reactions of Vomicine and Its Derivatives and Degradation Products .. A . Reactions Involving the Benzene Ring and the Phenolic (Hydroxyl Group (Not Including Oxidations) ................................ :....... B . Reactions Around Rings I11 and VII (Not Involving Oxidation or Reduction) ; the Deoxyvomicines ................................... C. Quaternizations : Hofmann and Emde Degradations . . . . . . . . . . . . . . . . . . D . Reductions ..................................................... E . Oxidations ...................................................... F . Dehydrogenations : Voniipyrine and Related Compounds .............. VII . Minor Alkaloids .................................................... A . The Australian Strychnos Alkaloids ................................. B . The Congo Strychnos Alkaloids ..................................... C. S . nuz-uomica Alkaloids ........................................... References .........................................................

647 648 648 649 653 656 659 663 663 663 665 665 666

Strychnine and brucine have been the subject of a very large number of degradative researches. nearly all of them before the advent of modern spectroscopic techniques. and the elucidation of their structure represents one of the major achievements of classical organic chemistry . The 591

592

G. F. SMITH

following treatment of this topic falls into two sections: the first is a concise historical survey in chronological order of the events which, in 1948, led to the final proof of the structure; the second section is an attempt to group the various more important transformations into reaction types to make more readily accessible t o the nonspecialist in the strychnine field the vast fund of degradative experience which this field holds.

I. Strychnine and Brucine: Historical Survey Strychnine, C21H2202N2, and brucine, C23H2604N2, were first isolated as far back as 1819 from the seeds and bark of Strychnos nuxvomica by Pelletier and Caventou (1) and were fully characterized some 20 years later by Regnault (2). The actual degradative work started in the 188O’s,and the finishing touches were published in 1948 in a memorable paper by Woodward and Brehm (3), the major contributioiis being made by H. Leuchs and his school and by Sir Robert Robinson and W. H. Perkin, Jr., and their collaborators. Robinson (4)and Huisgen (5) have written summarizing historical accounts of the work on strychnine and brucine. The paper by Woodward and Brehm (3) in 1948 contains an excellent and detailed discussion of the final stages of the structural proof. The first recognizable fragment, picric acid, was isolated by Shenstone (6) from the nitric acid oxidation of strychnine and was taken t o indicate the presence of a benzene ring in that base. Later oxidations with the same reagent ( 7 ) yielded a carboxylic acid, dinitrostrycholcarboxylic acid, which 30 years later was to play an important role. Tafel, as the result of studies involving alkaline hydrolysis, methylations, hydriodic acid deoxygenations, and electrolytic reductions (8), established the presence of a basic tertiary nitrogen, an inert oxygen, and the Ar-N-

I

1

C=O

system in the molecule. Hanssen (9) found that strychnine and brucine are both oxidized by chromic acid to the same amino acid, C16H2004N2, thus showing brucine to be a Bz-dimethoxystrychnine. From then on, of course, the degradation of both alkaloids became a single structural problem, and most of the reaction sequences were duplicated. I n some cases, a reaction could only be made to yield a crystalline product in one series : thus, slthough acetylbrucinolone-b gave crystalline curbine (CLX),the oxidation of acetylstrychninolone-b

17. Strychnos

593

ALKALOIDS

is not mentioned, presumably because it failed t o yield a homogeneous product (vide infra). The first insight into the nature of the nonaromatic portion was obtained by H. Leuchs and his school, whose exemplary skill and thoroughness led to the first stepwise degradation of these alkaloids. R

A

I I -N-CO

I 1 -N-CO

/ \ /

CO-CH=CH

CH I

R \ /

CHOR'

CO-CH2-CH

I

I/

I1

1 ) -N-CO

co I

I COzH

\ /

CHOAC

COzH

A

I I -N-CO

I co

I1

I CHz-C0zH

0

I / -N-CO

/

< l N /H

\CHOR'

/

\ CCOzH I6

\C H O H

CHOAc

I 1 -N-CO

ti

H

/

VI

\

/

\

/

CHOH

IV

Their contribution in terms of experimental work was great, but in terms of structural elucidation, it was essentially limited t o the demonstration that brucinolone-a and strychninolone-a have partial structures I (R = OMe and H, respectively; R' = H) and that the corresponding b-isomers have partial structures I1 (R = OMe and H, respectively; R' = H). The degradation sequence proceeded as follows. Brucine, C23H2s04N2, was oxidized by permanganate in acetone to a mixture of the monocarboxylic acids brucinonic acid, C Z ~ H ~ ~ Oand S N dihydro~, brucinonic acid, C23H2608N2 (10). The latter compound contained a secondary hydroxyl function which in the former was replaced by a

594

G . F. SMITH

ketone carbonyl group. Reduction of this ketone group with sodium amalgam, however, gave not dihydrobrucinonic acid but the epimeric secondary alcohol, brucinolic acid (11) which, on being treated with alkali, lost glycolic acid to give a neutral compound brucinolone-a, CzlHzzOsNz ( I ; R = OMe, R’ = H ) that was easily isomerized to brucinolone-b (11;R = OMe, R’ = H ) (11, 12). Permanganate oxidation of acetylbrucinolone-b (11; R = OMe, R’ = Ac) gave the acid C23Hz409N2 (111),which on acid hydrolysis gave malonic acid and a base, curbine, ClsHzo05Nz (IV, CLX) (13). The acetyl derivative of strychninolone-a (1,R = H, R’ = Ac) was oxidized to the dicarboxylic acid CslHzoOsNz (V), which on hydrolysis gave oxalic acid and the amino acid C17H1804Nz (VI) (14).These data led Leuchs t o the correct partial structures for the brucinolones and strychninolones, the arguments being evident in formulas I-VI. On the other hand, the interpretation given by Leuchs (11)to the sequence, brucine + dihydrobrucinonic acid + brucinolone-a, and a t the time accepted by Perkin and Robinson (15), turned out to be quite wrong and need not be described in this account. The correct interpretation of this sequence of reactions and the definition of the complete chain of seven atoms between N, and Nb was achieved by Robinson and associates and described in two papers (16, 1 7 ) . This was a very great step forward. Not only was the true nature of the glycolic acid elimination as a simple /!-elimination with respect to the N, amide carbonyl realized, but also the solution of the appearance of a secondary hydroxyl group in the permanganate oxidation product, dihydrobrucinonic acid. This difficulty had previously been avoided by assuming this alcoholic grouping to be already present in the parent alkaloids, in spite of the inability of strychnine or brucine to form acetyl derivatives. The truly brilliant idea which occurred to the English group was that this secondary hydroxyl must arise by the hydrolytic fission of an a-hydroxy-B-keto-amide grouping, I l l 1 O=C--CCN(b) I II I OH0

Ha0 __f

I COzH

I I I CH--CN(b)

I

/I

OH 0

I

in which the nitrogen must be the alkaloid N,, the three oxygen atoms being introduced by the oxidizing agent : all this led to partial structure VII for strychnine, going by way of VIII to I X , which represented dihydrostrychninonic acid, which on being treated with alkali suffered /?-elimination of glycolic acid to give strychninolone-a (I;R = R‘ =.H). The position of the methoxyl groups in brucine was firmly proved by permanganate oxidation to 4,5-dimethoxy-N-oxalylanthranilic acid (18).

1 7 . Strychnos

595

ALKALOIDS

A t this stage (1930),the nature of 7 of the 21 carbon atoms of strychnine had still to be discovered, and the arrangement of the seven rings established. This formidable problem was gradually solved over the succeeding 18 years, the final solution being very largely owing to Robinson’s insight.

v I1

I I

XIV

xv

XVI

A number of guesses had already been made concerning the over-all structure of these alkaloids : all such structures contained a quinoline or isoquinoline ring system because of the unproved view, current for a long time, that dinitrostrycholcarboxylic acid was a quinoline or isoquinoline derivative. Thus Robinson, in his Bakerian lecture to the Royal Society in 1930 (IS)), proposed structure XIV1 for strychnine. Not lang ____ 1 Numbers X-XI11 have not been used.

596

G. F. SMITH

after this, dinitrostrycholcarboxylic acid was a t last proved to be an indole derivative (XV) (20). This, together with the formation of carbazole on dehydrogenation, the inclusion of a tryptamine unit on biogenetic grounds, and the assumed presence of a blocked dihydroindole system because of the failure of attempts to convert strychnine into an aromatic indole by removal of two hydrogens, led Robinson (21) to propose structure XVI for strychnine. This was very nearly the correct solution to this complex problem. Much unnecessary trouble was then caused by Leuch’s contention (22) that, since dioxonucidine formed a monobromo derivative said to be XVII, the ,€-carbonof the indole system could not be quaternary, and he proposed structure XVIII. Robinson accepted this criticism, and modified his structure to XIX, abandoning the notion of a tryptamine

@a & 0

0

XVII

0

XVIII

0

CLB 0

0

XIX

unit being present (23). Several years later, Holmes and Robinson (24) found that they could not repeat Leuchs’s bromination of dioxonucidine, and Leuchs himself reported that bromine did not after all react by substitution, but attacked the double bond of dioxonucidine ( 2 5 ) .This, of course, led to a return to structure XVI for strychnine which fit in well with the intervening actual isolation of tryptamine by the pyrolvtic alkaline degradation of certain strychnine derivatives (26, 271, which for 3 years had cast doubt on the correctness of XVIII and XIX. It was in 1946 that the accepted structure X X for strychnine was proposed for the first t8ime(28). This proposal was based mainly on the following arguments: (a) it explained the formation of ,€-collidineduring dehydrogenative pyrolyses ;

17. &rychnos

597

ALKALOIDS

(b) it accommodated the demonstration that ring D ( = ring VI in formula X X ) must be six-membered (29, 123); and (c)itshoweda structuralrelationship with cinchonine. The size of ring D was fixed by a study of the amino acid X X I : Prelog and Szpilfogel showed that decarboxylation in pyridine followed by crystallization from methanol yielded a

XXII

XXI

carbinolamine methyl ether which must be formulated as X X I I ; Robinson and co-workers showed that the product of oxidative decarboxylation of XXI, cuninecarboxylic acid (XXIII),quite readily formed a stable lactam. Both X X I I and X X I I I lactam must contain a ring D that has contracted to a five-membered ring; hence, the ring must originally have been six-membered.

XXIII

XXIV

xxv

That all this evidence still did not prove the structure X X was evident from the fact that two further structures were advanced in 1947. Areturn to structure X X was made when information on the relationship of Nb with the B-carbon of the indoline ring was derived from the realization (30) that strychnone is, in fact, an N-acylindole (XXIV). Strychnone (31) is formed by oxidation of pseudostrychnine (XXV), the carbinolamine oxygen of which becomes the N,, amide carbonyl oxygen of XXIV :this can be rationalized only if the carbinolamine carbon of XXV is attached directly to the /3-carbon of the indoline ring. One major difficulty connected with structure X X a t this stage was the position and degree of substitution of the olefinic bond in the neo series. Much experimental work had been carried out, but the results were difficult to interpret. This question was finally solved independently

598

G . F. SMITH

by Chakravarti and Robinson (32) and Woodward and Brehm (3). The former workers found that neostrychnine reacts with nitrous acid to give a neutral compound which can be hydrolyzed to a see-amino ketone, formic acid, and hydroxylamine, and interpreted these observations unambiguously as illustrated. I HNOz

"\CHo

HCOzH

dilute HCl ___f

-40

I

NHzOH I

The American workers took up the vexed question of the nature of the oxidation of methoxymethyldihydroneostrychnine: this compound (XXVI), formed by the action of methanolic potassium hydroxide on strychnine methiodide (33), is oxidized by perbenzoic acid to a ketone (XXVII) which, being neutral, must contain both nitrogens in amide groups (34). This oxidation product, named methoxymethylchanodihydrostrychnone, was found t o be resistant to hydrolysis by concentrated hydrochloric acid. This observation was the cause of a long series of incorrect speculations, for it led to the conclusion that the newly formed amide group must be part of a cyclic amide or lactam system.

.

KOH

oxidation

_ j

MeOH

XXVI

XXVII

The question was finally solved in one of the last important papers on the degradation of strychnine (3), when it was found that dilute mineral acid hydrolyzed this oxidation product quite readily to give formic acid. The series of reactions was thus formulated as has been shown, and the last obstacle to the acceptance of structure XX was removed. The degradation work had up to this point occupied itself mainly with the proof of the gross skeletal structure, and relatively little attention had been given to the stereochemistry of the molecule. It was at this point that this aspect of the problem was cleared by the determination of the structure of strychnine by X-ray crystallographic analysis independently by Robertson and Beevers (35) and Bijvoet et al. (36).

17. Sfrychnos ALKALOIDS

699

Later, X-ray crystallography, in the hands of Peerdeman, even provided the absolute configuration, (XXVIII, XXIX) (37),later confirmed by a chemical method by Schmid and his collaborators (37a) ( s w 1). 632)).

XXVIII

XXIX

The very interesting stereochemical arguments based on the chemical properties of strychnine and its degradation products, that were elaborated by Holmes and published in 1952 (38), did not prove the complete stereochemistry of the molecule, but they did show that all the previous observations that had stereochemical implications were in accord with the X-ray structure. The final stage was reached when Woodward and his collaborators announced the total synthesis of strychnine, a truly great achievement (see Section I V ) (39). 11. The Reactions of Strychnine, Brucine, and Their Derivatives and Degradation Products

A. ELECTROPHILIC SUBSTITUTION ON THE AROMATICRING

'

Strychnine undergoes electrophilic attack a t position 2, para to the aromatic nitrogen atom. Thus, 1 mole of chlorine on aqueous strychnine hydrochloride yields 2-chlorostrychnine, together with a small quantity of a trichlorostrychnine of unknown constitution (40). Under similar conditions, bromine yields 2-bromostrychnine (41).Nitration with nitric acid (D 1.41) a t 0" gives 2-nitrostrychnine (42). It is of interest to note that the action of concentrated or fuming sulfuric acids on strychnine has not given a homogeneous product (43). Strychnidine has not been halogenated or nitrated, but has been coupled with benzenediazonium chloride to give a 2-phenylazo derivative (44),reduced to 2-aminostrychnidine. Reduction of %nitrostrychnine gives 2-aminostrychnine (42).

600

G. F. SMITH

Noteworthy is the formation of 2-nitrosoisostrychnic acid (XXXI)by the action of alcoholic hydrochloric acid on N-nitrosoisostrychnic acid (XXX) ( 7 ) .

B. REACTIONS AROUND RINGSI11 AND V I I (NOTINVOLVING OXIDATIONOR REDUCTION) 1. Opening of the N, Lactam Ring III Strychnine is converted into strychnic acid (XXXII) under relatively mild alkaline conditions, such as sodium ethoxide in ethanol at 50" (8). Strychnic acid easily lactamizes back to strychnine by the action of hot dilute mineral acid (45). Brucine behaves in an entirely analogous manner (46).

bOzH XXX

I

COzH XXXI

More vigorous conditions, such as aqueous barium hydroxide a t 140" (47) or sodium hydroxide in boiling amyl alcohol give isostrychnic acid (XXXIII) (48), in which the configurations of C-12 and '2-13 have been inverted (49). This reaction probably proceeds by way of XXXIV, and the subsequent closure of the ether ring, since isostrychnine-I (XXXV; R = OH) also gives isostrychnic acid (44).It would seem that the

17. Strychnos ALKALOIDS

601

inverted configuration a t C- 13 gives a thermodynamically stabler molecule when the lactam ring 111 is open. I n this configuration, this ring cannot close again, for attempts to lactamize isostrychnic acid have failed. The fact that the Wieland-Gumlich aldehyde (LV, see Section 11, B, 4),on being made to react with malonic acid in the presence of pyridine and piperidine, gives isostrychnic acid and not strychnine (49)underlines the stabler configuration a t C-13 in isostrychnic acid. The conversion of strychninolone-a into strychninolone-c is entirely parallel (see Section 11, B, 3). The N, lactam in strychninonic acid (XXXVIII) (50) and brucinonic acid (51) opens with boiling 5 N hydrochloric acid for 30 minutes or with 1 N NaOH a t 0" for 48 hours. The alkaline hydrolysis was only reported for brucinonic acid : it is interesting t o note that, under these conditions, some elimination of glycolic acid also occurs (cf. Section II, B, 3 ) . Lactam hydrolysis under acidic conditions only seems to occur with compounds in which N1, is neutral; thus, strychninolones-a and -b are converted into the amino acids, isolated as the methyl esters, by warming with concentrated hydrochloric acid a t 100" (50, 52). Strychninolone-c cannot be hydrolyzed under these conditions, which is in keeping with it being the thermodynamically most stable of the three isomers ( 5 2 ) .

2. Opening of Ring V I I This ring opening is simply an acid- or base-catalyzed elimination of the ether oxygen made possible by the N, lactam carbonyl, to which the ether oxygen is in a /3 relationship the key step is removal of a proton from C- 11. This reaction is undergone by strychnine, dihydrostrychnine, and dihydrobrucine: strangely, it has not been reported for brucine. Strychnine is converted into isostrychnine-I (XXXV; R = OH), which

xxxv

XXXVI

is best produced by the action of water a t 180" (53),methanolic ammonia a t 15C" (54),or hydrobromic acid in glacial acetic acid a t the boiling point (55). The last reagent, however, gives as main products two isomeric bromodeoxyisostrychnine hydrobomides (XXXV, R = Br ;

602

G . F. SMITH

XXXVI, R = Br) (56).The position of the double bond in isostrychnine-I follows from the UV-spectrum, which shows it not t o be in conjugation with the lactam carbonyl (57). Alcoholic potassium hydroxide causes a conversion back to strychnine in 20 yoyield, presumably by isomerization of the double bond back into conjugation with the lactam carbonyl followed by p-addition of the ally1 alcohol oxygen. No product corresponding to a double-bond isomer of isostrychnine could be isolated, or even detected UV-spectroscopically in the noncrystalline part of the reaction mixture (57). Dilute HBr hydrolysis of the less soluble bromodeoxyisostrychnine (XXXV; R = Br) yields isostrychnine-I (56, 58), whereas hydrolysis of the more soluble salt gives the isomeric isostrychnine-I1 (XXXVI ; R = OH), the position of the double bond being fixed by its ease of reduction by sodium amalgam. Isostrychnine-I1 has not been directly obtained from either strychnine or isostrychnine-I. Dihydrostrychnine gives only one isodihydrostrychnine : the position of the double bond does not seem to have been proved. Dihydrobrucine, on the other hand, gives three isomeric isodihydrobrucines (52). Treatment of brucine with hydrobromic acid and glacial acetic acid a t the boiling point gave bisapomethylbromodeoxyisobrucine, which was hydrolyzed by dilute hydrobromic acid to bisapomethylisobrucine. The position of the double bond in ring I11 has not been determined (58).

3. Glycolic Acid Elimination from Strychinolones and Related Compounds This is very closely related to the elimination leading to isostrychnine. Strychinolic acid (XXXIX) gives a nearly quantitative yield of strychninolone-a (XL) with 1 N NaOH a t room temperature for 1 day (59). Dihydrostrychninonic acid (XXXVII) undergoes an analogous elimination, but much less readily, for isostrychninolone-I (XLT) is produced in only about 10% yield after 1 week in 1.5 N NaOH a t room temperature (41).The reaction of brucinonic acid (2,3-dimethoxy-XXXVIII) with aqueous alkali, however, besides about 10% of the amino acid formed by opening of ring 111, gives about 30% of the glycolic acid isolated as the zinc salt (51): it seems likely, therefore, that the failure to isolate brucinone from this reaction is mainly owing to its failure to crystallize. That the N, lactam carbonyl plays a leading role in acidifying the (3-11 protons is demonstrated by the total resistance of dihydrobrucidinonic acid (XLII, see Section 11,D, 4)t o alkaline glycolic cleavage (60).This, of course, parallels the failure of strychnidine to give an isostrychnidine. It has been suggested (61) that the great ease of glycolic acid elimination in the case of strychninolic acid is owing to facilitation of the

17,

&WJCh'?%OS

603

ALKALOIDS

separation of the alkoxy anion through internal hydrogen bonding with the Czl hydroxyl group (XLIII). If one accepts this very plausible idea, then the stereochemistry of strychninolic acid must be as in XXXIX and that of dihydrostrychninonic acid as in XXXVII, the hydroxyl group in the latter being well away from the ether oxygen when ring VI is a boat form (XLIV) (62). The orientation of this hydroxyl will be

x LI

XLIL

XL

1 CHzCOzH

I

4; I5

XLIV

N

XLIII

0 -

XLV

L

604

G. F. SMITH

determined by the mode of addition of proton to the enolate anion XLV which is the immediate product of the fission of the ,B-dicarbonylsystem I -CO-COH-CO-N-

I

A stereoelectronic argument (62) would suggest quasi-axial addition leading t o the right configuration (XLIV)for dihydrostrychniaonic acid. The reduction of strychninonic acid to the epimeric strychninolic acid is governed simply by the more ready accessibility of the “outer” side of the ketone carbonyl group (62). More recent experiments bearing on the assisted elimination of glycolic acid show that it occurs by acid catalysis as well (63). It is argued that, in this case, since the mechanism must involve protonation of the C-12 oxygen, hydrogen bonding with the C-21 hydroxyl cannot be involved in the assistance (63). Stabilization of the oxonium ion by increasing the ‘‘ solvation ” can, however, be visualized. The reactions are the hot concentrated HCl hydrolysis of ring VI in hydroxydihydrostrychnone-a (XLVI)to the amino acid (XLVII) with retention of the C-12 hydroxyl, and similar hydrolysis of hydroxydihydrostrychninolone-a(XLVIII) to XL with elimination of the C-12 hydroxyl.

XLVI

XLVII

XLVIII

The double bonds in the strychninolones and brucinolones are subject to base- and acid-catalyzed migration. Thus strychninolone-a (XL) is converted into strychninolone-b (XLIX) in 0.5 N NaOH a t room temperature for several days (64) and inio strychninolone-c (L) on heating to 100”with 1 N NaOH (63) or with methanolic NH3 a t 100’ (64). These changes are irreversible, and represent increases in stability of the molecules, that is, mainly decreases in steric strain in the skeleton. The formation of strychninolone-b before the thermodynamically more stable strychninolone-c is an example of kinetic control, the addition of a proton to the common intermediate anion taking place more rapidly a t the center of the mesomeric system (62). The structures follow from the oxidation experiments of Leuchs et al. (14, 65) and from the observation that catalytic hydrogenation of a and c gives different dihydro derivatives (52, 66), both of which are produced by hydrogenation of b.

17. Xtrychnos

605

ALKALOIDS

Further confirmation of the position of the double bonds comes from the UV-spectra, which easily differentiate between conjugated and nonconjugated systems (66). The stereochemistry follows from the order of formation : strychninolone-a must have the same configuration as strychnine a t C- 13, and strychninolone-c the inverse. As an example of the acid-catalyzed migration of the double bond, we have the conversion of strychninolone-a into acetylstrychninolone-b with acetic anhydride and dry hydrogen chloride a t 100" (64). 4. Reactivity of the N , Lactam a-Methylene (C-11) Strychnine condenses with amyl nitrite in the presence of sodium ethoxide to give 11-oximinostrychnine (LI) (67, 68). This oxime undergoes Beckmann rearrangement with thionyl chloride to give Lwo products, L i I and LIII, the first of which gives the amino acid LIV on

NOH LI

L1I

LIII

1

I

LIV

LV

&OzH LVI

hydrolysis and the second of which is hydrolyzed with loss of Cog and HCN to give the Wieland-Gumlich aldehyde (LV), the key compound in the elucidation of the struetare of the curare alkaloids. The WielandGumlich aldehyde exists as the lactol (49). Very remarkable is the different behavior of dihydrostrychnine toward amyl nitrite and sodium ethoxide. The product is very unstable

606

G . F. SMITH

and was not obtained in the pure state. It is an amino acid, contains an N, nitroso group, and is given structure LVI (67). This spontaneous opening of the lactam ring is the more remarkable since dihydrostrychnine itself is not hydrolyzed to dihydrostrychnic acid by alkali. The ease of attack of the N, amide carbonyl by the hydroxyl ion will very much depend on the extent t o which the stereochemistry ofring I11 allows the amide mesomerism to develop. Presumably, this is at its greatest in dihydrostrychnine, is less developed in strychnine itself, and the combination of tetrahedral C-21 and trigonal C-11 greatly reduces it, thus making the amide group most susceptible to nucleophilic attack. The C-11 methylene also condenses with benzaldehyde : strychnine reacts in hot aqueous alcoholic KOH t o give the yellow 1l-benzylidenestrychnine (LVII) (69). Likewise, neostrychnine gives 1l-benzylideneneostrychnine (33).

OH II

CHPh

CHPh

LVII

LVIII

LIX

CHPh

CHPh

t LX

CHzPh

= CHPh

LXI

LXII

Under mild conditions, that is, with piperidine as catalyst, isostrychnine condenses to give the yellow compound LVIII (57). Under more vigorous conditions, however, the colorless pyridone derivative (LIX) (70) is produced by double-bond migration (4).This easy isomerization was of structural value in demonstrating clearly the presence of a hydrogen atom on C-8. Isodihydrostrychnine and the isodihydrobrucines also give yellow and colorless benzylidene derivatives ( 7 1). N-methyl-sec-pseudostrychnine (CCVI) can be made to condense with

17. Strychnos

ALKALOIDS

607

one benzaldehyde molecule over a short reaction time to give LX (72) and over a longer reaction time to give the dibenzylidene derivative LXI (73). The value of this was t o demonstrate the presence of a methylene group next t o the potential C-16 carbonyl of the pseudostrychnine. The secondary alcohol LXII, of course, only gave a monobenzylidene derivative.

5 . Nucleophilic Addition to the C-11 to (7-12 Double Bond Compounds with the a,P-unsaturated N, lactam system tend to add nucleophiles ; thus, strychninolone-a (XL)gives LXIII (R = CHSO)with methanolic sodium methoxide (74), and LXIII (R = NH2) with methanolic ammonia at 100' (together with strychninolone-c) (64). The conjugated double bond in the c isomers is, however, less reactive, for although brucinolone-c adds ammonia, it does not add methanol (75). The conversions of isostrychnine into strychnine and isostrychnic acid (see Sections 11, B, 2 and IV) also involve nucleophilic addition at C-12. Hydroxydihydrostrychnolone-a (LXIII ; R = OH), as has been mentioned, easily loses water : the reverse reaction, nucleophilic addition of OH to strychninolone-a, occurs as a minor reaction in the alkaline cleavage of strychninolic acid to strychninolone-a and in the basecatalyzed isonierization of strychninolone-a to strychninolone-b (63). C. THEQUATERNARY SALTSAND THEIRREACTIONS WITH AND VON BRAUN NUCLEOPHILIC REAGENTS : THEHOFMANN DEGRADATIONS Pyrolysis of the quaternary chlorides of strychnidine, neostrychnidine (76), dihydrostrychnidine (44), and neostrychnine (33) lead to the

regeneration of the original bases. Strychnine metho salts, when heated with aqueous alkali, suffer only N, lactam opening to give strychnic acid methyl betaine (LXIV) ( 7 6 , 7 ? ) but, when either the betaine or the metho salts are treated with hot methanolic sodium methoxide, methoxylating fission occurs to give methoxymethyldihydroneostrychnine (LXV) (76, 77, 3). The fission is probably preceded by migration of the double bond to the neo position (C-20 to C-21), for neostrychnine methiodide (LXVI) can be isolated from the reaction as a by-product (33). Migration of the double bond to the neo position makes C-20 no longer susceptible to nucleophilic attack, and perhaps even facilitates the SN2type of attack on C-18. No products corresponding to either of the alternative fissions appear t o be formed. Neostrychnine metho salts, of course, also give LXV. A parallel series of

608

G. F. SMITH

changes occurs when strychnidine N,-mono- and N,, Nb-dimetho salts are treated with sodium methoxide : both give methoxymethyldihydroneostrychnidine (LXVII) (76). The (3-18 to Nb bond is still the most easily broken one when the C-21,C-22 double bond is reduced. Thus, the action of sodium methoxide on the dimethochloride of dihydrostrychnidine-A gave about equal proportions of methoxymsthyltetrahydrostrychnidine (LXVIII ; R = CHs), identical with the product of hydrogenation of LXVII, and methyl dihydrostrychnidine-A, which was not investigated further (44).

LXIII

LXVI

LXV

pyrolysis of chloride

60,LXIV

LXVII-A

LXVII

When, however, the Hofxnann degradation is effected by pyrolysis of the quaternary hydrogen carbonate, a separab!e mixture of unchanged dihykostrychnidine-A ( 3 5 y 0 ) , methoxyniethyltetrahydrostrychnidine (LXVIII ; R = CH3) (5y0), hydroxymethyltetrahydrostrychnidine (LXVIII; R = H) (30yo),desbase A (LXIX) (loyo),and desbase D (LXX) (15%)is obtained (78). The C-18 to N, bond is thus still by far the most easily broken one, but only by displacement and not by olefinforming elimination. I n the context of the Hofmann degradation, A refers to fission of C-20,Nb, D to fission of C-16,Nb, and v to fission cf C-13,Nh. Desbase A was originally given a structure with the double

1 7 . Strychnos >A,KALOIDS

LXXlX

609

LXVIII :a

,

LXXVIII Dihydrostrychnidine-D metho salt

LXXI Didesbase AD :np 73"

LXXIII Didesbase AD mp 113"

LXXV Desazastrychnidine (a)

Didesbaae DV

LXXVI Des~zaatrychnidine(b)

CHARTI. Hofmann degradation of dihydrostrychnidine-A.

610

G. F. SMITH

bond in the tetrasubstituted 14,21 position : recent work, however, shows that no migration of the double bond occurs (78a). Hofniann degradation of desbase A gives three products (79)"- : one produced by a niethoxylating fission, to which structure LXXI is ascribed, and two produced by D-type fission, didesbase AD, mp 73" (LXXII) and didesbase AD, mp 113' (LXXIII). Both these bases give the same tetrahydro derivative, hence they differ only in the position of the double bond. That V-type fission has not occurred is clear from the Kuhn-Roth oxidation of the tetrahydro derivative, which still shows the presence of only one C-methyl. Hofmann degradation of the monomethochloride of desbase D fails, and regeneration of the initial base occurs. Ring opening, however, occurs when the dimethochloride is used. The two products are didesbase AD, mp 73" (also produced from desbase A) and a new product, didesbase DV. This gives a negative result in the Kuhn-Roth oxidation, but its tetrahydro derivative gives acetic acid corresponding to one C-methyl group :these results make structure LXXIV very likely for didesbase DV. Hofmann degradation of didesbase AD, mp 113", and didesbase DV gives two products (79, 8 0 ) , desazastrychnidine-a and -b, for which structures LXXV and LXXVI have been proposed, but not confirmed. A similar series of Hofmann degradations has been achieved with dihydrobrucidine (81, 82, 83). It is noteworthy that, in all the many Hofmann degradations involving an N, quaternary metho salt, fission of ring I11 has never been observed. Several ring closures which, in effect, are reversals of the corresponding ring openings have been reported. The most important of these, which first led to the isolation of neostrychnine (LXXVII-A),was achieved by the treatment of methoxymethyldihydroneostrychnine with hot 20y0 aqueous sulfuric acid, which yielded neostrychnine metho salt with loss of methanol (33). Similarly, neostrychnidine metho salt is produced by heating methoxymethyltetrahydrostrychnidine with concentrated HC1 (33). Related to the foregoing is the thermal (230") ring closure of the hydriodide of methoxymethyltetrahydrostrychnidine to tetrahydrostrychnidine methiodide by loss of methanol (76). Another important type of ring closure involves the conversion of desbase D (LXX) into an approximately 2 : 1 mixture of a new ring system, dihydrostrychnidinium-D metho salt (LXXVIII) and the normal dihydrostrychnidinium-A metho salt (LXXIX) (84). This 2 Structures LXXI to LXXVI are those originally proposed in the 1930's. In none of them have the positions of the double bonds been rigorously proved.

17. Strychnos

611

ALKALOIDS

conversion must involve the simple nucleophilic addition of Nb to the protonated C-15,C-l6 double bond, for it occurs in glacial acetic acid a t the boiling point and not at all in the presence of strong mineral acid (84a). It was first observed to occur in acetic acid under hydrogenation conditions, when about 10% of 15,16-dihydrodesbase D is also produced (84, 84a). The interesting observation that desbase D (LXX), with a pK, of 5.1, is a much weaker base than closely related bases such as desbase A (LXIX, pK, = 6.95) and didesbase AD (pK, = 7.95) has been interpreted Br

08<,\o/ I

LXXX

/ LXXXI

1

Pt-Hz

Br \

LXXXIII

LXXXII

in terms of a transannular interaction between the unshared electrons of

N,, and the double bond. The resulting polarization of the C-15 to C-16 double bond is furthermore believed to be responsible for the difficulty encountered in its catalytic hydrogenation (84b). The von Braun degradation (85, 67) was reinvestigated and fully worked out by Boit (86). The reaction of strychnine with cyanogen bromide in hot benzene leads to the formation of two products, the amorphous bromocyanamide-I (LXXX) and the crystalline bromocyanamide-I1 (LXXXI). Dihydrostrychnine, on the other hand, gives only one product (LXXXII), which is identical with the product of catalytic hydrogenation (platinum in ethyl acetate) of bromocyanamide-11. The corresponding brucine compounds behave analogously.

612

G . F. SMITH

The structure of LXXXII was demonstrated by its conversion by zinc and acetic acid into tetrahydro- 18,19-chanostrychnine (LXXXIII ; R = H), methylation of which gave LXXXIII (R = CHs), which gave one equivalent of acetic acid on Kuhn-Roth oxidation and which was found not to be identical with 21,2%-dihydraXC (see Sectim I T , D, 1). The amorphous bromocyanamide-I was not further investigated.

D. REDUCTIONS 1. Catalytic Hydrogenation and Hydrogenolysis There are many examples of the straightforward hydrogenation of double bonds. Thus, hydrogen and palladium-charcoal in aqueous acetic acid reduce strychnine to 21122-dihydrostrychnine, and reduce tetrahydrostrychnine (LXXXIV)to hexahydrostrychnine (44). Strychnidine, strangely enough, was not easily reduced with this catalyst, and good results were obtained only when Adams at02 catalyst was used in glacial acetic acid (87). Perhydrogenations were observed when the PtO2 catalyst was wed in dilute hydrochloric acid a t 50"-60" : strychnidine gave octahydro(LXXXV) and decahydrostrychnidine (LXXXVI), and brucidine gave octahydrostrychnine, with hydrogenolysis of both methoxyl groups (88, 89). The formation of decahydrostrychnidine must involve hydrogenolysis of the allyl ether system. Methoxymethyldihydroneostrychnidine was another base that was perhydrogenated (89). Neostrychnine (LXVII-A) is more difficult to hydrogenate than strychnine and yields dihydrostrychnine : a tsmperature of 50" in aqueous acetic acid is necessary (33) (cf. electrolytic reduction). There are also many examples of the hydrogenolysis of the allyl ether system: 3% of the tetrahydrostrychnine (LXXXVII) may be isolated from the mother liquors of the hydrogenation of strychnine to dihydrostrychnine (90) ; brucine behaves similarly (91). N-Acetylisostrychnic acid gives a tetrahydro derivative (LXXXVIII) (92); N-nitrosoisostrychnic acid gives isotetrahydrostrychnine (LXXXIX) (48). Hydrogenolytic fission of the quaternary N,, allyl ammonium system occurs more or less readily, and has played a part in the structural problem. This is also effectively achieved by sodium amalgam (see later discussion). The simplest case is that of the N,,-methylstrychninium ion: with palladium as a catalyst, the acetate undergoes normal allylic hydrogenolysis to give XC as the main product, together with small quantities

17.

A%VJC~%OS

613

ALKALOIDS

of the 14,21 double-bond isomer and the two epimeric 21,22-dihydro derivatives (93); Adams catalyst, however, behaves differently in giving only 3% of allylic hydrogenolysis in water a t 70°, but 40% of the dihydro derivative of XC in methanolic ammonia, the remainder being 21,22dihydrostrychnine metho salt (86). Strychnidine methochloride likewise gives tetrahydromethylstrychnidine (XCI) (93). Likewise, 2,3-dioxonucidine methoperchlorate (XCII) was hydrogenated to give a mixture of the quaternary salt (XCIII) and the base (XCIV) (94). No hydrogenolysis occurred with 2,3-dioxodihydronucidinemetho salt.

~ H ~ O H LXXXIV

LXXXV

LXXXVI

LXXXVII

Hydrogenolysis of isostrychnine methochloride (XCV ; R = H) and isopseudostrychnine methochloride (XCV; F, = OH) leads to the same product (XCVI) (95,96).On the other hand, 0-methyl pseudostrychnine metho salt (XCVII; R = H) gives mainly the corresponding quaternary dihydro derivative, and only a small yield of base XCVIII (R = H) (97). Allylic hydrogenolysis, however, again predominates in the case of

614

0. F. SMITH

XCII

XCIII

XCIV

xcv

XCVI

XCVII

XCVIIl

XCIX

C

17. StryChnOs

ALKALOIDS

615

0-methyl-sec-pseudobrucine methoperchlorate (XCVII ; R = OCH3), which is converted into base XCVIII (R = OCH3) (98), The corresponding enol ether is converted into base XCIX (97). Hydrogenolysis of a C-Br bond occurs when bromodeoxyisostrychnine hydrobromide (XXXV; R = B r ) is shaken with Adams PtOa and hydrogen in glacial acetic acid, and the product is tetrahydrodeoxyisostrychnine (C) (56), probably also obtained by the action of red phosphorus and hydriodic acid on strychnine (99).

2. Sodium Amalgam This is the classical reagent for Emde reductions, and has been successfully applied in the strychnine field. It was first used on strychnine methosulfate, and gave a complex mixture of products. The main product is most probably GI, and is formed by double allylic hydrogenolysis. This then can undergo further transformations such as loss of water and migration of the double bond (100). The reduction of strychnidine methosulfate is much simpler and leads t o dihydromethylchanostrychnidine (CII), together with a base containing two hydrogens more, different from the dihydro derivative of CII to which structure C I I I was given (101). Dihydrostrychnidine methochloride is not reduced by sodium amalgam ( IOl), and neostrychnidine methosulfate was simply reduced tci strychnidine methosulfate with no ring fission (28). The greatest contrast between catalytic and sodium amalgam hydrogenolysis is shown by 0-methylpseudobrucine metho salt (XCVII ; R = OCH3) which is hydrogenolyzed catalytically to XCVIII (R = OCH3) (98), but which is converted by sodium amalgam into CIV (102).

Sodium amalgam has also been used in many instances to reduce carbonyl groups, as in the important conversion of strychninonic (XXXVIII) into strychninolic acid (XXXIX) (59). Conjugated double bonds have also been reduced, as in the conversion of Hanssen C19 acid (CXXVII) into its dihydro derivative by reduction of the maleic acid moeity (103), the conversion of 11-benzalstrychnine (LVII) into 11benzylstrychnine ( 104), and the reduction of isostrychnine-I1 (XXXVI; R = OH) to its 11,12-dihydro derivative (58).

3. Zinc and Acid This has effected the reduction of carbinolamine systems, as in the conversion of pseudostrychnine (XXV) into strychnine (23). An analogous simple reduction is that of neostrychnine (LXVII-A)to 21,22-dihydrostrychnine by zinc and hot dilute hydrochloric acid (177). A more

616

G. I?. SMITH

complex case is the reduction of CXLII (R = H) (see Section 11,E, 2) to the trio1 CV: this involves reduction of the carbonyl group, elimination of the ether oxygen /3 to the N, lactam carbonyl, and reduction of the a,p-unsaturated lactam (105). Of two “ Clemmensen reductions,” probably the only two in strychnine chemistry, neither turned out, in

CI

cv

CII

CVI

17. Strychnos ALKALOIDS

617

fact, to involve a conversion of CO into CH2. The first one is the important case of the reduction of methoxymethylchanodihydrostrychnone (XXVII) (106), the product of perbenzoic acid oxidation of methoxymethyldihydroneostrychnine (34). The product, methoxymethylchanodihydrostrychnone, was shown to have structure CVI (3), and to be probably formed by elimination of the /%oxygen to give CVII, followed by reduction of the a,fi-unsaturated ketone, hemiketal formation, and reduction to CVI. Simple reduction of XXVII to CVIII was achieved by formation of the diethyl mercaptal, followed by desulfurization with Raney nickel (3).

CIX

CXII

CXI

CX

CXIV

Zn-Hg

CXIII

As was indicated in the historical introduction, the most misleading aspect of the chemistry of XXVII is the complete resistance of the N-formyl group to alkaline hydrolysis, and the failure of concentrated hydrochloric acid to effect hydrolysis. The formyl group in XXVII and in CVI, CVIII, and the alcohol derived from XXVII by catalytic hydrogenation is, however, easily and quantitatively hydrolyzed by 2 N sulfuric acid (3). The failure of the hydrolysis under Clemmense? conditions may be ascribed to the low activity of water in the reaction mixture, and the failure of alkaline hydrolysis is most probably because of steric hindrance of approach by OH- ion, a view which is supported by a study of models (3). The second case of a “failed” Clemmensen reduction is that of the conversion of strychanone (CIX) into dihydroisostrychanol (CX),

a. F.

618

SMITH

presumably by way of isostrychanol (CXI), which is the carbinolamine tautomer of the amino ketone CXII (107). The key step here must be hydrogenolysis of the C-N bond in H +N--CC=O

which is the analog of the hydrogenolysis of a ketol in acid. The milder reducing agent, aluminum amalgam and dilute acid, leads to isostrychanol (CXI): sodium borohydride, an the other hand, reduces the carbonyl group simply to give the corresponding secondary alcohol, strychanol (CXIII) (108). Zinc dust and hydrobromic acid in glacial acetic acid have been used to debrominate XXXV (R = Br) and its 21,22-dihydro derivative to deoxyisostrychnine (CXIV) and its dihydro derivative (56).

4 . Electrolytic Reduction This technique was introduced by Tafel(7, 8) and was used frequently whenever i t was desired t o reduce the N, lactam carbonyl. It was about the only way of achieving this type of reduction before the discovery of lithium aluminum hydride. It was first applied t o the reduction of tetrahydrodeoxystrychnine t o the corresponding oxygen-free base, and then applied to the reduction of strychnine t o an approximately 2 : 1 mixture of strychnidine (CXV) and tetrahydrostrychnine (LXXXIV) (7 ; cf. 76, 87). Brucine behaves similarly, and strychnine methosulfate

cxv

CXVI

yields strychnidine methosulfate (76) with apparently no tetrahydrostrychnine methosulfate. Isostrychnine-I (XXXV ; R = OH) is reduced to the base CXVI (44). Isotetrahydrostrychnine (LXXXIX), the product of the catalytic hydrogenation of isostrychnic acid (XXXIII, seeSectionI1, B, l),isreducedsmoothlytobaseCXVI1 :now since isotetrahydrostrychnine is readily dehydrated and base CXVII is not, this fixed

17. Strychnos ALKALOIDS

619

the position of the hydroxyl on C-12 (48),and led t o the correct structure for isostrychnic acid. Brucinonic acid (2,3-dimethoxy XXXVIII) is reduced to two isomeric dihydrobrucidinonic acids (XLII), C-21 epimers, in which the N, lactam carbonyl, but not the Nblactam carbonyl, has been reduced; the ketone carbonyl is concurrently reduced to the epimeric alcohols (60, 109). It would seem that electrolytic reduction t o CH2 is a special property of the N, lactam carbonyl group.

CXVII

The other use to which electrolytic reduction was put was the reduction of the double bond in bhe neo series. The reducibility of the neo double bond under these conditions was, in fact, taken to be support for its enamine character. By this method, neostrychnidine was reduced a t 100" to dihydrostrychnidine-A and methoxymethyldihydroneostrychnidine (LXVII) a t 15" to methoxymethyltetrahydrostrychnidine (LXVIII; R = OCH3) (76). The conversion of CIX into CXI has also been carried out electrolytically (107).

5. Lithium Aluminum Hydride This reagent reduces strychnine smoothly to strychnidine (1lo), dihydrobrucine to dihydrobrucidine, and the two colubrines to the corresponding colubridines (111). Brucine, however, behaves differently and yields 10,ll-dehydrobrucidine (CXVIII) by only partial reduction. The reason for this anomalous behavior may well be that the salt produced by the initial attack by LiAlH4 is so insoluble that it separates out of solution and becomes immune to further reduction (111). 10,ll-Dehydrobrucidine (CXVIII) readily and reversibly adds methanol to form the carbinolamine ether (CXIX),and is hydrogenated t o brucidine (111). I n this connection, it may be mentioned that tetrahydrostrychnine (LXXXIV), formed in the electrolytic reduction of strychnine, must arise from the intermediate carbinolamine (10-hydroxystrychnidine, CXX) which is in equilibrium with the open-chain amino-aldehyde, the aldehyde group of which is then reduced to primary alcohol.

620

G . F. SMITH

Other cases of LiAlH4 reduction are the normal conversion of N methyl-see-pseudostrychnine (CCVI) into the base CXXI (112) to be contrasted with the abnormal behavior of N-methylisopseudostrychnine

CXVIII

CXIX

cxx

CXXI

CXXIV

(CXXII), which gave CXXIII; that is, no reduction of the C-16 ketone carbonyl and only partial reduction of the lactam carbonyl. N-Methylisopseudostrychnidine (CXXIV),on the other hand, smoothly gave the C-16 secondary alcohol (112).

17. Strychnos

62 1

ALKALOIDS

E. OXIDATIONS

1. Chromic Acid Chromic acid in aqueous sulfuric acid attacks the benzene ring in a large number of strychnine and brucine derivatives, leading to the isolation of products in which the aromatic ring has been oxidized away. Doubtless, as the low yields indicate, attack of the other, nonaromatic, portion of the molecules also occurs, but this is presumably so indiscriminate that the isolation of homogeneous products has so far failed. The following is a selection of the more important and interesting chromic acid oxidations. Both strychnine (9, 113) and brucine (9, 114, 115) are oxidized a t 50"-85" to C17 Wieland acid (2,3 dioxonucine dihydrate) (CXXV) and CIS Hanssen acid (nucinecarboxylic acid) (CXXVI). The yields from brucine (about l2-18% in each case) are much better than those from strychnine (about 5% of the C17 acid and 1 % of the C16 acid), as might be expected. These two acids are also produced by oxidation of a wide range of compounds differing only in the benzene ring area, for example 2 aminostrychnine (42) and C19 Hanssen acid (CXXVII) (113). The

I

CO2H

cxxv

CXXVI

CXXVII

CXXVIII

CXXIX

cxxx

oxidation of strychnidine proceeds similarly, t o give 2,3 dioxonucidine (CXXVIII) in about 20% yield ( 1 15). Leuchs and Wegener introduced the nucine nomenclature ( 116) to facilitate the naming of these oxidation products: they defined nucine as CXXIX (R = 0),nucidine as CXXIX ( R = HZ), and aponucine as CXXX.

a.

622

F. SMITH

The strychnine and brucine sulfonic acids, which probably contain a sulfonic acid group on a carbon a to Nb,are oxidized to products which correspond to sulfonic acids of the C16 and C17 acids ( 1 17, 118). Further examples are the oxidation of CXXXI (R = H or OCH3) to CXXXII in 30-40% yield (119), and of brucinonic acid (2,3 dimethoxy XXXVIII) to two C13 amino acids (CXXXIII and CXXXIV) in the extremely poor yield of about 0.5% (22, 120). This reaction also involves the oxidative breakdown of the Nb a-ketoamide ring.

CXXXI

CXXXIII

CXXXVII

CXXXIV

CXXXVIII

cxxxv

CXXXIX

The preceding oxidations reveal the resistance of the 21,22 double bond and of the environs of Nb to oxidative attack by chromic acid. Under milder conditions, brucine ( 12 1)and the corresponding dihydric phenol, bisapomethylbrucine (122), are oxidized to bruciquinone (CXXXV). Of the several instances of the simple oxidation of an alcoholic grouping, that of strychninolone-a (XL) to strychninone (CXXXVI) is

17. Strychnos ALKALOIDS

623

typical (123).Only one other need be mentioned, and that is the oxidation of hydroxymethylchanotetrahydrostrychnine(CXXXVII) to the corresponding aldehyde, which exists as a tautomeric mixture [CXXXVIII and CXXXIX (see Section 11,F)] (124).

2. Permanganate This oxidizing agent was used very extensively and played the leading role in the establishment of the peripheral sequence -CHI I +Na-CHz-CH-O-CHz-CHd-CH2-Nb-

I

In practically all the instances of its use, attack of a double bond and of C-20 (between Nb and the C-21 to C-22 double bond) occurs. There is no instance of the simple oxidative disruption of a benzene nucleus; indeed, the most vigorous conditions, aqueous alkaline permanganate at 50"-70") lead to the oxidation of strychnine to N-oxalylanthranilic acid in about 8% yield. Brucine likewise gives 4,5-dimethoxy-N-oxalylanthranilic acid, but in much lower yield (0.1 to 0.2%) (18). a-Colubrine (3-methoxystrychnine) gave a 2.7 yoyield of 4-methoxy- and /3-colubrine (2-methoxystrychnine) a 5.5% yield of 5-methoxy-N-oxalylanthranilic acid (125). These oxidations fixed the positions of the methoxyl groups in the last three alkaloids. The mildest conditions hydroxylate the C-21 to C-22 double bond and leave the Nb methylenes unchanged : thus, strychnine, with permanganate in neutral acetone, gives some of the diol CXL'( 126), and 18-0x0strychnine gives the diol CXLI (127). The next stage of the oxidation of strychnine, the ketol CXLII (R = H) is obtained in 28% yield by working in dilute aqueous mineral acid; under these conditions, pseudostrychnine (XXV) behave similarly, to give the ketol CXLII (R = OH) (105). 0-Methylpseudostrychnine (XXV with OCH3 instead of OH) with permanganate in acetone yields compounds CXLIII (128) and CXLIV (128, 129). Compound CXLIII may be oxidized to CXLIV by alkaline ferricyanide (128). Similar oxidation occurs when the benzene ring is absent, as in dihydro-Clg-Hanssen acid (CXLVI), which is oxidized in aqueous potassium bicarbonate at 0" to CXLVII (130). The oxygenation at C-20, C-21, and C-22 in compound CXLIV is of particular interest, because it corresponds to that of the immediate precursor of the key compound in the formation of dihydrostrychninonic acid (XXXVII): this key compound, CXLV, has never been isolated, presumably because of the ease with which it hydrolyzes t o dihydrostrychninonic acid (see Section I).

624

G. F. SMITH

CXL

CXLII

CXLI

CXLIII

CXLIV

CXLV

CXLVI

CXLVII

H CXLVIII

17. Strychnos

625

ALKALOIDS

Oxidations of strychnine straight through to mixtures of strychninonic acid (CXLVIII; R = H) and dihydrostrychninonic acid (XXXVII) and the parallel oxidation of brucine have, of course, been worked out in great detail: in the oxidation of strychnine, the best yield is 24% of a 2 5 : 1 mixture of strychninonic and dihydrostrychninonic (XLVIII) acids (131).About 1%of 12-hydroxydihydrostrychninolone-a

AcO/’

AGO/‘

CXLIX

CL

CLIV

CLI

CLIII

CLII

is isolated from the nonacidic fraction. Strychnine N-oxide also gives strychninonic acid as a main product (1321, although more recently, conditions for the oxidation ofstrychnine N-oxide to as much as 37% of dihydrostrychninonic acid and only small amounts of strychninonic acid have been described (132a): parallel results were obtained with brucine N-oxide. Analogous oxidations occur with pseudostrychnine and 0-methylpseudostrychnine, which give CXLVIII (R = O H or OCH3,

626

G . F. SMITH

respectively) (128) : these oxidations were important in establishing the position of the hydroxyl group in pseudostrychnine. 0,N-Diacetyltetrahydrostrychnine (CXLIX) yields CL (133). Of very special interest is the oxidation of N-acetyl-sec-pseudostrychnine (CLI)to the keto acid CLII (134).That Nbis already acylated

OH CLVIII

CLVII

bOzH CLIX

CLX

inhibits oxidation of C-20. The interesting feature of this reaction is that there is no reason why intermediate ketol CLIII should hydrolyze to CLIV, since there is no C-20 carbonyl to stabilize a negative charge on C-21 (cf. XLV). One must therefore suppose that the ring fission does not proceed hydrolytically, that is, not via CLIV, but by straightforward oxidative cleavage leading directly from CLIII to CLII. It is noteworthy that no successful oxidation of strychnidine or brucidine has been reported ; dihydrostrychnidine has, however, been

17. Strychnos

627

ALKALOIDS

oxidized in hot acetone to an Nb-oxo derivative in about 15% yield. The position of the lactam carbonyl has not been fixed ( 6 9 ) . The other structurally significant oxidations were carried out on the strychninolones and brucinolones. Here oxidation resulted in simple fission of the C-11 to C-12 or C-12 to C-13 double bonds, and enabled H. Leuchs to determine the environment of these bonds and their relation

CLXI

CLXII

to N, (see Section I ) . Acetyl strychninolone-a (CLV) is oxidized in acetone at 10" in 30-40y0 yield to acetylstrychninolone-a acid (CLVI) together with about 1yo of 11,12-dihydroxydihydrostrychninolone-a (CLVII) (14).Similar oxidations were carried out on acetylbrucinolone-a (14) and acetylstrychninolone-c (74). Acetylbrucinolone-b (CLVIII) was oxidized to about 5% of the 12,13-dihydroxy derivative and to about 10% of the acid CLIX, hydrolysis of which gave curbine (CLX), which rather surprisingly does not seem to have been investigated further (13).

628

G . F. SMITH

The corresponding degradation of strychninolone-b has not been described, probably because of the failure to obtain crystalline products. Dihydrodeoxyisostrychnine-I(CLXI) is oxidized in acetone a t 0" to nordihydrofluorocurarin (2.5%) (CLXIII) and, after esterification, t o dihydroakuammicine (3.5%) (CLXIV), almost certainly by way of the cr-pyridone CLXII, which was also isolated, albeit in the very low yield of about 0.2%. If the total oxidation product is treated with hot mineral acid, deformylation and decarboxylation occur, and a 10-15% yield of the base CLXV can be isolated (135, 136). Neostrychnine (CXVII-A) has been oxidized to strychnone (CLXVI ; R = 0) (137) and neostrychnidine to strychnidone (CLXVI; R = H2) (76). Methoxymethylchanodihydroneobrucidine(cf. LXVII) is oxidized in acetone a t - 10" to a mixture of probably stereoisomeric dioxy compounds CLXVII (81).

Pd\O CLXVIII

CLXVII

CLXIX

Also of interest is the oxidation of benzylidenedihydrostrychnine to a small yield of the ketol CLXVIII and a moderately good yield of the lactone acid CLXIX. The analogous products have been obtained in the brucine series (138, 139). Last, an unusual oxidation occurs when methoxymethyltetrahydrostrychnidine-B is oxidized in moist acetone a t room temperature : the N,-methyl group is oxidized to formyl. The structure of dihydrostrychnidine-B is not known. It is produced by the reduction of strychnidine-A with concentrated hydriodic acid and red phosphorus (69).

17. Xtrychnos ALKALOIDS

629

3. Hydrogen Peroxide a n d Peroxidic Reagents This reagent has been used very frequently, either with barium hydroxide or with formic acid, t o oxidize a-keto acids as either acids or lactams : thus, strychninonic acid (XXXVIII) is converted by H ~ 0 2 / Ba(0H)z into CLXX, in this case with concurrent base-catalyzed elimination of the glycolic acid residue (140).Very similar is the oxidation of dihydrostrychninone (CLXXI)in formic acid and hydrogen peroxide

$... ::I---b

0- F ' P O/

'.'H CLXXII

CLXXIV

'0

'-H

CLXXIII

CLXXV

to the amino acid CLXXII, which lactamizes on sublimation to CLXXIII (141) (see Section I). The a-ketocarboxyl group in Wieland acid (CXXV) is cleaved with loss of COz to give c16 Hannsen acid (CXXVI): the same result is achieved with chromic acid at 85" ( 1 15). Hydrogen peroxide on its own converts basic Nb into the N-oxide; thus, strychnine may be converted in high yield into strychnine N-oxide (142). Brucine behaves similarly but, under much more vigorous conditions (hydrogen peroxide and formic acid at SO'), oxidation of the benzene ring occurs, and a 48% yield of Hannsen acid is obtained, a, much higher yield than is obtained with CrOs (143).

630

G . F. SMITH

A rearrangement of strychnine-N-oxide catalyzed by sodium dichromate in boiling water gives almost quantitative conversion into two products in equal amounts : one of these is pseudostrychnine (XXV) and the other is 18-oxostrychnine. It is interesting that no product corresponding to oxidation of C-20 seems to be formed (143a, 127).

CLXXVI

CLXXVII

0

CLXXVIII

0 II

CLXXIX

CLXXX

Peroxidic reagents may dehydrogenate C-8-N,. The example of this involves the conversion of isostrychnic acid (XXXIII) by hydrogen peroxide in formic OT acetic acids in the presence of catalytic quantity of cobalt salt or by potassium nitrosodisulfonate, into the lactone bases CLXXIV (R = H and OH) (144, 145). I n both cases, the initially formed simple 3-H indole is oxidized further, probably by way of the tautomeric enamine, to the 13-oxy derivative, which then lactonizes. A reaction which was of great importance in determining the relation of Nb to the /3-indoline carbon (C-7) was the oxidation of pseudostrychnine (XXV) by acidic hydrogen peroxide (146) to strychnone (XXIV)

t

T

3. 3.

t t

I

63 1

632

Q. F. SMITH

(30). This reaction was regarded by Woodward et al. (30) as involving a Baeyer-Villiger-type oxidation of the C- 16 carbonyl present in the pseudostrychnine tautomer, leading to the intermediate CLXXVI, which then by elimination and lactonization gives strychnone (XXIV). The simpler alternative interpretation of this reaction draws an analogy with the oxidation of isostrychnic acid and with the tautomeric equilibrium, CLXXIV + CLXXV (145) : dehydrogenation a t C-8 would lead to an intermediate CLXXVII, which by the straightforward movement of electrons shown, would lead directly to strychnone. Closely related to the foregoing oxidations, but with cupric sulfate or mercuric acetate as oxidizing agents, is the conversion of strychanone (CIX) into 3-oxo-d4,16-dehydrostrychindole(CLXXVII-A), probably by way of the intermediates shown (37a). This opened the way to the determination of the absolute configuration of strychnine by a chemical method: a number of simple reactions converted CLXXVII-A into CLXXVII-B, which was then synthesized with diethylsuccinic acid of known absolute configuration as starting material (146a). This synthesis is closely related to that of 2 -dihydrocorynantheane by van Tamelen et al. (147). Perbenzoic acid converts strychnine and dihydrostrychnine into the corresponding N-oxides, with no attack on the C-21 to C-22 double bond a t all (137). On the other hand, when the Nb is nonbasic, as in 18-oxostrychnine, reaction occurs to give the C-21 to C-22 epoxide (CLXXVIII) (127). It is interesting to note that this epoxide is so sterically hindered to nucleophilic attack, that prolonged boiling with dilute sulfuric acid does not affect it. Methoxymethyldihydroneostrychnine (XXVI) is another molecule in which Nb is, if not nonbasic, a t least very weakly basic : in this case, too, perbenzoic acid cleaves the double bond t o give methoxymethylchanodihydrostrychnone (XXVII) (34). 4. Ozone

This reagent has not been of any value in strychnine work. With strychnine itself, the only isolable product is pseudostrychnine, and no product corresponding to double-bond fission can be isolated (148).With neutral Nb in 18-oxostrychnine, on the other hand, specific attack ofthe double bond occurs to give the epoxide CLXXVIII, also obtained by perbenzoic acid oxidation, and the lactone CLXXIX : the production of this lactone, which involves the loss of C-20, C-22, and (2-23, is very anomalous, and a mechanism for its formation has been suggested (127). I n two instances, in the case of CLXXI and of CLXXIII, ozone has been used t o destroy the benzene ring: subsequent treatment with

17. Strychnos ALKALOIDS

633

alkaline peroxide to oxidize u-ketocarboxylic acid groups led to the isolation of the amino acid CLXXX (141). The only straightforward double-bond ozonolysis was achieved with great experimental skill in the case of dihydrodeoxyisostrychnine-I (CLXI). The reaction was carried out in methanol at -7O", and two molar equivalents of ozone were used. Very careful work-up, which included an acid hydrolysis, led t o a 50% yield of strychanone (CIX) (108).

5 . Miscellaneous Oxidizing Agents Osmium tetroxide converts strychnine into the 21,22-diol (CXL) in unspecified yield (149). The sodium chlorate-osmium tetroxide reagent has been used to oxidize dihydrodeoxyisostrychnine-I(CLXI) and its %nitro derivative to the corresponding 12,13-glycols: the yield in the case of the nitro derivative, 45%, is quite high (132a).

cLxxxIII

CLXXXIV

Lead tetraacetate has cleaved three glycols : dihydrostrychnine-2 1,2% diol (CXL) is converted into the keto-aldehyde CLXXXI, further oxidation of which with permanganate gives strychninonic acid (149); the Nb lactam diol CXLIV is likewise oxidized to the corresponding keto-aldehyde (150); the glycol CLXXXII, but not the corresponding lactam, has been converted into 2-nitrostrychanone (cf. CIX) in 25 yo yield (132a). Periodate has not been used very successfully (132a, 127). Bromine and hydrobromic acid a t loo", extraordinarily, seem to cleave the 21,22-double bond in compounds in which the benzene ring

634

G . F. SMITH

has been oxidized away. Typical of such oxidations is the conversion of C19 Hannsen acid (CXXVII) into an aldehyde by the addition of two oxygens : the product has been given structure CLXXXIII, the result of an internal aldol condensation of the intermediate keto-aldehyde ( 151, 152).An analogous oxidation is that of 3-oxy-2-oxonucidine (CLXXXIV)

CLXXXV

CLXXXVI

to the aldehyde CLXXXV in about 60% yield ( 1 16). When Nb is quaternary, as in the methobromide of CXXVII, the reaction does not go as far, and the two products isolated are the bromohydrin CLXXXVI and the corresponding vinyl bromide formed by dehydration (152). The interpretation of these reactions is not fully proved, and a reinvestigation using modern spectroscopic techniques would be of considerable interest.

\

AB' I -

- /

CLXXXVII

cxc

cLxxxvIII

NH

CHO

CLXXXIX

LXVII

Superficially similar, but mechanistically almost certainly different, is the conversion of neostrychnine (LXVII-A) into the aldehyde CLXXXVII by bromine in aqueous sulfuric acid at room temperature (32,153): the reaction most probably runs by way of the " bromohydrin " CLXXXVIII and the tautomeric aldehyde CLXXXIX, which then cyclizes by nucleophilic displacement of bromide by the NbH. Quite as

1 7 . Strychnos

635

ALKALOIDS

interesting is the regeneration of neostrychnine by acid treatment of the primary alcohol CXC produced by reduction of the aldehyde CLXXXVII : this must involve a Wagner-Meerwein type of rearrangement in which neutral nitrogen migrates to a primary carbonium atom. Bromine and HBr have also been used to break down the quinonoid ring I in cacotheline (CXCI) to give C19 Hannsen acid (CXXVII) in about 30% yield (154, 103): the other products are bromopicrin and COz, and the reaction probably proceeds by initial bromonium attack a t C-4, addition of OH a t (2-5, ring fission to CXCII, further bromination of C-4 to bromopicrin, and oxidation fission of the C-2 :C-3 a-ketoacid system. If excess bromine is used, then further oxidation a t C-21 to C-22 occurs (see above).

6OzH CXCI

CXCII

CXCIV

CXCIII

cxcv

N-Bromosuccinimide in CC14 dehydrogenates acetylstrychninolones-a, -b, and -c to the same acetyldehydrostrychninolone(CXCIII), which has a UV-spectrum very similar to that of N-phenyl a-pyridone. Catalytic hydrogenation yields dihydrostrychninolone-c, showing that no skeletal rearrangement has occurred a t the oxidation stage. This was used as proof of the six-membered nature of ring 111 (155). Mercuric acetate in hot acetic acid is very effective in the conversion of acetylstrychninolone-b into CXCIII, but does not dehydrogenate strychninolone-a acetate (156). Nitrous acid has been used to effect yet another fission of the enamine double bond of neostrychnine (LXVII)to give an Na-formyl-21-oximino compound (see reaction sequence, Section I) (32): this reaction is related to the oxidative nitrosating ring fission of ketones, a well-known example of which is the conversion of CXCIV into CXCV ( 1 5 7 ) .

636

C. F. SMITH

Vigorous nitric acid oxidation of strychnine gives 5,7-dinitroindole2,3-dicarboxylic acid (XV) (158). Its only other application as an oxidizing agent has been in the partial oxidation of the aromatic ring i n brucine to give bruciquinone (CXXXV) and its 4-nitro derivative a t - 5" (121) and t o give cacotheline (CXCI) a t 50"-60' (159). Oxygen in the presence of cupric salts provides yet another convenient method for the oxidation of strychnine to pseudostrychnine (XXV) and strychnone (XXIV) (31, 160). Manganese dioxide interacts with a hot aqueous solution of strychnine sulfite in the presence of excess sulfur dioxide to give a mixture of three sulfonic acids (43, 161) and brucine behaves likewise (143a). The position of the sulfonic acid group in these compounds is not known : the benzene positions are excluded because the ring can be oxidized with chromic acid to give acids which correspond in composition with sulfonated and Hannsen C16 acids. Likewise, position 11 is excluded Wieland because these compounds form benzylidene derivatives (143a).The most likely positions for the sulfonic acid residue are on the carbons adjacent to Nb-their formation would then involve dehydrogenation followed by bisulfite addition : more experimental work is called for.

F. PSEUDOSTRYCHNINE (XXV) AND

N-METHYL-SEC-PSEUDOSTRYCHNINE (CCVI, CCVII) The reactions of pseudostrychnine and of its derivatives which involve the CWN, carbinolamine grouping are of considerable interest ( 162). IR-spectra of pseudostrychnine in the solid phase and in solution indicate that it exists almost entirely in the carbinolamine form CXCVI. Two other forms might be expected t o exist in equilibrium with CXCVI in solution : the immonium form CXCVII, produced by proton-catalyzed loss of the hydroxyl group, is totally excluded on steric grounds; the keto-amine form CXVIII, on the other hand, almost certainly exists in equilibrium with CXCVI in solution, but in concentrations which escape detection by physical methods. Its presence is deduced from some of the chemical reactions of pseudostrychnine. The easy formation of 0-alkyl ethers (125) by interaction with methanol or ethanol even at room temperature almost certainly involves nucleophilic attack of the alcohol molecule on the carbonyl carbon in CXCVIII or in the 0-protonated form CXCIX (R = H) to yield a hemiacetal as a first step; likewise, the formation of 16-cyanostrychnine (XXV; CN instead of OH) must involve addition of cyanide ion t o CXCVIII, and the reaction with

17. Strychnos

637

ALKALOIDS

cyanoacetic acid to give 16-cyanomethylstrychnine (XXV ; .CH,CN instead of OH) almost certainly proceeds analogously (163). On the other hand, reactions leading to substitution on Nb are best understood by visualizing direct electrophilic addition t o CXCVI to give CC. Nitrosation and acetylation thus proceed by the steps CC + CCI or CXCIX + CCII (R = NO or COCH3, respectively) : the acetylation product then has structure CLI (134). Reaction with methyl iodide proceeds by CC -+CCI -+ CCIII, and the main product is thus CCVIII (102). I i 0 4 :NH I I

1 Ie CH30-G-N-CHs I 1

CXCVIII

CCIII

T I I HO--CN: 1 1

+

I I@ HO--CN-R

I

I

I I@ 0-C-N-R I 1 0

CXCVI

cc

CCI

It

li

Jt

e l l HO=C :N-R

I 1 C=N@ I I

I

CXCVII

I

CXCIX

CCIV

I 1 0 4 :N-R I I CCII

ccv

Whereas there is no difficulty in writing a structure for pseudostrychnine, N-methyl-sec-pseudostrychninepresents certain complications. Basically, the structure is CCVI: the problem consists in defining the

638

G . F. SMITH

degree of interaction between the tertiary nitrogen and the carbonyl groups. When two such groups are held very closely together across a medium-sized ring, then each is considerably affected by the other. The first cases to be discussed were vomicine (CCXXXVII) (164), cryptopine (CCIV), and N-methyl-sec-pseudostrychnine (CCVI) ( 165, 166). It was observed that the availability of the unshared electrons on the nitrogen was reduced, this being reflected in the relatively weakly basic nature of these compounds (pK, of N-methyl-see-pseudostrychnineis 6. l ) , and that the carbonyl-stretching frequency was lowered (5.97 p, or 1675 cm-1 for cryptopine and 6.02 p, 1660 em-1 for N-methyl-sec-pseudostrychnidine). I n these first discussions of the problem, the analogy with amide mesomerism was drawn, and the situation was described by the partial structure (165) I O=C 6- I

0 1

:NI r3a

and more precisely (164) as a case of mesomerism involving the two structures CCVI and CCVII as canonical forms. The carbonyl group in such compounds fails to react with carbonyl reagents. When, however, the nitrogen no longer has unshared electrons, as in CCXI, then in spite of the close spatial relation, the carbonyl group reacts with hydroxylamine to form an oxime (98). The carbonyl group is also normally ketonic when the nitrogen is no longer permanently held very close by being part of a medium-sized ring : base CCV is an example of this ( 1 12). The relatively simple view that N-methyl-sec-pseudostrychnine is a resonance hybrid of CCVI and CCVII is questioned by Hendrickson (162), who argues that the stereochemistry of the two structures is probably too different to allow any appreciable resonance to occur: C-16 in CCVI is trigonal and in CCVII is tetrahedral, and the distance between C- 16 and Nb in CCVI is much greater than in CCVII. A situation closely resembling tautomerism is proposed, where the molecule can exist in two discrete forms, one being a slightly polarized modification of CCVI, in which the carbonyl group still exists, but is no longer purely ketonic in character, and the other is more or less equivalent to CCVII. The main support for this comes from a study of the aldehyde CXXXVIII (124) : crystallization of this compound from water gives a lower-melting form, the solid state IR-spectrum of which shows an almost normal aldehyde G O band a t 1710 em-1 (5.84p ) ; crystallization from benzene gives a higher melting form showing no carbonyl absorption

17. Strychnos

CCVI

639

ALKALOIDS

CCVII

640

G . F. SMITH

a t all ;the solution spectrum shows the presence of a mixture of the two forms. This is most simply interpreted in terms of an equilibrium in solution between the carbonyl and zwitterionic forms (CXXXVIII + CXXXIX), and the actual isolation in the crystalline state of each tautomer as a homogeneous entity (124). The carbonyl group of the aldehyde can swing well away from Nb, and this probably accounts for the almost normal carbonyl frequency in the crystalline low-melting tautomer. It has not been possible so far to isolate two forms of N-methyl-seepseudostrychnine. However, the IR-spectrum of the solid shows it to contain a modified carbonyl group, hence the structure of the molecule in the crystal must approximate tc; CCVI. In solution, the carbonyl absorption disappears rapidly and completely, which means that when surrounded by solvent, and presumably solvating, molecules, the zwitterionic tautomer becomes the stabler one (162).It is this zwitterionic form which is protonated on oxygen to give salts of type CC, and is likewise methylated to give the 0-methylquaternary salt CCVIII. The zwitterionic form CCVII is also believed to be responsible for the allylic hydrogenolysis of N-methyl-see-isopseudostrychnine(CXXII) to XCVI (95) and for the failure of CXXII to be reduced a t (3-16 by lithium aluminum hydride. Until relatively recently, structure CCXI was erroneously assumed to be that of the product of methylation of N-methyl-see-pseudostrychnine. This necessitated the postulation of a series of migrations of alkyl from oxygen to nitrogen and back in order to account for various reactions of this salt ( 7 2 , 23). Boit and Paul have now clearly shown the structure of this quaternary salt to be CdVIII by the synthesis of structure CCXI by an unambiguous route (98). The action of sodium methoxide on the quaternary salt CCVIII had long been known to give a tertiary base to which the correct structure CCIX had been assigned (23) (this on the old view necessitated migration of methyl from nitrogen t o oxygen). Treatment of this base with acid regenerated the original quaternary salt CCVIII (on the old view, migration of methyl from oxygen back to nitrogen). The base, however, reacts with methyl iodide to give the quaternary salt CCX, acid hydrolysis of which gives the quaternary salt CCXI which contains a carbonyl group (IR) reacting normally with hydroxylamine and which is not converted into a tertiary base by sodium methoxide (98). The complete independence of Nb and C-16 in the quaternary salt CCXI is nicely demonstrated by its hydrogenolysis and hydrogenation to base CCXII: this is to be contrasted with the hydrogenolysis of CCVIII to XC (98).

17. Strychnos

641

ALKALOIDS

G. HIGH-TEMPERATURE DEGRADATION OF STRYCHNINE, BRUCINE, AND DERIVATIVES Almost all these degradations were carried out with alkaline reagents at temperatures around 300". Distillation of strychnine from lime gave skatole and 8-picoline (167). Slaked lime on brucine gave about 2% of mixed P-picoline and p-ethylpyridine (168). Soda lime gave 0.5% of carbazole from strychnine, as well as skatole and P-picoline (169). It is interesting to note that pyrolysis of strychnine with moist potassium hydroxide gave indole, not skatole (170). Indole (and carbazole) was also obtained by the KOH decomposition of methylstrychnine (LXIV) (76). On the other hand, other workers using the same reagent but different pyrolysis conditions obtained tryptamine from strychnine (171, 271, pseudostrychnine (1481, and strychninonic acid and strychninolone (171). This a t the time was of very considerable structural significance (see Section I). N-Methyl-see-pseudostrychnine gave N-methyltryptamine ( 148), and ethoxymethyl dihydroneostrychnine (LXV with EtO instead of CHsO) gave tryptophol ethyl ether (172). Other products of the KOH decomposition of strychnine were 3-ethylindole (27), 4-methyl-3-ethylpyridine (27, 173), and base CloHllN of as yet unknown structure : this latter base does not appear to be P-ethylpyrrocoline (CCXIII) (174).

CCXIII

CCXIV

CCXV

A relatively mild dehydrogenation not involving any skeletal change is the conversion of carboxyaponucidine (CCXIV), the product of hydrogen peroxide oxidation of nucidine, into base CCXV by palladium black a t 290" (175). H. NEOSTRYCHNINE (LXVII-A) This compound was first obtained by the pyrolysis of methylneostrychnidinium chloride (see Section 11, C). This method of preparation was rather tedious: however, the discovery of the much more direct method of isomerization of strychnine with Raney nickel in boiling

a.

642

F. SMITH

xylene has made neostrychnine a relatively readily available compound (176). Other reactions which lead to neostrychnine are the action of Raney nickel on dihydropseudostrychnine methyl ether (177) and the action of selenium on strychnine at 250" (178). Neostrychnine is an extremely weak base with a pK, of 3.8. Enamines are usually more strongly basic, for they protonate on the /I-carbon to give immonium salts, - C =G N -

I

+ -CH-C=N-

83

I

Neostrychnine, however, cannot protonate on carbon, because the resulting immonium system would carry a bridgehead double bond (166). The reactivity of the neodouble bond (see Sections I, and 11,E , 5), however, indicates that the nitrogen still can exert a polarizing effect. This effect may still be mesomeric (166) or mainly inductive (162), although in the latter case one would think that the nitrogen would attract rather than release negative charge.

111. a- and p-Colubrines These two alkaloids were first isolated from the mother liquors of strychnine manufacture from Strychnos nux vomica seeds (125). That a-colubrine is 3-methoxystrychnine and /I-colubrine is 2-methoxystrychnine follows from permanganate oxidation t o the corresponding methoxyoxalylanthranilic acids (12 5 ) and from chromic acid oxidation of the two colubridines to diketonucidine (CXXVIII) (111).

IV. The Total Synthesis of Strychnine The total synthesis was achieved in 1954 by Woodward and his collaborators (156). Before this, various unsuccessful attempts had been made to make a start toward a synthesis of strychnine (179, 180, 181, 182,) but these are now of little interest, with the exception of Robinson's idea ( 1 82) to emulate the postulated biosynthesis by attempting to synthesize the dialdehyde CCXVI, which then might be induced t o cyclize by a combination of Mannich and aldol type condensations t o the Wieland-Gumlich aldehyde (LV): the synthesis of CCXVI unfortunately was not realized. Much more recently, however, this idea has been used by van Tamelen et al. (184),who successfully synthesized the dialdehyde CCXVII and converted it in aqueous acetic acid-sodium

17. 8trychnos

643

ALKALOIDS

acetate solution into an uncharacterized aldehyde, borohydride reduction of which yielded a crystalline alcohol claimed to have structure CCXVIII (R = 0). Lithium aluminum hydride then yielded a noncrystalline alcohol (characterized as the crystalline cyclic urethane) to which structure CCXVIII (R = H2) with undefined stereochemistry was assigned. Further developments along these lines will be of considerable interest.

OCH3 CCXVIII-A

CCXVI

N/\

NH

-

\/ N

\/\

CH3

\/‘OH I OH \/”OH I OH CCXIX

CCXX

A reaction between N-tosyloxindolylethylamine and 3,4-dimethoxyphenylacetaldehyde, designed to simulate a possible biosynthetic step, led, under very mild conditions, to the linking up of the indoline /3-carbon, C-16, and Nb: a subsequent POC13 ring closure gave CCXVIII-A (185).

a. F.

644

SMITH

Another recent observation bearing on possible simplified strychnine syntheses is the rearrangement of CCXIX in boiling concentrated hydrochloric acid to CCXX (186). Woodward’s synthesis of strychnine stands out as a major synthetic achievement (156).The plan of this synthesis was also strongly influenced by biogenetic considerations, but no short cuts were taken, and the synthesis proceeds step by step, a t all stages fully under control. The synthesis begins with a Fischer reaction which leads t o 2-veratrylindole (CCXXI).A tryptamine side chain is next built on by conventional methods by way of the Mannich reaction with formaldehyde and dimethylamine to CCXXII, the methiodide of which with sodium cyanide gives the acetonitrile, reduced to 2-veratryltryptamine (CCXXIII). Now the real work of the synthesis begins. An attempted direct formation of ring V by a “biogenetic type” nilannich reaction between the try-ptamine w-nitrogen and the indole 13-position with ethyl glyoxylate yielded only the Schiff base CCXXIV : treatment of this with toluenesulfonyl chloride and pyridine, however, close the ring to give the 3H-indole CCXXV. Reduction with sodium borohydride followed by acetylation gave the acetylindoline CCXXVI. CHeO

___f HN(CHs)s

o

\

-/CHZN(CH~)Z ~H ) v C H /

CCXXI

CCXXII

OCN

3

OCH3

__f

IIa

17. Strychnos ALKALOIDS

645

0 -

CCXXVII

CCXXVIII

0 -

0 -

ccxxx

CCXXIX

I

0 CCXXXI

CCXXXII

-m;zz 0’

CCXXXIII

CCXXXIV

n

ccxxxv CHART 11. The synthesis of strychnine (162).

646

a. F. SMITH

The next step, the ozonolysis of the veratrole ring to give CCXXVII is extraordinary and is probably the most original in the whole synthesis. It was perhaps conceived in parallel with the then extremely fruitful, if now no longer widely accepted, biogenetic notion of the Woodward fission. The fission in the synthesis of course plays a completely different structural role. The yield in this step is 29%. The action of acidic methanol on CCXXVII yields the pyridone CCXVIII. This pyridone ring will be playing a very important part in the final stages of the synthesis; in the meantime, it is a stable system which allows the building of rings V and V I to proceed unhindered. The next stages see the closing of ring I V : the tosyl group is replaced by acetyl by successive hydrolysis with aqueous hydriodic acid and red phosphorus, acetylation in pyridine and acetic anhydride, and esterification with diazomethane. A Dieckmann ring closure then gives the pentacyclic system CCXXIX. The reductive removal of the enolic hydroxyl proved t o be tricky, and was achieved by conversion of CCXXIX by way of the enol tosylate followed by sodium benzylmercaptide into the thioenolbenzyl ether (-SC7H7 for -OH), Raney nickel desulfurization to the a$-unsaturated ester, and catalytic hydrogenation to give stereospecifically the correct stereoisomer CCXXX (addition of hydrogen from the least-hindered side). This acid was resolved and found to be identical with the corresponding product formed by degradation of strychnine. This degradation involved simple barium hydroxide hydrogen peroxide cleavage of dehyd,rostrychninone (CCXXXII) (obtained by hydrolysis and oxidation of acetyldehydrostrychninolone,CXCIII, Ref. 155) to the amino acid which was then acetylated. At this stage, optically active CCXVII obtained by degradation of strychnine was used as a relay to complete the synthesis. The further progress of the synthesis involves the closing of ring VI. This was found to be very difficult to achieve, quite contrary, one supposes, to anticipation. All the obvious methods failed, and the successful method turned out to involve conversion of the acid CCXXX into the methyl ketone CCXXXI by pyridine and acetic anhydride, and selenium dioxide oxidation of this ketone to dehydrostrychninone (CCXXXII). The final stages are reaction with sodium acetylide, and reduction of the ethinylcarbinol with Lindlar catalyst to CCXXXIII. At this stage, a most tricky and ingenious reaction was achieved: reduction of the pyridone ring with lithium aluminum hydride gave a 30% yield of CCXXXIV with the ring I11 double bond nicely placed the correct stereochemistry at (2-8, the unwanted Nb lactam carbonyl reduced to

17. Strychnos

ALUOIDS

647

CH2, and the N, lactam carbonyl intact. Unbelievable! Vigorous acid treatment finally caused an ally1 rearrangement and gave isostrychnine (CCXXXV). Since this had already been converted into strychnine (see Section 11,B, Z), the total synthesis was complete.

V. Vomicine: Historical Survey

Nearly all the work on this alkaloid was carried out by Wieland and his collaborators. Their task was an exceptionally hard one, as is evident in the following discussion. They were able to unravel the greater part of the structure by making use of analogies with the reactions of strychnine. Wieland’s working hypothesis in 1948 was structure CCXXXVI, in which the nature of the fourth oxygen and of Nb and its environs are in error (175). One of the misleading observations in this connection was a negative Herzig-Meyer determination on vomicine, which seemed to speak clearly for the absence of an N-methyl group (187) : later work resulted in positive N-methyl determinations, but these were held to be spurious (188). The main single structural contribution was made by Robinson and his collaborators who, speculating on a structural relationship between vomicine and N-methyl-sec-pseudostrychnine (CCVI), oxidized N-methyl-sec-pseudobrucine with chromic acid and obtained a acid identical with that obtained by the chromic acid oxidation of vomicine. This established the correct structure, CCXXXVII, for vomicine (189), with interaction between the (2-16 carbonyl and the Nb as in N-methyl-sec-pseudostrychnine(see Section 11, F). The position of the phenolic hydroxyl group has not been proved by a classical method of degradation, but by spectroscopic evidence (190) combined with the observed cryptophenolic properties (191). Most of the reactions and degradations carried out on vomicine find their counterpart in strychnine chemistry : thus, there is electrolytic reduction to vomicidine, formation of vomicinic acid and isovomicine, condensation of vomicine with benzaldehyde and with amyl nitrite, oxidations with chromic acid leading to products in which the benzene ring has been destroyed. The one main difference is that permanganate oxidation has not been of any value a t all: it will be remembered that permanganate oxidation played the leading part in the elucidation of the structure of strychnine ; its failure in the vomicine series was thus quite a handicap. This failure was simply owing to oxidation of the phenolic system as well as of the Nb-C-21, C - 2 2 region :furthermore, protection of the benzene ring against permanganate oxidation by 0-methylation was

648

G. F. SMITH

not possible because of the strong hydrogen-bonding between the phenolic hydroxyl and the lactam carbonyl. The Hofmann and Emde degradations led to many interesting transformations, none of which, however, helped to shed light on the structural problem.

kOzH CCXXXVI

CCXXXVII

CCXXXVIII

bop CCXXXIX

CCXL

CCXLI

VI. The Reactions of Vomicine and Its Derivatives and Degradation Products

A. REACTIONS INVOLVING THE BENZENE RINGAND THE PHENOLIC HYDROXYL GROUP(NOTINCLUDING OXIDATIONS) That vomicine contains a phenolic hydroxyl group is not immediately obvious, for the alkaloid is insoluble in alcoholic alkali, does not give a ferric chloride color, and is not easily methylated or acetylated on the phenolic oxygen ( 1 8 7 ) . Later work led to the isolation of what may be 0-methylvomicine in small yield (191) and t o 0-acetylvomicine by the action of acetic anhydride sodium acetate (192). The presence of a phenolic grouping is evident, however, in vomicinic acid (CCXXXVIII ;

17. Strychnos

ALKALOIDS

649

R = R' = H), a compound which autoxidizes very readily t o give deepgreen solutions (187). The phenolic group is also evident in vomicidine (CCXXXIX), which easily forms a methyl ether (192, 193), and O-acetyl and O-benzoyl derivatives ( 193). It was Wieland and Calvet who proposed position 4 for the phenolic hydroxyl group, and interpreted the cryptophenolic properties in terms of an extra ring, as in partial structure CCXL (191). This was later replaced by the correct structure, which involves strong hydrogenbonding between the phenolic group and the lactam carbonyl oxygen. This was confirmed by spectroscopic evidence (190): the UV-spectrum of vomicine is very closely similar t o that of aspidosine, which contains a phenolic hydroxyl ortho to an N-acetyl group and is also similar to the UV-spectrum of N-acylindolines, such as strychnine (chromophore CCXL would be expected to resemble that of, say, O-methoxyaniline, and be quite different). Further spectral evidence for the presence of the lactam carbonyl comes from the IR-spectrum, which shows a band at 1632 cm-1 (6.12 p ) and no hydroxyl absorption in the 3300-3500 cm-1 ( 3 p ) region : this is exactly duplicated in the spectrum of aspidosine. The 1632 cm-1 band is not present in vomicidine (190). Vomicine is easily monobrominated, probably in position 3, ortho to the phenolic group (187). Mononitration may be achieved with dilute aqueous nitric acid, but is difficult and gives varying yields; dihydrovomicine on the other hand gives a dinitro derivative in very good yield (194). Nitric-concentrated sulfuric acid mixture a t - 18' oxidizes the aromatic ring (see Section IV, E, 2).

B. REACTIONS AROUND RINGSI11 AND VII (NOTINVOLVING OXIDATIONOR REDUCTION) ; THE' DEOXYVOMICINES

1. Opening of the N , Luctum This occurs readily with warm alcoholic potassium hydroxide : vomicine gives vomicinic acid (CCXXXVIII ; R = R' = H) and dihydrovomicine gives dihydrovomicinic acid (187). These acids are very readily autoxidized because of the ortho amino-phenol system they contain, and must therefore be handled in the total absence of oxygen. The ring opening is easily reversed by mineral acid (191). Vomicine methiodide is likewise hydrolyzed by warm alkali to the betaine CCXLI (195). Vomicinic acid is converted by successive treatment with methyl iodide and alkali into a mixture of N,-methyl- and O,N,-dimethylvomicinic acids and methyl esters (191).

a. F.

650

SMITH

2. Reactions at C-11, cc to the N , Carbonyl There is complete parallelism here with strychnine ; thus, dihydrovomicine gives an 11-benzylidene derivative (196). Deoxyvomicine (CCXLII) gives a product which may be the simple benzylidine compound CCXLIII or the rearranged a-pyridone CCXLIV (197): deoxydihydrovomicine behaves similarly.

I

CHPh

CHzPh

CCXLIII

CCXLIV

Vomicine condenses with ethyl nitrite in the presence of alkoxide to give 11-isonitrosovomicine, which with thionyl chloride undergoes Beckmann rearrangement to give the nitrile CCXLV, easily hydrolyzed to norvomicinic acid (CCXLVI) (198).

CCXLV

CCXLVI

3. Opening of Ring V I I : Isovomicine, Yellow and Colorless Deoxyvomicines, and Neodeoxyvomicine Isovomicine (CCXLVII) is formed by the action of hydrobromic acid and red phosphorus in glacial acetic acid on vomicine : the corresponding 23-bromo compound surprisingly was not isolated (199,200),contrasting with the behavior of strychnine (see Section 11,B, 2 ) . The bromo compound CCXLVIII is, however, the main product of the action of hydrobromic acid on dihydrovomicine (199). The action of hydriodic acid and red phosphorus in glacial acetic acid on vomicine leads to yellow deoxyvomicine, CzzHz403N2, with loss of the ether oxygen (187, 197). The structure of this compound is not yet firmly established ( 1 62) : structure CCXLIX has been proposed for it

17. Xtrychnos

ALKALOIDS

651

(201), but this does not satisfactorily account) for its UV-absorption. Yellow deoxyvomicine is easily converted into colorless deoxyvomicine, whose reactions are fully in accord with structure CCX L I I (202), by alkali in pyridine (196), long refluxing in organic solvents, high vacuum distillation, and by zinc chloride or sodium acetate in acetic acid (201). The last conditions also give rise to 20% of a third isomer, neodeoxyvomicine (CCL) : this base is the main product of the action of potassium iodide and red phosphorus in phosphoric acid on vomicine (192). There is no explanation yet for the formation of these different products. Dihydrovomicine is converted by hydriodic acid and red phosphorus in glacial acetic acid into a colorless iodotetrahydrovomicine hydriodide, basification of which gives, with loss of 2H1, to deoxydihydrovomicine (CCLI), which is colorless : a yellow isomer is not formed (187, 197). A minor product in the reaction of vomicine with hydriodic acid and red phosphorus in glacial acetic acid is 12-iododihydrodeoxyvomicine (CCLII ; R = I) (197)) which is reduced to isodihydrodeoxyvomicine (CCLII; R = H) by zinc and acetic acid. The structure of colorless deoxyvomicine (CCXLII) follows from (a) ozonolysis, which gives acetaldehyde in high yield; (b) the easy formation of a benzylidene derivative, in accord with a CH2 group a t C - l l ( 1 9 7 );and (c)the UV-spectrum, which is similar to that of vomicine, showing that there is no double bond in conjugation with the N, lactam system (201, 202). The structure of neodeoxyvomicine (CCL) follows from (a) the failure of ozonolysis to give any acetaldehyde, (b) the low pK, of 5.16, as compared with the high value of 7.4 for deoxyvomicine, (c) the resistance of Nbto quaternization. These three pieces of evidence support the presence of a double bond in the neo position. The position of the other double bond is given by the failure of neodeoxyvomicine to form a benzylidene derivative and by the UV-spectrum, which is similar to that of N crotonyl-O-aminophenol (CCLIII) (201). The structure of yellow deoxyvomicine must be closely related to that of colorless deoxyvomicine because of the ease of conversion of the former into the latter already mentioned. Since ozonolysis yields about 60% of acetaldehyde, yellow deoxyvomicine must contain the ethylidene system (201). Structure CCXLIX is the only one which is closely related to deoxyvomicine, and explains the conversion quite plausibly by analogy with the conversion of strychninolone-a into strychinolone-b. The stability of the 11 , l %double bond in neodeoxyvomicine is explained by assuming a relationship with strychninolone-c, that is, by visualizing an epimerized CIS. This stereochemical difference is taken to be the cause of the very marked difference between the UV-absorption of yellow- and

652

0.F. SMITH

neodeoxyvomicine, the longer wavelength absorption of the former owing to a higher degree of conjugation (201).This does not sound really convincing, for the UV-absorption of N-crotonyl-0-aminophenol (CCLIII), in which there is nothing t o prevent the double bond from conjugating fully with the rest of the chromophore, is almost identical with the shorter wavelength absorption of neodeoxyvomicine (201,162).

CCXLVII

CCXLVIII

CCXLIX

CCXLII

CCL

CCLI

CCLII

17. Strychnos

ALKALOIDS

653

The absorption of yellow deoxyvomicine is, in fact, much more closely similar to the absorption of dehydrostrychninolone (CXCIII) (155, 162). Clearly, more experimental work is called for. Of particular interest is the normally basic character of yellow and colorless deoxyvomicines. These bases have a pK, of the order of 7.5, which is closely similar to that of strychnine but much higher than that of vomicine, which is 5.8 (201). Huisgen et al. (202) argue that the higher basicity indicates a higher degree of interaction between Nb and the C-16 carbonyl carbon, presumably leading to a greater partial negative charge on the C-16 oxygen which would be the site of proton addition. I n support of this, he quotes the observations that deoxyvomicine and isovomicine are hydrogenolyzed with Nb-C20 Emde fission to products containing a CIS-Nb bond, and that vomicine, with a lower degree of Nb-carbonyl interaction is only hydrogenated to the 2 1,22-dihydro derivative. It would be of particular interest to know whether yellow and colorless deoxyvomicines show a C-16 carbonyl band in the IR. C. QUATERNIZATIONS : HOPMANN AND EMDE DEGRADATIONS Vomicine does not react with methyl iodide, but does so with dimethyl sulfate in benzene to give the quaternary salt CCLIV (195). (The Munich school throughout give structures for the quaternary salts later shown to be erroneous in the analogous pseudostrychnine and its derivatives by Boit. The structures in this review are brought up to date.) Dihydrovomicine, however, reacts normally with methyl iodide to give the analogous quaternary salt. Vomicidine gives the dimethiodide CCLV (193). This salt undergoes an extraordinary reaction on being treated with diazomethane: demethylation of the oxygen on C-16 apparently occurs to give the betaine CCLVI. Subsequent pyrolysis of the betaine in mcuo leads in a normal manner to 0-methylvomicidine (193). The sodium amalgam (Emde) reduction of vomicine metho salts (CCLIV) in dilute acetic acid gives two products (195, 203): base I , which has structure CCLVII (R = CH3) (202),and base 11,the structure of which is not proved (202, 203). Why allylic hydrogenolysis does not occur is not known. Base I is demethylated to a secondary alcohol CCLVII (R = H) (203), which could not be oxidized by the Oppenauer method back to vomicine (202), nor could vomicine be reduced t o it. Base I is quaternized to CCLVIII, which now undergoes the normal hydrogenolysis with Na/Hg to give base CCLIX, Hofmann degradation of which gives trimethylamine and a deaza compound which could not be characterized (203).

a.

654

F. SMITH

Catalytic hydrogenolysis of methyl vomicinium salts leads to a mixture of the corresponding dihydrovomicinium quaternary salt and of two isomeric tertiary bases, probably epimeric at C-21, t o which structure CCLX has been given (195, 202). The reduction in this case then proceeds by way of a normal allylic hydrogenolysis, and is to be contrasted with the abnormal Na/Hg reduction.

CCLIV

CCLV

CCLVI CHa-N-CH3

CCLVII

ccL vIII

CCLIX

The quaternization of colorless deoxyvomicine is not straightforward. Reaction with methyl iodide at 100" gives methiodide A (203): it was later found that reaction with dimethylsulfate gives a methosulfate which with K I is converted into an isomeric methiodide B (202). Methiodide B is isomerized to methiodide A by a trace of iodine in boiling methanol (202). Methiodide B is believed to be derived normally from the colorless deoxyvomicine with no double bond migrations and to have structure CCLXI (202, 162). It behaves on reduction in a manner entirely parallel with that of methylvomicinium salts; that is, it undergoes an anomalous Emde reduction with Na/Hg to give base CCLXII, and a normal allylic hydrogenolysis followed by further hydrogenation and hydrogenolysis of the methoxyl group t o give a mixture of two stereoisomeric bases CCLXIII (195, 202). The structure of the rearranged methiodide A is not yet clearly proved. The Munich school suggested that the isomerization of B into A involves the migration of the ethylidene double bond into the neo position, giving structure CCLXIV for methiodide A (201, 202). This

17. Strychnos

CCLX

655

ALKALOIDS

CCLXI

CCLXII

CCLXIV

CCLXV

CCLXIII

CCLXVJII

CCLXVI

CCLXVII

CCLXIX

656

G. I?. SMITH

structure does not easily account for the Emde reduction to two isomeric bases, I and 11,of which base I yields acetaldehyde on ozonolysis. But it does account for the observations that the base I contains two N-methyl groups, forms a diacetyl derivative, and is hydrogenated catalytically to a dihydro derivative which no longer gives acetaldehyde on ozonolysis (202). Catalytic hydrogenation of methochloride A in very dilute HC1 gives a hexahydro tertiary base C23H3203N2 containing two N-methyl groups, reduction of which with Na/Hg gives an alcohol C23H3403N2 (202), not identical with dihydro base I already mentioned. An alternative formulation for methiodide A, CCLXV (162) accounts for the allylic hydrogenolysis, and requires double-bond migration for the formation of base I, which would have structure CCLXVI : base I1 would have structure CCLXVII, the dihydro derivative of base I would be a stereoisomer of CCLXVIII, as would the hydrogenolysis and Na/Hg reduction product C23H3403N~. Either formulation CCLXV or CCLXVI for methiodide A requires that the iodine catalyzed rearrangement involves migration of methyl from oxygen (in CCLXI) to nitrogen. This takes one back to earlier controversies associated with the structure of pseudostrychnine quaternary salts. More work is required. The methiodide of CCLXVI loses NMe3 and HI in a high vacuum a t 270' to give a 23% yield of crystalline deazadeoxyvomicine (CCLXIX). The two stereoisomeric tetrahydro derivatives of CCLXVI behave similarly (202). Yellow deoxyvomicine reacts with dimethyl sulfate in boiling benzene to give a pale yellow quaternary methosulfate. Although the direct isomerization of this to the colorless quaternary salt derived from colorless deoxyvomicine and methylsulfate is not reported, it probably occurs under the alkaline conditions of the Emde reduction with Na/Hg, for the main product there is identical with base CCLXII, obtained from the quaternary methochloride of colorless deoxyvomicine (CCLXI) (202).

D. REDUCTIONS Normal electrolytic reduction of vomicine t o vomicidine (CCXXXIX) occurs : no product corresponding to tetrahydrostrychnine (LXXXIV) is formed (192, 193).The electrolytic reduction of the N, lactam carbonyl is also achieved with colorless deoxyvomicine and several other compounds (196, 204), but does not proceed normally with yellow deoxyvomicine, the main product gives vomicidine color reactions, but does not crystallize. A minor crystalline product is obtained (198); this

17. Strychnos

657

ALKALOIDS

probably has structure CCLXX. Electrolytic reduction of the enamine double bond in " tetrahydro " deoxyvomicine A (CCLXXVIII) also occurs to give CCLXXIX (196, 205). Straightforward catalytic hydrogenation of the 21,22-double bond in vomicine (187) and in vomicidine (193) occurs with platinum in acetic acid. The catalytic hydrogenation of the deoxyvomicines, deoxyvomicidine, and related compounds, however, presents a more complex picture (201, 162). The earlier reports contain a number of contradictory results, and a critical and summarizing account containing much new work and full reference to previous work is given in Reference 201. The nature of the hydrogenation products depends very much on solvent, temperature, and activity of the catalyst. The over-all conclusions are best grasped by examining structures CCLXX-tCCLXXXI.

CCLXX A (Dihydrodeoxyvomicine-11)

CCLXXIII DandE

CCLXXII C

CCLXXI B

Yellow deoxyvomicine gives the greatest number of hydrogenation products, five in all, A-E. The primary product A is dihydrodeoxy vomicine I1 (CCLXX), also obtained in very poor yield by the electrolytic reduction of yellow deoxyvomicine or by the zinc and acetic acid reduction of the iodo compound obtained as a minor product in the preparation of yellow deoxyvomicine. Product B (CCLXXI) simply arises by reduction of the 21,22-double bond: further reduction of B does

658

G . F. SMITH

Colorless deoxyvomicine (CCXLII)

CH3 ‘CH3

0”

cC LX xIv

CCLXXV G and H

F

CCLXXVI Dihydrodeoxyvomicine-I

CCLXXXII Dihydrodeoxyvomicidine

CCLXXXIII “Tetrahydro” deoxyvomicidine C

“

CCLXXVII K and L

CCLXXVlII Tetrahydro ” deoxyvomicine A (K)

CCLXXX “Tetrahydro ” deoxyvomicine B (L)

CCLXXIX Tetrahydrodeoxyvomicidine A

CCLXXXI Tetrahydro” deoxyvomicidine B “

17. Strychnos

ALKALOIDS

659

not occur. Product C (CCLXXII) arises by allylic hydrogenolysis of the zwitterionic form of CCLXX, and is further hydrogenated to bases D a.nd E, epimeric a t C-21 (CCLXXIII). Colorless deoxyvomicine is reduced by allylic fission, together with reduction of the 12,13-double bond, to give F (CCLXXIV). Without definite experimental evidence, Huisgen claims P and C to be C-13 epimers. Further reduction gives bases Q and H (CCLXXV),epimeric at C-21, and stereoisomeric with bases D and E (CCLXXIII). The main difference between the hydrogenation of yellow and colorless deoxyvomicines thus is that, in the latter, simple hydrogenation of the 21,22-double bond does not occur as a first step, so that products still containing the C-16 carbonyl are not formed. The hydrogenation of dihydrodeoxyvomicine I (CCLXXVI), the product of zinc and acetic acid reduction of the HBr adduct of colorless deoxyvomicine (199) and of the bronio compound CCXLVIII, leads to two products claimed by Huisgen to be tetrahydrodeoxyvomicines (CCLXXVII) epimeric at C- 13 reduced electrolytically to the corresponding deoxyvomicidines. A reinterpretation of this work by Woodward (205, 206) claims that hydrogenation of CCLXXVI does not occur: compounds K and L are double-bond isomers. One of them is given structure CCLXXVIII, and its electrolytic reduction product is given the structure of a true tetrahydrodeoxyvomicidine (CCLXXIX). This extra reduction occurs because of the enamine nature of the 8,13double bond. The other “ tetrahydro ” deoxyvomicine is given structure CCLXXX, and its electrolytic reduction product is a dihydrodeoxyvomicidine (CCLXXXI). Electrolytic reduction of dihydrodeoxyvomicine I (CCLXXVI) gives dihydrodeoxyvomicidine (CCLXXXII), catalytic “hydrogenation ” of which gives the isomer CCLXXXIII. The enamine position of this double bond follows from the failure of this compound to give the typical color reactions of vomicidine in acid media. This is explained by protonation of C-13 leading to nonoxidizable salts (205). Q

+-N==C--CH

E. OXIDATIONS

I. Chromic Acid The oxidation of vomicine yields three products :the acid CCLXXXIV, analogous to dioxonucine dihydrate (Wielands’ C17 acid, CXXV), the acid CCLXXXV, the analog of carboxyaponucine (Hanssen’s CIS acid, CXXVI), and the base CCLXXXVI, which has no analog in strychnine

660

G . P. SMITH

chemistry and is produced by the decarboxylation ofthe acid CCLXXXV (187, 207). This easy decarboxylation of acid CCLXXXV is readily understood, since the C-16 carbonyl is p to the carboxyl group. Acid

CCLXXXIV

CCLXXXV

CCLXXXVI

CCLXXXV is also formed by oxidation of N-methyl-see-pseudobrucine (189), which led t o the elucidation of the structure of vomicine (see Section V). The oxidation of dihydrovomicine gives the dihydroderivatives of CCLXXXIV, CCLXXXV, and CCLXXXVI, which may also be obtained by catalytic hydrogenation of the vomicine oxidation products (207). The oxidation of vomicidine leads t o one product only in about 10% yield (188). This acid is given structure CCLXXXVII. In contrast t o acid CCLXXXV, it does not decarboxylate at all readily and is converted, by heating to 200°, into a mixture of compound CCLXXXVIII

’ 0 CCLXXXVII

HOzC\ OC’l\

CCLXXXVIII

i;.“r3

CCLXXXIX

I

CCXCII

17. Strychnos

661

ALKALOIDS

and base CCLXXXIX. One of the main reasons for preferring structure CCLXXXVII to the alternative structure CCXC is the failure of the compound to show any carbonyl reactivity which would be expected of CCXC, by analogy with CXXV, CCLXXXIV, and with CCLXXXVIII, which forms a hydrazone converted by the Wolff-Kishner method into the normal reduction product CCXCII. The simple thermal cyclization of CCLXXXVII t o CCLXXXVIII, involving acylation of position 7 by the oxalyl carboxyl group, is unusual: it would be interesting t o know if the N-oxalyl derivative of CCLXXXIX could likewise be cyclized. The main difficulty, which applies to both structures CCLXXXVII and CCXC, is the resistance to thermal or aqueous alkali-catalyzed decarboxylation. Strong hydrogen-bonding of the p-keto acid carboxyl in CCLXXXVII with the N, amide carbonyl is the most plausible explanation : this would hinder hydrogen bonding with the C-16 carbonyl, which is a desirable prelude to easy decarboxylation (162). An alternative explanation is that the structure of the acid is CCXCI, in which there is no /3-keto acid system ( 1 7 7 ) : it is difficult to see why the same situation should not arise in the case of acid CCLXXXV, although it might be argued that the -COT in CCXCI is stabilized by hydrogen bonding with the oxalyl carboxyl. The oxidation of colorless deoxyvomicine (CCXLII) fails to give a crystalline product, but oxidation of deoxyvomicidine (CCXCIII) gives the base CCXCIV in about 3% yield (204). This oxidation finds no counterpart in strychnine chemistry, for, in addition to the destruction

CCXCIII

CCXCIV

ccxcv

of the benzene ring, it involves dehydrogenation of ring I11 to a simple aromatic pyridine system. The acid corresponding t o CCXCIV (by analogy with CCLXXXV) is not isolated presumably because of very easy decarboxylation owing to a double activation by the (3-16 carbonyl and by the cc-pyridine carbon. Oxidation of deoxydihydro vomicidine (CCLXXXII) gives CCXCV (208). Under very much milder conditions, with controlled quantities of chromic acid, oxidation of vomicinic acid (CCXXXVIII; R = R’ = H)

662

Gi. F. SMITH

and of all compounds containing a basic N, atom gives deep-purple products reduced by zinc and acid to crystalline dimeric leuco bases ( 187, 209, 210) ; thus vomicidine gives divomicidyl (CCXCVI) (193).

CCXCVI

CCXCVII

2. Miscellaneous Oxidations The only other oxidative degradation of any importance leading t o a large fragment is that of vomicine by nitrating mixture at - 18', which gives an acid formulated as CCXCVII (allowing for the known structure of vomicine) (194). Ozonolysis has been of value in determining the presence or absence of ethylidene groups (197).

+ I OH CCLXXXIX

CCXCIX

CCXCVIII

T

ccc

17. Strychnos ALKALOIDS

663

F. DEHYDROGENATIONS : VOMIPYRINE AND RELATED COMPOUNDS Palladium dehydrogenation of the base C I ~ H Z ~ O (CCLXXXIX) ~N~ leads to the formation of vomipyrine (CCXCVIII)and smaller quantities of base CCXCIX, shown to be an intermediate by being converted into vomipyrine by further treatment with palladium (188, 211). The structure of vomipyrine was proved by synthesis (212). Palladium dehydrogenation of base CCC, produced by catalytic hydrogenation and hydrogenolysis of the acid CCLXXXV, gives oxyvomipyrine (CCCI) converted into vomipyrine by zinc dust distillation (211, 200). Milder dehydrogenation of base CCC with sulfur gives compound CCCII (200). A completely different reaction course occurs when base CCXCV is dehydrogenated: loss of CHzO occurs, but this involves loss of the N-methyl, and gives a base C15H20N2, which may well be identical with base CCXV (206, 175).

VII. Minor Alkaloids A. THEAUSTRALIAN Xtrychnos ALKALOIDS Of the four species of Xtrychnos found in Australia, S. lucida R. Br. was found t o contain strychnine (0.1%) and brucine (1.3%) as well as a further 0.5% of unresolved alkaloids (213). An alkaloid, iucidine-S, has been reported from the same source (214), but may,be present only in the young plants, as its presence could not be confirmed in the older samples (213). The leaves of S. psilosperma F. Mueli. were claimed to contain strychnine, brucine, and strychnicine (214) ;a later investigation (213) did not confirm this, but resulted in the isolation of two new alkaloids, strychnospermine (0.9%) and spermostrychnine (0.5%). Strychnospermine, CzzHzsNz03, was found to contain one methoxyl and at least two C-methyl but no N-methyl groups (213). One nitrogen was tertiary, forming a methiodide; the other was present as N-acetyl, with a UV-absorption very similar to that of 13-colubrine,which yielded an N-nitroso derivative after hydrolytic deacetylation. The remaining oxygen atom was unreactive and assumed to be an ether bridge. It is a tribute to the elucidative power of the biogenetic mechanism .that it sufficed, with these few data, to predict the formula (CCCIII; R = OCH3) for strychnospermine (213), which has been virtually established by the experiments described here (215). Spermostrychnine, C Z ~ H Z ~ N ~ O Z , contains no methoxyl group and shows the UV-spectrum of strychnine.

664

G . F. SMITH

It was felt that this alkaloid represented simply the demethoxylated derivative (CCCIII; R = H) of strychnospermine (213); this was proved by oxidation of each alkaloid to the same C15H20N203 base with chromic acid. The chemical similarity to dioxonucidine in the strychnine series allowed formulation of this compound as CCCIV. Deacetylspermostrychnine was treated vigorously with hydrogen bromide in acetic acid and the crude product reduced with zinc and acetylated to yield deoxydihydrospermostrychnine (CCCV). The identical substance was produced by reduction of the Wjeland-Gumlich aldehyde (216, 217) with potassium borohydride and hydrogen over palladium-charcoal, followed by the same hydrogen bromide, zinc dust, acetylation series as before.

CCCV

Thus, the carbon skeletons of these two new alkaloids are established (215). The size of the ether ring was confirmed by NMR-spectroscopy [F. A. L. Anet, quoted in (217a)], and the mass spectra of both alkaloids were found to be in complete accord with structures CCCIII (217a, p. 331). The position of the methoxyl group in strychnospermine was further confirmed by rhodamine dye colors produced on fusion of phthalic anhydride with demethyldeacetylstrychnospermine or N-ethyldemethyldeacetylstrychnospermine, which showed the presence of a metahydroxyaniline, and, more conclusively, by comparison of UVspectra with various model compounds (bz-methoxy-hexahydrocarbazoles) (215). Thus, the relationship of spermostrychnine and strychnospermine corresponds to that of strychnine and ,&ohbrine, respectively, and adds new data to the already strong evidence for the biogenetic schemes.

665

B. THECONGOStrychnos ALKALOIDS In recent years, a Belgian group has been examining the alkaloids of several species of Strychnos in this region, having identified strychnine in S. icaja in the amount of 6. 6% in the branch bark, thus making this by far the richest source yet discovered (218). Two new alkaloids, "B" and "C," were isolated by chromatography (219, 220). The formulas C23H34N208 and C23H34N207 were given as well as color reactions and UV-spectra similar to those of brucine for both compounds. Methoxyl but no methylenedioxy groups were observed. S. angolensis also yielded strychnine and two other alkaloids by paper chromatography (221). From S. holstii four new alkaloids were separated by chromatography : holstiine, holstiline, condensamine, and retuline ( 2 2 2 ) . Holstiine, C22H26N204, mp 248"-250", [a] + 268.9", contains one N-methyl but no methoxyl group and is not catalytically hydrogenated (223). The UV-spectrum is very similar to that of strychnine but significantly different from that of vomicine or isovomicine. The compound is reported to be soluble in excess 1 yoNsOH but not in NH40H (222), but the UV-spectrum, on the other hand, is unchanged in alkaline medium (223). The IR-spectrum shows a carbonyl doublet just above 6.0 p, one peak (or both) of which is presumably due t o a lactam, probably located as in strychnine. The lower carbonyl of the doublet may be due to an unreactive C- 16 carbonyl as in vomicine, but, compared with vomicine (pK= 5.88) or strychnine (pK= 7.37) (166), holstiine possesses an unusually strong basicity at pK 8.8, which would seem to contravene this hypothesis. Holstiline, C23H30N204, mp 219"-220",, has a similar UV-absorption but has methoxyl and no N-methyl groups. Condensamine, C24H28N205, mp 262"-265", also has methoxyl without N-methyl but shows a UVabsorption similar t o a-colubrine. Retuline, C21H26N202, mp 165"-170°, has no N-methyl or methoxyl and shows a UV-spectrum almost identical with that of strychnine. Its formula, solubilities, and meiting point suggest it is identical with the tetrahydroneostrychnine of Robinson et al. (33) (reported mp i67'-168"), although the comparison has not been made.

C. S . nux-vomica ALKALOIDS The alkaloid novacine has recently been isolated from S. nux-vomica and proved to be N-methyl-sec-pseudobrucine (224), thus establishing another biogenetic link between the strychnine and vomicine series and

666

G . F. SMITH

demonstrating that oxidation of the former to pseudostrychnine can actually occur in the plant and need not be simply an artifact of oxidation during isolation. This follows from the fact that strychnine metho salts are not oxidized to the pseudo series, so that oxidation foIlowed by methylation must occur in the plant.

REFERENCES 1. Pelletier and Caventou, Ann. Chim. (Paris) 10, 144 (1819); 12, 117 (1820); cf. Fliickiger, Arch. Pharm. 230, 345 (1892). 2. V. Regnault, Ann. 26, 17, 35 (1838). 3. R. B. Woodward and W. J. Brehm, J . Am. Chem. SOC. 70, 2107 (1948). 4. Sir Robert Robinson, in “Progress in Organic Chemistry” (J.W. Cook, ed.), Vol. I, p. 12. Butterworths, London, 1952. 5. R. Huisgen, Angew Chem. 62, 528 (1950). 6. W. Shenstone, J . Chem. SOC.47, 139 (1885). 7. J. Tafel, Ann. 301, 285 (1898). 8. J. Tafel, Ann. 264, 33 (1891); 268, 229 (1892); 301, 288 (1898). 9. H. Hanssen, Ber. 18, 1917 (1885). 10. H. Leuchs, Ber. 41, 1713 (1908). 11. H. Leuchs and L. E. Weber, Ber. 42, 770 (1909). 12. H. Leuchs and H. Rauch, Ber. 47, 378 (1914). 13. H. Leuchs and G. Peirce, Ber. 45, 2654 (1912). 14. H. Leuchs and G. Schwaebel, Ber. 47, 1552 (1914); 48, 1009 (1915). 15. J. W. Perkin, Jr. and Sir Robert Robinson, J . Chem. SOC. 97, 305 (1910). 16. J. W. Perkin, Jr., Sir Robert Robinson, and R. C. Fawcett, J . Chern. SOC.131, 3082 (1928). 17. J. W. Perkin, Jr., Sir Robert Robinson, and K. N. Menon, J . ChemSoc. 133, 830 (1930). 18. E. Spgth and H. Bretschneider, Ber. 63, 2997 (1930). 19. Sir Robert Robinson, Proc. Roy. SOC.Ser. A 130,431 (1931). 20. K. N. Menon and Sir Robert Robinson, J . Chem. SOC. 134, 773 (1931). 21. K. N. Menon and Sir Robert Robinson, J . Chem. SOC.135, 780 (1932). 22. H. Leuchs and W. Baur, Ber. 65, 1230 (1932). 23. B. K. Blount and Sir Robert Robinson, J . Chem. SOC.135, 2305 (1932). 24. H. L. Holmes and Sir Robert Robinson, J . Chem. SOC.p. 603 (1939). 25. H. Leuchs and H. Griinow, Ber. 72,679 (1939). 26. M. Kotake, Proc. Imp. Acad. Tokyo 12, 99 (1936). 27. G. R. Clemo,J. Chem.Soc. p. 1695 (1936). 28. L. H. Briggs, H. T. Openshaw, and Sir Robert Robinson,J. Chem. SOC.p. 903 (1946). 29. V. Prelog and S. Szpilfogel, Helv. Chim. Acta 28, 1669 (1945). 30. R. B. Woodward, W. J. Brehm, and A. L. Nelson, J . Am. Chem. SOC.69,2250 (1947). 31. H. Leuchs and F. Riick, Ber. 73, 731 (1940). 32. R. N. Chakravarti and Sir Robert Robinson, Nature 160, 18 (1947). 33. 0. Achmatowicz, G. R. Clemo, W. H. Perkin, Jr., and Sir Robert Robinson, J . Chem. SOC.135, 767 (1932). 34. L. H. Briggs and Sir Robert Robinson, J . Chern. Soc. p. 590 (1934).

17. Strychnos

ALKALOIDS

667

35. J. H. Robertson and C . A. Beevers, Acta Cryst. 4,270 (1951);Nature 165,690 (1950). 36. C. Bokhoven, J. C . Schoone, and J . M. Bijvoet, Acta Cryst. 4, 275 (1951). 37. A. F. Peerdeman, Acta Cryst. 9, 824 (1956). 37a. C. Weissmann, H. Schmid, and P. Karrer, Helv. Chim. Ada 45, 62 (1962). 38. H. L. Holmes, i n “The Alkaloids” (R. H. F. Mankse s a d H.-L. Holmes, eds.), Vol. 11,p. 536. Academic Press, New York, 1952. 39. R. B. Woodward, M. P. Cava, W. D. Ollis, A. Hunger, H. V. Daeniker, and K. Schenker, J . Amer. Chem. SOC.76, 4749 (1954). 40. H. Leuchs and K . Steinborn, Ber. 71, 1579 (1938). 41. H. Leuchs and D. Ritter, Ber. 52, 1585 (1919). 42. H. Leuchs and F. Krohnke, Ber. 62,2176 (1929). 43. H. Leuchs and W. Schneider, Ber. 41, 4394 (1908). 44. A. E. Oxford, W. H. Perkin, Jr., and Sir Robert Robinson, J. Chem. SOC.130, 2389 (1927). 45. W. F. Loebisch and H. Schoop, Monatsh. Chem. 7, 83 (1886). 46. N. Moufang and J. Tafel, Ann. 304, 26 (1899). 47. M. Oesterlin and G. Imoudsky, Ber. 76, 176 (1943). 48. H. G. Boit, Ber. 84, 16 (1951). p. 982 (1952). 49. Sir Robert Robinson and J. E. Saxton, J . Chem. SOC. 50. H. Leuchs and P. Reich, Ber. 43, 2421 (1910). 51. H. Leuchs and L. E. Weber, Ber. 42,3706 (1909). 52. H. Leuchs, W. Diels, and A. Dornow, Ber. 68, 107 (1935). 53. A. Bacovescu and A. Pictet, Ber. 38,2787 (1905). 54. H. Leuchs and R. Nitschke, Ber. 55, 3171 (1922). 55. H. Wieland and R. G. Jennen, Ann. 545, 99 (1940). 56. H. Leuchs and H. Schulte, Ber. 75, 573 (1942). 57. V. Prelog, J. Battegay, and W. I. Taylor, Helw. Chim. Acta 31, 2244 (1948). 58. H. Leuchs and H. Schulte, Ber. 75, 1522 (1942). 59. H. Leuchs and W. Schneider, Ber. 42, 2498 (1909). 60. H. Leuchs and F. Krohnke, Ber. 65, 218 (1932). 61. R. B. Woodward, quoted in reference 38, p. 517. 62. J. T. Edward, Tetrahedron 2, 356 (1958). 63. J. T. Edward and Sir Robert Robinson, Tetrahedron 1, 29 (1957). 64. H. Leuchs and W. Bendixsohn, Ber. 52, 1445 (1919). 65. H. Leuchs and J. F. Brewster, Ber. 45,217 (1912). 66. V. Prelog, 8 . Szpilfogel, and J . Battegay, Helw. Chim. Acta 30,366 (1947). 67. H. Wieland and W. Gumlichi A7m. 494, 191 (1932). 68. H. Wieland and K. Kaziro, Ann. 506, 60 (1933). 69. W. H. Perkin, Jr. and Sir Robert Robinson, J . Chem. Soc. 132,964 (1929). 70. H. Leuchs and H. Schulte, Ber. 76, 1038 (1943). 71. H. Leuchs and A. Dornow, Ber. 69, 1838 (1936). 72. H. Leuchs, H. Grunow, and K. Tessmar, Ber. 70, 1701 (1937). 73, B. K. Blount and Sir Robert Robinson, J . Chem. Soc. 135,2305 (1932). 74. H. Leuchs and R. Nitschke, Ber. 55, 3741 (1922). 75. H. Leuchs, J. Gruss, and H. Heering, Ber. 55, 3731 (1922); H. Leuchs, E. Hellriegel, and H. Heering, Ber. 54, 2177 (1921). 76. G. R. Clemo, W. H. Perkin, Jr., and Sir Robert Robinson, J . Chem. SOC.130, 1589 (1927). 77. J. Tafel, Ber. 23, 2733 (1890); Ann. 264, 56 (1891). $8. 0. Achmatowicz and Sir Robert Robinson, J . Chem. Soc. p. 581 (1934).

668

G . F. SMITH

7%. 0. Achmatowicz and S. Achmatowicz, Bull. Aead. Polon. Sci. Ser. Sci. Chim. 10, 555 (1962). 79. 0. Achmatowicz, J. Chern. Soe. p. 1472 (1938). 80. 0. Achmatowicz and C. Dybowski, J . Chem. SOC. p. 1483 (1938). 81. J. M. Gulland, W. H. Perkin, Jr., and Sir Robert Robinson, J . Chem. SOC.130, 1627 (1927). 82. 0. Achmatowicz, P. Lewi, and Sir Robert Robinson, J. Chem. Soc. p. 1685 (1935). 83. 0. Achmatowicz and C. Dybowski, J . Chem. Soc. p. 1488 (1938). 84. 0. Achmatowicz and Sir Robert Robinson, J . Chem. SOC. p. 1467 (1938). 84a. 0. Achmatowicz and S. Achmatowicz, Bull. Acad. Polon. Sci. Ser. Sci. Chim. 10, 555 (1962). 84b. 0. Achmatowicz, A. Achmatowicz, J . Skolic, and M. Wiewiorowski, Personal communication (1963). 85. H. Leuchs and H. S. Overberg, Ber. 66, 79 (1933). 86. H. G. Boit, Ber. 86, 133 (1953). 87. H. I,. Holmes and Sir Robert Robinson, J . Chem. SOC.p. 807 (1939). 88. H. Leuchs and H. S . Overberg, Ber. 66, 951 (1933). 89. H. Leuchs, H. Beyer, and H. S. Overberg, Ber. 66, 1378 (1933). 90. H. Leuchs, Ber. 77, 676 (1944). 91. H. Leuchs and M. Mengelberg, Ber. 82, 247 (1949). 92. H. Leuchs, M. Mengelberg, and L. Hemmann, Ber. 77, 737 (1944). 93. 0. Achmatowicz, Roczniki Chem. 13, 25 (1933); 0. Achrnatowicz and S. Achmatowicz, Bull. Acad. Polon. Sci., Ser. Sci. Chim. 10, 595 (1962). 94. H. Leuchs, Ber. 67, 1607 (1934). 95. H. G. Boit, Ber. 83, 217 (1950). 96. H. G. Boit, Ber. 85, 106 (1952). 97. H. G. Boit, Ber. 84, 923 (1951). 98. H. G. Boit and L. Paul, Ber. 88, 697 (1955). 99. J. Tafel, Ann. 268, 245 (1892). 100. W. H. Perkin, Jr., Sir Robert Robinson, and J. C. Smith, J . Chem. SOC.135, 1239 (1932). 101. W. H. Perkin, Jr., Sir Robert Robinson, and J. C. Smith, J . Chem. Soc. p. 574 (1934). 102. H. G. Boit, Ber. 82, 303 (1949). 103. H. Leuchs, H. Mildbrand, and W. R. Leuchs, Ber. 55, 2403 (1922). 104. H. Leuchs and H. S. Overberg, Ber. 66, 1711 (1933). 105. V. Prelog and A. Kathriner, Helw. Chim. Acta 31, 505 (1948). 106. T. M. Reynolds and Sir Robert Robinson, J . Chem. SOC. p, 592 (1934). 107. C. Weissman, H. Schmid, and P. Karrer, Helv. Chim.Acta 43, 2201 (1960). 108. C. Weissmann, 0. Heshmat, K. Bcrnauer, H. Schmid, and P. Karrer, Helv. Chim. Acta 43, 1165 (1960). 109. 0. Achmatowicz, W. H. Perkin, Jr., and Sir Robert Robinson,J. Ghem.Soc. 135,486 (1932). 110. P. Karrer, C. H. Eugster, and P. Waser, Helw. Chim. Acta 32, 2381 (1949). 111. S. P. Findlay,J. Am. Chem. Soe. 73, 3008 (1951). 112. H. G. Boit and L. Paul, Ber. 87, 1859 (1954). 113. F. Cortese, Ann. 476, 283 (1929). 114. H. Wieland and W. Munster, Ann. 469, 220 (1929). 115. H. Leuchs and F. Krohnke, Ber. 63, 1045 (1930). 116. H. Leuchsand W. Wegener, Ber. 63, 2215 (1930). 117. H. Leuchs, G. Schlempp, and W. Baur, Ber. 65, 1121 (1932).

17. Strychnos 118. 119. 120. 121. 122. 123. 124. 125. 126.

ALKALOIDS

669

H. Leuchs, G. Schlempp, and A. Dornow, Ber. 66,743 (1933). H. Leuchs and H. Beyer, Ber. 67, 1577 (1934). H. Leuchs, Ber. 71,2237 (1938). H. Leuchs, H. Seeger, and K. Jaegers, Ber. 71, 2023 (1938). H. Leuchs and R. Anderson, Ber. 44,2136 (1911). H. L . Holmes, H. T. Openshaw, and Sir Robert Robinson, J . Chem. Soc. p. 908 (1946). R. Hall, Thesis, Harvard Univ., Cambridge, Massachusetts (1950). K. Warnat, Helv. @him.Actu 14, 997 (1931). M. Kotake and T. Mitsuwa, Chem. Abstr. 30, 5228 (1936) [J. Chem. Soc. Japan 57,

228 (1936)l. 127. P . J. Scheuer,J. Am. Chem. Soc. 82, 193 (1960). 128. V. Prelog and M. Kocbr, Helv. Chirn. Acta 30, 359 (1947). 129. A. Kogure, T. Sakan, and M. Kotake, Chern. Abstr. 47, 6959 (1953) [J.Inst. Polytech. Osaka City Utaiv. C2, 67 (1952)l. 130. H. Leuchs and F. Krohnke, Ber. 62, 2598 (1929). 131. H. Leuchs and G. Schwaebel, Ber. 46, 3693 (1913). 132. E. Oliveri-Mandala and G. Comella, Gazz. chim. Ital. 53, 283 (1923). 132a. P. Rosenmund and T. Wieland, Ber. 93, 775 (1960). 133. H. Leuchs, Ber. 47, 539 (1914). 134. H. Leuchs, Ber. 73, 1392 (1940). 135. K. Bernauer, W. Arnold, C. Weissmann, H. Schmid, and P. Karrer, Helv. Chirn. Acta 43, 717 (1960). 136. C. Weissmann, H. Schmid, and P. Karrer, Helv. Chim. Acta 44, 1877 (1961). 137. M. Kotake and T. Mitsuwa, Chem. Abstr. 30, 5229 (1936) [J.Chem. SOC. Jupan 57, 236 (1936)l. 138. H. Leuchs and H. Beyer, Ber. 67, 459 (1934). 139. M. Kotake and T. Mitsuwa, Ann. 505, 203 (1933). 140. H. Leuchs and W. Diels, Ber. 69, 47 (1936). 141. V. Prelog and M. Kocor, Helv. Chim. Acta 31, 237 (1948). 142. A. S. Bailey and Sir Robert Robinson, J . Chem. Soc. p. 703 (1948). 143. H. J. Teuber and E. Fahrbach, Ber. 91, 1301 (1958). 1438. H. Leuchs and W. Geiger, Ber. 44, 3050 (1911); H. Leuchs, Ber. 67, 1082 (1934). 144. H. J. Teuber and E. Fahrbach, Ber. 91, 713 (1958). 145. J. A. Joule and G. F. Smith, Proc. Chem. Soc. p. 322 (1959). 146. H. Leuchs, E. Tuschen, and M. Mengelberg, Ber. 77,408 (1944). 146a. K. Nagarajan, C. Weissmann, H. Schmid, and P. Karrer, Helv. Chim. Acta 46, 1212 (1963). 147. E. E. van Tamelen, P. E. Aldrich, and T. J. Katz, Chem. fnd.(London)p. 793 (1956). 148. M. Kotake, T. Sakan, and S. Kusumoto, Chem. Abstr. 33, 5858 (1939) [ S c i . Papers Inst. Phys. Chem. Res. (Tokyo)35, 415 (1939)l. 149. A. Kogure and M. Kotake, Chem. Abstr. 46, 6131 (1952) [J. Inst. Polytech. Osaka City Univ. C2, 39 (1951)l. 150. A. Kogure, T. Sakan, and M. Kotake Chem. Abstr. ,45, 7129 (1951) [ J . Chern. Soc. Japan 70, 182 (1949)l. 151. H. Leuchs and K. Taube, Ber. 58, 1729 (1925). 152. H. Leuchs and H. L. Louis, Ber. 72, 490 (1939). 153. R. N. Chakravarti, K. H. Pausacker, and Sir Robert Robinson, J . Chem. Soc. p. 1554 (1947). 154. A. Hanssen, Ber. 20, 452 (1897). 155. V. Prelc-g,M. Kocor, and W. I. Taylor, Helv. Chirn. Acta 32, 1052 (1949).

670

G . F. SMITH

156. R. B. Woodward, M. P. Cam, W. D. Ollis, A. Hunger, H. U. Daeniker, and K. Sehenker, J. Am. Chem. Soc. 76, 4749 (1954); Tetrahedron 19, 247 (1963); R. B. Woodward, Ezperientia Suppl. 11, 12, 213 (1955). 157. R. 13. Woodward and W. E. Doering, J . Am. Chem. SOC.66, 849 (1944). 158. J. Tafel, Ber. 26, 333 (1893). 159. H. Leuchs, F. Osterburg, and H. Kaehrn, Ber. 55, 564 (1922). 160. H. Leuehs, Ber. 70, 1543 (1937). 161. H. Leuchs and W. Schneider, Ber. 42, 2682 (1909). 162. J. B. Hendrikson, in “The Alkaloids” (R. H. F. Manske and H. L. Holmes, eds.), Vol. VI, p. 189. Academic Press, New York, 1960. 163. H. Leuchs, H. Flammersfeld, and G. Villain, Ber. 76, 1065 (1943). 164. R. Huisgen, H. Wieland, and H. Eder, Ann. 561, 193 (1949). 165. F. A. L. Anet, A. S. Bailey, and Sir Robert Robinson, Chem. I d . ( Lo n d o n )p. 944 (1955). 166. V. Prelog arid 0. Hiifliger, Welu. Ch,im. Acta 32, 1851 (1949). 167. C. Stoehr, Bcr. 20, 810, 1108, 2727 (1887). 168. L. Berend and C. Stoehr, J . Prakt. Chem. 42, 416 (1890). 169. W. F. Loebisch and H. Malfatti, Monatsh. Chem. 9, 626 (1888). 170. H. Coldschmidt, Ber. 15, 1977 (1882). 171. M. Kotake, Chem. Abstr. 30, 7579 (1936) [Proc. Impericd Acczd. ( T o k y o )12,99(1936)]. 172. M. Kotake, T. Sakan, and T. Miwa, Chem. Abstr. 45, 7582 (1951) [ J . Inst. Polytech. Osaka City U n i u . C1, 7 (1950)]. 173. G. Clemo and R. Raper, J . Chem. Soc. p. 891 (1946). 174. R. T. Holden and R. Raper, J . Chsm. SOC. p. 2545 (1963). 175. H. Wieland, R. Huisgen, and R. Bubenik, Ann. 559, 191 (1948). 176. R . N. Chakravarti and Sir Robert Robinson, J . Chem. SOC. p. 78 (1947). 177. K. H. Pauaacker and Sir Robert Robinson, J . Chem. SOC.p. 951 (1948). 178. M. Kotake and M. Yokohama, Cliem. Abstr. 30, 5229 (1936) [ J . CJLem. Soc. Jnpnn 57,240 (1936)l. 179. H. T. Openshaw and Sir Robert Robinson, J . Chem. SOC.p. 941 (1937). 180. H. T. Openshaw and Sir Robert Robinson, J . Chem. Soc. p. 912 (1946). 181. H. L. Holmes, H. T. Openshaw, and Sir Robert Robinson,J. Chem. SOC. p. 910 (1946). 182. Sir Robert Robinson and J. E. Saxton, J . Chem. 9oc. p. 2598 (1953). 183. A. R. Katritzky, J . Chem. SOC. p. 2581, 2586 (1955). 184. E. E. van Tamelen, L. J. Dolby, and R. G . Lawton, Tetrahedron Letters No. 19, 30 (1960). 186. J. B. Hendrickson and R. A. Silva, J . Am. Chem. SOC.84, 643 (1962). 186. J. Harley-Mason and W. R. Waterfield, Chem. I n d . (London)p. 1477 (1960). 187. H. Wieland and G. Oertel, L4nn.469, 193 (1929). 188. H. Wieland and L. Homer, Ann. 528, 73 (1937). 189. A. S. Bailey and Sir Robert Robinson, Nature 161, 433 (1948). 190. B. Witkop, quoted in reference 162. 191. H. Wieland and F. Calvet, Ann. 491, 117 (1931). 192. H. Wieland and W. W. Moyer, Ann. 491, 129 (1931). 193. H. Wieland, F. Holscher, and P. K. Bose, Ann. 507, 69 (1933). 194. H. Wieland and F. Holscher, Ann. 491, 149 (1931). 195. H. Wieland and 0. Miiller, Ann. 545, 59 (1940). 196. H. Wieland and J. Kimmig, Ann. 527, 151 (1937). 197. R. Huisgen and H. Wieland, Ann. 555, 9 (1944). 198. H. Wieland and G. Varvoglis, Ann. 507, 82 (1933).

17. Strychnos

ALKALOIDS

671

H. Wieland and R. G. Jennen, Ann. 545,99 (1940). H. Wieland and It. Huisgen, Ann. 556, 157 (1944). R. Huisgen, H. Eder, L. Blatzejewicz, and E . Mergenthaler, Ann. 573, 121 (1951). R. Huisgen, H. Wieland, and H. Eder, Ann. 561, 193 (1949). H. Wieland and W. Weisskopf, Ann. 555, 1 (1943). H. Wieland and 0. Sohmauss, Ann. 545, 72 (1940). R. 13. Woodward, Footnote 26 in reference 3. R. B. Woodward, Lectures 1946, quoted by E. Crane, i n Thesis, Harvard Univ., Cambridge, Massachusetts, 1949. 207. H. Wieland, F. Holscher, and F. Cortese, Ann. 491, 133 (1931). 208. H. Wieland and R . G. Jennen, Ann. 545,86 (1940). 209. H. Wieland, F. Calvet, and W. W. Moyer, Ann. 491, 107 (1931). 210. H. Wieland and F. Holscher, Ann. 500, 70 (1933). 211. H. Wieland and L. Homer, Ann. 545, 112 (1940). 212. Sir Robert Robinson and A. M. Stephen, Nature 162, 177 (1948). 213. F. A. L. Anet, G. K. Hughes, and E . Ritchie, Australian J. Chem. 6 , 58 (1953). 214. F. H. Shaw and I. S. de la Lande, Awtralian J . Ezpcpll. Biol. Med. Sci. 26, 199 (1948). 215. F. A. L. Anet and Sir Robert Robinson, J . Chem. Soc. p. 2253 (1955). 216. H. Wieland and W. Gumlich, Ann. 494, 191 (1932). 217. H. Wieland and K. Krtziro, Ann. 506, 60 (1933). 217a. K. Biemann, “Mass Spectroscopy.” McGraw-Hill, New York, 1962. 218. A. Denoe1,J. Phamn. Belq. [N.S.] 5,59 (1950). 219. F. Jaminet, J. Pharm. Belg. [N.S.] 8, 339 (1953). 220. F. Jaminet, Lejeunia 15, 9 (1951). 221. A. Denoel, F. Jaminet, E. Philippot, and M. J. Dallemagne, Arch. Intern. Physiol. Biochinz. 59, 341 (1951). 222. J. Bosly,J. Phurm. Belq. [N.S.] 6, 150,243 (1951). 223. M. M. Janot, R. Goutarel, and J. Bosly, Compt. Rend. Acad. Sci. 232, 853 (1951). 224. W. F. Martin, H. R. Bentley, J. A. Henry. and F. S. Spring, J . Chem. Soe. p. 3603 (1952). 199. 200. 201. 202. 203. 204. 205. 206.

This Page Intentionally Left Blank

-CHAPTER

18-

ALKALOIDS OF HAPLOPHYTON CIMICIDUM J. E. SAXTON The University, Leeds, E?tgland

The Mexican "cockroach plant," Haplophyton cimicidum A.DC. (Apocyanaceae), was first investigated by Snyder and his collaborators, who isolated two insecticidal alkaloids, haplophytine and cimicidine (1-4). Both alkaloids were reported t o be toxic to a wide variety of insects, but most of the toxicity of the plant appeared to be owing t o the haplophytine conLent ( I ) . Haplophytine, C27H31N30s1 mp 290"-293" (dec.), [ccI~D~" + 109" (chloroform), occurs t o the extent of 0.034% and is an amphoteric base containing two methoxyl groups, one methylimino group, possibly one C-methyl group, and two active hydrogen atoms. It gives rise to a series of diacidic salts ; however, titration of the dihydrochloride with alkali reveals that one of the two basic nitrogen atoms is much more basic than the other. All three nitrogen atoms are tertiary; the IR-spectrum shows little absorption in the region 3760-3000 cm-1, indicating the absence of ordinary imino (or hydroxyl) groups. Reaction of haplophytine with acetyl chloride gives an 0-acetate hydrochloride ; an N-acetyl derivative is not formed. Attempted Hofmann degradation fails, sirice haplophytine does not give a methiodide under normal conditions, and, although the alkaloid reacts with cyanogen bromide, the von Braun degradation also gives inconclusive results (1, 2). Haplophytine is a very weak acid, but it is apparent from its behavior that it contains neither carboxylic acid nor simple phenolic groupings. For example, chloroform extracts haplophytine from its solution in 0.2 N sodiuril hydroxide, but normal sodium hydroxide removes haplophytine from its solution in chloroform. Evaporation of its solutions in baryta or ammonia yields only hapiophytine (2). That the weakly acidic center is contained in a cryptophenolic grouping is established by the following observations. Haplophytine reacts slowly with diazomethane to give a nonacidic 0-methyl ether, mp 288"-291", [ c c ] ~+' 12" (chloroform), which contains only one active hydrogen ; this transformation is accompanied by the disappearance of an IR-band a t 1656 cm-1 in haplophytine, and the appearance of a band a t 17 15 cm-1 in the methyl ether. Similarly, the acetylation of haplophytine results in disappearance of the band a t 673

674

J. E. SAXTON

1656 em-’, and the appearance of absorption a t 1765 cm-1 (phenolic acetate) and 1710 cm-1. This behavior is reminiscent of o-hydroxyacetophenone, in which the carbonyl absorption is also displaced to lower frequencies by hydrogen bonding. The UV-spectrum of haplophytine is in accord with this conclusion. I n neutral solution, the UV-spectrum exhibits maxima a t 220 and 265 mp; the latter is displaced to 306 mp in alkaline solution, owing to formation of the corresponding phenoxide ion. I n contrast, the UV-spectrum of 0-methylhaplophytine is unchanged in alkaline solution, and bears a marked similarity to the spectrum of aspidospermine. Similarly, the spectrum of haplophytine itself closely resembles that of vomicine. Hence, haplophytine probably contains the chromophore I . This position of the hydroxyl group is consistent with the absence of normal hydroxyl absorption in the IR-spectrum and with the inconclusive Otto color reaction; in contrast, 0-methylhaplophytine gives a positive Otto reaction. The position of the amide carbonyl absorption in the IR-spectrum suggests that the amide grouping is contained in a five-membered lactam ring. The sensitivity of haplophytine to aerial oxidation in the presence of acids or bases is readily understood on the basis of the partial structure I ; as expected, haplophytine is stable to alkali in the absence of oxygen (3, 4).

I

Of the five oxygen atoms in haplophytine, four can be accounted for by the phenol, lactam, and two methoxyl groupings. The fifth oxygen atom appears to be contained in a carbonyl group responsible for I R absorption a t 1751 cm-1. Reduction of haplophytine with sodium borohydride yields a noncrystalline, hygroscopic product, which no longer exhibits absorption a t 1751 cm-l, but instead exhibits absorption owing to a hydroxyl group. The formation of an alcohol in this reduction is confirmed by the conversion of the product into an (unstable) 0-acetate. The high frequency of the carbonyl absorption in haplophytine suggests the presence of a four or five-membered cyclic ketone but no evidence confirming such a function could be obtained, since haplophytine is unreactive toward all carbonyl reagents. Other puzzling features of the chemistry of haplophytine are the resistance of

18.

ALKALOIDS OF

Haplophyton cimicidum

675

its borohydride reduction product to oxidation (Oppenauer) or catalytic hydrogenation, although the substance is not identical with the hydrogenation product of haplophytine (3). The nature of the unsaturation present in haplophytine is not yet clear. Hydrogenation over Adams' catalyst a t room temperature gives a rather unstable, amorphous product which appears to be the result of addition of two hydrogen atoms to the molecule ; this is confirmed by analysis of its picrate and hydrochloride. Repeated recrystallization of the base itself eventually affords a new product, C27H33N306, mp 1980-200°, [ m ] g O + 160' (chloroform), which is evidently formed by aerial oxidation of the primary hydrogenation product. The presence of a double bond in haplophytine cannot be confirmed by the usual reagents, since either no reaction occurs or the reaction is anomalous. The use of alkaline hydrogen peroxide leads to a dimorphic product, C21H24N205, or 250'-260", + 112' (chloroform). Potassium mp 216'-219' permanganate oxidation yields a weakly basic, acidic product, C27H29N306, but the course of this oxidation cannot be adequately explained by any of the normal oxidation processes involving loss of two hydrogen atoms and gain of one oxygen atom (2). Cimicidine, C23HzsN205, mp 259'-262' (dec.), [a]$" + 123' (chloroform), occurs to the extent of 0.003% in Haplophyton cimicidum, and is a monoacidic, amphoteric base containing one methoxyl but no 227.5 and 260 mp) and IR-spectra methylimino groups. Its UV- ,,A,(, resemble those of haplophytine, and i t is probable that the two alkaloids have several common structural features. Although not specifically proved, it can provisionally be assumed that the acidity of cimicidine stems from a cryptophenolic grouping of the type indicated in partial structure I. Thus, the IR-spectrum of cimicidine exhibits no hydroxyl absorption, but does exhibit a complex absorption pattern in the carbonyl region. Acetylation of cimicidine yields a readily hydrolyzed acetyl derivative, in which the band a t 1632 cm-1 is replaced by bands a t 1671 and 1774 cm-1 (phenolic acetate). The reduction of cimicidine by sodium borohydride exactly parallels that of haplophytine. The product is a hygroscopic, solvated dihydro derivative, whose composition approximates C23H30NzOsHzO ; this derivative does not exhibit the carbonyl absorption at 1751 cn-l, observed in the spectrum of cimicidine, but exhibits hydroxyl absorption. The nature of this carbonyl group is unknown; like haplophytine, cimicidine is inert toward carbonyl reagents. I n contrast to haplophytine, cimicidine does not react with cyanogen bromide, nor with neutral potassium permanganate at room temperature. Reaction with methyl iodide leads to a quaternary salt,

676

J. E. SAXTON

C24H31N20513, which corresponds to addition of one molecule of iodine to the methiodide. Hydrogenation of cimicidine by use of Adams' catalyst yields a dihydro derivative, obtained as its dihydrate from methanol and characterized as its picrate and hydrochloride. The reaction of cimicidine with alkaline hydrogen peroxide is also different from that of haplophytine; the product is an almost nonbasic, acidic substance of composition C23H28N206 (1, 3). The structures of haplophytine and cimicidine are still obscure; in a recent investigation of the constituents of H . cimicidum five additional bases were isolated, but no further data concerning haplophytine and cimicidine were reported ( 5 ) . The additional bases include the two epimers eburnamine (11;R = a-OH) and isoeburnamine (11; R = P-OH), which are characteristic alkaloids of Hunteria eburnea Pichon, and 0-methyleburnamine (11; R = a-OMe). The constitution of this last base was established by its conversion by boiling alcoholic yicric acid into the picrate of eburnamenine (III),and by its oxidation with chromic oxide in pyridine to the lactam eburnamonine (11; R = =O). Two new bases were also isolated, namely, haplocine and haplocidine.

11; R = a-OH Eburnamine 11; R = P-OH Isoeburnemine 11; R = wOMe 0 -Methyleburnamine

I11

Haplocine, C22H28N203, mp 186O-187', [a],, + 196' (chloroform), like haplophytine, is based on a 7-hydroxy-N-acylindoline nucleas, since its UV-spectrum has maxima a t 219.5, 258, and 292 mp, and its IRspectrum has maxima in the carbonyl region a t 1631, 1603, and 1570 cm-1; in contrast, 0-acetylhaplocine exhibits IR-absorption a t 1758 cm-1 (phenolic acetate) and 1667 cm-1 (amide carbonyl). The NMRspectrum of haplocine contains a strongly hydrogen-bonded hydroxyl group, and a triplet ( 3 protons) and quartet (2 protons) which form the ethyl group in an N-propionyl function no other C-methyl groups are present, nor are there any methoxyl or methylimino groups in the molecule. Reduction of haplocine by sodium borohydride or hydrogenation by use of a platinum catalyst gives a dihydrohaplocine, Cz~H30N203, mp 175O-176', [aID +117', which exhibits a UV-spectrum almost

18.

ALKALOIDS OF

Hupbophyton cimicidum

677

identical with that of haplocine and shows the same pattern of I R absorption in the carbonyl region. Dihydrohaplocine is identical with limaspermine (IV), an alkaloid of Aspidospermu limae Woodson (6); hence, haplocine, which does not contain an aliphatic hydroxyl group, must be the related carbinolamine ether (V; R = Et) (5).

IV

V; R = Et Haplocine V;R =Me Haplocidine

The companion base, haplocidine, C21H26N203, mp 183"-184", ["Iu (chloroform), exhibits I R - and UV-spectra closely similar to those of haplocine, and clearly contains the same functional groups. Its NMR-spectrum differs from that of haplocine only in that the quartet and triplet characteristic of the propionyl group are replaced by a sharp singlet ( 3 protons) attributable t o an N-acetyl function. That haplocidine (V; R = Me) is simply the N-acetyl compound related to haplocine is established by the prolonged alkaline hydrolysis of both alkaloids, which affords the same desacyl base (VI),C19H24N202, mp 250" (dec.). Acetylation of VI regenerates haplocidine, and propionylation regenerates haplocine ( 5 ) . Huplophyton cimicidum thus joins the interesting group of plants which produce both the aspidospermine and the biogenetically intimately related eburnamine types of alkaloids.

+ 231'

VI

VII; R = H Cimicine VII; R = O M e Cimicidine

678

J. E. SAXTON

Addendum Since the preceding account was written, the structures of cimicine, a newly discovered constituent of H . cimicidum, and cimicidine have been elucidated (7).Cimicine, CzzHzsNz04, mp 229"-231", [@II) + 113" (chloroform), exhibits IR- and NMR-spectra which show a close similarity to those of haplocine. Its NMR-spectrum contains absorptions owing to a strongly hydrogen-bonded hydroxyl group and an N-propionyl function ; no methoxyl or N-methyl groups are present. The IR-spectrum exhibits three bands in the carbonylregionat 1631,1603, and 1570cm-l, identical in position with corresponding bands in the spectrum of haplocine, assigned to the 7-hydroxy-N-propionylindolinechromophore. Two additional carbonyl bands are also present, a t 1770 and 1742 cm-l; these are attributed to a y-lactone grouping, the presence of which is presumably responsible for the solubility of cimicine in aqueous alkali (contrast haplocine). These data, and the relationship of the molecular formulas of cimicine and haplocine, suggest that cimicine has the structure V I I (R = H) ;this is established by the oxidation of haplocine with chromium trioxide in pyridine, which affords cimicine in low yield ( 7 ) . The molecular formula of cimicidine, C Z ~ H ~ ~ Nthe Z Opresence ~, of one methoxyl group, and the similarity of its IR-spectrum to that of cimicine, suggest that it is a methoxycimicine. This conclusion is confirmed by a comparison of the NMR-spectra of cimicidine and cimicine, which differ only in the presence of signals corresponding t o two vicinal aromatic protons and an aromatic methoxyl group in the spectrum of cirhicidine, whereas cimicine exhibits signals corresponding to three adjacent aromatic protons. Hence, the methoxyl group in cimicidine must be situated in position 14 or position 16. The latter is preferred, since to date no alkaloids possessing a methoxyl group in position 14 of the aspidospermine ring system have been isolated ;in contrast, alkaloids bearing oxygenated substituents a t positions 16 and 17 are well known. Hence, cimicidine (VII ; R = OMe)is regardedas 16-methoxycimicine (7). REFERENCES 1. E. F. Rogers, H. R. Snyder, and R. F. Fischer, J . Am. Chem. SOC.74, 198i (1952). 2. H. R. Snyder, R. F. Fischer, J. F. Walker, H. E. Els, and G. A. Nussberger,J. Am. Chem. SOC.76,2819 (1954). 3. H. R. Snyder, R. F. Fischer, J. F. Walker, H. E. Els, and G . A. Nussberger,J. Am. Chem. SOC.76, 4601 (1954). 4. H. R. Snyder, H. F. Strohmayer, and R . A. Mooney,J. Am. Chem. Soc. 80,3708 (1958). 5. M. P. Cava, S. K. Talapatra, K. Nomura, J. A . Weisbach, B . Douglas, and E. C. Shoop, Chem. Ind. (London)p. 1242 (1963). 6. M. Pinar, W. van Philipsborn, W. Vetter, and H. Schmid, Helu. Chim. Acta 45, 2260 (1962). i . M. P. Cava, S. K. Talapatra, P. Yates, M. Rosenberger, A. G. Szabo, B. Douglas, R. F. Raffauf, E. C. Shoop, and J. A. Weisbach, Chem. Ind. (London)p. 1875 (1963).

-CHAPTER

19-

THE ALKALOIDS OF GEISSOSPERMUM SPECIES R.H. F. MANSKEAND W. ASHLEYHARRISON Dominion Rubber Research Laborutory, Guelph, Ontarao, C w w d a

........

679

11. Geissoschizoline (Pereir

681

........................................ . . . . . . . . . . . . . . . . . . .................. .................................................... Geissospermine ...................................................... Other Alkaloids ..................................................... A. Vellosimine. . . . . . . . . . . . . . . . . . ................................ B. Vellosiminol . . . . . . . . . . . . . . . . . ........... .......

I. Introduction.

IV. V.

683 685

687 687 687

C. Geissolosimine (Alkaloid Dz) . . . . . . . . . . . . . . . . .

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

VI. Flavopereirine. References

690

I. Introduction The genus Geissospermum (Apocynaceae) is represented by only a few tropical and subtropical species and of these only two have been chemically examined. These are G. sericeum Benth. and Hook. f., indigenous to Guiana, and G. laeve (Vellozo) Baillon (synonymous with G. laeve Miers. and with 0.vellozii or G. vellosii Aliem.). The latter is the more thoroughly investigated plant and ail of the alkaloids described here have been isolated from it. The only alkaloid so far reported from G. sericeum is geissospermine. The alkaloids so far isolated along with some of their properties are given in Table I. Two of these alkaloids have been briefly discussed in Volume VII (p. 138) and the subsequent events are chronologically somewhat as follows : 1958. Rapoport and co-workers (1)isolated several new alkaloids, one of which was identical with geissoschizoline, one of the products obtained by scission of geissospermine with concentrated hydrochloric acid. They characterized two other cleavage products, namely, geissoschizine and apogeissoschizine, and showed that geissoschizoline has an indoline UV-spectrum while the others are indoles. 679

680

R. H. F. MANSKE AND W. ASHLEY HARRISON

Bertho et al. (2)showed that a, base, pereirine, which they isolated from

G.vellosii, was identical with one of the two cleavage products which they obtained from geissospermine. From spectral studies they concluded t'hat it was indolinic, whereas the other cleavage product was an indole. 1959. Janot and associates (3) obtained geissoschizoline and geissoschizine by hydrolysis of geissospermine with 2 N hydrochloric acid as Bertho had done. Geissoschizine on treatment with concentrated hydrochloric acid was converted into apogeissoschizine. Puisieux et al. (4) subsequently reported the conversion of geissoschizine into corynantheidol, thereby establishing its structure and absolute configuration. They also proposed a structure for apogeissoschizine and later published a review ( 5 ) . TABLE I ALKALOIDS ISOLATED FROM CeissospeTmum SPECIES Alkaloid Geissospermine Geissoschizoline (Pereirine) Flavopereirine Vellosine ( ? )"

Di Geissolosimine (Dz) Ei Vellosimine Vellosiminol

Formula

mP

[UID

2 13-2 14" 126"

- 101" (EtOH)

2 3 3-2 35' 189" 237-238" 140" 163-165" 305-306" 273-275"

Inactive

+ 32'

(EtOH)

-

+

70.4" (MeOH) -51" (EtOH)

+ 48" (MeOH) + 36' (MeOH)

a Isolation reported in the early literature (see Volume 11, p. 455) but not confirmed by more recent work (1, 2).

Contemporarily, Rapoport et al. (6) were able to establish the structure, but not the stereochemistry, of geissoschizine by dehydrogenation experiments. 1960. Rapoport et al. ( 7 ) gave a detailed account of some earlier work on geissoschizine and although they attempted its conversion to corynantheidol and were probably successfuI, their results were not entirely conclusive. They presented evidence supporting the structure proposed for apogeissoschizine. I n the meantime, Janot et al. (8) succeeded in relating geissoschizoline to akuammicine.

19.

THE ALKALOIDS

OB

Geissospermum SPECIES

681

1961. Bertho and Koll (9, 10) published independent evidence estalolishing the structure of pereirine (geissoschizoline).Puisieux and Le Hir (11) proposed a structure for geissospermine, and Janot (12) published a review on this alkaloid and its cleavage products (not including Bertho’s most recent work). Edwards and Smith (13) established the structure of tetrahydroakuammicine and found that it could be reduced to geissoschizoline. 1962. Rapoport and Moore (14) established the structures of three more Qeissospermum alkaloids. 11. Geissoschizoline (Pereirine) The name pereirine as used by Bertho et al. (2) is synonymous with geissoschizoline but it is not necessarily synonymous with amorphous and probably impure bases previously obtained from Geissospermum species (15). Geissoschizoline (11)contains a C-ethyl group, has a basic tertiary nitrogen, two active H atoms, and its UV-spectrum is typical of a

OL I

/

H

I

CHG

I

H O ~ H I\ ~ I1

CH~OH

CH~OH I11

IV

3,3-disubstituted indoline. It forms a diacetyl derivative whose IRspectrum exhibits acetoxyl and amide absorption and whose UVspectrum resembles that of strychnine ( 2 ) . Zinc dust distillation generates 3-ethylpyridine ( 2 ) , and Oppenauer oxidation gives an aldehyde whose UV-spectrum indicates the chromophore I which Karrer and co-workers recognized as that of fluorocurarine (16). These facts are not inconsistent with structure I1 for geissoschizoline.

682

R. H . F. MANSKE AND W. ASHLEY HARRISON

Oxidation of geissoschizoline with permanganate in acetone solution leads t o an unsaturated base formed by loss of only two hydrogen atoms. The IR-spectrum of this new base (C19H240N2) exhibits absorption characteristic of indolenines ( 1 7). The evidence relating to the position of the double bond in the new base is ambiguous and appears to require both partial structures I11 and IV. UV-absorption is consistent with the sum of both chromophores and this '' delocalization " of the double bond explains why the reactions carried out with this derivative generally proceed in more than one way (12). Another important derivative of geissoschizoline is obtained by first tosylating it (tosyl chloride in pyridine) and then reacting the amorphous tosylate with alcoholic alkali (12). The product is anhydrogeissoschizoline (C19H24N2, mp 128", [a]=+ 48")which differs from its progenitor only by loss of water. It cannot be catalytically hydrogenated with Adam's catalyst, nor can i t be acetylated. The IR-spectrum no longer shows the presence of an NH group and active hydrogen is absent from the new base. The suggested structure (V)is consistent with all spectral data, and furthermore, its formation sheds some light on the stereochemistry of geissoschizoline. An examination of models shows that formation of V is possible only if the C-2 and (3-16 hydrogens are cis and that its formation is facilitated if these hydrogens are trans to the one on C-15. Thering system is such that the configurations a t C-3 and (2-7 are fixed by that of C-15. The stereochemistry and absolute configuration of geissoschizoline were determined by relating it t o akuammicine (VI)and, in particular, t o 2,16-dihydroakuammicine (VII) and tetrahydroakuammicine (VIII). The establishment of the structure and stereochemistry of these compounds is discussed elsewhere in this volume. 2,16-Dihydroakuammicine (VII), obtained by reducing the natural base VI with zinc and sulfuric acid (13, 18), was further reduced with lithium aluminum hydride to the corresponding primary alcohol (1x1 and this, on catalytic reduction, generated geissoschizoline (8, 12). Geissoschizoline was also obtained by lithium aluminum hydride reduction of tetrahydroakuammicine (VIII) (13), which is prepared by catalytic reduction of VI. It was thereby established that geissoschizoline (11)has the same stereochemistry and absolute configuration as tetrahydroakuammicine. I n addition, it has been found (12) that catalytic reduction of the permanganate oxidation product of geissoschizoline (111 and/or IV) yields a base X that is also formed by catalytic reduction of the doubly unsaturated base XI, which is obtained from akuammicine on reduction with lithium aluminum hydride (18).

19.

THE ALKALOIDS OF

Geissospermum SPECIES

683

dCH3 VI

V

dCH3 VIII

X

IX

XI

111. Geissoschizine Geissoschizine (XII), C Z ~ H Z ~ (mp O ~ 194"-196", N~ ["ID + 115"), is the second cleavage product when geissospermine is treated with hydrochloric acid (3). It is amphoteric and has one 0-methyl and two active hydrogens and its IR-spectrum shows the presence of C=O, NH, and OH groups. The presence of an ethylidene group was demonstrated by

684

R. H. F. MANSICE AND W. ASHLEY HARRISON

ozonolysis and by Kuhn-Roth oxidation ( 7 ) . Its UV-spectrum relegates it to an indole structure upon which is superimposed a pH-sensitive chromophore. When geissoschizine is heated with dilute hydrochloric acid, it loses carbon dioxide, presumably after hydrolysis to its corresponding acid, to generate an amorphous base (C1gHzzONz2) (6, 7 ) , which on sodium borohydride reduction yields an alcohol (ClgH240Nz) and on WolffKishner reduction yields an oxygen-free base (ClgH24NZ). Since the former of these contains one C-methyl and the latter two C-methyl groups, their amorphous progenitor must he an aldehyde, probably formed by hydrolysis and loss of carbon dioxide from the chromophore, H, I ,C=C-C,

HO

/o OM0

I ( t o O=CH-CHz)

whose presence is indicated spectrally. Hydrolysis of geissoschizine with N Z ) by alcoholic alkali leads to an acidic product ( C ~ ~ H Z Z O Zformed hydrolysis of the ester function and loss of carbon monoxide, and this on reduction with lithium aluminum hydride gives the same alcohol (ClgH240N~) described above (4, 12).

XI1

Dehydrogenation experiments carried out on the oxygen-free base (CigH24N2)and on derivatives of geissospermine enabled Rapoport and co-workers (6, 7 ) to estabiish the structure of this alkaloid, but the

19.

THE ALKALOIDS OF

Qeissospermum

SPECIES

685

evidence did not establish its stereochemistry. The same structure as well as the stereochemistry of geissoschizine (XIL) was elucidated by Janot and co-workers (4)by the simple expedient of relating it to corynantheidine. The alcohol, ClgH240N2, obtained as above from geissoschizine, was reduced catalytically and the product, XIII, was identical with corynantheidol. Corynantheidine on acid hydrolysis undergoes reactions parallel to those already described for geissoschizine and the product, corynantheidal, on sodium borohydride reduction is converted into corynantheidol (XIII). The noncrystalline base, apogeissoschizine (C21H2202N2, mp of B.HC1, 145"), which is obtained from geissoschizine or from geissospermine by the action of concentrated hydrochloric acid, is formed by loss of water and retains the ester and ethylidine functions of the former. The aldehyde function and active hydrogens are absent so that structure XIV seemed a priori very probable (4,12). This has been confirmed by spectral studies of it and derivatives and by comparison with model compounds containing similar chromophores ( 7 ) .

IV. Geissospermine Geissospermine, C40H4803N4, contains one active hydrogen, a methyl ester group, and has a UV-spectrum consistent with the sum of the spectra of an indole and an indoline.

Unless deep-seated rearrangements have occurred in the cleavage fragments (I1and X I I ) of geissospermine during or subsequent to that cleavage, it is possible to envisage two structures for this alkaloid which take into account the fact that the chromophore Me02C-C=C-O of geissoschizine is absent (partial structures XV and XVI). Since the two

686

R . H. F. MANSKE AND W. ASHLEY HARRISON

fragments undergo partial recombination in 10% acetic acid (19), the possibility of rearrangements having occurred is improbable. Although formula XVI explains the presence of one active hydrogen, it does not accommodate the fact that no acetyl derivative can be prepared, whereas one can be prepared from geissoschizoline. Furthermore, the presence of a tetrahydrometoxazine ring as in XV receives confirmation from a study of the condensation of a simple aldehyde (acetaldehyde) with geissoschizoline (11). The latter in 10%

XVII

XVIII

XIX

xx

acetic acid reacts with acetaldehyde t o give XVII, which can be isolated as its crystalline methiodide (mp 245’) and which on hydrolysis with dilute hydrochloric acid regenerates its progenitors. Its structure follows from the fact that its reduction product (XVIII)with lithium aluminum hydride is identical with that obtained by the action of the same reducing agent on diacetylgeissoschizoline. The reduction of geissospermine with the same reducing agent leads to a parallel result with the formation of geissospermol (XIX), whose correct formula is C39H5002N4 (19) and not C39H4802N4 as previously proposed (3). Structure XX is consequently the accepted one for geissospermine.

19.

THE ALKALOIDS OF

Geissospermum

687

SPECIES

V. Other Alkaloids

A. VELLOSIMINE Vellosimine (14), ClgH200N2, has a UV-spectrum typical of an indole and its IR-spectrum indicates the presence of NH and an unconjugated aldehyde. The study of its NMR-spectrum in liquid sulfur dioxide indicated that the benzene nucleus of the indole portion was unsubstituted. The other features of structural fragments a, b, and c were also recognized by this means, taken in conjunction with the IR- and UVdata.

I 8

b

C

Wolff-Kishner reduction of vellosimine yields the oxygen-free base, (mp 307°-3080), whose NMR-spectrum indicates a proton on the carbon adjacent to the new methyl group and also suggests that the methyl group is directed away from the aromatic ring. On the basis of the foregoing data, and in analogy with other indole alkaloids, Rapoport and Moore (14) proposed structure XXI. C19H22N2

B. VELLOSIMINOL This naturally occurring base, ClgH~20N2,is obtainable from vellosimine by reduction of the aldehyde function to a primary alcohol by

XXI; R = CHO XXII;R = CH2OH

4

XXIII

R.

688

H. F. MANSKE AND W. ASHLEY HARRISON

means of sodium borohydride (14). Therefore, on the basis of the structure proposed for vellosimine, it was assigned structure XXII. This structure is also that of normacusine-B (20), of desformylakuammidinol (21),and of tombozine (22),and comparison ofthe reported physical data indicated that vellosiminol is in fact identical with those bases (14).

C. GEISSOLOSIMINE (ALKALOID D2) Geissolosimine, C38H440N4.BH20, in concentrated hydrochloric acid at room temperature is cleaved into vellosimine and geissoschizoline, and the fragments will recombine to yield the original base in dilute (1.5N ) acetic acid. Bearing in mind the structure ofgeissospermine and the ease with which geissoschizoline reacts with aldehydes, it follows that the structure of geissolosimine is XXIII. UV-, IR-, and NMR-spectroscopic data are consonant with this structure (14).

VI. Flavopereirine The occurrence and structure (XXVIIIb) of this alkaloid have been detailed (Volume VII, p. 139) (23). I n the meantime, no less than six separate syntheses of this base have been reported. Le Hir and co-workers (24)condensed tryptamine with BrCH2 . CHEt CHz .CH2. C02Et to obtain a moderate yield of compound XXIV which

.

XXIV

xxv

XXVI a,R=H b, R = Et

on ring closure with phosphorus oxychloride generated hexahydroflavopereirine (XXV). The last on dehydrogenation with palladium charcoal yielded flavopereirine (XXVIIIb). Prasad and Swan ( 2 5 ) first prepared compound XXVIa by refluxing the phenylhydrazone of 1,2,3,4 tetrahydro-1-oxopyridocoliniumbromide (XXVIIa) in ethanolic hydrogen chloride.

19.

THE ALKALOIDS OF

Geissospermum SPECIES

689

H XXVII a,R=H b, R = Et

XXVIII a,R=H b, R = Et b, Flevopersirine

XXIX R = Ph.CO b,R=H &,

Subsequent oxidation in acetic acid with chloranil yielded XXVIIIa whose spectra resemble very closely those of sempervirine. A parallel procedure via the phenylhydrazone of XXVIIb which was obtained by a known series of reactions from 2-cyano-5-ethylpyridine gave first XXVIb and finally XXVIIIb isolated as perchlorate. The last was identical with flavopereirine perchlorate of natural origin. Thesing and Festag (26) condensed a quaternary salt of gramine with N-phenacyl-3-ethylpyridinium bromide to yield XXIXa from which the benzoyl group was removed by the action of alcoholic alkali forming XXIXb. The latter was hydrogenated to a tetrahydropyridine compound which spontaneously cyclized to XXX and this in turn was dehydrogenated to flavopereirine. Ban and Seo (27) condensed 2-chloro-5-ethylpyridinewith 3-(2bromoethy1)-indole and thus obtained in one step the compound XXVIb which on dehydrogenation gave flavopereirine.

xxx Kaneko (28) followed essentially the method described by Prasad and Swan (25). The ketone XXVIIb was, however, prepared by a different route and its phenylhydrazone on heating with hydrogen bromide in ethanol yielded a bromine-containing product. This was first dehydrogenated with palladium charcoal in p-cymene and subsequently debrominated with palladium charcoal and hydrogen to form flavopereirine. Wenkert and his co-workers seem to have a special fondness for synthesizing flavopereirine, in that they reported no less than three separate routes. In the first (29), they heated the hydrochloride of

690

R. H . F. MANSKE AND W. ASHLEY HARRISON

compound XXXI with palladium charcoal at 300" for 20 minutes and obtained flavopereirine as well as its tetrahydro derivative and the obviously possible isomer. I n the second method (30), stress was placed upon obtaining compounds of the type of XXXI. Tryptophyl bromide readily reacts with pyridines to yield the corresponding quaternary compounds, which on reduction with sodium borohydride gave tetrahydropyridine derivatives. These could be fully reduced catalytically and ultimately dehydrogenated. The third method (31) utilized mercuric acetate as the agent which brought about oxidative coupling between the piperidine and the indole nucleus of compound XXXI. By-products and isomers are formed, but the fully hydrogenated (rings C and D) flavopereirine is accessible in practical yields. It can of course be dehydrogenated to flavopereirine. REFERENCES 1. H. Rapoport, T. P. Onak, N. A. Hughes, and M. G. Reinecke, J . Am. Chem. SOC.80, 1601 (1958). 2. A. Bertho, M. Koll, and M. I. Ferosie, Ber. 91, 2581 (1958). 3. M. M. Janot, R. Goutarel, A. Le Hir, and F. Puisieux, Compt. R e n d . Acad. Sci.248, 108 (1959). 4. F. Puisieux, R. Goutarel, M. M. Janot, and A. Le Hir, Compt. Rend. Acad. Sci. 249, 1369 (1959). 5. F. Puisieux, A. Le Hir, R. Goutarel, M. M. Janot, and J. Le Men, Ann. Pharm. Franc. 17, 626 (1959). 6. H. Rapoport, R. J. Windgassen, N. A. Hughes, and T. P. Onak, J . Am. Chem. SOC. 81, 3166 (1959). 7. H. Rapoport, R. J. Windgassen, Jr., N. A. Hughes, and T. P. Onak, J . Am. Chem. SOC.82, 4404 (1960). 8. M. M. Janot, J. Le Men, A. Le Hir, J. LBvy, and F. Puisieux, Compt. Rend. Acad. Sci. 250,4383 (1960). 9. A. Bertho and M. Koll, Naturwissemchaften 48, 49 (1961). 10. A. Bertho and M. Koll, Ber. 94, 2737 (1961). 11. F. Puisieux and A. Le Hir, Compt. Rend. Acad. Sci. 252, 902 (1961). 12. M. M. Janot, Tetrahedron 14, 113 (1961). 13. P. N. Edwards and G . F. Smith, J. Chem. SOC.p. 152 (1961). 14. H. Rapoport and R. E. Moore, J . Org. Chem. 27, 2981 (1962). 15. 0. Hesse, Ann. 284, 195 (1895). 16. W. v. Philipsborn, H. Meyer, H. Schmid, and P. Karrer, Helv. Chim. Acta 41, 1257 (1958). 17. K. Bernauer, W. Arnold, C . Weissmann, H. Schmid, and P. Karrer, Helw. Chim. Acta 43, 717 (1960). 18. J. LBvy, J. Le Men, and M. M. Janot, Bull. SOC.Chim. France p. 979 (1960). 19. F. Puisieux, R. Goutarel, M. M. Janot, J. Le Men, and A. Le Hir, Compt. R e n d . Acad. Sci.250, 1285 (1960). 20. A. R. Battersby and D. A. Yeowell, Proc. Chem. SOC. p. 17 (1961).

19.

THE ALKALOIDS O F

Geissospermum

SPECIES

69 1

J. LQvy, J. Le Men, and M. M. Janot, Compt. Rend. Acad. Sci. 253, 131 (1961). D. Stauffacher, Helu. Chim. Acta 44, 2006 (1961). N. A. Hughes and H. Rapoport, J. Am. Chem. SOC.80, 1604 (1958). A. Le Hir, M. M. Janot, and D. van Stolk, Bull. SOC.Chim. France p. 551 (1958). K. B. Prasad and G. A. Swan, J. Chern. SOC. p . 2024 (1958). J. Thesing and W. Festag, Ezperientia 15, 127 (1959). Y. Ban and M. Seo, Tetrahedron 16, 5 (1961). K. Kaneko, J. Pharm. SOC.Japun 80, 1374 (1960). E. Wenkert and J. Kilzer, J. Org. Chem. 27, 2283 (1962). E. Wenkert, R. A. Massy-Westropp, and R. G. Lewis, J. Am. Chem. SOC.84. 3732 (1962). 31. E. Wenkert and B. Wickberg, 3. Am. Chem. SOC.84, 4914 (1962).

21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

This Page Intentionally Left Blank

~ H A P T E R 20-

ALKALOIDS OF PSEUDOCINCHONA AND YOHIMBE* R . H . F. MANSKE Dominion Rubber Research Laboratory. Guelph. Ontario. Canada

. .

I Introduction

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

I1 Yohimbane ......................................................... A . Yohimbine and Its Isomers ........................................ B Syntheses ....................................................... C. ..Yohimbine .................................................... D . 3-Epicorynanthine................................................ E cr-Yohimbine (Rauwolsche)........................................ F. Alloyohimbine and 3-Epialloyohimbine . ............................. G. Seredine ........................................................

.

.

694 695 695 700 704 704 704 705 707

I11. Heteroyohimbane ................................................... A . Occurrence ...................................................... B . Stereochemistry .................................................. C Correlations ..................................................... D . Syntheses ....................................................... E . Biosyntheses ..................................................... F . Serpentinine ..................................................... G . Raumitorine ..................................................... H . Rauvanine ...................................................... I Raunitidine and Rauniticine ....................................... IV . Corynane (17,18-Secoyohimbane) . A . Occurrence ...................................................... B . Corynantheine ................................................... C . Correlation with Cinchona Alkaloids ................................

716 716 716 718

V . Corynoxane ........................................................ A Occurrence ...................................................... B . Structure ........................................................

720 720 721

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

721

.

.

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

.

References

707 707 708 710 713 713 713 714 714 715

* Much of this chapter is based upon a manuscript received from M.-M. Janot and J Poisson 693

.

.

694

R. H. F. MANSKE

I. Introduction The alkaloids of the two genera (family Rubiaceae) named in the title have been previously reviewed by L. Marion in Volume I1 and by J. E. Saxton in Volume VII. For the present purpose they are conveniently classified into four structural groups which are dealt with in the four following sections. The structures I-IV represent the basic skeleton of the yohimbane, heteroyohimbane, corynane, and corynoxane groups, respectively. The nomenclature is in part new, and no stereochemistry is suggested in these structures. Once the absolute configuration is established, the terms a and /3 have the meanings ascribed to them in steroids. The numbering is the same as in previous volumes. A number of alkaloids related to the above four groups occur in other plants, and particularly in the genus RauwolJia (Apocynaceae). They are treated in the relevant chapters of this volume.

Yohimbane

I1 Heteroyohimba,ne

111 Corynane

1V Corynoxane

I

Mass spectrometry has been studied in some members of this group of alkaloids ( 1 ) . Chart I represents the main fragments, and their molecular peaks, which have been identified from /3-yohimbine (2). Methoxyl derivatives show the expected changes, but stereochemical differences do not appear t o have pronounced effects upon the mode or extent of scission.

20. ALKALOIDS

OF

Pseudocinchona AND Yohimbe

a; 170

b ; 169

c ; 184

d ; 156

695

\n I '"

'I OH

e ; 295

OH f; 325

CHART I. Main fragments isolated from 8-yohimbine.

11. Yohimbane

A. YOHIMBINE AND ITSISOMERS Table I is a list of yohimbine and its known isomers along with their sources and some physical properties. The extensive isomerism arises from the fact that there are five ssymmetric centers in the yohimbane skeleton. The structure and absolute stereochemistry of yohimbine (V) have been known for some time (Volume VII, p. 50). The absolute configuration at C-17 has been determined by the method of Prelog, as follows: the phenylglyoxalate of yohimbine (VI), on treatment with methyl magnesium bromide, generated an ester (VII) which on hydrolysis gave a substantial yield of L-( + )-atrolactic acid (VIII) (3). Correlations have also been established between the rotatory dispersion curves of yohimbone (IX) and a tricyclic ketone (X) of known absolute configuration (4). The symmetry of the curves suggests a trans junction of rings D/E and the a-configuration at (2-15.

696

R . H. F. MANSKE

TABLE I AND ITSISOMERS YOHIMBINE

Type

Name

A110

La’

(pyridine)

quebrachine

235-236

+ 106

corynanthine

rauhimbine

225-226

- 85

8-yohimbine

amsonine

236-237

-54

268

+27

252-256

- 13

203-204

- 83

135-140

-84

pseudoyohimbine epi-3-corynanthine epi-3-8yohimbine alloyohimbine a-yohimbine

Epiallo

FP

( C)

Normal yohimbine

Pseudo

Synonyms

epi-3-alloyohimbine epi-3-ayohimbine

corynanthei- 235-236 dine rauwolscine 224 epi-3-rau225 wolscine isorauhimbine

- 18

+ 151 -104

Configuration Occurrence‘

3a, 15a, 208, 16a, 17a 3a, 15a, 208, 168, 17a 3a, 15a, 208, 16a, 178

5,13 1, 3, 5, 7, 8, 14, 15

38, 15a, 208, 16a, 17a 38, 15a, 208, 168, 17a 38, 15a, 208, 16a, 178

1,5

3a, 15a,20a, 168, 17a 38, 15a, 20a, 168, 17a

1,5

38,15a, 20a, 168, 178 38, 15a,20a, 168, 17a

Numbers refer to : 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

1-12

Pausinystalia yohimbe Pierre (Corynanthejohimbe K. Schum.) P . trillesii Beille Corynanthe paniculata Welw. C. macroceras K. Schum. Rauwolfia spp. Vincaspp. Aspidosperma quebracho-blanco Schlecht. Diplorhynchus condylocarpon Pich. Alchorea jloribunda Muell. -Arg. Lochnera lancea K. Schum. Pouteria sp. (Fam. Sapotaceae). Hunterk eburnea Pich. Pseudocinchona africuna A. Chev. Amsonia elliptica Roem. et Schult. Aspidosperma oblongum A.DC. Aktonia constricts F. Muell.

1, 5,6, 13, 16

5

20. ALKALOIDS

OF

Pseudocinchona

Yohirnbe

AND

697

0-CO-CO-Ph

VI

1

BrMgCHs

HOzC I HO-G-CH3

'"0

M e 0 zC.,. H80 t

I

Ph

VIII

CH3

I

0-CO--CPh I OH

VII

II

0 X

IX

It has been possible to correlate yohimbine with the corynane group by opening ring E (5). The reduction product (XI) of yohimbone (IX) with sodium borohydride was converted as shown to the corresponding chloride (XII).Elimination of hydrogen chloride from this and oxidation of the unsaturated base (XIII) yielded a dialdehyde (XIV) whose bissemicarbazone on reduction with hydrazine gave dihydrocorynantheane (XV),the absolute configuration of which had already been established (Vol. 7, p. 37).

IX

--i,4>A BH4Na

collidine'

-----+'

.

' OH f XI

POCls-Py

-'

1. OSO4 2. HI04

Y c1 XI1

XI11

/k

>

I

LHEHO XIV

698

R. H. I?. MANSKE

Isomerization of yohimbine a t position 3 is achieved by f i s t dehydrogenation at positions 3,4 with N-halogeno compounds (6,7) or with mercuric acetate ( 7 ) and then reduction with zinc and acetic acid. The

xv product is a mixture of regenerated yohimbine and pseudoyohimbine (6). Sodium borohydride reduction regenerates only yohimbine ( 8 ) , but catalytic reduction under favorable conditions will yield mostly pseudoyohimbine. Epimerization at C-3 to the extent of about 15% takes place when yohimbine is heated for prolonged times in acetic acid. The reverse epimerization is a practical one (8) and has been utilized in the analogous transformation of reserpine into isoreserpine (9). Other isomers of yohimbine, particularly those of the all0 series (15,20 c i s ) , behave somewhat differently (8). I n a specific case it has been possible t o achieve isomerization at the (3-15 to C-20 junction. Dehydration of yohimbic acid gives apoyohimbic acid (XVIb) which on subjection to the Schmidt reaction yields the ketone, ( + )-16-yohimbone (XVII) (10,ll). When the same series of reactions is applied to 3-epi-a-yohimbine (XVIII),there is first generated the ketone X I X which spontaneously epimerized to X X which is (-)-16-yohimbone (12). The epimerization of yohimbine a t C-17 to 8-yohimbine (XXI)without attack a t other asymmetric centers has been achieved by four separate routes. ( 1 ) Apoyohimbine (XVIa) is transformed by the action of sodium methylate into 17-O-methy1-/3-yohimbine(XXII), which on demethylation with hydrobromic acid, followed by esterification, generates /3yohimbine (12). (2) Yohimbic acid reacts with ethyl chloroformate to yield the /3-lactone XXIII, which reacts with lithium bromide in acetone to generate the bromoacid XXIV with inversion a t C-17. Esterification and replacement of the bromine by hydroxyl via acetyl (XXV)generated 8-yohimbine (13). (3) Oppenauer oxidation of yohimbine yields yohimbinone which on reduction with sodium borohydride gives /3-yohimbine, the preferential formation of which is rationalized on the

20. ALKALOIDS OF Pseudocinchona AND Yohimbe

699

H

XVI

XVII

a: R = CH3

xx

mp 254"-256", [=ID f 85"

mp 256", [=ID -86"l

b:R=H

OH XVIII

XIX

assumed greater stability of the hydroxyl in the equatorial conformation, rather than in the axial conformation of yohimbine (13,14). Yohimbinone is also obtainable by the oxidation of yohimbine with chromic oxide in acetone a t 0" (15). (4) The simplest method of epimerizing yohimbine to /3-yohimbine is to heat the alkaloid in benzene solution with potassium t-butylate for 1 hour (13).

XXIII

XXIV

xxv

700

R . H. F. MANSKE

B. SYNTHESES A number of syntheses of the yohimbane skeleton have been described. These include some leading to more or less unsaturated derivatives and finally to yohimbine by the timely introduction of the requisite five centers of asymmetry.

1. Derivatives of Hexadehydroyohimbane Several modifications of the Hahn-Werner method (16) consist in condensing tryptamine (XXVI) with a phenylpyruvic acid. 3-Hydroxy4-methoxyphenylpyruvic acid (XXVII) then yields a benzylharmane A -H J2-Jf

__f

OCCOzH

xxvl

D

O M I OH

OLQH

“I

I

OWMe

e

I OH

XXVII

XXVIII

OH XXIX

XXXII

xxx

XXXIII

XXXI

XXXIV

lH+

XXXI

20.

ALKALOIDS OF

Pseudocinchona AND Yohimbe

701

(XXVIII) which wher, >'- .ed by means of formaldehyde generates 17- hydroxy- 18-methoxy- 15,16,17,18,19,20-hexadehydroyohimbane (XXIX) (17). Yohimbanes with a variety of substituents in ring E have thus been prepared (18,19), some of which are simple but inactive models of reserpine. The condensation of isocoumarincarboxylic acid (XXX) with tryptamine gave unsatisfactory yields of the expected product, XXXI (20). The condensation of 3-acetylindole (XXXII) with isoquinoline in the presence of iodine generates a quaternary ketonic compound (XXXIII), and this upon reduction with lithium aluminum hydride furnishes the unstable enamine XXXIV which finally, in an acid medium, cyclizes to XXXI (21, 22).

2. Xempervirine This alkaloid (XXXV) from Gelsemium sernpervirens Ait. and from G . elegnas Benth. is an octadehydroyohimbine. Reaction of /3-bromethylindole (XXXVI) with a 3-halogenotetrahydroisoquinoline (XXXVII; x = Br or C1) in benzene solution in the presence of aluminum chloride leads to sempervirine (XXXV) in three stages (23).

XXXVII

XXXVI

xxxv

XXXVIII

- CQ I \3 0

H+

XL

xxxv

702

R . H. F. MANSKE

The Fischer indole synthesis has also been employed. The nitrile (XXXVIII) reacts with the Grignard reagent prepared from y-ethoxypropyl bromide to give an intermediate (XXXIX)which in the presence of acid generates the cyclic ketone XL. The phenylhydrazone of the last can be cyclized to sempervirine (XXXV) (24). Catalytic hydrogenation of sempervirine gives dl-alloyohimbine (Volume VII, p. 55).

3. dl- Yohimbane Analogous to the method of van Tamelen (Volume VII, p. 51) and following that due t o Corsano and Panizzi ( 2 5 ) ,a stereospecific synthesis of dl-yohimbane has been achieved ( 1 7 ) as shown in the sequence XLI to XLVI. Though none of the reactions are essentially new, the experimental skill necessary to manipulate often intractable substances was first-rate. Inversion of configuration at the stage XLII to XLIII was a welcome bonus.

v XLI

COzEt

COzEt

COzEt

1. SOCla 2. CHnNa

0

(COaEth +EtOK

XLII

XLIII

v

COzH

o=c’

COzEt

+

v

H=O

XLIV

H

v XXVI

XLV

\/ XLVI

4 . Yohimbine This synthesis, which involved nearly as many collaborators as there were stages, is an example of a “tour de force ” requiring strict attention t o stereochemistry as well as manipulative competence of a high order (26).The reactions are indicated in the sequence XLVII-LVIII (to VII), the starting material (XLVII) being obtained by reducing the condensation product of butadiene and p-quinone with zinc in an acid medium. There followed a Darzens condensation and after several more stages there resulted the chloride of an octalinonecarboxylic acid (L)which was condensed with tryptamine. The resulting amide was subjected in

20.

Pseudocinchona

ALKALOIDS OF

AND

Yohimbe

703

CO2Et

I

C0Cl

0

0

XLVII

XLVIII

0

XLIX

L

iv’l -Y

0

LI

XXVI

LIII

LIVa (R = 0 )

I

OH LV

LVI

LVII

LVIII Pseudoyohimbine

__f

V

704

R. H. F. MANSKE

succession to oxidation, reduction, oxidation again, condensation, and reduction, each with specific reagents as shown. The resultant product (LIVb) had one superfluous ring (F)in which, however, the atoms were so placed that future chemical manipulation proceeded as anticipated. The olefin LVI obtained from LIVb by conversion to the acetate (LV) followed by loss of acetic acid was subjected t o oxidation with a sequence of reagents, and final esterification generated dl-pseudoyohimbine (LVIII). Resolution of the last with ( - )-camphorsulfonic acid provided the d-isomer, easily transformed into yohimbine (V). A further synthesis of yohimbine starting from yohimbone (IX) has been reported (26). This introduction of the carbethoxyl by means of ethyl carbonate in the presence of sodium ethylate was stereospecific, as was the subsequent catalytic reduction to LX. Yohimbone had already been synthesized (Volume VII, p. 45).

IX

LIX

LX

C. jl-YOHIMBINE The stereochemistry of 13-yohimbine (XXI) has been discussed, epimerization of yohimbine a t the hydroxyl yielding the 13-isomer. Oppenauer oxidation of p-yohimbine generates yohimbinone ( 14).

D. 3-EPICORYNANTHINE This isomer (LXI) was obtained from corynanthine by the procedure used to obtain pseudoyohimbine from yohimbine (14). E. a-YOHIMBINE (RAUWOLSCINE) The structure and stereochemistry of this isomer (LXII) have been detailed (Volume VII, p. 58, 7 0 , 7 1 ) . Subsequent work on reserpine, deserpidine, alloyohimbane, and 3-epialloyohimbane completely confirms these assignments (27), but the absolute configuration is not yet certain (28,29).

20. ALKALOIDS

OF

Pseudocinchona AND Yohimbe

H

LXI 3-Epicorynanthine

'

705

I_,-

I

LXII a-Yohimbine

F. ALLOYOHIMBINE AND

3-EPIALLOYOHIMBINE

The close relation of alloyohimbine (LXIII) to a-yohimbine is proved by the observation that Oppenauer oxidation of either gives alloyohimbone (LXIV). Their difference is due to the orientation of the hydroxyl a t C - 1 8 , axial /Iin alloyohimbine and equatorial cc in a-yohimbine (Volume VII, p. 56). The latter configuration is the more stable since potassium t-butoxide in benzene transforms alloyohimbine into CCyohimbine ( 1 3 , 2 5 ) .

LXIII

LXIV

LXV

' Alloyohimbine

The stereochemistry a t positions 3,15, and 20 is preserved in alloyo, himbone (LXIV) and its reduction product, alloyohimbane ( ~ c c15a,20cryohimbane, LXV), of which several syntheses have been reported (Volume VII, p. 58) (30). I n a recent synthesis, tryptamine (XXVI) was condensed with 4-methoxyhomophthalic anhydride (LXVI) to the amide LXVII. This in the five stages shown was converted to LXVIII and the latter, through another series of reactions, converted to LXX consisting of two epimers which were separable. Tosylation of the hydroxyl and ultimate reduction with lithium aluminum hydride generated dl-alloyohimbane (LXV) (31). If the keto group of LXIX is blocked by ketolization, it is then

R . H. F. MANSKE

706

possible t o hydrogenate the product (palladium on charcoal) and form dl-alloyohimbone (LXIV) (32a). Oxidation of alloyohimbine with mercuric acetate leads t o a 3,4dehydro base, which on reduction with zinc and acetic acid is largely converted to 3-epialloyohimbine (LXXII) (14).

LXIX

LXVIII

1. NaBH4 2. NiRaney 3. resolution

OH(e\ LXX

LXXI

LXXII

20, ALKALOIDS OF Pseudocinchona

AND

Yohimbe

707

G. SEREDINE The roots of a variety of Rauwolfia vomitoria Afzel. collected in Guinea yielded a small amount of a base, seredine, C23H3005N2,mp 308", [.In- 1' (CHCls), and pK,' 6.69 (33). It contains three methoxyls, three active hydrogens, its UV-spectrum is of the 5,6-dimethoxyindole type, and its IR-spectrum is essentially that of a superposition of the IR-spectra of 5,6-dimethoxy-2,3-dimethylindole (LXXIII) and of a-yohimbine (LXXIVa). Oppenauer oxidation gives seredone (LXXVb) similar to alloyohimbone (LXXVa). A monoacetyl derivative may be prepared. Molecular rotation studies of derivatives of seredine and of a-yohimbine show a convincing parallelism, and furthermore the rotatory dispersion curves of the two bases are virtually identical. Seredine is therefore regarded as 10,11-dimethoxy-ayohimbine. The mass spectrum of seredone is identical with that of alloyohimbone except for a displacement of + 60 mass units (20Me) of the fragment containing the indole nucleus (1).

Me0yA-jcH3

Me0/6\

N H

'CH3

il

OH LXXIII

LXXIV a: R = H a-yohimbine b: R = OMe seredine

LXXV a:R=H b : R = OMe

111. Heteroyobbane

A. OCCURRENCE The natural representative of this group is 8-yohimbine or ajmalicine first found in Yohimbe and in Rauwoljia serpentina Benth. ex Kurz (Volume VII). Since 1957 five new alkaloids belonging to this group have been described : rauvanine, raunitidine, rauniticine, neoreserpiline, and holoinine. Other sources, as given, have also been found for : ajmalicine, RauwolJia sumatrana (Miq.) Jack. (33) and R. javanica Koord et Val (34); tetrahydroalstonine, Lochnera Zancea Boj. (ex A.DC.) (35) and

708

R. H. F. MANSKE

Alstonia constricta F. Muell. (36); alstonine, R. vomitoria (37) ; aricine, Aspidosperma marcgravianum Woodson (37); reserpiline, R. decurva Hook. (38); isoreserpiline, R. cambodiana Pierre (39), R. decurva, and Ochrosia poweri F. M. Bailey; serpentinine, R. vomitoria (37) and R. javanica (34). Of the five new alkaloids, the first four are tertiary bases. Holoeinine is a quaternary base from R. samdwicensis A.DC. identified as the Nb methyl quaternary derivative of isoreserpiline (40). [a],,- 78' (EtOH), from Neoreserpiline, C Z ~ H ~ ~ Omp ~ N120'-131', Z, R. peralcensis, is a dimethoxyheteroyohimbane of unknown structure (41). The stereochemistry of the heteroyohimbanes has largely been elucidated, and a total synthesis of ajmalicine has been achieved. Mass spectra of these alkaloids give useful structural information. The molecular ion, LXXVI, of ajmalicine is characteristic. The other fragments represent the indole portions following rupture and loss of ring E (1).

B. STEREOCHEMISTRY Aside from the fact that some of the heteroyohimbanes (LXXVII) carry substituents a t positions 10 and/or 11, they differ in their configuration a t the four centers of assymetry, 3, 15, 30, and 19, to which may be added the possibility of conformational isomerism. Table I1 summarizes the eight possible types of isomerism, their relations to the corresponding yohimbanes, and the alkaloids known in each type. The solutions of the various stereochemical problems were arrived at by a combination of methods, which follow.

1. Position 3 and Junction of Rings C / D This is in fact a quinolizidine system, the stereochemistry of which has been reviewed by Bohlmann (42). The configuration of C-3 must be considered in its relation to that of Nb where the bond C-3 to Nb is defined as either cis or trans with reference to the bond C-3 to H and the free electronic doublet of Nb. a. The IR-spectra of bases with a C/D trans junction show bands at 2940 and 2700 em-1 absent from the corresponding cis bases (43, 44). b. The NMR-spectra of the Nb methiodides are somewhat dependent upon the configuration of C-3 (45). The protons of (+)-NCH, have a signal at much lower field values in the C/D cis than in the C/D trans compounds (44).

20. ALKALOIDS OF Pseudocinchona

AND

Yohimbe

709

c . The NMR-spectra of the tertiary C/D cis bases show a very sharp signal a t 4.5 ppm absent from those with a C/D trans junction (46). d. The reactions which lead to isomerization in the yohimbanes are equally applicable here (8). TABLE I1 HETEROYORIMBANES LXXVII

Corre-

Junc- Juncsponding CH3 tion tion Type yohimbane (19) C/D D/E type

Alkaloids

A

A110

trans

cis

Tetrahydroalstonine Aricine Reserpinine Isoreserpiline

B

Epiallo

cis

cis

Akuammigine Isoreserpinine Reserpiline

C

A110

trans

cis

Rauniticine Raunitidine

D

Epiallo

trans

cis

Mayumbine Isoraunitidine

E

Normal

trans

trans

Ajmalicine

R1

Rz

Natural corresponding quaternary bases

-c Tetraphylline

F

Normal

trans

trans

Raumitorine Rauvanine

G

Pseudo

cis

trans

Epi-3-sjmalicine

H

Pseudo

cis

trans

(I

f

Epimeric at (2-20 of serpentine XC. Chloride of N, methylisoreserpilinium. Absolute configuration, 15a,20g (ajmalicine) (50). Tetradehydro-3,4,5,6-ajmalicineXC. Compare XCI. Absolute configuration (47)(ajmalicine, serpentine).

2. Positions 15 and 20 and Junction of Rings DIE a. There is sc. r - 2 indication that the rate of dehydrogenation of rings C and D with palladium and maleic acid is dependant upon the nature of

710

R . H. F. MANSKE

their junction. Compounds of the type 3-epiallo (C/D cis-DIE trans dehydrogenate very slowly (8). b. I n general the pK values show that bases of the C/D trans-DE cis type, corresponding to the alloyohimbanes, are the least basic, although the differences are not always unequivocal (44). c. On the other hand the rates of formation of the methiodides are greatly influenced by the spatial environment around Nb. Weak bases of the tetrahydroalstonine type have the lowest velocity of reaction. The higher rates are observed in the ajmalicine type (C/D trans-DIE trans, corresponding to the “normal” yohimbanes) in which the Nb is freely accessible (44). d. I n the NMR-spectra the coupling constants of the protons a t positions 19 and 20 are influenced by their relative configuration so that it is not possible to ascertain the configuration of one unless the other is known (46). e. The IR-spectra in the 1200 cm-1 region point to the presence of four different heteroyohimbane skeletons, but their relegation to specifictypes does not seem possible.

3. Position 19 a. The orientation of the methyl at C-19 presents a somewhat difficult problem. If the configuration of the rest of the molecule is known, NMRspectral studies of methiodides give useful information. b. The rate of methiodide formation is influenced by the proximity of the’C-19 methyl (44). c. The absolute configuration a t C-19 has been determined in both serpentine and ajmalicine (py-tetrahydroserpentine) by transforming ring E into a five-membered lactone. Hudson’s rule was applied to the determination of its stereostructure, the results being in agreement with speculations based on NMR-studies (47). I n support of the structures there determined, it has been possible to effect a number of transformations by methods already detailed for the yohimbanes. The following may be mentioned ; tetrahydroalstonine to akuammigine (8); reserpiline to isoreserpiline and the reverse (48, 49); and ajmalicine to isoajmalicine (8). C. CORRELATIONS Ajmalicine (LXXVIIIa) has been correlated with the bases in which ring E is not closed. Saponification of the alkaloid and decarboxylation of the acid, LXXVIIIb, yielded the hemiacetal LXXIX which on

20.

ALKALOIDS OF

Pseudocinchona

AND

Yohimbe

711

Wolff-Kishner reduction gave the open-chain alcohol LXXX. Elimination of the hydroxyl via the corresponding ketone generated dihydrocorynantheane (XV) (50). Contrary to earlier belief (50), epimerization at C-20 does not occur during the Oppenauer oxidation of LXXX (44,46).

LXXVI

LXXVII

OH LXXVIII a: R = CHn aimalicine b:R=H

LXXIX

LXXX 1. Oppenauer 2. Wolff-Kishner

xv The absolute configuration of ajmalicine is thus established through its conversion to dihydrocorynantheane, which itself has been related to cinchonamine (50-53). By the same series of reactions, tetrahydroalstonine (LXXXI) has been related to dihydrocorynantheane (46), but this time by reversal of the configuration a t C-20 by the Oppenauer oxidation on the open-chain alcohol, the configuration equatorial being the more stable. Some naturally occurring oxindoles are related to the heteroyohimbanes in that they have rings D and E in common. I n consequence, the

712

R . H. F. MANSKE

oxidation of ajmalicine (LXXVIIIa) with t-butyl hypochlorite gives a mixture of mitraphylline and isomitraphylline (LXXXII, epimers a t position 7 ) (54-56). Similarly, it has been shown that carapanaubine from Aspidosperma curupanauba is one of the oxindoles derived from isoreserpiline (57,58). An examination of the NMR-spectra has indicated that mayumbine from Pseudocinchona mayumbensis R-Hamet (Volume VII, p. 6 1) is related to formosanine (59).

LXXXII Mitraphyllines

MeOzCYI LXXXIII

LXXXIV

LXXXV

LXXXVIII

LXXXVII

LXXXVI

J. ( d l )-Ajmalicine

20. ALKALOIDS

OF

Pseudocinchona

AND

Yohimbe

713

D. SYNTHESES The condensation of methyl acetoacetate with dimethyl glutaconate generates compound LXXXIII and this, when reacted with tryptamine (XXVI) in the presence of formaldehyde, gives the amide LXXXIV, on which ring closure is effected by the action of POC13 to LXXXV. When subjected to a series of reactions involving catalytic reduction, elimination of carboxyl from p-ketonic ester, sodium borohydride reduction (LXXXVI), lactone formation brought about by dicyclohexylcarbodiimide (to LXXXVII), and condensation with methyl formate to a hydroxymethylene derivative (LXXXVIII),LXXXV ultimately yields dl-ajmalicine by rearrangement in an acid medium (52). An abortive attempt to synthesize alstonine has been reported (60).

E. BIOSYNTHESES That tryptamine, or more correctly, tryptophan, is the biogenetic precursor furnishing the tryptamine moiety has been shown by feeding tryptophan-(2)C14 (LXXXIX) to RauwolJia serpentina seedlings. The Serpentine (XC) which was isolated was shown by appropriate degradation to have the C14 a t position 5 (61).

F. SERPENTININE The structure of this alkaloid, first found in R. serpentina and subsequently in other related species, has been somewhat of an enigma although a close relation to serpentine was suspected (Volume VII, p. 96). Following a tentative suggestion that it is a hydroxyserpentine (62), it was shown to have twice the molecular weight ( C ~ Z H ~ ~ O ~ N ~ H Z O ) resulting from a coupling of serpentine and ajmalicine (63).The structure XCI is a satisfactory representation of all of its known reactions and accords with an exhaustive examination of its NMR-spectrum.

Q-!75’ rJQ9:: \

H

Iq H

,&\

Lb,,cH

MeOzC LXXXIX

\/o

xc

Serpentine

714

R . H. F. MANSKE

XCI Serpentinine

XCII a: R1 = OCH8; Rz = H, raumitorine b : R1,Rz = OCH3, rauvanine

G. RAUMITORINE The structure (XCIIa) of this alkaloid has long been known (Volume 11, p. 100) (63,64).Its UV- and IR-spectra as well as a number of chemical reactions clearly indicate that it is 10-methoxyajmalicine. Selenium dehydrogenation yields 5-methoxyalstyrine and the stereochemical criteria as already outlined confirm C/D and D/E trans junctions and a, C-19 equatorial methyl.

H. RAUVANINE Rauvanine, C23H2805N2.5H20, mp 129"-135", [a]=+ 32 (CHC13) has been isolated from R.vomitoria (65). It contains three methoxyls and its UV-, IR-, and NMR-spectra indicate clearly that it is a 5,6-dimethoxy heteroyohimbane. Stereochemical criteria relegate it to ajmalicine and consequently its structure is XCIIb, indicating a very close relation $0 raumatorine which accompanies it in the plant (65,48).

20.

ALKALOIDS OF

Pseudocinchona AND Yohimbe

715

I. RAUNITIDINE AND RAUNITICINE Two new alkaloids have been isolated from R. nitida Jaeq. which show spectra characteristic of heteroyohimbanes (49): raunitidine, C22Hz&4Nz, mp 276"-278", [aID - 30.6 (py); and rauniticine, C21H2403N2, mp 233"-235", [.ID - 6.6 (py). .COzMe

OMe

XCIV

I R

XCIII a : R = OMe, ramitidme b : R = H. rauniticine

LXXXI

Selenium dehydrogenation of rauniticine gives alstyrine similarly obtained from ajmalicine. These alkaloids differ only in that raunitidine is a monomethoxyrauniticine. Their IR-spectra (3.4p ) indicates a C/D trans junction with the C-3-H axial. That the D/E junction is cis follows

716

R . H. F. MANSKE

from a study of methylation rates a t N,, from their NMR-spectra, and from the irreversible epimerization in acid media. However, these data do not agree unequivocally with an all0 type structure as exemplified by tetrahydroalstonine (LXXXI). Consequently, the methyl a t C-19 is assigned the axial configuration since this tends to some shielding of Nb and accordingly raunitidine is XCIIIa and rauniticine is XCIIIb. Raunitidine when treated with acetic anhydride yields isoraunitidine, mp 259'-261') [ a ] D + 131 (py), a change which has generated a C/D cis junction with retention of configuration a t all other centers. The axial bonded H a t C-3 is now /3 and axial in the isomer so that it is in fact 3-epiraunitidine (XCIV) (49). It differs therefore from other bases of the type 3-epiallo with a C/D cis junction and a C-3 to H equatorial confirmation exemplified by reserpiline (44).

IV. Corynane (17,18-Secoyohimbane)

A. OCCURRENCE To the three alkaloids of this group already known (corynantheine, corynantheidine, and dihydrocorynantheine) three additional ones have now been isolated as follows : dihydrocorynantheal (XCV), ClgHz60N2, mp 181°-1830, [.ID -19 (CHClS), from A . marcgravianum and A . auriculatum Markg. (34); N,-methyldihydrocorynantheol (XCVI), B .HC1, CzoH290NzC1, mp 296"-297") [aID + 101 from Hunteria eburnea Pichon (66,67); and demethyldihydrocorynantheine (XCVII), C2lHZS03NZjmp 174', [ a ] ~ - 84 (py) (48,68).*

B. CORYNANTHEINE The total structures of this alkaloid (XCVIIIa) and its dihydroderivatives (XCVIIIb) as well as a synthesis of the former have been reviewed (Vol. VII, p. 42). I n the meantime, a total synthesis of dldihydrocorynantheine (CIIb) has been reported (69,70). The condensation of tryptamine (XXVI) with the glutaric ester XCIX in the presence of Raney nickel and hydrogen yields a pair of isomeric lactams (Ca and Cb) of which the former on cyclization with phosphorus oxychloride followed by catalytic reduction gave the

* Geissoschizine, a scission product of geissospermine, belongs t o this group of bases (see Chapterl9).

20. ALKALOIDS

OF

Pseudocinchona AND Yohimbe

717

c1"

A-/\

\

\CH~OH xcv

CHaOH

XCVI Nb-Methyldihydrocoryriant heol

Dihydrocorynantheol

XCVIII Corynantheine a: R = -CH=CHz b : R = -CHz-CHa

XCVII

EtO2C' CI

EtOZC'\ CII a: R = O H

b: R = OM0

CHR

718

R. H. F. MANSKE

pyridocoline CI. Finally, the last was condensed with methyl formate in the presence of sodium triphenylmethyl to the oxymethylene compound CIIa, which on methylation was transformed into dihydrocorynantheine CIIb.

WITH Cinchona ALKALOIDS C. CORRELATION

Dihydrocinchonamine (CIIIb) is tosylated to CIV which spontaneously quaternizes t o CV ( X = C1, changed to TSO by reacting with AgTSO). The same quaternary base is obtained by refluxing the tosylate of dihydrocorynantheol (XCV) in dimethylformamide (50).

CIII

cv

CIV

Cinchonamine a: R = CH=CHZ b : R = CHz-CHs

\

CHzOR

xcv

R = Ts

Cinchonine (CVI) can be transformed into dihydrocorynantheane (XV) by a series of reactions which first changes it into 9-benzoyl-2oxyhexahydrocinchonine (CVII). The latter reacts with cyanogen bromide to give CVIII and this in turn under alkaline conditions suffers reduction and rearrangement to CIX, which on further reduction with lithium aluminum hydride followed finally by Oppenauer oxidation generates 3-epidihydrocorynantheane (CX). This can be isomerized by known means t o dihydrocorynantheane (XV) ( 7 1). The conversion of quinine (CXI)to 10-methoxydihydrocorynantheane (CXII) by the above series of reactions has also been described (72).

20. ALKALOIDS

OF

Pseudocinchona

AND

Yohimbe

719

CVII

CVI Cinchonine

CVIII

CIX

cx Since the configuration of the Cinchona alkaloids is known (73), the above correlations confirm the configurations of dihydrocorynantheine, of corynantheine (3a,15a, 20p), and of corynantheidine (3a,15a, 20a) epimeric at C-20 of dihydrocorynantheine (74). The pair corynantheine-corynantheidineplays an important role in that a number of alkaloids of other structural types can be transformed into derivatives of either of this pair. The configuration a of C-15,H can thus be assigned t o ajmaline (CXIII) because it can be transformed into two isomers of CXV, one of which on further transformation yields a tetracyclic derivative (CXV) related to corynantheidine (75). The same sequence of reactions applied to isoajmaline (epimeric a t (2-20 with ajmaline) gives a derivative of corynantheine ( 7 5 ) . Sarpagine (CXVI)

720

R. H. P. MANSKE

L CXI

'

CXII

Quinine

CXIII

CXIV

cxv

Ajmaline

CXVI Sarpagine

can also be transformed via a deoxyajmalol (CXIV) into a corynantheidine derivative (75).

V. Corynoxane A. OCCURRENCE Two new bases were isolated from some residues derived from Pseudocinchona africana A. Chev. : corynoxeine, CzzHzs04Nz, mp 210°, [ a ] ~ + 23 (py),containing two methoxyls but neither NMe nor CMe; hydrogenation produces dihydrocorynoxeine, CzzHzg04Nz, mp 2 lo", [elD - 17 (CHC13), containing an ethyl side chain ; corynoxine, CzzHzs04Nz, mp 166°-1680, [elD - 14 (py), containing two methoxyls, no NMe, but one ethyl side chain (76).

20. ALKALOIDS

OF

Pseudocinchona A N D Yohimbe

721

B. STRUCTURE The UV- and IR-spectra indicate the presence of an a-oxindole chromophore superimposable without conjugation upon the MeO&=COR chromophore characteristic of those of corynantheine and of ajmalicine. Hydrolysis followed by Wolff-Kishner reduction of dihydrocorynoxeine and of corynoxine yields two bases, dihydrocorynoxeinine (CXIX) and corynoxinine (CXX), respectively, which possess only the oxindole chromophore. The former base, dihydrocorynoxeine, has been shown to be identical with rhynchophylline (CXVIIb) (77), which possesses the ethyl side chain and hence corynoxeine (CXVIIa) has the vinyl chain (76). The parallelism between this pair of alkaloids and the corynantheinecorynantheidine pair suggests that the latter may be convertible into the former. This has in fact been achieved and consequently the structures assigned to corynoxeine (CXVIIa) and to corynoxine (CXVIII) are the correct ones (54, 32a).

CXVII corynoxeine b: R = -CH~-CHI, rhynchophylline a: R = -CH=CHz,

CXVIII Corynoxine

cxx

CXIX

REFERENCES 1. I. D. Antonaccio, N. A. Pereira, B. Gilbert, H. Vorbrueggen, H. Budzikiewicz, J. M. Wilson, L. J. Durham, and C. Djerassi, J . Am. Chem. SOC.84, 2161 (1962). 2. G. Spiteller and M. Spiteller-Friedmann, Monatsh. Chem. 93, 795 (1962). 3. Y. Ban and 0. Yonemitsu, Chem. I d . (London)p. 948 (1961). 4. C. Djerassi, R. Riniker, and B. Riniker, J . Am. Chem. SOC. 78, 6362 (1956). 5. Y. Ban and 0. Yonemitsu, Tetrahedron Letters No. 5 , 181, (1962). 6. W. 0. Godtfredsen and S. Vangedal, Acta Chem. Scand. 10, 1414 (1956). 7. Danish Patent 87,658 (1959); Chem. Abstr. 54, 7765f (1960).

722

R . H. F. MANSKE

E. Wenkert and D. K. Roychaudhuri, J . Am. Chcm. Soc. 80, 1613 (1958). E. Wenkert and L. H. Liu, Ezperientia, 11, 302 (1955). E. Wenkert, E. W. Robb, and N. V. Bringi, J . Am. Chem. SOC.79, 6570 (1957). R. K. Hill and K. Muench, J . Org. Chem. 22, 1276 (1957). W. 0. Godtfredsen and S. Vangedal, Acta Chem. S c a n d . 11, 1013 (1957). A. Le Hir and E. W. Warnhoff, Compt. R e n d . Acad. Sci.246, 1564 (1958). M. M. Janot, R. Goutarel, E. W. Warnhoff, and A. Le Hir, Bull. Soc. Chim. France p. 637 (1961). 15. F. Puisieux, Personal comrnunication (1963). 16. G. Hahn and H. Werner, Ann. 520, 123 (193.5). 17. S. Corsano, A. Romeo, and L. Panizzi, R i c . Sci. 28, 2274 (1958); Chem. Abstr. 53, 20107e (1959). 18. A. Buzas, C. Hoffmann, and G. Rcgnier, Bull. Soc. Chim. France p. 643 (1960). 19. M. Onda and M. Kawanishi, 1.Phurm. Soc. J a p a n 76, 966 (1956); Chem. Abstr. 51, 2824d (1967). 20. C. Ribbens, Sci. Cornmun. Reseurch Uept. N . 17. Koninkl. Pharm. FabrieEen vlh Broendes-Stheeman Pharrnacia. 10, 9 (1960-1) ; Chem. Abstr. 56, 737% (1962). 21. D. R. Liljegren and K. T. Potts, Proc. Chem. Soc. p. 340 (1960). 22. D. R. Liljegren and K. T. Potts, J . Org. Chem. 27, 377 (1962). 23. Y. Ban and M. Seo, Telrahedron 16, 11 (1961). 24. G. B. Kline, J . Am. Chem. SOC. 81, 2251 (1959). 25. S. Corsano and L. Panizzi, A n n . Chim. ( R o m e ) 48, 1025 (1958). 26. E. E. Van Tamelen, M. Shamma, A. W. Burgstahler, J. Wolinsky, R. Tamm, and P. E. Aldrich, J . Am. Chem. Soc. 80,5006 (1958). 27. P. E. Aldrich, P. A. Diassi, D. F. Diekel, C. M. Dylion, P. D. Hance, C. F. Huebner, B. Korzun, M. E. Kuehne, L. H. Liu, H. R. Macphillamy, E. W. Robb, D. K. Roychaudhuri, E. Schlittlcr, A. I?. St. Andre, E. E . Van Tamelen, F. L. Weisenborn, E. Wenkert, and 0. U‘intersteiner, J . Am. C‘hem. Soc. 81, 2481 (1959). 28. G. B. Mitra, S. K. Mitra, and M. G. Basak, 2. Krist. 108, 385 (1957); 110, 165 (1958). 29. L. Ray, I n d i a n J . Phann. 18, 199 (1986); Chem. Abstr. 52, 5079h (1958). 30. G. Stork and R. K. Hill, J . Am. Clzem. Soe. 79, 495 (1957). 31. K. Naito and 0 . Nagase,J. Phann. Soc. J a p a n 80, 629 (1960); Chem. Abstr. 54,22700d (1960). 32. J. Poisson, Unpublished data (1963). 33. A. Chaudhury and A. Chatterjee, J . Sci. I n d . R e s . ( I n d i a ) 18B, 398 (1959); Chem. Abstr. 54, 7070i (1960). 34. S. K. Talapatra, J . Sci. I n d . Res. ( I n d i a ) 21B, 198 (1962); Chem. Abstr. 57, 11555d (1962). 35. M. M. Janot, J. Le Men, and Y . Gabbai, Ann. Pharm. Franc. 15,474 (1957). 36. G. H. Svoboda, J . Am. Pharm. Assoc. Sci. Ed. 46.508 (1957). 37. S. Siddiqui and M. Manzur-I-Khuda, P a k i s t a n J . Sci. I n d . R e s . 4, 1 (1961); Chem. Abstr. 56,27158 (1962). 38. G. K. Atal, J . Am. Pharm. Assoc. 48, 37 (1959). 39. D. A. A. Kidd, J . Chem. SOC.p. 2432 (1958). 40. J. Scheuer and J. T. Hamamoto-Metzger, J . Org. Cliem. 26, 3069 (1961). 41. K. A. Kim, in “Collection of Physical Data of Indole and Dihydroindole Alkaloids” (N. Neuss, ed.), Vol. 11, 1962. Eli Lilly and Co., Indianapolis, Indiana. 42. F. Bohlmann, Angew. C h e m . 69, 641 (1957). 43. M. Shamma and J. B. Moss, J . Am. Chem. SOC.83, 5038 (1961). 44. M. Shamma and J. B. Moss, J . Am. Chem. SOC.84, 1739 (1962). 8. 9. 10. 11. 12. 13. 14.

20. ALKALOIDS

OF

Pseudocinchona

AND

Yohimbe

723

45. T. M. Moynehan, K. Schofield, R. A. Y. Jones, and A. It. Katritzky, Proc. C'hem. SOC. p. 218 (1961); J. Chem.Soc. p. 2637 (1962). 46. E. Wenkert, B. Wickberg, and C. L. Leicht, J . Am. Chem. SOC. 83, 5037 (1961). 47. H. Fritz, Ann. 655, 148 (1962). 48. J. Poisson and R.Bergoeing, Unpublished data (1963). 49. R. Salkin, N. Hosanky, and R. J a r e t , J . Pharm. Sci. 50, 1038 (1961). 50. E. Wenkert and N. V. Bringi, J . Am. C'hem. SOC.80, 3484 (1958); 81, 1474 (1959). 51. N. News and H. E. Boaz, J . Org. Chem. 22, 1001 (1957). 52. E. E. Van Tamelen and C. Placeway, J . Am. Chem. Soc. 83, 2594 (1961). 53. E. Ochiai and M. Isjikawa, Tetrahedron 7, 228 (1959). 84. N. Finch and W. I. Taylor, J . Am. Chem. SOC. 84, 3871 (1962). 55. J. B. Hendrickson, J . Am. Chem. SOC. 84, 650 (1962). 56. J. Shave1 arid H. Zinnes, J . Am. Chem. SOC.84, 1320 (1962). 57. B. Gilbert, J. Aguayo Brissolese, N. Finch, W. I. Taylor, H. Budzikiewicz, J. M. Wilson, and C. Djerassi, J . Am. Chem. SOC.85, 1523 (1963). 58. N. Finch, C. W. Gemenden, I. H. C. Hsu, and W. I. Taylor,J. Am. Chem.Soc. 85,1520 (1963). 59. E. Wenkert, B. Wickberg, and C. Leicht, Tetruhderon Letters p. 822 (1961). 60. S. L. Reid, Dissertation Abstr. 19, 1208 (1958). 61. E. Leete, Tetrahedron 14, 35 (1961). 62. A. Chatterjee and S. Bose, Sci. and Cult. (Calcutta) 25, 84 (1959); Chem. Abstr. 54, 11065b (1960). 63. J. Poisson, A. Le Hir, R. Goutarel, and M. M. Janot, Compt. Rend. Acad. Sci. 239, 302 ( 1954). 64. J. Poisson, Thesis, Univ. Paris (1958). 65. R. Goutarel, M. Gut, and J . Parello, Compt. Rend. Acad. Sci. 253, 2589 (1961). 66. J. D. M. Asher, J. M. Robertson, G. A. Sim, M. F. Bartlett, R. Sklar, and W. I. Taylor, Proc. Chem. SOC. p. 72 (1962). 67. M. F. Bartlett, B. Korzun, R. Sklar, A. F. Smith, and W. I. Taylor, J . Org. Chem. 128, 1445 (1963). 68. T. H. Van Dcr Meulen, Thesis, Univ. Leiden (1962). 69. E. Van Tamelen, P. E. Aldrich, and T. J. Katz, J . Am. Chem. SOC.79, 6426 (1957). 70. E. Van Tamelen and J. B. Hester, Jr., J . Am. Chem. Soc. 81,3805 ( 1 959). 71. E. Ochiai andM. Isjikawa, Chem. Pharm. Bull. (Tokyo) 7, 386 (1959); Chem. Abstr. 55, 5554d (1961). 72. E. Ochiai and M. Ishikawa, Chern. Phrcrm. Bull. ( T o k y o )7,256 (1959); Chcm. Abstr. 55, 5553e (1961). 73. V. Prelog and E. Zalan, Helv. Chim. Acta 27, 535 (1944). 74. R. Goutarel, M. M. Janot, V. Prelog, and W. I. Taylor, Helv. Chim. Acta 33, 150 (1950). 75. M. F. Bartlett, J . Am. Chem. Soc. 84, 622 (1962). 76. N. An Cu, R. Goutarel, and M. M. Janot, Bull. SOC.Chim. France p. 1292 (1957). 77. Y. Ban and T. Oishi, Tetrahedron Letters p. 791 (1961).

This Page Intentionally Left Blank

---CHAPTER

21-

THE ERGOT ALKALOIDS A . STOLL AND A . HOFMANN Pharmaceutical- Chemical Research Laboratories. Sandoz Limited. Basel. Switzerland

I . The Biology of Ergot and a Short History of Its Active Principles Up to the DiscoveryofErgotamine ............................................

726

I1. Structural Types with Tables o f the Natural Ergot Alkaloids . . . . . . . . . . . . . . 729 111. Lysergic Acid and Isolysergic Acid ................................... A . Elucidation of Structure ......................................... B . Stereochemistry of Lysergic Acid and the Dihydrolysergic Acids . . . . . . . C. Syntheses in the Dihydrolysergic Acid Series ........................ D . Synthesis of Lysergic Acid ................... .................

734 735 737 742 744

I V . Simple Lysergic Acid Amides ........................................ A . Ergine.Erginine, C16H170NZ ( 2 6 7 . 3 ) . .............................. B . Lysergic Acid Methylcarbinolamide, C I S H Z I O ZN(31 ~ 1.4). ............ C . Ergometrine.Ergometrinine, C I S H ~ ~ O(Z3 2N5 ~ .4). ...................

746 746 747 747

V . Peptide Alkaloids .................................................. A . Elucidation of Structure of the Peptide Portion ..................... B . The Synthesis of Ergotamine and Its Stereochemistry ............... C. Characterization of the Ergot Alkaloids of the Peptide Type . . . . . . . . . .

748 748 753

--l o b >

V I . The Alkaloids of the Clavine Series ................................... A . Structural Relationships ......................................... B . The Clavine Alkaloids ...........................................

760 760 766

VII . Biogenesis of the Ergot Alkaloids ....................................

766

.........

C . Substitutions in the Ring System of Lysergic Acid . . . . . . . . . . . . . . . . . . . D . Biological Oxidation of Lysergic Acid Derivatives . . . . . . . . . . . . . . . . . . .

769 570 771

IX. The Pharmacology and Thcrapeutic Use of Ergot Alkaloids and Their

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

772

References ........................................................

779

Derivatives

726

726

A. STOLL AND A. IXOFMANN

I. The Biology of Ergot and a Short History of Its Active Principles Up to the Discovery of Ergotamine

The isolation of therapeutically active compounds from metabolic products of fungi is no longer regarded as an extraordinary event. Antibiotics constitute excellent examples of the results produced by modern biochemical research. The fungus ergot has been employed in medicine for centuries, and the quest for its active principles commenced nearly 150 years ago. The term ergot or Xecale cornutum designates the dark brown, hornshaped pegs, projecting from the ripening ears of rye in place of the rye grains. These tuberous bodies are collected before and during the harvesting or are separated from the threshed rye and represent one of the most remarkable drugs of the therapeutic armamentarium. Histologically speaking, they consist of compactly interwoven hyphae of a filamentous fungus [Claviceps purpurea (Fries) Tulasne] ; biologically speaking, these compact grains are sclerotia, the form in which the fungus passes the winter. Warm weather in the spring causes the ergot, which has become swollen owing to moisture, to germinate and put out bundles of hyphae and, later, long-stalked stromata. The surfaces of these stromata carry numerous perithecia (radially arranged, pearshaped cavities) in which filamentous ascospores are formed. Upon exposure of the terminal knob or capitulum of the stroma to the light, the ascospores are expelled into the air, are carried upward by rising convection currents, and settle on open rye flowers. Here, by using the stigmata of the flower as nutrient, they form a mycelium. Very soon, conidia are formed from the fungus filaments by abstriction, and these are surrounded by a vast quantity of sweet fluid, the so-called honeydew, which is secreted simultaneously. The infectious secretion is transferred to other rye flowers by insects or when neighboring ears are brought into contact by the wind, so that an infection, caused by the conidia suspension, results. After a few weeks, the mycelium solidifies to the externally darkcolored, internally white pseudoparenchyma, which forms the so-called sclerotium known as ergot. The fungus Claviceps purpurea and related species also attack other plants of the Gramineae (1) family and form sclerotia, the shape and size of which vary with the species of the host plant. The official form of ergot, however, is the product that forms on the ears of rye. It is not within the scope of this article to describe how devastatingly the population of vast areas was poisoned as a result of the presence of ergot in the grain used for making bread. It should, however, be noted

21. THE

ERGOT ALKALOI1)S

727

that the cause of the poisonings was still unknown, even after ergot was already being used in small doses by midwives as a proved means of inducing pains, as described by Adam Lonicer in his “Kreuterbuch” as early as 1582. The history of ergot and the ergot poisonings has been exhaustively described by Barger in his comprehensive monograph “Ergot and Ergotism,” Gurney and Jackson, London, 1931. Ergot, however, has only been used in official medicine since the American physician John Stearns (2) reported on the drug’s contractive action on the uterus in his publication, “Account of the Pulvis Parturiens, a Remedy for Quickening Child-birth ’) in 1808. The first pharmaceuticalchemical investigation was published in 1816 by the French pharmacist Vauquelin ( 3 ) . However, this publication and the numerous publications which appeared in the following 100 years gave no convincing data on the chemical nature of the specific active principles of ergot. Opinions regarding the chemical nature of the active principles changed frequently up to the beginning of the present century, even after the French pharmacist Tanret (4,5 ) succeeded in isolating a crystalline alkaloid preparation ‘‘ ergotinine cristalliske ” in 1875 and the English research workers Barger and Carr (6, 7 ) and the Swiss pharmacist Kraft (8, 9) simultaneously isolated compounds from ergot in 1906 which they named ergotoxine and hydroergotinine, respectively, and which were later found t o be identical. Our seventh treatise on the ergot alkaloids (10) reported the history of ergotoxine and ergotinine in great detail and showed that the two preparations, described in the literature, were not of a uniform nature but consisted of variable components. It is probable that the resulting fluctuations in activity were partly responsible for the fact that ergotinine and ergotoxine preparations found no lasting clinical application. Nevertheless, it was with ergotoxine preparations that the pharmacologist Dale ( 1 1) was first able to demonstrate the uterotonic effect typical of ergot. He also noted that they inhibited those functions controlled by the sympathetic nervous system and that they exhibited a specific antagonistic action to adrenaline. These observations were of the greatest importance to the subsequent practical application of ergot alkaloids. On the basis of disappointing toxicological observations, however, Kraft (12) advised that the alkaloids of ergot extracts should be removed even more carefully than had hitherto been the case. Even the 1923 edition of the British Pharmaceutical Codex still expressed the opinion that ergotoxine was not the specific active component of ergot. Many of the prescriptions given in the pharmacopoeia led t o preparations which, though containing small and varying quantities of a water-soluble active principle, did not contain alkaloids of the ergotoxine type. Thus, the necessity of obtaining an ergot preparation for the treatment

728

A. STOLL AND A. HOFMANN

of post partum uterine atony, constant and reliable in its action, remained. This need could only be satisfied by producing the pure natural active principle. The well-known fact that storage caused the activity of ergot to decrease proved that the active principle of ergot could decompose. Furthermore, upon administration of the ergot extracts, the contracting action occurred only after a latent period, so that the molecule constituting the active principle could be assumed to be relatively large (13, 14) and would thus undergo oxidative and fermentative changes upon storage. Mild conditions of the type developed in the preparation of chlorophyll, for example, suitably adapted to the isolation of the active principles of ergot, were thus a prerequisite for the success of the isolation. We worked on the assumption that the specific active principle of ergot was a base, e.g., an alkaloid. In order to protect this substance against conversion or oxidation, it was left under the protection of the amphoteric cell material. This was made weakly acid, and all the soluble, acid, and neutral components were exhaustively extracted with an inert solvent, e.g., benzene. By making the cell material alkaline, e.g., by treating it with ammonia, it was possible t o extract, with the aid of the same solvent, the basic components of the drug, relatively free from impurities. These components could then be obtained by evaporating the extract. Crystallization of the resulting crude alkaloid from aqueous acetone resulted in diamond-like, glistening crystals. These were homogeneous and possessed all the typical biological properties of ergot. The alkaloid was named ergotamine (15). The susceptibility of ergotamine t o light and t o air and the fact that, under the influence of acid, it is easily converted to a difficultly soluble isomeric form, ergotaminine, explains why previous research workers, who had not taken the same precautions, were unsuccessful in its isolation. Ergotamine was thus the first homogeneous specific active principle of ergot, and its production formed the basis for dependable clinical research. As little as 0.25 mg of the tartrate form in 0.5 ml of isotonic solution generally causes powerful contractions of the human uterus and arrests dangerous post partum hemorrhage. Pure ergotamine enabled pharmacological and clinical investigations t o expand and intensify in the field of the vegetative nervous system, thus providing the basis for the widespread use of ergot’s active principles in internal medicine and neurology. With the preparation of pure ergotamine, chemical research on ergot also entered a new phase so that, after many decades, it finally became possible to elucidate the structure of the ergot alkaloids and t o achieve their total synthesis. This will be shown in the remainder of this chapter.

21.

THE ERGOT ALKALOIDS

729

It should be noted a t this stage that, in some ergot types, other bases having a similar structure, e.g., those of the ergotoxine group, may occur in addition to or in place of ergotamine. Furthermore, mention should be made of the water-soluble, low molecular weight alkaloid ergometrine (ergobasine), which exhibits a powerful and quick-acting constrictor action on the smooth muscle of the uterus but is practically devoid of actions on the vegetative nervous system. I n addition, a new group of ergot alkaloids, found especially in ergot grown on wild grasses, was later discovered. These alkaloids belong to a new type of structural group, the so-called clavine type of ergot alkaloids. Detailed analytical investigations of commercial samples of ergot showed that the alkaloid content varied from 0.00 to 0.30%. The source of the ergot strongly influences the type of compounds contained therein, i.e., whether they are alkaloids of the ergotamine or of the ergotoxine type. This was a further reason necessitating the production and use of pure compounds for pharmacological investigations and clinical use. After the report on the chemical results, we shall include a final section as a summary of the pharmacological and clinical results obtained. 11. Structural Types with Tables of the Natural Ergot Alkaloids

If the stereoisomer forms are regarded as a single alkaloid in each case and those bases which, though shown to exist by means of paper chromatography, have not had their structure elucidated are ignored, a total of approximately two dozen ergot alkaloids have been described to date. All these alkaloids are formed from the same tetracyclic ring system that Jacobs and Gould (16) named ergoline (I).

5 I y . p (pJ p;il H I Ergoline

On the basis of their structural differences, the ergot alkaloids may be divided into two main groups: one group to include all lysergic acid derivatives of the acid amide type, and the other to include the so-called clavine alkaloids. Further structural groups may be recognized within these main groups, as is shown by Tables I and 11.

730

A . STOLI, AND A . HOFMANN

TABLE I THENATURAL ERGOTALKALOIDS OF COR

THE

LYSERGIC ACIDSERIES

H General formula ~~

Alkaloid

Formula

Section

A. Simple lysergic acid and isolysergic acid amides

R

=

NHz IV, A

Ergine Erginine

CH3

R

=

I

NHCH I OH

Lysergic acid methylcarbinolamide

CH3

I

R = NHCH

I

CHzOH Ergometrine (Ergobasine) Ergometrinine (Ergobasinine)

B. Derivatives of lysergic acid and isolysergic acid of the peptide type Ergotamine group

Ergotamine Ergotaminine

IV, B

21.

731

THE ERGOT ALRALOIDS

TABLE I-contiiaued Alkaloid

Ergosine Ergosinine

Formula

Section

v, c, 2

Ergotoxine group I

Ergocristine Ergocristinine

v, C, 3

Ergocryptine Ergocryptinine

v, C, 4

732

A. STOLL AND A. HOFMANN

TABLE I-continued Alkaloid

Formula

Section

Ergocornine Ergocorninine

TABLE I1 THE NATURAL ERGOT ALKALOIDS OF

THE

CLAVINE SERIES

Rz CHzRi

H General formula Alkaloid

Formula

Section

Ergolene-(8) derivatives

R1= Rz = R3 = H Agroclavine

VI, B, 1

R1= OH; Rz =-; R3 = H Elymoclavine

VI, B, 2

21. THE ERGOT ALKALOIDS

733

TABLE 11-continued

Alkaloid

Molliclavine

Formula

Section

VI, B, 3

Ergolene-(9) derivatives

R1= R2 = R3 = H Lysergine

VI, B, 4

R1= OH; Rz = I t 3 = H Lysergol

VI, B, 5

R1= R2 = -; R3 = H Lysergene

VI, B, 6

R1= H ; R2 2 OH; R3 = H SetocIavine Isosetoclavine

VI, B, 7 VI, B, 8

Penniclavine Isopenniclavine

VI, B, 9 VI, B, 10

Ergoline derivatives

Festuclavine Pyroclavine Costaclavine

VI, B, 11 VI, B, 12 VI, B, 13

734

A. STOLL AND A. HOFMANN

TABLE 11-continued Alkaloid

Formula

Section

. I

Fumigaclavine A

CisHzzOzNz

VI, B, 14

Fumigaclavine B

Ci6HzoONz

VI, B, 15

CHzOH With open ring D:

5-qHCH.3

OJ’ I

H Chanoclavine Chanoclavine

CieHzoONz

VI, B, 16

111. Lysergic Acid and Isolysergic Acid

The structural element common to all the alkaloid pairs shown in Table I is lysergic acid or its stereoisomer, isolysergic acid. Fresh ergot almost exclusively contains the pharmacologically highly active lysergic acid alkaloids. These are all levorotatory or only slightly dextrorotatory. By using old ergot or unsuitable methods of isolation, however, considerable quantities of the isomeric alkaloid forms are obtained. Isolysergic acid is the characteristic component of these alkaloids which have only a weak activity and are strongly dextrorotatory. The names of all the members of this group end in -inine. Lysergic acid and isolysergic acid, as well as the alkaloids derived therefrom, are readily reversibly interconvertible (17, 18), and an equilibrium of the isomeric forms is reached especially quickly in an alkaline medium ( 19). Lysergic acid, C16H1602N2, was first obtained by vigorous alkaline hydrolysis of ergot alkaloids (20, 21, 22). Recently, it was also found in

2 1. THE ERGOT

ALKALOIDS

735

its free form in saprophytic cultures of the ergot fungus (23). It crystallizes from water in thin, long, hexagonal leaflets, containing 1 mole of water of crystallization, and it has a melting point of 238" that is not well defined; [cL]$) + 32" in pyridine. Isolysergic acid, C16H1602N2, is somewhat more easily soluble in water than is lysergic acid, and it crystallizes therefrom with 2H20; mp 218" (dec.), [a]E2,, + 368" in pyridine (18). The natural dextrorotatory lysergic acid was named d-lysergic acid and its is0 form, d-isolysergic acid (24). Their optical antipodes do not occur in nature but may be produced by racemization of d-lysergic acid (24) or via the racemic hydrazides resulting from the splitting of lysergic or isolysergic acid amides with hydrazine (25). Naturally, total synthesis of lysergic acid (26) also yields racemates. Lysergic acid, isolysergic acid, and their derivatives give typical color reactions. The color reactions most commonly used are those of Keller (271, i.e., with concentrated sulfuric acid and glacial acetic acid containing iron chloride, and of van Urk-Smith (28, 29), i.e., with p-dimethylaminobenzaldehyde in a sulfuric acid solution. A blue color results from both reactions and is used for the qualitative and quantitative determination of ergot alkaloids in galenical preparations and in drug analysis (30, 31, 32).

A. ELUCIDATION OF STRUCTURE By oxidizing ergot alkaloids and lysergic acid, Jacobs and Craig obtained quinoline-betaine-tricarboxylicacid (33, 21, 34))and by fusion of dihydrolysergic acid with alkali, they obtained methylamine, propionic acid, l-methy1-5-aminonaphthalene,and 3,4-dimethylindole (35, 36). These cleavage products led to the deduction that lysergic acid contained, as its main structural entity, a new tetracyclic ring system which was named ergoline (I). By comparing the UV-spectra of lysergic acid and isolysergic acid on the one hand and dihydrolysergic acid (Fig. 1) on the other, it was deduced that lysergic acid and isolysergic acid possessed a double bond situated outside the indole system, but conjugated therewith (37). The position of the carboxyl group was ascertained from pK measurements (38) and the results of a /3-aminocarboxylic acid cleavage of dihydrolysergic acid (39). Removal of the asymmetry center a t the C-8 position, in which the carboxyl group is situated, yielded identical products in the case of lysergic acid and isolysergic acid. This proved lysergic acid and isolysergic acid to be diastereomers a t '2-8, as shown by structures I1 and I11 (40).

736

A. STOLL AND A. HOFMANN

H

COOH

I

H

I1 Lysergic acid

H 111 Isolysergic acid

The isomerization proceeds via an intermediate compound that is symmetrical at the C-8 position and that exhibits a continuous conjugated system of double bonds stretching from the enol double bond up to the indole system. Thus, isomerization of dihydrolysergic acids, in

FIG.1. UV-spectrum in methanol. I: Lysergic acid; 11: dihydrolysergic acid.

which this continuous conjugation is interrupted, as well as that of ergolene derivatives, which have no carbonyl function that could be enolated, is more difficult or even impossible. The speed of the isomerization and the position of the equilibrium are both strongly influenced by the nature of the substituent at the carboxyl radical of lysergic acid (19).

21. T H E

73 7

ERGOT ALKALOIDS

B. STEREOCHEMISTRY OF LYSERGIC ACIDAND THE DIHYDROLYSERGIC ACIDS Investigations on the stereoisomer dihydro acids and application of the “conformation theory” (41, 19) allowed the question of the configurative relationships between the two asymmetrical centers in lysergic acid and in isolysergic acid to be solved. The results of catalytic hydrogenation of the Ag-lO-doublebond of lysergic acid, isolysergic acid, and their derivatives (42, 43)are depicted by formulas IV to IX. The absolute configurations, discovered by methods to be discussed later, have already been taken into account in these formulas. HOOC

H

H

COOH

H

H

V Isolysergic acid

IV Lysergic acid

HOOC H

H COOH

HOOC H

mH

H-1‘l0 1

.

VI

Dihydrolysergic mid-I

/ VII Dihydrolysergic acid-I1

II

H COOH \.’e

/

VIII Dihydroisolysergic acid-I

IX Dihydroisolysergic acid-I1

tI

Hydrogenation of lysergic acid and its alkaloids yielded only one dihydrolysergic acid. I n the is0 series, however, both dihydroisolysergic acid-I and dihydroisolysergic acid-11 were formed, the latter stereoisomer being predominant upon rapid hydrogenation. In an alkaline medium, dihydroisolysergic acid-I was irreversibly rearranged t o form dihydralysergic acid-I. Dihydrolysergic acid-I1 could hitherto only be

738

A. STOLL AND A . TIOFMANN

produced in very small yield by alkaline saponification of dihydroisolysergic acid-I1 hydrazide. Conversely, saponification of dihydrolysergic acid-I1 hydrazide yields practically only dihydroisolysergic acid-11. I n this case the equilibrium is on the side of the is0 compound (44).Dihydrolysergic acid-I and dihydroisolysergic acid-I, as well as dihydrolysergic acid-I1 and dihydroisolysergic acid-11, differ only in their steric configuration a t C-8. Further proof of this is the formation of an identical lactani upon heating of dihydrolysergic acid-I and dihydroisolysergic acid-I with acetic acid anhydride (40), which removes the asymmetrical center a t C-8. The spatial configurations a t C-10 are identical for all of these acids. The question of cis- or trans-linkage of rings C and D may be solved by noting the behavior upon hydrogenation. The fact that only one stereoisomer forms upon hydrogenation of lysergic acid, namely, dihydrolysergic acid-I, can be explained by assuming that the carboxyl group and the hydrogen atom are on the same side of the molecule in lysergic acid, thus screening the double bond on one side (45). The attack of the hydrogen atom must then necessarily occur from the other side, so that the trans compound results. On the other hand, the preponderant formation of dihydroisolysergic acid-I1 under conditions which favor the cis compounds (rapid hydrogenation in glacial acetic acid with Pt catalyst), indicates a cis configuration in the -11series. As epimerization conditions favor the formation of dihydrolysergic acid-I and dihydroisolysergic acid-11, it may be assumed that these compounds carry their carboxyl group in equatorial arrangement. I n agreement is the fact that, in comparison with their epimers, they are less hindered in the C-8 position during saponification and condensation reactions. Furthermore, their derivatives exhibit a greater tenacity in chromatography. Dihydrolysergic acid-I thus has the C/D-trans-%equatorial formula VI, whereas the epimeric dihydroisolysergic acid-I has the C/D-trans-8axial arrangement (VIII). I n the case of dihydrolysergic acid-I1 (structure VII) and dihydroisolysergic acid-I1 (structure I X ) , both of which have a cis-linkage of rings C and D, the carboxyl radical is in the axial and the equatorial position, respectively. Thus, the carboxyl group is in an equatorial position in lysergic acid (IV) and in an axial position in isolysergic acid (11).Consideration of the space model shows that such a constellation is only possible if the ring D has a pseudo-chair form in both epimers. Measurement of pK values of lysergic acid and dihydrolysergic acid derivatives and the analysis of their IR-spectra confirm that these structural configurations are indeed correct (19).

21. THE ERGOT

ALKALOTIPS

739

The foregoing deductions relate only to the relative spatial arrangement a t the various asymmetric centers of isomeric lysergic and dihydrolysergic acids. The absolute configuration as depicted by formulas IV-IX was elucidated, on the one hand, by analysis of rotation-dispersion curves (46), and on the other, by chemical degradation of lysergic acid to an amino acid derivative of known absolute configuration (47). Comparison of the four isomeric lysergic acids (Fig. 2) shows that only the steric relationships a t the carbon atom in the 5 position (linkage of

(MO 300 400 500 600 790 mp FIG.2. Rotation-dispersion curves. 1 :d-Isolysergic acid; 2: d-lysergic acid; 3:&lysergic acid; 4 : I-isolysergic acid.

rings C and D) are decisive for a positive or negative Cotton effect and that the configuration at C-8 has only an additive or substractive effect. Model compounds, containing rings C and D, or A, C, and D of lysergic acid, were compared with polycyclic compounds of the steroid series, whose absolute configuration, with respect t o the spatial arrangement corresponding to C-5 of lysergic acid, was known. ( + )-d4~4a-N-Methyioctahydroquinolin-2-one (X), ( + )-d',lob-Nmethyihexahydrobenzo[ f Iquinoiin-2-one (XI), and the corresponding

7 40

A. STOLL AND A. HOFMANN

quinolin-2-01 (XII) show a positive Cotton effect (see Figs. 3-5), as does testosterone (XIII).It is thus justifiable to assign the same configuration as occurs in testosterone, namely, the hydrogen atom in /?-position, to

x

XI11

these compounds at the center of asymmetry, which corresponds to C-5 in lysergic acid. Since the tricyclic alcohol base XI1 corresponds in structure to lysergic acid, save for the absence of the pyrrole ring and the functional group at C-8 which, however, do not affect the direction of rotation of the rotation-dispersion, it was assumed, as a result of analogous rotation-dispersion curves of XI1 on the one hand, and of d-lysergic acid and d-isolysergic acid on the other, that the configurations were also analogous. Thus, it may be concluded that natural dextrorotatory lysergic acid has its hydrogen atom at C-5 in the /?-position. This fact was confirmed by oxidative degradation of the d-lysergic acid lactam (XIV) t o a D-aSpartiC acid derivative (47). d-Lysergic acid as such cocld not be used for the degradation, as it would yield an aspartic acid (XVIII), dialkylated a t the nitrogen atom, which was known t o racemize quickly. To obtain a sterically stable degradation product having a secondary nitrogen atom, N-norlysergic acid would have had to be used as starting material. As, however, demethylation of lysergic acid was unsuccessful, the lactam XIV, produced by heaticg lysergic acid in acetic acid anhydride for a short time, was employed (40). The lactam XIV was ozonized in a methylene chloride-methanol-water solution and then oxidized with hydrogen peroxide with the addition of anhydrous formic acid. As attempts to isolate the anticipated tricarboxylic acid XV proved unsuccessful, the reaction mixture was treated with 1 N EC! t o split off the oxalyl radical. After subsequent

1 . :i! A

I:

..

6000. 6000

40001

0

1

_/‘--- _ _ _--______ ________

#I

2000:

2000,

+0-

(-

2000:

2-

2000.

200 3p0 400 500 600 700 m p

4000.

1290 300 400 500 600 7 p rnp

FIG. 4. Rotation-dispersion curves. 1 : ( - ) d1,lOb-N - Methylhexahydrobenzo[,f]quinolin-2-one; 2 : ( + )-dl,lobN -niethylhexahydrobenzo[ f lquinolin%one (XI). ~

2000-

4000,

I200 XX, 400 500 600 700 m,u

ERGOT ALKALOIDS

t

4000.

21. T3E

2000.

4000.

FIG. 5. Rotation-dispersion curves. . A l J O b - N Methylhexahydrobenzo[f]quinolin-2-01; 2: ( +)-dl*’obN-methylhexahydrobenzo[ f Iquinolin2-01 (XII). 1 : ( -)

~

741

742

A. STOLL A N D A. H O F M A N N

esterification of XVI with n-propanol, a yield of 4% of an optically active amino acid ester was obtained in pure form. This proved to be identical with authentic D-( + )-N-methylaspartic acid di-n-propylester (XVII). The natural lysergic acid thus has the (5R:8R) configuration depicted by formula IV. HOO&&O I

HOOC) -

3

1 1

H

xv

IV

lHi

I I

.1 9

COOH Is

COOC3H7 H-LN'H l \CH3 CHz

CHs-CHz-CHzOH

<

H+

COOC3H7

XVIII

C. SYNTHESES IN

XVII

XVI

THE

DIHYDROLYSERGIC ACIDSERIES

Numerous intermediate stages led to the total synthesis of lysergic acid. The syntheses by Jacobs and Gould in 1937 (16) of ergoline (I),a saturated tetracyclic ring system, constitute the first basic step. This was followed in 1945 by the synthesis ofracemic dihydrolysergic acid by Uhle and Jacobs (48);in 1950, the optically active dihydrolysergic acids were synthesized by Stoll et al. (49); and finally Kornfeld et al. (50) synthesized natural lysergic acid in 1954. 4-Aminonaphthostyril (XIX), in which the rings A, B, and C of the ergoline system are already present, was used as starting material for the first synthesis of racemic dihydrolysergic acid (48). By condensing with cyanomalonic dialdehyde, cyclizing with zinc chloride, and saponifying the nitrile group with HC1, the ring structure (XX) of lysergic acid, containing a carboxyl group in the 8-position, was obtained. Catalytic hydrogenation of the chloromethylate of X X yielded tetrahydro-N-

21.

743

THE ERGOT ALKALOIDS

methylcarboxylic acid (XXI) which, upon reduction with sodium and butanol, yielded a small quantity of racemic dihydrolysergic acid (XXII). COOH I

H

xx COOH

COOH

2!3-cH3 =d/rJ

2. Redn.

H

H

XXI

XXII rac. -Dihydrolysergicacid

The naphthostyril derivative X X I was later produced by other processes (51,52), thus clearing the way for further syntheses of dihydrolysergic acid. Compound X X was also produced by a further process, COOCzH5

I

o=c /=

N 'H

-1. ZnHCl

2. CHaOH/HCl

COOCH3

COOCHs

I

or

I\ O N / c = o

V N / = o

I

I H

I

H

H

A

XXIII

fiS7

XXIV

COOH

XXIV H

XXV T ~ C-Dihydronorlysergic .

acid

B

744

A. STOLL AND A . HOFMANN

namely, by starting with 1-hydroxymethylene-1-phenyl-2-propanone as ring A and successively adding rings D, C, and B (53). An improved synthesis of dihydrolysergic acid, which yielded the homogeneous raceinates and furthermore led to the optically active dihydrolysergic acids, was effected via racemic dihydronorlysergic acids (54). As a starting material for this novel synthesis, Stoll and Rutschmann used the quinolone X X I I I which had already been described by Gould and Jacobs ( 5 5 ) . To ensure a successful reaction, the butanol used must contain a trace of water so that the methyl ester group of XXIV is saponified before reduction to the corresponding alcohol occurs. Compound XXV is a racemic mixture of three stereoisomers which, after esterification with methanol, may be resolved chromatographically to give racemic dihydronorlysergic acid-I methyl ester, racemic dihydronorisolysergic acid-I methyl ester, and racemic dihydronorisolysergic acid-I1 methyl ester. By reducing the tricyclic naphthostyril system to benz[c,d]indoline before addition of the D ring (56, 57), further variations for the synthesis of raceinic dihydronorlysergic acids may be obtained. The dihydronorlysergic acids were converted to the corresponding stereoisomer racemic dihydrolysergic acids by migration of the methyl radical. This migration was effected by heating the methyl esters of the dihydronorlysergic acids or by reduction in the presence of formaldehyde (49, 44, 58). By resolution of the resulting dihydrolysergic acid-I racemates in the form of the L-norephedride, Stoll et al. (49) succeeded in producing d-( - )-dihydrolysergic acid-I. This is the basic constituent of all the dihydrogenated natural ergot alkaloids.

D. SYNTHESIS OF LYSERGIC ACID Various groups of research workers unsuccessfully attempted the synthesis of lysergic acid (59, 60, 61, 62, 63, 64, 65). Most of Ihe attempts a t synthesis failed because of the inability to introduce the double bond in the 9,lO-position. This appears to be impossible when using the naphthostyril or benz[c,d]indoline system. The latter has a resonance energy which is approximately 20 kcal greater than that of the ergolene system of lysergic acid. Strong acidic reagents irreversibly rearranged lysergic acid derivatives XXVI to form benzindoline derivatives (XXVII) (56), which were crystallized in the form of their stable acyl compounds.

2 1. THE

745

ERGOT ALKALOIDS

The production of lysergic acid via compounds of the type XXVII, for which syntheses were developed (56, 51), was thus not possible. R

R

253-cH3 /==<

- & - \ * - C

H3

5

I\/( 1 y'1

\\/\*/CHz

H XXVI

I Ac

XXVII

R = COOH, COOCH3, CHzOAc

The first, and hitherto only, synthesis of lysergic acid was effected by the research group of Kornfeld et al. (50, 26). I n this particular synthesis a N-acyl-2,3-dihydroindolederivative was used as starting material, thus allowing the formation of rings C and D by classical methods. Dehydrogenation of the 2,3-dihydroindole system to the indole system was only effected in the last stage so that the formation of the benz[c,d]indoline system was prevented.

HOOC-CH2 \

XXVIII

xxx

XXIX

COOH

CN

0

I

I

I1

-CH3 1. AcaO

I H

XXXI

4. NaCN

Ac

XXXII

HeO-NazHAsOd

I

H XXXIII ruc. -Lysergic acid

N-Benzoylindolinyl-%propionic acid (XXVIII) was condensed to form the tricyclic ketone X X I X via the acid chloride. The bromination

746

A . STOLL AND A. HOFMANN

to the a-bromoketone was followed by the conversion with methylamineacetone ethylene ketal to form compound XXX. After acid hydrolysis to form the corresponding methylketone, this was cyclized with sodium methylate t o give the tetracyclic, unsaturated ketone XXXI. The liberated secondary amino group was again protected by acetylation and the ketone then reduced with NaBH4 to the secondary alcohol that was converted to the corresponding chloride in SO2 solution with thionyl chloride. Conversion with sodium cyanide in liquid hydrocyanic acid yielded the nitrile X X X I I which was converted by methanolysis to the corresponding ester. This was then subjected to alkaline saponification to form 2,3-dihydrolysergic acid. Dehydrogenating with deactivated Raney nickel in an aqueous solution in the presence of sodium arsenate yielded racemic lysergic acid (XXXIII). As the resolution of the racemic lysergic acid into its optical antipodes and the synthesis of ergometrine had been described earlier by Stoll and Hofmann (25,66),this meant that not only had the synthesis of d-lysergic acid been achieved but also the first total synthesis of an ergot alkaloid. This synthesis has hitherto found no industrial application, as various steps are difficult to effect on a technical scale, especially since the dehydrogenation of 2,3-dihydrolysergic acid gives only a small yield.

IV. Simple Lysergic Acid Amides A. ERGINE-ERGININE, C16H170N3 (267.3) 0

HzN-C

11

H

+;yH3 '/

y H XXXIV Ergine

H

::

CNHz

P-cH LJ \/

xy

H

xxxv Erginine

Ergine crystallizes from methanol in prisms, mp 242" (dec.), [a]? 0" [ t ~ ] E 2 ~+ ~ 15" ( _ +2") (c = 0.5 in pyridine). Erginine is readily formed by rearrangement of ergine, e.g., by recrystallization of the latter from methanol. Long massive prisms containing 1 mole of methanol result; mp 132"-134" (dec.), [a]: t-480" ( 2"), 608" ( f 2") (c = 0.5 in pyridine).

(T2"))

+

21.

THE ERGOT ALKALOIDS

747

Lysergic acid amide (XXXIV) and isolysergic acid amide (XXXV) which, for a long time, were only known as the hydrolysis products of ergot alkaloids (67, 18, 68), were recently isolated as the main constituents of the alkaloid mixture from ergot of Paspalurn distichurn L (69). They are thus also to be regarded as genuine ergot alkaloids. The term ergine, which was originally used for the first characteristic hydrolysis product of ergot alkaloids (67), before this was identified as isolysergic acid amide, is now correctly used in connection with lysergic acid amide. I n accordance with the conventional nomenclature used in the case of ergot alkaloids, the isolysergic acid amide has been named erginine. Recently, ergine and erginine were also isolated as main components, together with other alkaloids of the clavine group, from seeds of Rivea corymbosa (L.) Hall. f. and Ipomoea tricolor Cav. (70, 71). These seeds were used centuries ago by Central American indians as a magic drug under the name of " Ololiuqui." The occurrence of lysergic acid alkaloids in the plant family of Convolvulaceae is a completely unexpected phytochemical discovery as they had hitherto only been found in the lower fungi of the genus Claviceps but recently also in the genera Aspergillus and Rhizopus (72).

B. LYSERGIC ACID METHYLCARBINOLAMIDE, C ~ ~ H Z ~ (31 OZ 1.4) N~ (33.3

1 H-C-HN--C

o II

H

H

XXXVI

Lysergic acid methylcarbinolamide (XXXVI) crystallizes from chloroforminlongprisms, mp 135" (dec.), [a]E + 29" (c = 1.Oindimethylformamide). The alkaloid, which was isolated together with ergine and erginine from saprophytic cultures of Claviceps paspali (69), easily decomposes in a weak acid solution to form ergine and acetaldehyde. C19H2302N3 (325.4) C. ERGOMETRINE-ERGOMETRININE, Ergometrine (XXXVII) crystallizes from ethyl acetate in massive tetrahedrons, mp 162" (dec.). From chloroform, in which the alkaloid is

748

A. STOLL A N D A. HOFMANN

difficultly soluble, it is obtained with 1 mole of crystalline solvent. Upon crystallization from acetone, dimorphism was observed (73). Aside from that form which melts at 162", long needles having a mp of 212" (dec.) . I 3 + 41°, [a];!,, + 60" (c = 1.0 in ethanol). resulted; [ CHzOH 0

I

H-C--HN-C

I1

H

H

H

XXXVII

XXXVIII Ergometrinine

Ergometrine

I n 1935 ergometrine was discovered in four different laboratories almost simultaneously and described under four different names (74, 75, 76, 77). Although the names ergometrine and ergobasine have remained in use in Europe, ergonovine was adopted in the USA as the official nomenclature for the specific uterotonic ergot alkaloid. Ergometrinine crystallizes from acetone in the form of prisms, mp + 414", [ C C ]+; ~520' ~ ~(c = 1in chloroform). Only a small 196" (dec.), quantity of it occurs in the ergot drug together with ergometrine (78). However, it may easily be produced by rearrangement of ergometrine (79). The relatively simple structure of ergometrine was elucidated by Jacobs and Craig, who showed that, upon alkaline hydrolysis, lysergic acid and L( + )-2-aminopropanol result (80). The synthesis of ergometrine from these two components was effected by Stoll and Hofmann (81, 66). This was the first synthesis of an ergot alkaloid. By use of rac.isolysergic acid hydrazide, which could be resolved into its optical antipodes with di-(p-toluy1)tartaric acid (25), all eight theoretically possible stereoisomer forms of ergometrine were synthesized via the corresponding azides (66). V. Peptide Alkaloids

A. ELUCIDATION OF STRUCTURE OF THE PEPTIDE PORTION The majority of alkaloids produced from ergot fungi are peptides of lysergic acid. Upon hydrolysis they decompose to give lysergic acid, two

21.

THE ERGOT ALKALOIDS

749

amino acids (one of which is always proline), one a-keto acid, and one equivalent of ammonia (82, 83, 84, 85, 13). L-Phenylalanine, L-leucine, and L-valine were found as variable amino acids, and pyruvic acid or dimethylpyruvic acid as a-keto acids. The amino acid proline, common to all peptide alkaloids, was obtained in the D-form upon acid hydrolysis or as L-proline under mild alkaline conditions. Table I11 gives a summary of the hydrolysis products. The two alkaloid pairs yielding pyruvic acid have been named the ergotamine group after their most prominent representative. The three alkaloid pairs yielding dimethylpyruvic acid upon hydrolysis are known as the ergotoxine group. This nomenclature is due to the fact that, for numerous decades, mixtures of ergocristine, ergocryptine, and ergocornine, which contained varying proportions of these constituents, were assumed to be single homogeneous individuals and were named “ergotoxine” by Barger and Carr in 1906 (6). It was only in 1943 that Stoll and Hofmann succeeded in separating ergotoxine into its three components with the aid of di-p-toluyl-L-tartaric acid (10). I n the “periodic system ” of ergot alkaloids, the alkaloid corresponding to ergocornine and having valine as the second amino acid is still missing from the ergotamine group. This alkaloid, which hitherto has not been found in nature, was, however, recently produced synthetically (169) (see Section V, B). Recently a new alkaloid was discovered in the ergot of rye which, upon hydrolysis, yields a-keto butyric acid as a-keto acid, while the other cleavage products correspond to those of ergotamine and ergocristine. The new alkaloid, of which only a very small quantity is present together with ergotamine, was named ergostine whereas the isomer, which results from a transposition of said ergostine, was called ergostinine (175). This alkaloid pair forms the first representative of a further alkaloid group which, when using nomenclature analogous to that employed for the ergotamine and ergotoxine groups, could be called the ergostine group. The a-keto acids are not present as such in the alkaloids as no free keto group can be detected. It was thus assumed that an a-hydroxy-a-amino acid group was present in the alkaloids, which would decompose to the corresponding a-keto acid and 1 equivalent of NH3 (82, 86). As neither a free carboxyl group nor a basic amino group could be detected in the peptide portion, a cyclic configuration of the various constituents had to be assumed. As somewhat milder alkaline hydrolysis yielded mainly lysergic acid amide instead of lysergic acid, it was obvious that an acid-amide linkage was present between the lysergic acid and the a-hydroxy-a-amino acid racidals. As a result of these

II

A . STOLL A N D A. HOFMANN

II

II

21. THE: ERGOT ALKALOI1)S

751

considerations, Jacobs and Craig (86) postulated structural formula XXXIX and Barger (87) postulated formula XL for the peptide alkaloids. R1

R1

R1

\C/H

R1

\CG

Lys. CONH-LO-CO-Ck I CO-N

I

XXXIX

‘-co

RZ \

NH

/

I

Lys .CONH-C-0-CO

I 7 GO-NH-CH-CO

IJ N

I

R2

XL

These hypothetical structures (XXXIX and XL) differ only in the sequence of the amino acids. This was determined as a result of large cleavage products obtained by a partial hydrolysis of the alkaloids or their dihydro derivatives. Thus, for example, the cleavage of dihydroergotamine and of dihydroergocristine with hydrazine yielded, aside from dihydrolysergic acid hydrazide, propionyl-L-phenylalanyl-Lproline, or isovaleryl-L-phenylalanyl-L-proline(88). The structure of these compounds was confirmed by synthesis. The keto acid component was reduced by hydrazine to form the corresponding fatty acid. Upon mild hydrolysis with one equivalent of alkali in alcohol, however, the or keto acid as such remained, and pyruvoyl-L-phenylalanyl-L-proline dimethylpyruvoyl-L-phenylalanyl-L-proline (68) was obtained from the above-mentioned alkaloids. Thus i t was that the sequence of the amino acids was found t o be that depicted by formula XL. This, however, did not explain subsequent observations, all of which suggested that the peptide portion must necessarily contain a diketopiperazine ring. Cleavage with hydrazine and hydrolysis with one equivalent of alkali yielded, aside from the already mentioned acyl dipeptides, a considerable quantity of diketopiperazine, consisting of proline and the different amino acids, i.e., phenylalanylproline lactam in the case of ergotamine and ergocristine. Even more informative were the results of (1) the reduction with lithium aluminum hydride and (2) the thermocleavage. These results are depicted by formulas XLI-XLVII (89). It was to be expected that the lactone group in XL would be split with LiAlH4 to form a primary and a tertiary hydroxyl function. Instead, however, the reduction products XLII, XLIII, and XLIV, the structure and configuration of which were confirmed by the synthesis, were obtained. I n the case of the polyamines XLII, all the C and N atoms of

152

A. STOLL AND A . HOFMANN

the relative alkaloids are still present, whereas all the oxygen has been removed by reduction.

XLIII

Ri

Ri

\C/H

RCONHz

I + C/O

Ri

RI HzC---CHz (D)l

1 -.OH I

1

/C\O

\C/H

I

I

.,c/C\N/CH~

O=(I--N\

XLIV

OHo M

CO COOH

( L )Rz 9

XLV

XLVI

HzC-CHz

+

(D)

I

I

OC/C\N/CHZ H \ H I H N \ /CO

WX,

XLVII

R = ergolenyl or ergolinyl radical R1 = H or CH3 Rz = benzyl, isobutyl, or isopropyl radical

The thermocleavage of the dihydro alkaloids yielded, aside from the dihydrolysergic acid amide, the corresponding pyruvoyl- or dimethylpyruvoyl-diketopiperazines (XLV). These products could later also be prepared by total synthesis, i.e., by cyclizing the corresponding acidic pyruvoyl- or dimethylpyruvoyl-peptides (90). Proline is present in the D form in compound XLV. Inversion must have occurred during the thermocleavage for, in the alkaloids, the proline radical has the L configuration, as may be seen from the isolation of XLII and XLIV. In XLV the carbonyl function of the cr-keto acid radical cannot be detected chemically; obviously the capacity for reaction is impaired by its proximity t o the neighboring lactam carbonyl group. The above findings may be satisfactorily explained by ascribing

21. THE ERGOT

ALKALOIDS

753

formula XLI to the peptide portion. I n this formula the nine-membered lactam-lactone ring of XL is divided into a five- and a six-membered ring. The structure resulting from cyclizing and migration of a hydrogen atom from the nitrogen atom of an amide group to the oxygen atom of a neighboring lactone group was named the cyclol grouping. This name stems from Wrinch (91), who used it to characterize cyclic peptide structures which were, a t that time, still hypothetical. The orthocarboxylic acid form of proline is the basis of the novel amino acid grouping of the ergot alkaloids. One of the three hydroxyls has a lactonelike linkage, the other a lactam-like linkage, and the third is free. The last-mentioned hydroxyl still exhibits very weak acidic properties and is responsible for the solubility of the peptide alkaloids in strong aqueous sodium hydroxide solution. Owing to the complete elucidation of the constitution of the compounds obtained by total reduction with LiAlH4 and by thermocleavage, it was possible to propose structural formulas for the peptide alkaloids as early as 1951 (89). These formulas were confirmed by total synthesis of ergotamine after a further 10 years of intensive laboratory research (92).

B. THESYNTHESIS OF ERGOTABIINE AND ITSSTEREOCHEMISTRY The major problem in the synthesis of the peptide portion of the ergot alkaloids was the assembly of the extremely labile a-hydroxy-a-amino acid grouping and the formation of the cyclol structure which was, a t that stage, unknown in organic chemistry. Extensive investigations on the production of a-hydroxy-a-amino acid derivatives were conducted by Shefnyakin et al. (93, 94). They succeeded in incorporating derivatives of this type into di- and tripeptides but, owing to the labile nature of the a-hydroxy-a-amino acid group, all attempts a t cyclization were unsuccessful. The reverse route, however, i.e., formation of the cyclol group followed by the introduction of the a-amino acid fraction, produced the desired compound (92). It was observed t,hat cyclol formation occurred spontaneously if certain structural conditions were fulfilled and that the system underwent considerable stabilization upon cyclolization. On this basis, Hofmann et al. synthesized the peptide portion of ergotamine via stages XLVIII-LIII and, by linking the resulting peptide portion with lysergic acid, synthetically produced this alkaloid (92). Methylbenzyloxymalonic acid ester, which may be produced from methylmalonic acid diethylester by bromination and conversion with

7 54

A. STOLL AND A. HOFMANN

XLVIII

XLIX

-I

CH3

I

H~C~OOC--COCH~C~HS I

o=c

I

/

k,ckh=O

a: mp 180"-183" b: mp 131"-133"

sodium benzylate, was used as a starting material for the production of the a-hydroxy-a-amino acid moiety. By saponification of one of the ester functions and subsequent treatment with thionyl chloride, the semiester acid chloride XLVIII resulted. This was reacted with L-phenylalanyl-Lproline lactam (XLIX)in pyridine to produce the acylated diketopiperazine (L). As compound XLVIII was in its racemic form, compound L consisted of a mixture of two diastereoisomer forms. The acyl radical of compound L, however, is easily split off again hydrolytically. For this

2 1. THE ERGOT

ALKALOIDS

755

reason, compound L was immediately treated with palladium-hydrogen so as to remove the benzyl group. The resulting compound LI, having a free hydroxyl group, cyclized spontaneously to form the stable cyclolcarboxylic acid ester LII. The cyclol structure in L I I was detected by the presence of an acid hydroxyl group and by analysis of the IR- and NMR-spectra. As compound L I I was stable it could be resolved into the homogeneous stereoisomer forms by fractional crystallization. Two stereoisomers resulted, LIIa having a melting point of 135"-136" and LIIb having a melting point of 202"-204". These compounds differ in their configuration a t the C-2 atom (compare numbering in LIII). The occurrence of only two stereoisomer forms of L I I shows that the cyclol formation, as a result of which a new center of asymmetry is formed a t C-12, takes place stereospecifically. I n both isomers the carbethoxy group was converted to the amino group by means of a Curtius degradation. Owing to the stability of the cyclol system, the ester could be saponified with sodium hydroxide solution to form the corresponding acid. This acid was reacted, in the form of the sodium salt, with oxalyl chloride to yield the acid chloride which, upon reaction with sodium azide, yielded the corresponding acid azide. A Curtius rearrangement of the azide, effected by heating with benzyl alcohol, yielded benzyl urethane which, upon hydrogenolytic cleavage, decomposed to form the amino cyclol LIII. Although compound L I I I is not stable as a free base, it was crystallized in the form of its hydrochloride. The hydrochloride of the amino cyclol of the ester LIIa melts at, 180"-183" and that of LIIb a t 131"-133". The entire peptide portion of ergotamine is present in compound LIII. Conversion of isomer LIIIa with lysergic acid chloride hydrochloride (IVa) in chloroform, with the addition of tributylamine, yielded a compound whose chemical, physical, and pharmacological properties were identical with those of the natural alkaloid ergotamine. Thus, the total synthesis of this alkaloid was accomplished, lysergic acid having already been synthesized at an earlier date ( 5 0 ) . Acylation of compound LIIIb with IVa yielded an alkaloid (LIV) which is an isomer of ergotamine, differing therefrom only by the spatial arrangement a t the C-2 atom of the peptide portion. By means of a substantially similar process, ergosine as well as an alkaloid of the ergotamine group which had hitherto not been found in nature, viz., that alkaloid having L-valine as variable amino acid and which could be referred to as "ergovaline," were produced synthetically (175).

An improved process for the production of ergotamine (173) utilizes the optically active acid chloride of the methylbenzyloxymalonic

756

A. STOLL AND A. HOFMANN

acid semiester (XLVIII)as starting material. As the absolute configuration of the corresponding optically active semiester had been elucidated (17 4 ) , the absolute configuration at the C-2' atom of the peptide portion 0

I

H

IVa

LIIIa

I

H LIT' Ergotarnine

of ergotamine is likewise known. Furthermore, with the aid of degradation reactions of intermediate products formed in the synthesis of the peptide portion, the configuration a t the C-12' atom could also be elucidated so that the absolute configuration is now known for all six centers of assymetry of ergotamine, as is shown by the formula LIV (173). The similar optical rotation values obtained could lead to the conclusion that the remaining ergot alkaloids of the peptide type exhibit the same stereochemistry as does ergotamine.

C. CHARACTERIZATION OF THE ERGOT ALKALOIDS PEPTIDE TYPE

OF THE

A total of five alkaloid pairs of the peptide type, whose structures are depicted by formulas LIV-LVIII, have been isolated. A further alkaloid pair, ergosecaline and ergosecalinine, the structure of which has, as yet,

21.

THE ERGOT ALKALOIDS

757

not been definitely elucidated, also appears to have a peptide character (95).

Rz a : Lysergic acid radical (IV)

b: Isolysergic acid radical (V) R1 = H ; Rz = C H ~ C ~ Hergotamine S: (LIVa), ergotaminine (LIVb)

CH3 R1 = H ; Rz = CHzCH/ . ergosine (LVa), ergosinine (LVb) \CH3' R1 = CH3; Rz = C H z C & , : ergocristine (LVIa), ergocristinine (LVlb) ,CH3 Ri = CH3; Rz = CHzCH, : ergocryptine (LVIIe), ergocryptinine (LVIIb) CH3 CH3 Ri = CH3; Rz = CH/ ' ergocornine (LVIIIa), ergocorninine (LVIIIb) 'CH3'

1. Ergotamine-Ergotaminine, C33H3505N5 (581.7) Ergotamine (LIVa) ( 13) crystallizes especially easily and typically from a 90% aqueous acetone solution in the form of truncated polyhedral prisms having the following composition : C33H3505N5.2CH3COCH3.2H20. They disintegrate rapidly upon exposure to the atmosphere, mp 180" (dec.). From an 800-fold quantity of boiling benzene, long thin prisms - 160", [ C L ] ; ~-~192" ~ (c = 1.0 in are obtained, mp 212"-214" (dec.),$I.[ CHC13); [ a ] E - 12.7", [ C C ] ; ~ ~8.6" ~ (c = 1.0 in pyridine). Xalts of ergotarnine. Ergotaminetartrate (C33H3505N~)~. C4H606. 2CH30H, from methanol thick rhombic slabs, mp 203", is the salt of ergotamine most used in pharmaceutical preparations (e.g., Gynergen, Bellergal) and this salt is the one which has been adopted by the pharmacopeias. Ergotaminehydrochloride has a, mp 212" (dec.); ergotaminehydrogen maleate, mp 195"-197" (dec.); ergotaminephosphate, mp 200" (dec.) ; ergotaminesulfate, mp. 207" (dec.); ergotaminemethanesulfonate, mp 210" (dec.). Ergotamine was isolated from Swiss ergot by Stoll in 1918. It was the first chemically homogeneous and fully active ergot alkaloid and found widespread medical application. The great variety of clinical uses t o which it is being put a t preeent will be described in greater detail in the

758

A . STOLL AND A . HOFMANN

last section of this article. The major portion of ergotamine is currently obtained from ergot cultivated in the Swiss midlands by artificial mechanical infection of rye cultures. A process (14, 13), developed by Stoll in as early as 1918, has been found especially effective for the industrial production of ergotamine. Ergotaminine (LIVb) is very difficultly soluble in most solvents. Approximately 1500 parts of boiling methanol are required before it will dissolve and, upon cooling, it crystallizes in thin rhombic plates; mp 241"-243" (dec.); [a]'," +369", [a]:'&, +462" (c = 0.5 in CHC13); [a]? + 397", [a];&l + 497" (c = 0.5 in pyridine). As a result of the difficulty with which it dissolves, it crystallizes rapidly from the equilibrium set up between ergotamine and ergotaminine in hydroxyl-containing solvents. This leads to practically complete conversion of ergotamine to ergotaminine. On the other hand, the extreme difficulty with which ergotamine sulfate dissolves in glacial acetic acid may be used for the reversion of ergotaminine to ergotamine (13). As is the case with most of the isolysergic acid derivatives, ergotarninine is practically ineffective pharmacologically (96). Long storage of ergotamine or ergotaminine a t room temperature in acid solution or boiling for a short time causes rearrangements also of the peptide portion, and an equilibrium among ergotamine, aci-ergotamine, ergotaminine, and aci-ergotaminine is set up (97). The aci-isomers have an amphoteric nature and, unlike the alkaloids from which they stem, dissolve not only in dilute acids but also in dilute aqueous alkali solutions. Aci-eigotamine, C33H350sN5, yields needles from methanol ; mp 185"187", [a]? - 32" (c = 1.2 in pyridine). Aci-ergotaminine, C33H3505N5, yields fine needles from methanol-ether, mp 203" (dec.), [a]'," +258" (c = 1.2 in pyridine). Pharmacologically speaking (96), the aci-isomers are only weakly active.

2. Ergosine-Ergosinine,C30H37OsNs (547.6) Ergosine (LVa) (98) crystallizes from ethyl acetate in rectangular plates; mp 220"-230" (dec.), [a]: - 183", [a]&, - 220" (c = 1.0 in CHCI,); [a]'," - 8", [a]f:61 - 1" (c = 1.0 in pyridine). Ergosine.HC1 crystallizes from acetone in plates, mp 235" (dec.), ergosine. CH3S03H crystallizes from methanol in clusters of needles, mp 217"-218" (dec.). Ergosinine (LVb) crystallizes from acetone in the form of blunt prisms: mp 228" (dec.), [a]:: + 420", [a]:&l + 522" (c = 1.0 in CHC13). Ergosine and ergosinine, which were isolated from Iberian ergot by Smith and Timmis in 1937 (98), have so far found no medicinal application.

21. THE

ERGOT ALKALOIDS

759

3. Ergocristine-Ergocristinine, C35H3905N5 (609.7) Ergocristine (LVIa) crystallizes from acetone in the form of prisms containing crystalline solvent and melting a t 160"-175" (dec.); [a]? -217" (c = 1.0 in CHC13); [a]? -93", [a]& - 107" - 183", [ c L ] ~ ' & ~ (c = 1.0 in pyridine). Ergocristine . HCI crystallizes from alcohol ether in long slabs ; mp 205" (dec.); ergocristine. H3P04 crystallizes from alcohol in hexagonal plates ; mp 195" (dec.); ergocristine. C2H5S03H crystallizes from acetone in the form of hexagonal slabs; mp 207" (dec.). Ergocristine was isolated from Iberian ergot by Stoll and Burckhardt (99) in 1937 and later found by Stoll and Hofmann (10) to be a constituent of the alkaloid mixture known as ergotoxine (7). As a result of the strong sympathicolytic act'ion of its dihydro derivative, it has found clinical application, e.g., as a component of Hydergin. Ergocristinine (LVIb) crystallizes from alcohol in long thin prisms ; mp 226" (dec.) (10); [a]$' +366", [a]f!61 +460" ( c = 1.0 in CHC13); [ ~ ] +g 462", [a]'$,, + 576" (c = 1.0 in pyridine). 4 . Ergocryptine-Ergocryptinine, C32H4105N5 (575.7) Ergocryptine (LVIIa) crystallizes from a concentrated methyl alcoholic solution in truncated prisms; mp 212"-214" (dec.);[a]g - 190", [a]g261 -226" (c = 1.0 in CHC13); [ a ] E - 112", - 133" (c = 1.0 in pyridine). Ergocryptine . H3P04 crystallizes from 90% alcohol in hexagonal plates ; mp 198"-200" (dec.); ergocryptine .C2H5S03H) prisms from alcohol-ether ; mp 204" (dec.). Ergocryptine was discovered by Stoll and Hofmann as a component of the ergotoxine complex (10). It is the main alkaloid in ergot of Japanese (100) and South American (101) wild grasses. Its dihydro derivative is used as a constituent of Hydergin. Ergocryptinine (LVIIb) was produced by rearrangement of ergocryptine in boiling methyl alcohol (10). Thin prisms result; mp 240"242" (dec.); [ a ] g + 408"; [ajzi,, + 508" (c = 1.0 in CHC13); [ a ] g +479", [ E ] E ~ , ~+ 596" (c = 1.0 in pyridine).

5. Ergocornine-Ergocornin~ne,C31H3905N5 (561.7) Ergocornine (LVIIIa) crystallizes from methanol in the form of poly-226" (c = 1.0 in hedra; mp 182"-184" (dec.); [a]'$' - 188") 122" (C = 1.0 in pyridine). Ergocornine. CHC13); [a]? - 105", H3P04crystallizes from 90% alcohol in the form of pointed prisms which combine to form clusters; mp 190"-195" (dec.); ergocornine. C Z H ~ S O ~long H , needles from alcohol; mp 209" (dec.). Ergocornine was first discovered by Stoll and Hofmann (10) upon

760

A. STOLL AND A. HOFMANN

separation of the ergotoxine complex. It is used medically in the form of its dihydro derivative as a constituent of Hydergin. Ergocorninine (LVIIIb) was obtained by isomerization of ergocornine (10). It crystallizes as pointed prisms from methanol; mp 228" (c = 1.0 in CHC13); [a]: +500", (dec.); [a19 +409", [ C ( ] E ~ ~ +512" ~ + 624" (c = 1.0 in pyridine).

VI. The Alkaloids of the Clavine Series The first representatives of this second main group of ergot alkaloids, which differ from the classical lysergic acid alkaloids in that the carboxyl group of the lysergic acid has been reduced to the hydroxymethyl or methyl group, were discovered by Abe and collaborators in 1951 in Japan in the ergot of various species of grass growing in the Far East (100). These compounds were agroclavine, obtainedfrom the ergot of Agropyrum semicostatum and elymoclavine from Elyrnus mollis ergot. All alkaloids of this type that were isolated later were given names ending in -clavine to show that they form part of the same structural group. Recently, with the aid of the new, efficient methods of detection, e.g., paper and thin-layer chromatography, clavine alkaloids have also been found to occur in ergot obtained from rye. Furthermore, in 1960, Hofmann and Tscherter surprisingly found the occurrence of alkaloids of the clavine type in higher plants, i.e., in genera of the family of twining plants (Convolvulaceae) (70, 7 1).

A. STRUCTURAL RELATIONSHIPS The structural and configurative relationships between the alkaloids of the clavine group on the one hand, and the derivatives of lysergic acid alkaloids (LXII, LXIII) on the other, are shown by the illustrated formula scheme (LIX to LXVI). Catalytic hydrogenation of elymoclavine (LIX) (102) yielded a mixture of d-dihydrolysergol-I (LXII) ( 103) and d-dihydroisolysergol-I (LXIII) (103). It has the same ring system and the same configuration a t the C-5 and C-10 atoms as d-dihydrolysergic acid. Reduction of compound LIX with sodium in butanol yielded a mixture of agroclavine (LX), lysergine (LXI), pyroclavine (LXIV), festuclavine (LXV), and costaclavine (LXVI) in addition to d-dihydrolysergol-I (LXII) and d-dihydroisolysergol-I (LXIII) (104, 105).

2 1. THE ERGOT ALKALOIDS

u

761

762

A. STOLL AND A. HOFMANN

HOHzC H \-

LIX

CHz

'L

Na-butylate

H

H

LXVII Lysergol

LXVIII Lysergene

H

H

LIX Elymoclavine

LX Agroclavine

CrzO?a-

Crz07*

HBC OH \,,*

HO

CH3 \,,,

HOHzC

OH

\,,-

HO

CHzOH

\,-

c3H-cH3

H LXIX Setoclavine

(H

H LXX Isosetoclavine

H LXXI Penniclavine

H LXXII Isopenniclavine

The position of the isolated double bond in agroclavine, for which the 7-8 or 8-9 position came into consideration, was determined from the fact that agroclavine gave nearly the same pK value as its dihydro derivatives (106). Heating of agriclavine in a sodium butylate solution produced only lysergine, whereas the same treatment in the case of elymoclavine yielded a mixture of lysergol (LXVII) and lysergene (LXVIII) (105). The correctness of these deductions was confirmed by the extensive investigations of Schreier (107).

2 1. THE

763

ERGOT ALKALOIDS

The structures of setoclavine (LXIX), isosetoclavine (LXX), penniclavine (LXXI), and isopenniclavine (LXXII) were, for the major part, obtained from the oxidative formation of these alkaloids from agroclavine (LX) and elymoclavine (LIX). Oxidation of compound LX with potassium dichromate in dilute sulfuric acid yielded a mixture of setoclavine and isosetoclavine (108) which has the same UV-spectrum as lysergic acid. The double bond in the 8-9 position of compound LX has shifted t o the 9-10 position, As a result of the tertiary character of the hydroxyl group, the pK values, and chromatographic behavior, the structure depicted by LXIX was attributed to setoclavine and that of LXX to isosetoclavine. Similarly, elymoclavine yielded the isomeric pair, penniclavine and isopenniclavine (109, 108). The structures depicted by LXXI and LXXII were deduced for these alkaloids which contain a glycol grouping. 0 H3C

H

H3C

H

H

H

H3C

LXXIII

H LXXIV

Fumigaclavine A

Fumigaclavine B

H LXI Lysergine

Deacetylation caused fumigaclavine A (LXXIII) to be converted to fumigaclavine B (LXXIV). The ease of deacetylation, as well as the IRbands at 1241 and 1725 cm-1, indicated the presence of an ester group in compound LXXIII. Heating of compound LXXIV with NaOH caused water to be split off and resulted in the formation of lysergine (LXI). The configurations at the C-9 and C-10 atoms of compounds LXXIII and LXXIV have not, as yet, been elucidated. The negative optical rotation suggests a trans-linkage of rings C and D (72). Chanoclavine (LXXV) (108) is the only known ergot alkaloid in which the D ring of the ergoline system is open. LXXV easily forms an 0 , N CHzOH

HsC B H

/C---CH3 I

HC<

H - C

H

NHCH3

&i= H

LXXV Chanoclevine

__3

Ly H

LXV Bestuclavinn

3

-1 Q,

TABLE IV

ip

TEE CLAVINE ALKALOIDS Alkaloid

Formula

I[.

mp (solvent)

Source; ergot of:

Reference

("C) Agroclavine

Ci~Hi8Nz vba

205-206 (acetone)

[a];'

- 182' pyridine 0.5

- 155' CHC13 0.9 [a];"p - 152" pyridine 0.9 [a];n

Elyrnoclavine

Ci6Hi80N2 vb

245-249 (methanol)

Agropyrum semicostatum A . ciliare Fr. Pennisetum typhoideum Rich.

110 111 112

E l p u s mollis Tri. Pennisetum sp. Rivea corymbosa (L.) (seed) Nal1.f.

100 112 71

Elymus mollis

113

Agropyrum sp. ex elymoclavine ex agroclavine ex lysergene

114 104 104 107

Rivea corymbosa Elymus sp.

71 114, 115

ex elymoclavine E1yrnu.s sp.

104, 107 114, 115

Ci6Hi80zNz g

253 (methanol)

[alg

+ 30" pyridine 0.2

[a]&

+42" pyridine 0.2

Ci6HisNz vb

286 (methanol, ethyl acetate)

[a];'

+ 65" pyridine 0.5

Lysergol

253-255 (ethanol)

[a];'

Lysergene

247-249

Setoclavine

229-234 (acetone, methanol)

[a];'

+ 174" pyridine 1.0 +232" pyridine 1.0

Pennisetum typhokieum Elymus mollis

108 109

Isosetoclavine

234-237 (methanol)

[a];' [a]&,

+ 107" pyridine 0.5 + 147" pyridine 0.5

Pennisetum sp. Japanese grasses

108 116

Molliclavine Lysergine

+ 54" pyridine 0.3 + 87" pyridine 0.3 [a]tO + 504' pyridine 0.4

p g

8

* 3

?

5 2

+

[a];' 151" pyridine 0.5 [a]:&1+20lo pyridine 0.5

Penniclavine

Ci6Hi80zN~ 222-225 g (acetone)

Isopenniclavine

Ci6Hi80zN2 g

163-165 (water)

[a]:' [a]:&1

Festuclavine

C16H20N2 vb

242-244 (methanol)

[a];' - 110" pyridine 0.5 [a]:&1- 128" pyridine 0.5 [a]:' - 70" CHC13 0.5 [a]:&, - 83" CHC13 0.5

Pyroclavine

Ci6HzoN~ vb

204 (methanol, benzene, ethyl acetate)

[a];',

Costaclavine

Ci6H2oNz

182 (ethyl acetate)

[a];'

+ 146" pyridine 0.7 + 198' pyridine 0.7

Fumigaclavine A

CigHZzOzNz 84-85 [a]'& (methanol -HzO)

Fumigaclavine B

C16H2oONz

- 90" pyridine 0.2 - 105" pyridine 0.2

+ 44" pyridine 0.2 + 59" pyridine 0.2

[a]:&1- 113' pyridine 0.6

220-222 (methanol, acetone)

[a];' [a]'&

- 240" pyridine 1.0 -294" pyridine 1.0

N-acetylchanoclavine

226-228

[a];'

- 80" pyridine 0.5

0,N-diacetylchanoclavine

174-175

[a]:'

- 55" pyridine 0.9

Chanoclavine

Ci6HzoONz vb

Keller's color reactions: g = green; b = blue; vb = violet-blue.

Agropyrum and Phalaris spp. Aspergillus fumigatus Fres. ex agroclavine ex ML. from festuclavine ex agroclavine

Agropymcrn spp. ex agroclavine and elymoclavine

- 56.7" methanol 1.5 Aspergillus fumigatus

244-245 and 265-267 (ethanol-HzO)

[a]:&1

Pennisetum $yphoideum Japanese grasses Rye Pennisetum sp.

Aspergillus fumigatus

108,112 109 117 108 111 72 104,107, 118 116 104,107

116 104

e 72 72

-6.3" methanol 1.2

Pennisetum typhoideum Ergots Rivea corymbosa

8 bw Lg z!

108 116,119, 120 71

4

oa cl1

7 66

A. STOLL AND A. HOFMANN

diacetyl derivative in which the N atom in the 6 position is acetylated. The configuration of LXXV was deduced by comparison with festuclavine (LXV), a small yield of which is obtained by catalytic hydrogenation of chanoclavine.

B. THECLAVINEALKALOIDS Table I V (p. 764) contains chemical and physical data and indications on the source of the clavine alkaloids.

VII. Biogenesis of the Ergot Alkaloids I n all hypotheses relating to the biogenesis of the ergoline-lysergic acid moiety of ergot alkaloids, it was assumed that tryptophan constituted a main structural element. This assumption was later shown by experiment t o be correct. Injection of ~,~-tryptophan-t%C14 into the internodes of rye plants yielded ergot alkaloids whose lysergic acid portion was radioactive (121). Upon the addition of ~,~-tryptophan-/%C14 to saprophytic cultures of an ergot strain producing clavine alkaloids it was found that 10-39y0 was assimilated. It was further found that the addition of pyridoxalphosphate increased the yield. Experiments with tryptophan-C1400H, on the other hand, yielded practically inactive alkaloids, from which it may be deduced that the carboxyl group of the tryptophan is not incorporated in the ergoline structure (122). A large quantity of D,Ltryptophan-fl-Cl4 was also found t o be assimilated in the case of a Claviceps strain producing lysergic acid alkaloids ( 123).By incorporation of D-tryptophan labeled with tritium, it was proved that the unphysiological D-form of this amino acid was also utilized by the fungus (124). It was also possible t o incorporate indole-2-CI4 into the ergot alkaloids. Tryptamine-P-C14, however, is not utilized. Thus, it is improbable that decarboxylation of tryptophan occurs prior to its incorporation into the alkaloid molecule. An assimilation of 1.35yo for L-methionine-methylC14 shows that the N-methyl group of the ergot alkaloids is derived from methionine via a transmethylation reaction (125). In saprophytic cultures of the Pennisetum ergot fungus, only tryptophan, deuterated in the 5 or 6 position, was used for the synthesis of the clavine alkaloids without loss of the deuterium. Deuterium in the 4 position was lost. These experiments showed that the hypotheses, according to which 5-hydroxytryptophan would be an intermediate stage in the biosyntheses (126, 127), could not be correct (128).

21.

767

T H E ERGOT ALKALOIDS

Various hypotheses were forwarded concerning the nature of the structural element which was required in biosynthesis in addition to the tryptamine portion (127, 126, 129, 130). The following experiments show that, aside from tryptophan, an isoprenoidal 5-C compound participates in the formation of the ergoline structure. This was first postulated by Mothes (121). C

Mevalonic acid, the precursor of the active isoprene, was used in the incorporation studies. Both mevalonic acid-2-Cl4 and mevalonic acid2T or -4T were utilized by the fungus for the synthesis (131, 132). By means of a stagewise degradation of radioactive alkaloids formed with the aid of mevalonic acid-2-Cl4 and localization of the C14, it could be shown that mevalonic acid is incorporated into the molecule in the manner depicted by the accompanying reaction scheme (133, 134, 135). *CH~COOH

I

HO--CCH3 / NHz H2C\ CH2OH d C O O H

*C H ~ R

*COOH

The addition of mevalonic acid-1-04 to a pyroclavine- and festuclavineproducing fungus strain yielded inactive alkaloids which, in agreement with the scheme, showed that the carboxyl group of the mevalonic acid is not incorporated. Lowering of the assimilation of mevalonic acid-2-C14 by the addition of dimethylallylpyrophosphate or isopentenylpyrophosphate supported the assumption that mevalonic acid enters the alkaloid molecule via one of these activated isoprene radicals. This was confirmed by the incorporation of deuterated isopentenylpyrophosphate in alkaloids of the clavine type in saprophytic cultures of a Cluviceps strain (128). Experiments with alkaloids labeled with C14 showed that, in saprophytic cultures, agroclavine is converted into elymoclavine, penni-

768

A. STOLL AND A. HOFMANN

clavine, and isopenniclavine, whereas only penniclavine and isopenniclavine are produced by labeled elymoclavine. Labeled penniclavine and isopenniclavine on the other hand caused no formation of elymoclavine or agroclavine. The biogenesis thus appears to consist of a progressive hydroxylation (136).

MIL Derivatives of Ergot Alkaloids The potent and versatile pharmacological activity of natural ergot alkaloids prompted investigations on the relationship between their chemical structure and physiological action. It also caused the natural alkaloids to be modified chemically and the resulting variations of pharmacological and therapeutic action to be studied.

A. LYSERGIC ACIDAND DIHYDROLYSERGIC ACIDDERIVATIVES OF THE ACIDAMIDETYPE The azide process, employed in the first partial synthesis of ergometrine (81,66),was used in the production of a great number of lysergic acid and dihydrolysergic acid derivatives of the acid amide type. Aside from the homologs and analogs of ergometrine, resulting from the condensation of isolysergic acid azide with the corresponding amino alcohols (66, 137), many unsubstituted mono and dialkyl amides of lysergic acid and dihydrolysergic acid (138, 139, 140, 141), as well as cycloalkylamides and a great number of derivatives of the peptide type, in which the lysergic acid or dihydrolysergic acid is linked with amino acids or with di- or tripeptides, were produced (142, 143). I n the last few years, further processes for the partial synthesis of ergometrine and other lysergic acid derivatives of the amide type have been discovered. In accordance with the method of Garbrecht (144), the lithium salt of lysergic acid is converted in dimethylformamide solution with SO3 to form the mixed lysergic acid sulfuric acid anhydride which reacts with primary or secondary amines to give a good yield of the corresponding lysergic acid amides. Similarly, the process according to Pioch (145) is effected via a mixed anhydride, namely, via the mixed lysergic acid trifluoroacetic anhydride. I n this case, however, the yields of lysergic acid amide are usually not as good as in the SO3 process.

21. THE ERGOT ALKALOIDS

769

A further process, which is used on a technical scale and which, like the azide method, had its origin in the SANDOZ laboratories, employs lysergic acid chloride hydrochloride as the activated form (146,' 92). Furthermore, the method developed in peptide chemistry, using N,N'carbonyldiimidazole as condensation agent (147), may be used to produce acid amides of lysergic acid and dihydrolysergic acid (148). Of all these derivatives, two compounds have found medical application. D-Lysergic acid ( )-butanolamide(2), which is the next highest homolog of ergometrine, finds wide application in obstetrics, under the name of Methergine, owing to its uterotonic action and ability t o arrest post partum hemorrhage. &Lysergic acid diethylamide, which is characterized by its exceptional hallucinogenic and psycholytic action, has become known in experimental psychiatry as LSD 25 or under the trade name of Delysid and is also used as a drug aid in psychotherapy.

+

B. AMINOAND CARBAMIC ACIDDERIVATIVES O F 6-METHYLERGOLENE AND 6-METHYLERGOLINE The isomeric 6-methyl-8-amino-ergolenes(LXXVIa) and ergolines (LXXVIIa) were produced by a modified Curtius degradation of dlysergic acid azide, of d-isolysergic acid azide, or of the isomeric d-dihydrolysergic acid azides (149). R

R

I

H LXXVI

H LXXVII a: R = NHz b : R = NHCOOR' ,R' C : R = NHCON R '

Conversion of the corresponding isocyanates, resulting from the heating of the azides in benzene, with alcohols of general formula R'OH, yielded the carbamic acid esters LXXVIb (150), and with amines R"R"NH, the urea derivatives LXXVIc and LXXVIIc (151, 152).

770

A . STOLL AND A. HOFMANN

c. SUBSTITUTIONS IN T H E RINGS Y S T E M O F LYSERGIC ACID Positions 1 and 2 of the ring structure of lysergic acid are especially reactive and various substituents may be introduced. I n derivatives of lysergic acid and dihydrolysergic acid, in which the carboxyl group has an ester-like or acid amide-like substituent, the indole nitrogen atom could be acetylated with ketene (LXXVIIIa and b ) substituted with the hydroxymethyl radical (LXXVIIIc) using formaldehyde, by the dialkylaminomethyl radical (LXXVIIId) by means of a Mannich reaction (153), or by the cyanoethyl radical (LXXVIIIf) by the alkaline catalyzed addition of acrylonitrile (154). 0 II C-R

I

X LXXVIII a: X b: X C: X d: X e: X f:X

= = = = = =

COCH3 COCHzCOCHs CHzOH CHzN(a1kyl)z alkyl, benzyl CHzCHzCN

The hydrogen atom a t the indole nitrogen atom of lysergic acid derivatives was furthermore replaced by the methyl group and by larger alkyl radicals, e.g., ethyl, propyl, allyl, and benzyl radicals (LXXVIIIe). The alkylation is effected in liquid ammonia by reaction with the corresponding alkyl halides (155) and can be carried out most advantageously with the free carboxylic acids (156, 148). The N-methyllysergic acid butanolamide is a highly active serotonin antagonist and is widely used in the form of Deseril to cure stubborn headaches. Derivatives of lysergic acid or dihydrolysergic acid, in which the 1 position is free or substituted with an alkyl radical, may be halogenated in the 2 position with bromino- or iodosuccinimide, N-B,B-trichloro-4nitroacetanilide, and similar mild reagents (157) to compounds LXIX, a-c. Halogenated lysergic acid and dihydrolysergic acid derivatives may easily be differentiated from the starting materials containing no

21. THE

771

ERGOT ALKALOIDS

halogen since the van Urk color reaction, which is based on the oondensation of p-dimethylaminobenzaldehyde with the free 2 position of the indole compounds (158), is negative in the case of halogenated derivatives. 0 11

C R i

Rz LXXIX a : X = c1 b: x = Br c:X= J

Saturation of the double bond in the 9-10 position by catalytic hydrogenation has been discussed in Section 111,B. One molecule of water could also be added at the double bond in the 9-10 position. This reaction is effected in an acid aqueous solution of the alkaloids under intense irradiation with UV-light. The structures and configurations of the so-called lumi derivatives are depicted by formulas LXXX-LXXXIII (159, 160, 161, 162, 163, 164). R

0 /I C ,..H (-)-CHz

H07--/H LXXX Lumi-lysergic acid-I derivative

0

0

II

/I

H ,A-R

/

T

H

LXXXI Lumi-isolysergic acid-I derivative

II

r)-CH3 HO - H

(7-CHs HO----

0

H ,.-C-R

/

LXXXII Lumi-lysergic acid-I1 derivative

\

LXXXIII Lumi-isolysergic acid-I1 derivative

By reduction with zinc dust in hydrochloric acid, the lysergic acid derivatives may be hydrogenated selectively in the 2-3 position and dehydrogenated with mercuric acetate in glacial acetic acid t o form the starting materials (165).

D. BIOLOGICAL OXIDATIONOF LYSERGIC ACIDDERIVATIVES d-Lysergic acid diethylamide is converted with a microsome preparaacid tion, obtained from guinea pig liver, to 2-oxo-2,3-dihydro-d-lysergic

772

A. STOLL AND A. HOFMANN

diethylamide (LXXXIV). The constitution of this compound was ascertained by its partial chemical synthesis from d-lysergic acid diethylamide (166,167). 0

H LXXXIV

CH2OH

H LXXXV

On the other hand, ergometrine and d-lysergic acid diethylamide are secreted by the rat through the gall in the form of glucuronides of hydroxy derivatives (16s).The hydroxyergometrine LXXXV isolated was found to be identical with the 12-hydroxyergometrine obtained by oxidation of 2,3-dihydroergometrine with potassium nitrosodisulfonate (1 69).

IX. The Pharmacology and Therapeutic Use of Ergot Alkaloids and Their Derivatives The literature on pharmacological and clinical investigations of the ergot alkaloids and their derivatives is very comprehensive. There have already been more than one-thousand reports on lysergic acid diethylamide (LSD 25) ( 6 6 ) ,which was prepared only 20 years ago ;the publications on the pharmacological and clinical effects of ergotamine (13, 14), which was isolated in 1918, are far greater in number. I n this review it will therefore not be possible to go into details; it will be based on a schematic representation provided by Cerletti (170) and further elaborated by Hofmann (171). The pharmacodynamic properties of the ergot alkaloids cover a relatively broad spectrum of activity which can be classified as in the accompanying tabulation (see next page). All natural ergot alkaloids possess these six principal effects to a greater or lesser degree. The vasoconstrictor effect and the contractile effect on the smooth muscle of the uterus are the most important peripheral effects. The latter accounts for the classical indication of the ergot alkaloids in obstetrics where they are used to treat post partum hemorrhage and to accelerate uterine involution in the puerperium. The wide-

21. THE

ERGOT ALKALOIDS

773

Direct peripheral effects (on smooth muscles)

(1) Uterine contraction ( 2 ) Vasoconstriction

Indirect peripheral (humoral) effects

( 3 ) Serotonin antagonism (4) Adrenergic blockade

Central nervous effects

( 5 ) Bulbomedullary components: Vomiting, bradycardia, inhibition of the vasomotor center, and of the baroceptive reflexes (6) Mesodiencephalic components : Syndrome of ergotropic excitation with mydriasis, hyperglycemia, and hyper thermia

spread and successful use of ergotamine to mitigate migraine attacks is due to its tonifying effect on the smooth muscle of the blood vessels. The neurohumoral effects of the ergot alkaloids are manifested in an antagonism t o adrenaline and noradrenaline on the one hand and to 5-hydroxytryptamine (serotonin) on the other. The adrenolytic effect accounts for the use of the ergot alkaloids in internal medicine for the treatment of sympathetic overexcitation. The antagonism of the natural alkaloids to serotonin was discovered only in recent years, as it is not present to a marked degree. However, as will be shown subsequently, certain derivatives exhibit a marked and specific antagonism to serotonin. The effects of the ergot alkaloids on the central nervous system are very diverse as sites of action are situated in the vasomotor Center and the cardiac inhibitory center in the medulla oblongata as well as in the sympathetic structures of the diencephalon, particularly the hypothalamus. The inhibition of the vasomotor center and of t6e baroceptive reflexes and the stimulation of the vagal nuclei are raspphible for the vasodilator, hypotensive, and bradycardic effects, espddially in the case of the peptide type of alkaloid. Some also h a v w ktimulating effect on the vomiting center. Most ergot alkaloids stimulate the Sympathetic structure;. of the mesencephalon and diencephalon, particularly the hypothalamus, leading t o a syndrome of excitation with mydriasis, hyperglycemia, tachycardia, etc, This syndrome may be closely related t o the psychotomimetic effects of certain alkaloid derivatives, e.g., lysergic acid diethylamide. Considerable quantitative shifts in the pharmacodynamic effects can be produced by chemical changes in the periphery of the lysergic acid

7 74

A. STOLL AND A. HOFMANN

moiety of the ergot alkaloids. The six principal effects remain intact but some are so markedly modified that they are practically no longer manifest, whereas others are so enhanced that they determine the character of action of the substance and a t the same time restrict its range of action. Cerletti (170) made a graphic representation of the activity spectra of the various ergot alkaloids, making use of the symbols Site of oction:

Symptoms:

peripherol

central

FIG.6. Activity spectra of ergot alkaloids.

illustrated in Fig. 6. If the relative dosage scale is recorded on the ordinate and the principal effects are entered on the abscissa in a series 1-6 from left to right, more or less comparative diagrams are obtained (Fig. 7) which clearly show the quantitative differences in the various activity ranges for the individual substances. Figure 7a depicts the spectrum of activity of ergotamine. This possesses the typical effect of the ergot alkaloids in a well-balanced manner. Ergotamine exerts a full-strength contractile effect on the uterus and on the smooth muscle of the vessels,

2 1. THE

775

ERGOT ALKALOIDS

reduces adrenergic activity, and elicits central effects by inhibiting the vasomotor centers. The stimulation of higher nervous structures is less pronounced and occurs only after the administration of toxic doses. Based on this spectrum of activity, ergotamine is an excellent hemostatic in obstetrics while it is also used in internal medicine and neurology as a sympathicolytic for the treatment of sympathicotonic conditions (01

Ergotamine

(b) Ergometrine

I

(c) Dihydroergotomine

(d) LSD 25

FIG. 7. Comparative diagrams showing the quantitative differences of activity of various ergot products.

and as a vasoconstrictor in migraine and related vascular headaches. The single therapeutic dose in the form of ergotamine tartrate is 0.251.0 mg. The other natural polypeptide alkaloids, e.g., the alkaloids of the ergotoxine group (ergocristine, ergokryptine, and ergocornine) have a, spectrum of activity similar t o that of ergotamine but their toxic effects are more pronounced. For this reason they have not attained the clinical importance of ergotamine.

776

A. STOLL AND A. HOFMANN

The activity spectrum of ergometrine (Fig. 7b) is quite different from that of ergotamine (Fig. 7a). It exerts a marked uterotonic effect, whereas its adrenolytic action is practically insignificant and the central nervous effects are only manifest after high doses. Ergometrine is therefore used mainly in obstetrics. I n recent years it has been demonstrated that ergometrine exerts a pronounced antiserotonin effect. Partial synthesis of a great number of ergometrine analogs, in which the amino-propanol moiety is replaced by other amino alcohols or by simple primary and secondary amines, results in significant shifts in the pharmacodynamic profile ; these shifts are especially evident quantitatively, but also qualitatively. Substitution of the propanolamine moiety in ergometrine by butanolamine yields methylergometrine, which exerts a greater effect than the natural alkaloid on smooth muscle and is therefore used on a large scale under the name Methergin as a uterotonic and hemostatic in obstetrics for the management of the third stage as well as in gynecology. Lysergic acid diethylamide (LSD 25), trade name Delysid, which is partly synthetic, exhibits a quite unexpected type of effect as can be seen from Fig. 7d where comparison is made with ergometrine. LSD 25 possesses a clear-cut uterotonic effect but exerts practically no adrenolytic effect although it is a potent antagonist of serotonin. This product elicits marked excitation of central nervous structures. I n minimal doses it elicits mydriasis, hyperthermia, and hyperglycemia. This syndrome of central excitation is closely related to the psychotomimetic effects of LSD 25, which has become of considerable importance in experimental psychiatry and has given a substantial impetus to modern psychopharmacology. The most striking effects of LSD are colored vision and hallucinations. LSD 25 is still regarded as an experimental tool in psychotherapy and its use as an adjuvant in psychoanalysis meets a growing interest. Active doses usually range from 30 to 150 pg. Hydrogenation of the double bond in position 9-10 of the lysergic acid moiety results in fundamental changes in the pharmacodynamic action, as a comparison of the activity spectra of ergotamine (Fig. 7a) and dihydroergotamine (Fig. 7c) makes evident. The vasoconstrictor and uterotonic effects of the dihydrogenated derivative have been markedly attenuated, as has the stimulation of central structures, so that these effects are hardly elicited by therapeutic doses. On the other hand the adrenolytic effect and the central inhibition of the vasomotor centers are markedly enhanced. This is manifested clinically in vasodilatation, hypotension, and a certain sedative action. These properties are still more marked in the case of the hydrogenated derivatives of the ergotoxine alkaloids. A combination of dihydro-

21. THE

ERGOT ALKALOIDS

777

ergocristine, dihydroergokryptine, and dihydroergocornine in equal proportions has been widely used under the name Hydergine for the treatment of peripheral and cerebral vascular disorders and of essential hypertension. (a)

2-Br-LSD

I /

(c) Br - Hydergin

(d) I-Methyllysergic acid butanolamide (UML 491)

FIG. 8. Comparative diagrams showing the autivity of ergot substances made by substitutions in the indole nucleus.

Substitutions a t the nitrogen and in position 2 of the indole portion of lysergic acid have a striking influence on pharmacodynamic properties. Many such substitution products have been prepared and pharmacologically investigated. The activity spectra of four such substances (Fig. 8) demonstrate how very diverse the effects of ergot substances can be made by substitutions in the indole nucleus.

778

A . STOLL AND A. HOFMANN

Bromination of LSD 25 yields 2-bromolysergic acid diethylamide which exerts practically no psychotomimetic effect (Fig. 8a) ; the central excitation syndrome is reduced, whereas the antiserotonin effect remains intact, but the effect on smooth muscle is practically absent. This compound is almost as potent as LSD 25 in antagonizing serotonin but does not elicit hallucinogenic effects. Bromination of ergotamine to yield 2-bromoergotamine (Fig. 8b) enhances the adrenolytic effect and reduces the oxytocic effect to such a degree that it cannot be demonstrated, e.g., on the rat uterus. The vasoconstrictor, hypertensive, and central effects of bromoergotamine are also weakened, resulting in significantly lower toxicity. The introduction of bromine in the hydrogenated derivatives of the ergotoxine alkaloids (Hydergine) reduces their adrenolytic properties, but the central hypotensive effects remain intact (Fig. Sa). The other central effects, particularly the emetic effect, and toxicity are decreased. Methylation a t the indole nitrogen group specifically enhances the serotonin antagonism exhibited by all ergot alkaloids. For example, 1-methyl-LSD and 1-methyl-2-bromo-LSD exhibit antiserotonin activity several times greater than that of their nonmethylated parent compounds, The same phenomenon is observed when ergotamine is methylated. It is still more marked when ergometrine or methylergometrine (Methergin) is methylated a t the indole nitrogen. 1-Methyl-D-lysergic acid propanolamide is 2.5 times more active than lysergic acid diethylamide (LSD 25) in antagonizing serotonin ; 1-methyl-D-lysergic acid butanolamide (Deseril, Sansert) is, depending on the test employed, 4-6 times more powerful than LSD 25 as a serotonin antagonist and is therefore the most potent serotonin antagonist so far discovered (Fig. 8d). This antagonism is very specific ; Deseril (Sansert) inhibits serotonininduced potentiation of barbiturates but has no effect on phenothiazineinduced potentiation. On the other hand the oxytocic effect of Deseril is 15-20 times weaker than that of Methergin. Concurrently, the toxicity of the 1 -methyl compound is markedly decreased. 1-Methyl-D-lysergic acid butanolamide does not exert any significant adrenolytic effect ; it does not possess the vasoconstrictor and pressor properties so typical of the natural ergot alkaloids. As it also lacks psychotomimetic effects, it can be used in far higher doses than methylergometrine or LSD 2 5 . As investigations in recent years have shown that serotonin may play an important role as a transmitter substance, Deseril will undoubtedly contribute to the elucidation of the functions of serotonin in the organism. Investigations have already shown that it exerts very beneficial effects in the preventive treatment of migraine and other vascular headaches

21.

THE ERGOT ALKALOIDS

779

and that it offers promising results in the treatment of rheumatic disorders and certain symptoms in the carcinoid syndrome. However, several years will be required before the therapeutic range of l-methylmethylergometrine can be definitely established. All ergot alkaloids which have so far been used therapeutically are lysergic acid derivatives. Representatives of the second main group, the clavine alkaloids, have also been found to be pharmacodynamically active, but as yet, none has been found to exert effects that can be utilized in therapy. The uterotonic and sympatholytic actions are less prominent in their pharmacological spectra of activity but, for example, elymoclavine and agroclavine have a pronounced central excitatory action which is attributed to their stimulation of sympathetic centers ( 172). The ergot alkaloids, especially those containing the lysergic acid radical, occupy a special position among the indole alkaloids not only by virtue of their origin and their peptide structure, but also by virtue of the diversity of pharnincodynamic properties which is not often found among plant bases. Relatively slight chemical modifications yield derivatives in which the properties present in the natural substances are so selectively enhanced that new types of drugs for more closely defined ranges of indications are obtained. The ergot alkaloids have proved not only to be highly interesting natural substances from the chemical point of view but also a veritable treasure house for new types of drugs. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

L. R. Brady, Lloydia 25, 1 (1962). J. Steams, Med. Repository N . Y . 5, 308 (1808). M. Vauquelin, Ann. Chim.Phys. 3, 337 (1816). C. Tanret, Compt. Rend. Acad. Sci. 81, 896 (1875). C. Tanret,J. Pharm. Chim. 23, 17 (1876). G. Bsrger and F. H. Carr, Chem. News 94, 89 (1906). G. Barger and F. H. Carr, J. Chem. Soc. 91, 337 (1907). F. Kraft, Arch. Pharm. 244, 336 (1906). F. Kraft, Arch. Pharm. 245, 644 (1907). A. Stoll and A. Hofmann, Helv. Chim. Acta 26, 1570 (1943). H. H. Dale, J . Physiol. (London) 34, 163 (1906). F. Kraft, Arch. Pharm. 244, 359 (1906). A. Stoll, Helv. Chim. Acta 28, 1283 (1945). SANDOZ Ltd., Swiss Patent 79,879 (1918); SANDOZ Ltd., Ger. Patent 357,272 (1922). A. Stoll, Schweiz. Apotheker-Ztg. 60, 341, 358, 374 (1922). W. A. Jacobs and R. G. Gould, Jr. J . Biol. Chem. 120, 141 (1937). A. Stoll, Schweiz. Apotheker-Ztg. 60, especially see p. 364 (1922). S. Smith and G. M. Timmis, J . Chem. SOC. p. 1440 (1936).

780

A. STOLL AND A. HOFMANN

19. A. Stoll, T. Petrzilka, 3. Rutschmann, A. Hofmann, and H. H. Gunthard, Helv. Chim. Acta 37, 2039 (1954). 20. W. A. Jacobs and L. C. Craig, J . Biol.Chem. 104, 547 (1934). 21. W. A. Jacobs and L. C. Craig, J . Biol.Chem. 106, 393 (1934). 22. S. Smith and G. M. Timmis, J . Chem. Soc. p. 674 (1934). 23. C. I. Abou-Char, L. R. Brady, and V. E. Tyler, Jr., Lloydia 24,89 (1961). 24. A. Stoll and A. Hofmann, 2. Physiol. Chem. 250, 7 (1937). 25. A. Stoll and A. Hofmann, Helv. Chirn. Acta 26, 922 (1943). 26. E. C. Kornfeld, E. J. Fornefeld, G. B. Kline, M. J . Mann, D. E. Morrison, R. G. Jones, and R. B. Woodward, J. Am. Chem. SOC.78,3087 (1956). 27. C. C. Keller, Schzueiz. Wochschr. Chem. Pharm. 34, 65 (1896). 28. H. W. van Urk, Pharm. Weekblad. 66, 473 (1929). 29. M. I. Smith, Public Health Rept. (U.S.) 45, 1466 (1930). 30. N. L. Allport and T. T. Cocking, Quart. J . Pharm. Pharmacol. 5 , 341 (1932). 31. F. Schlemmer, P. H. A. Wirth, and H. Peters, Arch. Pharm. 274, 16 (1936). 32. N. v. BBkBsy, Biochem. 2. 302, 187 (1939). 33. W. A. Jacobs, J . Biol. Chem. 97, 739 (1932). 34. W. A. Jacobs and L. C. Craig, J . Biol. Chem. 113, 767 (1936). 35. W. A. Jacobs and L. C. Craig, J . Biol. Chem. 111, 455 (1935). 36. W. A. Jacobs and L. C. Craig,J. Biol. Chem. 128, 715 (1939). 37. W. A. Jacobs, L. C. Craig, and A. Rothen, Science 83, 166 (1936). 38. L. C. Craig, T. Shedlovsky, R. G. Gould, Jr., and W. A. Jacobs, J. Biol. Chem. 125, 289 (1938). 39. W. A. Jacobs and L. C. Craig, J . A m . Chem. SOC. 60, 1701 (1938). 40. A. Stoll, A. Hofmann, and F. Troxler, Helv. Chin%.Acta 32, 506 (1949). 41. J. B. Stenlake, Chern. Ind. (London)p. 1089 (1953). 42. A. Stoll and A. Hofmann, Helv. Chim. Acta 26, 2070 (1943). 43. A. Stoll, A. Hofmann, and T. Petrzilka, Helv. Chim.Acta 29, 635 (1946). 44. A. Stoll and J. Rutschmann, Helv. Chim. Aeta 36, 1512 (1953). 45. R. C. Cookson, Chem. I d . (London)p. 337 (1953). 46. H. G. Leemann and S. Fabbri, Helv. Chim. Acta 42, 3696 (1959). 47. P. Stadler and A. Hofmann, Helv. Chim. Acta 45, 2005 (1962). 48. F. C. Uhle and W. A. Jacobs, J . Org. Chem. 10, 76 (1945). 49. A. Stoll, J. Rutschmann, and W. Schlientz, Helv. Chim. Acta 33, 375 (1950). 50. E. C. Kornfeld, E. J. Fornefeld, G. B. Kline, M. J. Mann, R. G. Jones, and R. B. Woodward, J . Am. Chem. Soe. 76, 5256 (1954). 51. F. R. Atherton, F. Bergel, A. Cohen, B. Heath-Brown, and A. H. Rees, Chem. l n d . (London)p. 1151 (1953). 52. C. A. Grob and E. Renk, Welv. Chim. Acta 44, 1531 (1961). 53. G. N. Walker and B. N. Weaver, J . Org. Chem. 26, 4441 (1961). 54. A. Stoll and J. Rutschmann, Helv. Chim. Acta 33, 67 (1950). 55. R. G. Gould, Jr. and W. A. Jacobs,J. Am. Chem. SOC.61,2891 (1939). 56. A. Stoll and T. Petrzilka, Helv. Chim. Acta 36, 1125 (1953). 57. A. Stoll and T. Petrzilka, Helv. Chim. Acta 36, 1137 (1953). 58. A. Stoll and J. Rutschmann, Helv. Chim. Acta 37, 814 (1954). 59. C. A. Grob and J. Voltz, Helv. Chim. Acta 33, 1796 (1950). 60. F. C. Uhle, J . Am. Chem. SOC. 73, 2402 (1951). 61. J . A. Barltrop and D. A. H. Taylor, J . Chem. SOC. p. 3399, 3403 (1954). 62. H. Plieninger, Ber. 86, 25, 404 (1953). 63. H. Plieninger and T. Suehiro, Ber. 88, 550 (1955).

21. THE 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87.

88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101.

102. 103. 104. 105.

ERGOT ALKALOIDS

781

H. Plieninger and G. Werst, Ber. 89, 2783 (1956). A. Stoll, T. Petrzilka, and J. Rutschmann, Helw. Chim. Acta 35, 1249 (1952). A. Stoll and A. Hofmann, Helw. Ch.im. Acta 26, 944 (1943). S. Smith and G. M. Timmis, J . Chem. SOC. p. 763 (1932). A. Stoll and A. Hofmann, Helw.Chim. Acta 33, 1705 (1950). F. Arcamone, C. Bonino, E. B. Chain, A. Ferretti, P. Pennella, A. Tonolo, and L. Vero, Nature 187, 238 (1960). A. Hofmann and H. Tscherter, Ezperientia 16, 414 (1960). A. Hofmann, Planta M e d . 9, 354 (1961). J. F. Spilsbury and S. Wilkinson, J . Chem. Soc. p. 2085 (1961). R. L. Grant and S. Smith, Nature 137, 154 (1936). H. W. Dudley and C. Moir, Brit. Med. J . I, 520 (1935). A. Stoll and E. Burckhardt, Compt. Rend. Acad. Sci. 200, 1680 (1935). M. S. Kharasch and R. R. Legault, Science 81, 388, 614 (1935). M. R. Thompson, Science 81, 636 (1935). S. Smith and G. M. Timmis, Nature 136, 259 (1935). S. Smith and G. M. Timmis, J . Chem. SOC. p. 1166 (1936). W. A. Jacobs and L. C. Craig, Science 82, 16 (1935). A. Stoll and A. Hofmann, 2. Physiol. Chew$.251, 155 (1938). W. A. Jacobs and L. C. Craig, J . Biol. Chem. 110, 521 (1935). W. A. Jacobs and L. C. Craig, J . Am. Ghem. SOC. 57, 960 (1935). W. A. Jacobs and L. C. Craig, J . Org. Chem. 1, 245 (1936). A. Stoll, A. Hofmann, and B. Becker, Helw. C h im. Acta 26, 1602 (1943). W. A. Jacobs and L. C. Craig, J . Biol. Chem. 122, 419 (1938). G. Barger, in, " Handbuch der experimentellen Pharmakologie " (E. Abderhalden, ed.) Ergaenz. Werk Vol. VI, 84, 222. Springer, Berlin, 1938. A. Stoll, T. Petrzilka, and B. Becker, Helw. Chim. Acta 33, 57 (1950). A. Stoll, A. Hofmann, and T. Petrzilka, Helv. Chim.Acta 34, 1544 (1951). A. Stoll, A. Hofmann, H. G. Leemann, H. Ott, and H. R. Schenk, Helw. Chim. Acta 39, 1165 (1956). D. M. Wrinch, Nature 137, 411; 138, 241 (1936). A. Hofmann, A. J. Frey, and H. Ott, E z p e r i e d a 17, 206 (1961). M. M. Shemyakin, E. S. Tchaman, L. I. Denisova, G. A. Ravdel, and W. J. Rodionow, Bull. SOC.Chim. France p. 530 (1959). W. K. Antonov, G. A. Ravdel, andM. M. Shemyakin, Chimia (Aaruu)14,374 (1960). M. Abe, T. Yamano, S. Yamatodani, Y. Kozu, M. Kusumoto, H. Komatsu, and S. Yamada, Bull. Agr. Chem. SOC. J a p a n 23,246 (1959). W. Schlientz, R. Brunner, A. Hofmann, B. Berde, and E. Stiirmer, Pharrn. Acta Helw.36, 472 (1961). W. Schlientz, R. Brunner, F. Thudium, and A. Hofmann, Experientia 17,108 (1961). S. Smith and G. 81.Timmis, J . Chem. SOC. p. 396 (1937). A. Stoll and E. Burckhardt, 2. Physiol. Chem. 250, 1 (1937). M. Abe, T. Yamano, Y. Kozu, and M. Kusumoto, J . Agr. Chem. SOC. Japan 25, 458 (1952). G. L. Szendey, K. H. Renneberg, and P. Hartmann, Naturwissenschaften 48, 223 (1961). S. Yamatodani and M. Abe, Bull. Agr. Chem. SOC. Japan 19, 94 (1955). A. Stoll, A. Hofmann, and W. Schlientz, Helv. Chim. Acta 32, 1947 (1949). S. Yamatodani and ill. Abe, Bull. Agr. Chem. SOC. Japan 20, 95 (1956). S. Yamatodani and M. Abe, Bull. Agr. Chem. SOC. J a p a n 21, 200 (1957).

782

A. STOLL AND A. HOFMANN

M. Abe, J. Agr. Chem. SOC. Japan 28, 44 (1954). E. Schreier, Helv. Chim. Acta 41, 1984 (1958). A. Hofmann, R. Brunner, H. Kobel, and A. Brack, Helv. Chim. Acta 40,1358 (1957). M. Abe, S. Yamatodani, T. Yamano, and M. Kusumoto, Bull. Agr. Chem. SOC. Japan 19, 92 (1955). 110. M. Abe, Ann. Rep. Takeda Res. Lab. 10, 73 (1951). 111. M. Abe and S. Yamatodani, J . Agr. Chem. SOC. J a p a n 28, 501 (1954). 112. A. Stoll, A. Brack, H. Kobel, A. Hofmann, and R. Brunner, Helv. Chim. Acta 37, 1815 (1054). 113. M. Abe and S. Yamatodani, Bull. Agr. Chem. Soc. J a p a n 19, 161 (1955). 114. M. Abe, S. Yamatodani, T. Yamano, and M. Kusumoto, Agr. Biol. C'hem. (Tokyo)25, 594 (1961). 115. S . Yamatodani, Ann. Rep. Takeda Bes. Lab. 19, 1 (1960). 116. M. Abe, 8. Yamatodani, T. Yamano, and M. Kusumoto, Bull. Agr. Chem. SOC. Japan 20, 59 (1956). 117. M. Semonsk9, M. Beran, and K. Macek, Colleclion Czech. Chem. Commun. 23, 1364 (1958). 118. M. Abe, Ann. Rep. Takeda Res. Lab. 10, 145 (1951). 119. H. Kobel, R. Brunner, and A. Brack, Ezperientia 18, 140 (1962). 120. D. Groger, V. E. Tyler, Jr., and J. E. Dusenberry, Lloydia 24, 97 (1961). 121. K. Mothes, F. Weygand, D. Groger, and H. Grisebach, 2. Naturforsch. 13b, 41 (1958). 122. D. Groger, H. J. Wendt, K. Mothes, and F. Weygand, 2. Naturforsch. 14b, 355 (1959). 123. W. A. Taber and L. C. Vining, Chcm. I n d . (London)p. 1218 (1959). 124. D. Groger, K. Mothes, H. Simon, H. G. Floss, and F. Weygand, 2. Naturforsch. 16b, 432 (1961). 125. R. M. Baxter, S. I. Kandel, and A. Okany, Chem. Ind. (London)p. 1453 (1961). 126. J.Harley-Mason, Chem. Ind. (London)p. 251 (1954). 127. E. E. van Tamelen, Ezperientia 9, 457 (1953). 128. H. Plieninger, R. Fischer, G. Keilich, and H. D. Orth, Ann. 642,214 (1961). 129. N. L. Wendler, Ezperientia 10, 338 (1954). 130. A. Feldstein, Ezperientia 12, 475 (1956). 131. D. Groger, K. Mothes, H. Simon, H. G. Floss, and F. Weygand, 2. Naturforsch. 15b, 141 (1960). 132. E. H. Taylor and E. Ramstad, Nature 188, 494 (1960). 133. A. J. Birch, B. J . McLoughlin and H. Smith, Tetrahedron Letters No. 7, 1 (1960). 134. S. Bhattacharji, A. J. Birch, A. Brack, A. Hofmann, H. Kobel, D. C . C . Smith, H. Smith, and J. Winter, J . Chem. Soc. p. 421 (1962). 135. R. M. Baxter, S. I. Kandel, and A. Okany, Tetrahedron Letters No. 17, 596 (1961). 136. S. Agurell and E. Ramstad, Tetrahedron Letters No. 15, 501 (1961). 137. M. Semonskj., A. Cernj., and V. ZikLn, Collection Czech. Chem. Commun. 21, 382 (1956). 138. A. Stoll and A. Hofmann, Helv. Chim. Acta 38, 421 (1955). 139. M. Semonskj., V. Zikan, and Z. Votova, Collection Czech. Chem. Commun. 22, 1632 (1957). 140. v. Z i k h and M. Semonskj., Collection Czech. Chem. Commun. 24, 1274 (1959). 141. M. Semonskj. and V. ZikLn, Collection Czech. Chem. Commun. 25, 2038 (1960). 142. A. Stoll, A. Hofmann, E. Jucker, T. Petrzilka, J. Rutschmann, and F. Troxler, Helv. Chim. Acta 33, 108 (1950). 106. 107. 108. 109.

2 1. THE ERGOT

ALKALOIDS

783

143. A. Stoll and T. Petrzilka, HeZw. Chim. Acta 35, 589 (1952). 144. W. L. Garbrecht, J . Ory. Chem. 24, 368 (1959). 145. R. P. Pioeh, U.S. Patent 2,736,728 (1956). 146. SANDOZ Ltd., Fr. Patent 1,308,758. 147. R. Paul and G. W. Anderson, J . Am. Chem. Soc. 82, 4596 (1960). 148. A. Cernf and M. Semonsky, CoZlection Czech. Chem. Commun. 27, 1585 (1962). 149. A. Hofmann, Helw. Chim. Acta 30, 44 (1947). 150. F. Troxler, Helw.Chim. Acta 30, 163 (1947). 151. V. Zikan and M. Semonsky, CoZZection Czech. Chem. Commun. 25, 1922 (1960). 152. SANDOZ Ltd., Belg. Patent 607,502; 609,010. 153. F. Troxler and A. Hofmann, Helw.Chim. Acta 40, 1706 (1957). 154. SANDOZ Ltd., Swiss Patent appl. 13071/61 (Nov. 10, 1961) and Appl. for a patent of addition 11100/62 (Sept. 9, 1962). 155. F. Troxler and A. Hofmann, HeZw. Chim. Acta 40, 1721 (1957). 156. SANDOZ Ltd., Belg. Patent 607,294; Fr. Patent 1,297,632. 157. F. Troxler and A. Hofmann, Helw. Chim. Acta 40, 2160 (1957). 158. H. W. Dibbern and H. Rochelmeyer, Arzneimittel-Forsch. 13, 7 (1963). 159. A. Stoll and W. Schlientz, Helv. Chim.Acta 38, 585 (1955). 160. H. Hellberg, Acta Chem. Scand. 11, 219 (1957). 161. H. Hellberg, Pharm. Weekblad. 93, 1 (1958). 162. H. Hellberg, Acta Chem. Scand. 12, 678 (1958). 163. H. Hellberg, Acta Chem.Scand. 13, 1106 (1959). 164. H. Hellberg, Acta Chem. &and. 16, 1363 (1962). 165. P. Stadler, A. J. Frey, and F. Troxler, Chimia ( A a r a u )15,575 (1961). 166. K. Freter, J. Axelrod, and B. Witkop,J. Am. Chem. Soc. 79, 3191 (1957). 167. F. Troxler and A. Hofmann, Helv. Chim. Acta 42, 793 (1959). 168. M. B. Slaytor and S. E. Wright, J . Med. Pharm. Ghem. 5, 483 (1962). 169. P. A. Stadler, A . J . Frey, F. Troxler,andA. Hofmann, Helw.Chim. Acta47,76G (1964). 170. A. Cerletti, Proc. Intern. Congr. h'euro-Psychopharm. lst, Rome, 1958, p. 117 (Pub. 1959). 171. A. Hofmann, Australasian J . Pharm. N.S. 42, 7 (1961). 172. T. Yui and T. Takeo, Japan. J . Pharmacol. 7, 157 (1958). 173. A. Hofmann, H. Ott, R. Griot, P. A. Stadler, and A. J. Frey, Helv. Chim. Acta 46, 2306 (1963). 174. P. A. Stadler, A. J. Frey, and A. Hofmann, HeZv. Chim. Acta 46, 2300 (1963). 175. SANDOZ Lt,d., Basel, Pharm. Chem. Res. Lab. Results to be published in HeZw. Chim. Acta.

This Page Intentionally Left Blank

---CHAPTER

22-

THE AJMALINE-SARPAGINE ALKALOIDS W. I. TAYLOR Research Department, C Z B A Pharmaceutical Company, Division of C Z R A Corporation, Summit, New Jersey

I. The Ajmaline Group ................................................ A. Determination of the Structure of Ajmaline .......................... B. The Stereochemistry of Ajmaline ............. ... C. Demethylation of Ajmaline Derivatives. ............................. ....................................... D. Congeners of Ajmaline. . . E. The 16-CarbomethoxytetraphyllicineSubgroup. . . . . . .....

7 89 789 792 796 797 800

It. TheSarpagineGroup ................................................. A. The Structure of Sarpagine. . . B. The 16-Carbomethoxy-10-deo

804 805

111. Mass Spectra of the Ajmaline-Sarpagine Alkaloids. .......................

808

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

81 1

IV. Pharmacological Notes, References

812

Detailed knowledge of the ajmaline-sarpagine group of alkaloids is of quite recent origin, being intimately connected with the extensive investigations of the Rauwol$a bases. As the structure of ajmaline began to unfold, it was guessed to conceal a new variant of the yohimbinoid skeleton (1).This hypothesis was not only correct, but it also contributed to the structural solution of the molecule ( 2 ) . 'Ajmaline, therefore, was simply a new member of an increasing number of compounds, all of which were thought to originate from the same starting materials, the amino acids tryptophan and dihydroxyphenylalanine [or in a minor variant its hydroaromatic precursors (3), e.g., prephenic acid (4)]and formaldehyde or their equivalents ( 5 ) . According to these ideas, the genesis of all the indole alkaloids with the sole exception of the yohimbines required a fission of the equivalent of a dihydroxyphenylalanine ring. Because the yohimbines were the only group of alkaloids t o contain an intact carbocyclic ring E and to avoid the difficulties inherent in the insertionof their carboxy groups, it has been suggested (6),as a result of visual dissection, that it would be simpler to view the origin of the hydroaromatic moiety of these and related alkaloids from an open chain 3C plus 6C - 1C unit. This progenitor in the form of I can be recognized in 785

786

W. I. TAYLOR

the hydroaromatic moieties of all the indole alkaloids shown in Chart I. Whether these units are already combined before the biosynthesis of an individual alkaloid or introduced stepwise is not known, but experimental support for this idea has come from feeding labeled malonic (3C), acetic (BC),and formic (1C) acids to certainindolealkaloid-yieldingplants, in particular to RuuwolJia serpentina ( 7 ) . Detailed mechanisms for the transformations implied above can be written, but their plausibility cannot be taken as a proof of their reality, and experimental results from those laboratories engaged in these aspects of biochemistry are bound to

Pseudoindoxyl analogs

(HOOC) Yohirnbines

Ring E oxygen heterocycles

I Yohimbinoid preriirsor

I/

0 Ring E Seco compounds

Ajmaline group

*

I

CHARTI. Classes of yohimbinoid alkaloids (excluding the strychnoid and gelsemine subgroups).

22. THE AJMALINE-SARPAGINE ALKALOIDS

787

turn up unexpected consequences of the general scheme. There seems to be little uncertainty about the biosynthetic steps which lead to ajmaline from I, since many isolated alkaloids--e.g., vellosimine (see Chart IV), deoxysarpagine (I11; R = H), and vomilenine (XX1V)-can be regarded as trapped intermediates. Most of the ajmaline-sarpagine group of alkaloids have been interrelated and their detailed structures rest on three separately derived proofs. Two of these were X-ray crystallographic analyses of sarpagine derivatives and the third was a facile unfolding of the ajmaline molecule by stereospecific processes which led to a ring E-seco compound of known absolute configuration (Chart 111). The ajmaline-sarpagine group TABLE I THEAJMALINE-SARPAQINE ALKALOIDS

I1

I11 Sarpagine

Aj rnaline

RauwoZfia alkaloids (details, see Table I, Chapter 13) A. Known structures Ajmaline (11; rauwolfine, neoajmaline, raugalline) ; isoajmaline (20-epi-21-epiajmaline) ; ajmalidine (ajmaline-17-ketone); vomalidine (12-methoxyajmalidine) ;sandwicine (17-epiajmaline) : vomilenine (l-demethylajmalin-l,19-diene-I7-O-acetate) ; perakina (XXV; rearrangement product of vomilenine). Tetraphyllicine [21-deoxyajmaline-19-ene, semperflorine (?)I ; rauvomitine (tetraphyllicine-17-0-trimethoxybenzoate) ; mauiensine (17-epitetraphyllicine) ; purpeline ( 12-methoxytetraphyllicinone); mitoridine ( 12-hydroxytetraphyllicinone) . Sarpagine' (111; R = OH).

B. Alkaloids of unknown structure Name Sandwicensine Ajmalinine Rauwolfinine Neosarpagine

Formula

mP

ClgHzzNzO 260"-262" C ~ o H z ~ N ~180"-181" 03 Cz0Hz6N202 235"-236' 390" C19HzzNzOz

[alD

+ 56" (MeOH) - 97" (CHC13)

Observations

1-Demethylmauiensine? Possibly 3-epi-a-yohimbine - 35" (EtOH) Probably impure isoajmaline Probably sarpagine

788

W. I. TAYLOR

TABLE 1-ontinued Non-Rauwolfia Alkaloids

Name

Formula

mP

Vincamajine (16-carbo- CzzH26NzO3 225" methoxy-2-epitetraphyllicine Vincamajine-17-0Cz4NzsNzO4 185" acetate (vincamedine) 1-Demethylvincamajine Cz1Hz4N203 276" (quebrachidine) Sarpagine- 10-methyl C ~ O H Z ~ N Z202"-203" OZ (lochnerine) Lochnerine nietho salt CalHz7NzOzI 235"-238" (lochneram) 10-Deoxysarpagine CzoHzzKzO 233"-235" (tombozine; or 275" vellosiminol) 10-Deoxysarpagine C2oHz5NzOCl248"-249" metho salt (mscusine

B) Vellosimine (10-deoxy- C1gHzoNzO 305"-306" sarpagine aldehyde) Akuammidine (16Cz1H24Nz03 234"-236" carbomethoxy - 10deoxysarpagine) C Z I H Z ~ N T243"-246" ~O~ Polyneuridine ( 16carbomethoxy-16-epi10-deoxysarpagine) Polyneuridine metho C ~ Z H Z ~ K Z 248" O~C~ salt (macusine A) 1-Methyl polyneuridine CzzHz6Nz03 223"-224" (voachalotine)

a

ID

Source*"(structured)

- 55" (alc.)

Vds, Vm9 (11) Tllo

- 66 (CHC13)

Vd12, Vm9 ( 11)

+ 54" (CHClJ) + 72 (EtOH)

Aq13

(13)

C14, Vr15

(16)

+41" (EtOH)

C17

+ 48" (MeOH)

GvZ0

(20)

+24" (MeOH) Pn24, Vd25 (11, 27) Rs26

- 3" (CHC13)

V C ~ ~ (11,30)

Also occurs in Vinca difformis.

' Plant key: Ap, Aspidosperrna polyneuron Mull. Arg.; Aq, Aspidosperma

quebrachoblancho Schlecht. ; C , Calebash curare; Dc, Diplorrhynchus condylocarpon (Mull. Arg.) Pichon ssp. mossumbicensis (Benth.) Durign. ; Gv, Geissospermurn vellosii Allem. ; Pn, PieraLima nitidu Th. et H. Dur.; Rs, Rhazya strictu Decaisne; St,Strychnos toxijeru Schomb.; T1, Tonduzin long
therefore, had the predicted (7a) stereochemistry a t the carbons (asterisked in Chart I)equivalent t o C-15 of yohimbine as do the other yohimbinoid alkaloidal systems pictured. Table I lists pertinent data on the various alkaloids with the exception

22.

THE AJMALINE-SARPAOINE ALKALOIDS

789

of the details of the Rauwoljia bases which are to be found in Table I of Chapter 13. The compact multiring systems have a number of very interesting properties which will be discussed below, and although some progress has been made toward synthesis of these heterocycles, none has yet been completed.

I. The Ajmaline Group A. DETERMINATION OF THE STRUCTURE OF AJMALINE The chemistry of ajmaline [ I I ; rauwolfine (31)] was finally put on a proper basis in 1954 ( l ) ,isoajmaline appearing to be its C-ethyl epimer, and neoajmaline being simply anhydrous ajmaline. While isoajmaline may be a natural product, it could have been produced artificially from ajmaline in those cases where it was reported (32). Anhydrous ajmaline is not easy to obtain directly, since it usually comes out of methanol in the forin of a highly crystalline solvate which has a characteristic but indefinite melting point. This behavior is a function of the type of melting point apparatus and the rate of heating. This is probably the reason that ajmaline methanolate has been a t least once thought to be a new methoxy alkaloid, viz., raugalline (33). Ajmaline heated above its melting point is transformed into isoajmaline, a process which can be carried out preparatively by refluxing the parent alkaloid in ethanolic alkali (1). The presence of an indoline nucleus was established by color reactions, substitution reactions, e.g., bromoajmaline, mp 192")and UV-absorption spectroscopy. Upon catalytic hydrogenation, the benzene ring was reduced with remarkable ease to yield hexahydroajmaline, mp 149"-150", + 92" (CHC13); there was, however, no other unsaturation apart from this (1). The N-methyl was fixed on the indoline nitrogen since potassium permanganate oxidation of ajmaline in acetone gave N-methylisatinacetone (34), and soda lime or zinc dust distillation afforded N-methylharman (1). It was from the study of the environment of the second nitrogen of ajmaline that its unusual nature first became apparent. It was a carbinolamine which had the nitrogen a t a bridgehead; herice it could not be converted into an iminiuni cation so that it showed many of the normal reactions of a tertiary amine and an unhindered alcoholic group. Other reactions were better interpreted as taking place via the ring-opened tautomeric amino aldehyde which we shall call chanoajmaline. It should be mentioned here that ajmaline methiodide exists in the closed form a t least in the solid state, since its IR-spectrum in a nujol dispersion con-

790

W. I. TAYLOR

tains no carbonyl. I n solution, however, i t has reacted as if it were N-methylchanoajmaline hydriodide, since it had pKi 9.2 (in water), gave with base N-methylchanoajmaline, mp 124", and was reducible to N-methyldihydrochanoajmaline (B.HI, mp 224"-226") (1). Whenever the formation of the chano form was allowed, the appropriate consequences were observed ( l ) ,e.g., formation of chanoajmaline oxime, mp 220", and from it by dehydration the corresponding nitrile, mp 254"-255", ["ID + 29" (CHCl3); Wolff-Kishner reduction of ajmaline to deoxydihydrochanoajmaline, mp 185"-186", [aID+ 39" (CHCl3); reduction of ajmaline with potassium borohydride to furnish dihydrochanoajnialine, mp 200" (34); formation of 2 1-alkylaminoisoajmaline derivatives by heating ajmaline with primary amines (35); and epimerization of ajmaline in the presence of alkali to the more stable isomer isoajmaline (1). ti ._

qNCHzOH I

21 -L)eoxyajmaline

/\Et Di hydrochanoaj maline

T Ajmaline

2 steps

Chanoaj maline Rn-/

"Et 21 -L)eoxyisoajmaline

Isoajmaline

xylene

H //N CH=NR

I

A E t Decarbonochanoajmaline

A E t Chanoajmaline oxime

(R

=

OH)

21-Alkylamino isoajmaline ( R = alkyl)

CHART 11. Some reactions of ajmaline about N-4 (partial formulas).

I n an experiment which has already been of value in biosynthetic studies ( 7 ) , the C-21 carbon was eliminated as carbon monoxide upon reflux of ajmaline (or isoajmaline) with Raney nickel in xylene; the residue consisted of decarbonochanoajmaline, mp 203"-204" (34). I n another experiment which may also be of use today, chromic acid oxidation of deoxydihydrochanoajmaline gave methylethyl ketone, in complete agreement with the chemistry sketched in Chart I1 (1).

2 2.

THE AJMALIN E- SARPAGIN E ALKALOIDS

791

The environment of the 21-hydroxyl was recognized by the foregoing experiments, but the nature of the second hydroxyl was not well defined a t first. Although ajmaline was readily acetylated [diacetylajmaline, mp 130"-132" or 187"-189", ["ID + 16" (CHC13)Iand the alkaloid could be recovered on hydrolysis, the hydroxyl group in, for example, 21-deoxyoctahydrochunoajmaline apparently resisted oxidation and was therefore thought to be tertiary (1). The carbinolamine hydroxyl was certainly more reactive, since brief treatment with an acylating agent or one mole equivalent of same gave the 21-0-acylate [e.g., acetate, mp 190", [aID + 100" (MeOH)] in high yield (1, 35, 36). Ajmaline-17-0-acetate, mp 212"-214", could be prepared either by heating the diacetate monohydrochloride in water at 100" (35) or by treating the diacetate with cyanogen bromide (36). The 21-hydroxyl could be eliminated from the nucleus by a number of routes, all of which proceeded via ring closure of chuno derivatives, the most efficient route being reduction of ajmaline by borohydride to dihydrochanoajmaline whose hydrobromide at 300" ring-closed quantitatively to 2 1-deoxyajmaline hydrobromide (34),from which the parent base, mp 310", [a]D+124" (MeOH-CHC13), was obtained. The more recent work on ajmaline has made almost exclusive use of this 21-deoxy derivative, which we shall refer to also as deoxyajmaline. Parenthetically, it should be pointed out that aside from the configuration of the ethyl group, all the properties and reactions of isoajmaline run parallel to those of ajmaline, so that is0 derivatives need be made only to settle questions concerning the C-2 1stereochemistry. A renewed study of deoxyajmaline showed that the alcoholic function was secondary, since upon Oppenauer oxidation, it furnished a ketone which according to the IR-spectrum was in a five-membered ring. Furthermore, by the action of lead tetraacetate in acetic acid, deoxyajmaline was converted into an indole aldehyde (XII, deoxyajmalal-B) (vide infra), the reaction being correctly interpreted [cf. conversion of pseudostrychnine to strychnone (37)] as shown in Chart 111, X - t X I I . This result, coupled with biogenetic ideas and the dehydrogenation results to be described below, enabled the partial formula IV to be completed as in V. Lead tetraacetate oxidation of dihydrochunoajmaline gave a hydroxy

OT%" Mf?

IV

:lTJH Me

V Ajmaline

792

W. I. TAYLOR

aldehyde which cyclized readily to the hemiacetal, VI (38). It is worth pointing out that there is a remarkably close relationship between this " an compound and Alkaloid C, mp 168"-169", [ c ~ ] ~ + 2 0 0(EtOH), oxindole alkaloid (VII) isolated fmm Alstoniu muelleriuna Domin (A. villosa Benth.) (39).

VII Alkaloid C

VI

I n agreement with the formula V, palladium charcoal dehydrogenation (2) of deoxydihydrochanoajmaline afforded N-methylharman, ajarmine (VIII; R =Me), and ajmyrine ( I X ; R =Me), whose structures were proved by synthesis (40). Decarbonochanoajmaline under the same dehydrogenating conditions gave the corresponding nor compounds (VIII and I X ; R = H) (41).

W

MeN

VIII

7

IC'r

Me

IX

B. THE STEREOCHEMISTRY OF AJMALINE The stereochemistry of ajmaline (11)has been the subject of discussion (41, 42), although no later experimental evidence to support these

deductions has been formulated. It is, of course, clear that except for centers a t 17, 20, and 21, it is largely a problem of the absolute stereochemistry, since the hexacyclic system can only be put together in two ways, epimerization being possible a t C-2. The absolute stereochemistry has been derived (16) (Chart 111)as a result of a stereospecific chemical unfolding of 21-deoxyisoajmaline to l-methyltetradehydrodihydrocorynantheane (XVI ;epimeric 20-ethyl) whose absolute stereochemistry

22.

THE AJMALINE-SARPAGINE ALKALOIDS

793

had already been established via a correlation of dihydrocorynantheol with cinchonamine and quinine (43). The degradation carried out using 21-deoxyajmaline (X) in place of isoajmaline gave the expected end product, N,-met hyltetradehydrocor ynant heidane (XVI),which not only complemented the above degradation, but also put the absolute stereochemistry of corynantheidine itself on a firm basis for the first time. Isoajmaline was converted into deoxyisoajmaline, mp 295", ["ID + 159" (MeOH-CHCIs), which upon treatment with lead tetraacetate in benzene gave deoxyisoajmalal-A, mp 143"-144", [aID+ 126" (MeOH), readily epimerizable into deoxyisoajmalal-B, mp 179"-180", ["ID + 80" (MeOH). The driving force for this change probably lay in the fewer diaxial interactions in th3 quinuclidine moiety in the B compared with the A form. Reduction of the B aldehyde with sodium borohydride gave deoxyisoajmalol-B, mp 243"-245", ["ID + 86" (MeOH), tosylate, mp 167"-168", upon reflux in collidine gave d-trans-3-ethyl-l,2,3,4-tetrahydro-12methyl-2 -vinylindolo [2,3-a]quinolizinium tosylate characterized as its perchlorate, mp 167"-168", ["ID + 64" (MeOH). This was reduced catalytically to N,-methyltetradehydrodihydrocorynantheaneperchlorate, mp 201", ["ID + 16" (MeOH), identical with material prepared from corynantheine. Deoxyajmaline gave similarly deoxyajmalal-A, mp 180°-181", ["ID + 40'; deoxyajmalal-B, mp 213"-215', ["ID - 1" (MeOH) [identical with the indole aldehyde prepared by oxidation of X with Pb(0Ac)s in acetic acid (2)]; deoxyajmalol-B, mp 217"-218", ["]D-3', tosylate, mp 149"Z-cis-3-ethyl-1,2,3,4-tetrahydro-12-methyl-2-vinylindolo[2,3-a] 151"; quinolizinium perchlorate (xv),mp 201"-202", ["ID - 27'; and N,methyltetradehydrocorynantheidane perchlorate (XVI),mp 212"-2 13", [ajD- 26" (MeOH). Of the asymmetric centers not yet defined by the above degradations, the configuration of the C- 17-hydroxylwas assigned to that shown in I1on the following grounds. Deoxyajmalone (XVIII),in which the side before the viewer is unhindered, yielded not 2 1-deoxyajmaline, but an epimeric alcohol, mp 262"-264", ["ID + 367" (CHC13), upon catalytic reduction (2, 16). The degree of hindrance to rearward attack must be quite severe, since it was the only isolable product of sodium borohydride or lithium aluminum hydride reductions (16). Support for this conclusion was to be found in the NMR-spectra of a number of ajmaline derivatives which showed that coupling of the C-17 proton in the five-membered ring to the C-16 proton was minimal in the normal series (J = < 2 cps and at about J = 9 cps in the 17-epi series) (16). Theoretically, the orientation at C-2 should be capable of solution

794

8 / h

W.

I. TAYLOR

c)

1

Y

u

1( I I -

-

\ /

22. THE

795

AJMALINE-SARPAQINE ALKALOIDS

provided that asuitableindoleninecould be prepared. Thus, anindolenine, e.g., XX, would be expected to be reduced catalytically from the top

XVIII

XIX

XX

side, rather than from the highly hindered underside. It was discovered that oxidation of deoxyajmaline-0-acetate, mp 102"-108", [a]=+ 14" (MeOH-CHC13), with either chromic oxide in pyridine (16) or lead tetraacetate in benzene (35) gave not only 2-hydroxydeoxyajmaline-170-acetate (XIX),mp 95"-98" or 182"-184', [aID- 108" (MeOH), but also l-demethyl-dl-deoxyajmaline-O-acetate (XX),mp 184"-185", [aID- 12" (MeOH). The former compound could not be converted into an indolenium salt, since if forcing conditions were used, it simply decomposed into a mixture of indole aldehydes ( X I and XII). The indolenine XX, however, was suitable for the purpose a t hand, since catalytic hydrogenation followed by N-methylation gave a new compound, 2-epideoxyajmaline (XXI), mp 242"-243", [a]=- 10" (MeOH--CHC13), isomeric but not identical with deoxyajmaline. Its epimeric nature a t (2-2 was supported fully by the optical rotatory dispersion curve which was in a mirror image relationship to that of deoxyajmaline, reflecting accurately the stereochemical change adjacent to the chromophoric moiety. Finally, comparison of the dissociation constants

XXI

of the epimeric pair showed that X, where the proton in the conjugate acid was bonded to the N-1 atom, was a stronger base ( p K , 8.44) than XXI ( p l y f a7.80) (16) where this was not possible.

796

W. I. TAYLOR

The chemistry of ajmaline as discussed above has indicated that in many cases it must be in tautomeric equilibrium with the open chain amino aldehyde, chanoajmaline (Chart 11). I n the solid state, however, the IR-spectrum shows no carbonyl band and the C-21 hydroxyl has presumably taken up the more stable configuration trans to the neighboring ethyl group as shown in 11. I n the alternative struct,ure there would result an increase in steric interactions due to an additional 1,2diaxial interaction between adjacent groups. I n isoajmaline (11,epimeric 20-ethyl)for the same reasons, the C-21 hydroxyl would be expected to be trans to the neighboring ethyl. It is still not clear why ajmaline should isomerize so completely into isoajmaline, although examination of accurate models would suggest that the quinuclidine moiety is skewed with the result that the ethyl in ajmaline is the more hindered. Examination of the NMR-spectra of the ajmaline and isoajmaline derivatives supports the above deductions in the case of ajmaline, but in isoajmaline the results were ambiguous, since judging from the models the observed coupling of the C-20 and C-21 protons would be theoretically of the same order for either configuration of the hydroxyl (35).

C. DEMETHYLATION OF AJMALINE DERIVATIVES Many of the properties about N-1 of the ajmaline system would be expected to resemble those of analogous dialkyl anilines. This would be especially true of oxidative attack, since the quinuclidine moiety with a bridgehead nitrogen would be unaffected. As has already been described above, chromic acid oxidation of suitable ajmaline derivatives gave the 2-hydroxy as well as 1-demethyl-A'- derivatives, but the yields were low and did not lend themselves to synthetic work. When lead tetraacetate in a nonpolar solvent was used as the oxidant the reactions proceeded stepwise in high yield to give either the 2-hydroxy compound or the indolenine, depending on the amount of oxidant used. This has opened up a useful route to 10-deoxysarpagine-typecompounds from ajmaline. It has also made possible a facile proof of structure for the alkaloid vomilenine (XXIV) and its derived product perakine (XXV) (wide infra) via the lead tetraacetate oxidation of diacetyl ajmaline. The mechanism of action of lead tetraacetate is not known; it may, however, be the same as that which has enabled dialkylanilines to be converted in acetic anhydride to the alkylacetanilide and the appropriate aldehyde in quantitative yield (44).Whether the fission of deoxyajmaline to deoxyajmalal-A proceeds via attack a t N-1 and/or via collapse of the lead triacetate alcoholate cannot be decided from the available evidence (35).

22.

THE AJMALINE-SARPAGINE ALKALOIDS

797

Since hydrogenation of the indolenines yielded almost exclusively 2-epi derivatives, this was not a practical route to l-demethyl ajmaline derivatives. 1-Demethyldeoxyajmaline, mp 307", has been obtained in

azoAc of' OOH,

\

/

H Irdolenine

10-Deoxysarpagine derivative

very poor yield by the pyrolysis of deoxyajmaline hydriodide, but once again not in a preparative yield (16).

D. CONGENERS OF AJMALINE

Rauwolfia species have been the source of other bases closely related to ajmaline, and the structures of some of them can be regarded as trapped intermediates of the plausible, but as yet still hypothetical, biogenetic route t o the major alkaloid, ajmaline itself. R

XXIiI Sandwicine

Thus, ajmalidine (XXII)and sandwicine (XXIII, 17-epiajmaline) by sodium borohydride reduction followed by lead tetraacetate oxidation gave one and the same hemiacetal, V I (45).

798

W. I. TAYLOR

The structure of the indolenine, vomilenine (XXIV),became clear (46) when i t was found that its 19,20-dihydro derivative was identical in all respects with the compound prepared from ajmaline-17,21-O,O-diacetate by oxidation with two mole equivalents of lead tetraacetate in benzene followed by alumina chromatography (35).Although there has been only one recorded isolation of vomilenine from RauwolJia vomitoria (47), perakine has been obtained not only from reserpine manufacturing mother liquors of the same plant (48),but also from R . perakensis (49). The close relationship between vomilenine (a potential a$-unsaturated aldehyde) and perakine (Michael addition product of an amine to an a,p-unsaturated aldehyde) would suggest that one of these was being converted to the other during the isolation procedure. The conversion of vomilenine to perakine by brief treatment with boiling acetic acid would appear to make it quite possible that perakine is an artifact, especially as no other analogs of this system have been observed. The structure of perakine had been deduced originally from its physical properties, NMR-spectra comparisons with model indolenines, and biogenetic speculation (48).

%H

OAc XXIV Vomilenine

\

%cH3

CHO

OAc XXV fernkine

The structure of the hydroaromatic portion of vomalidine (XXII; R = OMe) rests entirely on the similarity of its properties, physically and chemically, with ajmalidine (XXII; R = H). With lithium aluminium hydride or sodium borohydride it gave a dihydroderivative (12-methoxysandwicine) which gave an 0,O-diacetyl derivative and a chano oxime which could be dehydrated to a nitrile. Vomalidine also underwent a Wolff-Kishner reduction analogous to that of ajmaline and gave the expected 12-methoxydihydrodeoxychnnoajmalinewhich showed two C methyls in the Kuhn-Roth determination (47).Vomalidine was originally assigned (47) an 11-methoxy group on the basis of its UV-absorption spectrum, but this opinion has since been revised (50, 51) and is in complete agreement with the interpretation of its NMR-spectrum (50). Several alkaloids are closely related to tetraphyllicine (XXVI, 21deoxy-419-ajmaline), whose structure was proved ( 5 2 ) essentially by

22.

THE AJMALINE-SALPAGINE ALKALOIDS

799

hydrogenation to 21-deoxyajmaline (X).Tetraphyllicine-O-trimethoxybenzoate (rauvomitine) has also been recognized as a natural product (52).The structure of mauiensine or 17-epitetraphyllicine was determined not only by its synthesis from tetraphyllicine via catalytic reduction of tetraphyllicinone (XXVII ; R = H),but also by reducing its double bond and isolating 17-epideoxyisoajmaline (53). The latter was prepared for comparison purposes in four steps from isoajmaline (removal of its 2 1-hydroxyl and epimerizing the 17-hydroxyl). R

XXVI Tetraphyllicine Rauvomitine ; OH= OCOPh(0Me)s

XXVII Purpeline; R =OMe Mitoridine; R=OH

The remaining members of this group are purpeline (XXVII ;R = OMe, 12-methoxytetraphyllicinone)and mitoridine ( 12-hydroxytetraphyllicinine). The structure of purpeline, apart from the usual analytical data, depended almost entirely upon comparison of the NMR-spectrum of purpeline with suitable models which were for the aromatic ring, vomalidine and deacetylaspidospermine and for the hydroaromatic portion, tetraphyllicinone (XXVII ; R = H). Borohydride reddction of purpeline gave an alcohol [hydrochloride, mp 235"-238", [a]=+ 109" (HzO)] which with lead tetraacetate gave an indole aldehyde, mp 175"-176", analogous to deoxyajmalal-B (XII). The NMR-spectrum of this aldehyde had a pattern for the chemical shifts of its aromatic protons identical with that (50). observed for 8-methoxy-9-methyl-l,2,3,4-tetrahydrocarbazole The structure of the phenolic alkaloid, mitoridine (XXVII; R = OH), was secured by the demethylation of purpeline (50). The most puzzling of the Rauwolja alkaloids is rauwolfinine, mp - 35"(EtOH), which was first reported in 1954 (54) from a 235"-236", variety (Cochin) of R.serpentina and has been the subject of three additional papers (55) culminating in an analysis of its NMR-spectrum (56). Rauwolfinine was said t o be an indoline alkaloid with one G-methyl, and upon zinc dust or selenium dehydrogenation gave N-methylharman. Rauwolfinine with 4 N sulfuric acid (55)was supposed to have afforded an indole [specifically, an indole aldehyde characterized as its 2,4-dinitro-

800

W.

I. TAYLOK,

phenylhydrazone, mp > 200” (55. 56) stated in 1961 “ t o be virtually identical to that prepared from deoxyajmaline by the action of 5 N sulphuric acid ”!I. The same school has also made the claim (58)reiterated in 1962 (57)that ajmaline heated with 2 N sulfuric acid gives rise to an indole and since neither this observation nor the one involving deoxyajmaline could be confirmed (35), the initial observation with rauwolfinine may also be incorrect. It is also worth pointing out that despite the vast amount of work carried out by others on the alkaloid content of R. serpentina, rauwolfinine has not been reisolated, except by the same authors from R. perakensis (59),but even this was not confirmed by later work (49). If the recorded optical rotation of rauwolfinine is an error, it may very well turn out to be an impure mixture of alkaloids, consisting mainly of isoajmaline. It should also be emphasized that so far all naturally occurring 2 1-deoxy compounds have a 20-ethylidine group which makes the 2 1-deoxy-20-ethyl moiety in the latest rauwolfinine formula ( 5 6 ) exceptional.

E. THE 16-CARBOMETHOXYTETRAPHYLLICINE

SUBGROUP

Apart from RauwolJia species, other plants of the family Apocynaceae, principally Vinca species, have given rise to derivatives of the ajmaline its 17-0-acetate, and type, viz., 16-carbomethoxy-2-epitetraphyllicine, 1-demethyl analog. Their structures have been determined by their conversion into 10-deoxysarpagine derivatives, although in a recent paper mass spectrometry has been effectively used (13) now that significant model compounds are available. The important, interrelationships are given in Chart IV, the compounds of established structure being macusine A and akuammidine methiodide (structures by X-ray crystallographic analysis), and the deoxysarpagines (vide infra, structures and absolute stereochemistry implied by degradations in Chart 111). The oxidative ring opening of vincamedine which Ieads to polyneui-idine has an exact model in the same oxidation of 21-deoxyajmaline17-0-acetate (Section I, C). Also, reconstitution (60) of vincamajine by the catalytic hydrogenation of its corresponding indole aldehyde (Chart IV) follows from the fact that 21-deoxyajmalai-A (XI) under similar conditions furnished 2-epi-21-deoxyajmaline (XXVIII), i.e., the 17hydroxyl takes up the thermodynamically more stable of the two possibilities. Acetylation of deoxyajmal-A in acetic anhydride containing hydrogen chloride also gave the expected 2-hydroxy-2 1-deoxyajmaline-17-@acetate (XXIX) (61). The ease with which advantage can be taken of the nucleophilic reaciivity of the indolic ,B position toward a

22.

THE AJMALINE-SARPAGINE ALKALOIDS

801

suitably placed group was further illustrated by the behavior of 17-0tosyl-21-deoxyajmalol-A (XXX)toward heat or pyridine which resulted in 2-hydroxy-17,21-dideoxyajmaline (XXXI) (61).

A second proof for the configuration of the 17-hydroxyl in the above alkaloids came out of quebrachidine (1demethyivincamajine) chemistry. Its lithium aluminium hydride reduction product gave in a facile manner

XXXI

XXXIl

an isopropylidene derivative (XXXII) which would not have been expected from the isomeric 17-epihydroxy compound (13).

802

R ’

0

@ \ I

W. I. TAYLOR

/

X

0,

e

u,

X

2

ld

t-

8

22.

d

THE AJMALINE-SARPAGINE ALKALOIDS

803

804

W. I. TAYLOR

11. The Sarpagine Group

A. THE STRUCTURE OF SARPAGINE There is a small group of 5-hydroxyindoles, namely sarpagine (XXXIIIa), lochnerine (10-0-methylsarpagine),and its quaternary salt (lochneram), which were regarded as derivatives of a precursor of ajmaline. Their structures, however, were guessed solely from their interrelationships, determination of chromophoric moieties, functional group analysis, and obtention of dehydrogenation products with the expected UV-absorbing properties ( 14. 62). The correctness of these guesses and the establishment of the relative and absolute stereochemistry of sarpagine was proved by its conversion into l-methyl-10deoxydihydrosarpagine identical in all respects with one of the four possible 81-deoxyajmalols, viz., 21-deoxyajmalol-B (XIII) (16). So far the occurrence of the possible 16-epi compounds has not been recorded.

H XXXIII-R1 = Rz R3 H ; R3 = AC XXXIV-RI = H ; R z XXXV-R10 == H; Rz = 13; R3 = BC XXXVI-R10 = H ; Rz = Me; R3 = Ac: XXXVII-R10 = H ; Rz = Me; R3 = H ~

~

Tosylation of monoacetyl sarpagine (XXXIVa), mp 302"-305", D I . [ 4" (Py), followed by reductive cleavage (Raney nickel) gave 17-0-acetyl-10-deoxysarpagine(XXXVa), mp 213"-215", whose N,sodio derivative was methylated with methyl iodide in liquid ammonia to the 1-methyl derivative (XXXVI), mp 172"-176", hydrolysis of which gave 1-methyl-10-deoxysarpagine(XXXVIIa), mp 190"-192". The latter two compounds, although resisting hydrogenation, proved to be useful for the correlation of voachalotine and vincsmajine with ajmaline (Chart IV). Repetition of the above sequence, starting with dihydrosarpagine (XXXIIIb), mp 350", [.IU + 31" (MeOH), gave successively 17-0-acetyldihydrosarpagine (XXXIVa), nip 177"-278", [a]1) + 29"

+

22.

THE AJMALINE-SARPACINE ALKALOIDS

805

(MeOH); 17-0-acetyl-10-deoxydihydrosarpagine(XXXVa), mp 253"254", [aID 1" (MeOH); and 21-deoxyajmalol-B (XXXVII), mp 213"214", [.ID 2" (MeOH).

+ +

The stereochemistry of the ethylidene group could not be decided from these experiments but it came from the results of the X-ray work carried out on the correlated compounds (Chart IV), macusine A (28) and akuammidine ( 2 7 ) . Deoxysarpagine (tombozine)itself has been obtained from Diplorrhynchus condylocarpon and from Geissospermum vellosii along with its corresponding aldehyde (veliosimine). The latter compound was also obtained as an acidic fission product of the indole-indoline dimer geissolosimine, which could be reconstituted if the products were allowed to stand a t room temperature for several days in 1.5 N acetic acid (20). Useful NMR-data is also to be found in this paper. The ethylidene moiety in these alkaloids does not hydrogenate readily in alcohols if platinum is used, but goes well in the presence of acetic acid, hydrogen usually coming in such a way as to end up with the ethyl in the isoajmaline coiifiguration (20, 23, 63). However, with palladium as a catalyst in alcohol, sarpagine has apparently given only the alternative dihydro product (16, 62).

B. THE 16-CARBOIKETHOXY-~O-DEOXYSARPAaINE SUBGROUP The interrelationships of these bases are indicated in Chart IV, and except for akuammidine, they are 16-carbomethoxy-16-epi-10-deoxysarpagine derivatives. Upon attempted Oppenauer oxidation, they undergo a lacile extrusion of the hydroxymethyl residue (retro-aldolization), but the desired 16-aldehyde can be obtained either directly by chromic acid oxidation (21) or indirectly by oxidation [chromic acid (11) or better, lead tetraacetate (13)]of the corresponding 16-carbomethoxytetraphyllicine. Pyrolysis of the alkaloids would also appear to be a direct method of eliminating the hydroxymethyl residue, since upon attempted vapor phase chromatography voachalotine broke down in the inlet system to the dehydroxymethyl compound (30), as also the quebrachidine derivative (XXXVIII) collapsed into vellosimine as it was sublimed into the mass spectrometer (13). The hydroxymethyl in akuammidine is apparently (21) also the source of the methyl iodide obtained in the Herzig-Meyer estimation of N-methyl (26) and this has been reconfirmed with polyneuridine (21). Some of the transformations indicated in Chart IV can be regarded as mimicking a plausible biosynthetic route (Chart I)and the only important

806

W. I. TAYLOR

laboratory step still to be carried out is the facile generation from XXXIX of the carbinolamine, XL, which may then be converted into the sarpagine system (111).

% OHC

CHzOH \

XXXIX

XXXVIII

MeOOC*CHO XI,

Upon hydrogenation (30) of voachalotine, the ethyl formed is in the same configuration as in ajmaline, which stands in contrast to the course of the,hydrogenation evinced by deoxysarpagine itself. The nature of the N-methyl in voachalotine has been thoroughly studied by physical methods and compared with many model compounds (64). Some interesting results have come out of a study of the reactions about Nbof voachalotine. Theaction ofcyanogen bromideon dehydroxymethylvoachalotine (XLI) (30) gave a bromine-free apo product which had the UV-spectrum of a vinylindole. This reaction, reminiscent ( 6 5 ) of the behavior of ibogaine toward the same reagent, could have had structure X L I I or XLIII (R = H), the latter being preferred. I n the case of voachalotine itself, the product XLIV had an indolic UV-spectrum, and this could be interpreted as being formed via an acid-catalyzed ether formation between the primary product, the 3-vinylindole XLIII, and the 17-hydroxyl. The same workers also showed that chromic acid oxidation of voachalotine gave not the expected aldehyde (cf. polyneuridine) but an ether formulated as XLV. This may be mechanistically related to the course of a lead tetraacetate oxidation carried out on deoxyajmalol-A (XLVI)which had as its aim the preparation of XLVII which might have been rauwolfinine (35). Substance XLVII rearranged upon heating in

22.

T H E AJMALINE-SARPAGINE ALKALOIDS

807

dilute sulfuric acid to the indole XLVIII which was analogous in constitution to XLV. The mechanism of the lead tetraacetate oxidation and the exact route by which the acid-catalyzed rearrangement takes place has not been worked out. For example, does the first reaction proceed via

% R

R XLI

COOMe

COOMe

XLII

IBrCN

%% XLIV

X1,III

0

\ (‘OOMe

%w XLVIII

XI>\‘

-

CH2 HOHzC XLVI

XLVII

808

W. I. TAYLOR

thelead triacetate alcoholate or the lead complex with C-7 (66),and in the second step, is it fission of the ether followed by readdition, or does it proceed via a 1,2 shift to the 2-ether followed by an allylic rearrangement to the 7 pmition? 111. Mass Spectra of the Ajmaline-Sarpagine Alkaloids

One of the earliest applications of mass spectroscopy to indole alkaloid chemistry was the definition of the sarpagine skeleton by comparing a pertinent derivative (XLIX; R1 = OMe, Rz = H, Iz3 = Me) with an ajmaline degradation product of known structure (XLIX; R1 = H, Rs = Me, R3 = Me) (67). The two compounds behaved in essentially the same manner upon fragmentation under electron impact. I n this (68)and a related study on iboga alkaloids (69) the general methods for interpreting the course of the fragmentations were well illustrated. These involved the running and comparing of relevant derivatives and deuterium-labeled compounds and the detection of metastable peaks. The latter can only be used to elucidate those fissions which take place after the ions have left the slit (70). Fragmentation has to occur in an energetically favourable way to yield stable positive ions or radicals, and it is not always possible to decide between alternative paths from the results of a single measurement. The ajmaline-sarpagine alkaloids pose some interesting problems in this regard. All the sarpagine derivatives showed a strong M- 1 peak which could not for steric reasons have originated from a carbon adjacent to the basic nitrogen and was therefore by a process of elimination thought to be a C-6 proton (ion stabilized by the aromatic system). I n the dihydrosarpagine series, besides the M and M-1 ion, there are also observed peaks for M-CH3 [ G I 8 or C-17 (in X L I X ; R = CHs)] and M-CHzOH [when present (21)], none of which gave rise to further prominent peaks. The principal route to the stable P-carboline molecular ions is shown in Chart V. I n the sarpagines (ethylidine instead of ethyl) there are M, M-1, M-CH3, M-HzO, M-CHzOH, and where applicable also M-COOCH3 and M-COOCH3-HzO peaks, and as expected, the fragmentation path to the /3-carbolineswas different because no peak corresponding to 221 + R1+ Rz in Chart V was observed (13, 21, 79). The route to the ,8-carbolines appeared to start by expulsion of the bridging group C-16, leading to the products in Chart VI. This could take place as depicted in L and/or L I t o generate the 248 + R ion. The former route may be more favoured in the case of polyneuridine and voachalotine than in akuammidine.

22

22.

pz"

I

THE AJMALINE-SARPAGINE ALKALOIDS

+

d

aI

809

d

810

z-

W. I. TAYLOR

\

J E 2

5

81 1

22. THE AJMALINE-SARPAOINE ALKALOIDS

The situation with the ajmaline skeleton has not been spelled out in detail (13), but with the availability of models, it is now certain that by mass spectroscopy the difference between C-2 epimers can be detected. The method, however, is not sensitive to the c - 1 7 stereochemistry. The spectra were characteristic and involved deep-seated fissions, since the

p

HOHzC

L

COOMe

LI

compact highly branched system in itself cannot support stablized charges without decomposition. Thus, for example, quebrachidine split t o yield strong peaks at M-130, 130 (the dihydroindole + 1 carbon atom), and M-130-CH30H. Vincamajine-17-0-acetate split similarly to furnish the observed strong peaks a t M-144,144 (the N-methyldihydroindole + 1 carbon atom), and M-144-CH30H (13).

IV. Pharmacological Notes The individual alkaloids derived from RauwolJia and other apocynaceous plants have been subjected to a thorough scrutiny by many pharmacologists in the hope of discovering another drug as useful as reserpine (71). I n spite of the intensive interest shown by many pharmaceutical houses during the last 13 years no such drug has been recognized from natural sources. Ajmaline has long been known to possess antifibrillatory properties (72), but it does not seem t o be superior to quinidine which is used clinically for this purpose, although it is far from an ideal drug. Ajmaline produces no sedation or tranquilization and its blood pressure properties are unremarkable in the various specialized situations in which it has been placed (73). Similar remarks hold for isoajmaline (74). Sarpagine (75) and deoxysarpagine (76) have only fleeting effects on the blood pressure and the pressor amines. Voachalotine, like ajmaline, has some cardiotonic properties (77). Akuammidine is inactive in bird malaria (78). The alkaloids derived from Vinca species are probably

812

W. I. TAYLOR

uninteresting, since work on the crude alkaloidal extract has been unrewarding (see Chapter 12). REFERENCES 1. F. A. L. Anet, D. Chakravarti, R. Robinson, and E. Schlittler, J . Chem. Soc. p. 1242 (1954). 2. R . B. Woodward and K. Schenker, Angew. Chem. 68, 13 (1956). 3. R . B. Turner and R. B. Woodward i n “The Alkaloids” (R. H. F. Manske and H. L. Holmes, eds.), Vol. 111, p. 55. Academic Press, S e w York, 1983. 4. E. Wenkert, Experientia 15, 165 (1959). 5. Sir Robert Robinson, “Structural Relations in Natural Products,” Clarendon Press, Oxford, 1955. 6. E. Schlittler and W. I. Taylor, Experientia 16, 244 (1960). 7. E. Leete, Chem. and I d . p. 692 (1960); J . Am. Chem. SOC.82, 6338 (1960); P . N. Edwards and E. Leete, Chem. m z d Ind. p. 1666 (1961) ; E. Leete, G. Ghosal, and P. N. Edwards,J. A m . Chem. SOC.84, 1068 (1962). 7a. A. K. Bose, B. G. Chatterjee, and R. S. Iyer, Znd.J. Phnrm. 8, 185 (1956) ; E . Wenkert, E. W. Robb, and N. V. Bringi, J . Am. Chem. 8oc. 79, 6571 (1957). 8. M. Gabbai, These Doct. Univ. Pharm. Paris, s h i e U, No. 291, 1958. 9. M.-M. Janot and J . Le Men, Compt. Rend. 241, 767 (1955). 10. S. Goodwin and E. C. Homing, Chem. and I n d . p. 846 (1956). 1 1 . M.-M. Janot, J. Le Men, J . Gosset, and J. LBvy, BUZZ.SOC. Chim. France, p. 1079 (1962) ; J . Levy, J. Le Men, and M.-M. Janot, Compt. R e n d . 253, 131 (1961). 12. M.-M. Janot, J. Le Men, and Y. Hammouda, Compt. Rend. 243, 85 (1956). 13. M. Gorman, A. L. Burlingarne, and K. Biemann, Tetrahedron Letters p. 39 (1963). 14. W. A. Arnold, W. von Philipsborn, H. Schmid, and P. Karrer, Helv. Chim. Actu 40, 705 ( 1957). 16. W. B. Mors, P. Zaltzman, J. J. Beerebom, S. C. Pakrashi, and C. Djerassi, Chem. and ’ Ind. p. 173 (1956). 16. M. F. Bartlett, R. Sklar, W. I. Taylor, E. Schlittler, R. L. S. Amai, P. Beak, S. V. Bringi, and E. Wenkert, J . Am. Chem. SOC.84, 622 (1962). 17. W. F. Arnold, F. Berlage, K. Rernauer, H. Schmid, and P. Karrer, Helv. Chim. Actn 41, 1505 (1958). 18. D. Stauffacher, HeEu. Chim. Acta44, 2006 (1961). 19. R. Goutarel, X . Monseur, and J. Le Men, Compt. R e n d . 253, 485 (1961). 20. H. Rapoport arid R. E . Moore, J . Org. Chem. 27, 2981 (1962). 21. L. D. Antonaccio, N. A. Pereira, B. Gilbert, H. Vorbrueggen, J . M. Wilson, H . Budzikiewicz, L. J. Durham, and C. Djerassi, J . Am. Chem. SOC. 84, 2161 (1962). p. 1848 22. A. R . Battersby, R. Binks, H. F. Hodson, and D. A. Yeowell, J . Chem. SOC. (1960). 23. A. R. Battersby and D. A. Yeowell, Proc. Chem. SOC.p. 17 (1961). 24. T. A. Henry, J . Chem. Soc. p. 2759 (1932). 25. J . Gosset, J . Le Men, and BI.-M. Janot, A n n . Pharm. Frang. 20, 448 (1962). 26. A. Chatterjee, C. R. Ghosal, S. Adityachaudhury, and S. Ghosal, Chem. and Znd. p. 1034 (1961). 25. S. Silvers and A. Tulinsky, Tetrahedron Letters, p. 339 (1962). 28. A. T. McPhail, J. M. Robertson, G. A. Sim, A. R. Battersby, H. F. Hodson, and D. A. Yeowell, Proc. Chem. SOC. p. 223 (1961).

22.

THE AJMALINE-SARPAGINE ALKALOIDS

81 3

29. J. Pecher, N. Defay, M. Gauthier, J. Peeters, R. H. Martin, and A. Vandermeers, Chem. and I n d . p. 1481 (1960). 30. N. Defay, M. Kaisin, J . Pecher, and R. H. Martin, Bull. Soc. Chim. Belg. 70, 475 (1961); J. Pecher, Symp. Organic Chem. Natural Products, Brussels, June, 1962. 31. D. Mukherji, R. Robinson, and E. Schlittler, Ezperientia 5, 215 (1949). 32. S. Siddiqui, J. I d . Chem. Soc. 16, 421 (1939); R. Paris, Ann. Pharm. F r u q . 1, 138 (1943). 33. J.-P. Le Gall, Ann. Pharm. F r a q . 18, 817 (1960); W. I. Taylor, P. Potier, and J. Le Men, ibid. 21, 321 (1963). 34. Sir Robert Robinson, Chem. and Ind. p. 285 (1955). 35. W. I. Taylor and M. F. Rartlett, unpublished observations. 36. S. Siddiqui, S. A. Warsi, M. Alauddin, and V. Ahmad, Pakistan J . Sci. I n d . Res. 2, 86 (1959). 37. H. Leuchs, E. Tuschen, and M. Mengelberg, Ber. 77, 403 (1944); R. B. Woodward, W. J. Brehm, and A. L. Nelson, J. Am. Chem. Soc. 69, 2250 (1947). 38. K. Schenker and R. B. Woodward, cited by J. E. Saxton, Quart Rev. 10, 108 ( 1956). 39. C. E. Nordman and K. Nakatsu, J . Am. Chem.Soc. 85,353 (1963); R. E. Gilman, Ph.D. Thesis, Univ. of Michigan, 1959. 40. R. B. Woodward and Yang, cited by J. E. Saxton, Quart. Rev. 10, 108 (1956). 41. Sir Robert Robinson, Angew. Chem. 69, 40 (1957). 42. S. K. Talapatra and A. Chatterjee, Naturwissenschaften 45, 58 (1958); W. Venkateswaran and A. Chatterjee, J . I n d i a n Chem. Soc. 35, 363 (1958). 43. E. Wenkert and K. V. Bringi, J.Am. Chem. Soc. 81, 1474,6535 (1959); E. Ochiai and M. Ishikawa, Pharm. Bull. ( J a p a n ) 6, 208 (1958); V. Prelog and E. ZalBn, Helv. Chim. Acta 27, 545 (1944). 44. L. Horner, E. Winkelmann, K. H. Knapp, and W. Ludwig, Ber. 92, 288 (1959). 45. M. Gorman, N. Seuss, C. Djerassi, J. P. Kutney, and P. J . Scheuer, Tetrahedron I, 328 (1957). 46. W. I. Taylor, A. J. Frey, and A. Hofmann, Helv. Chim. Acta 45, 611 (1962). 47. A. Hofmann and A. J . Frey, Helv. Chim. Acta 40, 1866 (1957). 48. P. R. Ulshafer, M. F. Bartlett, L. Dorfman, M. A. Gillen, E. Schlittler, and E. Wenkert, Tetrahedron Letters, p. 363 (1961). 49. A. K. Kiang and A. S. C. Wan, J . Chem. Soc. p. 1394 (1960); Proc. U N E S C O S y m p . Phytochem. S.E.A. Science Cooperation Offcce Djakarta p. 181 (1957). 50. W. I. Taylor and P. R. Ulshafer, unpublished observations. 51. A. Hofmann and A. J . Frey, unpublished observations. 52. C. Djerassi, M. Gorman, S. C. Pakrashi, and R. B. Woodward, J . Am. Chem. Soc. 78. 1259 (1956); C. Djerassi, J . Fishman, M. Gorman, J . P . Kutney, and S. C. Pakrashi, J . Am. Chem. Soc. 79, 1217 (1957); J . Poisson, R. Goutarel, and M.-M. Janot, Compt. Rend. 241, 1840 (1955). 53. P. J. Scheuer, M. Y. Chang, and H. Fukami, J . Org. Chem. 28, 2641 (1963). 54. S. Bose, J . I n d i a n Chem. Soc. 31, 47 (1954). 55. S. Bose, J . Iruiian Chem. Soc. 31, 311 (1954); ibid. 31, 691 (1954); ibid. 35, 72 (1958). 56. A. Chatterjee and S. Bose, J . I n d i a n Chem. Soc. 38, 403 (1961). 57. A. Chatterjee and A. B. Ray, Sci. and Cult. 21A, 515 (1962). 58. A. Chatterjee and S. Bose, Sci. and Cult. 20, 606 (1955). 59. A. Chatterjee and S. K. Talapatra, Xaturwissenschaften 42, 182 (1955); A. Chatterjee, S. C. Pakrashi, and G. Werner, Fortschr. Org. Naturstoffe 13, 346 (1956). 60. J. Le Men, unpublished observations.

814

W. I. TAYLOR

61. M. F. Bartlett, B. F. Lambert, H. M. Werblood, and W. I. Taylor, J. Am. Chem. Soc. 85, 475 (1963). 62. J. Poisson, J. Le Men, and M.-M. Janot, Bull. SOC.Chim. France p. 610 (1957); D. Stauffacher, A. Hofmann, and E. Seebeck, Helv. Chim. Acta 40, 508 (1957). 63. J. Gosset, J . Le Men, and M.-M. Janot, Bull. SOC.Chim. France 1033 (1961). 64. J. Pecher, R. H. Martin, IT.Defay, M. Kaisin, J. Peeters, G. V. Rinst, N. Verzele, and F. Alderweireldt, Tetrahedron Letters p. 270 (1961). 65. M. F. Bartlett, D. F. Dickel, and W. I. Taylor, J . Am. Chem. SOC. 80, 126 (1958). 66. W. I. Taylor, Proc. Chem. SOC. p. 247 (1962). 67. K. Biemann, Tetrahedron Letters No. 15, 9 (1960). 68. K. Biemann, J. Am. Chem. Soc. 83, 4801 (1961). 69. K. Biemann and M. Friedmann-Spiteller, J . Am. Chem. Soc. 83, 4805 (1962). 70. K. Biemann, “Mass Spectrometry, McGraw-Hill, New York, 1962. 71. R. E. Woodson, Jr., H. W. Youngken, E. Schlittler, and J . A. Schneider, “Rauwolfia, Botany, Pharmacognosy, Chemistry and Pharmacology,” Little, Brown, Boston, 1957. 72. K. Van Dongen, Arch. Intern. Pharmacodyn. 53, 80 (1936). 73. R. N. Chopra, N. N. Das, and S. N. Mukherjee, I d . J . Med. Res. 24, 1125 (1937); R. N. Chopra, B. C. Bose, J. C. Gupta, and I.C. Chopra, I d . J . Med. Rcs. 30,219 (1942) ; 31, 71 (1943); H. J. Bein and F. Gross, Verhandl. Deutsch. Ges. Kreislauf-Forsch. 19, 277 (1953). 74. B. B. Bhatia and R. D. Kapur, Ind. J . Med. Res. 32, 177 (1944). 75. J. D. Achelis and D. Kroneberg, Arzneimittel-Forsch. 5, 204 (1955). 76. A Quevauviller and Y. Takenaka, J . Physiol. (Paris) 54, 403 (1962). 77. J. La Barre, Actualities Phurmacol. 14, 109 (1962). 78. J . A. Goodson, T. A. Henry, and J. W. S. MacFie, Biochem. J . 24, 874 (1930). 79. E. Clayton, R. I. Reed, and J. M. Wilson, Tetrahedron 18, 1449 (1962).

AUTHOR INDEX Numbers in parentheses are reference numbers and are included to assist in locating references where the author's name is not mentioned in t h e text. Numbers in italics refer t o the page of the chapter on which the reference is listed.

A 200), 324(185, 195, 200, 207, 208), 325 (195), 332, 333 Abdusalamov, B. A., 48(21a), 52 Abe, M., 757(95), 759(100), 760(100, 102, Allauddin, H., 294(120), 330 104, 105), 762(105, 106), 763(109), Allport, N. L., 735(30), 780 764(100, 104, 109, 110, 111, 113, 114, Almirante, L., 325(212), 333 116), 765(104, 109, 111, 116, 118), Amai, R. L. S., 485(177), 511, 534(71), 781, 7x2 535(71), 537(71), 577, 788 (16), Abou-Chaar, C. I., 735(23), 780 792(16), 793(16), 795(16), 797(16), Abramovitch, R. A,, 14, 25, 39(45), 40(46), 804(16), 805(16), 812 45 Amarisingham, R. D., 444( 108), 509 Amiel, Y., 361(24), 362(24), 443(24), Achelis, J. D., 811(75), 814 Achmatowicz, A., 611 (84b), 668 506 Achmatowicz, O., 598(33), 606(33), An Cu, N., 720(76), 721 (76), 723 607(33), 608(78), 610(33, 78a, 79, 80, Anderson, G. J., 305(163), 332 82, 83, 84), 611(84, 84a, b), 612(33), Anderson, G. W., 769(197), 783 613(93), 619(109), 665(33), 666, 667, Anderson, R., 622(122), 669 668 Anderson, R. C., 95 (53), 116 Achmatowicz, S., 610(78a), 611 (84a), Andres, W. W., 303(158), 331 613(93), 667, 668 Anet,E. F. L. J., 588(31), 589 Adams, R., 97(70), 117 Anet, F. A. L., 293(82), 329, 542(89), 578, 638(165), 663(213,215), 664(213,215), Aditya Chaudhury, N., 120(20), 156, 161(35), 194(35), 195(35), 199, 670,671,785(1,789(1),790(1), 791(1), 491(62), 505(62, 197, 199, 200), 508, 812 513 Angermann, L., 50(34), 53 Aghoramurthy, K., 120(18), 123(31), Anthony, W. C., 9(68), 13, 18(68), 23 124(31), 135(31), 146(31), 156, Antonaccio, L. D., 82(81), 83(83), 91, 191 (go), 201, 270(54), 281 (54), 284, 134(44), 157, 262(26), 266(44), 267, 466 (70), 508 268, 337(10), 390(10, 52), 401(10), Aguayo Brissolese, J., 398(40), 403(40), 404(10),410(49),411(10),414(10,52), 484(171), 507, 511, 712(57), 723 423(80), 426(80), 429(49), 432(79), 433(80), 434(49), 445(52), 447(52), Agurell, S., 768(136), 782 Ahmad, V., 791 (36), 813 448(52), 467(125a), 468(125a),469(49, Akabori, S., 49(28), 50(28), 52 125a), 482(163, 165), 484(163, Alauddin, M., 791 (36), 813 165), 485(165), 490(165), 491(165), Albertson, N. F., 5 (51), 23 495(180), 506, 507, 508, 510, 511, 512, Alderweireldt, F., 806(64), 814 535(77), 578, 694(1), 696(1), ,707(1), 721, 788(21), 805(21), 808(21), 812 Aldrich, P. E., 300(134), 331, 632(147), 669, 702(26), 704(26, 27), 716(69), Antonor, W. K., 753(94), 781 722, 723 Applegate, H. E., 322(196), 324(196), 333 Allais, A., 316(184, 185), 318(185), Arai, G., 211(31), 233 321 (185), 322(185, 195), 323(185, 195, Arcamone, F., 747(69), 781 815

816

AUTHOR INDEX

Archer, A. A. P. G., 337(10), 367(29), 389(29), 390(10), 401(10), 404(10), 410(49), 411(10,29), 414(10), 415(29), 429(49), 434(49), 469(49), 506, 507 Archer, S., 5(51), 2 3 Armen, A,, 5(49), 23 Arnaud, M., 238(4), 246 Arnold, W., 126(33),1 5 6 , 4 1 8 ( 6 7 ) ,466(67), 508, 520(27, 28), 534(27, 28), 545(93), 547(93), 552(109), 569(131, 132), 570(131, 132), 571(132), 573(132), 576, 578, 579, 628(135), 669, 682(17), 690 Arnold, W. A,, 804(14), 812 Arthur, H. R., 327 Arundale, E., 241 ( 2 5 ) , 247 Arya, V. P., 505(197), 505 Asahina,Y., 51(39), 53,56(2, 3, 7, 8, 9, I l ) , 57(2, 3, 13, 14), 58 Asero, B., 13(103), 14(103), 24 Ash, A. S . F., 13(105), 25 Asher, J. D. M., 262(28), 264(28), 268, 482(164), 511, 716(66), 723 Asma, W. J., 326(227, 228), 334 Asmis, H., 520(23, 24, 30), 521(23, 24, 30, 39), 523(23, 30), 539(39), 540(24, 30, 39), 541(24), 553(111), 555(24, 30), 564(111), 574(23, 39), 575(30), 576, 577,579 Atal, G. K., 328, 708(38), 722 Atherton, F. R., 743(51), 745(51), 780 Augustine, R. L., 246(33), 247 Axelrod, J., 772(166), 783 Ayer, W. A., 451(113f), 509

B Bacon, R. F., 161, 162(29), 199 Bacovescu, A., 601 (53), 667 Bader, F., 160(12), 199, 328, 329, 331 Bader, F. E., 166(55), 200, 293(76), 300(139), 316(183), 329, 331, 332, 529(52), 577 Badger, G. M., 48(20), 52, 60(14), 61(14), 64(14), 75(14), 85(14, 84), 86(84), 87(84), 89, 91 Badhwar, R. L., 272(7), 283 Biichli,E., 48(21), 52, 520(23, 24), 521(23, 24, 36, 40), 523(23), 528(40), 531(40), 532(40), 533(40), 534(40), 540(24), 541(24), 555(24), 561(122), 574(23,

36, 40), 475(36, 40), 576, 577, 579 Baechtold, H. P., 327(230), 334 Bahr, K., 466(133), 510, 517(12, 13), 520(12, 13), 521(12, 13), 539(13), 548(12), 560(13), 568(12), 575(13), 576 Bahr, K., 540(83), 578 BBrwald, L., 51 (43), 5 3 Bailey, A. S., 629 (142), 638 (165), 647 (189), 660(189), 669, 670 Baillon, H., 232 (70), 235 Ban, Y., 78(80), 79(80), 91, 173(68a), 200, 302(149), 331, 499(186i), 512, 533(63), 577, 689(27), 691, 695(3), 696(3, 5), 697(5), 698(3), 701(23), 721 (77), 721, 722, 723 Bardsley, W. G., 33(30), 45, 585(23), 589 Barger, G., 3(17), 4(17), 22,29(13), 30(15), 32(13), 41, 44, 60(19), 61(19), 64(19), 75 (19), 85 (19), 89, 582 (3), 583 (12), 588, 589, 727(6, 7), 749, 751, 759(7), 779, 781 Barltrop, J. A,, 744(61), 780 Barnes, A. J., Jr., 204(12), 218(12), 233, 270(22, 23), 271(22, 23, 30), 283, 508, 510 Barrett, W. E., 311(174, 175), 312(179), 314(182), 332 Barriga Villalba, A. M., 9 (76), 24 Bartlett, M. F., 120(22a), 148(56), 156, 157, 201 (92d), 201, 207(25), 209(25, 30), 213(25), 216(25), 233, 251 (7), 253(7, 11),254(7, 11, 32, 40), 255(47), 268, 256(7, 13), 257(7), 258(7), 259(7), 262(27, 28), 264(27, 28), 265(32), 266(40), 267, 268, 281(57), 284, 303 (153),328,331,482 (164),485 (1777, 495(183), 511, 512, 534(71), 535(71), 537(71), 577, 716(66, 67), 719(75), 720(75), 723, 788(16), 790(35), 791(35), 792(16(, 793(16), 795(16,35), 796(35), 797(16), 798(35, 48), 800(35, 61), 801(6l), 804(16), 805(16), 806(35, 65), 812, 813, 814 Barton, W. E., 294(119), 330 Basak, M. G., 704(28), 722 Basu, N. K., 270(21), 283, 288(113), 330 Battegay, J., 602(57), 604(66), 605(66), 606(57), 667 Battersby, A. R., 238(1), 246, 266(37),

AUTHOR INDEX

268, 423(95, l02), 426(95, 102), 427(95), 429(95, 102), 431(109), 434(92), 435(95), 438(92, 95), 441(109),444(109),445(109),485(172, 173), 491(173, 178), 509, 522, 522, 518(15), 519(15, 21), 520(15, 29, 31), 521(15), 526(15), 533(68), 534(15, 72), 535(72, 80, 81), 536(81), 539(15), 540(15, 31, 84), 541(31), 542(31), 543 (31), 547(31,96), 554( 114),555( 15, 31, 84, 113), 556(96, 113), 559(96), 567(15, 114), 574(15), 575(15), 576, 577, 578, 579, 688(20), 690, 788(23, 28), 805(23, 28), 822 Batty, J. E., 12(94), 24 Bauer, A., 560(120), 561(120), 579 Bauer, S., 257(15), 261(15), 267, 276(42), 284 Bauerovit, O., 257(15), 261(15), 267, 276(42), 284 Baur, W., 596(22), 622(22, 117), 666, 668 Baxter, R. M., 766 (125), 767 (135), 782 Beak, P., 485(177), 512, 534(71), 535(71), 537(71), 577, 788(16), 792(16), 793( 16), 795 (16), 797 (16), 804(16), 805(16), 822 Becker, B., 749(85), 751(88), 781 Beckett, A. H., 61 (26a, 26c), 85(26a, 26c), 86(26a, 26c), 87(26a, 26c), 88(26a), 89(26c), 90 Beecham, A. F., 48(20), 52 Beer, C. T., 271(12), 272(2, 3, 12), 282, 283 Beerebom, J. J., 270(16), 283, 534(70), 577, 812 Beevers, C. A,, 598, 667 Bein; H. J., 811(73), 824 Bejar, O., 498(186e), 512, 533(57), 577 Belohlav, L., 277(38), 283 Bender, M. L., 44(58), 45 Bendixsohn, W., 604(64), 605(64), 607(64), 667 Bendz, G., 240(16), 241(16), 247 Benedict, R. G., 11(83b, 83c), 24 Bennett, F. C . , 272(11), 283 Bentley, H. R., 665(224), 672 Bentley, K. W., 33(26), 44 Bentley, R., 108(91), 217 Benyon, J. H., 359(20), 390(20), 391(20), 392(20), 407(20), 456(20), 506 Beran, M., 765(117), 782

817

Berde, B., 758(96), 781 Berend, L., 641 (168), 670 Bergel, F., 743(51), 745(51), 780 Berger, A., 44(61), 45 Bergoeing, R., 710(48), 714(48), 716(48), 723 Berlage, F., 128(38), 156, 466(124), 510, 520(28), 534(28), 541(85), 543(85), 546(85, 95), 547(97), 548(85, 98), 562(112), 576, 578, 579 Bernal, R. M., 328 Bernard, S . A., 44(6l), 45 Bernasconi, R., 175(82), 178(82), 179(82), 180 (82), 182(82), 187(82), 190(82), 201 Bernauer, K., 126(33), 128(38), 156, 418(67), 466(67, 124), 508, 510, 520(28), 534(28), 541(85, 86), 572(86, 87), 543(85, 86), 545(94), 546(85, 95), 547(86,97), 548(85,98, 101), 552(108, log), 553(110), 561(121, 122), 562(112, 124, 125), 563(124, 127), 564(110, 127), 565(127), 566(127), 567(127), 569(131), 570(131), 576, 578, 579,618(108), 628(135), 633(108, 127), 668, 669, 682(17), 690 Bernini, G., 14(107), 25 Bersch, H. W., 42(50), 45 Bertho, A., 680, 681(2, 9, lo), 690 Berton, A., 96(63), 216, 238(8), 247 Bertrand, G., 17(120), 25 Besch,E., 466(132), 510,523(47), 549(105), 551(105, 107). 562(47), 577, 579 Besendprf, H., 327(230), 334 Beyer, H., 612(89), 622(199), 628(138), 668, 669 Bhatia, B. B., 811 (74), 824 Bhatnagar, S. S . , 271(8), 272(8), 283 Bhattacharji, S., 767(134), 782 Bhettacharya, A., 429(103), 441 (103, 104, 107), 444(106, 107), 509 Bhattacharya, A. K., 441 (105), 443(105), 509 Bhattacharyya, N. K., 67(67), 91 Bianchi, E., 399(37), 432(37), 435(37), 473(37), 485(37), 505 (37), 507 Bianculli, J. A., 281(67), ZS2(67), 284 Bickel, H., 300(139), 316(183), 332, 332, 521(42), 523(42), 524(42,49), 525(49), 526(49), 577 Biemann, K., 212(34), 221 (34), 224(43a),

818

AUTHOR INDEX

234, 259(19), 262(19), 267, 271(28), 280(51), 283, 284, 337(4, 9), 359(18, 18a, e l ) , 361(18), 365(18), 367(9, 18a), 390(4, 18, 21, 28, 28b, 51, 51a), 391(9, 21, 28), 393(21), 394(21), 395(4, 21, 28, 51a), 396(9, 28, 51, 51a), 397(18, 28, 51a), 401(9), 404(9), 407(4, 21), 419(28,72a), 431 (21),440(97),453(28, 51a, 118), 456(21, 28, 51a), 461(118), 484(51, 166, 167, 168, 169), 490(167), 491(179b), 494(179b), 495(28, 51, 179b),497(21,51,185),504(192e), 510, 511, 512, 664(217a), 671, 788(13), 800(13), 801(13), 805(13), 808(13, 67, 68, 69, 70), 811(13), 812, 814 Bijvoet, J. M., 598(36), 667 Binks, R., 485(173), 491 (173), 511,518(15), 519(15,21),520(15),521(15),526(15), 534(15), 529(15), 540(15), 555(1), 567(15), 574(15), 575(15), 576, 812 Binst, G. V., 806(64), 814 Birch, A. J., 176(76), 177(76, 81), 180(76), 183, 184(81), 186(86), 187(86), lSS(S6), 189(86), 200, 201, 767(133, 134), 782 Bisset, N. G., 439(88), 508 Bizzi, L., 326(223), 334 Blaha, L., 318(189), 319(191), 321(191), 322(189, 194), 325(197), 333 Blaise, H., 60, 90 Blanpin, O., 232(68, 69), 235 Blatzejewicz, L., 651 (20l), 652(201), 653(201), 654(201), 657(201), 671 Blomster, R. N., 282(84), 285 Blossey, E. C., 434(48), 445(48), 457(48), 462(48), 469(48), 474(48), 477(48), 482(48), 507 Blount, B. K., 596(23), 607(73), 615(23), 640(23), 666, 667 Boaz,H. E., 167(59), 173(70), 200,273(13), 275(13), 283, 293(83, 84, 104), 329, 330, 711(51), 723 Bodendorf, K . , 293(64,IOI), 329, 330 Boegemann, W. H., 293(21), 327, 512 Boehm, R., 517, 576 Boehringer, A., 294(122), 330 Boehringer, C., 294(122), 330 Boehringer, R., 20(136), 25 Boekelheide, V., 518 (18), 519 (18), 520 (18), 521(18), 543(92), 549(18), 568(129), 570(129), 576, 578, 579

Boesler, W., 50(33), 53 Bogdanikowa, B., 280(78), 285 Bogdanski, D. F., 16(115), 25 Bohlmann, F., 302(145), 331, 708(42), 722 Boit, H. G., 600(48), 611, 612(48), 613(95, 96, 97), 615(97, 98, 102), 619(48), 620(112), 637(102), 638(98, 112), 640(95, 98), 667, 668 Bokhoven, C., 598(36), 667 Boller, A., 548(100), 578 Bommer, P., 390(28b), 506 Bonino, C., 747(69), 781 Borisyuk, Y. G., 277(50), 284, 288(116), 330 Borkowski, B., 48(18), 50(18), 52 Bose, A. K., 332, 788(7a), 812 Bose, B. C., 811 (73), 814 Bose, P. K., 429(103), 441(103), 509, 649(193), 653(193), 656(193), 657(193), 670 Bose, S., 56(5), 57(5), 58, 293(105, 107), 308(166), 327, 328, 330, 713(62), 723, 799(54, 55, 56), 800(55, 56, 58), 813 Bosly, J., 665(222, 223), 670 Boutroux, L., 238(5), 246(5), 246 Bovet, D., 516(2), 520(25), 523(25), 576 Bovet-Nitti, F., 516(2), 576 Bowden, K., 6(53), 12(94), 23, 24, 406(50), 507 Boyer, P. D., 44(58, 59), 45 Brack, A., lO(81, 83), 11(83), 12(83, 90, 93a), 24, 763(108), 764(108, 112), 765(108,112,119), 766(108), 767(134), 782 Brady, L. R., 11(83b, 83c), 24, 726(1), 735(23), 751(1), 779, 780 Brandt, K., 4(34), 5(34), 22 Braun, L., 13(100c), 24 Breccia, A., 6(55), 23 Brehm, W. J., 592, 597(30), 598, 607(3), 617(3), 632(30), 666, 791(37), 813 Bretschneider, H., 594( 18), 666 Brewer, €I W., . 390(35, 36), 392(36), 405(35, 36), 406(36), 407(35, 36), 507 Brewster, J. F., 604(65), 667 Breyer-Brandwijk, 272 (7), 283 Briggs, L. H., 596(28), 598(34), 615(28), 617(34), 632(34), 666 Bringi, N. V., 169(62), 200, 238(11), 247, 264(30), 268, 308(165), 332, 485(177), 511, 534(71), 535(71), 537(71), 577,

AUTHOR INDEX

819

696(10), 698(10), 709(50), 711(50), 102), 430(81), 432(37), 434(82), 718(50), 722,723,788(7a, 16), 792(16), 435(37, 82), 438(82), 445(52), 447(52), 793(16,43), 795(16),797(16),804(16), 448(52), 450(113e), 456(115a), 457 (115a, 117), 460( 117), 461 (117), 805 (16), 812, 813 Brissolese, J. A,, 82(82), 83(82), 84(82), 463(122), 465(115a, 122), 466(122), 85(82), 91, 434(48), 445(48), 457(48), 467(125a, 126, 127), 468(125a), 462(48), 469(48), 474(48), 477(48), 469(125a, 127), 473(37), 482(165), 484(165, 170, 171), 485(37, 122, 165), 482(48), 507 Britten, A. Z., 138(50), 139(50, 53), 490(165, 165a), 491(165), 495(170, 142(50), 144(50), 145(50), 147(50, 180), 497(170), 503(192g), 505(37), 53), 148(53), 149(53), 151(50, 53), 505, 506, 507, 508, 509, 510, 511, 512, 152(51a),154(55), 157,491(179e),512 513, 535(77), 578, 694(i), 696(i), Brizzolara, A., 365(27b), 506 707(1), 721,788(21), 805(21), 808(21), Brodie, B. B., 16(115), 25, 327(230), 334 812 Brook, P. R., 175(82), 178(82), 179(82), Budzikiewicz, J. M., 270(32), 271 (29), 180(82), 182(82), 187(82), 189(82), 276(40), 277(49), 281(64), 283, 284 201, 361 (24), 362(24,26), 443(24),506 Bulard, C., 13(96), 24 Brossi, A,, 327 (229), 334 Bumpus, F. M., 17(127), 25 Brown, B. G., 12(94), 24 Burckhardt, C. A., 204(2), 207(23), 233 Burckhardt, E., 748(75), 759, 781 Brown, K. S., 451(113g), 509 Brown, K. S., Jr., 450( 113e), 452 (113i), 509 Burger, A., 65(62), 91 Burgstahler, A. W., 702(26), 704(26), 722 Bruce, D. W., 13(100), 24 Bruecke, F., 326(228), 334 Burlingame, A. L., 224(43a), 234, 259(19), Brunner, R., 328, 758(96, 97), 763(108), 262(19), 267, 390(51), 396(51), 453( 118), 461 (118), 484(51), 764(108, 112), 765(108, 112. 119), 491(179b), 494(179b), 495(51, 179b), 766(108), 781, 782 Brutcher, F. V., 9(69), 23 497(51), 504(192e), 507, 510, 512, Bubenik, R., 641(175),647(175), 663(175), 788(13), 800(13), 801(13), 805(11), 670 808(13), 811(13), 812 Burnell, R. H., 451(113f), 509 Bucharova, D. A,, 282(85), 285 Buchi, G., 266(39), 267(39), 268, 419(72a), Buzas, A., 325(211), 333, 505(194), 513, 469(142), 471(142), 476(158), 701 (18), 722 477(158), 480(158), 498(158, 186d), 499(186d), 502(186d), 508, 510, 511, C 512 Bucourt, R., 9(70b), 23, 316(184, 185), Cahen, R., 95(44), 116 318(185), 321(185), 322(185), Cahill, W. M., 3 ( l a ) , 22 323(185), 324(185), 332, 333 Cailleux, R., 12(89, go), 24 Budzikiewicz, H., 82(81, 82), 83(82, 83), Cais, M., 288(114), 330, 400(41), 450(41), 50 7 84(82), 85(82), 91, 134(44), 139(51a), 147(51a), 148(51a), 149 (514, Calvet, F., 647(191), 649(191), 662(209), 670, 671 151(61a),152(51a),153(51a),157 (57), 157, 191(92), 192(92), 194(92), 201, Cameran, N., 273(14), 283 224(44), 225(44), 234, 258(18), Camerman, N., 220(42c), 234 260(18), 261(18), 262(18, 25, 26), Campello, J., 434(48), 445(48), 457(48), 462(48), 469(48), 474(48), 477(48), 265(42, 43), 266(44), 267, 268, 337(3, 482(48), 507 6,7), 390 (6,7, 28a, 32,36,52), 392 (36), 393(6), 399(37), 405(6, 35, 36), Campos, J. S., 328 406(36), 407(35, 36), 414(52), 416(3, Cannon, J. R., 250(4), 267 7), 418(7), 419(36), 420(32), 423(81, Caprio, L., 325(212), 333 82, 102), 426(81, 82, 102), 429(81, Carcamo, V., 507

820

AUTHOR INDEX

Carpenter, R. D., 48(22), 52 Carr, F. H., 727(6, 7), 749, 759(7), 779 Carrazzoni, E. P., 434(48), 445(48), 457(48), 462(48), 469(48), 474(48), 477(48), 482(48), 507 Carter, J. H., 44(61), 45 Carter, P. H., 470(143), 510 Cartier, P., 13(98), 24 Carvalho-Ferreira, P., 511 Caserio, M. C., 313(181), 332 Casinovi, G. C., 517(7), 519(7, 20), 520(7), 521(7, 431, 523(20, 43), 548(7), 576, 577 Catalfomo, P., 17(126a), 25 Cava, M. P., 416(113c), 450(113b, c, j), 497(113b), 504(192f), 509, 512, 599(39), 635(156), 642(156), 644(156), 667, 670, 676(5), 677(5), 678(7), 678 Caventou, 592, 666 Ceder, O., 518(18), 519(18), 520(18), 521(18), 543(92), 549(18), 568(129), 570(129), 576,578,579 Ceglowski, M. J., 296(126), 307(164), 311(126, 164), 330, 332 Cekan, Z., 259(20), 261(22, 24), 267, 276(33, 35, 36, 45, 47, 48), 277(36, 46, 48), 281(36), 282(33), 283, 284, 495(174), 512 Cerletti, A., 772, 774, 783 Cernq, A., 768(137), 769(148), 770(148), 782, 783 Chabasse-Massonneau,J., 531 (56), 577 Chaigneau, M., 227(56), 230(56), 231 (56), 234 Chain, E. B., 747(69), 781 Chakrabarti, J. K., 102, 103(79), 104(79),

330,332,429(103), 441(103, 107, 110), 444(106, 107), 491(62), 505(62, 197, m),508, 509, 707(33), 713(62), 722, 723, 792(42), 799(56), soo(56, 57, 58, 59), 813 Chotterjee, B. G., 308(166), 332, 788(7a), 812 Chaudhury, N. A., 161(36), 195(36), i96(36), 199,327, 329,707(33), 7 2 2 Chen,A.L.,48(12),52,55(1), 58 Chen,K. K., 17(122),25,48(12), 52,55(1), 58,95(45,53,54,55), 116, 162(46),200 Cheng, W. O., 95(22), 115 Chesher, G. B., 13(95), 24 Chi, Y. F., 95(21, 23), 96(21), 115 Chiavarelli, S., 523(45), 577 Chick, O., 238(5), 246(5), 246 Chisholm, M. D., 6(57), 23 Chopra, I. C., 811 (73), 824 Chopra, R. N., 246(35), 247, 272(7), 283, 811(73), 814 Chou, T. Q., 94(15, 20), 95(22, 54), 96(58, 61), 97(58), 107(15), 110(14), 115, 116 Christensen, B. V., 95(35), 116 Chu, T. T., 96(58, 61), 97(58), 116 Claassen, V., 326(228), 334 Clark, V. AT., 585(21), 589 Clayton, E., 134(45), 157, 511 Clemo, G. R., 51(45), 53, 596(27), 598(33), 606(33), 607(33, 76), 608(76), 610(33, 76), 612(33), 618(76), 628(76), 641(27, 76, 173), 665(33), 666, 667, 670 Clinquart, E., 19, 25, 119, 155 Cocking, T. T., 735(30), 780 Cohen, A., 743(51), 745(51), 780 Cohen, L. A., 359(17), 506 iir Cole, S. W., 50(32), 53 Chokravarti, D., 182(83), 201, 293(82), Collera, O., 204(11), 216(11), 217(11), 329,785(1),789(1),790(1),791(1),812 224 (1l), 227 (1I), 233, 337 (8),358 (8), Chakravarti, H., 162(44), 200 416(66a), 457(8), 505(8), 506, 508 Chakravarti, R. N., 182(83), 201, 634(32, Collier, H. 0. J., 13(95), 24 153), 635(32), 642(176), 666, 669, 670 Colb, V., 13(103), 14(103), 24 Chalmers, J. R., 462(120a), 510 Combes, G., 326(222), 328, 334 Chang, M. Y., 329, 799 (53), 813 Comella, G., 625 (132), 669 Chang-pai, C., 243(26, 27, 28), 247 Cone, N. J., 204(10, 12), 218(12), 225(10), Chatterjee, A., 56(5), 5 7 ( 5 ) , 58, 105(80), 226(10,48),228(lO), 233,234,255(46), 127, 120(20), 156, 161(34, 35), 268, 273(13), 275(13), 283, 419(72a), 177(78, 79, 80), 178(78, 79, 80), 474(148), 508, 511 191(91), 194(35), 195(35), 197(95), Conroy, H., 102, 103(79), 104(79), 117, 199, 200, 201, 202, 270(18), 283, 175(82), 178(82), 179(82), 180(82), 293(19, 96, 103, lob), 327, 328, 329, 182(82), 187(82), 190(82), 201,

AUTHOR INDEX

361 (24), 362(24, 26), 443(24), 506 Constantine, G. W., 327(232), 334 Cook, J. W., 60(14), 61(14), 62(51), 64(14, 55), 75(14), 85(14), 89, 90 Cookson, R. C., 738(45), 780 Coppola, J., 211(31), 233 Corral, R. A., 390(36), 392(36), 405(34, 35, 36), 406(36), 407(35, 36), 507 Corsano, S., 701 (17), 702(25), 705(25), 722 Cortese, F., 621(113), 660(207), 668, 671 Corvillon, O., 195(94), 201, 523(44), 575(44), 577 Cowley, R. C., 272(11), 283 Coxworth, E., 40(48), 41(49), 45 Crabb, T., 543(92), 578 Craig, L. C., 734(20,21), 735(21, 34, 35, 36, 37, 38, 39), 748(80), 749(82, 83, 84, 86), 780, 781 Crane, E., 659(206), 663(206), 671 Cranwell, P. A., 480(159, 160), 511 Creveling, C. R., 13(99), 24 Cross, A. D., 61(37), 80(37), 82(37), 90, 266 (45), 268 Crow, W. D., 160(10), 199, 288(112), 330, 434(87), 435(89), 438(89), 439(87, 88, 90), 440(89, 97), 444(89), 508, 509 Culvenor,C. C. J., 240(16), 241(16, 23, 24), 243(23, 24), 247 Curtis, R. G., 160(11), 199, 288(112), 330 Cushny, A. R., 94(6), 115 Cutts, J. H., 272(2, 3), 282, 283

D Daeniker, €1. U., 599 (39), 635 (156), 642(156), 644(156), 667, 670 Dale, H. H., 727, 779 Dalev, P., 282(72), 285 Dallemagne, M. J., 665(221), 671 Daly, J. W., 359(17), 506 Danilova, A., 195(93), 201 Das, N. N., 811 (73), 814 Das, P. K., 13(64b), 23 Das Gupta, B. M., 162(44), 200 d a Silva, P., 328 Dastoor, N., 482(113d), 509 Dave, K. G., 476(154), 511 David, J. C., 246(34, 35), 247 Deb, A., 441(107, l l o ) , 444(106, 107), 509 de Espanes, E. M., 95(38, 43, 44), 116

82 1

Defay, AT., 225(58), 226(46), 234,491 (179), 512, 535(76), 578, 788(30), 806(30, 64), 813, 814 De Feo, J. J., 265(41), 268, 498(186c), 512 Deger, E. C., 293(37), 328 de Groot, S., 326(216), 333 d e Jough, D. K., 246(34), 247 Delahunt, C. S., 327(232), 334 d e la Lande, I. S., 663(214), 671 De Langhe, J., 507 Delay, J., 12(91, 92), 24 Delourme-HoudB, J . , 204(3), 233(3), 233 Delvaux, E., 19, 25 de Moerloose, P., 237(3), 246 Denis, P., 59(4), 60(11, 12), 89 Denisova, L. I., 753(93), 781 Denoel, A., 665(218, 22l), 671 de Sa, J., 271(8), 272(8), 283 Deulofeu, V., 3(21), 17(123, 124), 22, 25, 327, 328, 403(47), 405(147), 469(141), 474(14i, i47), 505(147), 5 0 7 , 5 1 0 , 511 de Vrij, J. E., 240(14), 241 (la),247 Deyrup, J . A., 204(10), 225(10), 226(10), 228(10), 233, 474(148), 511 Diassi, P. A,, 293(62), 300(134, 135, 141), 302(135), 329, 331, 704(27), 722 Dibhern, H. W., 771 (158), 783 Dickel, D. F., 204(6), 205(6), 207(25), 209(25, 30), 213(25), 216(25), 223(6), 224(6), 225(45), 233, 234, 293(76, 89), 294(121), 300(134), 329, 330, 331, 704(27), 722, 806(65), 814 Diels, W., 601 (52), 602(52), 604(52), 629(140), 667, 669 Diepolder, E., 50(34), 53 di Paco, G., 326(219), 334 Djerassi, C., 82(82), 83(82, 83), 84(82), 134(44), 139(51a), 85(82), 91, 147(51a), 148(51a), 149(51a), 151(51a),152(51a), 153(51a), 157(57), 157, 160(17), 191 (92), 192(92), 194(92), 199, 201, 224(44), 225(44), 234,258(18), 260(18),261(18),262(18, 25, 26), 265(42, 43), 267, 268, 270(16, 3 2 ) , 27 1 ( 2 9 ) , 2 7 6 ( 4 0 ) , 2 7 7 ( 4 9 ) , 2 8 1 ( 6 4 ) , 283, 284, 288(114), 294(39), 327, 328, 329, 330, 337 (3, 6, 7, 9, lo), 367 (9,29), 389(10, 29), 390(6, 7, 10, 28b, 32, 35, 36, 52), 391(9), 392(36), 393(6), 396(9), 398(40), 399(37,40a), 400(41), 401(9, lo), 403(40), 404(9, lo), 405(6,

822

AUTHOR INDEX

34, 35, 36), 406(36), 407(35, 36), 410(29), 411(10, 29), 414(10, 40a, 52), 415(29), 416(3, 7), 418(7), 419(3, 6), 420(32), 423(80, 81, 82, 102), 426(80, 81, 82, 102), 429(49, 81, 102),430(81), 431 (98, 99), 432(37), 433(80), 434(48, 49), 435(37, 82), 438(82), 445(48, 52), 447(52), 448(52), 450(41, 113e), 452(113i), 456(115a), 457(48, 115a, 117), 460(117), 461(117), 462(48), 463(122), 465(115a, 122), 466(122), 467 (125a, 126, 127), 468 (125a), 469(48, 49, 125a, 127), 473(37, 143a, 143b), 474(48), 477(48), 482(48, 163, 165), 484(163, 165, 170, 171), 485(37, 122, 165), 490(165, 165a), 491(165), 495(170), 497(170), 503(192g), 505(37), 505, 506, 507, 508, 509, 510, 511, 513, 534(70), 535(77), 577, 578, 694(1), 696(l, 4), 699(4), 707(1), 721, 788(21), 797(45), 798(52), 799(52), 805(21), 808(21), 812 Dodo, T., 244(30), 247 Doering, W. E., 635 (157), 670 Dolby, L. J., 642(184), 670 Dolfini, J. E., 361(27a), 365(27a), 367(26a), 506 Dorfmen, L., 296(125), 300(125),303(155), 325(155), 328, 330, 331, 798(48), 813 Dornow, A., 601 (52), 602(52), 604(52), 606(71), 622(118), 667, 669 Douglas, B., 450(113b), 451(113b, j), 497(113b), 498(186b), 503(192d), 504(192d, 192f), 509, 512, 676(5), 677(5), 678(7), 678 Doy, C. H., 2(9), 21 Doy, F. A., 288(117), 330, 480(161), 482(161), 505(161), 511 Doyne, T., 300(137), 331 Draper, M. D., 293(61, log), 328, 329, 330 Draus, F. J., 281(67), 282(67, 84), 284, 285 Drew, W. 13. M., 271(8), 272(8), 283 Drill, V. A., 43(54), 44(54), 45 Duarte, A. P., 434(48), 445(48), 457(48), 462(48), 469(48), 474(48), 477(48), 482(48), 507 DfibravkovB, L., 261 (23), 267, 276(45, 87), 277(45, 86), 280(86, 8 7 ) , 284, 285, 337(10a), 420(73, 74a), 506, 508

Ducke, A., 520(25), 523(25), 576 Dudek, G. O., 504(192f), 512 Dudley, H. W., 748(74), 781 Duprat, E., 17(124), 25 Durandin, M. C., 12(91), 24 Durham, L. J., 83(83), 91, 134(44), 139 (51a), 147 (51a), 148 ( 5 1 4 , 149(51a), 151(51a), 152(51a), 153(51a), 157, 258(18), 260(18), 261(18), 262(18, 25, 26), 265(43), 267, 268, 271(29), 276(40), 277(49), 283, 284, 337(3, 7), 390(7, 32), 399(37), 416(3, 7), 419(3), 420(32), 423(81), 426(81), 429(81), 430(81), 432(37), 435(37), 473(37), 482(165), 484(165, 170), 485(37, 165),490(165), 491 (165), 495(170), 497(170), 505 (37), 505, 506, 507, 508, 511, 535(77), 578, 694(1), 696(1), 707(1), 721, 788(21), 805(21), 808(21), 812 Dusenberry, J. E., 765(120), 782 Dutcher, J. D., 516(4), 576 Dutt, S., 61(38), 80(38), 90 D u d , C., 246(39), 247 Dybowski, C., 610(80, 83), 668 Dybowski, J., 204(1), 206(1), 233 Dyer, E., 60(19), 61(19), 64(19), 75(19), 85(19), 89 Dylion,C. M., 300(134,135,141), 302(135), 331, 704(27), 722 Dziemian, R. L., 209(28), 213(35), 233, 234, 303(151), 304(151, 160, 162), 305( 162), 307 (164), 311 (164), 312(151), 331, 332

E Earl, A. E., 311 (175), 332 Eber, W., 28(7), 44 Ebnather, A., 568(128), 579 E d e r , H . , 293(64, 101), 329, 330, 638(164), 651(201,202),652(201),653(201,202), 654(201, 202), 656(202), 657(201), 670, 671 Edward, J. T., 603(62), 604(62, 63), 607(63), 667 Edwards, 0. E., 61(28), 67(28), 70(28), 73(28),76(28), 78(28), 90,108(88),I17 Edwards, P. N., 126, 128(35), 129(39), 130(39,40), 131 (40), 138(50), 139(50), 142(50), 144(50), 145(50), 147 (50),

AUTHOR INDEX

151(50), 156, 157, 231(66), 235, 405(77), 465(123), 466(77, 123, 137a), 508, 510, 533(67), 477, 681, 682(13). 690, 786(7), 812 Ehrenberg, A,, 28(4), 44 Eichler, O., 95(46), 116 Eiter, K., 582(6), 584(14, 15, 16), 586(6, 26), 588, 589 Elderfield, R. C., 160(8), 162(8, 32a), 163(49), 164(54), 165(54), 166(55). 170(65, 66), 172(66, 67, 68), 173(68, 70), 199, 200, 529(51), 577 Els, H. E., 451(1131, m), 510, 673(2, 3), 674(3), 675(2, 3), 676(3), 678 Elze, F., 1 (2), 2(2), 21 Engelhardt, A , , 264(31), 268 Englhart, E., 43(52), 45 Enslin, P., 164(50, 51), 200 Erdtman, H., 5(38, 41, 42), 22 Ernest, I., 318(187), 319(190), 321(190, 193), 322(190, 193), 333 Erspemer, V., 13(103), 14(103), 24 Espejo, 0. P., 328 Eugster, C . H., 619(110), 668 Everett, A. J., 362(27), 400(27), 506 Evers, N., 95(32), 116 Evstigneeva, R. P., 243(26, 27, 28), 247 Ewins, A. J., 9, 19(66), 23, 395(38), 507

F Fabbri, S., 739(46), 780 Fabing, H. D., 16(118), 25 Fahrbach, E., 629(143), 630(144), 669 Fan, C., 271(27), 281(56), 283, 284, 416(64), 418(64), 494(179c), 508, 512 Farnsworth, N. R., 272(9), 281(67), 282(67, 84), 283, 284, 285 Fawcett, R. C., 594(16), 666 Feldstein, A., 767(130), 782 Fellion, E., 420(76), 508 Fernandes, F., 271(8), 272(8), 283 Ferosie, M. I., 680(2), 681(2), 690 Ferrari, C., 396(42), 403(42), 445(42), 507 Ferreira, 5 . M., 265 (43), 268, 399 (37, 40a), 414(40a), 423(80, 81), 426(80, 81), 429(81), 430(81), 431 (37), 433(80), 435(37), 452(113g), 473(37), 485(37), 505 (37), 507, 508, 509 Ferreira, M., 96(59), 116 Ferretti, A., 747 (69), 781

823

Festag, W., 533(61), 577, 689(26), 691 Fichter, F., 20(136, 137, 138), 25 Field, E., 59(2), 61(2), 62(2), 89 Figueiredo, A., 452 (113g),509 Finch, N., 60(14a), 66(66), 69(66), 70(66), 77(66), 79(66), 80(66), 82(82), 83(82), 84(82), 85(82), 89, 91, 265(42), 268, 309(171), 310(172), 327, 332, 423(82), 426(82), 434(82), 435(82), 438(82), 484(171), 502(187, 189), 503(187, 189, 192% 192g), 5 0 8 , 5 1 1 , 5 1 2 , 5 1 3 , 7 1 2 ( 5 4 , 57, 58), 721 (54). 723 Findlay, S. P., 619(111), 642(111), 668 Finger, K. F., 327(232), 334 Finocchio, D., 311(174, 175), 332 Fischer, B. A., 172(68), 173(68), 200 Fischer, F. E., 23(101), 24 Fischer, O., 49(25, 29), 50(29, 31, 33, 34), 52, 53 Fischer, R., 766(128), 767(128), 782 Fischer, R. F., 451(113k, 1, m), 509, 510 673(1, 2, 3), 674(3), 675(2, 3), 676(1, 3), 678 Fish, F., 227(52), 232(52), 234, 511 Fish, M. S . , 8(65), 9(75), 10(75), 16(75), 19(75), 23, 24 Fishman, J., 329, 798(52), 799(52), 813 Fitzgerald, J. S., 4(30a), 22 Flagler, M. B., 512 Flammersfeld, H., 63?(163), 670 Fleming, A. M., 294(119), 330 Flores, S. E., 271(29), 283, 337(3), 399(37), 416(3), 419(3), 432(37), 434(48), 435 (37), 445 (48), 457 (48), 462(48), 469(48), 473(37), 474(48), 477(48), 482(48), 485(37), 505(37), 505, 507 Floss,H. G., 2(10), 21, 766(124), 767(131), 782 Fluckiger, 592(1), 666 Folkers, K., 3 (22, 23, 24, 25, 26), 22 Fondovila, M., 3(21), 22 Fong, H. H. S., 282(84), 285 Forbes, J. W., 293(83, 84, 104), 329, 330 Fornefeld, E. J., 735(26), 742(50), 745(26, 50), 755 (50), 780 Forsyth, W. G. C., 94, 96(16), 97(16), 107(16), 110(16), 115 Foy, J. M., 13(100a), 24 Freter, K., 772(106), 783 Frey, A., 10(83), l l ( 8 3 , 84), 12(83), 24, 311(177), 323(199), 324(199), 332, 333

824

AUTHOR INDEX

Frey, A. J., 293 (71,72), 300( 139), 316( 183), 329, 331, 332, 749(169), 753(92), 755(173), 756(173, 174), 769(92), 771(165), 781, 783,798(46,47,51), 813 Friedman, T. T., 294(119), 330 Friedmann-Spiteller, M., 212(34), 221 (34), 234, 390(28), 391 (28), 395(28), 396(28), 397(28), 419(28), 453(28), 456(28), 482(264a), 484(168, 169), 495(28), 506, 511, 808(69), 814 Friedrich, W., 94, 112, 113(19), 115 Fritsche, J., 48(1), 49(1), 52 Fritz, H., 466(132), 510, 523(47), 549(103, 105), 551 (105, 106, 107), 560(120), 561(120),562(47),570(133), 573(135), 577, 578, 579, 709(47), 710(47), 723 Fiirst, A., 548(100), 578 Fukami, H., 329, 799 (53), 813 Fukuda, T., 60(15), 89 Fukumoto, K., 9(70a), 13(70a), 23 Furlenmeier, A., 276(34), 283, 296(125), 300(125), 329, 330

G Gabbai, M., 281(6l), 284, 285, 707(35), 722, 812 Gabhai, Y., 160(19), 199, 282(75), 285 Gailey, R. M., 64(55), 91 Gambarin, F., 61 (36), 90 Ganguli, G., 197(95), 201, 505(197, 199, 200), 513 Garbers, C. F., 400(44), 411(44), 463(44), 50 7 Garbrecht, W. L., 768(144), 783 Garcia, F., 416(66a), 508 Garratt, J. H., 113, 115(104), 127 Garrattini, S., 311 (176), 332 Garria, P., 328 Gates, M., 106(81), 107(81), 117 Gauthier, M., 226(46), 234 Gazet du Chatelier, G., 280(76), 285 Gearien, J. E., 326(221, 224), 334 Geffroy, Y., 13(98), 24 Geiger, W., 630(143a), 636(143a), 669 Geissman, T. A., 5(49), 23 Gelbrecht, H., 264(31), 268 Gellhrt, E., 19(135), 20(135), 25, 95(26), 116, 207(23), 233, 508 Gemenden,C. W., 310(172), 332,502(189), 503(189), 512, 712(58), 723

Gentile, R. A , , 3(27), 22 Genvresse, P., 238(5), 246(5), 246 George, T., 265(42), 268, 337(10), 367(29), 389(29), 390(10), 399(37), 401(10), 404(10), 411(10,29), 414(10), 415(29), 423(82), 426(82), 432(37), 434(82), 435(37, 82), 438(82), 473(37), 485(87), 505 (37), 506, 507, 508 Georgiev, V., 282 (72), 285 Gerrard, A,, 93, 115 Ghatak, N., 2(12, 13), 22 Ghosal, C. R., 120(20), 156, 491(62), 505 (62, 200), 508 Ghosal, S., 105(80), 117, 120(20), 156, 161(34), 177(77, 78, 79, SO), 178(77, 78, 79, 80), 191(91), 197(95),199, 200, 201, 231 (66), 235, 491 (62), 505(62), 508, 533(67), 577, 786(7), 812 Ghose, R., 182(83), 201 Ghosh, A,, 56(5), 57(5), 58 Ghosh, B. K., 162(43), 200 Ghosh, B. P., 327 Ghosh, S., 272(7), 283 Gibson, M. S., 96(60), 116 Giesbrecht, E., 520(23), 521 (23, 36, 42), 523(23, 42), 524(42), 574(23, 36), 575(36), 576, 577 Gilbert, B., 82(82), 83(82, 83), 84(82), 85(82), 91, 134(44), 157, 262(26), 265(42, 43), 267, 268, 337(9, lo), 367(9, 29), 389(29), 390(10, 52), 391 (9), 336(9), 398(40), 399(37, 40a), 401 (9,10), 403 (40),404(9,10),410 (49), 411(10, 29), 414(10, 40a, 52), 415(29, 423(81, 82). 426(81, 82), 429(49, 81), 430(81), 432(37), 434(48, 49, 82), 435(37,82), 438(82), 445(48), 445(52), 447(52), 448(52), 457(48), 462(48), 467 (125a), 468 (1254, 469 (48, 49, 125a), 473(37,143b), 474(48), 477(48), 482(48, 163, 165), 484(163, 165, 171), 485(37, 165), 490(165), 491(165), 505(37), 506, 507, 508, 510, 511, 535(77), 578, 694(1), 696(1), 707(1), 712(57), 721, 723, 788(21), 805(21), 808(2l), 812 Cillen, M. A., 328, 798(48), 813 Gillo,L., 204(9),227(9, 54), 232(9,54), 233, 234 Gilman, R. E., 161(38), 197(38), 200, ?92(39), 813

AUTHOR INDEX

Godtfredsen, W. O., 303(150, 167), 331, 696(6, 12), 698(6, 12), 721, 722 Goebel, F., 48(2), 49(2), 52 Gogolak, G., 326(228), 334 Goldman, L., 302 (144), 331 Goldner, M., 94, 115 Goldschmidt, H., 641 (170), 670 Goldsworthy, L. J., 240(16), 241(16, 23, 24), 243(23, 24), 247 Gomez, L., 328 Goncalves de Lima, O., 10 (79), 24 Goodson, J. A,, 120(6), 155, 161(30, 32), 162(32, 39), 174(32), 191(30), 199, 200, 510, 811 (78), 814 Goodwin, S., 240(17), 247, 264(29), 268, 477(156), 505(156), 511 Gopinath, K. W., 584(17), 589 Gordin, H. M., 581(1, 2), 582(1), 588 Gordon, J., 40(48), 45 Gorman, M., 204(10, 12), 216(40), 218(12, 40), 219(40, 42, 42a), 220(42a, 42b), 221 (42a), 223(42a, 42b), 225(10, 42), 226(10), 228(10), 233, 234, 270(6, 22, 23, 31), 271(6, 12, 22, 23, 28, 29, 30), 272(6, 12), 273(6, 13, 15), 275(6, 13), 283, 294(39), 328, 337(3, 4, 5), 390(4), 395(4), 407(4), 416(3), 419(3, 5, 72a), 431 (99), 474( 148), 491 (179b), 494(179b), 495(179b), 505, 506, 508, 509, 510, 511, 512, 788(13), 797(45), 798(52), 799(52), 800(13), 801(13), 805(13), 808(13), 811(13), 812, 813 Gorter, K., 439(85), 508 Goschke, R., 587(30), 589 Gosset, J., 120(22), 133(43), 134(43), 156, 157, 225(58), 234, 281 (58, 59, 64), 284, 337(6), 390(6), 393(6), 405(6, 74), 419(6, 74), 490(165a), 491(165, 176), 506, 511, 535(74), 537(74), 578, 788(11), 805(11, 63), 812, 814 Goto, M., 10(78), 24 Gould, R. G., Jr., 729, 735(38), 744(55), 779, 780 Goutarel, R., 48(16), 52, 94, 97(66), 98(71), 99(66), lOO(66, 71), 101(71), 107(86), 108(86), 112, 113(19), 115, 116, 117, 204(2, 8), 205(5, 7), 206(21), 207(24, 26), 209(26, 29), 211(5), 213(36), 215(37,39),216(39), 227(8, 50, 51, 53, 56), 230(56, 62), 231(39, 56), 232(68), 233, 234, 235, 237(6), 238(6, lo),

825

240(6), 246(6), 246,247,293(100), 328, 329, 330, 463(121, 122), 465(121, 122), 466(121, 122), 485(121, 122), 498(186e), 504(193), 510, 512, 513, 531(55, 56), 533(57), 577, 665(223), 671, 680(5), 683(3), 684(4), 685(4), 686(19), 690, 696(14), 699(14), 704(14), 706(14), 713(63), 714(63, 65), 719(74), 720(76), 721(76), 722, 723, 798(52), 799(52), 813 Govindachari, T. R., 175(72, 72a, 74, 75), 176(72, 75), 177(72, 74), 180(74, 75), 181(75), 186(75), (87), 187(75, 87), 188(87), 200, 201, 202, 427(101), 431(100),441 (100,101),443(101, l l l ) , 444(100, ,111, 113), 480(160a), 503(113), 509, 511, 584(17), 589 Gramling, L. G., 95(35), 116 Grant, I. J., 32(24), 44, 587(29), 589 Grant, R. L., 748(73), 781 Graul, E. H., 327(230), 334 Gray, A. P., 165(54), 200, 529(51), 577 Greet, Y. M., 160(10), 199, 288(112), 330, 439(88), 508 Gregory, H., 431 (log), 441 (log), 444(109), 445( log), 509 Greshoff, M., 3, 22, 265(33), 268, 272(10), 283, 441 (83), 508 Grewal, K. S., 61 (40), 90 Griffiths, P. J. F., 246(37), 247 Griot, R., 755(173), 756(173), 783 Grisebach, H., 766(121), 767(121), 782 Grob, C. A,, 743(52), 744(59), 780 Groger, D., 2(10), 21, 765(120), 766(121, 122, 124), 767(121, 131), 782 Gross, F., 811(73), 814 Griinow, H., 596(25), 607(72), 640(72) Griiss, J., 607(75) Gudjons, H. F., 51(40), 53 Gueorguieff, V. P., 282(85), 285 Guercio, P. A., 294(81), 329 Guggisberg, A,, 444(113), 503(113), 509 Gulland, J. M., 610(81), 628(81), 668 Gumlich, W., 542(88), 578, 605(67), 606(67), 611(67), 664(216), 667, 671 Gunthard,H. H., 734(19),736(19),737(19), 738(19), 780 Gupta, J. C., 811 (73), 814 Gurevich, E. L., 48(10), 52 Gurewitch, H., 281 ( 6 5 ) , 284 Gut, M., 328, 714(65), 723

826

AUTHOR INDEX

H Haack, E., 293(51, 68), 297(51, 95), 328, 329, 330 Habgood, T., 98(72), 99(74), 100(75), 101 (75), 117 Haeck, H. H., 326(227), 334 Hafliger, O., 638(166), 642(166), 665(166), 670 Hagen, J., 4(33), 22 Hahn, G., 51(40, 42, 43, 44), 53, 251(6), 267, 325(209), 333, 696(16), 700, 701 (l6), 722 Hall, R., 623(124), 638(124), 640(124), 669 Haller, A., 204(1), 233 Hamamoto-Metzger, J. T., 708(40), 722 Hamilton, J. A,, 32(23), 44, 184(84), 201, 535(79), 578 Hamlin, K. E., 13(101), 27 Hammouda, Y., 281 (63), 283(73), 284, 285 H a m o r , T . A,, 32(23, 24), 44, 184(84), 201, 585(19, 20), 597(29), 589 Hance, P. D., 300(134, 141), 331, 704(27), 722 Handley, G. J., 160(11), 199, 288(112), 330 Handovsky, H., 17, 25 Hano, J., 280(76, 78), 285 Hansch, C. H., 495(181), 512 Hansel, A., 51 (44), 53, 251 (6), 267 Hansen, J. H., 106(82, 83), 113, 115(104), 117 Hanssen, H., 592, 621(9), 635(154), 666, 669 Hargreaves, C. C., 95(31), 116 Hargrove, W., 419(72a), 508 Harley-Mason, J., 14(109), 18(109), 25, 37(42), 45,585(21),589,644(186), 670, 766(126), 767(126), 782 Harmor, T. A., 535(79), 578 Harnack, E., 161 (25, 26, 27), 199 Harrison, J. W. E., 293(93), 330 Hartman, P. J., 9(70), 23 Hartmann, P., 759(101), 781 Hartmann, R. E., 303(158), 331 Harvey, D. G., 50(37), 53 Hasenfratz, V., 107(84, 85), 117 Hasspacher, K., 560(120), 561 (120), 579 Hawkins, J. R., 16(118), 25 Hawkins, W. L., 160(8), 162(8), 199 Haynes, H. F., 250(3), 267

Heath-Brown, B., 743(51), 745(51), 780 Heckel, E., 204(l), 233 Heering, H., 607(75), 667 Heilbron, I. M., 406(50), 507 Heim,R., l O ( 8 1 , 82, 83), l l ( 8 2 , 83), 12(83, 86, 87, 88, 89, 90, 93), 24 Hein, G. E., 44(58), 45 Heinemann, H., 193(133), 331 Heinzelmann, R. V., 13(102), 24,312(178), 332 Hellberg, H., 771(160, 161, 162, 163, 164), 783 Hellriegel, E., 607(75), 667 Hellstrom, H., 4(32, 33, 34), 5(32, 34, 38), 22 Hemmann, L., 612(92), 668 Henbest, H. B., 5(46), 23 Henderson, F. G., 95(45), 116 Hendrickson, J. B., 61(26b), 62, 63(52), 66(65), 73(65), 85(26b), 86(26b), 87(26b),90,91,587(30), 589,636(162), 638(162), 640(162), 642(162), 643(185), 650(162), 652(162), 653(162), 654(162), 656(162), 657(162), 661(162), 670, 712(55), 723 Hendriksen, T. W., 326(228), 334 Henry, J. A., 665(224), 671 Henry, T'. A., 119(1), 120(6), 123(5), 130, 131(5), 134(5), 145(1, 5), 155, 161(32), 162(32, 39), 174(32), 199, 200, 238(9), 240(9), 247, 300(140), 331, 811(78), 812, 814 Herran, J., 416(66a), 508 Herter, C. A,, 1 (6), 21 Hertle, F., 95(46), 116 Hertting, G., 326(228), 334 Herz, W., 52(47), 53 Heshmat, O., 545(94), 578, 618(108), 633 (108), 668 Hesse, A., 1(l),2 (l),21 Hesse, G., 17(129), 25 Hesse, M., 201 (92b), 201, 545(93), 547(93, 96), 556(96), 559(96), 563(127), 564(127), 565(127), 566(127), 567(127), 570(134), 578, 579 Hesse, O., 2 8 ( l ) , 44, 48(5), 52, 60, 90, 160(4,5, 6), 161 (21,22, 23, 24, 28, 31), 174, 199, 240(14), 241(14), 247, 456(59), 508, 681 (15), 690 Hester, J. B., Jr., 716(70), 723 Hetzel, H., 8(63), 23

AUTHOR INDEX

Hill, R. K., 31(20), 37, 41, 44, 308(168), 332, 6 9 6 ( l l ) , 698(11), 705(30), 722 Hiltebrand, H., 545(93), 547(93, 96), 555(116), 556(96, 116), 559(96), 563(127), 564(127), 565(116, 127), 566(127), 567(127), 578, 579 Hino, T., 33(31, 32b), 37, 45, 586(28), 588 (33, 34), 589 Hirt, R., 469(138), 470(138), 510 Hochstein, F. A., 9(77), 10(77), 24, 48(13), 49(13), 52, 160(18), 199, 266(38, 39), 267(39), 268, 288(114), 293(21), 294(21), 327, 330, 400(41), 450(41), 476(158), 477(157, 158), 480(158), 498(158, 186d), 499(186d), 502(186d), 507, 509, 511, 512 Hodson,H.F.,33(28,29,32a),45, 175(73), 176(76), 177(76,81), 180(76), 183(81), 184(81), 186(86), 187(86), lSS(S6), 189(86), 200, 201, 485(173), 491(173, 178), 511, 512, 518(15), 519(15), 520(15, 31), 521(15), 526(15),533(15), 535(80), 539(15), 540(15, 31, 84), 541(31), 542(31), 543(31), 547(31), 554(114), 555(15, 31, 84, 113), 556(113), 562(126), 567(15, 114), 574(15), 575(15), 576, 5 7 8 , 5 7 9 , 5 8 2 ( 7 , 8), 586(7, 8, 25, 27), 589, 788(28), 805(28), 812 Hoffmann, C., 325(211), 333, 701(18), 722 HoffmannovB, J., 276(47), 284 Hofmann,A., lO(81, 82, 83), l l ( 8 2 , 83, 84, 85), 12(83, 90, 93a), 18(131), 24, 25, 293(2, 38, 64, 71, 72, 87, 99, loo), 311(177), 323(199), 324(199), 327, 328, 329, 330, 332, 333, 534(69), 577, 727(10), 734(19), 735(24, 25, 40), 736(19), 737(19, 42, 43), 738(19), 740(40, 47), 746(25, 66), 747(68, 70, 71), 748(25, 66, 81), 749(10, 85, 169), 751(68, 89), 752(90), 753(89, 92), 755(173), 756(173, 174), 758(96, 97), 759, 760(70, 71, 103), 763(108), 764(71, 108, 112), 765(71, 108, 112), 766(108), 767(134), 768(66, 81, 138, 142), 769(92, 149), 770(153, 155, 157), 772(66, 167, 169, 171), 779, 780, 781, 782, 783, 798(46, 47, 5!), 804(62), 805(62), 813, 814 Hohl, J., 160(19a), 199, 520(26), 528, 529(26), 576

827

Holden, C. L., 204(6), 205(6), 223(6), 224(6), 233 Holden, R. T., 641 (174), 670 Holker, J. S. E., 288(114), 330, 400(41), 405(34), 450(41), 507 Holmes, H. L., 33(27), 45, 596(24), 597(123), 599(38), 612(87), 618(87), 623 (123), 666, 667, 668, 669 Holscher, F., 649(193, 194), 653(193), 656(193, 194), 657(193), 660(207), 662(193, 194, 210), 670, 671 Holt, R. J. W., 51 (45), 53 Holubek, J., 259(20), 267, 276(35), 283, 495(184), 512 Hooper, D., 59(1), 89 Hopkins, F. G . , 50(32), 53 Homer, L., 647(188), 660(188), 663(188, 211), 670, 671, 796(44), 813 Homing, E. C., 9(75), 10(75), 16(75), 19(75), 24, 477(156), 505(156), 511, 812 Hosanky, N., 710(49), 715(49), 723 Hosansky, N., 293(93), 327, 328, 330 Hoshino, T., 3(15), 9(67), 18(130), 22, 23, 25, 35, 45 Hou, H. C., 95(50, 51), 116 Hovden, R. A., 13(106), 25 Howard, B. F., 238(5), 240(14), 241(14), 246(5), 246, 247 Hsing, C. Y., 5(43), 22 Hsu, I., 209(28), 233, 303(151), 304(151, 160, 162), 305(162), 310(172), 312(151), 331, 332 Hsu, I. H., 502(189), 503(189, 192g), 512, 513, 712(58), 723 Huang, Y. T., 95(21), 96(21), 115 Huebner, C. F., 110(99), 117, 293(89), 296(125), 297(131), 300(125, 134, 136, 142), 303(161), 304(161), 305(152), 308(161), 315(152), 3 2 9 , 3 3 0 , 3 3 1 , 3 3 2 , 704(27), 722 Hiirlimann, H., 233(75), 235 Hug, E., 3(21), 22 Hughes, G. K., 250(4), 267, 588(31), 589, 663(213), 664(213), 671 Hughes, N. A., 498(186f), 512, 533(58), 577, 679(1), 680(1, 6, 7), 684(6, 7), 685(7), 688(23), 690, 691 Huisgen, R., 592, 638(164), 641(175), 647(175), 650(197, 200), 651(197, 201, 202), 652(201), 653(201. 202),

828

AUTHOR INDEX

654(201, 202), 656(202), 657(201), 662(197), 663(175, 200), 670, 671 Hundeshagen, H., 327(230), 334 Hunger, A., 599(39), 635(156), 642(156), 644(156), 667, 670 Hunzicker, F., 469(138, 139), 470(138), 474(139, 140), 476(139), 477(139), 480(139), 510

37a), 131(42), 133(42, 43), 134(43), 135(48), 136(48), 137 (48), 139(51a), 147(15, 48, 51a), 148(15, 51a), 149(15, 51a), 151(51a), 152(51a), 153(51a), 156, 157(57), 157, 160(19), 191(92), 192(92), 194(92), 199, 201, 204(2, 8), 205(5, 7), 206(21), 207(24, 26), 209(26), 211 (5), 213(36), 215(37, 39), 216(39), 225(58), 227(8, 51, 53, 56), 230(56, 62), 231(39, 56), 232(68), I 233, 234, 235, 237(6), 238(6, 8, lo), Iacobucci, G., 327,328,474(153), 477(157), 240(6), 246(6), 246, 247, 258(18), 511 260(lS), 261(18), 262(18, 25), 267, Ikan, R., 175(82), 178(82), 179(82), 270(26, 54), 271(27, 29), 276(40), 180(82), 182(82), 187(82), 190(82),201 277(49), 280(76),281 (52,53,54, 56,58, Ikeda, T., 61(27, 30), 64(58), 70(30), 59, 62, 63, 64, 66), 282(73, 74, 75, 83), 71(30, 58, 69, 72), 74(27), 90, 91 283, 284, 285, 293(100), 328, 329, 330, Imoudsky, G., 600(47), 667 337(3, 6, 7), 390(6, 7, 32), 393(6), Inagaki, I., 160(18a), 199 405(6, 74), 416(3, 7, 64, 65), 418(7, Inamoto, N., 252(10), 267 64, 68, 69), 419(3, 6, 74), 420(32, 75, Ing, H. R., 59(6), 62(6), 89 76), 456(115a), 457(115a), 465(68, 69, Ingle, J. D., 1(7), 21 115a), 466(68, 69), 467(126), 484(170), Ingleby, R. F. J., 585(21), 589 490(165a), 491(174, 176, 179, 184b), Iorio,M. A,, 195(94),201,519(22), 520(25), 494 (179c), 495 (170), 497 (170), 521(38), 532(25, 44,45), 575(44), 576, 498(186e),505,506,508,510,511,512, 577 531(55, 56), 533(57, 59), 534(69), Irie, T., 56(9), 58 535(74), 537(74), 577, 578, 665(223), Ishidate, M., 327, 328 671, 680(5, 8), 681, 682(S, 12, 18), Ishikawa, M., 238(13), 243(13), 244(29, 683(3), 684(4,12), 685(4, 12), 686(19), 31), 247, 711(53), 718(71, 72), 723, 688(21, 24), 691, 696(14), 699(14), 793(43), 813 704(14), 706(14), 707(35), 713(63), Iyer, R. S., 308(167), 332, 788(7a), 812 714(63), 719(74), 720(76), 721(76), 722, 723, 788(11), 798(52), 799(52), J 804(62), 805(11, 62, 63), 812, 813, 814 Jackman,L.M., 85(84),86(84),87(84),91, Jaret, R., 327, 710(49), 715(49), 723 Jeffery, G. A., 211(31), 233 547 (96), 556(96), 559(96), 578 Jelyazkoff, D. K., 282(85), 285 Jackson, A. H., 14(109), 18(109), 25, 31(18), 32(18, 25), 34(18), 35(18), Jennen, R. G., 601(55), 650(199), 659(199), 661 (208), 667, 671 37(42), 44, 45 Jensen, K. A., 108(93), 117 Jackson, R. W., 3(16), 22 Jenssen, H., 17 (122), 25, 49(24), 52 Jacob, .A,. 582(3), 588 Jacobs, W. A,, 729,734(20,21), 735(21,33, Jilek, J. O., 318(186, 187, 188), 319(190), 321(190,193),322(190,193), 324(203), 34, 35, 36, 37, 38, 39), 742, 744(55), 748(80), 749(82, 83, 84, 86), 779, 780, 333 791 Jobst, J., 28(1), 44, 161(22), 199 Jaegers, K., 622(121), 636(121), 669 Johnson, I. S., 219(42), 225(42), 234, Jaminet, F., 665(219, 220, 221), 671 270(6), 271(6), 272(6), 273(6), 275(6), Janot, M. M., 94, 96(63), 97(66), 98(71), 283, 337(5), 419(5), 506 99(66), lOO(66, 71), 101(71), 107(86), Johnson,L.F., 337(9), 367(9,29), 389(29), lOS(SS), 112, 113(19), 115, 116, 117, 391(9),396(9),401(9),404(9),411(29), 120(15, 16, 17, 18, 19, 22), 128(36, 37, 506

AUTHOR INDEX

Johnson, N. M., 9(75), 10(75), 16(75), 19(75), 24 Johnson, W. S., 313(181), 332, 559(117), 5 79 Jolly, J . , 316(184,185), 318(185),321 (185), 322(185), 323(185), 324(185), 332, 333 Joly, R., 9(70b), 23, 316(184, 185), 318(185), 321(185), 322(185), 323(185), 324(185), 332, 333 Jones,E. R. H., 5(46), 23, 406(50), 507 Jones, G., 97(69), 98(69), 99(69), 117 Jones, R. A. Y., 708(45), 723 Jones, R. G., 735(26), 742(50), 745(26,50), 755(50), 780 Jones, W. J., 305(163), 332 Joshi, B. S., 63(53), 64(53), 90, 201 Joule, J. A., 136(49), 138(49, 50), 139(49, 50), 141(49),142(49, 50), 144, 145(50), 146,147(49,50), 151 (50), 157,281 (55), 284, 434(48), 445(48), 457 (48), 462(48), 469(48), 473(143a, 143b), 474(48), 477(48), 482(48), 491(184a), 5 0 7 , 5 1 0 , 5 1 1 , 5 1 2 , 6 3 0 ( 1 4 5 ) ,632(145), 669 Jucker, E., 768(142), 782 Julian, P. L., 35, 36, 41, 45 Jurd, L., 403(46), 507 Justoni, R., 15, 25

K Kaehrn, H., 636(159), 670 Kaiser, F., 293(51, 68, 208), 295(124), 328, 329, 330 Kaisin, M., 225(58), 234, 491(179), 512, 535(76), 578, 788(30), 803(30, 64), 813,814 Kakac, B., 318(186, 187, 188, 189), 322 (189, 194), 333 Kalberer, F., 12(93a), 24 Kaldor, A., 280(81), 285 Kamat, V. N., 271 ( 8 ) , 272(8), 283 Kametani, T., 9(70a), 13(70a), 23 Kandel, S. I., 766( 125), 767 (135), 782 Kane, G., 293(8), 327 Kaneko, H., 533(62), 577 Kaneko, K., 689(28), 691 Kao, Y. S., 95(21), 96(21), 115 Kapur, R. D., 811(74), 814 Karrer, P., 48(21), 52, 126(33,34), 128(38), 156, 164(50, 51), 200, 201 (92b), 201,

829

400(44), 411 (44), 418(67), 463(44), 466(67, 124, 125, 137), 507, 508, 510, 517(6), 518(16), 519(16, 19), 520(16, 23, 24, 27, 28, 30, 32, 33, 34, 35), 521(16, 23, 24, 30, 32, 33, 34, 35, 36, 37, 39, 40, 41, 42), 522, 523(23, 30, 35, 42), 524(42, 49), 525(49), 526(49), 527(34),528(40), 529(54),531 (40,54), 532(40), 533(40), 534(27, 28, 30), 537(82), 539(37, 39, 82), 540(24, 30, 35, 39, 82), 541(24, 85, 86), 542(86, 871, 543(85, 861, 545(16, 93, 94), 546(85, 95), 547(85, 93, 97), 548(33, 34, 35, 85, 98, 102), 549(102), 551 (102), 552 (108, log), 553 (110, 111),555(16,24,30,116),556(96,116), 559(96),560(32), 561 (35,82,121,122), 562(35, 112, 124, 125), 563(124, 127), 564(110, 111, 127), 565(116, 127), 566(127), 567(16, 32, 127), 568(128, 130),569(131,132), 570(131, 132, 134), 571(132), 573(16, 35, 132), 574(16, 23, 35, 36, 39, 40, 136), 575(16, 30, 34, 35, 36, 40, 41, 136, 137), 576, 577, 578, 579, 599(37a), 618(107, 108), 619(107, IlO), 628(135, 136), 632(37a, 146a), 633(108), 667, 668, 669, 670, 6R1(16), 682(17), 690, 804(14), 812 Kashiwaki, K., 56(2), 57(2), 58 Kataoka, H., 244(30), 247 Katchdski, E., 44(61), 45 Kates, M., 96(64, 65), 116 Kathriner, A., 616(105), 623(105), 668 Kato, T., 160(18a),199 Katritzky, A. R., 708 (45), 723 Katz, L., 495(181), 512 Katz, T. J., 585(21), 589, 632(147), 669, 716(69), 723 Kauffman, A., 3(21), 22 Kaul, R., 2(12), 22 Kavkova, K., 261 (22, 24), 267, 276(36, 44, 48), 277(36, 46, 48), 281(36), 284 Kawanishi, M., 325(215), 333, 701 (19), 722 Kawazoe, Y., 543(92), 578 Kaziro, K., 542(88), 578,605(68),664(217), 667, 671 Kebrle, J . , 518(16),519(16), 520(16,23,33, 35), 521(16, 23, 33, 35, 42), 523(23,

830

AUTHOR INDEX

35, 42), 524(42), 540(35), 545(16), 548(33,35),555(16),561 (35),562(35), 567(16), 573(16, 35), 574(16, 23, 35), 575(16, 35), 576, 577 Keck, J., 327 Keilich, G . , 766(128), 767(128), 782 Keller, C. C . , 735(27), 780 Keller,F., 293(51,109),296(132),297(132), 328, 329, 330, 331 Kendall, E. C., 308(169), 332 Keogh, P., 162(42),200 Kerigan, A., 220(42c), 234, 273(14), 283 Kermack, W. O., 49(27), 52 Kharasch, M. S., 748(76), 781 Khashakyan, L. V., 56(11a), 58 Kiang, A. K., 265(34), 268, 328, 444(108), 498(186b), 509, 512, 798(49), S00(49), 81 3 Kidd, D. A. A., 328, 329, 708(39), 722 Kiefer, B., 325(213), 333 Kierstead, R. W., 300(139), 316(183), 331, 332 Kiesel, R. J., 209(28), 233, 296(126), 303(151), 304(151, 160), 311(126), 312(151), 330, 331, 332 Kikumoto, R., 584(18), 589 Kilzer, J., 533(64), 577, 689(29), 691 Kim, K. A., 708(41), 722 Kimmig, J., 650(196), 651(196), 656(196), 6M (196), 670 King, F. E., 30(17), 31(19), 33, 34(19, 35), 35(39), 44, 45, 61(37), 80(37), 82(37), 90, 266(45), 268 King, H., 516(3), 517, 518(14), 520(14), 521(14), 526(14), 528(50), 539, 555(14), 560(14), 567(14), 574(14), 575(14), 576, 577 King, J. A., 293(8), 327 King, T. J., 61(37), 80(37), 82(37), 90, 266(45), 268 Kirberger, E., 13(100c),24 Kirby,K. S., 238(9), 240(9, 15, 16),241(15, 16, 22, 23, 24), 243(22, 23, 24), 247 Kirkpatrick, J. L., 503(192d), 504( 192d, f),

512 Kirkwood, S., 4(35), 22 Kline,G. B., 702(24),722,735(26),742(50), 745(26, 50), 755(50), 780 Klohs, M. W., 293(51, log), 296(132), 297 (132), 328, 329, 330, 331 Klyne, W., 302(147), 331

Knapp, K. H., 795(44), 813 Knowles, R. N., 543(92), 578 Kny, H., 337(13), 358(13), 359(17), 506 Kobayashi, T., 34, 35, 45, 584(18), 589 Kobel, H., lO(81, 83), 11(83), 12(83, 90, 93a), 24, 763(108), 764(108, 112), 765(108, 112, 119), 767(134), 782 Kocbr, M., 623(128), 626(128), 629(141), 633 (141), 635(155), 646(155), 653 (155), 669 Koepfli, J. B., 327 Kogure, A., 623(129), 633(149, 150), 669 Koll, M., 680(2), 681(2, 9, lo), 690 Kolosov, M. M., 34(36), 36, 45 Komatsu, H., 757(95), 781 KompiB, I., 257(15, 16), 260(16), 261(15, 16), 267, 276(42, 43, 87), 277(39, 86), ZSO(S6, 87), 283, 284, 285, 337 (lOa, b), 405(78), 420(73), 495(78), 506, 508 Kondo, H., 60(15), 61(27, 29, 30, 31), 64(59, 60), 65(63), 70(30, 31, 59), 71(30, 59, 70, 71, 73, 74), 75(27, 74), 76(74), 89, 90, 91 Konuiszy, F., 3(22, 23, 25), 22 Konz, W., 17(121), 18(121),25,517(10,11, 12, 13), 520(10, 11, 12, 13), 576 Kornblum, N., 305(163), 332 Kornfeld, E. C., 735(26), 742(40), 745(26, 50), 755(50), 780 Korol, B., 266(38), 268 Korsby, G., 303(157), 331 Korzun, B., 120(22a), 156, 262(27), 264(27), 267, ZSS(lll), 295(15), 300(134, 142), 327, 330, 331, 704(27), 722 Kotake, M., 596(26), 623(126, 129), 628(137, 139), 632(137, 148), 633(149, 150), 641(148, 171, 172), 642(178), 666, 667, 670 Kowanko, N., 237(2), 246 Kozu, Y., 757(96), 759(100), 760(100), 764(100), 781 Kraft, F., 727(8, 9, 12), 733(9), 779 Kralt, T., 326(227, 228), 334 Krause, E. F., 303(158), 331 Krohnke, F., 599(42), 602(60), 619(60), 621(42, 115), 623(130), 629(115), 667, 668, 669 Kroneberg, D., 811(75), 814 Krupka, R. M., 44(58), 45

831

AUTHOR INDEX

Kuehne,M. E., 213(35),234,300(134,142), 307(164), 311(164), 331, 332, 704(27), 722 Kuhlmann, J. G., 328 Kuhn, H., 5(44), 22 Kulka, M., 474(151, 152), 498(152), 511, 516(5), 576 Kumari, T. N. S., 160(16a), 199 Kummert, E., 1(3), 21 Kump, C., 265(36), 266(35, 36), 268 Kump, W. G., 201 (92c), 201, 423(95, 102), 426(95, 102), 427(95, 101), 429(95, 102), 434(91), 435(95), 438(91, 95), 439(96), 443(101),462(119), 466(119), 495(91), 497( 186), 498( 186), 504(91, 96), 509, 510, 512 Kupchan, S. M., 559(117), 579 Kurtz, R., 175(82), 178(82), 179(82), 180 (82), 182(82), 187(82), 190(82), 201 Kusserow, G. W., 296(132), 297(132), 329, 331 Kusumoto,M., 757(95),759(100),760(100), 763(109), 764(100, 109, 114, 116), 765(109, 116), 781, 782 Kusumoto, S., 632(148), 641(148), 669 Kutney, J. P., 220(42c), 234,273(14), 283, 294(39), 328, 431(99), 509, 797(45), 798(52), 799(52), 813 Kuzovkov, A. D., 48(11, 19), 49{11, 19), 52

L La Barre, J., 204(9), 227(9, 54, 57), 232(9, 54, 68, 69), 233, 234, 235, 811(77), 814 Labriola, R., 3(21, 27), 22, 507 Laidler, K. J., 44(58), 46 Lal, J. B., 61 (38),80(38), 90 Lambert, B. F., 800(61), 801(61),814 Landesman, H., 365(27b), 506 Landrin, E., 204(1), 206(1), 232(71), 233, 235 Lapibe, C., 3(28), 22 Lardy, H., 44(58, 59), 45 Larrieu, P., 61(39), 90 Lasslo, A,, 326(220),334 Lawson, W., 52(46), 53 Lawton, R. G., 642(184), 670

,,

Le Count, D. J., 266(", %68,423(95, 102), 426(95, 102), 427(95), 429(95, 102), 434(92), 435(95), 438(92,95), 509 Lederer, E., 49(30), 50(36),53,584(14),589 Lederer, M., 521 (38, 43), 523(43), 577 Lee, C. M., 61(26c), 85(26c), 86(26c), 87 (26c), 89 (26c), 90 Lee, C. B., 95(23), 115 Lee, D. H. K., 271(8), 272(8), 283 Lee, H. M., 95(55), 116 Lee, s. T., 95(23), 115 Lee, T.B., 164(62),zoo Leemanb, H. G., 739(46), 752(90), 780, 781 Leehe, E., 2(11), 6(54, 56), 22, 23, 231(66), 235,237(2), 246,533(67), 577,713(61), 723,786(7), 790(7), 812 Le Gall, J. P., 293 (97), 330, 789 (33), 813 Legault, R. R., 748(76), 781 Legler, G., 9(73a), 16(73a), 23 Le Hir, A., 128(37), 156,215(39), 216(39), 231(39), 234, 293(100), 328, 329, 330, 418(68), 465(68), 466(68), 498(186e), 508, 512, 533(57, 59), 577, 680(5, 8), 611 ( l l ) ,682(8),683 (3),684(4),685(4), 686(11, 19), 688(24),690, 691, 696(13, 14), 698(13), 699(13, 14), 704(14), 705(13), 706(14), 713(63), 714(63), 722, 723 Lehner,R,400(23,43),476(155),480(155), 506,507, 511 Leicht, C., 66(64), 87(64), 68(64j, 72(64), 73(64), 80(64), 91, 503(190, 191), 512 Leicht, C. L., 113, 115(104), 117, 122(29), 156, 169(61), 200, 709(46), 710(46), 711(46), 712(59), 723 LeMen, J., 120(15, 16, 17, 19, 22), 128(36, 37), 131(42), 133(42, 43), 134(43), 135(48), 136(48), 137(48), 139(51a), 147(15, 48, 51a), 148(15, 51a, 56), 149(15, 51a), 151(51a), 152(51a), 153(51a), 156, 157 (57), 157, 160(19), 191(92), 192(92), 194(92), 199, 201, 225(58), 234, 255(47), 258(18), 260(18), 261(18), 262(18, 26), 267, 268, 270(26, 54), 271(27,29), 276(40), 277(49), 280(76), 281(52, 53, 56, 56, 58, 59, 62, 63, 64, 66), 282(73, 74, 75, 83), 283,284,285,293(98), 330, 337(3, 7), 390(7, 32), 405(74), 416(3, 7, 64, 65), 418(7, 64, 68, 69), 419(3, 74), 420 (32, 76), 456 (115a), 457 ( m a ) ,

832

AUTHOR INDEX

463(121, 122), 465(68, 69, 115a, 121, 122), 466(68, 69, 121, 122), 467(126), 484(170), 485(121, 122), 490(165a), 491(174, 176, 179a, 184b), 494(179c), 495(170), 497(170), 505, 506,508, 510, 511, 512, 534(69), 535(74), 537(74), 578, 680(5, 8), 682(8, 18), 686(19), 688(21), 690, 691, 707(35), 722, 788 (1l), 789 (33), 800 (60), 804(62), 805(11, 62, 63), 812, 814 LempBriBre, T., 12(92), 24 Leonard, N. J., 163(49), 200 L&opold,A. C . , 13(96), 24 Lequime, J., 227(57), 234 Leroux, 60(7), 89 Lettenbauer, G., 297(95, 128, 129), 313(128, 180), 328, 330, 331, 332 Leuchs, H., 593(10), 594(11, 12, 13, 14), 596(22, 25), 597(31), 599(40, 41, 42, 43), 601(50,51,52, 54), 602(41, 51,52, 56, 58, 59, 60), 604(14, 52, 64, 65), 605(64), 606(70, 71), 607(64, 72, 74, 75), 611(85), 612(88, 89, 90, 91, 92), 613(94), 615(56, 58, 59, 103, 104), 618(56), 619(60), 621(42, 115, 116), 622(22, 117, 118, 119, 120, 121, 122), 623(130), 625(131), 626(133, 134), 627(13, 74), 628(138), 629(115, 140), 630(143a, 146), 634(116, 151, 152), 635(103), 636(31, 43, 121, 143a, 159, 160, 161), 637(134, 163), 640(72), 666, 667, 668, 669, 670, 791 (37), 813 Leuchs, W. R., 615(103), 635(103), 668 Le Vine, P. B., 326(224), 334 LCvy, J., 120(15), 128(36, 37), 131(42), 133(42,43), 134(43), 135(48),136(48), 137(48), 139(51rt), 147(15, 48, 51a), 148(15, 51a, 56), 149(15, 51a), 151(51a), 152(51a), 153(51a), 156, 157(57), 157, 225(58), 234, 255(47), 268, 281(59), 284, 418(68, 69), 456(115a), 457(115a), 465(68, 69, 115a), 466(68, 69), 490(165a), 491(165, 174, 179a, 184b), 508, 510, 511, 512, 535(74), 535(74), 578, 680(8), 682(S, 18), 688(21), 690, 691, 788(11), 805(11), 812 Levy, P. R., 586(24), 589 Lewi, P., 610(82), 668 Lewis, G. P., 16(114), 25 Lewis, R. G., 533(65), 577, 690(30), 691

Liebrecht, I., 294(122), 330 Liebrecht, J., 294(122), 330 Liljegren, D. R., 110, 117, 701(21, 22), 722 Lindlar, H., 327(229), 334 Lindner, A., 326(228), 334 Linstead, R. P., 31(21), 44 Liquori, N., 33(33), 45 List, P. H., 8(63), 23 Liu, L. H., 300(134), 331, 696(9), 698(9), 704(27), 722 Lodish, H. F., 265(42), 268, 423(82), 426(82), 434(82),435(82),438(82),508 Loebisch, W. F., 600(45),641 (169),667,670 Loewi, O., 43(52), 45 Lofgren, N., 4(33, 34), 5(34), 22 Logemann, W., 325(212), 333 Lorck, H., 303(157), 331 Loudon, J. D., 62(51), 64(55, 56), 65(56), 76(56), 85(56), 90 Louis, H. L., 634(152), 669 Louveau, G. 1 (5), 21 Lovell, F. M., 102(78), 117, 300(137), 331 Lovenberg, W., 8(64a), 23 Lucas, R. A,, 209(28), 213(35), 225(45), 233, 234, 293(76, 89), 296(125, 126), 300(125), 303(151), 304(151,160,162), 305(162), 307(164), 311(126, 164,174, 175), 312(151, 179), 329, 330, 331, 332 Ludewig, H., 51(42), 53 Ludwig, W., 796(44), 813 Lutomski, J., 48(17), 50(17), 52 Lyapunova, P. M., 277(50), 284, 288(116), 330 Lynch, G. R., 95(32), 116 Lynn, E. V., 48(15), 52 Lyttle, D. A., 5(52), 23

M McArthur, M., 232(74), 235 Macek, K., 765(117), 782 McCloxkey, P., 62(51), 90 McCurdy, 0. L., 172(67), 200 MacFie, J. W. S., 120(6),155, 162(39),200, 811(78), 814 Maciak, G., 271(12), 272(12), 283 McIntyre, A. R., 515(1), 528(1), 576 Macko, E., 288(115), 330 McLafferty, F. W., 416(54), 507 McLamore, W. M., 108(94), 109(95), 117

AUTHOR INDEX

McLean, S., 367(45), 396(30, 42), 400(30), 401(30, 35), 403(30, 42), 416(54a), 445(42), 506, 507 McLoughlin, B. J., 767 (133), 782 McMurray, W., 390(28b), 506 Mcl'hail, A. T., 491 (178), 512, 535(80), 5 6 o ( i i 8 ) , 578, 579, 788(2~),805(28), 812 MacPhillamy, H. R., 209(28), 213(35), 225(45), 233, 234, 250(5), 296(125), 300(125, 134), 303(151, 161), 304(151, 160, 161), 307(164), 308(161), 310(173), 311(164), 312(151, 179), 330, 331, 332, 704(27), 722 Madinaveitia, J., 8(58, 59), 23, 582(3), 583 (12), 588, 589 Magalhiies Alves, H . , 523(44), 575(44), 577 Magalhiies Alves, H . , 195(94), 201 Mahan, J. F., 97(70), 117 Mainil, J., 48(16), 52 Maj, J., 280(76, 78), 285 Majima, R., 9(67), 23 Major, R. T., 3(24), 22 Majumdar, S. G., 177(77, 78), 178(77, 78), 200 Malamud, W., 194(119), 330 Malesh, W., 209, 329, 330 Malfatti, H . , 641 ( l e g ) , 670 Malikov, V. M., 282(71), 284 Malone, M. H . , 10(83a), 13(83a), 24 Manas-Santos, F., 161 ( 3 7 ) , 195(37), 196(37), 200 Manda, K. L., 270(2l), 283 Manh, D. D., 258(18), 260(18), 261(18), 262 (18), 267, 276 (40), 284, 484( 170), 495(170), 497(170), 511 Mann,.M. J.,735(26), 742(50), 745(26, 50), 755(50), 780 Manning, R. E., 266(39), 267(39), 268, 419(72a), 498(186d), 499(186d), 502 (186d), 508, 512 Manohar, H., 32(23), 44, 184(85), 201 Manske, R. H. F., 19(132), 25, 33(27), 45, 49(23), 50(35), 52, 53, 56(8), 58, 474(151, 152), 498(152), 511, 582(4, 5), 583 (13), 585, 588, 589 Manzur-I-Khuda, M., 708(37), 722 Maranon, J., 3(19), 22 Marini-Bettolo, G. B., 195(94), 201, 474(145, 149), 476(149), 498(149), 511, 516(2), 517(7), 519(7, 22), 520(7,

833

2 5 ) , 521 (7, 383 43), 523(25, 43, 44, 45), 548(7), 575(44), 576, 577 Marion, L., 4(35), 6(53, 54, 55), 8(57c), 22, 23, 49(23), 52, 61(28), 64(57), 66(57), 67(28), 70(28), 73(28), 76(28, 7 8 ) , 77(78), 78(28), 90, 91, 94, 96(57, 64, 65), 97(67, 68), 98(72), 99(74), 100(75), 101(75), 108(8S), 110(18), 111(102), 112 (18), 113 (18), 114 (18), 115(18), 115, 116, 117, 396(30, 42), 400 (30), 401 (30), 403 (30,42), 445 (42), 506, 507, 509, 582(5), 583(13), 585, 588, 589 Markess, D. G., 65(62), 91 Narrian, S. F., 94, 96(16), 97(16), 107(16), 110(16), 115 Marsh, D. F., 5k7(6), 526(6), 576 Marshall, P. B., 13(100b), 24 Martin, R. H . , 225(58), 2,26(46), 234, 491 (179), 512, 535(76), 578, 788(30), 806 (30, 64), 813, 814 Martin, S., 3 (14), 22 Martin, W. F., 665(224), 671 Maschmann, L., 95(40), 116 Mason, S. F., 58.5(22), 589 Massagetov, P. S., 13(96a), 24, 48(9, 11, 19), 49(11, lo), 52, 195(93), 201 Massy-U'estropp, R. X., 533(65), 577, 690(30), 691 Massicot, J., 8(57c), 23 Massion, L., 61 (44), 90 Masuda, S., 252(10), 267 Mathes, K., 43(53), 45 Mathieu,,J., 316(185), 318(185), 321 (185), 322(185), 323(185), 324(185), 333 Mathys, F., 207(24, 26), 209(26), 233 Matteson, D. S., 40(48), 45 Mattox, V. R., 308 (leg), 332 Maxfield, R. C . , 204(6), 205(6), 209(30), 223(6), 224(6), 233 Maximowa, T., 4(36), 5(36), 22 Maxwell, R. A., 311 (174, 175), 332 Mayo,D. \V.,476(168), 477(158),480(158), 498(158), 511 Mayeda, S., 56(3), 57(3), 58 Mazzocco, P., 3(21), 22 Meade, E. M., 31 (21), 44 Medina, J . D., 451(113f), 509 Meier, H. L., 302(148), 331 Melera, A., 545(93), 547(93), 578 Meli, A,, 325(212), 333

834

AUTHOR INDEX

Mendive, J. R., 17(123), 25 Mengelberg, M., 612(91, 92), 630(146), 668, 669, 791 (37), 813 Menon,K. N., 594(17), 596(20, 21), 666 Menshikov, G. P., 13(96a), 24, 48(10), 52 Merck, F., 28(8), 44 Mergenthaler, E., 651 (201), 652(201), 653(201), 654(201), 657(201), 671 Merlis, S., 17(119), 25 Merz, H., 17(126), 25, 294(122), 330, 518(17), 519(17), 520(17), 521(17), 522(17), 560(17, 191), 574(17), 575(17), 576, 579 Metreveli, L. I., 34(36), 36(36), 45 Metzger, J. T. H., 512 Meyer, H., 466(137), 510, 520(34), 521(34, 36), 527(34), 548(34, 102), 549(192), 551(102), 573(135), 574(36, 136), 575(34,36,136),576,579,681(16), 690 Mezey, K., 160(14), 199 Michael, M., 434(87), 435(89), 438(89), 439(87),440(89),444(89), 508 Michiels, L., 60(7, 8, 9), 89 Middleton, P. M., 294(119), 330 Mikeska, L. A., 241 (25), 247 Mildbrand, H., 615(103), 635(103), 668 Millat, L., 59(3), 60(23), 61(3, 23, 46), 75(76), 89, 90, 91 Miller, F. M., 326(225), 333, 334 Mills, J. F. D., 337(11), 361(11, 22), 400(11), 506 Millson, M. F., 120(23), 121(23), 123(23), 134(23), 145(23,52), 146(23), 156,157 Minamidani, H., 160(18a), I99 Mineshita, T., 95(24), 96(25), 115, 116 Mitra, G. B., 704(28), 722 Mitra, S. K., 704(28), 722 Mitscher, L. A, , 302(144), 331 Mitsuwa, T., 623(126), 628(137, 139), 632(137), 669 Mittasch, H., 17(121, 129), 18(121), 25 Miwa, T., 641(172), 670 Moed, H. D., 326(227, 228), 334 Mohrbacher, R. J., 56(6), 57(6), 58 Moir, C., 748(74), 781 Mokri, I., 282(72), 285 MokrJi, J.,257(15, 16),260(16),261(15,16, 23), 267, 276(42, 43, 45, 87), 277(39, 45, 86), 280 (86, 87), 283, 284, 337 (lOa, lob), 405(78), 420(73, 74a), 495(78), 506, 508

Mollov, N., 282(72), 285 Monseur, X., 175(71), 200, 328, 463(121, 122), 465(121, 122), 466(121, 122), 485(121, 122), 510 Montecchio, G., 61 (36), 90 Monteiro, H. J., 399(37), 432(37), 435(37), 473(37), 485(37), 505(37), 507 Mookerjee, A., 293(73), 329 Mooney, R. A., 450(113a), 509, 673(4), 674(4), 678 Moore, B., 177(81), 183(81), 184(81), 186 (86), 187 (86), 188(86), 189 (86), 201 Moore, R. P., 288(117), 330, 477(161), 480(161), 482(161), 503(192d), 504(192d), 505(161), 511, 512 Moore, C. W., 94(8), 96(8, 56), 97(56), 98(56), 115, 116 Moore, M., 293(62), 329 Moore, R. E., 535(78), 578, 681, 687(14), 688(14), 690, 788(20), 805(20), 812 Moreau, J., 13(98), 24 Morris, M. J., 304(160), 332 Morrison, D. E., 735(26), 745(26), 780 Mors, W. B., 48(14), 52, 270(16), 283, 534(70), 577, 812 Mortari, A,, 311(176), 332 Moss, J. B., 169(63,64), 200, 503(192), 512, 708(43, 44), 710(44), 716(44), 722 Mothes, K., 2(10), 21, 766(121, 122, 124), 767(121, 131), 782 Motzel, W., 17(126), 25 Moufang, N., 600(46), 667 Moyer, W. W., 648(192), 649(192), 656(192), 662(209), 670, 671 Moynehar,, T. M., 708(45), 723 Moyse-Mignon, H., 95(27), 116, 270(17), 283 Mom, B. K., 270(25), 271 (25), 283 Mudd, S. H., 7(57b), 8(57b), 23 Mueller, J., 294(81), 312(178), 329, 332 Mueller, J. M., 288(7), 293(62), 296(125), 297(7), 300(125), 327, 329, 330 Muench, K., 696(11), 698(11), 722 Mukerji, B., 162(43, 45), 200 Mukherji, D., 789(31), 811 (73), 813, 814 Muller, G . , 316(184, 185), 318(185), 321(185), 322(185, 195), 323(185, 195, 200, 201, 205), 324(185, 195, 200, 207, 208), 325(195), 332, 333

835

AUTHOR INDEX

Muller, O., 649(195), 653(195), 654(195), 6 70 Miinster, W., 621(114), 668 Murai, K., 293(21), 327, 512, 327 Murayama, M., 36, 45 Myrbilck, K., 44(58, 59), 45

N Nachmansohn, D., 43(55, 56), 44(56), 45 Nagarajan, K., 427(101), 443(101, I l l ) , 444(111, 113), 503(113), 509, 632 (146a), 669 Nagase, O., 705(31), 722 Nagy, M., 584(16), 589 Nagy, Z., 280(77, 78), 285 Nagyvary, J., 569(131, 132), 570(131, 132), 571 (132), 573(132), 579 Nair, C. P. N., 270(24), 283 Nair, M. D., 61(28), 67(28), 70(28), 73(28), 76(28), 78(28), 90 Naito, K., 705(31), 722 Nakagawa, Y., 191(92), 192(92), 194(92), 201, 270(32), 283, 337(7), 390(7), 416(7), 418(7), 434(48), 445(48), 457(48), 462(48), 467(125a, 126,127), 468(125a), 469(48, 125a,) 474(48), 477(48), 480(48), 482(48), 506, 507, 510 Nakajima, H., 12(91), 24 Nakatsu, K., 197(96), 2/31, 792(39), 813 Nakazaki, M., 39, 45 Natsume, M., 568(129), 570(129), 579 Nauta, W. T., 325(214), 333 Nazir, B. N., 270(21), 283 Neeman, M., 313(181), 332 Neish, A. C., 6(57), 23 Nelson, A. L., 597(30), 632(30), 666, 791(37), 813 Nelson,E. R., 250(3), 251(8,9),252(9),267 Nenitzescu, C. D., 15(112), 25 Neugebauer, H., 95(30, 34), 116 Neuss, N., 160(17), 167(59), 199, 200, 204(4, 10, 12), 216(40), 218(12, 40), 219(40, 42, 42a), 220(42a, 42b), 221(42a), 223(42a, 42b), 225(10, 42), 226(10,48),228(10), 233,234,255(46), 268, 270(6, 22, 23, 31), 271(6, 12, 22, 23, 28, 29), 272(6, 12), 273(6, 13, 15), 275(6, 13), 283, 293(83, 84, 88, 100,

104), 294(39), 327, 329, 330, 337(3, 4, 5), 390(4), 395(4), 407(4), 416(3), 419(3, 5, 72a), 431(98, 99), 444(56), 457 (55), 474 (148), 505, 506, 507, 508, 509, 510, 511, 711(51), 723, 797(45), 81 3 Newcombe, F., 227(52), 232(52), 234 Nicolas-Charles, P., 12 (92), 24 Nitschke, R., 601(54), 607(74), 627(74), 667 Nitzberg, C., 28(3), 29(11), 36, 42(3, 51), 44,45 Noble, R. L., 272(2, 3), 282, 283 Nogradi, T., 325(210), 333 Noguchi, T., 10 (78), 24 Noland, W. E., 9(70), 13(106), 23, 25 NominB, G . , 316(184, 185), 318(185), 321(185), 322(185), 323(185), 324(185), 332, 333 Nomura, K., 416(113c), 450(113b), 451(113b, c). 497(113b), 509, 676(5), 677(5), 678 Nordman, C. E., 197(96), 201, 792(39), 813 Norkina, S., 4(36), 5(36, 37), 22, 281(65), 284 Novak, L., 318(187, 188), 319(190), 321(190,193),322(190,193),323(204), 333 Nozoye, T., 60(10), 61(31), 64(10, 59, 60), 65(60, 63), 66(10), 70(31, 59), 71(59, 70, 71, 73, 74), 72(75), 75(74, 7 7 ) , 76(74), 7 7 ( 7 9 ) , 80(79), 89, 90, 91 Nugent, R. H., 293(49), 297(42, 49), 300 (42), 303(155), 323(155), 328, 331 Nussberger, G. A., 451(1131, m), 510, 673(2, 3), 674(3), 675(2, 3), 676(3), 678 Nyburg, S. C., 337(11), 36(11,22), 4 0 0 ( l l ) , 506

0 Oberlin, L., 160(3), 199 Ochiai, E., 238(13), 243(13), 244(29, 30), 247,711 (53), 718(71,72), 723,793(43), 813 O’Connell, F. D., 48(15), 52 O’Connor, J. M., 301(143), 302(143), 304(143), 331

836

AUTHOR INDEX

O’Donovan, D., 6(56), 23 Oertel, G., 647(187), 648 (187), 650(187), 651(187), 657(187), 660(187), 662(187), 670 Oesterlin, M., 600(47), 667 Ogawa, K., 33 (326), 37, 45 Ogawa, M., 160(18a), 199 Ognyanov, I., 282(72), 285 Ogui, 28(5), 44 Ohashi, M., 473(143b), 490(165a), 511 Ohta, T., 56(9, 10, ll), 57(13, 14), 58 Oishi, T., 78(80), 79(80), 91, 721(77), 723 Oka, Y., 244(29), 247 Okada, M., 327, 328 Okamoto, T., 499(186j), 512 Okanishi, T., 95(37), 116 Okany, A., 766(125), 767(135), 782 Okaya, Y., 300(137), 331 Oletta, S. A., 204(18), 233 Oliver, A. T., 271(30), 283 Oliveri-Mandala, E., 625( 132), 669 Olivier, L., 120(15),139(51a), 147(15,51a), 148(15, 51a, 56), 149(15, 51a), 151(51a), 152(51a), 153(51a), 156, 157, 255 (47), 268, 491 (179a), 512 Ollis, W. D., 599(39), 635(156), 642(156), 644(156), 667, 670 Onak, T. P., 679(1), 680(l, 6, 7), 684(6, 7), 685(7), 690 Onda, M., 325(215), 333, 701(19), 722 Ondetti, M. A., 405(147), 469(141), 474(141, 147), 505(147), 510, 511 Ongley, P. A., 60(14), 61(14), 64(14), 75(14), 85(14), 89 Openshaw, H. T., 362(27), 400(27), 462(120a), 506,510,596(28),597(123), 615(28), 623(123), 642(179, 180), 666, 669, 670 Orazi, 0. O., 390(36), 392(36), 405(34, 35, 36), 406(36), 407(35, 36), 507 Ordway, H. W., 294(81), 329 Orechoff, A., 281 (65), 284 Orekhov, A. P., 4(36), 5(36, 37), 22 Orr, J. C . , 113, 115(104), 117 Orth, H. D., 766 (128), 767 (128), 782 Osade, S., 51 (39), 53 Oshima, K., 61(29), 90 Osowiecki, M., 505(194), 513 Osterburg, F., 636(159), 670 0% H., 10(83), 11(83, 84), 12(83), 24, 311(177), 323(199), 324(199), 332,

333, 752 (go), 753 (92), 755 (173), 756(173), 769(92), 781, 783 Oudemans, A. C., 240(14), 247 Ovejero, A. F., 48(3), 52 Overberg, H. S., 611(85), 612(88, 89), 615(104), 668 Owellen, R. J., 265(43), 268, 399(37, 50a), 414(40a), 423(80, 81, 102), 426(80, 81, 102), 429(81, 102), 430(81), 432(37), 433(80), 434(48), 435(37), 445(48), 457(48), 482(48), 485(37), 505 (37), 507, 508 Oxford, A. E., 600(44), 607(44), 608(44), 612(44), 618(44), 667

P Pachter, I. J., 5(39, 40), 10(40), 16(40), 22, 56(4, 6), 57(4, 6), 58(15), 58 Packman, E. W., 293(93), 330 Page, I. H., 17(127), 25 Pai, B. R., 201, 202(98), 427(101), 441(101), 443(101), 509 Pailer, M., 277(38), 284 Pakrashi, S. C., 160(17), 199, 270(16), 283, 293(96, 103), 327, 328, 330, 431(98), 509, 534(70), 577, 798(52), 799(52), 800(59), 813 Paladini, A. C., 507 Palazzo, G., 326(223), 334 Palm, C., 160, 199 Palmer, K., 396(30, 42), 400(30), 401(30), 403(30, 42), 445(42), 506, 507 Palmer, K. H., 505(196), 513 Pan, S. C . , 303(156), 331 Pandow, M. L., 297(42), 300(42), 328 Panizzi, L., 701(17), 702(25), 705(25), 722 Panov, P., 282 (72), 285 Paradies, A. M., 9(77), 10(77), 24, 48(13), 49(13), 52 Parello, J., 714(65), 723 Paris, R., 95(27), 116, 232(74), 235, 327, 789(32), 813 Paris, R. R., 48(16), 52, 270(17), 283 Parmerter, S. M., 495(181), 512 Parratt, J. R., 13(100a), 24 Paszek, L. E., 204(6), 205(6), 209(30), 223(6), 224(6), 233 Patel, M. B., 462(119), 466(119), 510 Patel, M. D., 227(55), 234

AUTHOR INDEX

Patrick, J. B., 98(73), 100(76), 117, 362(25), 395(25), 399(25), 400(25), 414(25), 506, 523, 577 Patterson, E. S., 303(158), 331 Paul, J. K., 302(144), 331 Paul, L., 615(98), 620(112), 638(98, 112), 640(98), 668 Paul, R., 769(147), 783 Pausacker, K. H., 615(177), 634(153), 642(177),661(177),6 6 9 , 6 7 0 Pavanaram, S. K., 542(87), 578 Pavolini, T., 61 (36), 90 Pecher, J., 225(58), 226(46), 234,491(179), 512, 535(76), 578, 788(30), 806(30, 64), 813, 814 Peerdeman, A. F., 543(91), 578, 599(37), 667 Peeters, J., 225(58), 226(46), 234, 806(64), 814 Peirce, G., 594(13), 627 (13), 666 Pelletier, 592, 666 Pelz, K., 319(191), 321(191), 325(197), 333 Penna, A,, 523(45), 577 Pennella, P., 747(69), 781 Pepinsky, R., 102(78), 117, 300(137), 331 Percheron, F., 48(16), 52, 205(5), 211(5, 32), 213(36), 215(39), 216(32, 39), 217(32), 227(56), 230(32, 56, 62), 231 (32, 39, 56), 233, 234 Pereira, N. A., 83(83), 91, 134(44), 157, 262(26), 267, 482(165), 484(165), 485(165), 490(165), 491(165), 511, 535(77), 578, 694(1), 696(1), 707(1), 721, 788(21), 805(21), 808(21), 812 Perezamador, M. C., 416(66a), 508 Perkin, W. H., Jr., 49(26, 27), 50(26, 35), 52(46),52,53, 594(15, 16, 17), 598(33), 600(44), 606(33, 69), 607(33, 44, 76), 608(44, 76), 610(33, 76, 81), 612(33, 44), 615(100, 101), 618(44, 76), 619(109), 627(69), 628(69, 76, 81), 641 (76), 665(33), 666, 667, 668 Pernet, R., 328 Perrot, E., 61 (39, 45, 46), 90 Perry, F., 238(5), 246(5), 246 Pessina, R., 15, 25 Petelot, R.,272(7), 283 Peters, H., 735(31), 780 Petit, A., 29 ( l o ) , 44 Petracek, F. J., 293(109), 330

837

Petrova, M. F., 13(96a),24 Petrzilka, T., 10(83),11(83, 84), 12(83),24, 311(177),323(r99),324(199),332,333, 734(19), 736(19), 737(19, 43), 738(19), 744(56, 57, 65), 745(56), 751(88, 89), 753(89), 768(142, 143), 780, 781, 782, 783 Peyer, J., 18(131), 25 Philippe, J., 328 Philippot, E., 665(221), 671 Philipsborn, von W., 128(38), 156 Phisalix, C., 17(120), 25 Pichon, M., 269, 282 Pichot, P., 12(92), 27 Pictet, A., 601 (53), 667 Pierson, W. G., 304(162), 305(162), 332 Pietra, S., 15(113), 25 Pikl, J., 35, 36, 41, 45 Pilet, P. E., 8(65a), 23 Pillay, P. P., 160(16a), 199, 270(19, 24), 283, 328 Pimenta, A., 520(25), 521(38), 523(25), 576, 577 Pinar, M., 389(31), 390(33), 394(33), 403(31), 416(33, 120), 462(120), 506, 507, 510, 677(6), 678 Pioch, R. P., 768, 783 Pistor, H. J., 466(133), 510, 517(11, l2), 520(11, 12), 521(12), 548(12), 568(11, 12), 576 Pittenger, 1'. S., 95(33), 116 Placeway, C., l68(60), 200, 711(52), 712(52), 723 Plant, S. G. C., 470(143), 510 Plat, M., 227(50), 234, 258(18), 260(18), 261(18), 262(18, 25), 267, 271(29), 276(40), 277(49), 283, 284, 337(3, 7), 390(7, 3 2 ) , 416(3, 7), 418(7), 419(3), 420(32, 76), 484(170), 495(170), 497(l'iO), 504(193), 505, 506,508, 511, 51 3 Platonova, T. F., 48(11, 19), 49(11, 19), 52 Plekhanova, N. V., 282(69), 284 Pletscher, A., 327(230), 334 Plieninger, H., 325(213), 333, 744(62, 63, 64), 766(128), 767(128), 780, 781, 782 Plummer, A. J., 294(121), 308(159), 311(174, 175), 312(179), 314(182), 326(159), 330, 331, 332 Podkowinska, H., 5(40a), 22

838

AUTHOR INDEX

Poisson, J., 191 (92), 192(92), 194(92), 201, 227(50, 52), 232(52), 234, 328, 329, 467(126), 504(193), 510, 513, 534(69), 577, 710(48), 713(63), 714(48, 63, 64), 716(48), 723, 798(52), 799(52), 804(62), 805(62), 813, 814 Polindexter, E. H., 48(22), 52 Polonovski, M., 30(14, 16), 32(14), 36, 42(3), 44 Polonovski, M., 28(3), 29(10, 11), 30(14, l6), 32(14), 42(51), 44, 46 Pomykacek, J., 333 Popelak, A., 293(51, 68, 108), 295(124), 297(51, 95, 128, 129), 313(128, 180), 328, 329, 330, 331, 332 Porter, G. R., 44(60), 45 Potier, P., 293(98), 330, 789(33), 813 Potts, H., 177(81), 183(81), 184(81), 201 Potts, K. T., 5(47), 23, 110, 117, 701(21, 22), 722 Pourrat, H., 416(65), 508 Pozzati, C., 326(223), 334 Prasad, K. B., 498(186g), 512, 533(60), 577, 688(25), 689(25), 691 Prein, N., 281 (65), 284 Prelog, V., 97(66), 98(71, 73), 99(66), lOO(66, 71), 101(71), 107(86), lOS(S6, 87), 116, 117, 207(24, 26), 209(26), 233, 237(6), 238(6, 10, 12), 240(6), 246(6), 246, 247, 302(148), 331, 597(29), 602(57), 604(66), 605(66), 606(57), 616(105), 623(105, 128), 626( 128), 629( 141), 633( 141), 635(155), 638(166), 642(166), 646(155), 653(155), 665(166), 666, 667, 668, 669, 719(73, 74), 723, 793 (43), 813 Preobrazhenskii, N A , , 243(26, 27, 28), 247 Preobrazhensky, N. A , , 34(36), 36(36), 45 Price, J. R., 250(3), 251(8, 9), 252(9), 267 Prins, D. A., 204(14), 211(33), 212(33), 215 (14), 216 (14), 217 (41), 225 (14), 227(14),228(59,60),229(59),233,234, 504(192e), 512 Probst, H., 20(137), 25 Protiva, M., 318(186, 187, 188), 319(190), 321(190, 193), 322(190), 323(204), 324(203), 333 Przybylska, M., l l l ( 1 0 1 , 102, 103). 117

Puisieux, F., 128(37), 156, 418(68), 465(68), 508, 680(5, 8 ) . 6 8 l ( l l ) , 682(8), 683(3), 684(4), 685(4), 686(11, 19), 690, 696(15), 699(15), 701 (15), 722

Q Quevauviller, A., 232(68, 69), 235, 280(76), 282(83), 285, 811(76), 814 Quinn, G. P., 327(230), 334 Quisuddin, M., 511

R Raab, N., 400(43), 507 Rachor, J., 505(195), 513 Rack, F., 597(31), 636(31), 666 Raffauf, R. F., 56(4), 57(4), 58, 288(115), 330, 451(113J), 509, 512, 678(7), 678 RQileanu, D., 15(112), 25 Raison, C. G., 59(6), 62(6), 89 Rajagopalan, S., 559(117), 579 Rajappa, S., 175(72, 72a, 74, 75), 176(72, 75), 177(72, 74), 180(74, 75), 181(75), 186(87), 187(75, 87), 188(87),189(87), 200,201,427(101), 431(100), 441(100, 101), 443(101), 444(101), 480(160a), 509, 511, 584(17), 589 Rajsner, M., 321 (193), 322(193), 324(203), 333 Rakshit, B., 329 Ramaseshan, S., 32 (23), 44, 184 (85), 201 Ramstad,E., 2(10), 21,767(132), 768(136), 782 Rangashari, P. N., 49(24), 52 Rao, C. V., 3(20), 22 Rao, D. S., 328 Rao, G. V., 555(113), 556(113), 579 Rao, K. V., 226(49), 234 Rao, P. S., 3(20), 22 Rao, S. B., 328 Raper, R., 641(173, 174), 670 Rapoport,H.,40(48),41(49), 45,498(186f), 512, 533(58), 535(iS), 577, 578, 679(1), 680(1, 6, 7), 681, 684(6, 7), 685(7), 687(14), 688(14, 23), 690, 691, 788(2O), 805(20), 812 Rausat, A., 227(50), 234, 504(193), 513 Ratnagiriswaran, A. N., 226(47), 234 Ratovis, R., 326(222), 334

AUTHOR INDEX

839

Rauch, H., 594(12), 666 Robb, E. W., 300(134), 331, 696(!C!:, Ravdel, G. A., 753(93, 94), 781 698(10), 704(27), 722, 788(7a), 6 i b Ray, A. B.,201, 202, 800(57), 813 Robbins, E. B., 95(53), I 1 6 Ray, L., 704(29), 722 Roberts, J. D., 313(181), 332 Raymond-Hamet, M., 19(135), 20(135), Robertson, J. H., 598, 667 21(139, 140), 25, 59(3, 5 ) , 60(13, 16, Robertson, J. M., 32(23, 24), 44, 184(84), 17, 18, 20, 21, 22, 2 5 ) , 61(3, 21, 22, 32, 201, 262(28), 264(28), 268, 482(164), 33, 34, 35, 39, 41, 42, 43, 45, 46, 47, 491(178), 521, 512, 535(79, 80), 578, 48, 49, 50), 62(5, 53), 64(53, 54), 585(19, 20), 587(29), 589, 716(66), 70(21, 32, 33), 75(54, 76), 8 5 ( 5 ) , 89, 723, 788(28), 805(28), 812 90, 95(41, 42, 47, 48,49, 52), 110(100), Robinson, R., 28(8a), 31 (22), 32(22), 116, 120(7, 8, 9, 10, 11, 12, 13, 14), 33 (as), 44, 45 121(24), 123(24), 131(14,41), 134(47), Robinson, K. W., 175(82), 178(82), 145(47),156,157,162(40,41), 191 (89), 179 (82), 180(82), 182(8 2 ) , 187 (82), 200, 201, 206(19, 20), 232(74), 190(82), 201 233(76), 233, 235, 238(7, 8), 240(7), Robinson, R., 49(26, 2 7 ) , 50(26, 35), 246(34, 35,36), 247, 253(11), 264(31), 52(46), 52, 53, 56(8), 58, 785(l), 265(41), 267, 268, 280(77), 285, 789(1, 31), 790(1), 791(1), 812, 813 498(186a), 512 Robinson, Sir Robert, 5(47), 23, 30(17), Reed, R . I., 134(45), 157, 490(165b), 511 31(19), 33(33), 34(19, 35), 35(39), 40, Rees, A. H., 743(51), 745(51), 780 44,45,96(60),11G, 119(1), 120(18,23), Rees, R., 587(30), 589 121(2, 23), 123(23, 30, 31), 124(31), Regnault, V., 592, 666 130, 134(23, 46), 135(31, 46), 145(23, Regnier, G., 325(211), 333, 701(18), 722 52), 146(23, 31, 46), 155, 156, 157, Reich, P., 601 (50), 667 182(83), 191(90), 201, 240(16), Reid, 8. L., 713(60), 723 241(16, 23, 24), 243(23, 24), 247, Reid, T. L., 64, 91 250(2), 267, 270(54), 281(54), 284, Reinecke, M. G., 679(1), 680(1), 690 293(82), 329,337(2), 418(70), 466(70), Relyveld, P., 505(196a), 507 505, 508, 510, 542(89), 578, 582(8), Remshagen, H., 570(133), 579 583 ( l o ) , 585 ( l o ) , 586(8,24), 589, 592, Renk, E., 743(52), 780 594(15, 16, 17), 595, 596(20, 21, 23, Renneberg, K. H., 759(101), 781 24, 2 8 ) , 597(123), 598(33, 34), 600(44, Renner, U., 204(14), 205(15, 17), 211(33), 49), 601 (49), 604(63), 605(49), 606(4, 212(33), 215(14), 216(14), 217(41), 33, 69), 607 (33, 44, 63, 73, 76), 608 (44, 225(14), 226(17), 227(14, 15), 228(17, 76,78), 6 1 0 ( 3 3 ,7 6 ,8 1 ,8 2 ,8 4 ) ,6 1 1 ( 8 4 ) , 59, 60), 229(59), 230(17), 233, 234, 612(33, 44, 87), 615(23, 28, 100, 101, 504(192e), 512 177), 617(34, l06), 618(44, 76, 87), Reuse, J., 327 (230), 334 619(109), 623(123), 627(69), 628(69, Reynolds, T. M., 617(106), 668 76, 81), 629(142), 632(34), 634(32, Ribbens, C., 325(214), 333, 701(20), 153), 635(32), 638(165), 640(23), 722 641(76), 642(176, 177, 179, 180, 181, Ribeiro, O., 5(40), 10(40), 16(40), 22, 182), 647(189), 660(189), 661(177), 56(4), 57(4), 58 663(212, 215), 664(215), 665(33), 666, Rieth, R., 48(7), 52 667, 668, 669, 670, 671, 785(5), Rinehart, R. K., 232(74), 235 789(34), 790(34), 791(34), 792(41), Riniker, B., 696(4), 699(4), 721 812, 813 Riniker, R., 696(4), 699(4), 721 Robison, M. M., 209(28), 233, 303(151), Risi, A,, 95(39), I16 304(151, 160, 162), 305(162), 312(151, Ritchie, E., 250(4), 267, 588(31), 589, 179), 331, 332 663(213), 664(213), 671 Robson, W., 50(37), 53 Ritter, D., 599(41), 602(41), 667 Rochelmeyer, H., 771 (158), 783

840

AUTHOR INDEX

Rodionow, W. J., 753(93), 781 Roe, A. M., 106(81), 107(81), 117 Rogers, E. F., 451(113k), 509, 673(1), 676(1), 678 Rohner, F., 20(138), 25 Roland, M., 328 Romeo, A., 701 (17), 722 Rose, H. A., 173(69),200 Rosecrans, J. A, 265(41), 268, 498(186c). 512 Rosen, W. E., 121(27), 156, 300(138), 301 (143), 302( 143), 304(143), 331 Rosenberger, M., 451 (113j), 509, 678(7), 678 Rosenberger, S., 101 (77), 117 Rosenmund, P., 625(132a), 633(132a) Rothen, ,4., 735(37), 780 Rothlin, E., 95(48), 116 Rottenberg, M., 44(62), 45, 395(39), 400(39), 507 Roussinoff, K. S., 282(85), 285 Rout, M. K., 362(26), 506 Rowson, J. M., 227(55), 234 Roychaudhuri, D. K., 110(98), 117, 121(25, 26, 28), 156, 167(56, 57, 58), 200, 300(134), 302(146), 331, 529(53), 577, 696(8), 698(8), 704(27), 709(8), 710(8), 722 Rubin, N., 277(37), 284 Rummel, L, 160(2), 199 Rusinov, K., 282(72), 285 Rutledge, R., 311(174), 314(182), 332 Rutschmann, J., 12(93a), 24, 734(19), 736(19), 737(19), 738(19,44), 742(49), 744(44, 49, 54, 58, 65), 768(142), 780, 781, 782 Ruveda, E. A., 507 Ruyssen, R., 237(3), 246 Rydon, H. N., 44(60), 45

S Sadykov, A. S., 48(21a), 52 Safrazbekyan, R. R., 56 ( 1 la), 58 Sager, R. W., 281(67), 282(67), 284 Saito, K., 49(28), 50(28), 52, 327, 328 Sakan, T., 623(129), 632(148), 633(150), 641 (148, 172), 669, 670 Salkin, R., 293(93), 327, 330, 710(49), 715(49), 723 Salway, A. H., 28(6), 44

Samsonova, G. A., 48(10), 52 Sandoval, A,, 204(11), 2 1 6 ( l l ) , 217(11), 224(11,44),225(44),227(11),233,234, 337(8), 358(8), 416(66, 66a), 457(8, 117), 461(117), 505(8), 506, 508, 510 Sandoz, S. A., 319(192), 333 Saner, H., 293(62, go), 329 Santhakumari, T . N., 270(19), 283 Santos, A. C., 161(37), 195(37), 196(37), 200 Santos, J. K., 3(19), 22 Sanyal, R. K., 13(64b), 23 Sargeant, K., 97(67, 68), 117 Sargeant, L. J . , 60(19), 61(19), 64(19), 75(19), 85(19), 89 Sarkar, B., 270(21), 283, 288(113), 330 Sastry, R. V . R., 326(220), 334 Savitri, T. S., 201, 202(98) Saxton, J . E., 33(30), 45, 293(92), 296(92), 329, 480(159, 160), 511, 585(23), 589, 600(49), 601(49), 605(49), 642(182), 667, 670, 792(38, 40), 813 Sayo, J. Y., 300(137), 331 Sayre, L. E., 94(9, 10, 11, 12, 13), 95(28, 29), 115, 116 Scane, J . G., 246(38), 247 Schales, O., 51 (43), 53 Schatzle, E., 44 (62), 45 Scheindlin, S., 277(37), 284 Schenk, H. R., 311(177), 323(199), 324 (199), 332, 333, 752 (go), 781 Schenker, E., 173(70), 200 Sohenker, K., 599(39), 635(156), 642(156), 644( 156), 667, 670, 785 (2), 792 (2, 38), 793(2), 812, 813 Scheuer, P. J., 294(39), 328, 329, 431(99), 509,512,623 (127),630 (127),632 (127), 633(127), 669, 708(40), 722, 797(45), 799(53), 813 Schindler, O., 505(194), 513 Schindler, W., 215(38), 234 Schlagdenhauffen, F. R., 160(3), 199 Schlager, L. H., 326(226), 334 Schleifer, M. J., 294(119), 330 Schlemmer, F., 735(31), 780 Schlempp, G., 622(117, 118), 668, 669 Schlientz, W., 742(49), 744(49), 758(96, 97), 760(103), 771(159), 780, 781, 783 SchIittIer, E., 19(135), 20(135), 25, 160(12, 19a), 164(53), 199, 200, 204(2),

AUTHOR INDEX

207 (23), 232(67), 2 3 3 , 2 3 5 , 2 5 0 ( 1 ) ,267, 276(34), 283, 293(62, 76, 82, 89,90, 91, 118), 294(81), 296(91, 125), 297(91, 131), 300(125, 134, 142), 302(91), 303(91, 161), 304(161), 308(161), 328, 329, 330, 331, 332, 395(39), 400(39), 485(177), 507, 508, 511, 520(26), 528, 529(26,54),531 (54), 533(68), 534(71), 535(7i), 537(17), 576, 577, 704(27), 722, 785(1, 6), 788(16), 789(1, 31), 790(1), 791(1), 792(16), 793(16), 795(16), 797(16), 798(48), 804(16), 805(16), 811(71), 812, 813, 814 Schmauss, O., 656(204), 661(204), 671 Schmid, H., 48(21), 52, 126(33, 34), 128(38), 156, 201(92b, 92c), 201, 265(36), 266(35, 36), 268, 389(31), 390(33), 394(33), 400(44), 403 (31), 411 (44), 416(33, l20), 418(67), 419(53), 423(95, 102), 426(95, 102), 427(95, 101), 429(95, 102), 434(91), 435(95),438(91,95),439(96),443(101, I l l ) , 444(111, 113), 461(118a), 462(118a, 119, 120), 463(44), 466(67, 119, 124, 125, l37), 482(113d), 495(53, 91), 497(186), 498(186), 503(113), 504(53, 91, 96), 506,507, 508, 509, 510, 512, 518(16), 519(16, 19), 520(16, 23, 24, 27, 28, 30, 32, 33, 34, 35), 521(16, 23, 24, 30, 32, 33, 34, 35, 36, 37, 39, 40, 41,42), 522,523 (23,30,35,42), 524 (42, 49), 525(49), 526(49),527(34), 528(40), 529(54). 531 (40,54),532(40), 533(40), 534(27, 28, 40), 537(82), 539(37, 39, 82), 540(24, 30, 35, 39, 8 2 ) , 541 (24, 85, 86), 542(86, 571, 543(85, 86),545(16, 93, 94), 546(85, 95), 547(85, 93, 97), 548(33, 34, 35, 85, 98, 102), 549(102), 551(102), 552(108, log), 553(110, l l l ) , 554(115), 555(16, 24, 30, 116), 556(96, 116), 559(96), 560(32), 561(35, 82, 121, 122), 562(35, 112, 124, 125), 563(124, 127), 564(110, 111, 127), 565(116, 127), 566(127), 567(16, 32, 127), 568(128, 130), 569(131, 132), 570(131, 132, 134), 571(132), 573(16, 35, 132), 574(16, 23, 35, 36, 39, 40, 136), 575(16, 30, 34, 35, 36, 40, 41, 136, 137), 576, 577, 578, 599(37a), 618(107, 108), 619(107), 628(135, 136), 632(37a, 146a),

841

633(108), 667, 668. 669, 677(6), 678, 681(16), 682(17), 690, 804(14), 812 Schmidt, I., 575(137), 579 Schmute, J., 400(23, 43), 469(138, 140), 470(138), 474(139, 140, 149, 150), 476(139, 149, 150, l55), 477(139), 480(139, 155), 498(149), 506, 507, 508, 510, 511 Schneider, J. A., 232(72,74), 235,296(127), 327(232), 330, 334, 811(71), 814 Schneider, W., 599(43), 602(59), 615(59) 636(43, l 6 l ) , 667, 670 Schnider, O., 327(229), 334 Schnoes, H. K., 259(19), 262(19), 267, 390(51), 396(51), 484(51), 495(51), 507 Schofield, J. A , , 44(60), 45 Schofield, K., 708(45), 723 Schoone, J. C., 598(36), 667 Schoop, H., 600(45), 667 Schopf, C., 57(12), 58, 308(167), 332 Schreier, E., 767, 764(107), 765(107), 782 Schroeder, H. D., 547(96), 5 5 5 ( l l 6 556(96, 116), 559(96), 565(116), 578, $79 Schulde, F., 5 (45), 9 (45), 23 Schuler, B. 0 . G., 204(13), 227(13), 230 ( l 3 ) , 233, 328 Schulte, H., 602(56, 58), 606('70), 615(56, 58), 618(56), 667 Schumann, D., 701 (Bib), 201, 461(118a), 462(118a), 510 Schwaebel, G., 594(14), 604(14), 625(131), 668, 669 Schwarz, H., 94, 95(26), 98(72), 110(18), 112 ( I S ) , 113 (18), 1l 4 ( 18), 115 (18), 115, 116, 117, 160(12), 164(53), 199, 200, 328 Schweer, M., 95(40), 116 Schwyzer, R., 294(81), 296(125), 300(125), 329, 330 Scott, C. C., 262(28), 264(28), 268 Scriabine, A., 327(232), 334 Seaton, J. C . , 61(28), 64(57), 65(57), 66(57), 67(28), 70(28), 73(28), 76(28, 78), 76(78), 78(28), 90, 91 Seba, R. A , 328 Seebeck, E., 205(16), 212(16), 233, 509, 534(69), 577, 804(62), 805(62), 814 Seeger, H., 622(121), 636(121), 669

842

AUTHOR INDEX

SefEoviE, P., 257(16), 260(16), 261(16, 23), 267, 276(43, 45, 87), 277(45, 86), 280(86, 87), 284, 285, 337(10a), 405(78), 420(73, 74a), 495(78), 506, 508 Seibl, J., 497(186), 498, 186), 512 Sela, M., 44(61), 45 Selzer, H., 326(228), 334 Semonsky, M., 765(117), 768(137, 139, 140, 141), 769(148, l S l ) , 770(148), 182, 183 Seo, M., 173(68a), 200, 499(186i), 512, 633(63), 577, 689(27), 691, 701 (23), 722 Seshadri, T. R., 3(20), 22 Shalitin, Y., 44(61), 45 Shamma, M., 169(63, 64), 200, 297(130, 133a), 331, 503(192), 512, 702(26), 704(26), 708(43,44), 710(44), 716(44), 722 Shapiro, D., 14, 25, 39(44), 45 Sharp, T. M., l l 9 ( l ) , 145(1), 155, 160(7), 161(33), 162(7, 48), 194(33), 195(33), 196(33), 199, 200 Shavel, J., 3 (25, 26), 22, 69 (68), 91, 293(8, 94), 310(172), 327, 330, 502(188), 512, 712(56), 723 Shaw, F. H., 162(42), 200, 663(214), 671 Shaw, G. E., 238(9), 240(9), 247 Shedlovsky, T., 735(38), 780 Shemyakin, M. M . , 753(93, 94), 781 Shenstone, W., 592, 666 Sheppard, H., 302(143), 304(143), 331 Sheppard, N., 585(21), 589 Shimamura, O., 252(10), 267 Shimata, A., 300(137), 331 Shimizu, M., 160(16), 199, 270(20), 283 Shimodaira, K., 18(130), 25 Shoolery, J. N . , 224(44), 225(44), 234, 258(17), 267, 300(138), 331, 337(9), 367(9, 29), 389(29), 391(9), 396(9), 401(9), 404(9, 29), 415(29), 447(114), 457(117), 461(117), 506, 510 Shoop, E. C., 450(113b), 451(113b, j ) , 497(113b),509,676(5), 677(6), 678(7), 678 Shore, P. A., 16(115), 25, 327(230), 334 Shrivastava, H. N . , 585(19), 589 Siddiqui, R. H., 293(86, 102), 329 Siddiqui, S., 48(4), 49(4), 52, 293(29, 86),

294(120), 327, 328, 329, 330, 708(37), 722, 789(32), 791 (36), 813 Siddons, L. B., 162(43, 44), 200 Sigg, E. B., 232(72), 235 Silva, R. A., 643(185), 670 Silverman, N., 362(26), 506 Silvers, S., 120(21), 134(21), 156, 281 (60), 284, 491(175), 511, 535(75), 537(75), 578, 805(27), 812 Silverton, J. V., 585( 19), 589 Sim, G. A., 32(23, 24), 44, 184(84), 201, 262(28), 264(28), 268, 482(164), 491(178), 511, 512, 535(79, 80), 560(118), 578, 579, 587(29), 589, 716(66), 713, 788(28), 805(28), 812 Simon, H., 2(10), 21, 766(124), 767(131), 782 Sims, J. J., 61(26b), 85(26b), 86(26b), 87(26b), 90 Sinha, S., 13(64b), 23 Sinha, Y. K., 13(64b), 23 Sjoerdsma, A., 8(64a), 13(64a, 99), 23, 24 Sklar, R., 85(84), 86(84), 87(84), 91, 120(22a), 148(56), 156,157,201 (92b), 201, 254(40), 262(27, 28), 264(27, 28), 266(40), 267, 268, 281(57), 284, 482(164), 485(177), 511, 534(71), 535(71), 537(71), 577, 716(66, 67), 723, 788 ( l 6 ) , 792 (16), 793 (16), 795( 16), 797( 16), 804( 16), 805(16), 812 Skolic, J., 611(84b), 668 Slaytor, M. B., 772(168), 7 8 3 Smith, A. F., 120(22a), 148(56), 156, 157, 201 (92d), 201, 209(30), 233, 254(40), 262(27), 264(27), 266(40), 267, 268, 477(156), 477(156), 505(156), 511, 716(67), 723 Smith, A. H., 8(63a), 11(83b), 13(63a), 23 24 Smith, C. W., 5(48, SO), 23 Smith, D. C. C., 767(134), 782 Smith, E., 293(93), 330 Smith, G. F., 5(46), 23, 33(28, 29, 30, 32a), 45, 124, 126(32), 128(35), 129(39), 130(39, 40), 131 (40), 136(49), 138(49, 50,51), 139(49,50,51,53,54), 141 (as), 142(49, 50), 144(51), 145(50, 51), 146, 147(49, 50, 53, 54), 148(53, 54), 149(51, 53, 54), 151(50, 51, 53, 54), 152(51, 53, 54), 154(55), 156, 157,

843

AUTHOR INDEX

175(73), 176(76), 177(76,81), 180(76), 183(81), 184 (81), 186 (86), 187 (86), 188(86),189(86,88),200,201, 265(34), 268, 281(55), 284, 337(12), 358(12), 359(19), 361(12), 362(27), 365(12), 367(19), 397(51b), 400(12, 27), 405(77), 419(51b), 453(19), 461(19), 462(120a), 465(19, 123), 466(77, 123, 137a), 491(179e, 184a), 506, 508, 510, 512, 562(126), 579, 582(7, 8), 585(23), 586(7, 8, 25, 27), 588(32), 589, 630(145), 632(145), 669, 681, 682(13), 690 Smith, H., 767(133, 134), 782 Smith, J. C., 615(101), 668 Smith, M. E., 327(232), 334 Smith, M. I., 735(29), 780 Smith, S., 734(18,22), 735(18), 747(18,67), 748(73, 78, 79), 758, 779, 780, 781 Smolik, S., 318(189), 322(189), 333 Sneeden, R. P. A., 97(66), 98(71), 99(66), lOO(66, 71), 101(71), 116, 117, 206(21), 233 Snyder, H. R., 5(48, 50), 23, 450(113a), 451(113k, 1, m), 495(181), 509, 510, 512,673(1,2, 3,4), 674(3,4),675(2,3), 676(1, 3), 678 Somers, T. C., 160(11), 199, 288(112), 330 Sonderhoff, R., 517(10), 520(10), 576 Sonnenschein, F. L., 93, 115 Spaeth, E. C., 495(181), 512 Spiith, E., 48(6, S), 49(30), 50(36), 52, 53, 582(9), 584(14), 589, 594(18), 666 Speetcr, M. E., 9(68), 13(102), 18(68), 23, 24 Spenser, I. D., 50(38), 53 Spiegel, L., 94(5), 115 Spilsbury, J. F., 747(72), 763(72), 765(72), 781 Spingler, H., 293(51, 68, 108), 297(51, 95), 328, 329, 330 Spiteller, G., 28 (Ba), 44, 138(50), 139 (50), 142(50), 144(50), 145(50), 147(50), 151(50), 154(55), 157, 201 (92b), 201, 359(18, 18a), 361(18), 365(18), 367(18a), 390(18, 28, 51a), 391(28), 395(28, 51a), 396(28, 51a), 397(18, 51a), 419(28), 441(107), 444(106, 107), 453(28, 51a), 456(28, 51a), 482(16%, 164a), 491(179e), 495(28, 179d),

505(199), 506, 507, 509, 611, 512, 694(2), 696(2), 698(2), 721 Spiteller-Freidmann, M., 201 (92b), 201, 290(51a), 397 (51a),

395(51a), 453 (514,

396(51a), 482 (162),

495(179d), 507, 511, 512, 694(2), 696(2), 698(2), 721 Spring, F. S., 665(224), 671 Stadler, P., 740(47), 749(169), 755(173), 756(173, 174), 771(165), 780, 783 Staib, I., 95(46), 116 St. Andr6, A. F., 288(111), 295(15), 296(125), 300(125, 134), 303(161), 304(161), 308(161), 327, 330, 331, 332, 704(27), 722 Stauffacher, D., 205(16), 212(16), 224(43, 43a), 231(64), 233, 234, 453(118), 460(116), 461 (118), 457 (116), 463(116), 466(116), 485(116), 509, 510, 534(69), 535(73), 549(73), 577, 578, 688(22), 691, 804(62), 805(62), 814 Steams, J., 727, 751(2), 769(2), 779 Stedman,E.,29(12, 13), 30(15), 32(13), 41, 44,45 Stesnhauer, A. J., 293(85), 329 Stein, O., 5(44), 22 Steinborn, K., 599(40), 667 St,enlake,J. B., 511, 737(41), 781 Stepanov, S. I., 9(71), 23 Stephen, A. M., 663(212), 671 Steuer, H., 57 (12), 58 Stevens, T. S., 94, 96(16), 97(16, 69), 98(69), 99(69), 107(16), 108(91), 110(16), 115, 117 Stevenson, A. E., 94(11, l 2 ) , 115 Stewart, J. M., 5(48), 23 St,oehr, C., 641 (167, 168), 670 Stoll, A., 18(131), 25, 293(2, 64), 327, 328, 329, 534(59), 577, 727(10), 728(13, 15), 734(17, 19), 735(24, 25, 40), T36( 19), 737 (19, 42, 43), 738( 19, 44), 740(40), 742, 744(44, 49, 54, 56, 57, 58, 65), 745(56), 746(25, 66), 747(68), 748(25, 66, 75, 81), 749(10, 13, 85), 751(68, 88, 89), 752(90), 753(89), 757(13), 758(13), 759(10, 99), 760( 103), 764( 112), 765( 112), 768(66, 81, 138, 142, 143), 771(159), 772(13, 66), 779, 780, 781, 782, 7 8 3

844

AUTHOR INDEX

Stoll, W. G., 204(14), 215(14), 216(14), 225(14), 233 Stork, G., 361(27a), 365(27a, 27b), 367(27a), 506, 705(30), 722 Stowe, B. B., 1( 8 ) , 2, 21 Strasky, E., 280(76), 285 Strating, J., 326(216), 333 Straus, F., 29(9), 44 Streuli, P., 583(12), 589 Stroh, W., 582(9), 584(14), 589 Strohmayer, H. F., 450(113a), 509, 673(4), 674(4), 678 Stromberg, V. L., 16(117), 25 Strouf, O., 259(20), 260(21), 261 (21), 22, 24), 267, 276(35, 36,41,44,47,48), 277(36, 46, 38), 281(36), 283, 284, 495(184), 512 Stumpf, W., 327(230), 334 Sturmer, E., 758(96), 781 Sudarsanam, V., 480(16Oa), 511 Suehiro, T., 744(63), 780 Sugasawa, S., 36, 45 Suginome, H., 40, 45 Suld, G., 58(15), 58 Siiltanov, If.B., 280(80), 282(85), 285 Suter, C. M., 5(51), 23 Svierak, O., 582 (6), 586 (6, 26), 588, 589 Svoboda, G. H., 160(9), 173(9), 199, 204(12), 218(12), 219(42), 225(42), 233, 234, 270(4, 6 , 22, 23, 31), 271(4, 5, 6, 12, 22, 23, 30), 272(4, 5, 6, 12), 273(6), 275(6), 283, 337(5), 419(5), 506, 508, 510, 708(36), 722 Swan, G. A., 108(89), 109, l l 0 ( 9 6 ) , 117, 164(52), 200, 498(186g), 512, 533(60), 577, 688(25), 689(25), 691 Swanholm, C. E., 265(43), 268, 423(81), 426(81), 429(81), 430(81), 508 Swanson, E. E., 95(31), 116 Szabo, A. G., 451(113j), 509, 678(7), 678 Sza,b6, Z., 280(77, 78, 81), 285 Szbra, St., 10(80), 24 Szasz, K., 280(77), 285 Szczeklik, E., 280(78), 285 Szendey, G. L., 759(101), 781 Szmuszkovicz, G., 312(178),332,365(27b), 506 Szpilfogel, S., 597 (29), 604 (66), 605 (66), 666, 667 Szporny, L., 280(77), 285

T Tabata, T., 220(42c), 234, 273(14), 283 Taber, W. A., 2(10), 21, 766(123), 782 Tackie, A. N., 61(26a, 26c), 85(26a, 26c), 86(26a, 26c), 87(26a, 26c), 88(26a), 89(26c), 90 Tafel, J., 592(7, 8 ) , 600(7, 8, 46), 607(77), 615(99), 618(7, 8), 636(158), 666, 667, 668, 670 Takahashi, M., 244(30), 247 Takenaka, Y., 811 (76), 814 Takao, T., 779(172), 783 Talapatra, S. K., 161(35, 36), 194(35), 195(35,36),196(36), 199,270(18),283, 293(19), 327, 328, 329, 416(113c), 450(113b), 451(113b, c, j), 497(113b), 504(192f), 509, 512, 676(5), 677(5), 678(7), 678, 707(34), 708(34), 716(34), 722, 792(42), 800(59), 813 Tamm, R., 702(26), 704(26), 722 Tani, H., 499(186j), 512 Tanret, C., 727(4, 5), 779 Tasaki, S., 1(1), 2(1), 21 Tatevosyan, G. T., 56(11a), 58 Tatsui, G., 51 (41), 53 Taube, K., 634(151), 669 Tauber, E., 49(25), 52 Tauro, C. S., 326(219), 334 Taylor, D. A. H., 227(55), 234, 744(61), 780 Taylor, E. H., 2(10), 21, 767(132), 782 Taylor, W. C., 250(4), 267 Taylor, W. I., 60(14a), 63(53), 64(53), 66 (66), 69(66), 70(66), 77(66), 79(66), 80(66), 82(82), 83(82), 84(82), 85(82), 89, 90, 91, 97(66), 99(66), 100(66), 116, 120(22a), 148(56), 156, 157, 201 (92b, 92d), 201, 204(6), 205(6), 207(22, 25), 209(25, 27), 213(25), 216(25), 223(6), 224(6), 231 (63), 232(67), 233, 234, 235, 237(6), 238(6, lo), 240(6, 20, 21), 246(6), 246, 247, 250(1), 251 (7), 252(7, 11, 12), 254(7, 11, 32, 40), 255(12, 47), 256(7, 13), 257(7), 258(7), 259(7), 262(27, 28), 264(27, 28), 265(32), 266(40), 267, 268, 281(57), 284, 293(49, 72,98, 118), 295(123),297(49), 308(170), 309(171), 310(172), 327, 328, 329, 330, 332, 400(43),482(164),484(171), 485(177),

845

AUTHOR I N D E X

495(183), 502(187, 189), 503(189, 192a, 192g), 507,511,c512,513,533 (68), 534(71), 535(71), 537(71), 577, 602(57), 606(57), 635(155), 646(135), 653(155), 667, 669, 712(34, 58), 716(66, 67), 719(74), 721(54), 723, 785(6), 788(16), 789(33), 790(35), 791(35), 792(16), 793(16), 795(16,35), 796(35), 797(16), 798(35, 46, 50), 799(50), 800(35,61), 801 (61), 804(16), 805(16), 806(35, 65), 808(66), 812, 813, 814 Tchaman, E. S., 753(93), 781 Terrell, R., 365(27b), 506 Terzyan, A. G., 56 (1l a ) , 58 Tessmar, K., 607(72), 640(72), 667 Teuber, H. J . , 101(77), 117, 583(10), 585(10), 589, 629(143), 630(144), 669 Thesing, J., 5(45), 9(45), 23, 533(61), 577, 689(26), 691 Thesis, Ph. D., 31(18), 32(18), 34(18), 35(18), 44 Thomas, A. F., 119(1), 120(15a, 23), 121(2, 23), 123(23, 30), 130, 134(23, 46), 135(46), 145(23), 146(23, 46), 147(15a), 155, 156, 157, 510 Thompson, F. A , , 93, 115 Thompson, M. R., 748(77), 781 Thudium, F., 758(97), 781 Thullier, J . , l2(91), 24 Thurkauf, M., 44(62), 45 Timmis, G. M . , 734(18, 2 2 ) , 735(18), 474(18, 67), 748(78, 79), 758, 779, 780, 781 Tishler, M., 5(50), 23 Titus, E., 17(125), 25 Tobita, M., 61(31), 64(60), 65(60), 70(31), 71(73), 90, 9 I Tocker, S., 52 (47), 5 3 Tomita, M . , 60(15), 89 Tomko, J., 257(15), 261(15), 267, 276(42), 284 Tomlinson, &I., 470( 143), 510 Tondeur, R., 64(57), 66(57), 91 Tonolo, A., 747(69), 781 Trcka, V., 326(218), 334 Tretter, J. R., 327(231), 334 TrojAnek, J . , 259(20), Z S O ( Z l ) , 261(21, 22, 24), 267, 270(25), 271 (25), 276(33, 35, 36, 41, 44, 47, 481, 277(36, 46, 48),

281 (36), 282(33), 283, 284, 495(184), 512 Trotter, J., 220(42c), 234, 273(14), 283 Troxler, F., 10(83), l l ( 8 3 , 84, 85), 12(83), 18(131), 24, 25, 311(177), 323(199), 324(199), 332, 333, 735(40), 740(40), 749(169),768(142),769(150),770(153, 158, 157), 771(165), 772(167), 780, 782, 783 Tsao, D. P. N., 265(41), 268, 498(186c), 512 Tscherter, H . , 747(70), 760(70), 781 Tschesche, R., 9(73a), 16(73a), 23, 49(24), 52 Ischuchihashi, R., 1(1), 2(1), 21 Tshugaeff, L., 238(5), 246(5), 246 Tsukamoto, H., 65 (63), 91 Tsuyuki, T., 252(10), 267 Tulinsky, A., 120(21), 134(21), 156, 281 (60), 284, 491(175), 5 I I , 535(75), 537(75), 578, 805(27), 812 Tuma, L., 14(108), 25 Tunman, P., 505(195), 513 Turley, J. LV., 300(137), 331 Turner, F. A., 326(221), 334 Turner, R. B., 246(32), 247, 785(3), 812 Turner, W. J . , 17(119), 25 Tuschen, E., 630(146), 669, 791(37), 813 Tyler, V.E.,Jr.,8(63a), 10(83a), 11(83b,c), 13(63a, 83a, 97), 17 (126a,b), 23, 24, 25, 735(23), 765(120), 780, 782

,.. .

U Uchimaru, F., 160(16), 199, 270(20), 283 Udelhofen, J . H . , 67(67), 91 Udenfriend, S., 8(64a), 13(64a, 99), 16(115), 17(125), 23, 24, 25 Uebel, H., 232(68), 235 Uhle, F. C . , 742, 744(60), 780 Uhlenbroock, K., 95(40), 116 Ullyot, G. E., 56(4), 57(4), 58 Ulshafer, P. R., 293(49), 294(81), 295(15), 297(42, 49), 300(42,79),303( 155,161), 304(79, 161), 308(161), 323(155), 327, 328, 329, 331, 332, 503(192a), 512, 798(48, 50), 799(50), 813 Unna, K. R., 232(73), 235 Uribe, R.,160(14, 15), 199 Utkin, L., 195(93), 201

846

AUTHOR INDEX

V Vairel, C., 232(74), 235 Valls, J., 9(70b), 23, 316(185), 318(185), 321(185), 322(185), 323(185), 324 (185), 333 Valsecchi, A., 311(176), 332 Valzelli, L., 311(176), 332 Vamvacas, C., 48(21), 52, 521 (40), 528(40), 529(54), 531 (40,54), 532(40), 533(40), 534(40), 574(40), 575(40), 577 van Bever, L., 175(71), 200 van Binst, G., 225(58), 234 van Camp, A., 173(69), 200 Vand, V., 300(137), 331 van den Driessen Mareeuw, W. P. H., 441 (84), 508 Vandermeers, A., 226 (46), 234 Van Der Meulen, T. H., 716(68), 723 Vanderwerff, W. D., 9(69), 23 Van Dongen, K., 811(72), 814 Vangedal, S., 303(150, 157), 331, 696(6, 12), 698(6, 12), 721, 722 van Heerswynghels, J., 227(57), 234 van Itallie, L., 293(85), 329 Van Os, F. H. L., 280(82), 285 van Proosdij -Hartzema, E. G., 246 (34), 247 van Romburgh, P., 3(17), 4(17, 30), 22 Van Stolk, D., 533(59), 577, 688(24), 691 VanTamelen,E. E., 168(60),200,295(123), 300(134, 141), 330,331,632(147), 642, 669,670,702(26), 704(26,27),711 (52), 713(52), 716(69, 70), 722, 723, 766(127), 767(127), 780, 781, 782 van Urk, H. W., 735, 780 Vareed, C., 246(34), 247 Varvoglis, G., 650(198), 656(198), 670 Vauquelin, M., 727, 779 Vaz, A., 271 ( 8 ) , 272(8), 283 Vee, A., 28(2), 44 VejdBlek, Z. J., 14(108), 25, 323(218), 334 Velluz, L., 316(184, 185), 318(185), 321(185), 322(185), 323(185,200,201, 202, 205), 324(185, 200, 202, 207, 208), 332, 333 Venkateswaran, W., 792(42), 813 Venkatochalam, K., 226(47), 234 Verbeek, A. A., 204(13), 227(13), 230(13), 233 Vercellone, A., 13(103), 14(103), 24

Vergara, B. U., 160(13), 199,328 Vero, L., 747(69), 781 Verzele, N., 806(64), 814 Vetter, W., 390(33), 394(33), 416(33), 507, 677(6), 678 Villain, G., 637 (163), 670 Villegas-Castillo, A., 328 Vincent, D., 232 (74), 235 Vining, L. C., 2(10), 21, 766(123), 782 Viswanathan, N., 201, 202 (98), 427(101), 431(100), 441(100, 101), 443(101), 444(101), 509 Vogel, G., 232(68), 235 Voltz, J., 744(59), 780 Volz, H., 549(104), 562(123), 579 v. Bbkbsy, N., 735(32), 780 von Euler, H., 4(32, 33, 34), 5(32, 34, 38, 41), 22 von Gorup-Besanez, 161 (20), 199 von Lippmann, E. O., 3, 22 von Miiller, F., 160(2), 199 von Philipsborn, W., 201 (92b), 201, 390(33), 394(33), 416(33), 466(124, 137), 507, 510, 520(27), 528(54), 531 (54), 534(27), 537(82), 539(82), 540(82), 541 (85), 542(87), 543(85), 545(93), 546(85), 547(93, 96, 97), 548(85, 98, 102), 549(102), 551(102), 552(108), 556(96), 559(96), 561 (82), 563(127), 564(127), 565(127), 566(127),567(127),568(130),569(131, 132), 570(131, 132), 571(132), 573(132), 576, 577, 578, 579, 677(6), 678,681(16), 690, 804(14), 812 von Soden, H., 1 (4), 21 Vorbrueggen, H., 83(83), 91, 134(44), 157, 262(26), 267, 482(165), 484(165), 485(165), 490(165), 491(165), 511, 535(77), 578, 694(1), 696(1), 707(1), 721, 788(21), 805(21), SOS(Zl), 812 Votova, Z., 768(139), 782

W Waalkes, T. P., 13(99), 24 Wahid, M. A., 397(51b), 419(51b), 507 Wakim, K. G., 162(46), 200 Walker, E. F., 297(130), 331 Walker, G. N., 744(53), 780 Walker, F. J., 451(1131,m), 510, 673(2, 3), 674(3), 675(2, 3), 676(3), 678

AUTHOR INDEX

Walls, F., 204(11), 216(11), 217(11), 224(11, 44),225(44),227(11),2,33,234, 337(8), 358(8), 416(66a), 457(8, 117), 460(117), 461(117), 505(8), 506, 508, 510 Walter, M., 327(229), 334 Wan,A. S. C., 328, 798(49), 800(49), 813 Wang, C. H., 95(22), 115 Warnant, J., 316(184, 185), 318(185), 321 (185), 322(185), 323 (185), 324(185),332, 333 Warnat, K., 623 (125), 636 (125), 642 (125), 669 Warnhoff, E. W., 469(142), 471(142), 510, 696(13, 14), 698(13), 699(13, 14), 704(14), 705(13), 706(14), 722 Warren, F. L., 204(13), 227(13), 230(13), 233, 328 Warsi, S. A., 294(120), 330, 791(36), 813 Waser, P., 520(33, 35), 521(33, 35, 39), 523(35), 539(39), 540(35, 39), 548(33, 35), 561 (35), 562(35), 573(35),574(35, 39, 136), 575(35, 136, 137), 576, 577, 579, 619(110), 668 Wasicky, R., 160(17), 199, 327, 431(98), 509 Watanabe, T., 10(78),24 Waterfield, W. R., 644(186), 670 Watt, J. M., 272(7), 283 Weaver, B. N., 744(53), 780 Webb, L. J., 439(86), 508 Weber, L. E., 594(11), 601(51), 602(51), 666, 667 Weedon, B. C. L., 406(50), 507 Wegener, W., 621 (116), 634(116), 668 Weichet, J., 318(189), 319(191), 321(191), 322(189, 194), 325 (197), 333 Weinberg, M. S., 326(225), 333, 334 Weinfeldt, F., 288(111), 330 Weisbach, J. A., 450(113b), 451(113h, j), 497(113b), 503(192d), 504(192d), 509, 512, 676(5), 677(5), 678(7), 678 Weisblat, D. I., 5(52), 13(102), 23, 24 Weisenborn, F. L., 293(62), 300(134, 135, 141), 302(135), 303(154, 156), 322(196), 324(196), 329, 331, 333, 704(27), 722 Weiss, R., 314(182), 332 Weissbach, H., 13(99), 16(115), 24, 25 Weisskopf, W., 653 (203), 546 (203), 671 Weissman, A., 327(232), 334

847

Weissmann, C., 126(33, 34), 156, 418(67), 466(67, 125), 508, 510, 545(94), 552(109), 563(127), 564(127), 565(127), 566(127), 567(127), 578, 579,599(37a),618(107,108), 619(107), 628(135, 136), 632(37a, 146a), 633(108, 127), 667, 668, 669, 681(17), 690 Wendler, N. L., 767(129), 782 Wendt, H. J., 2(10), 21, 766(122), 782 Wenkert, E., 7, 23, 64, 66(64), 67(64, 67), 68(64), 72(64), 73(64), 80(64), 85(84), 86(84), 87(84), 91, 106(82), 110(98), 113, 115(104), 117, 121(25, 26, 28), 122(29), 156, 167(56, 57, 58), l69(61, 62), 197(97), 200, 201, 231(65), 235, 238(11), 247, 256(14), 264(30), 267, 268,300(134,136),302(146), 308(165), 328, 331, 332, 337(14), 453(115), 469( 115), 476( 154), 485( 177), 495(183a), 503(190, 191), 506, 510, 511, 512, 529(53), 533(64, 65, 66), 534(71), 535(71), 537(71), 577, 689(29), 690(30, 31), 691, 696(8, 9, lo), 698(8, 9, lo), 704(27), 709(8, 46, 50), 710(8, 46), 711(46, 50), 712(59), 718(50), 722, 723, 785(4), 788(7a, 16), 792(16), 793(16, 43), 795(16), 797(16), 798(48), 804(16), 805(16), 812, 813 Werblood, H. M., 800(6l), 801(61), 814 Werner, G., 293(96), 328, 330, 800(59), 813 Werner, H., 51(43), 53, 325(209), 333, 696(16), 700, 701(16), 722 Werst, G., 744(64), 781 West, G. B., 8(64), 13(64), 23 Weygand,F., 2(10), 21, 766(121, 122, 124), 767(121, 131), 782 White, C. T., 272(7), 283 White, E. P., 8(60, 61, 62), 23 Wickberg, B., 66(64), 67(64), 68(64), 72(64), 73(64), 80(64), 91, 113, 115(104), 117, 122(29), 156, 169(61), 200, 256(14), 267, 503(190, 191), 512, 533(66), 577, 690(31), 691, 709(46), 710(46), 711(46), 712(59), 723 Wieland, H., 17(121, 129), 18(121), 25, 517(10, 11, 12, 13),520(10, 11, 12, 13), 521(12, 13), 539, 540(83), 542(88), 548(12),560(13),568(11, 12), 575(13), 576, 577, 578, 579, 601(55), 605(67,

848

AUTHOR INDEX

68), 606(67), 611 (67), 621 (114), 638(164), 641(175), 647(175, 187, 188, 191), 648(137, 191, 192), 649(187, 191, 192, 193, 194, 195), 650(187, 196, 197, 198, 199, 200), 651 (187, 196, 197, 202), 653(193,195,203), 654(195,202,203), 656(192, 193, 196, 198, 202, 204), 657(187, 193, 196), 659(199), 660(187, 188, 207), 661(204, 208), 662(187, 193, 194, 197, 209, 210), 663(175, 188, 200, 211), 664(216, 217), 667, 668, 670, 671 Wieland, T., 5(43), 17(126), 22, 25, 466(132, 133), 510, 518(17), 519(17), 520(17), 521(17), 522(17), 523(47), 549(103, 104, 105), 551(107), 560(17, 119, 120), 561(120), 562(47, 123), 570(133), 574(17), 575(17), 576, 577, 578, 579, 625(132a), 633(132a), 669 Wiewiorowski, M., 5(40a), 22, 611 (84b), 668 Wightman, F., 6(57), 23 Wilkinson, S., 16(116), 25, 747(72), 765(72), 766(72), 781 Williams, D. H., 390(28e), 506 Williams, R. E., 296(132), 297(132), 329, 331 Wilson, A. J. C., 102(78), 117 Wilson, J. M., 82(82), 83(82, 83), 84(82), 85(82), 91, 134(44, 45), 139(51a), 147(51a), 148(51a), 149(51a), 151(51a), 152(51a), 153(51a), 157, 191(92), 192(92), 194(92), 201, 224(44), 225(44), 234, 258(19), 260(18), 261(18), 262(18, 25, 26), 267, 270(32), 271 (29), 276(40), 277(49), 281(64), 283, 284, 337(3, 6, 7), 390(6, 7, 32, 52), 393(6), 399(37), 405(6), 414(52), 416(3, 7), 418(7), 419(3, 6), 420(32), 423(102), 426(102), 429(102), 432(37), 435(37), 445(52), 447(52), 448(52), 456( 115a), 457( 115a, 117), 460(117), 461(117), 463(122), 465(115e, 122), 466(122), 467(125a, 126, 127), 468 (125a), 469 (125a, 127), 473(37), 482(165), 484(165, 170, 171), 485(37, 122, 165), 490(165 165a, b), 491(165), 495(170), 497(170),505(37), 505, 506, 507, 510, 511, 535(77), 578, 694(1), 696(1), 707(1), 721, 788(21), 805(21), 808(21), 813

Windgassen, R. J., 680(6, 7), 684(6, 7), 685(7) Winkelman, E., 796(44), 813 Winkler, W., 229(61), 234 Winter, J., 767(134), 782 Wintersteiner, O., 300(134, 135, 141), 302(135), 331, 516(4), 576, 704(27), 722 Wirth, P. H. A , , 735(31), 780 Witkop, B., 31(20), 38, 41, 44, 96(62), 98(73), 100(76), 108(90, 92), 116, 117, 240(17, 18, 19), 241(19), 247, 264(29), 268, 337(13), 358(13, 15, 16), 359(17), 362(25), 395(25), 399(25), 400(25), 414(25), 498(186h), 506, 508, 512, 517(13), 520(13), 521(13), 523, 539(13), 540(83), 560(13), 575(13), 576, 577, 578, 647(190), 649(190), 670, 772 (166), 7 8 3 Wittwer, H., 474(150), 476(150), 511 Wohlfahrt, J., 213(36), 234 Wolinsky, J., 702(26), 704(26), 722 Woodson, R. E., Jr., 288(110), 289, 330, 507, 811 (71), 814 Woodward, R. B., 108(90, 94), 109(95), 117, 246(32), 247, 300(139), 316(183), 331, 332, 337(1), 474(153), 477(157), 505, 511, 542 (go), 578, 585, 589, 592, 597(30), 598,599(39), 602(61), 607(3), 617 (3), 632, 635 (156, 157), 642 (156), 644(156), 657(205), 659(205, 206), 663(206), 666, 667, 670, 671, 735(26), 742(50), 745(26, 50), 755(50), 780, 785(2, 3), 791(37), 792(2, 38, 40), 793(2), 798(52), 799(82), 812, 813 Wormley, T. G., 93, 115 Wragg, W. R., 13(105), 25 Wright, S. E., 772(168), 7 8 3 Wrinch. D. M., 753, 781 Wr6be1, J. T., 33(32a), 45, 124, 126(32), 156, 337(12), 358(12), 359(19), 361 (12), 365(12), 367(19), 400(12), 453(19), 461(19), 465(19), 506, 586 (27), 589 Wu, L. S., 95(36), 116 Wythe, S. L., 170(65, 66), 172(66),200

Y Yamada, S., 588(34), 589, 757(95), 781 Yamano, T., 757(95), 759(100), 760(100),

849

AUTHOR INDEX

763(109), 764(100, 109, 114, 116), 765(109, 116), 781, 782 Yamatodani, S., 757 (95), 760(102, 104, 105), 762(105), 763(109), 764(104, 109, 111, 113, 114, 115, 116), 765(104, 109, 111, 116), 781, 782 Yang, J. T., 792(40), 813 Yang, N. C., 585(21), 589 Yang, S. T., 162(47), 200 Yates, P., 451(113j), 509, 678(7), 678 Yeowell, D. A., 485(172, 173), 491(173, 178), 511, 512, 518(15), 519(15), 520(15, 29), 521(15), 526(15), 534(15, 7 2 ) , 535(72, 80, 81), 536(81), 539(15), 540(15), 547(96), 555(15, 113), 556(96, 113), 559(96), 567(15), 574(15), 575(15), 576, 578, 579, 688(20), 690, 788(23, 2 8 ) , 805(23,28), 812 Yokohama, M., 642(178) Yonemitsu, O., 302 (149), 331, 695 (3), 696 (3, 5), 697 (5), 698 (3), 721 Young, 2. H. P., 13(104), 25 Youngken, H. W . , Jr., 265(41), 268, 498(186c), 512, 811(71), 814 Yui, T., 779(172), 783 Yuldashev, P. K., 282(68, 69, 70, 71), 284

Yunusov, S. Y., 282(68, 69, 70, 71), 284 Yurashevskii, N. K., 9(71, 72, 73, 74), 23, 24

Z Zabolotnaja, E. S., 276(33), 282(33), 283 Zacharias, D. E., 5(40), 10(40), 16(40), 22, 56(6), 57(6), 58 Zahn, E., 238(12), 247 Zalan, E., 719(73), 723, 793(43), 813 Zaltzman, P., 48(14), 52, 270(16), 283, 534(70), 577, 812 Zambito, A. J . , 5(50), 23 Zeitschel, O., 1( l ) , 2 ( l), 21 Zetler, G., 232(73), 235 Zheliazkov, D. K., 280(79), 285 Zhelyazkov, D. M., 282(72), 285 ZikBn, V., 768(137, 139, 140, 141), 769(151), 782, 783 Zinnes, H., 69(68), 91, 310(172), 332, 502 (188), 512, 712 (56), 723 Zuong, N. D., 246(39), 247 Zurcher, A., 518(18), 519(18), 520(18), 521(18), 549(18), 569(129), 570(129), 576, 579 Zvacek, J., 318(189), 322(189), 333

This Page Intentionally Left Blank

SUBJECT INDEX Botanical names are printed in italics. Prefixes such as aci-, apo-, iso-, nor-, proto-, pseudo-, are printed in italics and disregarded for indexing purposes.

A C-Alkaloid B, 575 Abrus precatorius, 2 C-Alkaloid BL, 575 Acacia cultriformis, 8 Alkaloid C, 159, 161, 197 Acaciafloribunda, 8 C-Alkaloid C, 575 Acacia Zongifolia, 8 C-Alkaloid D, 520, 555 Acacia podulyriaefolia, 8 C-Alkaloid E, 521, 567, 573 Acacia pruinosa, 8 C-Alkaloid F, 521, 554, 560 Acacia vestita, 8 C-Alkaloid G, 521, 554, 567, 573 Acer rubrum, 5 C-Alkaloid H, 521, 554, 573 Acer saccharinurn, 5 G-Alkaloid I, 575 Acetylaspidoalbidine, 449 C-Alkaloid J, 574 Acetylcholinesterase, 43 C-Alkaloid L, 575 Acetyldehydrostrychninolone,635 C-Alkaloid M, 574 0-Acetylvincamajine, 281 C-Alkaloid 0 , 5 7 5 Adifoline, 61, 80,266 C-Alkaloid P , 575 Adina cordifolia, 266 C-Alkaloid Q, 575 A d i m rubrestiplata, 60 C-Alkaloid R, 575 Adinine, 61 C-Alkaloid S, 575 Adrenergie blockade, 773 G-Alkaloid T, 520, 534 Agroclavine, 732, 762 C-Alkaloid UB, 575 Agropyrum citiare, 764 C-Alkaloid Y, 521 Agropyrum semicostatum, 760, 764 Alkaloid-2, 575 Ajarmine, 792 Alstonia angustiloba, 161 Ajmalicine, 67, 166, 270, 282, 290, 353, Alstonia congensis, 161 482, 707, 710, 713 Alstonia constricta, 159, 288, 696, 708 Ajmalidine, 290, 787, 797 Alstonia cordifolia, 61 Ajmaline, 2, 290, 293, 495, 785, 787 Alslonia gillettii, 161 Ajmalinine, 290,293, 787 Alstonia mcrophylla, 161 Ajmyrine, 792 Alstonia rnuelleriana, 161, 792 Akuammenine, 155 Alstonia scholaris, 159 Akuammicine, 123, 350, 680 Alstonia somer8etensis, 161 Pseudoakuammicine, 130 Alstonia spatulata, 161 Akuammidine, 131, 281, 354,485,491, 535, Alstonia spectabilis, 159 788, 805 Alstoria venenata, 202 Akuammidinol, 131 Alstonia verticillosa, 161 Akuammigine, 120, 167, 709 Alstonia villosa, 161, 792 Pseudoakuammigine, 134 Alstonicine, 160 Pseudoakuammigol, 135 Alstonidine, 159, 173 Akuarnmiline, 155 Alstoniline, 159, 170 Alstonine, 159, 162, 270, 290, 708 Akuammine, 145, 270, 281 Alstovenine, 202 Abhoreafloribunda, 696 Alstyrine, 162, 529, 715 C-Alkaloid A, 521, 560, 567 851

852

SUBJECT INDEX

Amanita citrina, 17 Amanita mappa, 17 Amanita pantherina, 17 Amanita porphyria, 17 Amanita tomentelh, 17 Aminonaphthostyril, 742 2-Aminostrychnidine, 599 2-Aminostrychnine, 599 2-Aminostrychnine, 599 Amsonia elliptica, 696 Amsonia tabernaemontana, 357, 416 Amsoniaefoline, 290 Amsonine, 696 Arariba rubra, 48 Aricine, 290, 353, 482, 708 Arthrophytum leptocladum, 9, 48 Arundo donax, 4 Aspergillus fumigatus, 765 Aspidoalbine, 348, 445 Aspidocarpine, 339, 400, 445 Aspidomine, 343, 421, 432 Aspidofractine, 265, 344, 421, 429, 434 Aspidofractinine, 343, 420 Aspidolimidine, 348, 448 Aspidolimine, 339, 403 Aspidosine, 338, 649 Aspidosperma album, 357, 445 Aspidosperma auriculatum, 357, 7 16 Aspidosperma australe, 357, 505 Aspidosperma carapanauba, 82, 357, 502 Aspidosperma chakensis, 357, 405 Aspidosperma compactinervium, 357, 466 Aspidosperm cylindrocarpon, 357, 410 Aspidosperma dasycarpon, 357 Aspidosperma eburneum, 357 Aspidosperma fendleri, 357, 451 Aspidosperma gomezianum, 357 Aspidosperma hihrianum, 357 Aspidosperma limae, 357, 414, 448, 677 Aspidosperm longipetiolatum, 357 Aspidosperma marcgravianum, 357,708,7 16 Aspidosperma megalocarpon, 357, 500 Aspidosperma multijorum, 357 Aspidosperma neblinae, 452 Aspidosperma nigricans, 357, 467 Aspidosperma oblongum, 357, 696 Aspidasperma obscurinervium, 357, 452 Aspidosperma olivaceum, 357, 474 Aspidosperma paruifolium, 357 Aspidosperrna polyneuron, 266, 400, 485, 495, 788

Aspidosperma populifolium, 357, 433, 462 Aspidosperma pyricollum, 357 Aspidosperma pyrifolium, 357, 404, 429 Aspidosperma quebrachoblance, 259, 357, 395, 404, 453, 491, 696, 788 Aspidosperma quirandy, 357 Aspidosperma refracturn, 429 Aspidosperma sandwithianum, 357 Aspidosperma spruceanum, 357 Aspidosperma subincanum, 357, 477 Aspidosperma triternaturn, 357 Aspidosperma ulei, 469 Aspidospermatidine, 349, 452 Aspidospermatine, 349, 453 Aspidospermidine, 338, 396 Aspidospermine, 339, 359, 361, 395

B Banisteria caapi, 9, 48 Banisteriopsis inebrians, 48 Benzquinamide, 327 Brucine, 591 Brucinolone, 593 Brucinonic acid, 622 Brucinquinone, 622 Bufotenine, 9 Bufo vulgaris, 17 Burnamicine, 254, 264 Burnamine, 148, 255

C Cabi paraensis, 48 Caffeine, 266 C-Calebassine, 518, 521, 549, 553, 560 Isocalebassine, 561 Callichilia. barter;, 205 Callichilia stenosepala, 205 Callichilia subsessilis, 205 Callichiline, 203, 227 Calligonum minimum, 48 Carica papaya, 13 Calycanine, 583 Calycanthidine, 33, 585 Calycanthine, 581 Isocalycanthine, 582 Calycanthusjoridus L., 582 Calycanthus glaucus, 581 Calycanthus occidentalis, 582 Canembine, 290

SUBJECT INDEX

Canescine, 293 Canthin-6-one,250 Caracurine I, 575 Caracurine 11, 520, 555, 559 Caracurine 111, 575 Caracurine IV, 575 Caracurine V, 520, 546 Caracurine VI, 521 Caracurine VII, 520 Caracurine V I I I , 574 Caracurine IX, 574 Carapanaubine, 82, 355, 502 Carboline, 47 Carosidine, 271 Carosine, 271 Catharanthine, 203, 218, 270, 273 Pseudocatharanthine, 22 1 Catharanthus lanceus, 282 Catharanthus roseus, 272 Catharicine, 271 Celtis reticulosa, 1 Chaleupine, 290 Chandrine, 290 Chanoclavine, 734, 763 Cheirnnthus cheiri, 1 Chimonanthine, 32, 33, 582, 586 Chimonanthus fragrans, 582 Chlorogenine, 160 Chondrondendron tomentosum, 516 Cimicidine, 451, 673 Cimicine, 451 Cinchonamine, 238, 711 Cinchonidine, 243 Cinchonine, 243, 718 Claviceps purpurea, 2, 726 Cleavamine, 203, 218, 273 Cohoba, 16 Colubrines, 642 Compactinervine, 350, 466 Condensamine, 665 Conduramine, 229 Condurine, 229 Condylocarpine, 192, 349, 457 Conocybe cyanopus, 1I Conoduramine, 227 Conodurine, 227 Conopharyngia durissitna, 205, 225 Conopharyngia pachysiphon, 225 Conquinamine, 238, 241 Coprinus micaceus, 8 Coronaridine, 195, 225

853

Corymine, 265 Corynane, 7 16 Corynantheane, 792 Corynantheidine, 63, 685, 696, 716, 792 Corynantheine, 716, 793 Corynanthe macroceras, 696 Corynnnthe pnniculata, 696 Corynanthine, 290, 696 Epi-3-Corynanthine, 696, 704 Corynanthyrine, 164 Corynoxane, 720 Corynoxeine, 720 Corynoxine, 720 Costaclavine, 733 Croceocurine, 575 Cressopterine, 60 Cressopteryx kotschyana, 60 Crossoptine, 60 Cryptolepine, 19 Cryptolepis sanguinolentn, 19 Cryptolepis triangularis, 19 Cryptopine, 638 Curare, 516 C-Curarine, 517, 520, 553, 567 C-Cuararine 111, 521 Curbine, 594 Cylindrocarpidine, 341, 410 Cylindrocarpine, 410

D Deacetylaspidospermine, 395 Deacetylvindoline, 273 Decahydrostrychnidiue, 612 Deformylisoaspidofractine, 344 Demethoxyaspidospermine, 338, 398 Demethoxypalosine, 338, 398 11-Demethoxyreserpine, 293 Demethoxyvallesine, 338, 398 Demethylaspidocarpine, 339, 400 Demethylaspidospermine, 339 Deoxyajmaline, 791 Deoxyajmalone, 793 Deoxysarpagine, 787, 788 Deoxyvomicine, 650 Neodeoxyvomicine Deserpidine, 290, 297 Dichotamine, 348, 449 Dictyoloma incanescens, 16 Dihydroakuammicine, 466, 628 Dihydrocorynantheane, 632, 71 1, 718

854

SUBJECT INDEX

Dihydrocorynantheol, 482 Dihydroellipticine, 352, 477 Nordihydrofluorocurarin, 628 Dihydronorlysergic, 744 Dihydroobscurinervine, 452 Dihydroolivacine, 352, 474 Dihydrostrychnine, 612 C-Dihydrotoxiferine, 518, 548 C-Dihydrotoxiferine I, 521, 539, 553 C-lsodihydrotoxiferine I, 575 Nordihydrotoxiferine I, 521 4,5-Dimethaxycanthin1-6-one, 252 Dimethylpyruvic acid, 749 Dimethyltryptamine, 9 2,3-Dioxonucidine, 621 Dioxonucine dihydrate, 621 Diplorrhyncus condylocarpon, 805 Diplorrhyncus mossambicensis, 357, 463, 485 Dipterine, 9 Ditaine, 160 Ditamine, 161 Donaxarine, 8 Dregamine, 203, 225

E Eburnamenine, 253, 258, 266, 355 Eburnamine, 253, 258, 355, 450,495, 676 Isoebupamine, 253,258,355,450,495,676 Eburnamonine, 253, 276, 355, 495, 676 Echitamidine, 159, 191, 350 Echitamine, 32, 131, 159, 174 Echitenine, 161 Echites scholaris, 159 Eleagnus angustifolia, 48 Elliptamine, 356, 482, 505 Ellipticine, 352, 477 Elliptinine, 356, 504 Elymoclavine, 732, 762 Elymus mollis, 760, 764 Enteramine, 12 Ergine, 730, 746 Erginine, 730, 746 Ergobasine, 730, 748 Ergobasinine, 730 Ergocornine, 732, 759 Ergocorninine, 732, 759 Ergocristine, 731, 759 Ergocristinine, 731, 759 Ergocryptine, 731, 759

Ergocryptinine, 731, 759 Ergoline, 729, 735 Ergometrine, 730, 747 Ergometrinine, 730, 747 Ergonovine, 748 Ergosecaline, 756 Ergosecalinine, 756 Ergosine, 731, 758 Ergosinine, 731, 758 Ergostine, 749 Ergostinine, 749 Ergot, 726 Ergotamine, 728, 755, 757 Aci-ergotamine, 758 Ergotaminine, 730, 757 Aci-ergotaminine, 758 Ergotinine, 727 Ergotoxine, 727 Ergovaline, 755 Ervatamia divaricata, 205 Ervatamia coronaria, 205 Erythrina acanthocarpa, 3 Erythrina cristagalli, 3 Erythrina indica, 3 Erythrina pallida, 3 Erythrina subumbrans, 3 Erythrina variegata, 3 Escherichia coli, 2 Eseramine, 28 Eseridine, 28 Eseroline, 29 Eserine, 28, 136 Evodiamine, 57 Evodia rutaecarpa, 56 Excavatia coccirwa, 288, 357, 482

F Fedamazine, 574 Fendleridine, 451 Fendlerine, 348, 451 Festuclavine, 733, 763 Flavocarpine, 266, 499 Flavopereirine, 499, 533, 688 Flavopicraline, 154 C-fluorocurarine, 521, 548, 552 Norfluorocurarine, 351, 457, 466, 549 C-Fluorocurine, 521 Norfluorocurine, 524 #-Fluorocurine, 521 C-Fluorocurinine, 574

855

SUBJECT INDEX Folicanthine, 33, 136, 582, 586 Formosanine, 61, 69 Fruticosamine, 356, 444 Fruticosine, 356, 444 Fumigaclavine A, 734, 763 Fumigaclavine B, 734, 763

G Gabonine, 223 Gabunia eglundulosa, 205 Gambirine, 6 1 Geissolosimine, 535, 680, 688, 805 Geissoschizine, 683 Apogeissoschizine, 680 Geissoschizoline, 128, 680 Geissospermine, 680, 685 Geissospermum laeve, 679 Geissospermum sericeum, 679 Geissospernaum vellozii, 679, 788, 805 Gelsedine, 94, 112 Gelsemicine, 94, 111 Gelsemidine, 94 Gelseminine, 93 Gelsemine, 93, 95 Gelsemiurn elegans, 94 Gelsemiurn sempervirens, 93 Gelsemiurn rankinii, 95 Gelsemodine, 94 Gelseverine, 94, 115 Geneserine, 28 Geneseroline, 42 Girgensohnia diptera, 9 Gonioma kamassi, 357 Gossypium hirsutum, 13 Gramine, 4 C-Guaianine, 575 Guatambuine, 352, 474

H Haernadictyon amazonicum, 9 Hanadamine, 61 Haplocidine, 348, 450, 676 Haplophytine, 673 Haplophyton cimicidum, 357, 451, 673 Harman, 49, 51, 163, 532 Norharman, 49 Hsrman-3-carboxylic acid, 495

Harmaline, 49 Harmalol, 49 Harmidine, 49 Harmidol, 49 Harmine, 49 Apoharmine, 51 Harmol, 49 Hemitoxiferine I, 520 Herbaceine, 282 Heterophylline, 293 Heteroyohimbane, 707 Hexahydrostrychnine, 612 Hippophae rhamnoides, 13 Hodgkinsine, 582, 588 Hodykinsonia frutescens, 582 Holoinine, 707 Holstiine, 665 Holstiline, 665 Homoesermethole, 36 Homoeseroline, 36 Hordeum vulgare, 4 Hortia arborea, 56 Hortia braziliana, 56 Hortiacine, 57 Hortiamine, 57 Humanine, 95 Hunteracine, 255 Hunteramine, 254 Hunterburnine, 264 Hunteria corymbosa, 265 Hunteria eburnea, 148, 201, 206, 253, 263, 357, 495, 676,696, 716 Hunteria refractum, 357 Hunterine, 255 Huntrabrine, 255, 264 Hydroergotinine, 727 Hydroxyergometrine, 772 80-Hydroxy-1-vineadifformine, 277 Hypaphorus subumbrans, 3

1 Ibogaine, 203, 206 Ibogamine, 203, 213 Iboluteine, 209 Iboquine, 209 Iboxygaine, 203, 206, 211 Indole, 1 Ipomoea tricolor, 747 Isoajmaline, 787

856

SUBJECT INDEX

K a-Ketobutyric acid, 749 Kisantine, 223 Kimvuline, 2 11 Kopsamine, 344, 439 Kopsaporine, 356, 444 Kopseinc, 345 Kopsiafiavidcc, 439 Kopsia fruticosa, 357 Kopsin longijlora, 357, 439 Kopsia prunijormis, 439 Kopsia singapurensis, 357, 444 Kopsij€oreine, 344 Kopsiflorine, 344, 439 Kopsilongeine, 344 Kopsilongine, 344, 439 Kopsine, 346, 420, 441 Kopsingarine, 356, 444 Kopsingine, 356, 444 Kopsinilam, 266, 345, 423, 439 Kopsinine, 265, 344,434 Koumine, 94 Kouminicine, 94 Kouminidine, 94 Kouminine, 94 Kounidine, 95 Kromantine, 505 Kryptocurine, 574

L Lanceine, 282 Lens esculenta, 8 Leptocladine, 9 Leptactina densijtora, 48 Lespedeza bicolor, 10 Leurocristine, 271, 273, 419 Leurosidine, 21 1 Leurosine, 271, 419 Isoleurosine, 271 Leucine, 749 Limapodine, 341 Limaspermine, 341, 414, 677 Lochnera Inncea, 696, 707 Lochneram, 520, 534 Lochnericine, 270 Lochneridine, 270, 350 Lochnerine, 270, 534, 788, 804 Lupinus luteus, 5 Lysergene, 733, 762 Lysergic acid, 734, 744

Zsolysergic acid, 734 Lunzi-lysergic acid-I, 771 Lysergine, 733, 763 Lysergol, 733, 762

M Macralstonidine, 159, 161, 196 Macralstonine, 159, 161, 195 Macrophylline, 161, 195 Macrophylline A, 575 Macusine A, 520, 800, 805 Macusines A, R, and C, 534 Macusine B, 520 Normaeusine-B, 354, 457, 485, 688, Macusine C, 520 Mass spectral data, 376 Mauiensine, 290, 787, 799 6'-Mavacurine, 520 Mayumbine, 72, 709 Melinonine A, 520, 528 Melinonine B, 520, 528 Melinonine E, 574 Melinonine F, 521 Melinonine G, 521 Melinonine H, 575 Melinonine I, 575 Melinonine K, 574 Melinonine L, 574 Melinonine M, 574 Merutia praecox, 582 5-Methoxycanthin-6-one,251 Methoxyellipticine, 356, 480 5-Methoxy-N,N-dimethyltryptamine, 16 5-Methoxy-N-methyltryptamine,16 16-Methoxyminovincine, 420 16-Methoxy-20-0x0-1-vincadifformine, 277 Methoxyvincamine, 260, 276 N-Methylaspidospermidine, 396 0-Methyldeacetylaspidofiline, 433 0-Methyleburnamine, 450, 676 N-Methyltetrahydroellipticine,477 N-Methyltetrahydroharmol, 49 N-Methyltryptamine, 583 Methylkopsinine, 266 Methyl-D-lysergic acid butanolamide, 778 Methyl-D-lysergic acid propanolamide, 778 Micranthine, 293 Mimosa hostilis, 10

SUBJECT INDEX

Minorine, 276 Minovincine, 277, 342, 420 Minovincinine, 277, 342, 420 Minovine, 277, 342, 420 Mitoridine, 290, 787, 799 Mitragyna africana, 59 Mitragyna ciliatn, 60 Mitragyna diuersifolia, 61 Mitragyna inermis, 60 Milrallynu macrophylla, 60 Mitragyna parvqolia, 59 Mitrngyna rubrostipulaeea, 60 Mitragyna rotundifolia, 60 Mitragyna speciosa, 59 Mitragyna stipulosa, 60 Mitragynine, 59, 62 Mitragynol, 61, 85 Mitraphyllane, 66 Mitraphylline, 59, 64, 82, 712 Isomitraphylline, 67, 712 Mitraphyllol, 65 Mitraspecine, 59 Mitraversine, 61 Mitrinermine, 60, 75 Molliclavine, 733 Mossambine, 350, 457, 463 Mostuea buchholzii, 95 Mostuea stimulans, 95 Mucuna pruriens, 12 Muse paradisinca, 13 Muscaria, 17

N Narcissus jonquilla, 1 Neblinine, 452 Neburnamine, 255 Neoajimaline, 293, 787 Neoleurocristine, 271 Neoleurosidine, 271 Neosarpagine, 291, 787 Neostrychnine, 641 Nerifoline, 161 Neurospora, 2 2-Nitrostrychnine, 599 NMR data, 371 Novacine, 665 Nucidine, 621 Nucine, 621 Aponucine, 621 Nucinecarboxylic acid, 621

857

0 Obscuridine, 290 Obscurinervidine, 452 Obscurinervine, 452 Ochropamine, 503 Ochropine, 356, 503 Ochrosia coccinea, 357 Ochrosia elliptica, 357, 477 Ochrosiaformosana, 61 Ochrosia gambir, 6 1 Ochrosia glomerata, 357 Ochrosia guianensis, 60 Ochrosiu katuakamii, 61 Ochrosia moorei, 357 Ochrosia poweri, 288, 357, 503, 708 Ochrosia sandwicensis, 357, 477 Octahydrostrychnine, 612 Olivacine, 225, 352, 474 Ourouparia rhynehophyllu, 60 N-Oxalylanthranilic acid, 623 Oxokopsinine, 266 0x0-1-methylkopsinine, 266 18-0xostrychnine, 630 20-0x0-1-vincadifformine, 277 Oxyvomipyrine, 663

P Palosine, 339, 400, 416 Panaeolus acuminatus. 13 Panaeolus campanulatus, 13 Panaeolus foeniscii, 8, 13 Panaeolus linnaenus, 13 Panaeolus semiovatus, 13 Panaeolus sphinctrinus, 10 Panaeolus subalteatus, 13 Paspalum dist ichum, 74 7 Pnssi$ora foetida, 13 Passijlora incarnata, 48 Pausinystalia trillesii, 696 Pausinystalia yohimbe, 696 Peganum hnrmala, 47, 48 Pelirine, 290 Penniclavine, 733, 762 Isopenniclavine, 733, 763 Pennisetum typhoideum, 764 Peiataceras australis, 250 Perakine, 290, 293, 787, 796 Pereirine, 680 Perivincine, 281 Perivine, 270

858

SUBJECT INDEX

Petalostylis labicheoides, 48 Phalaris tuberosa, 16 Phenylalanine, 749 Physostigmu venenosum, 27 Physostigmine, 28 Isophysostigmine, 28 Physostigmol, 29 Physovenine, 28 Picrnlimw klaineana, 119, 357 Picralima nitida, 119, 357, 788 Picraline, 139, 147 Picralinol, 151 Pierasma ailanthoides, 252 Pierasma crenatu, 250 Piptadenia mucrocarpa, 9 Piptadenia peregrina, 9 pK data, 388 Pleiocarpamine, 201, 266, 344, 355, 504 Pleicarpu mutica, 201, 267, 357, 498, 504 Pleicarpa, 357, 497 Pleiocarpine, 266, 344, 423 Pleiocarpinidine, 266 Pleiocarpinilam, 266, 345, 423, 439 Pleiocarpinine, 266, 344, 423 Pleiocinine, 266 Pleiomutine, 266, 355, 504 Pleiomutinine, 266, 355, 504 Pleurosine, 27 1 Polyneuridine, 354, 485, 535, 788, 800 Polyneuron, 357 Porphyrine, 160 Pouterin spp., 696 Poweramine, 356, 505 Poweridine, 353 Powerine, 356, 505 Prephenic acid, 785 Prestonia nmazonica, 9 Proline, 749 Prosopsis juliJora, 8, 13 Pseudocjnchona africnnn, 696, 720 Pseudocinchom mayumbensis, 712 Psilocybe aztecorum, 10 Psilocybe baeocystis, 11 Psilocybe caerulescens, 10 Psilocybe cyanescens, 11 Psilocybe mexicana, 10 Psilocybe semperviva, 10 Psilocybe yungensis, 12 Psilocybe zapatecorum, 1d Psilocybin, 10 Pubescine, 281

Pultenaea altissima, 4 Purpeline, 290, 787, 799 Pyrifolidine, 337, 339, 395, 404 Pyrifoline, 265, 343, 421, 429 Pyroclavine, 733 Pyruvic acid, 749

Q Quebrachamine, 250, 280, 337, 338, 418 Quebrachacidine, 357 Quebrachidine, 355, 491, 801 Quebrachine, 696 Quinamine, 238 Isoquinamine, 241 Quidoline, 20 Quinine, 718

R Raubasine, 293 Raubasinine, 293 Raugalline, 293, 787 Raugustine, 291, 297 Rauhimbine, 293, 696 Isorauhimbine, 293, 696 Raujemidine, 291, 297 Raujemidine N-oxide, 291, 297 Raumitorine, 291, 709, 714 Raunamine, 291 Raunescine, 291, 297 Isoraunescine, 291, 297 Rauniticine, 291, 707, 715 Raunitidine, 291, 707, 715 Isoraunitidine, 709 Raupine, 293 Rauvanine, 291, 707, 714 Rauvomitine, 291, 787, 799 Rauwolfia a,ffiffilais, 289 Rnuwolfia amsoniaefolia, 289 Rauwolfia bahiensis, 289 RautcodJia heddomei, 289 Rauwolfia boliviana, 289 Rauwolfia caffra, 289 RauwolJa cnmbodinna, 289, 708 Rauwolfia canescens, 160, 288 Rauwolfia cuhan.n, 289 Rauwolfia cunaminsii, 289 Rauwolfia decurva, 289, 708 Rauwolfia degeneri, 289 Rauwolfia densi$ora, 289

859

SUBJECT INDEX

Rauwolfia discolor,289 Rauwolfia fruticosa, 289 Rauwolfia grandijlora, 289 Rauwolfia hirsuta, 289 Rauwolfia heterophylla, 289 Rauwoljia indecora, 289 Rauwolfia inebrians, 289 RauwolJcajavanica, 289, 707 Rauwolfia larnarckii, 289 Rauwolfia ligustrina, 298 Rauwoljia littoralis, 289 Rauwoljia macrophylla, 289 Rauwolfia rnannii, 289 Rauwoljia mauviensis, 289 Rauwolfia nzicranthu, 289 Rauwolfia momhasiuna, 289 Rauwoljia nana, 289 Rauwolfia natalensis, 289 Rauwoljia nitidu, 289, 7 15 Rauwolfia obscura, 160, 289 Rauwoljia paraensis, 289 Rauwoljia pentaphylla, 289 Rauwoljia perakensis, 289, 798 Rauwolfia rosea, 289 Rauwoljia salicofoliu, 289 Rauwolfia sandwicensis, 289, 708 Rauwolfia sarapiquensis, 289 Rauwolfia schueli, 289 Rauwolfia sellowii, 160, 289 Rauwolfia semperjlorens, 289 Rauwolfca serpentina, 2, 287, 707, 786, 799 Rauwolfia sprucei, 289 Rauwolfia sumatrana, 289, 707 Rauwolfia spp., 696 Rauwoljia ternifolia, 289 Rauwolfia tetraphylla, 289 Rauwolfia verticillata, 289 Rauwolfia viridis, 289 Rauwoljiu vomitoria, 160, 288, 707, 798 Rauwolfine, 291, 787 Rauwolfinine, 291, 787, 799 Rauwolscine, 293, 696, 704 Epi-3-rauwolsoine, 393, 696 Recanescine, 293 Refractalam, 429 Refractidine, 265, 343, 421, 429 Refractine, 344, 421, 434 Renoxidine, 291 Rescidine, 291, 297 Rescinnamine, 29 1 Reserpiline, 291, 353, 482, 708

Isoreserpiline, 291, 353, 482, 502, 708 Isoreserpiline-4-indoxyl, 503 Neoreserpiline, 707 Reserpine, 2, 160, 270, 277, 291 &Reserpine, 291, 297 Reserpinine, 281, 291, 353, 482, 709 Isoreserpinine, 291, 709 Reserpoxidine, 293 Retuline, 665 Rhazidine, 505 Rhazinine, 357, 505 Rhazya stricta, 357, 396, 419, 788 Rhetsine, 57 Rhetsinine, 57 Rhynchophyllane, 77 Isorhynchophyllane, 77 Rhynchophylline, 60, 75, 82 Isorhynchophylline, 61, 75 Rhynchophyllol, 76 Rivea corymbosa, 747, 764 Robinia pseudacacia, 1 Rotundifoline, 61, 85 Isorotundifoline, 61, 85 Rubradinine, 60 Rutaecarpine, 56

5 Sandwicensine, 291, 787 Sandwicine, 291, 787, 797 Sarpagine, 281, 291, 293, 534, 804 Secale cwnutum, 726 Semperflorine, 292 Sempervirine, 94, 107 Seredarnine, 292 Seredine, 292 Serotonin, 12, 773 Serotonin antagonism, 773 Serpentidine, 293 Serpentine, 2, 166, 270, 292, 293 Serpentinine, 292, 293, 708, 713 Serpine, 293 Serpinine, 293 Setoclavine, 733, 762 Isosetoclavine, 733, 762 Sickingia rubra, 48 Sitsirikine, 270 Speciofoline, 61, 85 Spegazzinidine, 340, 405 Spegazzinine, 340, 405 3-Spirocyclopropanooxindole,71

860

SUBJECT INDEX

Spermostrychnine, 663 Stemmadenine, 351, 457 Stemmadenia donnell-smithii, 205, 337, 357, 457 Stemmadenia galeottiana, 205 Stenamadenia pubescens, 357 Stemmadenia tomentosa, 357 Stropharia cubensis, 10 Strychanol, 618 Isostrychanol, 618 Strychanone, 545, 617, 633 Strychene, 466 Strychine, 591 Isostrychnic acid, 600 Strychnidine, 599, 619 Strychnidone, 628 Strychnine, 642 Isostrychnine-I, 602 Isostrychnine-11, 602 Pseudostrychnine, 630, 636 Strychninolic acid, 615 Strychninolone, 593 Strychninolone-a, 604 Strychninolone-b, 604 Strychninolone-c, 604 Strychnone, 597, 628 Strychnos mlinoniana, 48, 160 Strychnos toxifern, 485, 788 Strychnospermine, 663 Symplocarpus foetidus, 13 Symplocos racemosa, 48 Syrosingopine, 311

T Tabernaemontana alba, 357, 416 Tabernaemontana australis, 205 Tabernaemontana eitrifolia, 357 Tabernaemontana c o r o w r i a , 205 Tuberrzaemontana oppositijolia, 205 Tabernaemontana psychotrifolia, 205, 357, 474 Tabernaemontanine, 203, 225, 228 Tabernanthe iboga, 205 Tabernanthine, 203, 213 Tabersonine, 337, 342 Tetrabenzazine, 327 Tetrahydroalstonine, 159, 162, 270, 282, 292, 482, 707, 709 Tetrahy dronorharman, 49 Tetrahydroharmol, 49

Tetrahydrostrychnine, 612 Tetrahydroisoyobyrine, 107 Tetraphyllicine, 281, 292, 293, 787, 798 Tetraphyllicinoce, 799 Tetraphylline, 292, 709 Thebaine, 32 Thlaspi arvense, 1 Thrombocytin, 12 Tombozine, 535, 688, 788, 805 Tonduzia longifolfolia,288, 788 Toxiferine I, 518, 521, 539 Toxiferine 11, 518 Toxiferine 111, 574 Toxiferine IV, 567 Toxiferine VIII, 574 Toxiferine X I I , 575 Tryptamine, 8, 596 Tryptophan, 1 Tubifolidine, 466 Tubifoline, 466 d-Tubocurarine, 516 Tuboflavine, 355,497 Tubotaiwine, 349, 462 Tuboxenin, 419

U Uleine, 362 Uncaria gambier, 61 I’ncaricz kawakamii, 60 Uncaria rhynchophylla, 60 Uncaria tomentosa, 60 Uncarine-A, 61, 69 Uncarine-B, 61, 69 Urtica dioica, 13 Uterine contraction, 773 UV data, 368

v Valine, 749 Vallesia dichotoma, 288, 357, 400, 449 Vallesia glabra, 357, 400 Vallesine, 339, 395, 400 Vasoconstriction, 773 Velbanamine, 203, 218, 273 Vellosimine, 535, 687, 680, 787, 805 Vellosiminol, 680, 687, 788 Vellosine, 680 Venenatine, 202 3-Isovenenatine, 202 Villalstonine, 159, 161, 194

86 1

SUBJECT INDEX

Vinblastine, 419 Vinca difformis, 280, 788 Vinca herbacea, 282 Vinca Zaneea, 160, 282 Vincn major, 280, 788 Vinca minor, 257, 276, 288, 337 Vinca rosea, 160, 205, 270, 272, 288, 419, 788 Vinca spp., 696 Vincadifformine, 277, 281, 337, 342, 397, 419 Vincadine, 276, 280 Vincaine, 270 J7incaleukoblastine, 271, 273 Vincamajine, 133, 281, 535, 788, 800, 804 Vincamajoreine, 281 Vincamajoridine, 270, 281 Vincamedine, 133, 281, 494, 535, 800 Vincamicine, 271 Vincamidine, 277 Vincamine, 276, 281 Vincaminine, 261, 276 Vincaminoreine, 276, 280, 337 Vincaminoridine, 277 Vincaminorine, 276 Vincamirine, 281 Vincanidine, 282 Vincanine, 282 Vincanorine, 257, 276 Vincareine, 261, 276 Vincarodine, 271 Vinceine, 270 Vincine, 260, 276 Vincinine, 261, 276 Vincoridine, 277 Vincorine, 277 Vindolicine, 271 Vindolidine, 271 Vindoline, 271, 273, 337, 342, 419 Vindolinine, 271, 337, 342, 419 Vinine, 281 Virosine, 271 Voacafricine, 226 Voacafrine, 226 Voacamidine, 227 Voacamine, 225, 229 Voacanga africam, 205 Voacanga bracteata, 205 Voacanga chalotiana, 205, 491, 788 Voacauga dregei, 205 Voacanga schweinfiirthii, 205

Voacanga thouarsii, 205 Voacangine, 215 Voachalotine, 203, 226, 491, 535, 788, 804, 806 Voacorine, 227 Voacristine, 217 S'oacryptine, 203, 217 Vobasine, 203, 226, 228 Vobtusine, 227 Vomalidine, 292, 787, 798 Vomicidine, 647, 649 Vomicine, 638, 647 Isovomicine, 647, 650 Norvomicinic acid, 650 Vomilenine, 293, 787, 796 Vomipyrine, 663

w Wieland-Gumlich aldehyde, 126, 542, 605 U'uchuyine, 55

x Xanthocurine, 574

Y Yobrine, 107, 529 Yobyrone, 107 Yohimbane, 695 Yohimbine, 282, 292, 482, 695, 696, 702 a-Yohimbine, 293, 696 /3-Yohimbine, 292, 696 Epi-3-/3-yohimbine, 696 X-Yohimbine, 292, 704 Alloyehimbine, 696, 705 3-Epialloyohimbine, 705 3-Epi-a-yohimbine, 292, 696 3-Epi-x-yohimbine, 292 Pseudoyohimbine, 696 Yohimbol, 530 Yohimbone, 530

Z Zanthoxylum budrungo, 56 Zanthoxylum rhetsa, 56 Zanthoxylum suberosum, 250 Zygophyllum fabago, 48

This Page Intentionally Left Blank