Studies in Natural Product Chemistry, Volume 18: Stereoselective Synthesis, Part K (Studies in Natural Products Chemistry)

studies m Natural Products Chemistry Volume 18 Stereoselective Synthesis (Part K) Studies in Natural Products Chemist...

Author: Atta-Ur-Rahman

101 downloads 2003 Views 38MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

studies m Natural Products Chemistry Volume 18 Stereoselective Synthesis (Part K)

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

Vol. 1 Stereoselective Synthesis (Part A) Vol. 2 Structure Elucidation (Part A) Vol. 3 Stereoselective Synthesis (Part B) Vol. 4 Stereoselective Synthesis (Part C) Vol. 5 Structure Elucidation (Part B) Vol. 6 Stereoselective Synthesis (Part D) Vol. 7 Structure and Chemistry (Part A) Vol. 8 Stereoselective Synthesis (Part E) Vol. 9 Structure and Chemistry (Part B) Vol.10 Stereoselective Synthesis (Part F) Vol.11 Stereoselective Synthesis (Part G) Vol.12 Stereoselective Synthesis (Part H) Vol.13 Bioactive Natural Products (Part A) Vol.14 Stereoselective Synthesis (Part I) Vol.15 Structure and Chemistry (Part C) Vol.16 Stereoselective Synthesis (Part J) Vol.17 Structure and Chemistry (Part D) Vol.18 Stereoselective Synthesis (Part K)

studies in Natural Products Chemistry Volume 18 stereoselective Synthesis (Part K)

Edited by

Atta-ur-Rahman

H.EJ. Research Institute of Chemistry, University of Karachi, Karachi 75270, Pakistan

1996 ELSEVIER Amsterdam - Lausanne - New York - Oxford - Shannon - Tokyo

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

Transferred to digital printing 2005

ISBN: 0-444-82458-8 © 1996 Elsevier Science B.V. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science B.V., Copyright & Permissions Department, P.O. Box 521,1000 AM Amsterdam, The Netherlands. Special regulations for readers in the USA - This publication has been registered with the Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside of the USA, should be referred to the copyright owner, Elsevier Science B.V., unless otherwise specified. No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. This book is printed on acid-free paper. Printed and bound bv Anton\ Rowe Ltd, Eastboume

FOREWORD The present volume of this series should again provide highly interesting articles written by some of the most eminent organic chemists today. They range from stereocontrolled synthesis of complex natural products to structural studies on a variety of different types of natural products. It is hoped that this volume will be received with the same enthusiasm by the readers as the previous ones of the series. I wish to express my thanks to Miss Farzana Akhter and Syed Ejaz Ahmed Soofi for their assistance inthe preparation of the index. I am also grateful to Mr. Wasim Ahmad and Mr. Ahmed Ullah for the typing work and Mr. Mahmood Alam for secretarial assistance. Prof. Atta-ur-Rahman H.E.J. Research Institute of Chemistry University of Karachi

This Page Intentionally Left Blank

Vll

PREFACE Further developments in organic chemistry, natural products chemistry, and associated fields continue unabated. This high level of activity lies in sharp contrast to statements made during the past two decades by some prognosticators who had quite mistakenly predicted the rapidly approaching obsolescence of these fields of investigation. These predictions were based upon organic chemistry having reached a very mature level of development at a time when new areas of scientific inquiry were opening. Nevertheless, organic chemistry remains as vital and as active as ever in laboratories around the world. This continued activity may be attributed to many factors, including the development of new screening procedures for biologically active compounds, improvements in spectroscopic methods for determination of molecular structure, the availability of new, highly selective and often asymmetric methods for the synthesis of ever more complex, highly functionalized structures, and the applications of computer technology to chemistry. Another driving force for further work in organic chemistry continues to be the search for more effective pharmaceutical agents to treat many diseases such as cancer and other maladies that continue to plague humankind. In this same vein, continued searches are underway for new antibiotics to combat dangerous infectious bacterial strains that have become resistant to previously developed antibiotics. Organic chemistry has also been widely adopted as a tool for use in other areas of science, most notably in the biological realm wherein specially synthesized compounds can, for example, be used to probe the molecular details of cell function. In the most recent volume of this well-established series. Professor Atta-ur-Rahman again brings together the work of several of the world's leading authorities in organic chemistry. Their contributions demonstrate the rapid, ongoing development of this field by illustrating many of the latest advances in synthetic methods, total synthesis, structure determination, biosynthetic pathways, and biological activity. The opening chapter presents an overview of strategies for the synthesis of several classes of natural products with an emphasis on complex polycyclic systems. The next several chapters discuss the synthesis of specific classes of compounds, including morphine, polyketides, acetogenins, nonactic acid derivatives, complex spirocyclic ethers, 8-lactam and pyridone derivatives, inositol phosphates, sphingolipids, brassinosteroids, Hernandia lignans, and dimeric steroidal pyrazine alkaloids. Structure determination and biological function provide additional themes through many of these chapters. On the other hand, structure is discussed more exclusively in chapters on liverwort sesquiterpenoids, gymnemic acids, compounds of the Celastraceae plant family, fungal and protozoan glycolipids, and coumarins. Finally, the ever stronger links between chemistry and biology are reinforced by chapters on the origin and function of secondary metabolites, bioactive conformations of gastrin hormones, and immunochemistry. Professor Atta-ur-Rahman is to be congratulated for bringing together the present set of contributions as a continuation of this outstanding series. He has again met the goal of this series in demonstrating the strength, the vitality, and the diversity of organic chemistry as a central field of scientific investigation. Paul Helquist University of Notre Dame January 1996


CONTRIBUTORS

G. Adam

Department of Natural Products Chemistry, Institute Biochemistry, Weinberg 3, P.O. Box 250, D-06018 Halle/S.

of

Plant

N.L. Alvarenga

C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife, Espana.

Masao Arimoto

Osaka University of Pharmaceutical Sciences, 10-65 Kawai 2-Chome, Matsubara 580, Japan

Nancy S. Barta

Department of Chemistry, University of Michigan, Ann Arbor, Michigan 48109-1055, U.S.A.

I.L. Bazzocchi

C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife, Espana.

Eliana Barreto Bergter

Instituto de Microbiplogia da UFRJ, Centro de Ciencias da Saude-blocol 21.944-970-Cidade ijniversitaria, Rio de Janeiro-RJ

Maria Helena S. Villas Boas

Instituto de Microbiologia da UFRJ, Centro de Ciencias da Saude-blocol 21.944-970-Cidade Universitaria, Rio de Janeiro-RJ

Gabor Butora

Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061

Andre Cav6

Universite Paris-sud, Faculte de Pharmacie de Chatenay-Malabry, Laboratoire de Pharmacognosie, URA 1843 CNRS (BIOCIS)

Carsten Christophersen

Department of General and Organic Chemistry, The H.C. 0rsted Institute, K0benhavns Universitet, Universitetsparken 5, DK-2100 Copenhagen, Denmark

Helmut Duddeck

Institut fur Organische Chemie, Universitat Hannover, Schneiderberg IB, D-3000 Hannover 1, Germany

Bruno Figadere

Universite Paris-sud, Faculte de Pharmacie de Chatenay-Malabry, Laboratoire de Pharmacognosie, URA 1843 CNRS (BIOCIS)

Ian Fleming

Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 lEW, U.K.

Stephen P. Feamley

Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061, U.S.A.

A. Ganesan

Centre for Natural Products Research, Institute of Molecular and Cell Biology, National University of Singapore, 10 Kent Ridge Cresent, Singapore 0511

Manfred Gemeiner

Veterinar-Medizinische Universitat, Wien, Austria.

Sunil K. Ghosh

Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 lEW, U.K.

A.G. Gonzalez

C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife Espana.

Andrew G. Gum


Maria Helena

Institute de Microbiologia da UFRJ, Centro de Ciencias da Saude-blocol 21.944-970-Cidade Universitaria, Rio de Janeiro-RJ

GerdHiibener

Max-Planck-Institut fiir Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany

Tomas Hudlicky


Akitami Ichihara

Faculty of Agriculture, Hokkaido University, Kitata 9, Nishi 9, KJTAKU, Sapporo 060, Japan

Tadao Kamikawa

Department of Chemistry, Kinki University, Faculty of Science & Technology, Kowakae, Higashi, Osaka 577, Japan

Jiirgen Lutz


Shashi B. Mahato

Indian Institute of Chemical Biology, A Unit of C.S.I.R. Govt, of India, 4, Raja S.C. Mullick Road, Jadavpur, Calcutta-700-032, India

B. Mikhova

Institut fiir Organische Chemie, Universitat Hannover, Schneiderberg IB, D-3000 Hannover 1, Germany

Luis Moroder


Johann Mulzer

Institut fur Organische Chemie der Freien Universitat Takustraê 3, D14195, Berlin, Germany

O. Muhoz

Universidad de Chile, Facultad de Ciencias Casilla 653 Santiago, Chile

XI

S. Nishibe

Osaka University of Pharmaceutical Sciences, 10-65 Kawai 2-Chonie, Matsubara 580, Japan

Hideaki Oikawa

Department of Bioscience and Chemistry, Hokkaido University, Sapporo 060, Japan

Leo A. Paquette

Department of Chemistry, The Ohio State University, 120 West 18th Avenue, Columbus, OH 43210-1173, U.S.A.

A. Penaloza

Universidad de Chile, Facultad de Ciencias Casilla 653-Santiago, Chile

A. Porzel


A.G. Ravelo

C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La Laguna-Tenerife Espana.

J. Schmidt


of

B. Schneider


of Plant

Michele R. StabiV.


John R. Stille

Chemical Process Research and Development Eli Lilly and Company, Indianapolis, Indiana 46285-4813, U.S.A.

Motoo Tori

Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima 770, Japan

H. Toshima

Faculty of Agriculture, Hokkaido University, Kita 9, Nishi 9, KTTA-KU, Sapporo 060, Japan

B. Voigt


Yutaka Watanabe

Faculty of Engineering, EHIME University, 3, Bunkyo-cho, Matsuyama 790, Japan

H. Yamaguchi

Osaka University of Pharmaceutical Sciences, 10-65 Kawai 2-Chome, Matsubara 580, Japan

of Plant

Plant

of Plant


Xlll

CONTENTS Foreword

v

Preface

vii

Contributors

ix

Strategies for the StereocontroUed De Novo Synthesis of Natural Products L.A. PAQUETTE

3

A Historical Perspective of Morphine Synthesis T. HUDLICKY, G. BUTORA, S.P. FEARNLEY, A.G. GUM AND M.R. STABILE

43

New Developments in the Synthesis of Polyketides and of Chiral Methyl Groups J. MULZER

155

Total Stereoselective Synthesis of Acetogenins of Annonaceae : A New Class of Bioactive Polyketides B.nGADERE AND A. CAVE

193

The Synthesis of Nonactic Acid. Its Derivatives and Nonactin itself I. FLEMING AND S.K. GHOSH

229

Total Synthesis of Bioactive Natural Spiroethers, Tautomycin and Oscillatoxin D A. ICHIHARA, H. OIKAWA AND H. TOSHIMA

269

Aza-Annulation of Enamine Related Substrates with a,p-Unsaturated Carboxylate Derivatives as a Route to the Selective Synthesis of 5-Lactarns and Pyridones J.R. STBLLE AND N.S. BARTA

315

Selective Reactions and Total Synthesis of Inositol Phosphates Y.WATANABE

391

Synthesis of Phytosphingolipids T. KAMIKAWA

457

New Developments in Brassinosteroid Research G. ADAM, A. PORZEL, J. SCHMIDT, B. SCHNEIDER AND B. VOIGT

495

Structure Elucidation and Synthesis of the Lignans from the Seeds of Hemandia M. ARIMOTO, H. YAMAGUCHI AND S. NISHIBE Studies on the Absolute Configuration of Some Liverwort Sesquiterpenoids M. TORI

ovigera L. 551

607

XIV

Bioactive Gymnemic Acids and Congeners from Gymnema sylvestre S.B. MAHATO

649

TM

Theory of the Origin, Function, and Evolution Secondary Metabolites C. CHRISTOPHERSEN

677

The Celastraceae from Latin America Chemistry and Biological Activity O. MUNOZ, A. PENALOZA, A.G. GONZALEZ, A.G. RAVELO, I.L. BAZZOCCHI AND N.L. ALVARENGA

739

Structural Chemistry of Glycolipids from Fungi and Protozoa E.B. BERGTER AND M.H.S.V. BOAS

785

Potential Bioactive Conformations of Hormones of the Gastrin Family L. MORODER AND J. LUTZ

819

When Two Steroids are Better than One : The Dimeric Steroid-Pyrazine Marine Alkaloids A. GANESAN

875

Human IgGl Hinge-Fragment as a Core Structure for Immunogens L. MORODER, G. HUBENER AND M. GEMEINER

907

^^C-NMR Spectroscopy of Coumarins and their Derivatives : A Comprehensive Review B. MIKHOVA AND H. DUDDECK

971

Subject Index

1081

Stereoselective Synthesis


Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 © 1996 Elsevier Science B.V. All rights reserved.

Strategies for the Stereocontrolled De Novo Synthesis of Natural Products Leo A. Paquette

LEO A. PAQUETTE In no area of chemistry is stereoselectivity more often a necessary consideration than in the synthesis of structurally complex natural products. A practitioner in this field must be knowledgeable not only of many useful transformations and the mechanistic principles underlying their ability to bring about controlled chemical change, but also be capable of deploying the vast array of available reagents in that chemoselective, regioselective, and stereoselective manner appropriate to the target molecule under consideration. Although the achievements of the last three decades have in the minds of many caused these very important prerequisites to become highly developed, the demands placed on synthetic chemists are hardly exhausted. A recently pubhshed treatise entitled "Stereocontrolled Organic Synthesis" addresses many of the relevant issues from the viewpoint of how the field can expect to develop well into the 21st century [1]. There exists no doubt that the pace of progress has been breathtaking. Certainly, the fantastic advances in NMR spectroscopy and X-ray crystallography have greatly reduced the time needed to determine the structures of newly synthesized compounds. Notwithstanding, effective strategies remain the province of synthetic organic chemists, and it is in this arena where stereochemical elements are deployed with remarkable sophistication. In this chapter, we welcome the opportunity to provide an overview of some of the stereocontrolled syntheses successfully brought to completion in this laboratory in recent years. A. THE LYCOPODIUM ALKALOIDS MAGELLANINE AND MAGELLANINONE In a series of insightful papers, Castillo and MacLean established that the club mosses Lycopodium magellanicum and Lycopodium paniculatum produce alkaloids possessing structural features distinctively different from other metabolites known to arise from these and related sources. The three members of this small and unique subset were identified to share in common a central bicyclo[3.3.0]octane unit to which a functionalized cyclohexane and an Nmethylpiperidine ring were laterally fused. The occurrence of magcllaninc (1) [2],

magellaninone (2) [3], and paniculatine (3) [4] in nature has attracted significant attention [5-7], since all three represent challenging objectives for total synthesis.

HgC-N

H3C-N.

Our successful acquisition of both 1 and 2 gave particular attention to the requirement for strict stereochemical control at six of the eight carbons of the diquinane substructure by retrosynthetic disassembly of the two six-membered rings. The broadly defined goals were therefore to realize proper cyclohexannulation of enone 4 [8] in advance of a tandem vicinal difunctionalization process that would establish the heterocyclic ring. Disconnection of strategic bonds in this manner provided long term for the development of a new MichaelMichael ring-forming sequence as well as a novel means for incorporating the piperidinering[9]. The most expedient means for incorporating ring A involved the K2CO3promoted condensation of 4 with ethyl 5-ethoxy-3-oxo-4-pentenoate in tetrahydrofuran and ethanol containing alumina as a surface catalyst at room temperature. As a consequence of the somewhat folded conformation of 4, the face selectivity of the first conjugate addition proceeds syn to the angular hydrogen as in 5 for obvious steric reasons (Scheme I). Stereocontrol is not sacrificed in proceeding from 5 to 6 because the acceptor side chain is already positioned on the p surface and the diquinane segment possesses a latent thermodynamic preference for becoming cis- and not trans-fused. As a consequence, 7 is obtained in good yield. Following acid-catalyzed elimination of ethanol in 7, it proved possible to reduce the cyclopentanone carbonyl in 8 chemoselectively as expected. Noteworthy at this stage is the fact that borohydride attack occurs stereoselectively from the p face. Silyl protection of the resulting a alcohol afforded 9 and set the stage for unmasking of the second five-membered ring carbonyl. Recourse to thallium nitrate as the means for removing the dithiane moiety gave 10. The advantage of this strategy was that both ketone functional groups in 10 could be simultaneously modified now and at a later stage. Although the reduction of 10 with diisobutylaluminum hydride was not 100% stereocontroUed at -78 °C, the unwanted minor diastereomers could be separated chromatographically and reconverted quantitatively to 10 for recycling

O

o

o

COOEt

EtO

Qv

KgCOa.AlgOa,

>r^ Z^^^COOEt

COOEt

lb

THF. EtOH 25 «C

OH EtO,

COOEt

1. NaBH^, EtOH, CH2CI2,

(TsOH)

QOC

2. TBSOTf. imid, CH2CI2, n

TBSO

^

OH TI(N03)3

1. MOMCI. (/-POgNEt. CH2CI2

(/-Bu)2AIH. CH2CI2. -78 °C

TBSO

^

»

4

PCC/AÔa, CHgCI^ rt

MeOH, THF

2. B L ^ N ^ F "

OH 11

OMOM

HMPA. 3A MS rt

OMOM

OMOM 1. LiN(SiMe3)2, THF; PhSeCI 2. H202.py

CH2CI2 OMOM 12

OMOM

OMOM 13

14

Scheme I purposes. This simple tactic raised the efficiency with which 11 was produced to the 76% level and permitted its ready conversion via 12 to 13. It is significant in the context of what is to follow that hydride delivery to both carbonyl groups in 10 once again operates with a dominant p-face kinetic preference. Once 13 was in hand, enone 14 was generated through adaptation of conventional organoselenium technology for the purpose of incorporating the piperidine ring properly. The recognized propensity of the anion of (trimethylsilyl)acetonitrile to exhibit 1,4-addition to conjugated enones [10] was applied to 14. To our satisfaction, the diastereofacial guidance available to this reagent was identical to that provided to the reducing agents utilized earlier. Furthermore, the enolate intermediate thus formed proved entirely amenable to stereoselective C-acylation with methyl cyanoformate [11] and fumished 15 in a single laboratory operation (Scheme II). As a direct consequence of the relatively high acidity of the proton

OMOM

/

1. LICH(CN)SIMe3 HMPA, THF

^.

2. KF.aqCHgCN, OMOM 3. LDA, NCCOOMe

1. NaBH4, MeOH, -20 "C

MeOOC

2. COC^.py, THF; PhSeH

I

MeOOC. -^

OMOM 16

15

14

OMOM

/

O It PhSeCO

OMOM

(MegSij^SIH. AIBN'. CBHC A

MeOOC \ OMOM

NaBH4. CoClg, MeOH; *KOH, MeOH; HgO*

/ O,

17

Scheme II positioned central to the p-keto ester subunit of 15, enolization is facile. It is therefore not known whether the a orientation of the carbomethoxy substituent is the result of kinetic or thermodynamic control. Suffice it to indicate, however, that this stereogenic center has been improperly set and requires subsequent inversion. Since utilization of the ketone carbonyl was now complete, its removal was implemented via an efficient three-step sequence involving reductive cleavage of the derived selenocarbonate with tris(trimethylsilyl)silane [12] under free radical conditions [13]. With the acquisition of 17 in this manner, the serviceability of the reagent produced by adding sodium borohydride to cobaltous chloride for chemoselective reduction of the nitrile group [14] was assessed. Indeed, treatment of 17 in this manner, followed directly by basification with potassium hydroxide in methanol, secured 18. In this step as well as in the subsequent progression to the N-methyl derivative 19, no epimerization was seen within ring A. To our mind, the enolate of 19 should exhibit a decided kinetic bias for kinetically controlled protonation on its a face because of the steric encumbrance associated with p proton delivery. In actual fact, rapid introduction of its lithium salt into a 1:4 mixture of water and tetrahydrofuran at -78 °C resulted in its quantitative conversion to 20 (Scheme HI). Once the MOM groups had been removed, controlled oxidation with manganese dioxide led to 21, a very pivotal intermediate. To arrive at magellaninone (2), 21 was treated with methyllithium and the resulting unprotected diol 22 was directly reduced with lithium aluminum hydride. Subsequent Jones oxidation proceeded with the customary allylic rearrangement. The plan now called for producing mageUanine (1) by standard borohydride reduction of 2. However, in contrast to the directionality observed earlier for a

carbonyl group in this locale, only the p alcohol 23 was obtained perhaps because of the presence of the fused piperidine ring on the convex surface. In any event, Mitsunobu inversion [15] was successful in delivering the targeted alkaloid and in demonstrating that these unusual Lycopodium alkaloids can indeed be prepared in stereocontrolled fashion by three-fold annulation of 2-cyclopentenone. OMOM

0

/ {

\ \ ^^^ H 20 HaC,

1. UAIH4. THF. A

1\. » •^ H

OMOM

1. HOI. HgO.THF 2. MnOg, CHCb

0^ J H^--

\

if

\ ^•^*H

OH

21

-OH

CH3M

THF ^ > -78 °C

)

2. Jones oxid.

NaBH4 EtOH

OH

22

1. PhgP. DEAD HCOOH, THF 2. 10%KOH, H2O

H3C' 23

Scheme HI B. THE MOST HIGHLY CONDENSED PENTALENOLACTONE ANTIBIOTIC Ecological concerns have prompted chemists to become increasingly "atomeconomic" in their synthetic pathways. The goals associated with this concept are near-perfectly realized in the course of efficient isomerization reactions. Accordingly, we have incorporated a number of stereocontrolled rearrangements into our synthetic undertakings. Illustrated here is proper application of the oxadiTC-methane rearrangement to a total synthesis of pentalenolactone P methyl ester (24b) [16], the stable esterified form of naturally occurring 24a. Pentalenolactone P is the only member of the pentalenone family of antibiotics to possess a fused three-membered ring, which notably resides on the highly congested concave luifaeo or me moidouio [17].

24a.R - H b. R - C h ^

The central stereochemical issue in any projected synthesis of 24b is the establishment of a trans relationship between the cyclopropane and lactone rings. This being the case, we set out to develop a convenient route to the (J^y-unsaturated ketone 34 in advance of its triplet-state photoisomerization, which was projected [18] to generate the tetracyclooctanone 35 (Scheme IV). The Diels-Alder reactions of 1-methylcycloheptatriene (25) [19] with fumaroyl chloride followed by indirect hydrolysis was capable of producing large amounts of the dicarboxyUc acid 26. Necessary chemoselective differentiation of the functional groups in 26 was made feasible by oxymercuration. By this means, the role exercised by the methyl substituent on the steric course of the [4+2] cycloaddition was capitalized upon to considerable advantage. Moreover, the strained nature of lactone 27 allowed for smooth conversion to diol 28 by reduction with sodium borohydride, thereby effectively accomplishing suitable oxygenation of the proximal carbon of the original etheno bridge in 26. Buoyed by the ease of this oxygen atom transfer, we proceeded to generate the acetonide 29 and to advance the synthesis by implementing conversion to a,p-unsaturated ester 30 through deployment of oxidative elimination involving the a-phenylseleno derivative. Evidently, the significant strain introduced upon installation of the double bond accelerates acetal hydrolysis. Attention was next directed to regioselective chain extension and this maneuver was accomplished by sequential exhaustive silylation, reduction with diisobutylaluminum hydride, perruthenate oxidation to the aldehyde, and Wittig olefination. Once the conjugated diene 32 had been produced, it proved an easy matter to effect its conversion to 33 by regiocontroUed hydroboration and selective pivaloylation of die primary hydroxyl groups. Perruthenate oxidation of 33 efficiently delivered 34 whose irradiation in acetone solution with 3000 A light proceeded with full retention of stereochemistry to introduce a second cyclopropane ring as in 35. The stmctural assignment to 35, initially deduced on spectroscopic groups, was corroborated by X-ray analysis of the highly crystalline diol 36 produced by saponification.

a

CICO CHa

' •

\

^ ^COCI toluene, A

1. Hg(OAc)2. MeOH. rt;

COOH

2. CH3OH. py 3. NaOH, MeOH, H2O

NaBH^. -78 °C 2. CH2N2

COOH

25

COOMe

»>

26

NaBH4

Î^^X^COOMe

^

MeOH. rt

OH

COOMG

TsOH, THF

l^

I

OH

29

1. TBSCI.

VÔH

YT

2. Dibal4 CH2CI2

1. TRAP. N M O

Y\

OTBS OTBS

30

1. 9-BBN; NaBOg 2. PvCI. EtgN

OTBS 31

32

OPv

OPv (n-Pr)4NRu04,

3000 A

^

^>

3. 48% HF, CH3CN; PvCI, Et^. DMAP

^.

2. MCPBA NaHCOg. CH2CI2

28

J\

1.LDA,THF; PhSeBr

(CH3)2C(OMe)2

OPv OH

NMO, 4A MS,

33

acetone

34

OH

PvO

\

OPv

^

OPv NaOH H2O, Eton

OH

p^^^'^ô 36

35

Scheme IV The developments described above were predicated upon the expectation based on less highly substituted examples that the dissolving metal reduction of 35 would likewise result in regioselective rupture of the central bond of the threemembered ring conjugated to the carbonyl. No guidance was available to insure that the second cyclopropane would be insulated from electron transfer chemistry or that the a-pivaloyloxymethyl group would survive intact. Once the experiment

10

was carried out, it was made clear that the stereoelectronic factors operative in 35 were adequate to limit reduction to the dihydro level. Of the two products fomied, 37 was produced to a somewhat greater extent than 38 in ratios varying from 2:1 to 1:1 (Scheme V). hi this setting, it was opportune to acetylate the mixture and to effect P-elimination within esterified 37 to give 39. Careful saponification of this intermediate produced 38 in a high state of purity. If 38 was left too long under these alkaline conditions or a stronger base was employed, intramolecular Michael addition to the exomethylene ketone occurred prematurely. In order to craft the lactone ring, 38 was oxidized to 40 under Swem conditions in a prelude to intramolecular 1,4-addition of the hemiacetal anion [20] formed via nucleophihc attack by methoxide ion at the aldehyde site. With the availability of acetal 41, it became necessary to consider carefully whether to elaborate the epoxy lactone segment in advance of, or subsequent to, introduction of the a,p-unsaturated ester subunit. Since the latter option was considered more workable, 41 was transformed into the enol triflate and subjected to palladium(n) catalyzed methoxycarbonylation [21]. This methodology allowed for proper homologation of 42 to 43, and subsequent conversion to 44, in totally regiocontrolled fashion. The sector where theremainingcarbon atom needed to be introduced in 44 proved to be so sterically crowded that a number of standard methods for achieving lactone a-methylenation fared very poorly or, more often, worked not at all. Following these probe experiments, we found it possible to engage the neopentyl carbon in the capture of monomeric formaldehyde [22] as electrophile. The 10:1 mixture of epimeric hydroxymethyl products was directly dehydrated via the mesylates to deliver 46. The final oxidation could be effected either directly with m-chloroperbenzoic acid or by way of a three-step sequence involving DibalH, /-BuOOH with V0(acac)2, and TPAP with NMO [23]. Thus, 32 steps were required to reach pentalenolactone P methyl ester. The relative stereochemical relationship of its cyclopropane and lactone rings was immediately secured by Diels-Alder cycloaddition and maintained during the photoisomerization and reductive cleavage steps that followed. C. (+).IKARUGAMYCIN, AN UNUSUAL MACROCYCLIC TETRAMIC ACID ANTIBIOTIC As early as 1972, the culture broths of Streptomyces phaeochromogenes wererecognizedto be capable of producing a powerful and specific antiprotozoal and antiamoebic agent [24]. This dextrorotatory substance was determined to be the architecturally uncommon macrocyclic compound 47 and called ikarugamycin. The incorporation within 47 of a trans,anti.cis-AtcdiiyAxO'aS' indacene subunit, a largeringlactam, and an enoyltetramic acid prompted us [25] and others [26-28] to undertake its constmction in the laboratory.

-^sb^, ^

^^^ 37

35

ACgO, EtgN DMAP, CH2CI2

MeOH, H2O

Swern

OAc

40 NaOMe. MeOH

1. Pd(0Ac)2, PhgP, EtgN

LDA;

COatm MeOH, OMF 2. CH2N2

COOMe

OTf

PhNTfg THF

43

41 1. 10%HCI, THF 2. (n-Pr)4NRu04. NMO. 4A MS CH2CI2 HO

O 1. CH3SO2CI, EtgN, CHgCb

LDA. THF; COOMe

CH2O

^.

COOMe

2. DBU.CeHe

COOMe 46

MCPBA CH2CI2.A

Scheme V We saw in 47 an opportunity to deploy a triply convergent and enantioselective strategy. The challenge of obtaining the western half of the molecule, which was addressed first, was met with a concise route to racemic tricyclic hydroxy ketone 56 in six short steps [29] from readily available 48 [30]

12

(Scheme VI). The desirability of producing a major segment of the target molecule in racemic condition may appear illogical and is therefore deserving of comment. In brief, we were highly attracted to the possible deployment of a subsequent kinetic resolution of 56 by suitable application of Koga's chiral a,punsaturated aldimine methodology [31]. The superb success realized in the course of this adaptation is presented subsequently. CHgOvÔCHa

CH3OV.0CH3

48

49 KH. THF 25 "C; H2O (immed. / work-up)/^

H Q C H

H H

CHgC^^GCHg

CeCl2

3 OCH3 ^

*^*^« ''"^^' H2O (30 mInV

HQCH

KgCCb

I

• CH3OH

" OCH3

H H

1. Diba)-H. CH2CI2 ^--=-* 2. 3N HOI, ether

HO

^

53

52

& H H 54

HO

HO. KgCOs CH3OH

y^\^

NH3

A V ^ 56

55

Scheme VI An appreciation of the ability of 48 to attain appreciable levels of double diastereoselection when reacted with chiral (racemic) vinyl organocerium reagents had earlier been gained in this laboratory [32]. Consequently, it occasioned no surprise to observe that 49 [33] adds to this bicyclic ketone with customary endo stereoselectivity to deliver 50 and 51 in a relative ratio of 92:8. The major product, easily purified by chromatographic means, was smoothly isomerized to 52 under anionic conditions at room temperature. For structural reasons, this sigmatropic change is required to proceed via a boat-like transition state. The all-

13

cis tricyclic isomer must therefore be formed. However, if the quenched reaction mixture is left at 20 °C for 30 min, the basic environment promotes wholesale epimerization to 53. Consequently, only two steps need to be expended for stereocontroUed elaboration of the targeted framework having four stereogenic centers properly set in trans A/B-locked fashion. In order to invert the stereochemistry of those two carbon atoms that unite rings B and C, the ketone carbonyl was reduced and deketalization effected to give 54. Double bond migration to the intracyclic site in enone 55 and dissolving metal reduction completed the conversion to 56. The strongly acidic character of tetramic acids and their usual low solubility prompted us to delay the assembly of this heterocychc unit until very late in the synthesis. Accordingly, the appropriate ornithine segment was constructed next (Scheme VII). The known amino acid 57 [26b] was transformed via the fully protected derivative 58 to 59 by chemoselective unmasking of the y-amino group with formic acid. The remaining two substituents on the a-amino group are to be removed at different times, with the allyl carbamate destined to precede the 2,4dimethoxybenzyl functionality. 1. ArCHO. NaBHaCN. MeOH

2- 0 1 - ^ 0 ' 57 1. HCOOH. 10«C,3h

3. CH2N2

^3^" ^ ^ f

2. HOAc

Ar

- — ^ / ^ OMe OMe

59

Scheme VII The time had now arrived to append properly to 56 those sidearm substituents needed for elaboration of the macrocyclic ring. Rapid advance was realized when the silyl protected derivative 60 was formylated and O-aUcylated in situ to produce 61. Hydride reduction and acidic hydrolysis of this intermediate made available the a,p-unsaturated aldehyde 62 needed for evaluation of the potential usefulness of Koga's chemistry (Scheme Vni). Condensation of 62 with enantiopure L-rerr-leucine rerr-butyl ester led to an inseparable 1:1 mixture of the diastereomeric aldimines 63 and 64. Our expectations regarding the subsequent addition of vinylmagnesium bromide to this mixture were based on the recognized bidendate chelating ability of divalent magnesium to fix the nitrogen and oxygen atoms in a manner which significantly enhances conformational rigidity

14 OTBDMS %, P

OTBDMS

OTBDMS

^ ^^,,^,,, , 1 . KN(SIMe3)2. THF; HCOgEt

(/>Bu)2AIH.

2. (CH3)2CHI, 62

61

60 C02^Bu

b

CHO

HP*

HMPA

H

OTBDMS

OTBDMS

/ N

^

(HOAc). MgS04

^N>^C02^Bu r^f-Bu H H

Nv^COg^Bu H

H

^f-Bu 64

63 1.CH2=CHMgBr, THF

1. CH2=CHMgBr, THF

2. H3O*

2. H P *

OTBDMS

OTBDMS

H.>-^y=.

Br

Mg: 1

^Bu

.3^

T f-Bu

66

65

OTBDMS

OTBDMS

H. V - ^ H

0 67

(87:13)

68

Scheme Vin [31,34]. The relevant complexes are depicted as 65 and 66. The 1,4-addition in 65 is consequently relegated to the less sterically congested jc-face and should occur without complication. In contrast, 66 is the "mismatched" diastereomer lacking the ability to deliver the vinyl nucleophile well from the much more crowded concave direction. This competing process is kinetically disadvantaged to an extent such that the ratio of 67 to 68 obtained after citric acid quench is 87:13. When proper allowance is made for the quantity of unreacted 62 recovered, the efficiency of the vinylation was determined to be 48%. The enantiomeric purity of 67 was defined by chemical conversion to 69 and Mosher ester analysis to be 91% ee (Scheme IX). Three recrystaUizations of 67 provided enantiopure material. All eight of the stereogenic centers present in the westem sector of ikarugamycin had now been set in their proper absolute configuration.

15 OTBDMS

1. HC(OMe)3, (TsOH)

.

OTBDMS

1. PCC. NaOAc, CHgClg

^»

2. Disiamyiborane, THF; HgOij. NaOH, HjO

2. CBr4. PPhg,' CHÎg.py

OMe

OTBDMS I

J^COg. MeOH. HgO;

-^r-COOMe

70

OMe 71 OTBDMS I

-^=-CONH

COOMe

r

2,4,6-(CH3)3PhS02CI, THF:DMAP.59

OMe

OMe 1. (TsOH). acetone 2. KN{SiMe3)2,

r

OMe

o „V„ 72

{EtO)2h 2P

\A^p 73

OTBDMS PdCPPhg)^.

-CONH

"""i^^-ys^^

THF

0.^0 74 OTBDMS -CONH-

r\^-~.^y^ °x.°

OMe

75

OMe

Scheme IX With a bountiful supply of 69 at our disposal, the synthesis was continued by PCC oxidation to the aldehyde level and application of the Corey-Fuchs procedure [35] for chain homologation via dibromo olefin 70 to the acetylenic ester 71. Since amide bond construction next had to be implemented, this ester was saponified under mild conditions and the resulting carboxylic acid was activated by formation of a mixed anhydride with mesitylenesulfonyl chloride in advance of in situ condensation with 59. In order to preclude hydrolysis of the silyl ether functionality in 72, deacetalization had to be performed under anhydrous conditions in dry acetone containing a catalytic quantity of p-toluene-

16

sulfonic acid. This maneuver enabled condensation of the aldehyde so formed with phosphonate 73 [36] without encountering any detectable epimerization. The functional group array in 74 lent itself quite satisfactorily to chemoselective cleavage of the allyl carbamate residue by means of (tetrakistriphenylphosphine)palladium(O) [37], provided that acetic acid was present to inactivate the nucleophilic character of the liberated amine. The time had now arrived to effect the crucial macrocyclization. From the background experience gained by others [38], it was anticipated that the ketene 76 liberated by heating 75 in toluene for 4 h would be appropriately electrophihc. In addition, the extensive representation of diagonal and trigonal centers in 76 was expected to facihtate the desired intramolecular trapping. Indeed, the ring closure OTBDMS toluene 110*0 4h

OTBDMS

OTBDMS

CONH-

O

O

77 O^^NH1. Hg. 5% Pd-BaS04. quinoline

MeOGNSOgNEta.

2. 48% HF, CHgN O

O

COOMe

78 KOf-Bu (1 equjv)

TiuOH*^ O

O

COOMe

79

80

Q^NH CFgCOOH

/"~\

65 °G, lOmIn

-V^^ q J ^

^.

s^^

/ NH /

1

/ 1 1

H

Ar

.—^ÔMe OMe

47

1

Scheme X

COOMe

17

proceeded smoothly to deliver 77 with 94% efficiency (Scheme X). Successive semisaturation of the acetylenic bond by means of the Lindlar method, desilylation to liberate to hydroxyl group, and dehydration of alcohol 78 with the Burgess reagent [39] led most satisfactorily to introduction of the B ring double bond. Arrival at ikarugamycin from this vantage point was predicated upon the successful Dieckmann cyclization of 79. As a result of our awareness of the disastrous potential for base-promoted racemization of the proximal stereogenic center, 79 was treated with only one equivalent of potassium tert-huioxide and reaction was allowed to proceed for only 10 min at room temperature. These conditions provided enantiomerically homogeneous 80 in 66% yield. The major complication of the entire synthesis materiaUzed during subsequent removal of the 2,4-dimethoxybenzyl protecting group. After an exhaustive experimental search for proper conditions, it was recognized that heating 80 in trifluroacetic acid at exactly 62 °C and for precisely 10 min was uniquely effective in delivering 47. This successful enantioselective route to ikarugamycin demonstrates the latent capability of the anionic oxy-Cope rearrangement for highly dependable chirality transfer [40] and the potential for absolute stereochemical control offered by Koga's 1,4-asymmetric conjugate addition process. D. A REPRESENTATIVE FURANOSESQUITERPENE: (+)-PALLESCENSINA Nature has found it possible to assemble a wide range of furanosesqui- and diterpenes. Although it is quite clear that these substances are not biosynthesized via any sigmatropic scheme, the atom economy of such isomerization reactions appeared to us to warrant appHcation to this field. A thrust in this direction would require, however, that a furan ring be willing to utilize its n electrons in a manner suitable to rebonding. Precedent for an adaptation of this type was scarce [41]. Nonetheless, we have succeeded in developing a relatively concise enantioselective synthesis of natural (+)-pallescensin A (81), a marine metabolite first isolated in 1975 [42] and prepared earlier on several occasions [43-48].

81

Retrosynthetic considerations suggested that the obvious inducement for us was the opportunity to transform the known optically pure ketone 82 [49] into 83 in advance of an anionic Cope rearrangement (Scheme XI) [50]. Although 1,2addition of the cerate prepared from 3-furyllithium proceeded with appropriately high facial selectivity, subsequent isomerization of the potassium salt of 83

18

'\y

^ c „ .

Ho.î>

CeCb.THF -78 "C -* 0 X

KH 18-cr-6, diglyme, 100 "C

^

84

83

82

CH(OMe)2

CHO

Vô

MeOH, A

-O

85

LDA,

CH(OMe)2

NaHCO^ MeOH, H2O

/ K ^ O 87

86 O

CH(OMe)2

f-BuOOH

BF3'OEt2

LiAJHd.

NaOH. MeOH, A

CH2Cl2,25°C

AICI3, Et,0

••(p:.

89

88

90

H2. Pd-C. EtOAc. EtOH. Et2NH 91 \

H2. Pd-C,

X

EtOAc, EtOH, EtgNH

Scheme XI required elevated temperatures (100 °C) even when 18-crown-6 was present. Under these circumstances, the generation of enolate anion 84 was met with ensuing p-elimination of the alkoxide ion to give 85. This retro-Michael reaction is obviously facilitated by the resonance stabilization available to the leaving group. This development set the stage for chemoselective acetalization by heating 85 with ammonium chloride in methanol. Once 86 had been produced, it was possible to introduce further unsaturation as in 87, whose lone stereogenic center was to be the linchpin for establishing the proper absolute configuration of pallescensin A. In fact, the angular methyl group served as a stereocontrol element particularly well suited to introduction of the necessary trans ring fusion. Prior to that, the furan ring was concisely reconstructed by regioselective epoxidation of 87 to give 88 followed by exposure of this oxygenated intermediate to boron trifluoride etherate at room temperature [51]. In the presence of alkaline

19

tert'hutyl hydroperoxide, 87 experiences remarkably face-selective nucleophilic attack from the a direction at the more highly substituted enone double bond. These very accommodating steps were followed by reductive removal of the carbonyl group in 89 with alane [52]. NMR studies on 90 involving the use of Eu(dcm)3 as chiral shift reagent showed this advanced intermediate to be of 100% enantiomeric purity. Catalytic reduction of 90, necessarily performed in the presence of diethylamine to guard against the destructive effect of acid buildup, led via 91 to the targeted furanosesquiterpenoid. E. (-)-VULGAROLIDE, A HIGHLY OXYGENATED POLYCYCLIC METHYLENE LACTONE Our interest in developing the anionic oxy-Cope rearrangement into a powerful tool for the elaboration of structurally intricate natural products in a stereocontroUed manner has recently been successfully applied to the total synthesis of natural (-)-vulgarolide (92) [53]. Li addition to its highly rearranged isoprenoid framework, 92 features a central cyclooctanone ring to which tetrahydrofuran and y-lactone subunits are serially fused in trans-anti-trans fashion across the C-3 to C-6 positions [54]. The key elements of the stratagem designed to realize such twofold distal annulation involved initial addition of vinylmagnesium bromide to (+)-93 (100% ee), charge-accelerated [3,3] sigmatropy of the potassium salt of 94, and direct treatment of the enolate anion

produced regiospecifically witti ethyl iodoacetate (Scheme XII). Of particular relevance at this point was the fact that the new C-C bond in 95 had been installed from the p-surface. The next objective was to form lactone 96. Steric approach control operates during hydride reduction with the result that the configuration of the two cyclooctyl C-0 bonds are exactly opposite to those defined in the target. This feature was purposefully designed into the synthesis in anticipation that vulgarolide would be reached more concisely by double inversion. In fact, once the exo-methylene group had been introduced and the ahydroxyl substituent unmasked as in 97, the generation of a leaving group at the latter site and lactone hydrolysis was met by formation of oxirane 98. The bridgehead double bond was now selectively ozonolyzed. Spontaneous

20

cyclization occurred in a spectacularly facile manner to deliver vulgarolide (92) and its anomer 99 in a 1:1 ratio. Both hemiacetals converged to Omethylvulgarolide (100) during methylation, a maneuver that facilitated purification of the highly insoluble 92. Hydrolysis of 100 in turn produced predominantly vulgarolide.

^f-k-Ô

CHg-CHMgBr,

4Y

THF OSEM

.78«C-4ft

^f-^^^ OH OSEM OSEI

• 4Sô

1. KN(SiMe3)3. THF, A 2. ICHjCOOEt,

94

93

1. UAIH4

KX

2. TRAP, CHsCfe

j ^

\EUO SEMO

O—^ 96

98

SEMO

O

COgEt

95

1. LDA, CH2O THF,-78 to-25 X

1. MsCI, EtaN, DMAP

2. MsCI. EtgN, DMAP 3. DBU.CeHe 4. 5%HF, CH3CN

2. LiOH, CH3OH

a

97 A&O, GH3I CH2CI2

O3. CH2CI2;

HO

HMPA -78 "C -^ ft

10%HCI, THF HO

MeO 92,p-OH 99.a-OH

100

Scheme XII F. (+)-CEROPLASTOL I, A DICYCLOPENTA[a,d]CYCLOOCTANE SESTERTERPENE The Claisen rearrangement, a heteroatomic variant of the Cope process, holds equal appeal as a scaffolding element that is totally atom-efficient. We have addressed and defined those stereocontrol elements associated with a two-carbon intercalation tactic [55] in several contexts as, for example, in the preparation of (+)-ceroplastolI(101)[56].

21

To this end, it was opportune in light of background information to prepare 105 by sequential epoxidation of 102 [57], heating of the epoxy ketone with sodium methoxide in methanol containing a small amount of water [58], and Shapiro degradation of the ketone [59] in advance of acidic hydrolysis (Scheme Xm). Subsequently, 105 was oxidized with MPCBA to the ring-expanded epoxy lactone, heating of which at 175-180 °C in benzene solution (sealed tube) dehvered 106. FoUowing Wittig olefmation to give 107, a second carbon atom OCH3 1. NaBH^ NaOCH3. 2. MCPBA 3. PDC

H g C o ^

HgCO^

102

103

1. TSNHNH2. CH3OH, rt

l

\

H,CO \ H3U O ^ 107

104

1. MCPBA. NaHCO^CHzClg.A

2. CHaLi.THF, EtgO; NH4CI, H2O 3. HOAc, H2O.A

Y

CH3OH, (HgO)

r...o

HaCO^ 105

1. Cp2TI(CI)(CH2)AI(CH3)2 THF.(py) 2. 200 oQ

Ph3P=CH2

2. 175-180 "C. GeHg. sealed tube 106

K2CO3.

^- \ _ / f Y ' 0

CH^H, A

H3C

O ^

109

was introduced by means of the Tebbe reagent [60]. Heating this product at 200 °C in sealed, KOH-coated glass tubes resulted in conversion to the cyclooctenone 108, which was easily epimerized to the thermodynamically more favorable trans fused isomer 109. With this bicyclic intermediate available in sizeable amounts, ready advance to 111 could be conveniently accomplished prior to annulation of the second fivemembered ring (Scheme XIV). 1,3-Carbonyl transposition was realized by complete eradication of the original carbonyl by Ireland's method [60] followed by ally lie oxidation. Application of the Piers cyclopentannulation protocol [61] to 111 made 113 conveniently available. Introduction of a methyl group into ring B was brought about by treatment of the kinetically derived enol triflate [62] with lithium dimethylcuprate [63]. Hydrolysis of 114 gave the dienone, which was directly transformed into 115 by oxidation of its silyl enol ether with palladium acetate in acetonitrile [64].

22

Completion of the synthesis involved some adaptation of Boeckman's original route to 101 [65]. Introduction of the sidechain was accomphshed by copper-catalyzed conjugate addition of Grignard reagent 116 to 115. Nucleophilic attack occurred exclusively from the p-face with formation of a 3:2 mixture of 117 and its diastereomer. Once chromatographic separation had been accomplished, the carbonyl group in 117 was reduced by the action of sodium borohydride and zinc chloride on the tosylhydrazone [66]. Desilylation occurred during this step to deliver ceroplastol I (101) in a global overall yield of 0.13%. 1. LiAIH^ 2. n-BuU, HaCO^

CIP(0)(NMe2)2 3. Li, EtNHg, f-BuOH, ether

V-/f/c HaC

1. Se02,KH2P04, toluene, A 2. PDC

O ^

*"

V^T'O 111

110

109

1. KN(SiMe3)2, PhN(Tf)2 THF. -78 "C

C ' ~ \ /

HgC O ^

THF

HÔ

A ^^

113

112

1. (TsOH), acetone, H2O Jif

7'"0 HgC O 1 ^^^ ^^^

MgBr

CH3 116

^ 2. UN(SiMe3)2, MegSiCI 3. Pd(0Ac)2, CH3CN

^"^^^

2. (CH3)2CuLi. THF. -20 ''C

115

1. TsNHNH, (COOH)2, Eton

^.

CuBr*Me2S HMPA, THF, M©3SiCI, -78 °C

H j c X ^ '^'^''YÔHI CH3 1 »3C^y—\J

^

2. NaBHsCN, ZnClg, CHgOH. 90 °C 117

UjyHaC 101

Scheme XIV G. THE MARINE TOXIN (+)-ACETOXYCRENULIDE The crenulatan diterpenes, now believed to arise in the marine environment by solar-induced photoisomerization of dictyolactones [66], have been isolated from small brown seaweeds and the sea hares that feed on them. In light of the high survival rate of these species, the crenulatans were investigated and found to function as defensive agents [67,68]. The most bountiful of these toxins appears to be acetoxycrenulide (118), which features a central eight-membered ring and

23

fused cyclopropane and butenolide (or equivalent) subunits characteristic of this class [69-71]. H A H CeHil^^ OAc 118

In our initial studies aimed at the realization o^ an enantioselective, stereocontrolled synthesis of 118, the end game was to attach the methylheptenyl sidechain to position 3 quite late in the reaction sequence [72]. We saw in this undertaking an opportunity to again utilize the Claisen rearrangement as the

LDA

Vj^^-^

x'^VCHO 120

119

(1:1)

121

N-PSP. (TsOH), CH2CI2

H

1. Nal04. NaHCOg. MeOH, H2O

SePh H (1:1)

\

2. EtgNH. mesitylene, A

SePh

^-tS^ H

HC(0Me)3, (TsOH)^

2. {0^^)^r^, \

CH^2.CeHe

125

H

'^M H " O

123

124

.OH, A V ^ ^ - ^

O.

O ^ ' " ^ ] ' ^

W V - - A ^

^

1. KN(SiMe3)2. PhSeCI r^"^-'^-

2. Nal04. NaHCC^. MeOH. HÔ

126

127 H A H

5% HOI

128

Scheme XV

^ ^

24

means for elaborating the cyclooctane core. To this end, the well known lactone 119 [73-75] was engaged in aldol condensation with crotonaldehyde and intramolecular selenonium ion-promoted cyclization with participation by the neighboring hydroxyl group (Scheme XV). As illustrated, the aldol process is fully trans-selective, and provides an easily separable 1:1 mixture of 120 and 121. By making recourse to N-(phenylseleno)phthalimide [76], we were successful in transforming 120 into 122 and 123, both of which underwent elimination via the derived selenoxides to introduce an enol ether double bond exocyclic to the pyran ring. This transient species entered into the Claisen rearrangement exclusively via its chair conformation 124, whose adoption guaranteed not only the location of the intracyclic double bond but, most importantly, the absolute configuration of the carbon atom carrying the methyl substituent. The key intermediate 125 so produced was then ketalized and subjected to Simmons-Smith cyclopropanation [77]. The three-membered ring is introduced smoothly from the sterically less congested n surface to deliver 126. With the northern sector now completely constructed, the butenolide double bond was introduced by organoselenium technology and ketal hydrolysis implemented. Under these conditions, 128 was formed without any evidence of double bond migration. The rigid conformation adopted by 128, corroborated by X-ray crystallography, proved inimical to fmitful enolization at the "doubly activated" site in order to incorporate the Cg sidechain, thereby requiring that this unit be present from the outset. This realization led us to probe the consequences of incorporating this rather bulky substituent at an early stage instead [78]. Several issues were considered to hold relevance: (a) would the Claisen rearrangement such as that defined by 124 continue to prove serviceable in providing the means for delivering stereodefined 4-cyclooctenones? (b) would the ponderal effect of the Cg sidechain impact negatively on utilization of the chair-like arrangement in light of the fact that the configuration at C-3 would require this moiety to be axially disposed? (c) would the need to fix C-3 stereochemistry first be a deterrent to proper installation of the remaining stereogenic centers? To answer some of these questions, (5)-citroneUic acid (129), whose methyl-substituted carbon is enantiomeric to that in 118, was transformed via oxazolidinone 130 to the hydroxy acetate 132 (Scheme XVI) using quite standard reactions. When access was subsequently gained to lactone 134, it is made clear that the chiral auxiliary was deployed to set the absolute configuration of the stereogenic ring carbon properly [79]. At this stage, introduction of a hydroxymethyl substituent was undertaken. The conversion to 136, mediated by 135 and involving 1,4-addition of iPrOMe2SiCH2MgCl in the presence of copper(I) iodide and chlorotrimethylsilane [80], proved quite superior to the altematives which were examined. Following oxidative desilylation, it proved an easy matter to convert 137 into the thermody-

25 /-O OH O

1.(CCX)I)2 2. n-BuLi,

LDA. ak.

HNÔ.THF 11 O

129

^

1. LIAIH4

OAc

m

Br,

'^^

THF 130

131 OH 1. CrOg. H2SO4, acetone

9-BBN.THF; NaB03-4IÔ.

2. AcCI, py. CH2CI2

2. KOH.THF 3. TsOH.CeHe

HP

132

u

133

V I Ô-SiCHgMgCI,

1. KN(SiMe3)2; N-PSP, THF 2. H2O2. py, CH2CI2

Cul, MeaSiCI. THF

134

136 0/a= 85:15)

135 TsOH,

KHF2^ H2O2, DMF 137

138

Scheme XVI namically favored lactone 138. In order to safeguard the structural integrity of this pivotal intermediate, its oxidation with PDC to aldehyde 139 was effected as soon as possible (Scheme XVII). Although a variety of attempts to achieve chemoselective 1,2-addition to the aldehyde carbonyl in 139 proved troublesome because of steric shielding, these difficulties could be circumvented by the introduction of (phenylseleno)methyllithium under high dilution conditions [81]. As revealed by the product structure 140, the generation of an alkoxide ion in this manner was followed by intramolecular attack at the lactonic center. Fortunately, reconversion to the ylactone could once again be easily realized by acid-catalyzed isomerization. Protection of the hydroxyl group made it possible to effect aldol condensation with crotonaldehyde and subsequent ring closure to afford the bicyclic selenolactones 142 and 143. The selenoxide derived from 143 underwent both 1,2-elimination and Claisen rearrangement when heated in mesitylene containing

26 PDC. _4A_MS^ CHgCl/

PhSeCHjjU,

HO,^>v.Js^SePh

THF 140

1. TsOH, CeHe 2. = < ° ^ " 3 CH3. (POCI3)

A?"-'

1. LDA,

^^^^^

2. (TsOH). CeHg.A

142

141

SePh

1. Nal04, NaHCO»

°tCLsePh

9.

H

H2O. MeOH 2. EtgNH. mesityiene, A

143

145

Scheme XVH diethylamine to give 145 in 76% overall yield. Consequently, the associated [3,3] sigmatropic change unquestionably proceeds via the chair arrangement shown in 144. Furthermore, no adventitious epimerization operates and the p,'y-unsaturated double bond does not migrate into conjugation. The findings detailed in Scheme XVII provided important guidance and insight into the requirements necessary to the actual adaptation of this pathway for the production of 118. It is obvious that (/?)-citronellol must serve as the building block for the sidechain. Beyond that, however, it is not just a matter of producing the RJi'isomcT of 132, for this course of action will ultimately provide only the enantiomer of the target. This is because the remaining stereocenters are introduced under the fuU control of those present in this acetate. As a consequence, this intrinsic bias must be overridden by involving a chiral auxihary that is capable of properly establishing stereogenic centers in an absolute sense totally independent of those preexisting in the substrate. The successful realization of these objectives is outlined in Scheme XVIH. Since probe experiments disclosed convincingly to us that the double bond in the intact sidechain of acetoxycrenulide is more reactive to Simmons-Smith cyclopropanation than cyclooctenyl double bonds, the decision was made to introduce the isopropenyl group in the final stages of the synthesis. Accordingly, ester 146, which is easily produced from (/?)-citronellol [81], was transformed into

27 Br

1. LDA, RO'^'^""-^'"V"^COOMe

KN(SiMe3)2.

2. 0 » CH2CI2; Pfyp 3. NaBH^.MeOH

146

I o RO

b

k^-N' V ^

149

HzC^. py

147

V

kJ-..^'-'

PhSeCI;

RO^^^^-^^'VlW : H o

1. Og.CHgClg.

a

MeOH; MegS

rhBuU; THF -78 °C

2. HC{0Me)3, (TsOH),MeOH

148

150

(MaO)2CH

9^^)^^ H

L.1^^

O

1. 5 % H C I , OH

2. Pr^P-CHg

151

THF

2. PDC. 4AMS. CH2CI2

fÔ-

152

153

1. O ^ C H ^ I g ; PhaP SePh

2. n-BuLi, CH2(SePh)^

^SePh

THF, -78 '^C 3. CH3C(OMe)-CH2,

155

POCI3

1. Nal04. N a H C O j

O

H

^ /A

MeOH. H p 2. EtgN. CHg-CHOEt. MegNCOCHa, 220 °C

H

CH2I2,

7t'

(C2H5)Zn. CgHg

O

156

157 1. AC2O, DMAP

1. (^Bu)2AIH. CH2CI2. -78 ''C

2. P y H F , CH3CN. H2O

2. A g ^ O ^ - C e l l t e , CgHÂ

3. PDC. 4A MS

B

3. KN{SiMe^2. PhSeCI; NalO^. NaHC03

4. Ph3P«C{CH3)2. THF, -78 "C ^ rt

^^

158

159

0 0

XHI J

OAc

118

Scheme XVIII

28

147 by sequential C-allylation, ozonolysis, and borohydride reduction. Once butenolide 148 had been accessed, 1,4-addition of enantiopure allylphosphonamide 149 [82] was carried out. As hoped for, 150 was formed uniquely. The configurations at sites a and b arise therefore from the chirality inherent in 149 and not elsewhere. Immediate removal of the chiral auxiliary followed by functional group manipulation gave rise to 153, which was transformed by means of chemistry developed above into 155. Once this advanced intermediate had been oxidized to the selenoxide, more elevated temperatures than usual were found necessary to bring about the conversion to 157. More heating is presumably required because the sidechain in 156 must be projected axially if a chair-like sigmatropic transition state is to be utilized. Despite this steric inhibition, 157 was isolated in 55% yield. Particularly satisfying was the ease with which 157 was homologated to 158 (92%). Reduction of the ketone carbonyl in a chemoselective manner was not possible because of the steric protection it benefits from. This potential complication was skirted when it was found that the hydroxy lactol produced by diisobutylaluminum hydridereductionrespondedto the Fetizon reagent only at the five-membered site to deliver 159. With the stereochemistry of 159 securely established by NOE analysis, no obstacles were encountered during acetylation and the subsequent completion of sidechain construction. Thus, although this enantioselective approach to (+)-118 is linear, the five stereocenters that adorn the eight-membered ring are conveniently set in a fashion which could well prove useful in a wide range of synthetic settings. H. (+).CLEMEOLIDE, THE STRUCTURALLY UNIQUE DITERPENE LACTONE CONSTITUENT OF CLEOME VISCOSA The herb Cleome viscosa (syn. cleome icosandra), which is widely distributed in India, has long been recognized by the native population to serve as a rubefaciant, vesicant, and anthelmintic agent. As a consequence of these reputed properties, three research groups undertook almost simultaneously in the late 1970's to determine the principal active constituent of this sticky, odoriferous plant [82,83]. On the basis of the NMR, X-ray, and CD data, the substance was determined to be the macrocyclic diterpene lactone 160 and named cleomeolide. The structural features of this macrolide are unusual in several respects: (a) the double bond positioned a,p to the lactone carbonyl resides at a bridgehead site, a

=

'OH

160

uH3C„

CH3H \ H

^^^^^'

29

property shared in common with taxol and other select natural products [84]; (b) the nine-carbon chain cis-fused to the methylenecyclohexane subunit is projected diaxially from the six-membered ring such that three of the four groups pendant on the cyclohexane are oriented in this fashion; and (c) this three-dimensional arrangement provides extensive steric screening to the exocyclic methylene carbon, such that its introduction at a modestly advanced stage of the total synthesis should be viewed as problematical. The correctness of this assumption has been assessed experimentally [85]. Our enantioselective approach to cleomeolide began by controlled dithioketalization [86] of optically pure Wieland-Miescher ketone [87] in order to distinguish between the two carbonyl groups. The best means uncovered for the homologation of 161 to the cis-dimethyl ketone 163 involved 162 as an intermediate (Scheme XIX). The action of (methoxymethylene)triphenylphosphorane on 161 afforded a 7:1 cis/trans mixture of isomers, which were easily separated after sodium borohydride reduction to the primary carbinols. The major component underwent reductive conversion to 163 very smoothly. O

(

'. /—V S ^ "^

)

1. NaBH4.MeOH 2. CHaSDÎ,

"•• PhaPCHgOCI-^ Cl KN(SiMe3)2, THF

EtgN. CHjClg 3. LiBHEtg.THF 4. TI(N03)3«3H20, MeOH, THF

2. 10% HOI

^ 161

162

HC(OCH:^3, (TsOH),

OCH3

(

MeOH, DMF

OCH3

MCPBA

^

163

NaBH4

GeHe, hexanes, silica gel

MeOH

164

OCH3

1. TBSCI, imid, DMF 2. LiAIH4. THF

1. CH3SO2CI. EtgN. CH2CI2

OH

//

OTBS

OH 168

167

166

ON

2. KCN. 18-cr-6, DMF

H

ON

CI^-CHOCgHs.

xylene 200 °C

Hg(OCOCF3)2. EtgN 169

H3C

i^---^Sf? H

170

Scheme XIX

O

171

30

Since bond disconnection within 163 had to be implemented a to the carbonyl so as to maintain attachment of the methylene carbon, the dienol ether 164 was generated and treated with two equivalents of m-chloroperbenzoic acid. As a consequence of preferred electrophilic attack at the enol ether double bond and the high latent reactivity of the resulting oxygenated epoxide, ring cleavage occurred to deliver the aldehydo ester 165 [88]. Following the reduction of 165 to 167 via 166, the requisite additional carbon atom was introduced by cyanide ion displacement at a primary mesylate center. This transformation and the subsequent removal of the rerr-butyldimethylsilyl protecting group proceeded efficiently to provide 168. Transetherification of 168 with ethyl vinyl ether under strictly defined conditions [89] led to 169 and set the stage for the projected Claisen rearrangement. A reasonable rate of thermal isomerization was achieved at 200 °C in xylene under sealed tube conditions. The substitution plan in 169 led us to anticipate that transition state 172 would be kinetically favored ahead of 173 because of the development of destabiUzing 1,3-diaxial interaction in the latter. If 172 were indeed to be utilized, die new C-C bond would be appropriately installed in the desired cis manner as in 171. In actuality, the 170:171 ratio was found to be 2.7:1 and this product distribution was comparably found in analogues of 169 [90,91]. These findings suggest that a chair-like arrangement may not be strictly adopted during installation of the axial bond present in 170. H

v----Jr-CH3

172

Notwithstanding the adverse product distribution, contined use was made of this isomerization process because chromatographic separation of the epimers proved to be facile, and exceptional convergency could be subsequently reahzed during extension of the "lower chain". Thus, the two allylic alcohols 174 and 175 formed upon treatment of 171 with 2-propenylmagensium bromide were individually transformed into the identical aldehyde 176 following the implementation of a second Claisen rearrangement step (Scheme XX). Once this important finding was made clear, the three-step process could be streamlined by omitting all chromatography. Once acetal 177 had been produced [92], the cyano-substituted carbon was activated by conversion to 178 [93]. Cychzation of the unmasked aldehyde was efficiently realized with potassium carbonate and 18-crown-6 in toluene at room temperature [94], the ring closure occurring while both functionalized sidechains

31

HaC H,C.

o

H

THF, -78 "C

•

H3C

171

1. CH2.CHOC2H5. EtgN, Hg(OCOCF3)2 2. CBHe.170«'C

V >^V^^,„^^

•^JL MeO.

COsEt

Conditions: (a) i. BzCl, pyridine; ii. NaN02/AcOH; (b) H2, Pd/C; (c) FeCls; (d) i. SO2; ii. (MeO)2S02, K2CO3; (e) i. KOH; ii. NaN02/AcOH; iii. H2, Pd/C; iv. FeCls; (f) i. Et02CCH2CN, NEt3; ii. K3Fe(CN)6; (g) KOH; (h) butadiene, A. Scheme 2

65 mediated oxidation to an orthoquinone was followed by reduction and methylation to give dimethoxynaphthalene derivative 41. Saponification and repetition of this series of functionalizations afforded orthoquinone 42.

A Michael-type addition of ethyl

cyanoacetate, followed by reoxidation, hydrolysis, and decarboyxlation gave the necessary dienophilic component 43 for a simple intermolecular Diels-Alder reaction with butadiene to complete construction of ring C. An unusual catalytic hydrogenation, developed somewhat unexpectedly during model studies,^^ bestowed morphine's azacarbocyclic skeleton, although stereochemically incorrect at C14. The following mechanism to explain this fortuitous cychzation has been advanced. Figure 9M MeO.

MeO^

MeO. MeO

NH

Figure 9. Reductive Construction of Ring D

A modified Huang-Minion procedure, followed by N-methylation and amide reduction, yielded (±)-[3-A6-dihydrodesoxycodeine methyl ether 52, which after resolution proved identical to material derived from natural sources^^^'^ via manipulation of P-thebainone, and consequently served as a synthetic relay. (It is within an obscure portion of this series of rigorous structural studies and proofs that Evans later intercepted Gates' route and thus formaUzed his synthesis).

66 Having established the identity of synthetic material, Gates turned to the introduction of a Ce oxygen functionality. A series of unsuccessful studies culminated in a low yielding acid-catalyzed hydroxylation. Selective demethylation and oxidation gave P-dihydrothebainone 54, Scheme 3, and most other subsequent formal syntheses intercept MeO.

y^ ii ^^%^ II

^

45

MeO'

1 1\

^^

» i

^

J 'H

k>^ MeO^

C

52

f^\llII

V

HO"^

1 'H 0*^

54

Conditions: (a) H2, copper chromite; (b) i. KOH, N2H4; ii .NaH, Mel; iii .LiAlH4; (c) i. dibenzoyl tartrate resolution; ii. H2SO4/H2O; (d) i. KOH, ethylene glycol, 225 °C; ii. KO^Bu, Ph2CO. Scheme 3

at this point. Now, in Gates' own words, "... inversion at C14 to produce the cis fusion of rings II [B] and III [C] loomed large." Bromination followed by conversion to the 2,4-DNP hydrazone resulted in a cascade of events which eventually led to the correct Ci4 stereochemistry. However, it is in the proof of this epimerization, by thorough and extensive comparison with several other opiate alkaloids, that the truly complex character of this task comes to the fore. Having arrived at the final carbocyclic core of the molecule, an efficient closure of the C4-C5 ether bridge was required, but this proved impossible with several advanced systems, presumably because of unfavorable

67 stereochemical and conformational factors. In the end, it was necessary "...with some reluctance..." to reduce the enone moiety 57 to its corresponding ketone, Scheme 4.

MeO,

MeO.

C

NMe

'H

NMe

HO' NMe

NMe

56 Br

NMe

NMe

NMe

NMe

HO* Conditions: (a) Br2; (b) 2,4-DNPH, A; (c) HCl, aq. acetone; (d) i. H2, Pt02; ii. Br2 (2.0 eq.); iii. 2,4-DNPH; (e) HCl, aq. acetone; (f) LiAlH4; (g) pyHCl, 220 °C. Scheme 4

68 Treatment of the ketone with two equivalents of bromine, followed by hydrazone formation, finally effected closure of the ether bridge with concomitant elimination to hydrazone 58.^^^ Hydrolysis to the enone, followed by reduction, removed both the aryl bromide and introduced the alcohol stereoselectively to give natural codeine 2,^^'^*^ which was demethylated under established conditions'^ to yield morphine 1.^2,45 ^ summary of these studies can be divided into synthesis of 50,"*^ Cu-chromite reductive cyclization,"^^^ isomerization of C\4,^^^ reduction to codeine,"*^^ and a full paper sumary."^^^ 3b.

Ginsburg, 1954^7 A close second to Gates' momentous acheivement, Ginsburg set a lasting trend by

intercepting dihydrothebainone 75, thus rendering his synthesis formal. Despite an elegant stepwise construction of the ABC ring system, the closure of the C15-C16-N bridge proceeded with some element of luck. Initial condensation of cyclohexanone with ortho-hihiaitd veratrole, followed by elimination and an ingenious ally lie oxidation yielded enone 64, Scheme 5. Michael

X)

MeO. MeO'

60 62

61 MeO.

MeO.

MeO'

MeO'

0=N

HO-N

CI 63a

63b (continued)

69 (continued) MeO,

MeO^

70 (continued)

MeO,

^QQ

NMe

NH

74

75

Conditions: (a) i. "BuLi; ii. oxalic acid, toluene, A, Dean-Stark; (b) NOCl (from namylnitrite and HCl); (c) i. py, A; ii. aq. H2SO4, A; (d) dibenzyl malonate, KO^Bu; (e) i. H2, Pd; ii. A; (f) HF; (g) ethylene glycol, pTsOH, A, Dean-Stark; (h) i. AmONO, NaOEt; ii. H2, Pd, HCl; (i) acetoxyacetyl chloride, 2 eq. py , CHCI3; (j) ethylene glycol, pTsOH, A, Dean-Stark; (k) i. AmONO, NaOEt; ii. aq. acid; (1) hydrazine, ethylene glycol, A; (m) i. aq. acid, A; ii. LiAlH4; (n) i. CH2O, HCO2H, A; ii. benzophenone, KO^Bu; iii. resolution (d-tartaric acid). Scheme 5 addition, decarboxylation and Friedel-Crafts annulation led to the ABC tricycle 67. A selective protection-deprotection sequence allowed introduction of the amino group exclusively at C9, followed by a smooth acid chloride condensation which afforded acetoxyamide 70. Prior to reduction of these functionalities, an attempt to block C5 as its ketal gave unexpected results. Selective cleavage of the C4 methyl ether, not uncommon in higher morphine analogues, was accompanied by a spontaneous formation of the ethylamine bridge with the correct C13 quaternary center, accompanied by an undesired Cio ketahzation resulted in 71, containing the complete morphine carbon connectivity. Introduction of the required C6 ketone while removing those at C5 and Cio, reduction of the lactam followed by a necessary N-methylation and oxidation gave (±)dihydrothebainone 75. This was resolved to the 1-form with d-tartaric acid, and thus the synthesis was rendered formal."^^

71 3c.

Barton, 196349 Barton's approach was truly biomimetic, as it pursued an initially described

conversion of labelled reticuline 76 to radioactive morphine 77, thus supporting the proposed biosynthetic route, Figure 10. MeO^

papaver N^^Me

somniferum

N^*Me

MeO'

Figure 10. Transformation of Labelled Reticuline to Morphine An attempt to emulate this pathway in vitro followed, unfortunately hampered by a dismal Mn02 promoted oxidative coupling. However, a radioisotope dilution study did indeed suggest a 0.012% conversion to radioactive salutaridine 79, Scheme 6.

NMe

NMe MeO'

MeO'

NMe

NMe

Conditions: (a) Mn02, 0.012%; (b) NaBH4; (c) < pH 4, rt. Scheme 6

72

Reduction to a mixture of the epimeric diols, followed by acid catalyzed allylic displacement by the phenolic group, yielded radioactive thebaine 81 and thus formahzed "...a third and long sought synthesis of the morphine alkaloids...".^^ 3d.

Morrison, Wake, Shavel,

196739and Grewe, 196738

Pubhshing almost simultaneously with Grewe, this group adopted a bioanalogous synthesis starting from a substituted benzyltetrahydroisoquinoline 84, which underwent selective Birch reduction and monodeprotection to 85, Scheme 7. Treatment with acid

NMe

88 para coupled Conditions: (a) i. amidation; ii. Bischler-Napieralski; iii. reduction; methylation; (b) Na/^BuOH/NHs; (c) 10% aq. HCl, A, ^-3% andp-37%. Scheme 7

73

(refluxing 10% hydrochloric acid as reported by Morrison, Waite, and Shavel; phosphoric acid in the work of Grewe) causes enol ether hydrolysis and effects cychzation to both the para-coupltd flavinantine derivative 88 (37%) and dihydrothebainone 75 (3%), thus intercepting Gates' synthesis. The low regioselectivity of this process directly contrasts the effectiveness of the enzyme mediated biosynthesis.

3e.

Kametani, 196951 In this approach, a Pschorr-type cyclization was adopted in order to maximize

ortho-para

selectivty.

Diazotization of 2-aminobenzyltetrahydroisoquinoline 89

followed by thermal decomposition yielded racemic salutaridine 25 in a meager yield of 1.1%, Scheme 8. Other products isolated from the cychzation included benzaldehyde and

MeO.

MeO.

BnO NMe

NMe MeO'

MeO'

O

OMe MeO,

MeO,

NMe MeO'

25

NMe MeO

Conditions: (a) i. NaN02, H2SO4/ACOH; ii. 70 °C, (b) NaBH4, MeOH; (c) IM HCl. Scheme 8 laudanine 11 (3.5%), but no products of ortho-ortho couphng, which have been observed when zinc powder was employed.^^ Reduction followed by acid catalyzed ring closure

74

in the method of Barton^^^ gave racemic thebaine 8, and thus constituted a formal total synthesis of morphine. Resolution of 89 with di-p-toluoyl tartaric acid afforded entry to both enantiomeric series, allowing comparison of ORD/CD data.^^'^^

3f.

Schwartz, 197554 Schwartz successfully emulated the in vivo para-ortho coupling of N-

acylnorreticuline derivatives by the use of thallium tristrifluoroacetate (TTFA), Scheme 9. Treatment of ethoxycarbonyl derivative 91 with one equivalent of the salt afforded a

MeO^

MeO'

Conditions: (a) 1 eq. T1(TFA)3, CH2CI2, -78 to -20 "C; (b) LiAlH4; (c) IM HCl. Scheme 9

23% yield of the corresponding salutaridine 92. Once again, reduction followed by acid treatment yielded racemic thebaine and thus formalized syntheses of codeinone,^^ codeine,45b and morphine.46

75 This approach was later extended to an enantioselective synthesis,^"^^ Scheme 10. Unfortunately, some racemization occurred during preparation of substrate 94, but not during cyclization to 9 5 , this time mediated by iodosobenzene diacetate. Reduction/closure, followed by Barton radical decarboxylation yielded the known thebaine analogue 97.^6

MeO. .COoMe

MeO'

'^l"'''^

94

OH MeO,

MeO,

NCOjMe MeO

MeO

Conditions: (a) i. 3-benzyloxy-4-methoxyphenyl acetic acid, 1,1-carbonyldiimidazole, THF; ii. H2, Pd/C, EtOAc; iii. Me02CCl, CH2CI2, EtsN; iv. POCI3, MeCN; v. NaBH4, MeOH; vi. Me02CCl, Na2C03, MeOH; vii. Na2C03,H20; (b) PhI(OAc)2, TFA; (c) i. NaBfLt; ii. N,N-DMF dineopentenyl acetal; (d) Barton decarboxylation.^^^ Scheme 10 3g.

Beyerman, 1979^7 Beyerman has adopted a classic Grewe type bioanalogous approach, neatly

avoiding the problems of ortho-para temporary element of symmetry.

regioselectivity by introduction of a subtle

76 In the culmination^'^ of a series of related studies,^^ by now standard procedures yielded the chiral mono-Birch reduced benzyltetrahydroisoquinoline 100, Scheme 11.

OBn

OBn

MeO,

MeO.

BnO'

BnO'

NMe MeO'

MeO

r-"^P^NMe

NMe

M e O - ^ ' ^ ^ -'^

NMe NMe 75 Conditions: (a) CH2O, H2, Pt/C; (b) U/NH3, ^BuOH; (c) HCl, Et20,; (d) 5-chloro-lphenyltetrazole, K2CO3, DMF, 70 °C; (e) H2, Pd/C, 50-55 °C. Scheme 11

This substance underwent a smooth cycloalkylation, in aknost quantitative yield, to give morphinan 101. All that remained was the selective removal of the unwanted C2

77

hydroxyl. Presumed steric shielding indeed allowed a selective etherification to yield tetrazole derivative 102; a simple catalytic hydrogenolysis^^ afforded (-)dihydrothebainone 75, and thus a formal synthesis.^^^ The use of elevated temperature (55 °C) during fmal hydrogenolysis of the tetrazole seems to be critical since de Graw, arriving at the N-carbomethoxy derivative of 102 virtually via the same synthetic sequence, was not able to remove the C2 oxygen when performing the cleavage at room temperature.^^

3h.

Rice, 19802b,35 Rice employed bromo derivative 108, Scheme 12, to avoid undesired para

Ti

a

MeO.

HO'

iT'

NH,

MeO'

104

MeO.

MeO.

NCHO

MeO-^^^^jjjg

VO

êOv^^^^jv^Br

êO\^;^v^Br

^ i—\ ^^^^^>f^^NCHO \,0

106

•^

^•^

j ~ \

['''^'''î^^

107

108 (continued)

78 (continued) MeO,

MeO. g

h

HO

HO' NMe

NCHO

MeO,

NMe

111

Conditions: (a) i. 200 T ; ii. MeCN, POCI3, reflux; iii. NaCNBHa, 65%-MeOH, pH 4-5, A; (b) Li/NHs, ^BuOH; (c) i. 1.5 eq. PhOCHO, A; ii. cat. CH3SO3H, ethylene glycol, THF, quant; (d) N-bromoacetamide, 0 T ; (e) HCO2H/H2O; (f) 14% NH4F-HF, CF3SO3H, 0 "C; (g) 10:1 MeOH, aq. HCl, A; (h) H2, Pd/C, 2N AcOH, HCHO, NaOAc, quant; (i) i. Br2, AcOH; ii. CHCI3, IN NaOH; iii. as (h). Scheme 12 coupling. The idea of blocking C2 was poineered by Beyerman^^^ (C2-Me), who proposed^^g blocking with C2-halogen. Readily available amine 104 was subjected to Birch reduction and protection to give 106, prior to treatment with N-bromoacetamide at 0 °C. Deprotection afforded ketone 108 which underwent cyclization smoothly to give 109 in 60% yield. Hydrolysis and a one-pot hydrogenation to effect both reductive Nmethylation (using formaldehyde) and debromination led to dihydrothebainone 75, intercepting Gates' route. Alternatively, bromination, ether ring closure, and the same

79

hydrogenation protocol yielded dihydrocodeinone 111. This whole synthesis required isolation of only six intermediates, obtained sufficiently pure for immediate further use, and proceeded in 29% overall yield. It remains as the most practical preparation of morphine to date.^^'^^'^^ 3i.

Evans, 1982^2 Evans' initial steps are reminiscent of Ginsburg's, as orr/io-lithiated veratrole 115

was coupled with piperidone 116 to give after dehydration alkene 118, Scheme 13. COjH

^ :

CO2H

112 MeO^

M e O ^ ' ^ f ^ 115

^-^

Me 116

N^Me CIO4-

122 (continued)

CHO

80 (continued) MeO.

MeO.

MeO'

MeO' NMe

NMe

MeO. MeO' NMe

NMe

Conditions: (a) i. B2H6»THF, 25 "C; ii. PBrs, HBr (48%), CH2CI2; (b) ZnBri, C6H6, A; (c) Et20, 0 °C; (d)/?TSOH, toluene, A,; (e) i. "BuLi, THF, -10 T ; 114; ii. Nal, K2CO3, MeCN, A, (f) HCIO4, Et20, MeOH, ii, MeOH, 50 "C; (g) CH2N2, CH2CI2; (h) DMSO; (i) BF3»Et20, toluene, -10° C, (j) i. MsCl, NEts; ii. LiEtsBH; iii. OSO4, NaI04, THF, aq. AcOH. Scheme 13

Deprotonation to the enamine anion, selective coupling with the allylic terminus of dibromide 114, followed by an intramolecular enamine alkylation, afforded reduced isoquinoHne 119. A rather elegant conversion to aminoaldehyde 122 ensued. Immonium ion formation in 119 via protonation with perchloric acid at first yielded the kinetic trans isomer, which underwent equilibration upon reflux in methanol to give the corresponding crystalhne cis product 120. Diazomethane treatment led to aziridinium salt 121, which upon exposure to DMSO, ring opened with concomitant oxidation in a Komblum fashion to the aldehyde 122.^^ Treatment with Lewis acid effected B-ring closure, thus

81 completing the carbon framework. Reduction of the benzylic hydroxyl and LemieuxJohnson cleavage yielded Gates' ketone 124, thus formalizing the synthesis. A C14 epimerization procedure allowed verification by comparison with authentic epimeric 126, although conditions for the actual transformation of 124 to 125 and 126 are not given. 3j. Rapoport, 19836^ For a number of years prior to Evans' revelations, Rapoport had been involved in the development of a general methodology for the synthesis of several morphine structural analogues. These included both cis and trans 4a-aryldecahydroisoquinolines 127,^^ octahydro-lH-benzofuro-[3,2-e]-isoquinolines 128,^^ and novel octahydro-lH[l]-benzopyrano-[4,3,2-e,f|-isoquinoUnes 129:^'^ MeO^

MeO'

NMe

NMe

NMe

127

128

129

Although armed with a wealth of experience in the field,^^ several stereochemical problems proved unavoidable, and Rapoport finally resorted to interception of Evans' route, thus doubly formalizing his synthesis. However, the construction of the key intermediate, via an effective a-methylene lactam rearrangement, is markedly different. Starting from 2-hydroxy-3-methoxybenzaldehyde 130, Scheme 14, standard MeO.

MeO. MeO

EtOsC

(continued)

82 (continued)

"'°V^ J U -1^ MeO' Y

V "Tf^

r T 133

e MeO'

MeC^

A^COÊt

EtOjC

MeO,

MeO^

.C02Et

COjEl

0^"^ N ^ H 134

CN

MeO,

MeO.

MeO.

MeO'

MeO

MeO

COjEt

S

MeO.

MeO. j, k

MeO'

MeO'

MeO. MeO'

i/-C02H

HOjC

O2CH

NMe

O

Me

140

MeO,

:MeO » O,

'BUO2C

m NMe

141 (continued)

NMe

'BuOgC

0

142

83 (continued)

MeO Y ^ 0-^C^

TBDMSO^

MeO.

Br

166 MeO.

HO*'

MeO.

MeO,

HO*'

MeO

MeO.

MeO,

MeO*

(continued)

90 (continued)

MeO,

MeO,

NMe

MeO.

NMe

NMe

HO*^

Conditions: (a) BU3P, DEAD, THF; (b) i. (48%) HF, MeCN; ii. CrOs, H2SO4, aq. acetone, 0 T ; iii. DffiAL, THF, -78 to 25 °C; (c) "BuLi, THF, -78 °C; (d) OSO4, NMO, aq. acetone; ii. Pb(0Ac)4, CHCI3; (e) i. MeNHi-HCl, MeOH, NaCNBHs; ii. Me3SiCH2CH20COCl, aq. NaHCOs, (f) i. DMSO, TFAA, CH2CI2 then EtsN, -78 to 20 T ; u. (MeO)3CH, MeOH, pTsOH, 65 "C iii. TEOCCl, aq. NaHCOa; (g) KO^Bu, THF, (h) DDQ, pTsOH, CHCI3, H2O; (i) i. TFA; ii. CHCI3, aq. NaHC03; (j) HCl, Et20, CH2CI2 then 0.2 N HaOH, CHCI3; (k) NaBH4, MeOH; (1) BBr3, CHCI3. Scheme 19

at C5 and alcohol C6 cis., yielding the cycHzation precursor 166 (Attempted cyclization of the trans alcohol derived from 165 resulted in an "inseparable mixture"). Selective metal-halogen exchange at the aromatic ring induced an intramolecular conjugate

91 addition forming the C12-C13 bond, followed by alkylative closure at C14 to complete the A,B and C ring system, 167. Subsequent manipulation of the allyl moiety via oxidative cleavage, reductive amination and protection yielded the trimethylsilyethoxycarbonyl, ester, 169. Swem oxidation was followed by methyl enolether formation, 170, and base elimination of the sulfonyl moiety afforded the diene, 171. Subsequent DDQ oxidation yielded dienone 172 which upon TEOC-deprotection gave, via 1,6-addition, a mixture of racemic codeinone 27 and neopinone 28. Isomerization of the double bond as described by Rapoport and Barber,'^^ followed by reduction afforded (±)-codeine 2. Finally, this material was 0-demethylated following the conditions of Rice^^ to afford racemic morphine 1.

3n.

Tius, 19927 7 In this original and imaginitive approach, a rapid assembly of the phenanthrene

core of morphine, containing a novel non-aromatic A ring, was achieved in an intermolecular Diels-Alder reaction between quinone 173 (prepared from 3-methoxy-2hydroxy benzaldehyde in 7 steps and an overall yield of 35%) and diene 174 (from 1,4cyclohexanedione monoethylene ketal in 2 steps with an overall yield of 30%), Scheme 20. In one of several unsuccesful attempts to aromatize ring A, an unexpected tandem

O

92 (continued)

MeOoCN

MeOoCN

183

182

(continued)

93 (continued)

Conditions: (a) toluene, 100 °C; (b) i. PhSeCl, MeOH; ii. H2O2, THF; (c) aq. HCl, THF; (d) KN(TMS)2, THF, -78 °C, then 2-(p-toluenesulfonyl)-3-phenyloxaziridine, -78 °C; (e) H2, Pd, THF; (f) TFAA, DMSO, CH2CI2, -78 °C; (g) i. BF3»OEt2, -30 °C; ii. Mel, K2CO3, acetone; (h) i. PhSeCl, EtOAc; ii. H2O2, THF; (i) NaBH4, MeOH; (j) i- MeLi, THF, 0 °C; ii. aq. H2CO, NaCNBHs, MeCN; (k) Dess-Martin, CH2CI2; (1) Zn, NH4CI, aq. EtOH; (m) i. DIBAL, THF; ii. aq. HCl; iii. glac. AcOH, 100 °C. Scheme 20

selenocyclization and subsequent oxidative elimination gave urethane-aminal 176. Deprotection and kinetic enolization of the resultant ketone, followed by oxidation with Davis reagent, introduced the C4 oxygen and provided 178. Hydrogenation of the double

94 bond, followed by Swem oxidation of the C4 hydroxyl yielded acyloin 180. In a "...fortunate turn of events...", boron trifluoride-mediated rearrangement induced aromatization with simultaneous closure of the C4-C5 ether bridge. Methylation of the phenol to 181 followed by selenoxide elimination protocol produced enone 182, which was reduced to 183. This allowed cleavage of the carbamate with methyl lithium and reductive amination of the secondary amine afforded 184. The Cg hydroxyl was reoxidized under Dess-Martin conditions to give enone 185 which upon exposure to zinc dust and ammonium chloride underwent reductive cleavage of the aminal with concomitant closure at Cg to yield morphinan 186. Reduction of the Cg carbonyl, followed by acid catalyzed hydrolysis produced p-thebainone which was isomerized at Ci4 under acidic conditions yielding thebainone, 187, thus intercepting Gates' synthesis.

3o. Parker, 1993^8 80 The elegant formal total synthesis of morphine, accomplished by Parker, shows some similarities to that of Fuchs through analogous disconnections. In both syntheses, the core of the molecule was formed as a result of a tandem process; in this case as a result of a radical cascade.^^'80 The inmiediate cyclization precursor 191 was prepared via a Mitsunobu reaction between monoprotected cis-dio\ 189 (prepared in 8 steps from 2-((3-methoxyphenyl)ethylamine) in 47% overall yield) and phenol 188, followed by cleavage of the silyl ether. Scheme 21. The key step, homolytic cleavage of the Ci2-Br

« Me

MeO^ ^.^^

a 189

(continued)

95 (continued)

MeO,

MeO, SPh Br

Q NTs Me

TBDMSO^

H O ^ ^ ^

191

190 MeO,

MeO

SPh

Br

NTs Me MeO,

NMe

NMe

111

Conditions: (a) BU3P, DEAD, THF, 0 T ; (b) 10% HF, MeCN; (c) BusSnH, AIBN, benzene, 130 °C, sealed tube; (d) Li/NHs^BuOH, THF, -78 "C; (e) (COCl)2, DMSO, CH2CI2, -78 "C to 0 °C. Scheme 21

bond in 191, was mediated by BusSnH, and AIBN under sealed tube conditions. The aryl radical closed at C13 to form the dihydrofuran ring, yielding a new radical at C14 which in turn was trapped by the p-carbon of the styrene to give a resonance-stabilized intermediate. EUmination of the phenylthiol group afforded advanced intermediate 192, containing the tetracycHc carbon skeleton, with correct stereochemistry at C5, C13, and C14. Finally, a nitrogen radical anion, generated during cleavage of the tosyl group, was trapped by the C9-C10 double bond "....in an unprecedented closure....", completing ring D and setting the C9 absolute stereochemistry correctly, as in 193. Swem oxidation of

96 alcohol at C^ yielded racemic dihydrocodeinone 111, establishing the formal total synthesis of codeine'^^*^ and morphine 7^

3p.

Overman, 199381 The crucial step of Overman's approach is essentially a Grewe-type disconnection

but involves an intramolecular Heck reaction to complete ring B. An enantioselective reduction of 2-allyl-cyclohexeneone 195 introduced a chiral element. Condensation of the resultant S-alcohol, (196) with phenyhsocyanate, oxidation of the sidechain olefm with osmium tetroxide and acetonide protection afforded 198, Scheme 22. A copper^ «

Ph

N^g/ 194

^ ^

OH

195

196 OCONHPh

OCONHPh

197 SiMejPh

6^

MeO.

NHDBS

199

OBn OMe

200

(continued)

BnO

97 (continued)

MeO.

MeO^

BnO' NDBS

NMe

Conditions: ; (a) 194, catecholborane; (b) PhNCO; (c) i. Os04, R3NO; ii. acetone, acid catalysis; (d) i. THF, -30 "C; ii. "BuLi, Cu(Ph3P)2, 0 °C; iii. PiiMe2SiLi, 0 "C; iv. /7TsOH, MeOH; v. NaI04; vi. DBS-NH2, NaBHsCN; (e) Znl2, EtOH, 60 °C; (f) 10% Pd(OCOCF3)2(PPh3)2, 1,2,2,6,6,-pentamethylpiperidine, toluene, A; (g) i. BF3.0Et2, EtSH; ii. (as camphorsulfonate), 3,5-dinitrophenylperbenzoic acid, CH2CI2, 0 °C; (h) NMO, TPAP; (i) H2, Pd(0H)2, HCHO. Scheme 22

catalyzed suprafacial S N 2 ' displacement of the ally lie carbamate with lithium dimethylphenyl silane, deprotection and diol cleavage furnished the intermediate aldehyde, whose reductive amination with dibenzosuberyl amine

afforded 199.

Condensation of 200 (prepared in 7 steps from isovanillin in an overall yield of 62%) with allylsilane 199 at 60 °C in the presence of Znl2 was followed by iminium ion allylsilane cyclization to yield the advanced isoquinoline intermediate 201.^^ Palladium-

98 mediated coupling connected C12-C13 and afforded morphinan 202 with the correct stereochemistry at C9, C13, and C14. In the final steps, the phenolic oxygen was liberated, the double bond at C6-C7 was epoxidized on the P face, and intramolecular attack of the phenolic hydroxyl completed the dihydrofuran ring.^^ Oxidation, followed by reductive DBS cleavage in the presence of formaldehyde yielded (-)-dihydrocodeinone 111, established the latest reported formal total synthesis of (-)-morphine.

Using (S)-

oxazaborolidine catalyst for reduction of 195 establishes the formal total synthesis of (+)morphine.

4. SvnoDs of Apwoaches to the Ring Svstems of

Authormate

BIZ

Kcv Step

Starting Material

M rphine Final Product

MeO

d Angelo,

1990 Ref 84

M

lhcn 1 4 2 0 __c

65%

OMe

0-OBn

0-OBn

>95% e.e. from chiral Robinson

8 steps

Boger,

1982 Ref 85

Broka, 1988 Ref 86

0

Ho%

13 steps

&d 0

NCOfls

Ms

Ciganek, 1981 Ref 87,88

NMe

~

HO

Ciganek, 1981 Ref 87,88

0 ~

Ciufolini, 1993 Ref 89

T"ao4 steps

OAc

OAc

Hudlicky, I992 Ref 90

4-

to1ucne,

4 steps

*,'

THSO'"' THSO"" homochiral from 9 3 9 0 biooxidation

Jones, 1985

Ref 91

2

Kametani, 1986

OMe

OMe

Ref 92 NC

M;igtiiis,

I00 I

Ref 93 79%

Mamuno.

I993 Ref 94

OMe

I

McMuny.

I984 Ref 95.96

b

iPr0

8z I

W% c.c. from chiral formmidinc

I I

Monkovi ' c , 1973

Rcf 98.99

Monkovi-c, 1975

Ref 1 0 0

8. 0

I

Me0

Monkovi'c, 1978 Ref 101

Me0

4045' C __t

MeO

Noyori, 1987 Ref 102

__c

Me0

MeoB... '.;

"H NMs

Parsons,

1984 Ref 103

Schultz, 1976 Ref 104, 105

@NMe '"OH

PC5

Me0

OH

I Schultz,

1985 ief 106, 107

Shenvi, 1984

Ref 108

Stella, 1977

Ref 109, 110

Stella, 1983

Ref 1 1 1

($

yields connilio;. unknown

,..'

,o-NMeCl

NMe

X = CI. O h . 011

X

= CI. OAc. 011

Bz-y& 3. loo' C. 30 min

Tius,

19 steps

1986 Ref 112

e

Weller, 6 steps

1983 Ref 113, 114

0

e N' I

Me. I

Dalton I Costanzo 1988 Ref 1.8 (Appendix)

Hudlicky 1992 Ref 115

HO

13 stcps

Several steps

C02E1

I



I

107 5.

Conclusions This review attempted to collate all of the design elements inherent in the various

existing approaches to morphine. The authors hope that this information presented together in one document will make it easier for potential future contributors to this field to review the hterature and augment the existing approaches with their own. After nearly forty years of serious effort, it is evident that the field of morphine synthesis is still wide open. The pioneering synthesis of Gates and the most efficient one by Rice are accompanied by other ingenious approaches. What remains before the organic chemical community is the design and implementation of a truly practical approach.

6.

Acknowledgments The authors are grateful to Mallinckrodt Speciahty Chemicals for support of the

research work regarding their own approaches to morphine. We thank Kenner Rice (NIH) for reading the manuscript and for providing information connected to the use of morphine and derivatives and Professor David R. Dalton of Temple University for sharing with us a copy of a recent dissertation. Scott Richardson of Mallinckrodt Specialty Chemicals is acknowledged for sharing with us a review of morphine synthesis.

108 7.

1.

References

For a discussion of chemical dissonance and consonance see H.N.C. Wong, M.-Y. Hon, C.-W. Tse, Y.-C. Yip, J. Tanko, and T. HudUcky, Chem. Rev., 89, (1989), 165-198, and reference 19 therein, pertaining to an unpublished review on this topic by D.A. Evans.

2.

a. P.T. White and S. Raymer "The Poppy," National Geographic, Feb. 1985, vol. 167, pp. 142-188. b. See also K.C. Rice in The Chemistry and Biology of Isoquinoline Alkaloids, Phillipson et al, ed.; Springer-Verlag: Berlin, 1985, pp 191-203.

3.

C.E. Terry, and M.Pellens, The Opium Problem, Ch. 2, Bur. Soc. Hyg., New York, 1928.

4.

U.S. Drug Enforcement Agency, "Controlled Substance Aggregate Production Quota History," Federal Register, July 7, 1994.

5.

National Narcotics Intelligence Consumers Committee Report 1993, "The Supplies of Illicit Drugs to the United States," DEA-94066, August 1994.

6. 7.

O. Hesse, Ann., 220, (1883), 203. a. F. Santavy, Alkaloids, New York, 17, (1979), 385; b. For annual updates on chemistry of morphine, see Specialist Periodical Reports, Chem. Soc. London, "The Alkaloids", vol 1-13, (1970-1983).

8.

J.F. Deronse, Ann. Chim., 45, (1803), 257.

9.

M.A. Seguin, Ann. Chim., 92, (1806), 225.

10.

F.W.A. Sertiimer, Trommsdorjfs J. Pharm., 14, (1806), 47.

11.

J.M. GuUand and R. Robinson, Proc. Mem. Manchester Lit. Phil. Soc, 69 (1925), 79.

12.

M. Gates and G. Tschudi, J. Am. Chem. Soc, 74, (1952), 1109.

13.

D. Ginsburg, The Opium Alkaloids, Interscience Publishers: New York, 1962.

14.

A. Laurent, Ann. Chim. Phys., 19, (1847), 359.

15.

C.R.A. Wright, J. Chem. Soc, (1874), 1031.

16.

E. Grimaux, C. R. Acad. ScL, 93, (1881), 591.

17.

For an excellent overview of Pschorr phenanthrene synthesis see D.F. DeTar, Organic Reactions, 9, (1957), 409.

18.

M. Mackay and D.C. Hodgkin, J. Chem. Soc, (1955), 3261.

19.

G. Kartha, F.R. Ahmed, and W.H. Barnes, Acta Cryst., 15, (1962), 326.

109 20.

a. A.R. Battersby, R.C.F. Jones, and R. Kazlauskas, Tetrahedron Lett., 22+23, (1975), 1873, and references therein, b. M.L. Wilson and C.J. Coscia, J. Am. Chem. Soc.,97, (1975), 431.

21.

a. A.R. Battersby, R. Binks, R.J. Francis, D.J. McCaldin, and H. Rauz, / . Chem. Soc, (1964), 3600. b. D.H.R. Barton, G.W. Kirby, W. Steglich, G.M. Thomas, A.R. Battersby, T.A. Dobson, and H. Ramuz, J. Chem Soc, (1965), 2423, and references therein.

22.

M. Rueffer and M.H. Zenk, Z Naturforsch, 42c (1987), 319.

23.

a. T. Frenzel and M.H. Zenk, Phytochem., 29, (1990), 3505. b. R. Stadler and

24.

R.B. Herbert, The Biosynthesis of Secondary Metabolites, Chapman and Hall:

M.H. Zenk, Liebigs Ann. Chem., (1990), 555. London, (1989), 144. 25.

a. A.R. Battersby, D.M. Foulkes, and R. Binks, J. Chem. Soc, (1965), 3323. b.

P.R. Borkowski, J.S. Horn, and H. Rapoport, J. Am. Chem. Soc, 100,

(1978), 276. 26

W. De-Eknamkul and M.H. Zenk, Phytochem., 31, (1992), 813.

27.

D.H.R. Barton, and T. Cohen, Festschrift A., StoU; Birkhauser Verlag: Basel, (1957), 117.

28.

a. R. Gerardy and M.H. Zenk, Phytochem., 32, (1993), 79. b. For a related study, see: R. Stadler and M.H. Zenk, J. Biol. Chem., 268, (1993), 823.

29.

a. T. Amann and M.H. Zenk, Tetrahedron Lett., 32, (1991), 3675. b. M.H. Zenk, R. Gerardy, and R. Stadler, J. Chem. Soc. Chem. Comm., (1989), 1725.

30.

a. A.R. Battersby, Proc. Chem. Soc, (1963), 189. b. G. Stork, The Alkaloids, Vol 6; R. Manske and H. Holmes EDs.; Academic Press, Inc.: New York, (1960), 219.

31.

R. Gerardy and M.H. Zenk, Phytochem., 34, (1993), 125.

32.

H. Lotter, J. Gollwitzer, and M.H. Zenk, Tetrahedron Lett., 33, (1992), 2443.

33.

J. Gollwitzer, R. Lenz, N. Hampp, and M.H. Zenk, Tetrahedron Lett., 34, (1993), 5703.

34.

G.W. Kirby, S.R. Massey, and P. Steinreich, /. Chem. Soc, Perkin Trans. 1., (1972), 1642.

35.

K.C. Rice, J. Org. Chem., 45, (1980), 3135.

36.

R. Grewe and A. Mondon, Chem. Ber., 81, (1948), 279.

37.

For a review see: D.C. Palmer and M.J. Strauss, Chem. Rev., 11, (1977), 1.

[10 38.

a.

R. Grewe and W. Friedrichsen, Chem. Ber., 100, (1967), 1550. b. R.

Grewe, H. Fischer, and W. Friedrichsen, Chem. Ber., 100, (1967), I . e . R. Grewe and H. Fischer, Chem. Ber., 96, (1963), 1520. 39.

G.C. Morrison, R.O. Waite, and J. Shavel, Jr., Tetrahedron Lett., (1967), 4055.

40.

I. Fleming, Selected Organic Syntheses, Wiley Interscience: London, (1973).

41.

M. Gates, J. Am. Chem. Soc, 72, (1950), 228.

42.

R.B. Woodward, F.E. Bedar, H. Bickel, A.J. Frey, and R.W. Kierstead, Tetrahedron, 2, il95S), 1.

43.

M. Gates, R.B. Woodward, W.F. Newhall, and, R. Kunzli, J. Am. Chem. Soc, 72, (1950), 1141.

44.

G.M. Badger, J.W. Cook, and G.M.S. Donald, J. Chem. Soc, (1951), 1392.

45.

For a more complete account of these studies, see also: a. M. Gates, G. Tschudi, J. Am. Chem. Soc, 78, (1956), 1380. b. M. Gates, J. Am. Chem. Soc, 75, (1953), 4340. c. M. Gates and R. Helg, J. Am. Chem. Soc, 75, (1953), 379. d. M. Gates, and G. Tschudi, J. Am. Chem. Soc, 72, (1950), 4839. For information concerning the preparation of dihydrodesoxycodeine methyl ether as a synthetic relay, see: e. L. Small and G.L. Browning, J. Org. Chem., 3, (1939), 618. f. G. Stork, J. Am. Chem. Soc, 74, (1952), 768.

46.

H. Rapoport, C.H. Lovell, and B.M. Tolbert, J. Am. Chem. Soc, 73, (1951), 5900.

47.

D. Elad and D. Ginsburg, /. Am. Chem. Soc, 76, (1954), 312.

48.

For a fuller account of these studies, see also: a. R.O. Duthaler and D. Ginsburg, Helv. Chim. Acta, 69, (1986), 1559. b. D. Elad and D. Ginsburg, /. Chem. Soc, (1954), 3052. c. D. Elad and D. Ginsburg, J. Chem. Soc, (1953), 2664. d. D. Ginsburg and R. Pappo, J. Chem. Soc, (1953), 1524. e. D. Ginsburg and R. Pappo, /. Chem. Soc, (1951), 516. f. D. Ginsburg and R. Pappo, J. Chem. Soc, (1951), 939. g. E.D. Bergmann, R. Pappo, and D. Ginsburg, J. Chem. Soc, (1950), 1369.

49.

D.H.R. Barton, G.W. Kirby, W. Steglich, and G.M. Thomas, Proc Chem. Soc, (1963), 203.

50.

For related work, see: D.H.R. Barton, D.S. Bhakuni, R. James, and G.W. Kirby, J. Chem. Soc. (C), (1967), 128; see also reference 21b.

51.

T. Kametani, M. Ihara, K. Fukumoto, and H. Yagi, J. Chem. Soc. (C), (1969), 2030.

52.

I. Kikkawa, J. Pharm. Soc. Japan, 78, (1958), 1006.

53.

For applications of this methodology to the synthesis of numerous alkaloids, see: a. T. Kametani, A. Kozuka, and K. Fukumoto, J. Chem. Soc. fCj, (1971), 1021. b. T. Kametani, C. Seino, K. Yamaki, S. Shibuya, K. Fukumoto, K. Kigasawa, F. Satoh, M. Hiiragi, and T. Hayasaka, J. Chem. Soc. (C), (1971), 1043. c. T. Kametani, K. Shishido, E. Hayashi, C. Seino, T. Kohno, S. Shibuya, and K. Fukumoto, J. Org. Chem., 36, (1971), 1295. d. T. Kametani, T. Sugahara, H. Yagi, and K. Fukumoto, J. Chem. Soc. (C), (1969), 1063. e. T. Kametani, K. Fukumoto, A. Kozuka, H. Yagi, and M. Koizumi, J. Chem. Soc. (C), (1969), 2034. f. T. Kametani, K. Yamaki, H. Yagi, and K. Fukumoto, J. Chem. Soc. (C), (1969), 2602. g. T. Kametani, K. Fukumoto, and T. Sugahara, Tetrahedron Lett., 52, (1968), 5459. h. T. Kametani, K. Fukumoto, F. Satoh, and H. Yagi., J. Chem. Soc. Chem. Commun., (1968), 1398.

54.

a. M.A. Schwartz, and I.S. Mami, / Am. Chem. Soc, 97, (1975), 1239. b. M.A. Schwartz, and P.T.K. Pham, /. Org. Chem., 53, (1988), 2318.

55.

H. Conroy, /. Am. Chem. Soc, 77, (1955), 5960.

56.

For a number of ingenious related studies, see: a. D.G. Vanderlaan and M.A. Schwartz, J. Org. Chem., 50, (1985), 743. b. M.A Schwartz and M.F. Zoda, J. Org. Chem., 46, (1981), 4623. c. M.A. Schwartz and R.A. Wallace, Tetrahedron Lett., 35, (1979), 3257. d. D. H. R. Barton, Y. Herve, P. Potier, and J. Theirry, J. Chem. Soc Chem. Commun., (1984), 1298.

57.

T.S. Lie, L. Maat, and H.C. Beyerman, Reel. Trav. Chim. Pays-Bas, 98, (1979), 419.

112 58.

For a series of related studies, see also: a. H.C. Beyerman, J. van Berkel, T.S. Lie, L. Maat, J.C.M. Wessels, H.H. Bosman, E. Buurman, E.J.M. Bijsterveld, and HJ.M. Sinnige, Reel Trav. Chim. Pays-Bas, 97, (1978), 127. b. H.C. Beyerman, L. van Bommel, L. Maat, C. Olieman, Red Trav. Chim. Pays-Bas, 95, (1976), 312. c. H.C. Beyerman, T.S. Lie, L. Maat, H.H. Bosman, E. Buurman, E.J.M. Bijsterveld, and HJ.M. Sinnige, Reel Trav. Chim. Pays-Bas, 95, (1976), 24. d. H.C. Beyerman, E. Buurman, L. Maat, and C. Olieman, Reel Trav. Chim. Pays-Bas, 95, (1976), 184. e. C. Olieman, L. Maat, and H.C. Beyerman, Reel Trav. Chim. Pays-Bas, 95, (1976), 189. f. H.C. Beyerman, E. Buurman, T.S. Lie, and L. Maat, Reel Trav. Chim. Pays-Bas, 95, (1976), 43. g. H.C. Beyerman, E. Buurman, and L. Maat, J. Chem. Soe. Chem. Commun., (1972), 918. h. For a continuation of these studies, see also: C. Olieman, L. Maat, and H.C. Beyerman, Reel Trav. Chim.Pays-Bas., 99, (1980), 169.

59.

W.J. Musliner and J.W. Gates, Jr., /. Am. Chem. Soe., 88, (1966), 4271.

60.

J.I. DeGraw, J.C. Christensen, V.H. Brown, and M.J. Cory, J. Heteroeycl C/iem., 11, (1974), 363.

61.

K.C. Rice, and A. Brossi, J. Org. Chem., 45, (1980), 592.

62.

D.A. Evans, and C.H. Mitch, Tetrahedron Lett., 23, (1982), 285.

63.

An analogous opening with chloride followed by intramolecular Friedel-Crafts had served as a model study: D.A. Evans, C.H. Mitch, R.C. Thomas, D.M. Zimmerman, and R.L. Robey, J. Am. Chem. Soe., 102, (1980), 5955.

64.

W.H. Moos, R.D. Gless, and H. Rapoport, J. Org. Chem., 48, (1983), 227.

65.

D.D. Weller, R.D. Gless, and H. Rapoport, /. Org. Chem., 42, (1977), 1485.

66.


67.


113 68.

For some related studies see also: a. R.D. Gless, and H. Rapoport, J. Org. Chem., 44, (1979), 1324. b. D.D. Weller and H. Rapoport, /. Am. Chem. Soc, 98, (1976), 6650.

69.

J.D. White, G. Caravatti, T.B. Kline, E. Edstrom, K.C. Rice, and A. Brossi, Tetrahedron, 39, (1983), 2393.

70.

W. Ludwig, and H. J. Schafer, Angew., Chem., Int. Ed. Engl, IS, (1986), 1025.

71.

For an earlier, unsuccesful approach via anodic coupUng, see: H. Kliinenberg, C. Schaffer, and H. J. Schafer, Tetrahedron Lett., 23, (1982), 4581.

72.

J.E. Toth, and P.L. Fuchs, /. Org. Chem., 52, (1987) 473.

73.

Several model studies strengthened this approach: a. J.E. Toth, and P.L. Fuchs, /. Org. Chem., 51, (1986), 2594. b. P.R. Hamann, J.E. Toth, and P.L. Fuchs, J. Org. Chem., 49, (1984), 3865. For a full account, see: c. J.E. Toth, P.R. Hamann, and P.L. Fuchs, J. Org. Chem., 53, (1988), 4694.

74.

For an untethered variant of this strategy, see: J. Ponton, P. Helquist, P.C. Conrad, and P.L. Fuchs, J. Org. Chem., 46, (1981), 118.

75.

R.B. Barber and H. Rapoport, J. Med Chem., 19, (1976), 1175.

76.

K.C. Rice, J. Med Chem., 20, (1977), 164.

77.

M.A. Tins and M.A. Kerr, /. Am. Chem. Soc, 114, (1992), 5959.

78.

K.A. Parker and D. Fokas, J. Am. Chem. Soc, 114, (1992), 9688.

79.

Prehminary investigations of this radical approach were presented in: a. K.A. Parker, D.M. Spero, and J. Van Epp, / Org. Chem., 53, (1988), 4628. b. K.A. Parker, D.M. Spero, and in part K.C. Inman, Tetrahedron Lett., 27, (1986), 2833.

80.

A full account of this work has recently appeared: a. K.A. Parker, and D. Fokas, J. Org. Chem., 59, (1994), 3927. b. K.A. Parker and D. Fokas, J. Org. Chem., 59, (1994), 3933.

114 81.

C.Y. Hong, N. Kado, and L.E. Overman, J. Am. Chem. Soc, 115, (1993), 11028.

82.

For further details regarding this Mannich type cyclization, see: D.A. Heerding, C.Y. Hong, N. Kado, G.C. Look, and L.E. Overman, J. Org. Chem., 58, (1993), 6947.

83.

This methodology has recentiy been extended to a palladium catalyzed biscyclization in which the furan ring is formed directly: C.Y. Hong and L.E. Overman, Tetrahedron Lett., 35, (1994), 3453.

84.

H. Sdassi, G. Revial, M. Pfau, and J. d'Angelo, Tetrahedron Lett., 31, (1990), 875.

85.

D.L. Boger, M. Patel, and M.D. MulUcan, Tetrahedron Lett., 23, (1982), 4559.

86.

C.A. Broka and J.F. GerUts, J. Org. Chem., 53, (1988), 2144.

87.

E. Ciganek, J. Am. Chem. Soc, 103, (1981), 6261.

88.

E. Ciganek, U.S. Patent 4,243,668, (1981); Eur. Pat. Appl. 9780, (1980); Chem. Abstr. (1980), 93, 220720.

89.

M.A. CiufoHni, M.A. Rivera-Fortin, and N.E. Byrne, Tetrahedron Lett., 34, (1993), 3505.

90.

T. HudHcky, C.H. Boros, and E.E. Boros, Synthesis, (1992), 174.

91.

S. Handa, K. Jones, C.G. Newton, and D.J. Williams, J. Chem. Soc. Chem. Commun., (1985), 1362.

92.

T. Kametani, Y. Suzuki, and T. Honda, J. Chem. Soc. Perkin Trans. 1, (1986), 1373.

93.

P. Magnus and I. Coldham, / Am. Chem. Soc, 113, (1991), 672.

94.

Y. Genisson, C. Marazano, and B.C. Das, J. Org. Chem., 58, (1993), 2052.

95.

J.E. McMurry, V. Farina, W.J. Scott, A.H. Davidson, D.R. Summers, and A. Shenvi, J. Org. Chem., 49, (1984), 3803.

115

96.

For a successful application of this methodology to a total synthesis of 0-methylpallidinine, see: J.E. McMurry and V. Farina, Tetrahedron Lett., 24, (1983), 4653.

97.

A.I. Meyers and T.R. Bailey, J. Org. Chem., 51, (1986), 872.

98.

I. Monkovic, T.T. Conway, H. Wong, Y.G. Perron, I.J. Pachter, and B. Belleau, J. Am. Chem. Soc, 95, (1973), 7910.

99.

For the utilization of this methodology in the synthesis of several morphinan, isomorphinan, normorphinan and hasubanan derivatives of potential pharmacological interest, see also: a. I. Monkovic and H. Wong, Can. J. Chem., 54, (1976), 883. b. I. Monkovic, H. Wong, A.W. Pircio, Y.G. Perron, I.J. Pachter, and B. Belleau, Can. J. Chem., 53, (1975), 3094. c. I. Monkovic, H. Wong, B. Belleau, I. J. Pachter, and Y.G. Perron, Can. J. Chem., 53, (1975), 2515. d. T.T. Conway, T.W. Doyle, Y.G. Perron, J. Chapuis, and B. Belleau, Can. J. Chem., 53, (1975), 245. e. B. Belleau, H.Wong, I. Monkovic, and Y.G. Perron, J. Chem. Soc. Chem. Commun., (1974), 603. f. M. Saucier and I. Monkovic, Can. J. Chem., 52, (1974), 2736.

•

100.

I. Monkovic, Can. J. Chem., 53, (1975), 1189.

101.

I. Monkovic, C. Bachand, and H. Wong, J. Am. Chem. Soc, 100, (1978), 4609.

102.

M. Kitamura, Y. Hsiao, R. Noyori, and H. Takaya, Tetrahedron Lett., 28, (1987), 4829.

103.

M. Chandler and P.J. Parsons, J. Chem. Soc. Chem. Commun., (1984), 322.

104.

A.G. Schultz and R.D. Lucci, J. Chem. Soc. Chem. Commun, (1976), 925.

105.

For a full account see: A.G. Schultz, R.D. Lucci, W.Y. Fu, M.H. Berger, J. Erhardt, and W.K. Hagmann, J. Am. Chem. Soc, 100, (1978), 2150.

106.

A.G. Schultz, R.D. Lucci, J.J. Napier, H. Kinoshita, R. Ravichandran, P. Shannon, and Y.K. Yee, J. Org. Chem., 50, (1985), 217.

116 107.

For further developments see: A.G. Schultz, and PJ. Shannon, /. Org. Chem., 50, (1985), 4421.

108.

A.B. Shenvi, and E. Ciganek, J. Org. Chem., 49, (1984), 2942.

109.

L. Stella, B. Raynier, and J.-M. Surzur, Tetrahedron Lett., 31, (1977), 2721.

110.

For a full account, see: L. Stella, B. Raynier, and J.M. Surzur, Tetrahedron, 37, (1981), 2843.

111.

L. Stella, Angew. Chem. Intl. Ed. Engl., 22, (1983), 337, citing the thesis of J.L. Bourgeois, Marseille, 1980.

112.

M.A. Tius, and A. Thurkauf, Tetrahedron Lett., 11, (1986), 4541.

113.

D.D. Weller, E.P. Stirchak, and D.L. Weller, /. Org. Chem., 48, (1983), 4597.

114.

For preliminary results in this series, see: a. D.D. Weller, D.L. Weller and G.R. Luellen, J. Org. Chem., 48, (1983), 3061. b. D.D. WeUer and D.L. WeUer, Tetrahedron Lett., 23, (1982), 5239. c. D.D. Weller and G.R. Luellen, Tetrahedron Lett., 22, (1981), 4381.

115.

T. Hudlicky, G. Butora, S.P. Feamley, A.G. Gum, and M.R. Stabile, (pending release of publication from sponsors of research)

117 8. Appendix I. List of Dissertation Titles Concerning Morphine Synthesis 1.

The tandem radical cychzation synthesis of morphine alkaloids Fokas, Demosthenes (1993) 234 pp. Avail.: Univ. Microfihns Int., Order No. DA9406935 From: Diss. Abstr. Int. B 1994, 54(10), 5150

2.

Synthetic approaches to novel morphine analogs Turner, Stephen Michael (1991) 202 pp. Avail.: Univ. Microfihns Int., Order No. BRDX94736 From: Diss. Abstr. Int. B 1992, 52(10), 5283-4

3.

Part I. The total synthesis of two human urinary metaboHtes of delta-9-THC. Part n. The total synthesis of (d,l)- morphine Kerr, Michael Andre (1991) 387 pp. Avail.: Univ. Microfihns Int., Order No. DA9205864 From: Diss. Abstr. Int. B 1992, 52(9), 4733

4.

Intramolecular Diels-Alder cychzations in an approach to the morphine skeleton Wu, Chengde (1990) 254 pp. Avail: Univ. Microfihns Int., Order No. DA9128007 From: Diss. Abstr. Int. B 1991, 52(4), 2044

5.

Organic synthesis via palladium coupling reactions Pyatt, D. (1990) 167 pp. Avail.: Univ. Microfihns Int., Order No. BRD-92664 From: Diss. Abstt. Int. B 1991, 52(3), 1445

6.

A synthetic approach to morphine Ellwood, Charles Walter (1989) 141 pp. Avail.: Univ. Microfihns Int., Order No. BRDX91246 From: Diss. Abstr. Int. B 1991, 51(9), 4343

7.

New synthetic approaches towards the synthesis of morphine Spoors, Paul Grant (1989) 195 pp. Avail.: Univ. Microfihns Int., Order No. BRDX89587 From: Diss. Abstr. Int. B 1990, 51(4), 1836-7

8.

A radical cyclization approach to the synthesis of morphine and synthetic approaches to trialkoxyphthalic acid derivatives Spero, Denice Mary (1988) 143 pp. Avail.: Univ. Microfihns Int., Order No. DA8825202 From: Diss. Abstr. Int. B 1989, 50(3), 970

9.

A study directed at the total synthesis of (-)-codeine and (-)- morphine: synthesis via a novel asymmetric intramolecular Diels-Alder reaction Costanzo, Michael John (1988) 276 pp. Avail.: Univ. Microfilms Int., Order No. DA8818767 From: Diss. Abstr. Int. B 1989, 49(7), 2647

10.

Approaches to the synthesis of morphine. Wan, Barbara Yu Fong (1987) 112 pp. Avail.: Univ. Microfilms Int., Order No. DA8715769 From: Diss. Abstr. Int. B 1987, 48(6), 1692

11.

The total synthesis of (.+-.)- morphine Toth, John Eldon (1986) 682 pp. Avail.: Univ. Microfihns Int., Order No. DA8709865 From: Diss. Abstr. Int. B 1987, 48(1), 143

118 12.

A novel approach to the synthesis of morphine. Hinton, Michael (1987) 153 pp. Avail.: Univ. Microfihns Int., Order No. DA8711350 From: Diss. Abstr. Int. B 1987, 48(2), 441

13.

A contribution toward the synthesis of morphine. Rodriguez, Cesar (1986) 155 pp. Avail.: Univ. Microfihns Int., Order No. DA8617023 From: Diss. Abstr. Int. B 1987, 47(7), 2922

14.

A study of the phenoHc oxidative coupling reaction in the synthesis of morphine alkaloids. An approach to the asymmetric synthesis of codeine Pham Phuong Thi Kim (1985) 157 pp. Avail: Univ. Microfihns Int., Order No. DA8529558 From: Diss. Abstr. Int. B 1986, 46(11), 3851

15.

New aromatic annulation methods: total syntheses of juncusol, sendaverine, and morphine -related analgesics Mullican, Michael David (1984) 163 pp. Avail: Univ. Microfihns Int., Order No. DA8513829 From: Diss. Abstr. Int. B 1985, 46(4), 1175-6

16.

A study of the phenolic oxidative couphng reaction in the synthesis of morphine alkaloids Vanderlaan, Douglas George (1984) 105 pp. Avail: Univ. Microfihns Int., Order No. DA8428711 From: Diss. Abstr. Int. B 1985, 45(11), 3512

17.

An approach to the morphine alkaloids: synthesis of 9-methoxy-3-methyl2,3,4,4a,5,6-hexahydro-lH-benzofuro[3,2-e]i>oquinohne-7(7aH)-ones Weller, Doreen L. (1984) 109 pp. Avail: Univ. Microfihns Int., Order No. DA8402152 From: Diss. Abstr. Int. B 1984, 44(11), 3412

18.

Studies directed toward the total synthesis of morphine. Hamann, Phihp Ross (1983) 684 pp. Avail: Univ. Microfihns Int., Order No. DA8407547 From: Diss. Abstr. Int. B 1984, 45(3), 875

19.

New methods in organic synthesis. Part I. Regioselective conversion of ketones into olefins via vinyl triflates. Part U. An approach to the total synthesis of morphine Scott, Wilham Johnston (1983) 188 pp. Avail: Univ. Microfilms Int., Order No. DA8321902 From: Diss. Abstr. Int. B 1983, 44(6), 1832

20.

Approaches to the synthesis of morphine. Harris, David Jude (1982) 159 pp. Avail: Univ. Microfihns Int., Order No. DA8220104 From: Diss. Abstr. Int. B 1982, 43(4), 1102

21.

The apphcation of metalated enamines to the synthesis of morphine alkaloids Mitch, Charles Howard (1982) 159 pp. Avail: Univ. Microfilms Int., Order No. DA8218846 From: Diss. Abstr. Int. B 1982, 43(3), 731

22.

Heteroatom directed photoarylation. Approaches to the synthesis of morphine and the study of a stereospecific benzodihydrofuran photorearrangement Napier, James Joseph (1981) 309 pp. Avail: Univ. Microfihns Int., Order No. 8119452 From: Diss. Abstr. Int. B 1981, 42(4), 1458-9

119 23.

Approaches to the synthesis of morphine McGowan, Cynthia Baker (1981) 96 pp. Avail.: Univ. Microfilms Int., Order No. 8116395 From: Diss. Abstr. Int. B 1981, 42(2), 636

24.

Biomimetic syntheses of several morphine alkaloid analogs Zoda, Michael Francis (1981) 96 pp. Avail.: Univ. Microfihns Int., Order No. 8113273 From: Diss. Abstr. Int. B 1981, 42(1), 225

25.

Synthetic approaches to morphine and colchicine alkaloid analogs Wallace, Rebecca Abemathy (1979) 147 pp. Avail.: Univ. Microfihns Int., Order No. 7926834 From: Diss. Abstr. Int. B 1980, 40(7), 3179

26.

A biogenetically patterned synthesis of the morphine alkaloids Mami, Ismail Sadeg (1978) 78 pp. Avail.: Univ. Microfilms Int., Order No. 7917053 From: Diss. Abstr. Int. B 1979, 40(2), 755-6

27.

Heteroatom directed photoarylation. AppUcation toward the synthesis of morphine Lucci, Robert Dominick (1977) 189 pp. Avail: Univ. Microfihns Int., Order No. 7807790 From: Diss. Abstr. Int. B 1978, 38(12, Pt. 1), 5942

28.

Cell division and macromolecular synthesis in Tetrahymena pyriformis. Action of tetrahydrocannabinol, morphine, levorphanol and levalloiphan McClean, Daniel K. (1972) No pp. given Avail.: Natl. Libr. Canada, Ottawa Ont From: Diss. Abstr. Int. B 1974, 34(9), 4258-9

29.

Effect of cycloheximide, an inhibitor of protein synthesis on the development of tolerance to morphine Feinberg, Michael P. (1973) 140 pp. Avail.: Univ. Microfilms, Ann Arbor, Mich., Order No. 7323,480 From: Diss. Abstr. Int. B 1973, 34(4), 16

30.

Excitatory actions of morphine and synthetic surrogates Brister, Calvin Cotten (1972) 146 pp. Avail.: Univ. Microfihns, Ann Arbor, Mich., Order No. 72-20,226 From: Diss. Abstr. Int. B 1972, 33(1), 351-2

31.

Synthesis of morphine isomers Chang, Jaw-Kang (1969) No pp. given Avail.: Natl. Libr. Canada, Ottawa, Ont From: Diss. Abstr. Int. B 1970, 31(4), 2157-8

32.

Synthesis of some derivatives of (-)-3-hydroxy-6- oxomorphinan structurally related to known analgesics and analgesic antagonists of the morphine type. Neubert, Mary E. (1968) 211 pp. Avail.: 68-15,850 From: Diss. Abstr. B 1968, 29(5), 1612-13

120 8. Appendix n. List of references connected to synthetic transformations of morphine and derivatives and biological testing 1.

Synthesis and analytical characterization of dansyl derivatives of morphine-like substances Hosztafi, Sandor; Repasi, Janos Acta Pharm. Hung. (1994), 64(1), 22-5

2.

The NMDA receptor antagonists, LY274614 and MK-801, and the nitric oxide synthase inhibitor, NG-nitro-L-arginine, attenuate analgesic tolerance to the muopioid morphinebut not to kappa opioids Elliott, Kathryn; Minami, Nobuko; Kolesnikov, Yuri A.; Pasternak, Gavril W.; Inturrisi, Charles E. Pain (1994), 56(1), 69-75

3.

Nitric oxide (NO) synthase inhibitors attenuated naloxone-precipitated withdrawal Dzoljic, M. R.; Cappendijk, S. L. T.; de Vries, R. Regul. Pept. (1994), (Suppl. 1), S285-S286

4.

Inhibition of nitric oxide synthase attenuates the development of morphine tolerance and dependence in mice Majeed, N. H.; Przewlocka, B.; Machelska, H.; Przewlocki, R. Neuropharmacology (1994), 33(2), 189-92

5.

Involvement of the nitric oxide pathway in nociceptive processes in the central nervous system in rats Przewlocka, B.; Machelska, H.; Przewlocki, R. Regul. Pept. (1994), (Suppl. 1), S75-S76

6.

Synthesis of N,C 10-bridged morphine derivatives: 5H-10,13 iminoethanophenanthro[4,5-bcd]furan. I Fleischhacker, W.; Richter, B.; Voellenkle, H. Monatsh. Chem. (1993), 124(8-9), 909-22

7.

Synthesis and analgetic activity of nicotinic esters of morphine derivatives Hosztafi, S.; Kohegyi, I.; Simon, C ; Furst, Z. Arzneim.-Forsch. (1993), 43(11), 1200-3

8.

Biochemical characterization of a synthetic NPFF agonist with high affinity and resistance to brain peptidase inactivation Devillers, J. P.; Reeve, A.; Mazarguil, H.; AUard, M.; Zajac, J M.; Dickenson, A. H.; Simonnet, G. Regul. Pept. (1994), (Suppl. 1), S123-S124

9.

Structure activity relationships of synthetic and semisynthetic opioid agonists and antagonists Hosztafi, Sandor; Friedmann, Tamas; Furst, Zsuzsanna Acta Pharm. Hung. (1993), 63(6), 335-49

10.

Inhibitory effect of nitric oxide (NO) synthase inhibitors on naloxone-precipitated withdrawal syndrome in morphine -dependent mice Cappendijk, Susan L. T.; de Vries, Rene; Dzoljic, Michailo R. Neurosci. Lett. (1993), 162(1-2), 97-100

121 11.

Attenuation of some signs of opioid withdrawal by inhibitors of nitric oxide synthase Kimes, Alane S.; Vaupel, D. Bruce; London, Edythe D. Psychopharmacology (BerUn) (1993), 112(4), 521-4

12.

Manufacture of multilayered controlled-release transdermal patches Wick, John; Weimann, Ludwig J.; Pollock, Wayne C. Eur. Pat. AppL, 35 pp. HP 92-850190 920813 PRAI US 92-861534 920401

13.

Method for synthesizing glucuronides of 4,5-epoxymorphinans Mertz, Alfred Adophe Henri PCX Int. AppL, 25 pp. WO 9305057 Al 930318 WO 92-FR846 920904 PRAI FR 91-10927 910904

14.

Morphine alkaloids. 120. Synthesis of N-demethyl-N-substituted 14.beta.hydroxy-isomorphine and dihydroisomorphine derivatives Hosztafi, Sandor; Simon, Csaba; Makleit, Sandor Heterocycles (1993), 36(7), 1509-19

15.

Morphine suppresses DNA synthesis in cultured murine astrocytes from cortex, hippocampus and striatum Stiene-Martin, Anne; Hauser, Kurt F. Neurosci. Lett. (1993), 157(1), 1-3

16.

Biological synthesis of the analgesic hydromorphone, an intermediate in the metaboUsm of morphine, by Pseudomonas putida MIO Hailes, Anne M.; Bruce, Neil C. AppL Environ. Microbiol. (1993), 59(7), 2166-70

17.

Effect of genetic obesity and phenobarbital treatment on the hepatic conjugation pathways Chaudhary, Inder P.; Tuntaterdtum, Somsong; McNamara, Patrick J.; Robertson, Larry W.; Blouin, Robert A. J. Pharmacol. Exp. Ther. (1993), 265(3), 1333-8

18.

Synthetic and natural opiates interact with P-glycoprotein in multidrug-resistant cells Callaghan, Richard; Riordan, John R. J. Biol. Chem. (1993), 268(21), 16059-64

19.

Inhibition of nitric oxide synthase enhances morphine antinociception in the rat spinal cord Przewlocki, Ryszard; Machelska, HaUna; Przewlocka, Barbara Life Sci. (1993), 53(1), PL1-PL5

20.

Inhibition of the morphine withdrawal syndrome by a nitric oxide synthase inhibitor, NG-nitro-L-arginine methyl ester Adams, Michael L.; Kahcki, Joelle M.; Meyer, Edward R.; Cicero, Theodore J. Life Sci. (1993), 52(22), PL245-PL249

122 21.

Blockade of tolerance to morphine but not to .kappa, opioids by a nitric oxide synthase inhibitor Kolesnikov, Yuri A.; Pick, Chaim G.; Ciszewska, Grazyna; Pasternak, Gavril W. Proc. Natl. Acad. Sci. U. S. A. (1993), 90(11), 5162-6

22.

Enzymatic hydroxylation of arene and synunetry considerations in efficient synthetic design of oxygenated natur^d products HudUcky, Tomas; Fan, Rulin; Luna, Hector; Olivo, Horacio; Price, John Indian J. Chem., Sect. B (1993), 32B(1), 154-8

23.

Morphine regulates DNA synthesis in rat cerebellar neuroblasts in vitro Hauser, Kurt F. Dev. Brain Res. (1992), 70(2), 291-7

24.

Morphine alkaloids. 119. A new efficient method for the preparation of 2-fluoroN-propylnorapomorphine Berenyi, Sandor; Hosztafi, Sandor; Makleit, Sandor J. Chem. Soc, Perkin Trans. 1 (1992), (20), 2693-4

25.

Synthesis of a new morphine derivative with anorexogenic activity Berenyi, Sandor; Makleit, Sandor; Hosztafi, Sandor; Furst, Susanna; Friedmann, Tamas; Knoll, Jozsef Med. Chem. Res. (1991), 1(3), 185-90

26.

Synthesis of N-demethyl-N-substituted-14-hydroxycodeine and morphine derivatives Hosztafi, Sandor; Simon, Csaba; Makleit, Sandor Synth. Commun. (1992), 22(17), 2527-41

27.

NG-Nitro-L-arginine prevents morphine tolerance Kolesnikov, Yuri A.; Pick, Chaim G.; Pasternak, Gavril W Eur. J. Pharmacol. (1992), 221(2-3), 399-400

28.

An improved synthesis of noroxymorphone Ninan, Aleyamma; Sainsbury, Malcolm Tetrahedron (1992), 48(32), 6709-16

29.

Inhibition of estradiol-induced DNA synthesis by opioid peptides in the rat uterus Ordog, Tamas; Vertes, Zsuzsanna; Vertes, Marietta Life Sci. (1992), 51(15), 1187-96

30.

Morphine-induced downregulation of .mu.-opioid receptors and peptide synthesis in neonatal rat brain Tempel, Ann; Espinoza, Kathryn Ann. N. Y. Acad. Sci. (1992), 654(Neurobiol. Drug Alcohol Addict.), 529-30

31.

Design and synthesis of an opioid receptor probe: mode of binding of S-activated(-)-6.beta.-sulfhydryldihydromorphine with the sulfhydryl group in the .mu.-opioid receptors Kanematsu, Ken; Kaya, Tetsudo; Shimohigashi, Yasuyuki; Yagi, Kunio; Ogasawara, Tomio Med. Chem. Res. (1991), 1(3), 191-4

123 32.

Synthesis of N-substituted C-normorphinans and their pharmacological properties Takeda, Mikio; Inoue, Hirozumi; Noguchi, Katsuyuki; Honma, Yasushi; Okamura, Kimio; Date, Tadamasa; Nurimoto, Seiichi; Yamamura, Michio; Saito, Seiichi Chem. Pharm. Bull. (1992), 40(5), 1186-90

33.

Lx>ng term effects of morphine on mesangial cell proliferation and matrix synthesis Singhal, Pravin C ; Gibbons, Nora; Abramovici, Mirel Kidney Int. (1992), 41(6), 1560-70

34.

Facile syntheses of aporphine derivatives Hedberg, Martin H.; Johansson, Anette M.; Hacksell, Uh J. Chem. Soc, Chem. Conmiun. (1992), (11), 845-6

35.

Morphine alkaloids. Part 116. Synthesis of N-demethyl-N-substituted dihydroisomorphine and dihydroisocodeine derivatives Hosztafi, Sandor; Simon, Csaba; Makleit, Sandor Synth. Commun. (1992), 22(12), 1673-82

36.

Morphine alkaloids. Part 114. A stereohomogeneous synthesis of N-demethyl-N-substituted-14hydroxydihydromorphines Hosztafi, Sandor; Berenyi, Sandor; Toth, Geza; Makleit, Sandor Monatsh. Chem. (1992), 123(5), 435-41

37.

Structure-activity smdies of morphine fragments. El. Synthesis, opiate receptor binding, analgetic activity and conformational studies of spiro-[tetralin-l,4'-piperidines] Lawson, J. A.; Toll, L.; Polgar, W.; Uyeno, E. T.; Loew, G. H. Eur. J. Med. Chem. (1991), 26(8), 775-85

38.

Morphine alkaloids. 113. Synthesis of C-3 halogen-substituted apocodeines and apomorphines Simon, Csaba; Hosztafi, Sandor; Makleit, Sandor; Berenyi, Sandor Synth. Commun. (1991), 21(22), 2309-16

39.

Structure-activity studies of morphine fragments, n. Synthesis, opiate receptor binding, analgetic activity and conformational studies of 2-R-2(hydroxybenzyl)piperidines Loew, G. H.; Lawson, J. A.; Toll, L.; Polgar, W.; Uyeno, E. T. Eur. J. Med. Chem. (1991), 26(8), 763-73

40.

Controlled release pharmaceutical preparation and process for preparing same Zsuga, Miklos; Kelen, Tibor; Nagy, Jozsef; Barkanyi, Judit; Bene, Magdolna; Ondi, Sandor; Gulyas, Imre; Gyoeker, Istvan; Repasi, Janos; Repasi, Agota Eur. Pat. AppL, 6 pp. EP 463833A2 920102 AI EP 91-305669 910624 PRAI HU 90-4007 900627

41.

Coordination compounds as precursors for materials synthesis Langfelderova, H.; Papankova, B.; Makanova, D.; Gersi, P.; Kozisek,

124 J. Proc. Conf. Coord. Chem. (1991), 13th, 149-54 42.

Sustained-release pharmaceutical mucosal patches Scholz, Matthew T.; Scherrer, Robert A.; Marecki, Nelda M.; Barkhaus, Joan K.; Chen, Yen Lane PCT Int. AppL, 48 pp. PI WO 9106290 Al 910516 DS W: AU, BR, CA, JP, KR RW: AT, BE, CH, DE, DK, ES, FR, GB, GR, IT, LU, NL, SE AI WO90-US6505 901102 PRAI US 89-431664 891103

43.

Effect of opioids on the activity of some key enzymes involved in milk synthesis in manmiary gland of lactating rabbit Hossain, M. A.; GanguLL, N. C. Indian Vet. J. (1991), 68(7), 630-5

44.

Design and synthesis of HTV protease inhibitors. Variations of the carboxyterminus of the HTV protease inhibitor L-682,679 DeSolms, S. Jane; Giuhani, Elizabeth A.; Guare, James P.; Vacca, Joseph P.; Sanders, William M.; Graham, Samuel L.; Wiggins, J. Mark; Darke, Paul L.; Sigal, Irving S.; et al. J. Med. Chem. (1991), 34(9), 2852-7

45.

Inhibition of cell growth and DNA, RNA, and protein synthesis in vitro by fentanyl, sufentanil, and opiate analgesics Nassiri, M. Reza; Flynn, Gordon L.; Shipman, Charles, Jr. Pharmacol. Toxicol. (Copenhagen) (1991), 69(1), 17-21

46.

Opioid involvement in the control of melatonin synthesis and release Stankov, B.; Esposti, D.; Esposti, G.; Lucini, V.; Mariani, M.; Cozzi, B.; Scaglione, F.; Fraschini, F. Adv. Pineal Res. (1990), 4, 45-8

47.

Preparation of racemic and optically-active fatty amino acids, their homo- and hetero-oligomers and conjugates, as pharmaceuticals Gibbons, WilHam A. Brit. UK Pat. AppL, 55 pp. PI GB 2217319 Al 891025 AI GB 88-9162 880419

48.

New method for synthesis of tricyclic morphine analog Zhang, Yongmin; Zhang, Lihe; Liu, Weiqin; Thai, C ; Labidalle, S. Huaxue Xuebao (1990), 48(10), 1030-5

49.

Assay of semisynthetic codeine base with simultaneous determination of.alpha.-codeimethine and 06-codeine methyl ether as by-product impurities by high-performance Uquid chromatography Ayyangar, N. R.; Bhide, S. R.; Kalkote, U. R. J. Chromatogr. (1990), 519(1), 250-5

50.

Effects of morphine in arachidonic acid metaboHsm, of calcium-uptake and on cAMP synthesis in uterine strips from spayed rats Faletti, A.; Bassi, D.; Franchi, A. M.; Gimeno, A. L.; Gimeno, M. A.F.

125 Prostaglandins, Leukotrienes Essent. Fatty Acids (1990), 41(3), 151-5 51.

Design and synthesis of an opioid receptor probe: mode of binding of S-activated (-)-6.beta.-sulfhydryldihydromorphine with the sulfhydryl group in the .mu.-opioid receptor Kanematsu, Ken; Naito, Ryo; Shimohigashi, Yasuyuki; Ohno, Motonori; Ogasawara, Tomio; Kurono, Masayasu; Yagi, Kunio Chem. Pharm. BuU. (1990), 38(5), 1438-40

52.

Synthesis and analgesic activity of sulfur-containing morphinans and related comf)ounds Hori, Mikio; Iwamura, Tatsunori; Imai, Eiji; Shimizu, Hiroshi; Kataoka, Tadashi; Nozaki, Masakatsu; Niwa, Masayuki; Fujimura, Hajime Chem. Pharm. Bull. (1989), 37(5), 1245-8

53.

A novel synthesis of pyrimidobenzodiazepines Dlugosz, Anna Arch. Pharm. (Weinheim, Ger.) (1990), 323(1), 59-60

54.

Oxidative coupling of cis-3,N-bis(methoxycarbonyl)-N-norreticuline: an approach to the asymmetric synthesis of morphine alkaloids Schwartz, Martin A.; Pham, Phuong T. K. Adv. Biosci. (Oxford) (1989), 75(Prog. Opioid Res.), 121-4

55.

Synthetic opioids compared with morphine and ketamine: catalepsy, cross-tolerance and interactions in the rat Benthuysen, J. L.; Hance, A. J.; Quam, D. D.; Winters, W. D. Neuropharmacology (1989), 28(10), 1011-15

5 6.

Preparation and use of monomeric phthalocyanine reagents Stanton, Thomas H.; Schindele, Deborah C ; Renzoni, George E.; Pepich, Barry V.; Anderson, Neils H.; Clagett, James A.; Opheim, Kent E. PCT Int. Appl., 54 pp. PI WO 8804777 Al 880630 AI W 0 87-US3226 871211 PRAI US 86-941619 861215 US 86-946475 861224 US 87-61937 870612

57.

Synthesis and antinociceptive activity of thiohydantoin derivatives Xu, Guoyou; Yu, Zhengwei; Peng, Sixun Zhongguo Yaoke Daxue Xuebao (1988), 19(4), 245-8

58.

Enantioselective inhibition: strategy for improving the enantioselectivity of biocatalytic systems Guo, Zhi Wei; Sih, Charles J. J. Am. Chem. Soc. (1989), 111(17), 6836-41

60.

Attenuation of morphine withdrawal syndrome by macromolecular synthesis inhibitors in rats Copeland, Robert L., Jr.; Pradhan, S. N. Drug Dev. Res. (1989), 17(2), 169-74

60

Morphine alkaloids. 104. Synthesis and conversions of new epoxy derivatives

126 Gulyas, Gyongyi; Berenyi, Sandor; Makleit, Sandor Acta Chim. Hung. (1988), 125(2), 255-65 61.

Enzymic synthesis and immunocheniical characteristics of antigens of morphine -protein conjugates Kovalev, I. E.; Tomilin, V, A. Farmakol. Toksikol. (Moscow) (1989), 52(3), 62-6

62.

Synthesis of l,8-diaryl-2,7-di(hexamethyleneiniinomethyl)1,8-octanedione dihydrochlorides and their analgesic properties Petrosyan, L. M.; Gevorgyan, G. A.; Durgaryan, L. K.; Azhvyan, A.S.; Vlasenko, E. V.; Mndzhoyan, O. L. Khim.-Farm. Zh. (1988), 22(9), 1073-6

63.

8-Phenylmorphinones and 8-phenylcodeinones: synthesis and analgesic activity KaUnin, V. N.; Kazantseva, S. A.; Petrovskii, P. V.; Kobel'kova, N.I.; Ignatov, Yu. D.; Zvartau, E. E.; Dorokhova, M. I. Khim.-Fann. Zh. (1989), 23(1), 48-50

64.

Design and synthesis of suLfur-containing morphine and an opioid receptor probe Fujii, Dcuo; Togame, Hiroko; Yamamoto, Mayumi; Kanematsu, Ken; Takayanagi, Issei; Konno, Fukio Chem. Pharm. Bull. (1988), 36(6), 2282-5

65.

Peptides related to leucine-Zmethionine-enkephalinamides: synthesis and biological activities Sivanandaiah, K. M.; Gurusiddappa, S.; Suresh Babu, V. V. Indian J. Chem., Sect. B (1988), 27B(7), 645-8

66.

Synthesis of 8-aryldihydrocodeinones and -morphinones via palladium-catalyzed reactions Kalinin, V. N.; Kazantseva, S. A.; Petrovskii, P. V.; Kobel'kova, N.I.; Polyakov, A. v.; Yanovskii, A. I.; Struchkov, Yu. T. Dokl. Akad. Nauk SSSR (1988), 298(1), 119-22 [Chem.]

67.

Electrosynthesis of morphine-derivatives, isoquinolines, and cyclopentanoids Schaefer, H. J.; Schlegel, C ; Eilenberg, W.; Mueller, U.; Huhtasaari, M.; Becking, L. F.E.C.S. Int. Conf. Chem. Biotechnol. Biol. Act. Nat. Prod., [Proc.], 3rd (1987), Meeting Date 1985, Volume 1, 96-116 Publisher: VCH, Weinheim, Fed. Rep. Ger.

68.

Morphine analogs derived from tetra- and hex^ydrobenzofurans. II. Synthesis and stereoselective functionahzation of dibenzofurans Labidalle, Serge; Min, Zhang Yong; Reynet, Annick; Moskowitz, Henri; Vierfond, Jean Michel; Miocque, Marcel; Bucourt, Robert; Thal,Claude Tetrahedron (1988), 44(4), 1171-86

69.

Morphine analogs derived from tetra-and hexahydrobenzofurans. I. Synthesis and stereospecific epoxidation of arylalkylcyclohexane intermediates Labidalle, Serge; Min, Zhang Yong; Reynet, Annick; Moskowitz, Henri; Vierfond, Jean Michel; Miocque, Marcel; Bucourt, Robert; Thal,Claude

127 Tetrahedron (1988), 44(4), 1159-69 70.

Some pharmacological properties of a newly synthesized morphine derivative, (-)-6.beta.-acetylthiomorphine Takayanagi, I.; Konno, F.; Goromaru, N.; Koike, K.; Kanematsu, K.; Fujii, L; Togame, H. Arch. Int. Pharmacodyn. Ther. (1988), 294, 71-84

71.

Narcotic alkylating agents: synthesis, structure and biological activities Frigola, J.; Colombo, A.; Mas, J.; Pares, J. Farmaco, Ed. Sci. (1988), 43(4), 347-62

7 2.

Preparation of new echibohne derivatives having opioid properties Robinson, Brian; Rees, John Michael Hugh; Cox, Brian PCX Int. Appl., 40 pp. PI WO 8800193 Al 880114 WO 87-GB457 870630 GB 86-16089 860702

73.

Synthesis and carbon-13 NMR study of some podocarpic acid derivatives Ortellado, Maria AmeUa A. C ; MarsaioH, Anita J. J. Chem. Res., Synop. (1987), (10), 324-5

74.

Oxidative coupling of cis-3,N-bis(methoxycarbonyl)-N-norreticuline. An approach to the asymmetric synthesis of morphine alkaloids Schwartz, Martin A.; Pham Phuong Thi Kim J. Org. Chem. (1988), 53(10), 2318-22

75.

General asymmetric synthesis of morphine and its analogs via enantioselective hydrogenation Kitamura, Masahito; Hsiao, Yi; Noyori, Ryoji; Takaya, Hidemasa Tennen Yuki Kagobutsu Toronkai Koen Yoshishu (1987), 29, 385-92

76.

Alcohol and morphine affect protein synthesis of astrogUa-enriched primary cultures from various brain regions Roennbaeck, L.; Hansson, E. Neurol. Neurobiol. (1988), 39(Biochem. Pathol. Astrocytes), 215-17

77.

Synthesis of carbon-11 labeled diprenorphine: a radioUgand for positron emission tomographic studies of opiate receptors Lever, John R.; Dannals, Robert F.; Wilson, Alan A.; Ravert, Hayden T.; Wagner, Henry N., Jr. Tetrahedron Lett. (1987), 28(35), 4015-18

78.

Synthesis and conversion of new epoxy derivatives of morphine alkaloids Gulyas, Gyongyi; Berenyi, Sandor; Makleit, Sandor Magy. Kem. Foly. (1986), 92(11-12), 508-13

7 9.

Synthesis of the skeleton of the morphine molecule by mammalian hver Weitz, Charles J.; FauU, Kym F.; Goldstein, Avram Nature (London) (1987), 330(6149), 674-7

80.

Improved synthesis of 3-substituted 7-methoxybenzofurans. Useful intermediates for the preparation of morphine analogs of organic chemistry

128 Jung, Michael E.; Abrecht, Stefan J. Org. Chem. (1988), 53(2), 423-5 81.

Pharmacological evidence of morphine-induced inhibition of gastric mucus synthesis in rats Del Tacca, M.; Bemardini, C ; Corsano, E.; Bertelli, A.; Roze, C. Int. J. Tissue React. (1987), 9(5), 413-18

82.

The relative stabilities of substituted cis- and trans-l,2,3,4,4a,9,l0,10aoctahydrophenanthrenes, including configurational corrections in the EladGinsburg morphine synthesis Duthaler, Rudolf O.; Ginsburg, David Helv. Chim. Acta (1986), 69(7), 1559-66

8 3.

Synthesis of N- [methyl-11 C]hydromorphone by using multivariate strategies for optimization of radiochemical yields Rimland, Annika; Bergson, Goeran; Obenius, Ulf; Sjoeberg, Stefan; Laangstroem, Bengt Appl. Radiat. Isot. (1987), 38(8), 651-4

84.

Synthesis of metabolites of 3-O-tert-butylmorphine Mohacsi, E. J. Heterocycl. Chem. (1987), 24(2), 471-2

85.

Synthetic studies on morphine. Racemization of biaryl intermediates Weller, Dwight D.; Runyan, Mark T. Tetrahedron Lett. (1986), 27(40), 4829-32

86.

Synthesis, purification and biological evaluation of porcine corticotropin-releasing factor Guoth, J.; Olsen, D. B.; Kovacs, M.; Schally, A. V. Life Sci. (1987), 41(8), 1003-10

87.

Synthesis and pharmacological properties of morphine congeners having pendant crown ether as an opioid receptor probe Fujii, Ecuo; Togame, Hiroko; Kanematsu, Ken Chem. Pharm. Bull. (1986), 34(10), 4439-42

88.

In vivo tracer studies of glucose metabolism, cerebral blood flow, and protein synthesis in naloxone precipitated morphine withdrawal Geary, W. A., II; Wooten, G. F. Neurochem. Res. (1987), 12(7), 573-80

89.

Syntheses of 3-0- and 6-O-propanoylmorphine. A reinvestigation and correction Sy, Wing Wah; By, Arnold W.; NeviUe, George A.; Wilson, Wilham L. J. Pharm. Sci. (1986), 75(8), 787-9

90.

Further syntheses with nitroxide .alpha.,.beta.-unsaturated aldehydes and allyhc bromides Hideg, Kalman; Cseko, Jozsef; Hankovszky, H. Olga; Sohar, Pal Can. J. Chem. (1986), 64(8), 1482-90

91.

Synthesis of morphine analogs with a tetrahydrodibenzofuran skeleton

129 Labidalle, S.; Min, Zhang Yong; Reynet, A.; Thai, C; Moskowitz, H. Tetrahedron Lett. (1986), 27(25), 2861-2 9 2.

Synthesis of labelled morphine compounds Toth, Geza; Sirokman, Ferenc; Hosztafi, Sandor Izotoptechnika (1985), 28(4), 226-36

93.

Aporphines 65: Chemical, microbial synthesis and characterization by gas chromatography/mass spectrometry of (R)-(-)-10-hydroxy 11-methoxy-N-npropylnoraporphine, a potential metabolite of N-n-propylnorapomorphine Neumeyer, John L.; Abdel-Maksoud, Hamdy M.; Trainor, Thomas M.; Vouros, Paul; Davis, Patrick J. Biomed. Environ. Mass Spectrom. (1986), 13(5), 223-9

94.

Effects of morphine on DNA synthesis in neonatal rat brain Komblum, Harley I.; Loughhn, Sandra E.; Leshe, Frances M. Dev. Brain Res. (1987), 31(1), 45-52

95.

Long-lasting inhibitory effects of a synthetic enkephalin (FK 33-824) on evoked neuronal thalamic firing in the rat Braga, Pier Carlo; BieUa, Gabriele; Rigoh, Milena; Tiengo, Mario; Fraschini, Franco; Netti, Carmela; Guidobono, Francesca IRCS Med. Sci. (1986), 14(12), 1181-2

96.

Total synthesis of rac-salutaridine and sinoacutine [(-)-salutaridine] - a new method to prepare morphine Ludwig, Wolfgang; Schaefer, Hans J. Angew. Chem. (1986), 98(11), 1032-3

97.

Stimulation of brain-stem protein synthesis.by morphine Roennbaeck, Lars; Hansson, EHsabeth Biochem. Pharmacol. (1986), 35(21), 3685-92

98.

Precursors of the mammahan synthesis.of morphine: (-i-)-salutaridine and (-)thebaine from (-i-)-6-demethylsalutaridine, and (-)-N-13CH3-thebaine from (-)northebaine Dumont, Raymond; Newman, Amy Hauck; Rice, Kenner C; Brossi, A.; Toome, Voldemar; Wegrzynski, Bogda FEBS Lett. (1986), 206(1), 125-9

99.

Synthesis of N-p-azidophenylethyl-7,8-dihydronormorphine and its 7,8-ditritio analog. Potential opiate receptor photoaffmity labels Cooper, Geoffrey K.; Rapoport, Henry J. LabeUed Compd. Radiopharm. (1985), 22(11), 1201-7

100.

Status of opioids in bovine milk and its influence on enzymes in milk synthesis Hossain, M. Afzal; Kumar, Amalendra; Ganguh, N. C. Indian J. Anim. Sci. (1986), 56(5), 588-92

101.

Purifying codeine synthesized from morphine Cemy, Jozef; Proksa, Bohumil; Brezovsky, Zdenko Czech., 2 pp. PI CS 223100 B 860315 AI CS 82-2895 820423

130 102.

Effects of synthetic.analogs of enkephalins, morphine and their antagonists on the course of experimental traumatic shock Motin, V. G.; Yasnetsov, V. V. Farmakol. Toksikol. (Moscow) (1986), 49(3), 103-7

103.

Morphine enhances gastric mucus synthesis in rats Ho, Mai M.; Ogle, Clive W.; Dai, Soter Eur. J. Pharmacol. (1986), 122(1), 81-6

104.

A short, stereospecific synthesis of a morphine fragment via an intramolecular Diels-Alder reaction Handa, Sheetal; Jones, Keith; Newton, Christopher G.; WiUiams, David J. J. Chem. Soc, Chem. Commun. (1985), (19), 1362-3

105.

Amino acid incorporation during morphine intoxication: I: Dose and time effects of morphine on protein synthesis in specific regions of the rat brain and in astrogUa-enriched primary cultures Roennbaeck, L.; Hansson, E. J. Neurosci. Res. (1985), 14(4), 461-77

106.

Analysis of the analgesic and anti-inflammatory effects of rimazolium, a pyridopyrimidine derivative, compared with that of prostaglandin synthesis inhibitors and morphine Gyires, K.; Furst, S.; Miklya, I.; Budavari, L; Knoll, J. Drugs Exp. CHn. Res. (1985), 11(8), 493-500

107.

Studies directed at a synthesis of the morphine alkaloids. Regiocontrol in Robinson-type annulations of 2-(hydroxymethyl)-4-oxo-3-piperidinecarboxyUc acid .gamma.-lactones Schultz, Arthur G.; Shannon, Paul J. J. Org. Chem. (1985), 50(23), 4421-5

108.

Isolation, sequencing, synthesis, and pharmacological characterization of two brain neuropeptides that modulate the action of morphine Yang, H. Y. T.; Fratta, W.; Majane, E. A.; Costa, E. Proc. Natl. Acad. Sci. U. S. A. (1985), 82(22), 7757-61

109.

Morphine-like structures and semi-rigid side chains as molecular probes for the bioactive conformation of enkephalin. Cardinaux, Francis; Pless, Janos Pept., Proc. Eur. Pept. Symp., 18th (1984), 321-4. Editor(s): Ragnarsson, Ulf. PubHsher: Almqvist & Wiksell, Stockholm, Swed.

110.

Opiate dependence and withdrawal - a new synthesis Hendrie, Colin A. Pharmacol., Biochem. Behav. (1985), 23(5), 863-70

111.

A direct synthesis of 06-monoacetylmorphine from morphine Sy, Wing Wah; By, Arnold W.; Neville, George A.; Wilson, WilUam L. J. - Can. Soc. Forensic Sci. (1985), 18(2), 86-91

112.

1 -Substituted isoquinoUnes Hendrickson, James B.; Rodriguez, Cesar

131 U.S., 7 pp. PI US 4514569 A 850430 AI US 82-343407 820128 113.

Chemistry of natural products, LXXXXV. Synthesis of 4,5.alpha.-epoxy17,6',6'-trimethyl-6'H-morphin-7-eno[3,2-b]pyran- 6.beta.-ol via PTC reaction Reisch, Johannes; Dharmaratne, H. Ranjith W. Arch. Pharm. (Weinheim, Ger.) (1985), 318(4), 382-4

114.

Recent progress on the total synthesis of morphine Szantay, Csaba; Blasko, Gabor; Barczai-Beke, Marietta; Domyei,Gabor; Pechy, Peter Magy. Kem. Lapja (1985), 40(1), 37-42

115.

Biochemical correlates of morphine withdrawal. 1. Characterization in the adrenal medulla and locus ceruleus DiStefano, Peter S.; Brown, Ohver M. J. Pharmacol. Exp. Ther. (1985), 233(2), 333-8

116.

Preparation of optically active, functionalized cis-.DELTA.6-1 -octalones Boger, Dale L.; Mullican, Michael D. J. Org. Chem. (1985), 50(11), 1904-11

117.

Synthesis and determination of 10-oxomorphine, a decomposition product of morphine Proksa, B. Pharmazie (1984), 39(10), 687-8

118.

Metorphamide - a morphine-like peptide Barchas, Jack D.; Weber, Eckard; Evans, Christopher J. U.S., 5 pp. PI U S 4490363 A 841225 AI U S 83-508140 830624

119.

Comparative studies on morphine- and stress-induced analgesia and the development of tolerance to the effects: implication of protein synthesis mechanism in the process Takahashi, Masakatsu; Izumi, Reiko; Kaneto, Hiroshi Jpn. J. Pharmacol. (1985), 37(2), 197-202

120.

Studies directed at a synthesis of the morphine alkaloids. A photochemical approach Schultz, Arthur G.; Lucci, Robert D.; Napier, James J.; Kinoshita, Hideki; Ravichandran, Ramanathan; Shannon, Paul; Yee, Ying K. J. Org. Chem. (1985), 50(2), 217-31

121.

Synthesis of the new morphine partial structure- 15,16-secomorphinan Helv. Chim. Acta (1984), 67(6), 1598-602

122.

In vivo alteration in hypothalamic amino acid synthesis during perfusion of ethanol and morphine in unrestrained rat Noto, T.; Myers, R. D. Neurochem. Res. (1984), 9(11), 1653-65

132 123.

Profile of amino acid synthesis in rat hippocampus during push-pull perfusion of ethanol or morphine Noto, T.; Hepler, J. R.; Myers, R. D. Neuroscience (Oxford) (1984), 13(2), 367-76

124.

Participation of opioid receptors and protein- synthetic mechanisms in the development of tolerance to morphine Hirota, Noritake; YamazaM, Akira; Takahashi, Masakatsu; Kaneto, Hiroshi Yakubutsu, Seishin, Kodo (1984), 4(1), 85-6

125.

Anew synthesis of morphine-based analgesics Chandler, Malcohn; Parson, Philip J. J. Chem. Soc, Chem. Conmiun. (1984), (5), 322-3

126.

Morphine diesters. I. Synthesis and action on guinea pig ileum Owen, J. A.; Elliott, J.; Jhamandas, K.; Nakatsu, K. Can. J. Physiol. Pharmacol. (1984), 62(4), 446-51

127.

Diastereoisomeric quaternary morphinium salts: synthesis stereochemistry and analgesic properties lorio, Maria A.; DisciuUo, Anna; Mazzeo-Farina, Anna; Frigeni, Viviana Eur. J. Med. Chem. - Chim. Ther. (1984), 19(1), 11-16

128.

Increase in 5-HT synthesis in the dorsal part of the spinal cord, induced by a nociceptive stimulus: blockade by morphine Weil-Fugazza, Jeanne; Godefroy, Francoise; Le Bars, Daniel J. Brain Res. (1984), 297(2), 247-64

129.

Studies aimed at the synthesis of morphine. 7. Biomimetic total synthesis of (.-i-.)-pallidine Blasko, Gabor; Domyei, Gabor; Barczai-Beke, Marietta; Pechy, Peter; Szantay, J. Org. Chem. (1984), 49(8), 1439-41

130.

Identification and determination of by-products of the codeine synthesis Proksa, B.; Cemy, J. Chem. Zvesti (1983), 37(6), 837-42

131.

Dynorphin amide analogs Lee, Nancy M.; Loh, Horace H.; Chang, Jaw Kang Eur. Pat. Appl., 24 pp. PI EP96592A1 831221 AI EP 83-303313 830608 PRAI US 82-387005 820610

132.

Studies on the synthesis of morphine. VI. Recent progress on the biomimetic synthesis of morphine Szantay, Csaba; Blasko, G.; Barczai-Beke, Marietta; Domyei, G.; Pechy, P. Planta Med. (1983), 48(4), 207-11

133.

A new synthetic metiiod for tiie preparation of tert-butyl aryl ethers Mohacsi, Emo Synth. Commun. (1983), 13(10), 827-32

133 134.

DELTA. 16,17-Dehydroheroinium chloride: synthesis and characterization of a novel impurity detected in illicit heroin Allen, Andrew C ; Moore, James M.; Cooper, Donald A. J. Org. Chem. (1983), 48(22), 3951-4

135.

The effect of morphine on 5 -hydroxytryptamine synthesis and metabolism in the striatum, and several discrete hypothalamic regions of the rat brain Johnston, C. A.; Moore, K. E. J. Neural Transm. (1983), 57(1), 65-73

136.

Effect of prostaglandin synthetase inhibitors on non-analgesic actions of morphine Wallenstein, Martin C. Eur. J. Pharmacol. (1983), 90(1), 65-73

137.

Mechanism of inhibition of protein synthesis by morphine in rat brain and hver Retz, Konrad C ; Steele, WiUiam J. Mol. Pharmacol. (1982), 22(3), 706-14

138.

Alkaloids. PartXLV. Partial synthesis of (-)-butorphanol Mouralova, J.; Hajicek, J.; Trojanek, J. Cesk. Farm. (1983), 32(1), 23-6

139.

An inhibitory role for morphine on the release of dopamine into hypophysial portal blood and on the synthesis of dopamine in tuberoinfundibular neurons Reymond, Marianne J.; Kaur, Charanjit; Porter, John C. Brain Res. (1983), 262(2), 253-8

140.

Blockade of morphine dependence-related enhancement of secretory protein synthesis in the pons-medulla and striatum-septum by naltrexone Retz, K. C ; Steele, W. J. Neuropharmacology (1983), 22(2), 183-9

141.

Total synthesis of (.+-.)-3-deoxy-7,8-dihydromorphine, (.-i-.)-4-methoxy-Nmethylmorphinan-6-one and 2,4-dioxygenated (.+-.)-congeners Hsu, Fu Lian; Rice, Kenner C ; Brossi, Arnold Helv. Chim. Acta (1982), 65(5), 1576-89

142.

Stereospecific synthesis of a new morphine partial structure Haefliger, Walter; Kloeppner, Edgar Helv. Chim. Acta (1982), 65(6), 1837-52

143.

Tryptophan and tryptophan pyrrolase in heme regulation. The role of hpolysis and direct displacement of serum protein-bound tryptophan in the opposite effects of administration of endotoxin, morphine, palmitate, salicylate, and theophylline on rat hver 5-aminolevulinate synthase activity and the heme saturation of tryptophan pyrrolase Badaway, Abdulla A. B.; Morgan, Christopher J. Biochem. J. (1982), 206(3), 451-60

144.

How are morphine-Hke peptides synthesized?

134 Imura, Hiroo Gendai Kagaku (1982), 140, 28-31 145.

Myocardial protein synthesis in magnesium deficiency: effects of morphine Essman, Walter B.; Essman, Eric J. Monogr. Am. Coll. Nutr. (1982), 5(Nutr. Heart Dis.), 309-14

146.

Synthesis of some carbon-11 -labeled alkaloids Laangstroem, B.; Antoni, G.; Halldin, C.; Svaerd, H.; Bergson, G. Chem. Scr. (1982), 20(1-2), 46-8

147.

Biological activity of some synthesized compounds of the pyridine series Poddubnaya, L. V.; Olekhnovich, L. B. Izv. Sev.-Kavk. Nauchn. Tsentra Vyssh. Shk., Estestv. Nauki (1982), (1), 87-9

148.

Aporphines. 42. Synthesis of (R)-(-)-ll-hydroxyaporphines Ram, Vishnu J.; Neumeyer, John L. J. Org. Chem. (1982), 47(22), 4372-4

149.

Studies aimed at the synthesis of morphine. V. An economy approach to (.+-.)-reticuline from 3,4-dihydropapaverine Domyei, Gabor; Barczai-Beke, Marietta; Balsko, Gabor; Pechy, Peter; Szantay, Tetrahedron Lett. (1982), 23(28), 2913-16

150.

Synthesis and pharmacology of metabolically stable tert-butyl ethers of morphine and levorphanol Mohacsi, Emo; Leimgruber, Willy; Baruth, Herman J. Med. Chem. (1982), 25(10), 1264-6

151.

.beta.-Endorphin. Interaction of synthetic analogs having different chain lengths with morplune and enkephalin receptors in rat brain membranes Ferrara, Pascual; Li, Choh Hao Int. J. Pept. Protein Res. (1982), 19(3), 259-62

152.

Evaluation of the analgesic activity of morphine, enkephalins and their synthetic analogs Aloisi, P.; Scotti de Carolis, A.; Longo, V. G. Isr. J. Med. Sci. (1982), 18(1), 183-6

153.

Studies aimed at the synthesis of morphine. FV. A new approach to Nnorreticuline derivatives from homoveratronitrile Szantay, Csaba; Domyei, Gabor; Blasko, Gabor; Barczai-Beke, Marietta; Pechy, Peter Arch. Pharm. (Weinheim, Ger.) (1981), 314(12), 983-91

154.

Synthesis and biological properties of enkephalin-like peptides containing adamantylalanine in position 4 and 5 Kim Quang Do; Schwyzer, Robert Helv. Chim. Acta (1981), 64(7), 2084-9

155.

Studies aimed at the synthesis of morphine. 3. Synthesis of (.+-.)-salutaridine via phenoUc oxidative coupling of (.-i-.)-reticuline Szantay, Csaba; Barczai-Beke, Marietta; Pechy, Peter; Blasko, Gabor;

135 Domyei, Gabor J. Org. Chem. (1982), 47(3), 594-6 156.

Synthesis of serotonin at the spinal level in the rat: modifications induced by nociceptive somatic stimulation, associated or not with administration of morphine Weil-Fugazza, Jeanne; Godefroy, Francoise; Chitour, Djamel; Le Bars, Daniel C. R. Seances Acad. Sci., Ser. 3 (1981), 293(1), 89-92

157.

(-)-4-Hydroxymorphinanones: their synthesis and analgesic activity Manmade, Awinash; Dalzell, Haldean C ; Howes, John P.; Razdan, Raj K. J. Med. Chem. (1981), 24(12), 1437-40

158.

Restoration of pituitary prolactin synthesis and release by the administration of morphine to rats bearing a transplanted prolactin-secreting tumor Login, Ivan S.; Nagy, Ivan; MacLeod, Robert M. Neuroendocrinology (1981), 33(2), 101-4

159.

Studies related to the synthesis of morphine. n. Conversion of 06-demethylsalutaridine into salutaridine Horvath, G.; Makleit, S. Acta Chim. Acad. Sci. Hung. (1981), 106(1), 37-42

160.

Short total synthesis of dihydrothebainone, dihydrocodeinone and nordihydrocodeinone Rice, Kenneth C. U. S. Pat. AppL, 15 pp. Avail. NTIS. Order No. PAT-APPL-6-165 690 PI US 165690 801219 AI US 80-165690 800703

161.

Approaches towards a practical synthesis of morphine Brossi, Arnold Proc. Asian Symp. Med. Plants Spices, 4th (1981), Meeting Date 1980, Volume 1, 261-74. Editor(s): Kusamran, Kosan; Pohmakotr, Manat; Reutrakul, Vichai. Publisher: Aksom Charoen - Tat Publ. House, Bangkok, Thailand.

162.

Paper electrophoresis for the determination of some synthetic morphine substitutes in chemicotoxicological analysis Mikhno, V. V. Farm. Zh. (Kiev) (1981), (2), 49-51

163.

Restoration of pituitary prolactin synthesis and release by the administration of morphine to rats bearing a transplanted prolactin-secreting tumor Login, Ivan S.; Nagy, Ivan; MacLeod, Robert M. Neuroendocrinology (1981), 33(2), 101-4

164.

Morphiceptin (NH4-Tyr-Pro-Phe-Pro-CONH2): a potent and specific agonist for morphine (.mu.) receptors Chang, Kwen-Jen; Killian, Anthony; Hazum, Eli; Cuatrecasas, Pedro; Chang, Jaw-Kang Science (Washington, D. C , 1883-) (1981), 212(4490), 75-7

136 165.

The synthesis and agonist activity of some 14.beta.-substituted morphine and codeine derivatives Osei-Gyimah, Peter; Archer, Sydney Endog. Exog. Opiate Agonists Antagonists, Proc. Int. Narc. Res. Club Conf. (1980), Meeting Date 1979, 13-16. Editor(s): Way, E. Leong. Publisher: Pergamon, Elmsford, N. Y.

166.

Morphine analgesia and newly synthesized 5-hydroxytryptamine in the dorsal and the ventral halves of the spinal cord of the rat Weil-Fugazza, J.; Godefroy, F.; Coudert, D.; Besson, J. M. Brain Res. (1981), 214(2), 440-4

167.

Stimulation of dopamine synthesis and release by morphine and D-Ala2-D-Leu5enkephalin in the mouse striatus in vivo Urwyler, Stephan; Tabakoff, Boris Life Sci. (1981), 28(20), 2277-86

168.

Chemicalstudiesondrugmetabolism.lv. Synthesis and analgesic activity of morphine-7,8-oxide and heroin-7,8-oxide Miyata, Naoki; Uba, Kiyoko; Watanabe, Keizo; Hirobe, Masaaki Chem. Pharm. Bull. (1980), 28(12), 3722-4

169.

Morphine-like activities of synthetic enkephalin analogs Moritoki, H.; Kiso, Y.; Kageyama, T.; Matsumoto, K. J. Pharm. Pharmacol. (1981), 33(1), 54-5

170.

The effects of melanocyte-stimulating hormone release inhibiting factor and synthetic analogs on tolerance to and physical dependence on morphine Bhargava, Hemendra N.; Ritzmann, Ronald F.; Walter, Roderich Endog. Exog. Opiate Agonists Antagonists, Proc. Int. Narc. Res. Club Conf. (1980), Meeting Date 1979, 545-8. Editor(s): Way, E. Leong. Publisher: Pergamon, Elmsford, N. Y.

171.

Comparison of the analgetic action of inhibitors of PG synthesis, Chinoin-127 and morphine Gyires, Klara; Knoll, J. Adv. Pharmacol. Res. Pract., Proc. Congr. Hung. Pharmacol. Soc, 3rd (1980), Meeting Date 1979, Volume 5, Issue Opiate Recept. Neurochem. Correl. Pain, 227-33. Editor(s): Furst, Susanna. PubUsher: Pergamon, Oxford, Engl.

172.

Total synthesis of (.+-.)-3-deoxy-7,8-dihydromorphine. Preliminary communication Hsu, Fu Lian; Rice, Kenner C ; Brossi, Arnold Helv. Chim. Acta (1980), 63(7), 2042-5

173.

Studies aiming at the synthesis of morphine. n. Studies on phenoUc couphng of N-norreticuhne derivatives Szantay, Csaba; Blasko, Gabor; Barczai-Beke, Marietta; Pechy, Peter; Domyei, Gabor Tetrahedron Lett. (1980), 21(36), 3509-12

137 174.

Morphine-like peptides: their regulation in the neuroendocrine system and the effect of guanyl nucleotides and divalent ions on opiate receptor binding Simantov, R. Prog. Biochem. Pharmacol. (1980), 16(Endog. Pept. Cent. Acting Drugs), 22-31

175.

Effects of endogenous opioid peptides and their synthetic analogs on anterior pituitary function Halasz, B.; Marton, J.; Koves, K.; Nagy, G.; Vizi, E.; Bajusz, S.; Molnar, J.; Lukats, O. Recent Results Pept. Horm. Androg. Steroid Res., Proc. Congr. Hung. Soc. Endocrinol. Metab., 9th (1979), 67-73. Editor(s): Laszlo, F. A. PubUsher: Akad. Kiado, Budapest, Hung.

176.

Studies on the synthesis of morphine, n. Conversion of 06demethylsalutarichne into salutaridme Horvath, Geza; Makleit, Sandor Magy. Kem. Foly. (1980), 86(6), 260-3

177.

Studies aiming at the synthesis of morphine. I. Separation and characterization of the amide rotamers of 6'-halogeno-N-formylnorreticulines Szantay, Csaba; Blasko, Gabor; Barczai-Beke, Marietta; Domyei, Gabor; Radics, Lajos Heterocycles (1980), 14(8), 1127-30

178.

Synthetic and pharmacological studies with enkephalin analogs in relation to structural features of morphine DiMaio, J.; Schiller, P. W.; Belleau, B. Pept., Struct. Biol. Funct., Proc. Am. Pept. Symp., 6th (1979), 889-92. Editor(s): Gross, Erhard; Meienhofer, Johannes. PubUsher: Pierce Chem. Co., Rockford, 111.

179.

Apphcation of metalated enamines to alkaloid synthesis. An expedient approach to the synthesis of morphine based analgesics Evans, D. A.; Mitch, C. H.; Thomas, R. C ; Zimmerman, D. M.; Robey, R. L. J. Am. Chem. Soc. (1980), 102(18), 5955-6

180.

Some remarks on the development of synthetic analgesics with morphine-like effect Halbach, H. Wien. Z. Suchtforsch. (1980), 3(3), 27-8

181.

Effect of acute administration of morphine on newly synthesized 5-hydroxytryptamine in spinal cord of the rat Godefroy, Francoise; Weil-Fugazza, Jeanne; Coudert, Danielle; Besson, Jean Marie Brain Res. (1980), 199(2), 415-24

182.

Morphine differentially alters synthesis and turnover of dopamine in central neuronal systems Alper, R. H.; Demarest, K. T.; Moore, K. E. J. Neural Transm. (1980), 48(3), 157-65

183.

Influence of prostaglandin synthetase inhibitors on the analgesic activity of morphine in the rat

138 Poggioli, R.; Castelli, M.; Genedani, S.; Bertolini, A. Riv. FarmacoL Ter. (1980), 11(1), 11-14 184.

Synthesis of 10(S)-methylcodeine and 10(S)-methylmorphine Arzeno, Humberto B.; Barton, Derek H. R.; Davies, Stephen G.; Lusinchi, Xavier; Meunier, Bernard; Pascard, Claudine Nouv. J. Chim. (1980), 4(6), 369-75

185.

Modulation of ganglioside synthesis by enkephalins, opiates, and prostaglandins. Role of cychc AMP in glycosylation Dawson, Glyn; McLawhon, Ronald W.; Schoon, Gwendolyn; Miller, R. J. ACS Symp. Ser. (1980), 128(Cell Surf. GlycoUpids), 359-72

186.

Comparative effects of synthetic enkephalinamides and morphine on abstinence responses in morphine-dependent mice Bhargava, Hemendra N. Pharmacol., Biochem. Behav. (1980), 12(5), 645-9

187.

Activities of hepatic carbamoyl phosphate synthetase and argininosuccinate synthetase after morphinization Wong, S. C ; Yeung, D. IRCS Med. Sci.: Libr. Compend. (1980), 8(3), 151

188.

Synthesis of 4a-ary 1-decahydroisoquinolines Rapoport, Henry; Weller, Dwight D.; Gless, Richard D. U. S. Pat. AppL, 46 pp. Avail. NTIS. US 31749 791221 US 79-31749 790420

189.

Benzomorphans - clinically used synthetic analogues of morphine. A short review for nonpharmacologists Palmer, David C ; Strauss, Michael J. Ind. Eng. Chem. Prod. Res. Dev. (1980), 19(2), 172-4

190.

Syntiiesisof morphine alkaloid analogs. Hasubanans and 9,17-secomorphinans Schwartz, Martin A.; Wallace, Rebecca A. Tetrahedron Lett. (1979), (35), 3257-60

191.

Synthesis and kinetic study of the hydrolysis of 4-cyano-4-phenylpiperidines: access to morphine analogs Gervais, Christian; Anker, Daniel; Chareire, Martine; Pacheco, Henri Bull. Soc. Chim. Fr. (1979), (5-6, Pt. 2), 241-8

192.

Synthesis of morphine analogs. Part 2 Substituted pyrrolo[3,4-h]isoquinoUnes Lindsay Smith, John R.; Norman, Richard O. C ; Rose, Malcolm E.; Curran, Adrian C. w.; Lewis, John W. J. Chem. Soc, Perkin Trans. 1 (1979), (11), 2868-72

193.

Syntiiesisof morphine analogs. Parti. Syntiiesis of some 5-benzyl-2-methyloctahydroisoquinoUnes Lindsay Smitii, John R.; Norman, Richard O. C ; Rose, Malcohn E.; Curran, Adrian C. W. J. Chem. Soc, Perkin Trans. 1 (1979), (11), 2863-7

139 194.

Chemical studies on drug metabolism. Part 1. Stereospecific synthesis of codeine 7,8-oxide and codeinone 7,8-oxide Uba, Kiyoko; Miyata, Naoki; Watanabe, Keizo; Hirobe, Masaaki Chem. Pharm. Bull (1979), 27(9), 2257-8

195.

Analgesic narcotic antagonists. 1. 8.beta.-Alkyl-, 8.beta.-acyl-, and 8.beta.(tertiary alcohol)dihydrocodeinones and-dihydromorphinones Kotick, Michael P.; Leland, David L.; Polazzi, Joseph O.; Schut, Robert N. J. Med. Chem. (1980), 23(2), 166-74

196.

Synthesis and analgesic activity of somel4.beta.-substituted analogs of morphine Osei-Gyimah, Peter; Archer, Sydney J. Med. Chem. (1980), 23(2), 162-6

197.

Oxidative phenol coupling in the synthesis of morphine alkaloids and related compounds Schwartz, Martin A. Symp. Pap. - lUPAC Int. Symp. Chem. Nat. Prod., 11th (1978), Volume 4, Issue Part 2, 274-9. Editor(s): Marekov, N.; Ognyanov, I.; Orahovats, A. Publisher: Izd. BAN, Sofia, Bulg.

198.

Synthesis of 2-aminomorphine and 2-aminocodeine. Reduction of aromatic nitro groups with formamidinesulfmic acid Chatterjie, Nithiananda; Minar, Arlene; Clarke, Donald D. Synth. Commun. (1979), 9(7), 647-57

199.

Synthesis, chromatography and tissue distribution of methyl-11C- morphine and methyl-11 C-heroin Kloster, G.; Roeder, E.; MachuUa, H. J. J. LabeUed Compd. Radiopharm. (1979), 16(3), 441-8

200.

Presynaptic effect of morphine and haloperidol on dopamine synthesis Shyu, Bai-Chuang; Chiang, Yi; Wang, Wei-Kung BuU. Inst. ZooL, Acad. Sin. (1978), 17(2), 125-30

201.

Dual action of methadone on 5-HT synthesis and metabohsm Ahtee, Liisa; Carlsson, Arvid Naunyn-Schmiedeberg's Arch. Pharmacol. (1979), 307(1), 51-6

202.

Physiologically active nitrogeneous organic compounds. Synthesis of morphinetype analgesics Kametani, Tetsuji; Fukumoto, Keiichiro Kagaku, Zokan (Kyoto) (1979), (79), 77-97

203.

Morphine-like analgesia by a new dipeptide, L-tyrosyl-L-arginine (kyotorphin) and its analog Takagi, Hiroshi; Shiomi, Hirohito; Ueda, Hiroshi; Amano, Hiro Eur. J. Pharmacol. (1979), 55(1), 109-11

204.

Influence of morphine, .beta.-endorphin and naloxone on the synthesis of phosphoinositides in the rat midbrain Natsuki, R.; Hitzemann, R. J.; Loh, H. H. Res. Commun. Chem. Pathol. Pharmacol. (1979), 24(2), 233-50

140 205.

The synthesis rate and turnover time of S-hydroxytryptamine in brains of rats treated chronically with morphine Bhargava, Hemendra N. Br. J. Pharmacol. (1979), 65(2), 311-17

206.

Opiate-receptor mediated changes in monoamine synthesis in rat brain Garcia-Sevilla, J. A.; Ahtee, Liisa; Magnusson, T.; Carlsson, A. J. Pharm. Pharmacol. (1978), 30(10), 613-21

207.

Effects of morphine, naloxone, naltrexone and .beta.-endorphin on monoamine synthesis in rat brain Ahtee, Liisa; Garcia-Sevilla, J. A.; Magnusson, T.; Carlsson, A. Dev. Neurosci. (Amsterdam) (1978), 4(Charact. Funct. Opioids), 345-6

208.

The effects of synthetic and natural opioid peptides in isolated organs Ronai, A. Z.; Berzetei, L; Szekely, J. I.; Bajusz, S. Dev. Neurosci. (Amsterdam) (1978), 4(Charact. Funct. Opioids), 493-4

209.

In vivo antagonism by naloxone of morphine, .beta.-endorphin and a synthetic enkephalin analog Szekely, J. I.; Dunai-Kovacs, Zsuzs; Miglecz, Erzsebet; Ronai, A. Z.; Bajusz, S. J. Pharmacol. Exp. Ther. (1978), 207(3), 878-83

210.

The effect of morphine tolerance and dependence on cell free protein synthesis Craves, Frederick B.; Loh, Horace H.; Meyerhoff, James L. J. Neurochem. (1978), 31(5), 1309-16

211.

Pentapeptide with morphine-like activity Sarantakis, Dimitrios; Grant, Norman H. US 4128541 781205 US 76-754794 761227

212.

Effect of some prostaglandin synthesis inhibitors on the antinociceptive action of morphine in albino rats Srivastava, D. N.; Bhattacharya, S. K.; Sanval, A. K. Clin. Exp. Pharmacol. Physiol. (1978), 5(5), 503-9

213.

Polypeptides with morphine-like activity Sarantakis, Dimitrios US 4098781 780704 US 77-777181 770314

214.

Synthesis of opioid pentapeptides, Leu-enkephahn and Met-enkephalin by solution methods and their biological activities Wang, Kung-Tsung; Chen, Shui-Tien; Huang, Min-Che; Chang, Chuan-Chiung Proc. Natl. Sci. Counc, Repub. China (1978), 2(3), 232-7

215.

Morphine and related compounds - inducers of specific antibody synthesis(literature review) Kovalev, I. E. Khim.-Farm. Zh. (1978), 12(9), 3-14

216.

Polypeptides with morphine-like activity

141 Sarantakis, Dimitrios; Stein, Larry US 4097471 780627 US 77-800678 770526 217.

Improved synthesis of (+)- morphine, (+)-cocleine and (+)-heroine Rice, Kenner C ; lijima, Ikuo; Brossi, Arnold Symp. HeterocycL, [Pap.] (1977), 49. Editor(s): Kametani, Tetsuji. Pubhsher: Sendai Inst. Heterocycl. Chem., Sendai, Japan.

218.

Inhibition of protein synthesis and tolerance to opiates Delia Bella, D.; Frigeni, V. Factors Affecting Action Narc, [Proc. Meet.] (1978), Meeting Date 1976,403-12. Editor(s): Adler, Martin W.; Manara, Luciano; Samanin, Rosario. Publisher: Raven, New York, N. Y.

219.

Importance of in vitro measurements of adrenocortical steroid synthesis in assessing locus of morphine effects Essman, Walter B.; Rosenthal, Richard Factors Affecting Action Narc, [Proc. Meet.] (1978), Meeting Date 1976,125-31. Editor(s): Adler, Martin W.; Manara, Luciano; Samanin, Rosario. Publisher: Raven, New York, N. Y.

220.

Morphine and .beta.-endorphin on RNA synthesis Lee, Nancy M.; Loh, Horace H.; Li, Choh Hao Adv. Biochem. Psychopharmacol. (1978), 18(Endorphins), 278-88

221.

Alteration of brain chromatin and nuclear synthetic activity in morphine-tolerant rats Sprague, G. L.; Fong, D. W.; Castles, T. R. Res. Commun. Chem. Pathol. Pharmacol. (1978), 19(3), 553-6

222.

Pentapeptide with morphine-like activity Sarantakis, Dimitrios US 4075190 780221 US 77-777211 770314

223.

Studies on morphine and related compounds: synthesis of 2,5,9-trimethyl- and 2ethyl-5,9-dimethyl- 3,4:6,7-dibenzomorphans Sharma, Rajendra; Goyal, V. K.; Tyagi, R. P.; Joshi, B. C. J. Indian Chem. Soc. (1977), 54(7), 753-5

224.

Effects of inhibitors of catecholamine synthesisand catecholamine agonists on morphine- and hypoglycemia-induced release of growth hormone in the rat Bluet-Pajot, M. T.; Schaub, C. J. Endocrinol. (1978), 76(2), 365-6.

225.

Conversions of tosyl and mesyl derivatives of the morphine group, XX. Synthesis of mesyl esters Makleit, Sandor; Berenyi, Sandor; Bognar, Rezso; Elek, Sandor Acta Chim. Acad. Sci. Hung. (1977), 94(2), 161-3

226.

Studies in the (-i-)-morphinan series. 4. A markedly improved synthesis of {+)- morphine

142 lijima, Ikuo; Minamikawa, Junichi; Jacobson, Arthur E.; Jacobson, Arthur E.; Rice, Kenner C. J. Org. Chem. (1978), 43(7), 1462-3 227.

Effect of antiphlogistics, adrenaline, histone, morphine and other substances on prostaglandin synthesis by rabbit renal medulla sUces Bekemeier, H.; Giessler, A. J.; Hirschelmann, R. Congr. Hung. Pharmacol. Soc, [Proc] (1976), Volume Date 1974, 2(2, Symp. Prostaglandins), 183-91

228.

Synthesis and analgetic activity of 1,2,3,4,5,6-hexahydro1,6-methano-3-benzazocines Mazzocchi, Paul H.; Harrison, Aline M. J. Med. Chem. (1978), 21(2), 238-40

229.

.beta.-Endorphin: synthesis and morphine-like activity of analogs with D-amino acid residues in positions 1,2,4 and 5 Yamashiro, Donald; Tseng, Liang-Fu; Doneen, Byron A.; Loh, Horace H.; Li, Choh Hao Int. J. Pept. Protein Res. (1977), 10(2), 159-66

230.

Isolation, structure, synthesis and morphine -like activity of .beta.-endorphin from human pituitary glands Li, Choh Hao; Yamashiro, Donald; Chung, David; Doneen, Byron A.; Loh, Horace H.; Tseng, Liang-Fu Ann. N. Y. Acad. Sci. (1977), 297, 158-66

231.

A synthetic enkephalin analog with prolonged parenteral and oral analgesic activity Roemer, Dietmar; Buescher, Heinz H.; HiU, Ronald C ; Pless, Janos; Bauer, Wilfried; Cardinaux, Francis; Closse, Annemarie; Hauser, Daniel; Huguenin, Rene Nature (London) (1977), 268(5620), 547-9

232.

Synthesizing codeinone from thebaine Calvo, Fernando 4052402 771004 AI US 76-666663 760315

23 3.

New N-2-hydroxyalkyl-6,7-benzomorphan derivatives: Synthesis and preliminary pharmacology Rahtz, Dieter; Paschelke, Gert; Schroeder, Eberhard Eur. J. Med. Chem. - Chim. Ther. (1977), 12(3), 271-8

234.

Synthesis and antitussive activity of 3- azabicyclo[3.2.2]nonane derivatives Arya, V. P.; Kaul, C. L.; Grewal, R. S. Arzneim.-Forsch. (1977), 27(9), 1648-52

235.

Value of the vinyloxycarbonyl unit in hydroxyl protection: application to the synthesis of nalorphine Olofson, R. A.; Schnur, Rodney C. Tetrahedron Lett. (1977), (18), 1571-4

236.

Converting allyhc ethers to dienol ethers Rapoport, Henry; Barber, Randy B.

143 U. S. Pat. Appl., 17 pp. Avail. NTIS. US 724018 760916 US 76-724018 760916 237.

Influence of morphine on protein synthesis in synaptic plasma membranes of the rat brain Hitzemann, Robert J.; Loh, Horace H. Res. Commun. Chem. Pathol. Pharmacol. (1977), 17(1), 15-28

238.

A comparison of the effects of morphine and other psychoactive drugs on brain phosphohpid synthesis Natsuki, Reiko; Hitzemann, Robert; litzemann, Barbara; Loh, Horace Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 451-4. Editor(s): KosterUtz, H. W. PubUsher: North-Holland, Amsterdam, Neth.

239.

Differential effects of morphine on rates of translation in vivo in free and bound polysome compartments of rat brain regions Ramsey, James C ; Steele, William J. Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 435-8. Editor(s): Kosterlitz, H. W. Publisher: North-Holland, Amsterdam, Neth.

240.

Control of tolerance reversal by protein synthesis inhibitors DeUa Bella, Davide; Frigeni, Viviana; Sassi, Alessandro Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 353-60. Editor(s): Kosterhtz, H. W. Publisher: North-Holland, Amsterdam, Neth.

241.

Some morphine-like properties of a potent antinociceptive synthetic pentapeptide in relation to physical dependence in rodents Baxter, M. G.; Follenfant, R. L.; Miller, A. A.; Sethna, D. M. Br. J. Pharmacol. (1977), 59(3), 523P

242.

Alteration of RNA biosynthesis induced by chronic treatments of opiate Lee, N. M.; Loh, H. H. Opiates Endog. Opioid Pept., Proc. Int. Narc. Res. Club Meet. (1976), 423-6. Editor(s): Kosterhtz, H. W. PubUsher: North-Holland, Amsterdam, Neth.

243.

3H-Lysine accumulation into brain and plasma proteins in male rats treated with morphine, naloxone, or naloxone plus morphine in relation to controls Ford, D. H.; Rhines, R. K.; Joshi, M.; Cheslak, H.; Simmons, H.; Toth, J. Tissue Responses Addict. Drugs, [Proc. Workshop Sess. Int. Soc. Neuroendocrinol.] (1976), Meeting Date 1975, 489-507. Editor(s): Ford, Donald Herbert; Clouet, Doris H. PubUsher: Spectrum Publ., Inc., HolUswood, N. Y.

244.

The effects of morphine-like drugs and chlorpromazine on the synthesis of glyceroUpids by homogenates of rat Uver Sanderson, Robert F.; Dodds, Peter F.; Brindley, David N. Biochem. Soc. Trans. (1977), 5(1), 295-6

144 245.

The effect of morphine on the metabolic transport of 3H-lysine and incorporation into protein Levi, M. A.; Rhines, R. K.; Ford, D. H. Tissue Responses Addict. Drugs, [Proc. Workshop Sess. Int. Soc. Neuroendocrinol.] (1976), Meeting Date 1975, 471-88. Editor(s): Ford, Donald Herbert; Clouet, Doris H. Publisher: Spectrum PubL, Inc., Holliswood, N. Y.

246.

Effect of a potent synthetic opioid pentapeptide in some antinociceptive and behavioral tests in mice and rats Baxter, M. G.; Goff, D.; Miller, A. A.; Saunders, I. A. Br. J. Pharmacol. (1977), 59(3), 455P-456P

247.

Electrooxidation and biosynthesis of natural products Tobinaga, Seisho Kagaku To Seibutsu (1977), 15(4), 240-2

248.

Stereochemical studies on medicinal agents. 23. Synthesis and biological evaluation of 6-amino derivatives of naloxone and naltrexone Jiang, Jack B.; Hanson, Robert N.; Portoghese, Philip S.; Takemori, A. E. J. Med. Chem. (1977), 20(8), 1100-2

249.

Synthesis of 4a-aryldecahydroisoquinohnes. FunctionaUty in the carbocychc ring Weller, Dwight D.; Gless, Richard D.; Rapoport, Henry J. Org. Chem. (1977), 42(9), 1485-95

250.

Heteroatom directed photoarylation: synthesis of a tetracycUc morphine structural analog Schultz, Arthur G.; Lucci, Robert D. J. Chem. Soc, Chem. Commun. (1976), (22), 925

251.

N-Acyl-N-norsalutaridines Schwartz, Martin Alan U.S., 9 pp. US 4003903 770118 US 75-549491 750212

252.

Studies on the syntheses of heterocyclic compounds. Part DCLXIX. Studies on the syntheses of analgesics. XLI. Optical resolution of (.-i-.)-N-cyclopropylme3iyl-3-hydroxy-9azamorphinan Kametani, Tetsuji; Kigasawa, Kazuo; Hiiragi, Mineharu; Wagatsuma, Nagatoshi; Kusama, Osamu; Uryu, Tsuneo Chem. Pharm. Bull. (1976), 24(10), 2563-6

253.

Morphine effects on RNA synthesis in purified ohgodendroglial nuclei Stokes, K. B.; Lee, N. M. Proc. West. Pharmacol. Soc. (1976), 19, 48-54

254.

The inhibition of the effects of morphine by synthetic substance P

145 Stern, P.; Hukovic, S.; Radivojevic, M. Experientia (1976), 32(10), 1326-7 255.

Synthesis of morphine-cl5 and codeine-d8 Lawson, John A.; DeGraw, Joseph I.; Anbar, Michael J. Heterocycl. Chem. (1976), 13(3), 593-5

256.

Screening of new synthesized compounds for analgesic effect according to kiioll's method Nikolova, M.; Stefanova, D. Farmatsiya (Sofia) (1975), 25(4), 47-53

257.

Influence of morphine on protein synthesis in discrete subcellular fractions of the rat brain Hitzemann, Robert J.; Loh, Horace H. Res. Commun. Chem. Pathol. Pharmacol. (1976), 14(2), 237-48

258.

Response of division-synchronized protozoa to morphine and levorphanol McClean, Daniel K.; Zimmerman, Arthur M. Gen. Pharmacol. (1975), 6(2-3), 171-9

259.

Perinatal narcotic addiction in mice. Sensitization to morphine stimulation Shuster, L.; Webster, G. W.; Yu, G. Perinat. Addict., Proc. Conf. (1975), Meeting Date 1974, 277-91. Editor(s): Harbison, Raymond D. Pubhsher: Spectrum Publ., Inc., Holliswood, N. Y.

260.

Synthesis of new morphine derivatives Bognar, R.; Gaal, G.; Horvath, G.; Kerekes, P. Izv. Khim. (1975), 8(1), 194-202

261.

A synthetic peptide with morphine-like pharmacologic action Goldstein, Avram; Goldstein, Joshua S.; Cox, B. M. Life Sci. (1975), 17(11), 1643-54

262.

The preparation and synthetic utility of 0-substituted normethylmorphines Borowitz, Irving J.; Diakiw, Vladimir J. Heterocycl. Chem. (1975), 12(6), 1123-6

263.

Tolerance to morphine-induced calcium depletion in regional brain areas. Characterization with reserpine and protein syntiiesis inhibitors Ross, D. H. Br. J. Pharmacol. (1975), 55(3), 431-7

264.

Conversion of tosyl and mesyl derivatives of the morphine group. XV. Synthesis of pseudocodeine tosylate and the study of its nucleophilic reaction Makleit, Sandor; Somogyi, Gabor; Bognar, Rezso Magy. Kem. Foly. (1975), 81(11), 517-19

265.

Protein synthesis in mouse brain during development of acute morphine tolerance Nakajima, Tohru; Sasano, Hiroshi; Koida, Masao; Kaneto, Hiroshi Jpn. J. Pharmacol. (1975), 25(4), 367-74

266.

Influence of peptides on reduced response of rats to electric footshock after acute administration of morphine Gispen, Willem H.; Van Wimersma Greidanus, Tjeerd B.; Waters-Ezrin, Cheryl; Zimmermann, Emery; Krivoy, William A.; De Wied, David Eur. J. Pharmacol. (1975), 33(1), 99-105

267.

Mechanisms of appearance of morphine dependency about the synthesis of a morphine artificial antigen Aoki, Masatada; Kusama, Tadashi; Kuboyama, Noboru; Murakoshi, Yoshie Nippon Daigaku Yakugaku Kenkyu Hokoku (1974), 14, 8-14

268.

In vivo protein synthesis in morphinetolerant monkey brain Hollinger, Mannfred A.; KiUam, Keith F.; Deneau, Gerald A. Psychopharmacol. Commun. (1975), 1(3), 319-25

269.

Morphine. Synthesis of 2'-hydroxy-2,5-dimethyl3,4:6,7-dibenzomorphan. EI Pandey, R. K.; Pandey, P.; Joshi, B. C. Bull. Acad. Pol. Sci., Ser. Sci. Chim. (1975), 23(6), 469-72

270.

Synthesis of thebaine and oripavine from codeine and morphine Barber, Randy B.; Rapoport, Henry J. Med. Chem. (1975), 18(11), 1074-7

271.

Acute effects of heroin and morphine on newly synthesized serotonin in rat brain Perez-Cruet, Jorge; Nguyen Bich Thoa; Ng, Larry K. Y. Life Sci. (1975), 17(3), 349-62

272.

Synthesis of morphine-3-glucuronide Berrang, Bertold; Twine, Charles E.; Hennessee, G. L.; Carroll, F.I. Synth. Commun. (1975), 5(3), 231-6

273.

Syntheses of analgesics. DI. Synthesis of l,2,3,4,5,6-hexahydro-3-benzazocine derivatives. 2 Sawa, Yoichi; Kato, Takeshi; Morimoto, Akira; Tom, Masuda; Hori, Mikio; Fujimura, Hajime Yakugaku Zasshi (1975), 95(3), 261-8

274.

Elimination of phenolic hydroxyl groups in morphine alkaloids. Synthesis of 3-deoxydihydromorphine Bognar, Rezso; Gaal, Gyorgy; Kerekes, Peter; Horvath, Geza; Kovacs, Maria T. Magy. Kem. Foly. (1975), 81(2), 51-3

275.

Apomorphine and morphine stimulate prostaglandin biosynthesis ColHer, H. O. J.; McDonald-Gibson, Wendy J.; Saeed, S. A. Nature (London) (1974), 252(5478), 56-8

276.

Biogenetically patterned synthesis of the morphine alkaloids Schwartz, Martin A.; Mami, Ismail S. J. Am. Chem. Soc. (1975), 97(5), 1239-40

277.

Effect of morphine on the turnover and synthesis

147 3H-leucine-labeled protein and 14C-choline-iabeled phosphatidylcholine in discrete regions of the rat brain Loh, Horace H.; Hitzeman, Robert J. Biochem. Pharmacol. (1974), 23(12), 1753-65 278.

Synthetic compounds, pentazocine and Darvon, which have moiphine-like activity, and their effects on the pituitary-adrenocortical system. Morphine-like inhibitory effect on histamine ACTH-releasing action Chang, Wung Hua; Hiraga, Kogo Tokyo Jikeikai Dca Daigaku Zasshi (1973), 88(2), 314-21

279.

Qiolinergic mechanism of action of morphine and of some of its synthetic analogs Hadzovic, Jelena; Hadzovic, Safet Veterinaria (Sarajevo) (1973), 22(3), 347-51

280.

Synthesis of l,3,4,5,6,7,8,8a-octahydro-2-methyl-4a- phenylisoquinolin-6-ols. Novel fragments of the morphine molecule Finch, Neville; Blanchard, Louis; Puckett, R. T.; Werner, L. H. J. Org. Chem. (1974), 39(8), 1118-24

281.

Synthetic antigens Gross, Stanley J. Ger. Offen., 38 pp. PI DE 2324544 731129 PRAI US 72-253632 720515

282.

Morphine lethality in rats. Effects of inhibitors of brain catechol amine synthesis and methylation Davis, W. M.; Khalsa, J. H. Res. Commun. Chem. Pathol. Pharmacol. (1973), 6(3), 867-72

283.

Synthesis and pharmacological effect of morphine phosphate esters Mori, Masaaki; Oguri, Kazuta; Yoshimura, Hidetoshi; Kamata, Osamu Yakugaku Zasshi (1973), 93(10), 1302-7

284.

Late effects of perinatal morphine administration on pituitary-thyroidal and gonadal function Bakke, John L.; Lawrence, Nancy L.; Bennett, Jane Biol. Neonate (1973), 23(1-2), 59-77

285.

Thin-layer chromatography of drugs. V. Analysis of drugs which are synthetic substitutes of morphine Botev, B. Farmatsiya (Sofia) (1973), 23(3), 19-25

286.

Synthesisof B/C trans-fused morphine structures. VI. Mass spectrum, optical rotatory dispersion, and circular dichroism of B/C trans-morphine derivatives Inoue, Hirozumi; Takeda, Mikio; Kugita, Hiroshi Chem. Pharm. Bull. (1973), 21(9), 2004-13

287.

Acute effects of morphine on dopamine synthesis

148 and release and tyrosine metabolism in the rat striatum Gauchy, C ; Agid, Y.; Glowinski, J.; Cheramy, A. Eur. J. Pharmacol. (1973), 22(3), 311-19 288.

Effects of narcotic analgesic drugs on brain noradrenergic mechanisms Smith, Charles Bruce; Sheldon, M. I. Agonist Antagonist Actions Narcotic Analg. Drugs, Proc. Int. Symp. (1973), Meeting Date 1971, 164-75. Editor(s): KosterUtz, H. W. Publisher: Univ. Park Press, Baltimore, Md.

289.

Effect of chronic administration of morphine on mouse brain aminoacyl-tRNA synthetase and tRNA-amino acid binding Datta, Ranajit Kumar; Antopol, William Brain Res. (1973), 53(2), 373-86

290.

Action of levallorphan. Macromolecular synthesis and cell division Stephens, Ralph E.; Zimmerman, Arthur M. Mol. Pharmacol. (1973), 9(2), 163-71

291.

Isolation, x-ray analysis, and synthesis of a metabolite of (-)-3-hydroxy-N-allylmorphinan Blount, John P.; Mohacsi, Emo; Vane, Floie M. J. Med. Chem. (1973), 16(4), 352-5

292.

Fortral. Synthetic morphine derivative with morphine-antagonistic components Kubicki, St.; Haas, J.; Stoelzel, R. Verh. Deut. Ges. Inn. Med. (1972), 78, 1577-82

293.

Effects of inhibitors of protein synthesis in morphine tolerance and dependence Cox, Brian Martin Agonist Antagonist Actions Narcotic Analg. Drugs, Proc. Int. Symp. (1973), Meeting Date 1971, 219-31. Editor(s): Kosterlitz, H. W. PubUsher: Univ. Park Press, Baltimore, Md.

294.

Pharmacologic properties of synthetic.DELTA.9 tetrahydrocannabinol (THC) Kaymakcalan, Sulaii; Deneau, G. A. Acta Med. Turc, Suppl. (1972), No. 1, 27 pp.

295.

Structure related to morphine. Synthesis of.alpha.-2-N-heptyl-2'-hydroxy-5,9dimethyl-6,7-benzomorphan from 3,4-lutidine. II Ramachandran, Mrs. R.; Kishore, D.; Joshi, B. C. Def. Sci. J. (1972), 22(3), 201-3

296.

Brain chromatin activity of morphine-treated rats Hodgson, John R.; Lee, Cheng-Chun; Castles, Thomas R. Proc. Soc. Exp. Biol. Med. (1972), 141(3), 790-3

297.

[14C]-catechol amine synthesis in mouse brain during morphine withdrawal Rosenman, S. J.; Smith, C. B. Nature (London) (1972), 240(5377), 153-5

298.

Inhibition of development of tolerance to morphine by cycloheximide Feinberg, Michael P.; Cochin, Joseph

149 Biochem. Pharmacol. (1972), 21(22), 3082-5 299.

Inhibitory effects of chronic administration of morphine on uridine and thymidine incorporating abilities of mouse liver and brain subcellular fractions Datta, R. K.; Antopol, W. Toxicol. Appl. Pharmacol. (1972), 23(1), 75-81

300.

Incorporation of amino acids into proteins of synaptosomal membranes during morphine treatment Franklin, G. I.; Cox, B. M. J. Neurochem. (1972), 19(7), 1821-3

301.

Improvement in the technology of codeine synthesisfrom morphine Smimov, D. M.; Sigal, E. L.; Marechek, K. Ya.; Zakharov, V. P. Khim.-Farm. Zh. (1972), 6(5), 31-6

302.

Chemical synthesis and analgesic effect of morphine ethereal sulfates Mori, Masaaki; Oguri, Kazuta; Yoshimura, Hidetoshi; Shimomura, Kyoichi; Kamata, Osamu; Ueki, Showa Life Sci. (1972), ll(ll)(Pt. 1), 525-33

303.

Nucleic acid synthesis in brains from rats tolerant to morphine analgesia Castles, T. R.; Campbell, S.; Gouge, R.; Lee, C. C. J. Pharmacol. Exp. Ther. (1972), 181(3), 399-406

304.

Effect of morphine on cerebral glycogen content, glycogen synthetase, and incorporation of glucose into brain glycogen of mice Estler, C. J.; Mitznegg, P. Pharmacol. Res. Commun. (1971), 3(4), 363-7

305.

Dissociation of morphine tolerance and dependence from brain serotonin synthesis rate in mice Schechter, Paul J.; Lovenberg, Walter; Sjoerdsma, Albert Biochem. Pharmacol. (1972), 21(5), 751-3

306.

Effect of levorphanol tartrate on ribonucleic acid synthesis in normal and regenerating rat liver Becker, Frederick F.; Rossman, Toby; Reiss, Betti; Simon, Eric J. Res. Commun. Chem. Patiiol. Pharmacol. (1972), 3(1), 105-16

307.

Correlations between protein and serotonin synthesis during various activities of the central nervous system (slow and desynchronized sleep, learning and memory, sexual activity morphine tolerance, aggressiveness, and pharmacological action of sodium gamma-hydroxybutyrate) Laborit, H. Res. Commun. Chem. Patiiol. Pharmacol. (1972), 3(1), 51-81

308.

Minor alkaloids of morphine. VII. Synthesis of gnoscopine (dl-narcotine) Kerekes, Peter; Bognar, Rezso Magy. Kern. Foly. (1971), 77(12), 655-9

150 309.

Alkaloids associated with morphine. Vn. Synthesis of gnoscopines (DLnarcotine) Kerekes, P.; Bognar, R. J. Prakt. Chem. (1971), 313(5), 923-8

310.

Effect of morphine on protein synthesis in synaptosomes and mitochondria of mouse brain in vivo Kuschinsky, K. Naunyn-Schmiedebergs Arch. Pharmakol. (1971), 271(3), 294-300

311.

Amounts and turnover rates of brain proteins in morphinetolerant mice Hahn, D. L.; Goldstein, A. J. Neurochem. (1971), 18(10), 1887-93

312.

Biochemical pharmacology of tolerance to opioid analgesics Ginsburg, M. Sci. Basis Med. (1971) 305-19

313.

Morphine-associated alkaloids. 5. Synthesis and structure of narcotoline ethers Gaal, Gy.; Kerekes, P.; Gorecki, P.; Bognar, R. Pharmazie (1971), 26(7), 431-4

314.

Effect of p-chlorophenylalanine on the cardiorespiratory reflex response to morphine and serotonin in the rat Aldunate, Jorge; Prieto, Rafael Arch. Biol. Med. Exp. (1970), 7(1-2-3), 45-7

315.

Increase of brain tryptophan caused by drugs which stimulate serotonin synthesis Taghamonte, Alessandro; Tagliamonte, Paola; Perez-Cruet, Jorge; Gessa, Gian L. Nature (London), New Biol. (1971), 229(4), 125-6

316.

Correlations between protein synthesis and serotonin in various central nervous system activities. Slow and desynchronized sleep, memory training, sexual activity, morphine tolerance, aggressiveness, and sodium .gamma.hydroxybutyrate pharmacology Laborit, H. Agressologie (1971), 12(1), 9-24

317.

Structures related to morphine. Synthesisof .alpha.-2'-hydroxy-2-methyl-5-propyl-9-ethyl-6,7-benzomorphan. I Ramachandran, Reena; Joshi, Bhuwan C. Def. Sci. J. (1970), 20(4), 233-6

318.

Conversions of tosyl and mesyl derivatives of the morphine group. VIII. Synthesis and investigation of 6-deoxy-6-fluoroisocodeine Bognar, Rezso; Makleit, Sandor; Radics, Lajos Acta Chim. (Budapest) (1971), 67(1), 63-9

319.

Morphine alkaloids and related compounds. XX. Syntheses and pharmacology of some demethylated compounds related to 14-hydroxydihydro-6.beta.-thebainol4methyl ether (oxymethebanol), a new potent antitussive Seki, Isao; Takagi, Hiromu

151 Chem. Pharm. Bull. (1971), 19(1), 1-5 320.

Unchanged rate of brain serotonin synthesis during chronic morphine treatment and failure of p-chlorophenylalanine to attenuate withdrawal syndrome in mice Marshall, Ian G.; Grahame-Smith, D. G. Nature (London) (1970), 228(5277), 1206-8

321.

Tolerance to morphine-induced increases in [14C]-catechol amine synthesis in mouse brain Smith, Charles Bruce; Villarreal, Juhan E.; Bednarczyk, Janet H.; Sheldon, Murray I. Science (1970), 170(3962), 1106-8

322.

Is there a relation between protein synthesis and tolerance to analgesic drugs? Cox, Brian M.; Ginsburg, M. Sci. Basis Drug Depend., Symp. (1969), Meeting Date 1968, Volume Development of new p 77-92. Editor(s): Steinberg, Hannah. PubUsher: J. and A. Churchill Ltd., London, Engl.

323.

New neurotropic agents among synthesized compounds of the pyridine series Poddubnaya, L. V.; Olekhnovich, L. B.; Dorofeenko, G. N. Farmakol. Tsent. Khohnolitikov Drugikh Neirotropnykh Sredstv (1969), 317-24. Editor(s): Denisenko, P. P. Publisher: Leningrad. Sanit.-Gig. Med. Inst., Leningrad, USSR.

324.

Effects of morphine and pentobarbitone on acetylcholine synthesis by rat cerebral cortex Sharkawi, Mahmoud Brit. J. Pharmacol. (1970), 40(1), 86-91

325.

Synthesis of B/C transfused morphine structures. FV. Synthesis of B/C transisomorphine Inoue, Hirosumi; Takeda, Mikio; Kugita, Hiroshi Chem. Pharm. Bull. (1970), 18(8), 1569-75

326.

Synthesis of B/C trans-fused morphine structures. V. Pharmacological summary of trans-morphine derivatives and an improved synthesis of trans-codeine Kugita, Hiroshi; Takeda, Mikio; Inoue, Hirosumi J. Med. Chem. (1970), 13(5), 973-5

327.

Morphine alkaloids and its related compounds. XVUI. Syntiieses of N-substituted-morphinan dihydronormethines and 0-alkyhsoureas related to morphinan, norpethidine, or phenethylamine Seki, Isao; Takagi, Hiromu Chem. Pharm. Bull. (1970), 18(6), 1104-11

328.

Conversions of tosyl and mesyl derivatives of the morphine group. VI. Synthesis of acetylthio and mercapto derivatives Bognar, Rezso; Makleit, Sandor; Mile, Terez; Radics, Lajos Acta Chim. (Budapest) (1970), 64(3), 273-9

152 329.

Effect of morphine, nalorphine, naloxone, pentazocine, cyclazocine, and oxotremorine on the synthesis and release of acetylchohne by mouse cerebral cortex sUces in vitro Howes, John F.; Harris, Louis Selig; Dewey, William L. Arch. Int. Pharmacodyn. Ther. (1970), 184(2), 267-76

330.

Inhibition of morphine tolerance and physical dependence development and brain serotonin synfiiesis by cycloheximide Loh, Horace H.; Shen, Fu-Hsiung; Way, E. Leong Biochem. Pharmacol. (1969), 18(12), 2711-21

331.

Inhibition of the development of tolerance to morphine in rats by drugs which inhibit ribonucleic acid and protein synthesis Cox, Brian Martyn; Osman, O. H. Brit. J. Pharmacol. (1970), 38(1), 157-70

332.

Lack of a direct effect of morphine on the synthesis of pineal carbon-14 labeled indoles in organ culture Shein, Harvey M.; Larin, Frances; Wurtman, Richard J. Life Sci. (1970), 9(1), 29-33

333.

Conversion of tosyl and mesyl derivatives of the morphine group. V. of isocodeine and dihydroisocodeine Makleit, Sandor; Bognar, Rezso Magy. Kem. Foly. (1969), 75(5), 235

334.

Morphine alkaloids and its related compounds. XVI. Synthesis of 14hydroxyallopseudocodeine 8-ethers and its derivatives Seki, Isao; Takagi, Hiromu Chem. Pharm. Bull. (1969), 17(8), 1555-9

335.

Morphine derivatives, n. Stereochemistry of the by-products in the synthesis of 3-methoxy-N- methyhnorphinan Kawasaki, Kazuhiko; Matsumura, Hiromu Chem. Pharm. Bull. (1969), 17(6), 1158-74

336.

Metabohsm of drugs. LX. Synthesis of codeine and morphine glucuronides Yoshimura, Hidetoshi; Oguri, Kazuta; Tsukamoto, Hisao Chem. Pharm. Bull. (Tokyo) (1968), 16(11), 2114-19

337.

Synthesisof B/C trans-fused morphine structures. III. Synthesis of B/C trans-morphine Kugita, Hiroshi; Takeda, Mikio; Inoue, Hirosumi Tetrahedron (1969), 25(9), 1851-62

338.

Synthesis of B/C trans-fused morphine structures. II. Hydroboration of isoneopine, neopine, neopinone and thebaine Takeda, Mikio; Inoue, Hirosumi; Kugita, Hiroshi Tetrahedron (1969), 25(9), 1939-49

339.

Comparative action of morphine and a synthetic substance on behavior and pain in the baboon. Tolerance study Baetz, Pierre; Bourgoin, P.; Giono, Paulette; Giono, H.

Synthesis

153 Bull. Mem. Fac. Mixte Med. Pharm. Dakar (1967), 15, 270-4 340.

Conversions of tosyl and mesyl derivatives of the morphine series. Synthesis of acetylthio and mercapto derivatives Bognar, Rezso; Makleit, Sandor; Mile, Terez Acta Chim. (Budapest) (1969), 59(1), 161-4

341.

Synthesis of new morphine derivatives. 11. The preparation 0-benzoylmorphines with analgesic action and an O-benzylmorphine with a morphine-potentiating effect Selmeci, Gyorgy; Szlavik, Laszlo; Kaskoto, Zoltan; Lepenyene, Jilek Maria; Tothne, Aranyos Iren Khim.-Farm. Zh. (1968), 2(7), 19-23

342.

Morphine tolerance, physical dependence, and synthesis of brain 5-hydroxytryptamine Way, E. Leong; Loh, Horace H.; Shen, Fu-Hsiung Science (1968), 162(3859), 1290-2

343.

Elimination of the 4-hydroxyl group of the alkaloids related to morphine. XI. Syntiiesis of (-)-14-hydroxy-3metiioxy-N-methylmorphinan derivatives Sawa, Y. K.; Tada, H. Shionogi Res. Lab., Shionogi and Co., Ltd., Osaka, Japan Tetrahedron (1968), 24(20), 6185-96

344.

Synthesis of new morphine derivatives. I. Morphine derivatives substituted at the nitrogen and in position 3 Selmeci, G.; Szlavik, L.; Kaskoto, Z.; Jilek, L. M.; Maczko, I. Khim.-Farm. Zh. (1968), 2(6), 12-17

345.

Effect of morphine on acetylcholine release from rabbit brain tissue Sugano, Tsukasa; Takeno, Kazu; Yanagiya, Iwao Nippon Yakurigaku Zasshi (1967), 63(6), 494-500

346.

The synthesis of codeine and morphine D-glucuronides Yoshimura, H.; Oguri, K.; Tsukamoto, H. Tetrahedron Lett. (1968), (4), 483-6

347.

Synthesis and biological properties of l-dimethyl-amino-3-methyl-3-(3hydroxyphenyl)butane, a potential analgetic Pecherer, Benjamin; Sunbury, R. C ; Randall, Lowell O.; Brossi, Arnold J. Med. Chem. (1968), 11(2), 340-2

348.

Effect of morphine administration on the incorporation of leucine-14C into protein in cell-free systems from rat brain and hver Clouet, Doris H.; Ratner, Milton J. Neurochem. (1968), 15(1), 17-23

349.

Thin-layer-chromatographic distribution of opium alkaloids and some partially synthetic analogs Paris, R. R.; Sarsunova, Magda Pharmazie (1967), 22(9), 483-4

154 350.

Elimination of the 4-hydroxyl group of the alkaloids related to morphine. IX. Synthesis of 3-methoxy-N- methylisomorphin and derivatives Sawa, Yoshiro K.; Horiuchi, Masahiko; Tanaka, Katsura Tetrahedron (1968), 24(1), 255-60

351.

Alternate Route in the Synthesis of Morphine Morrison, Glenn Curtis; Waite, Ronald, O.; Shavel, John, Jr. Tetrahedron Lett. (1967), (41), 4055-6

352.

Synthesis of B/C-trans-morphine Kugita, Hiroshi; Takeda, Mikio; Inoue, Hirosumi Tetrahedron Lett. (1967), (14), 1277-81

353.

Biochemistry of poppy alkaloid synthesis. Phenolase complex in Papaver somniferum Kovacs, Peter; Jindra, Antonin; Psenak, Mikulas Abh. Dtsch. Akad. Wiss. BerHn, Kl. Chem., Geol. Biol. (1966), (3), 335-40

354.

Preparation of formaldehyde by photochemical condensation of carbon monoxide and tritiated hydrogen, and the synthesis of a tritiated morphine derivative Lane, A. C ; McCoubrey, Arthur; Peaker, R. J. Labelled Compd. (1966), 2(3), 284-8

355.

Analysis of medicine mixtures. VI. Microchemical detection of some alkaloids and synthetics with potassium-iron or potassium-copper-chloroiodide reagents Reisch, Johannes; Tittel, G. L.; Perlick, J. Dtsch. Apoth.-Ztg. (1965), 105, 575-6 From: CZ 1966, (22), Abstr. No. 1652


155

New Developments in the Synthesis of Polyketides and of Chiral Methyl Groups

Johann Mulzer Institut ftir Organische Chemie der Freien Universitat TakustraBe 3, D-14195, Berlin, Germany

Abstract: This review deals with recent advances in the synthesis of polypropionate structures. It focuses on the total synthesis of natural products (citreoviral, ACRL toxin IIIB) as well as on new synthetic methodology (chiral methyl branching, base induced 1,3-Hshift and chiral methyl groups).

Chiral Methyl Branching in Carbon Chains In the biosynthesis of polyketides the problem of chiral methyl branching is solved via enantiomer selective reduction of prostereogenic carbonyl groups (Scheme 1). The biosynthesis starts with the Claisen type condensation of the activated propionate 1 and methyl malonyl CoA 2 to give under elimination of carbon dioxide ^-ketoester 3 which has already a chiral methyl branching in 2-position. However, this center normally tends to racemization and is not configurationally stable. It may be safely assumed that it is the NADH mediated reduction of 3 to 4, which eventually defines the configurations of both the hydroxy and the methyl bearing stereogenic centers, the first one by direct chiral induction, the second one by kinetic resolution of a mobile equilibrium of the enantiomers of keto ester 3. The same process is then repeated for each new propionate subunit in the growing polyketide chain, e.g. from 4 via 5 to 6.

156

In this way hydroxymethyl (HM) or hydroxy-methyl-hydroxy-methyl (HMHM)subunits are generated in a stereodefined way (1).

CH3 CH3-CH2-C~SCoA

0

e.DoC

--C02

SCoA

CH,

OH

Reductase

O

II * // C113"" C112'"'C>"~ C H "" C

SCoA

+2

CH3-CH2-CH-CH-C' CH3 SCoA

-CO2

4 (stereodefined) OH 0 ^ 0 o ^ * I * II *2 /y Reductase GH3-CH2-CH-CH-C-CH-C ^ *

i

?

CH3 5

i

CH3

\

OH , OH 2 0 1 *4 1 *2 ^^ CH3-CH2-CH-CH~CH-CH-C *

SCoA

«

^

i

*

?

i

\

CH3 CH3 SCoA 6 (stereodefined)

Scheme 1: Polypropionate Biosynthesis

How far can this bioprocess be transferred to by in-vitro synthesis? An answer to this question has been given by R. W. Hoffmann et al. (2), who reduced ketoester 7, configurationally labile as discussed, with baker's yeast, hi fact the carbonyl reduction proceeded (5)-selectively, however, the kinetic resolution with respect to the enantiomers of 7 only led to an enantiomeric excess of 72% with respect to C-2 in 8/9.

Me-

OEt Me 7

baker's yeast (59%)

OH

0

Me-^ V ^

OH OEt

Me^ > ^

Me 8

0

Me 6.4: 1

9

"OEt

157

Apparently this method works unsatisfactorily in vitro so that purely chemical approaches appear advisable to generate HM- and HMHM-units in an enantiopure manner. As illustrated in Scheme 2, two possibilities (among others) may be envisaged to place a chiral methyl branching upon a pre-existing carbon chain: either by opening of a configurationally defined epoxide with a methyl cuprate reagent (equ. 1) or by alkylating a chirally substituted propionate type enolate (equ. 2). Alternatively, a suitably functionalized carbon chain may be methylated via the corresponding enolate.

Epoxide Opening:

f j y\yy^

1 j

0 CH3

•

M® i l j y"^^^^

l \

equ. 1

OH Enolate Alkylation of a Chiral Propionic Acid Derivative:

0

%

o Me

Scheme 2: How to Generate Chiral Methyl Branching in a Carbon Chain to Form an 1,2Hydroxy-Methyl (1,2-HM) Subunit

We describe some new methodology for both realizing equ. 1 and 2. Regarding the epoxide opening we reasoned that the v/c diol present in inexpensive carbohydrates such as Z)-mannitol would be highly suitable for the construction of HM- and HMHM-subunits. Specifically (Scheme 3) the D-mannitol diacetonide (10) gives the c/5'-epoxide 12 on treatment of tosylate 11 with base. With methyl cuprate regioselective ring opening occurs in favor of the HM-intermediate 13, which can easily be separated from the diastereomer 14. This opens a simple access to 5'j^/7-HM-structures.

158

Q K2CO3

Me2CuLi

6: 1

Scheme 3: syn-HM-Subunits from Epoxides

The a^2//-diastereomers 19 and 23 are available via the inversion of the 3- or 4-OHfunctions of 10 (Scheme 4). This is achieved via monoprotection of 10 to benzoate 15, oxidation to ketone 16, stereoselective reduction to 17 and formation of epoxide 18, which is C2-symmetrical and thus can only give the anti-UM derivative 19 on cuprate addition. Alternatively 17 is converted into 20 which gives the C2-symmetrical epoxide 22 via 21. Again only one HM-derivative, namely 23, can be formed on reaction with the dimethyl cuprate (3).

159

10

\. 2. OH-

o"

y^

^^^

^^

Me2CuLi 0

18 (C2-symmetricaI)

TBDPS-CI 17

22 (C2-symmetrJcal) Scheme 4: aA7//-HM-Subunits from Epoxides

HMHM-Subunits may be generated via bis-epoxides as demonstrated in Scheme 5. D-Mannitol is converted into the di-tosylate 24 which is cycHzed to 25 with base.

160

OTs OH

3 steps

D-Mannitol

f

•

:

HO""'^

TBDPS-CI

"^

p.^_^ K2CO3

^^^

.OTBDPS

TBDPSO'

"6 26 Me

TBDPso

i^r

Me OTBDPS

TBDPSO

3^^5^

OH 27 Scheme 5: HMHM-Subunits from Epoxides

Bis-protection of the terminal OH-functions of 25 furnished 26, to which Lipshutz' cuprate is added. The bulky silyl group directs the nucleophilic attack to the 3position to form 27, whereas smaller protective groups such as MOM lead to mixtures of 2,3-regioisomers. Monoepoxide 27 is not isolated but adds excess cuprate insitu to furnish 28 regioselectively. Once introduced the 3-methyl branching in 27 directs the cuprate attack towards the less hindered position at C-5.

OTs OH

TsO OH

K2CO3

OTs ^

'

29

2 Me2Cu(CN)Li2

Scheme 6: HMHM-Subunits from Epoxides

—

0

161

An alternative way to HMHM-structures such as 32 is shown in Scheme 6. The C2symmetrical ditosylate 24 is converted into the mono-acetonide 29 which forms the epoxide 30 on treatment with base. The cuprate attack on the epoxide generates a second epoxide 31 via tosylate ehmination. With excess cuprate 31 is opened to give 32 directly (4). Alkylation of Chiral Enolates As demonstrated in Scheme 7 the alkylation of chiral propionamide enolates has become a standard operation, since D. A. Evans introduced his unusually efficient oxazolidinone auxiliaries ®.

(1)

D. A. Evans O

SI

J

iPr

0

0

H

re

3 ) W. Oppolzer

1985

O

•••• Me"^ . . MPh

(4)' J. Rebekjr.

1984

T. Katsuki

0

^^

CH2OMOM

CH2OMOM

ds^ 100: 1

ds^ 90: 10

cfs « 95 : 5

O2

(2)

1981

1990

5)

K. Kimura

0

ds ^ 96 : 4

ds > 99 : 1 (one example!) ds > 500 : 1

Scheme 7: Alkylation of Chiral Propionamide Enolates

1992

162

This principle has found many variations and improvements, some of which are shown as (2) - (D in Scheme 7 (5).

0

NH2

Ph,^.OH

.L 1NHo

NaBH4

Cr U

33

1.

34

H ^ fBu

OTMS 2. HoC=C NTMS H

p

3: 1 Chromatographic Separation

35

3. E t - C ' CI (50%)

0

NHBoG

H

J^

MgBr

1. H + O

^^^

X

2. H ^ ^ f B u OTMS 3. H2C=C NTMS H 4.

Et-C CI (50%)

0 N'^fBu

r "0

'0 35a


Scheme 8: Chiral Dihydroxooxazines as Amide Auxiliaries

36a

163

In all cases except (2) the nitrogen is part of an imide system which forms a chelate complex after deprotonation to the enolate. We used a different approach by using the A^,(9-acetals 35/36. This was first performed in racemic form to test the diastereoselectivity of the enolate alkylation (Scheme 8). The diastereomers 35 and 36 are readily separable by chromatography. Analogously, amides 35a and 36a are prepared from 0-jV-Boc-amino benzaldehyde. Derivative 36a is crystalline and was submitted to an X-ray crystal structure analysis (Fig. 1).

\j.^'

V\tX

Figure 1:

^^^

Crystal structure of 36a

The amide function (Nl-Cl 1-0-11) is planar and exerts an allylic 1,3-strain effect on the adjacent stereogenic center (C-2), which forces the bulky /-butyl substituent into a pseudo-axial position with respect to the boat conformation of the acetal ring. The isopropenyl group at C-4 adopts another pseudo-axial position. Despite its axial rearrangement, the shielding effect of the /-butyl group on the enolate carbon C12 is low. Indeed, deprotonation of 35 and subsequent allylation proceeded with a stereoselectivity of 4:1 in 89% yield. If the phenyl and /-butyl group are on opposite ring faces as in 36, the deprotonation is severely hindered . Both yield and stereoselectivity of the allylation are low.

164

35

INaHDMS

(89%)

36

1. IDA, DMPU^

(30%)

We reasoned that the /-butyl group is still too small for an efficient chiral induction; therefore, the optically pure 0-TBDPS lactaldehyde was chosen for the formation of the A^,0-acetal. But only amides 39 and 41 are now non-racemic; 39 exhibits a satisfactory chiral induction on allylation, because the enolate carbon is shielded by the adjacent axial bulky substituent. In 41, both sidechains at C-2 and C-4 are equatorial and the stereocontrol drops significantly.The use of a chiral aldehyde for acetal formation even allows the use of the achiral 0-aminobenzylalcohol (43) as a template. Acetals 44 and 45 are formed and separated; due to the allylic 1,3-strain of the amide moiety both derivatives have axial sidechains (as detectable in the crystal structure of alkylation product 46d) (Fig. 2) and the chiral induction is similarly high in both cases. The chiral auxiliary is removed with lithium aluminium hydride without any racemization of the newly created stereocenter (6).

165 Ph

Ph 1. KHDMS

I

^0

Br

2.

N^2

OTBDPS

"^^^^'""Ô

39

^^^^^^

40 (cfs> 95:5)

E>h

Ph 1. KHDMS ,Br

OTBDPS

^

^

V

'0

OTBDPS

42 (ds = 80:20)

41

?

O ^ ^

.>o Gi

a

51^

Figure 2: Crystal Structure of 46d

IDi

166

, „V '"• OTBDPS

CC°"

OTMS

NHo

2. HoC=C

43 3.

Et-C

NHTMS

CI

(62%)

x^

N

OTBDPS

'0

0


44

45 R-X

R-X

46,47

x^

R

R,, ^ x ^ a b c d

ally! Et nPr -v-/^^^

OTBDPS

0

92:8 - 94:6

OTBDPS ds

ds

46

OTBDPS

47

88:12-92:8

Scheme 9: Achiral Dihydrooxazine Template

1,3-H-Shift. By serendipity we found a novel base induced stereocontrolled sigmatropic 1,3-Hmigration. Benzylation of the readily available alcohol 48 to benzyl ether 49 occurs under standard conditions at 25°C. By mere accident the student (G. Funk) raised the temperature of the mixture to 80 °C and left the reaction at that temperature for 14 h.

167

After the usual workup the rearrangement product 50 was isolated in quantitative yield as the pure (>98% ) ^-olefin (7).

OBn

BnCI / NaH DMF / 25°C

NaH DMF 80°C BnCI / NaH DMF / 80°C

On closer investigation J. Bilow found that the rearrangement requires sodium or potassium hydride as the base and DMF or tetramethylurea as the solvent (8). DMSO is also suitable, but inferior with respect to the yield. In the absence of benzyl chloride the rearrangement does not occur. However, benzylether 49 can be prepared first and then submitted to the rearrangement by treatment with NaH in DMF. Analogous isomerizations could not be found in the literature; it was only by personal communication that we came across a similar example from Prof. W. Kreiser's group (University of Dortmund, Germany), namely the rearrangement of steroid 51 into 52, although a different base and solvent were applied. The scope of "our" rearrangement is reasonably large. Me^

Ma,

MeT ^V^

MeT

^"V

/ v ^

tÂ^

Ha^

Me

Li / Ethylendiamine

MeT

^^/\

6min/120°C HO^^^^

H 51

f52

168

Scheme 10 shows some examples; the yields are uniformly higher than 85%. The conversion of 57 into 58 is noteworthy with respect to the presence of the n-butyl group. Allylic alcohol 61 does undergo the rearrangement, but the stereocontrol with respect to the ^-configuration is much lower than in the examples above (> 98%).

OBn

OH BnCI, NaH DMF, 80°C

BnCI, NaH OBn

DMSO, 60°C

BnO OBn

55

56

BnCI, NaH OH

DMF, 80°C 57 BnCI. NaH

"

DMF, 80°C

? 7 'Y'®

BnOÂ/>^Me 60

BnCI, NaH DMF, 80°C

OH 61

OBn

62 (E)/(Z) = 3:1 -8:1

Scheme 10: Further Examples of the Double Bond Migration

Detailed mechanistic investigations (Scheme 11) revealed that the formation of the Eolefin is the resuh of a kinetically and not of a thermodynamically controlled reaction.

169

This was shown by preparing the Z-olefm 63 independently and submitting it to the conditions. No isomerization to 50 was observed. We suspected that the rearrangement proceeded via an allylic anion as an intermediate which could possibly be trapped with deuterium. However, to our surprise, no H-D-incorporation was observed, when the reaction mixture was quenched with D2O . Similarly, no deuterium was exchanged on performing the reaction in dy-DMF. This means that the concentration of the anionic species, if present at all, must be very low throughout the reaction. We next turned to the question whether the reaction proceeds inter- or intramolecularly. To this purpose the monodeuteriated alcohols 68 and 69 were prepared as shown in Scheme 11 and submitted to the rearrangement.

NaH / DMF 7^

50

•

80°C

LLiAIH.

H.C-

^

OH

64

H3C

H

D

2. D2O

PBh

H3C

ÔH 65

66

CrCl2 NaH/DMF BnCI OBn

Scheme 11: Mechanistic Investigations

OBn

170

The deuterium was quantitatively transferred into the expected position to give 70 and 71 as the reaction products. No H-D-exchange of the substrate with the reaction medium was observed. As a final confirmation of the strictly intramolecular process a cross-over experiment of non- and dideuteriated material was performed (Scheme 12). MS-analysis clearly demonstrated that only do and d2-product (i.e. 75 and 76) was formedfi-om49 and 73, whereas in case of an intermolecular rearrangement also dj-material 77 and 78 should have been generated.

64

2. D2O

^

OBn D

D

1.LiAIH4

H,c\^

3. PBr3

Br

1

D

OBn D

+ 67 ^ CrCl2

72 OBn H

OBn D NaH

73

+

0'

^

6

H Me

DMF

0^

y^

-Ar-6

" T ^ "C:H2D +

Me

O'' > ^

—Vo

"T^

"CH3

Me

49

analogous result with 74

78 (MS-Analysis)

Scheme 12: Final Confirmation of Intramolecular Rearrangement: Negative Cross-Over Experiment with Dideuteriated and Undeuteriated Material

It was also interesting to know whether the intramolecular 1,3-H-shift follows a suprafacial (79) or an antarafacial (80). To distinguish between these pathways both deuterium and tritium had to be introduced into the C-6-position of olefin 49 in a stereodefined manner so that after the rearrangement a chiral methyl group could be

171

created. After oxidative removal from the rest of the molecule the chiral methyl group can be analyzed in form of the chiral acetic acids (R) or (6)-81 according to Arigoni's enzymatic method (9).

OBn

OBn H

.JV-6

(H)

OBn

B

or

Me H

Me

49

H

80 antarafacial

H

CO2H

H

(S)-81

CO2H

(R)-81

Thus, 49 was converted into dibromide 82 (Scheme 13) and submitted to a FritschButtenberg-Wiechell rearrangement to give the acetylide which was quenched with T2O (activity lOOmCi/ml) to give 83. Lindlar deuteriation furnished olefin 82 which was then rearranged to 85 under standard conditions. Lemieuix oxidation proceeded without racemization to furnish (5)-81 with an ee of 45% (10). Analogously, isomer 53 gave (i^)-81with 44% ee. This means that the suprafacial and the antarafacial 1,3shift compete with each other in a ratio of 73:27. Although the enantiomeric access is comparatively low it is sufficient for most labelling studies and; in view of the simple overall access to intermediate 85, may find application. Moreover, by functional group manipulation of 85 the chiral methyl group can be directly incorporated into polyketide structures and related natural products.

172 OBn 49

1.2BuLi

I.O3

^ 2. CBr4 / Zn / PPhg

OBn D KMn04

HO2C..D

Nal04 (S)-81

by enzymatic analysi (D. Arigoni)

Scheme 13: Synthesis of Chiral Acettic Acid (S)-81

For example, the method potentially opens an access to compounds with a doubly chiral isopropyl unit (Scheme 14). Li the pro-iS-selective enzymatic hydroxylation of isobutyric acid (88) to (5)-^-hydroxyisobutyric acid (89) the stereochemistry of the hydroxylation at C-3 is not known. It could be studied by preparing 88 in a doubly chiral form via stereocontrolled anti-S^l' reaction of dimethyl cuprate with the tosylate 90 to give 91 which is then degraded by Lemieux- and then Baeyer-Villiger oxidation to 88. In a final overview (Scheme 15) "our" 1,3-H-shift is compared with the one described by Cram some thirty years ago (11). It may be concluded that there are certain similarities, however Cram used a protic system and a C-H-acidic hydrocarbon as a substrate and observed a reversible rearrangment. Quite interestingly, he formulated a very similar transition state (94) which was termed a "guided tour mechanism".

173 . CH3 R-CH

. CHDT R-CH

CDs

CDs

86 (simply chiral isopropyl group)

87 (doubly chiral isopropyl gruop)

Application CHs

pseudomonas

HsC ' ^ C O s H

putida

CHs

88 "

(S)-89

Pros

Hydroxylation under Retention or Inversion at C-3 ?

Possible Synthesis: OTs D

Me D Me2CuLi

MeD

^

HO2C

SN2'

anti

Me*

double chiral 88

Scheme 14: Application to the Synthesis of Doubly Chiral Isopropyl Units

Me

KOt-Bu HOt-Bu

PK^

Me

Ph" 93

92 Me ^ / 0 ^ - B u .®

via

Characteristics of Cram's System: 1) Reversibility 2) Ar necessary 3) 6 - 56 % intramolecularly

94

4) Racemisation at C^ 5) Stereochemistry at C^ not tested

Scheme 15: Anionic Olefinisomerisation by Cram (1964)

174

Synthetic Applications of the Key Intermediate 50 The rearranged olefin 50 may be used in a variety of synthetic appHcations. For instance it can be converted into the novel di-bis-tetrahydroftiran-acetal 95 in a one-pot operation using trimethylsilyl iodide in dichloromethane at 22 °C (Scheme 16) (12). The mechanism involves the formation of an oxonium intermediate 96 which undergoes a Prins cyclization to form the cation 97. Subsequent pinacol rearrangement generates 98 which cyclizes to 99. This acetal dimerizes under elimination of trimethylsilylbenzyl ether and benzyliodide. The structure of 95 has been elucidated by X-ray analysis (Fig. 3) which nicely shows the C2-symmetry of the dimeric structure.

q

^

Figure 3:

Ci

013

Crystal Structure of 95

Another apphcation of 50 is the synthesis of ^,y-unsaturated amino acids such as 102 (Scheme 17). To this end, 50 was debenzylated with sodium in ammonia and then submitted to a Mitsunobu reaction. Clean SN2'-reaction with a«^/-stereochemistry occurred to furnish phthalimide 100 which was converted into the acid 102 by standard modifications (13).

175 OBn TMS-i, CH2CI2 RT, 5 min

TMS-I TMSOBn/-Bn

OBn

OBn

TMSO 0 99

tTMS.;; k 0

TMS-I ®

^-/OBn V-Ô

,©. 98

Scheme 16:Tandem-Rearrangement-Dinnerization of 50

NHPht 1. Na/NH3

I.HgO^

0 ^

50 2. Phthaiimid, Azoester, PPha

2. Pb (0Ac)4 3. CrOa

100 NHPht

NHo

NoH 2^4 H02C

HO2C 101

102 (ca. 50% overall yield

Scheme 17: Synthesis of a,^-unsaturated y-Amino-acids

176

Another application of intermediate 50 lies in the synthesis of citrovireal 103 which is a metabolite of citreoviridine (104). This is an interesting polyene-pyrone toxin which has been isolated from penicillium citreoviride cultures. Citreoviridine has been shown to cause the Beri Beri disease which is acquired from eating infected rice. Li effect the toxin 104 acts as an inhibitor of the enzyme ATPase. So far, several syntheses have been reported for optically active 103 which has thus served as a goal for developing new synthetic methodology (14).

HQ

OH

Our retrosynthetic analysis is shown in Scheme 18. Retro-Wittig reaction leads to aldehyde 105 which is generated from alcohol 106 by Swem oxidation. This tetrahydrofuran system might be generated by ring closure of epoxy alcohol 107 although this would involve an SN2 type attack of the hydroxyl function at the more hindered position of the epoxide. The diol unit in 107 was to be created by osmylation of an allylic alcohol as represented by precursor 108 (8). However, the stereochemistry of this osmylation would be opposite to Kishi's model which predicts an anti attack of the osmium tetroxide with respect to the 1-ORfunction as shown by the conversion of 109 into 110. To circumvent this problem it was necessaiy to introduce the 1-OR group first in the wrong configuration in order to

177

exert the desired a/7//-controlling stereodirection on the osmylation to form 112. Subsequently the configuration at C-1 has to be inverted to get the overall correct arrangement at C-1,2 and 3 in intermediate 113.

HQ

OH

HQ

OH

RQ

OH

O H C ^ Q - ^ 105 \

OH =>

OR

OH 107

RÔ OR 108

Scheme 18: Retrosynthesis of 103

The actual synthesis (Scheme 19) starts with 50 which is converted into the labile aldehyde 115 \ia diol 114. Aldehyde 115 adds methylmagnesium iodide with high (>95:5) chelate Cram selectivity to form 116, which is submitted to osmylation. Not surprisingly the stereocontrol in the sense of Kishi's model is high, as now two hydroxyl groups cooperate in the same direction. After ketalization intermediate 117 is obtained which after Swem oxidation and Wittig methylenation furnishes olefin 118. Debenzylation with sodium leads to allylic alcohol 119 which is epoxidized with high stereocontrol to form 120. Cyclization with acid generates 121 which fails to undergo selective oxidation at the primary or secondary position, hstead the keto aldehyde 122 is formed which turned out to be a dead end in the synthetic sequence.

178

TFA

Pb(0Ac)4

MeOH, RT (95%)

CH2CI2, 0°C

H

51

OBn 115

1. 1%0s04, NMO, MeMgl

OH -,

CH3CN, 60«C — •

1^

Et20, RT (72%)

OBn

2. DMP, CSA, CH2CI2, RT (85%)

116 (additional OH!)

fBuOOH, Na/NH3

1. Swern-Oxid.

^^ 2. PhaPCHs^Br", NaH,DMSO, RT

PhH, reflux

(95%)

(84%)

HO

TFA

^0

^

120 (> 98: 2)

^.

THF, -30°C

(41%)

MeOH, 50°C (78%)

\/0(acac)2

0^

OH

OH

HO

f^ 0 121

Oxid.

O H C - ^ o " ^ 122

Scheme 19: Synthesis of CItroviral from 51 (Part 1)

Therefore, despite lower stereoselectivity (3:1) in the epoxidation step the benzyl ether 118 (Scheme 20) was converted into 123 and then converted into tetrahydrofuran 124. After Swern oxidation a mixture of the aldehydes is generated; the isomer with the correct stereochemistry at C-2 cyclizes to the hemiacetal 125 whereas the second C-2 epimer did not cyclize and was thus easily removed by chromatography. By Wittig reaction 125 was transformed into 126 which was smoothly debenzydated under Hanesssian's conditions (15) to give alcohol 127. Inversion of configuration at C-2 was achieved by an oxidation reduction sequence with complete stereocontrol.

179

DEAL reduction of the ester to the allylic alcohol and oxidation to the aldehyde delivered citreoviral 103 eventually.

118

mCPBA Na2C03

TFA

CH2CI2, RT (78%)

MeOH. RT (78%)

PhMe, reflux (77%)

HO 0

BnQ

PH 0

SwernOxid.

EtOsC

PhSSiMeg Znl2, BU4NI

OH

0

»•

(CH2CI)2, 60°C (80%)

126 HO

Et02C

PH

124

Ph3PC(Me)C02Et

125

BnO

ISwern-Ox. OH2. ZnBH4, THF,-50°C ' Inversion 3. DIBAH

OHC

127 Scheme 20: Synthesis of Citroviral from 51 (Part 2)

First Total Synthesis of ACRL Toxin III B (128) (15) ACRL toxins form a family of metabolites of the microorganism altemaria citri rough lemon which is reponsible for the brown spot disease of citrus fruits. All these toxins are polyene pyrone polyketides in different oxidation levels. When we started the project only one synthesis of an ACRL toxin was known, namely that one of ACRL toxin I by Lichtenthaler et al. (16). Later two additional syntheses of 128 were reported (17). Our retrosynthetic disconnection of 128 is shown in Scheme 2L It results in the formation of three fragments 129-131. The first one contains the trisubstituted olefin unit which is accessible by the above-mentioned base induced 1,3-H-

180

shift. Fragment 130 can be preparedfi-omacid 89 with different patterns of protective groups and 131 is commercially available.

4" OMe ACRL Toxin III B (128)

OPG

0

OPG'

X 4 >r^2

129

130

131

Scheme 21: Retrosynthesis of ACRL Toxin II B

OTs

o-r^H ^

0

1. Crotylation

I.HgO^

2. TsCI

2. OMe"

132 1. Me2CuLi 2. TrCI 134

Scheme 22: General Synthesis of Triad Fragments

QH TrO^^^V^^'^V^^ Me

Me

135 (- 50% Yield overall)

181

So we concentrated on the synthesis of fragment 129. The two stereogenic centers at C-7 and C-8 were established from (i^)-2,3-isopropylidene glyceraldehyde 132 as shown in Scheme 22 via a sequence already employed in the total synthesis of erythronolide B (18). Stereotriad 135 is available in multigram quantities on this route via 133 and 134 (Scheme 22). After protection of the secondary OH as a pmethoxybenzyl (PMB) ether the base induced 1,3-rearrangement was achieved under standard conditions to furnish the desired olefin 137 (Scheme 23).

135

OPMB

PMB-CI

NaH / DMF

OPMB

•

OTr 136

80°C (85%)

OTr 137

Scheme 23:1,3-H-Shift of 136 to 137

An alternative route (Scheme 24) involved hydromagnesation of 2-butyne. Addition of the Grignard derivative 139 to aldehyde 140 resulted in the formation of a 1:1mixture of 141 and 142, which was oxidized to the ketone 143 under Swem conditions. With superhydride Felkin Anh controlled reduction occurred which led to alcohol 141 under high stereocontrol. After 0-protection compound 137 was formed, indistinguishable from the product obtained via the first route.

Superhydride

141 Felkin-Anh-Product

22: 1

142 af?^/-Felkin-Anh-Product

182 0

138

"'^^^^^^^^T^^^^Y^"ÔTr 141

(140) 139 (E) - selective OH

OH +

f^ÔTr

-^^^

1:1

+ H^^ÔTr

MgCI

/ BuMgCI, EfeO -—— ^> 1,5mol%Cp2TiCl2

>wern-

C

0

-

OTr

Oxid 142

143

Scheme 24: Alternative Synthesis of 137

Intermediate 143 having secured, the synthesis was carried on by deprotection to form the primary alcohol 144 (Scheme 25) which was oxidized to the aldehyde 145 and converted into the envisaged alkyne (corresponding to 139) via a Corey Fuchs chain elongation via dibromide 146. Deprotonation with /?-butyllithium and addition of aldehyde 148 generated alcohol 149 as a 2:l-diastereomeric mixture. Again the stereochemistry at the newly created center was corrected by an oxidation reduction sequence via ketone 151. This time the chiral reduction had to be performed with using Corey's oxazaborolidine catalysts (19). In this way both the (3i?)- and (35)-diastereomer of alcohol were available. LAH-reduction of (35)-149 led to the £-alkene 150 which was eventually oxidized to aldehyde 154 after protection-deprotection via 152 and 153. Addition of the potassium salt of pyrone 131 gave 155 as a 4:l-epimeric mixture. Removal of the PMB protective group led to selective destruction of the minor diastereomer, so that a 95:5mixture in favor of the desired stereoisomer 156 was obtained (Scheme 26).

183

141

PMBCI

137

•

OPMB

ZnBr2

SwernOxid.

OPMB

NaH, DMF (95%)

CBr4, PPha, ^ Zn, CH2CI2 (90%)

THF, -80°C (91%)

146

. H'^^v-'ÔTHP

A?BuLi

^

OTHP

(148) ,

THF, -80°C (65%)

OPMB

147

LAM, Et20

(95%) i49 OH

(3S;3R = 2:1) OPMB

SwernOTHP

OTHP

Oxid.

150 OPMB

chiral

OH

•

reduction according to Corey (20)

OTHP (3R)-150 with (S)-Oxa2aborolidine

""

OTHP (3S)-150 with (R)-Oxazaborolidine

Scheme 25: Synthesis of ACRL Toxin III B

Desilylation furnished the target compound which was crystalline in contrast to the natural product and was characterized by a X-ray crystal structure analysis (Fig. 4). The polyketide backbone forms an all anti zig-zag chain. Tlie 4-,8-, and 10-hydroxyl functions all point upward whereas the 3- and 7-methyl groups point downward and are on the opposite face of the chain, which in this manner shows a hydrophylic top face and a hydrophobic bottom face . Possibly this property is essential to the biological activity of the toxin. Additional interesting structural features are the allylic

184

1,3-strain relationships between C-17/C-5 C-6/C-8 and C-7/C-9, which help to rigidify the observed zig-zag-conformation.

OPMB

TBDPSCl Imidazol »

OH OTHP

OPMB

OTBDPS

•-

OTBDPS OTHP

CH2CI2 (84%)

(3S)-150

Eton

OPMB

152

^w®''"Oxid.^

OPMB

OTBDPS

'ÔH

PPTS (95%)

(85%) 153 O' ^

154

(131) PMBO

^ OMe

OH

0

DDQ OMe

KHMDS, THF -100°C; (55%)

HO

TBDPSO

155

TBDPSO

OH

TBAF

0 OMe

156

»-

CH2CI2 (59%)

128

AcOH.THF (95%)

(95:5)

Scheme 26: Synthesis of ACRL Toxin IN B: the End Game

A final comment has to be made on the reduction of ketone 150 with Corey's catalyst 157 (19). The mechanism (20) involves the formation of transition state complexes such as 158 in which, by interaction with the rest of the molecule the small substituent (Rs) of the ketone points upward and the large substuent (RL) downward. Remarkable, for a,^-unsaturated ketones the vinyl group is the large one and this is indeed confirmed by our case. The reduction is reagent controlled, but the substrate in-

185

fluence is still rather high. So we obtained a ratio of (3i?):(35)-149 of 96:4 in the reduction with (5)-158, whereas for (7?)-158 the ratio was 15:85. Obviously the first combination was ,âtched" and the second one „mismatched".

Figure 4:

Crystal Structure of ACRL Toxin ill B (128)

^

I^Ph

'^^B

OH

BH3

Keton

Me (S)-OxazaborolidJne (157)

HVî Rs

158 RL

159

186

Synthesis of the C-26-C-32-Tetrahydropyran Moiety of Swinholide A (161) Swinholide A is an interesting physiologically highly active marine metabolite with a macrocyclic diolide structure and a polyketide carbon skeleton. Recently the first total synthesis of 161 was reported by I. Paterson et al. (21). We focused on the synthesis of the tetrahydropyran part of the molecule as represented by compound 162. The particular feature of this ring is that it bears the largest substituent (at C-27) in an axial arrangement, as shown by the X-ray crystal structure of 161.

OMe

OMe

Me' ^^^^LST^'H MeO 162

Swinholide A (161)

X-ray strucute of 161: axial position of 27-substituent

187

This means that this substituent has to be arranged by a kinetically controlled stereo selective method, which in our case was a Hetero-Diels Alder reaction between a diene 164 and a glyoxylate 165. Ketone 166 is the precursor of 164 and tartaric ester 167 that of 165. The methyl ether 164 could not been made by deprotonation/methylation of 166 (Scheme 27).

.^^k^v-OH

MeO

COOR*

MeO

162

163

H ^ ^ ^ R * 0 165

V

0

OH OR*

R*0 OH 166

O

167

Scheme 27: Retrosynthesic Plan: Hetero-Diels-Alder-Reaction

Instead the silylether 168 was prepared and treated with 169 under thermal and Lewis acid mediated conditions. The stereochemical consequences were enormous. Under thermal conditions a 2:1-mixture of cis- and trans-isomcrs were formed, whereas the MgBr2-induced reaction furnished /ra/i^-cycloadduct 170 exclusively. In asymmetric

versions using menthyl-or 8-phenylmenthylglyoxalates 171 and 173, the cycloadducts 172 (l:l-diastereomeric mixture with respect to the absolute configurations around the tetrahydropyran ring) and 174 (> 97 % diastereomerically pure) were obtained (Scheme 28)

OEt

II 0

TMSO

0^^-^^C02Et

19

168

11

toluene / A >>

170 (cis:trans 2:1)

\^MgBr2/0°C^

170 (only trans\)

symmetric Versions:

168

^

"VY^ 0

Y

MgBr2 THF

0

^^

'C02Menthyl

172(1:1)

171 Ph 168

+

"YxJ) 173

MgBr2 THF

174 (97:3)

Scheme 28: Hetero-Diels-Alder Addition

The effect of the magnesium bromide is interpreted in terms of transition states 175 (small steric interaction of the bromide with the diene) and 176 (severe interaction).

189 exo

Br OTHF

endo

^ ^Mg-OTHF . Br 0

175 favorable H Br

^\ '

OTHF

^Mg-OTHF " "Br :0

176 unfavorable HC Br

172a was submitted to a X-ray crystal structure analysis (Fig. 5). As in 161 the largest substituent adopts an axial position which in this case may be due to an electrostatic repulsion between the lone pairs of the endocyclic oxygen and the ester oxygen atoms. Carbonyl reduction proceeds stereoselectively from the axial side. This is obviously a consequence of Cieplak's model (22). For nucleophilic additions to cyclohexanone with electron withdrawing substituents in the ^-position to the carbonyl an enhanced tendency towards axial attack is postulated.

Figure 5:

Crystal Structure of 172a

190 Although this model has so far been discussed for cyclohexanones only it apparently is also applicable to of tetrahydropyran ketones such as 172. For attachment to the acyclic polyketide chain 175 was converted into the bromide 176 (Scheme 29). The further synthesis of 161 is continued in our laboratories.

i

)'"'^^-''''^C02Menthyl 172b

NaBH4 / MeOH

axial attack contra sterically!

=

HO'' ^^•^^'^C02Menthyl 175 1. MeOTf 2. LAH

^^ . .^.^^^ V / ^ O H HO" 176

Scheme 29: Carbonal-Reduction

Acknowledgement. The results reported in this review have been achieved by a number of unusually capable and active young scientists: Dr. Jom Bilow, Dr. Catarina Pietschmann, Dr. Bemd Schollhom, Dipl.-Chem. Martin Hiersemann, Dr. Barry Bunn, Dr. Giinter Funk, Dr. Stefan Greifenberg and Dr. Susanne Dupre and Dipl.-Chem. Frank Meyer. The X-ray crystal structure analysis has been performed by Dr. Jlirgen Buschmann and Prof Dr. Peter Luger, FU Berlin. I thank all collaborators for their enthusiasm and experimental shill. 1 am also very indebted to Prof D. Arigoni and Dr. Martinoni, both ETH Zurich, for their splendid analysis of chiral acetic acids. Financial support from the Schering AG, Berlin and from the Deutsche Forschungsgemeinschaft is also greatfuUy acknowledged.

References See for example: J.D. Bu'Lock in: D.H.R. Barton and W.D. OlUs (Eds), Comprehensive Organic Chemistry, Vol. 5, Pergamon, Oxford, 1979, pp.927. S. Omura andY. Tanaka, m: S. Omura (Ed), Macrolide Antibiotics, Academic Press, Orlando, 1984, p. 199. J. Staunton, Angew. Chem. Int. Ed Eng., 30 (1991) 1302.

191 2

R.W. Hoffmann, W. Ladner, K. Steinbach, W. Massa, R. Schmidt and G. Snatzke, Chem. Ber., 114(1981)2786.

3

J. Mulzer, C. Pietschmann, B. Scholhom, J. Buschmann and P. Luger, Liebigs Ann. (1995) in press.

4

J. Mulzer and B. Schollhom, Angew. Chem. Int. Ed Eng.,29 (1990) 1476.

5

D.A. Evans and J.M. Takacs, Tetrahedron Lett., 21 (1980) 4233. P.E. Sonnet and R.R. Heath, J. Org Chem., 45 (1980) 3137. D.A. Evans, M.D. Ennis and D.J. Mathre, J. Am. Chem. Soc, 104 (1982) 1737. Y. Kawanami, Y. Ito, T. Kitagawa, T. Taniguchi, T. Katsuki and M Yamaguchi, Tetrahedron Lett., 25 (1984) 857. W. Oppolzer, P. Dudfield, T. Stevenson and T. Godel, Helv. Chim. Acta, 68 (1985) 212. W. Oppolzer, R. Moretti and S. Thomi, Tetrahedron Lett., 30 (1989) 5603. K.-S. Jeong, K. Parris, P. Ballester and J. Rebek, h., Angew. Chem., 102 (1990) 550. T.-H. Yan, V.-V. Chu, T.-C. Lin, C.-H. Wu and L.H.Liu, Tetrahedron Lett., 32 (1991)4959.

6

M. Hiersemann and B. Bunn, unpublished results, Freie Universitat Berlin, 1993-1995.

7

Gunter Funk, PhD Thesis, Freie Universitat Berlin, 1991.

8

J. Bilow, PhD Thesis, Freie Universitat Berlin, 1994.

9

J. Liithi, J. Retey and D.Arigoni, Nature, 221 (1969) 1213. J.W. Comforth, J.W. Redmond, H. Eggerer, W. Buckel and C. Gutschow, Nature, 221 (1969) 1212. Review: H.G. Floss and S. Lee, Ace. Chem. Res., 26 (1993) 116.

10

We thank Dr. Martinoni and Prof.Arigoni, both ETH Ziirich for the determination of the enantiomeric excess.

11

D.J. Cram and R.T. Uyeda, J. Am. Chem. Soc, 84 (1962) 4358. Review: A.J. Huber and H. Reimlinger, Synthesis, 1969, 97.

12

J. Mulzer, S. Greifenberg, J. Buschmaim and P. Luger, Angew. Chem. Int. Ed. Eng., 32(1993) 1173.

13

J. Mulzer and G. Funk, Synthesis, 1995, 101.

14

Y. Shizuri, S. Nishiyama, H. Shigemori and S. Yamamura, J. Chem. Soc, Chem. Commun. (1985)292. D.R. Williams and F.H. White, Tetrahedron Letters 26 (1985) 2529. M.C. Bowden, P. Patel and G. Pattenden, Tetrahedron Letters 26 (1986) 4793. M.J. Begley, M.C. Bowden, P. Patel and G.Pattenden, J. Chem. Soc, Perkin Trans. I (1991)1951 S. Nishiyama, Y. Shizuri, S. Yamamura, Tetrahedron Letters 26 (1985) 231.

192 B.M. Trost, J.K. Lynch and S.R. Angle, Tetrahedron Letters 28 (1987) 375. S. Hatakeyama, Y. Matsui, M. Suzuki, K. Sakurai and S. Takano, Tetrahedron Letters 26 (1985) 6485. H. Suh and C.S. Wilcox, 1 Am. Chem. Soc. 110 (1988) 470. S. Hatakeyama, K. Sakurai, H. Numata, N. Ochi and S Takano, J. Am. Chem. Soc. 110 (1988)5201. K. Wang, H. Venkataraman, Y.G. Kim and J.K. Cha, J. Org Chem. 56 (1991) 7174. 15

S. Hanessian and Y. Guindon, Tetrahedron Lett, 21 (1989) 2305.

16

J. Mulzer, S. Dupre, J. Buschmann and P. Luger, Angew.Chem.InlEd.Eng., 32 (1993) 1452.

17

F.W. Lichtenthaler, J. Dinges and Y. Fukuda, Angew. Chem. Int. Ed Eng, 30 (1991) 1339.

18

I. Paterson and D.J. Wallace, Tetrahedron Lett., 35 (1994) 9477. M.J. Munchhof and C.H. Heathcock, J. Org Chem., 59 (1994) 7566.

19

J. Mulzer, H. Kir stein, J. Buschmann and Ch. Lehmann, J. Am. Chem. Soc, 113 (1991)910.

20

E.J. Corey, Pure & ApplChem., 62 (1990) 1209.

21

T.K. Jones, D.C. Liotta, I. Shinkai and D.J. Mathre, J. Org Chem., 58 (1993) 799.

22

I. Paterson, K.S. Yeung, R.A. Ward, J.G. Gumming and J.D. Smith, J. Am. Chem. Soc, 116(1994)9391.

23

A.S. Cieplak, B.D. Tait and C.R. Johnson, J. Am. Chem. Soc, 111 (1989) 8447.


193

Total Stereoselective Synthesis of Acetogenins of Annonaceae : A New Class of Bioactive Polyketides Bruno Figadere and Andre Cave 1.

INTRODUCTION 1.1 Cl^$$ifigation Annonaceae, a family of tropical and subtropical trees, are known by populations of South

America either for their edible fruits (for species of the Armona genus) or for their uses in traditional medicine as pesticide, antiparasite, etc.... Until 1980, the chemical studies concerned mainly isoquinoline alkaloids and secondarily neutral compounds such as terpenes, fatty acids, flavonoids (1). In 1982, from Uvaria acuminata was isolated a compound with an original structure, uvaricin, displaying an antitumoral activity, which belongs to a new class of natural products, bistetrahydrofuranoid fatty acid lactones (2). The biogenesis of this product was discussed and the polyketide origin through acetyl-coenzymeA elongation process was admitted. In 1984, isolation of roUinicin, a related compound with an interesting cytotoxic activity, has been described (3) and the name of linear acetogenins was proposed for this type of natural product. Due to their specificity and natural source, the name of Annonaceous acetogenin is now systematically used. To date, about 100 related compounds have been isolated and characterized exclusively from the Annonaceae (4a-b).

TETRAHYDROFURAN MOIErY (l,2or3THF)

Y-LACTONE

R = O, OH, OAc

Acetogenins of Annonaceae These cytotoxic molecules possess 35 to 37 carbon atoms, in a long alkyl chain, bearing oxygenated functions (e.g. hydroxyl, acetoxyl, ketone), and/or double bonds, one to three tetrahydrofuran rings (THF), with a y-butyrolactone at the end. Because of the presence of these functionalized groups, acetogenins possess many stereogenic centres. These compounds have been classified in four main types A-D as a function of the number and position of the THF rings. Type A is characterized by the presence of one THF ring, a,a'-dihydroxylated, as for solamin 1 (5), type B by two adjacent THF rings, a,a'-dihydroxylated, as for isomolvizarin 2 (6), and type C by two THF rings separated by 4 carbon atoms as for otivarin 3 (7). Very recently a new acetogenin, goniocin 4 (8), representing the fourth type D has been characterized, bearing three contiguous THF rings, a-

194 hydroxylated. These compounds are further subdivided into three subtypes 1-3 as a function of the nature of the y-butyrolactone. Subtype 1 is characterized by the presence of an a,p-unsaturated ymethyl-y-lactone. Subtype 2 is characterized by an a-acetonyl-y-butyrolactone, and subtype 3 by an p-hydroxy-y-methyl-y-lactone.

f TYPE ) HO ^^ '

r OH

\ SUBTYPE

.O^

)

O

hr^ HO

^^^^

Ai 1 Bi 1

H

OH

\

SUBTYPE-1

hr

TYPEB HO

OH

r

B2 1 C2 1

SUBTYPE-2

HO

TYPEC .O^

OH

s

«°

Ô^

HO-Q^

A3 1 Ba 1

\ SUBTYPE-3

Acetogenins of Annonaceae OH

16\

OH

/19

goniocin(Dl)4

wmJk

195 In addition to these acetogenins, some new products have been recently isolated, which bear in place of THF rings, epoxy groups and/or double bonds as for corepoxylone 5 (9), and sometimes only oxygenated function (e.g. hydroxyl, ketone) as for reticulatamol 6 and reticulatamone 7 (10). These related compounds belong to the rapidly growing group of the biogenetic precursors and metabolites, as for muricatacin 19 (11), of annonaceous acetogenins .

Acetogenins precursors (and metabolites) 10

corepoxylone 5

X= OH,H : reticulatamol 6 X= O: reticulatamone 7

muricatacin 19

1.2 Isolation Extraction and isolation of acetogenins of Annonaceae from the seeds, bark, leaves, or roots are guided by bioassays and TLC (12). The methanolic extract is partitioned with solvents (e.g. hexane, H2O, CH2CI2) and several chromatographic steps are necessary in order to separate from the complex mixture, compounds with very close polarity. The use of HPLC is very helpful (normal or reverse phase), since Light Scanning Detection (LSD) allows one to trace acetogenins even if they lack a chromophore (13). From the analytical point of view, it is worth noting that gpc can be used with acetogenins which have been previously treated with TMSCl, in order to prepare the corresponding silyl ethers (14). Recentiy, it was shown that extraction conditions are crucial in order to avoid re-arrangements occurring with 4-hydroxy acetogenins, leading to the isolation of artefacts (e.g. acetogenins of subtype 2). Indeed, it is now admitted that acetogenins of subtype 2 are formed by the fran^-lactonization of 4-hydroxy acetogenins (15). Dosage measurements of acetogenins in a crude extract have been studied by mass spectrometry. The close examination of FAB spectra with mnitrobenzyl alcohol doped with LiCI performed on the crude extract, allows one to know the relative composition in acetogenins of the mixture (16).

1.3 Structural Elucidation The elucidation of acetogenin structures is rather difficult and requires, besides classical methods such as UV, IR, proton and carbon NMR and mass spectrometry, some innovative mass strategies such as mass-tandem or colhsion-induced-dissociation (CID) of [M-hLi]"*" ions using linked

196 scan analysis at constant B/E (17). Concerning the determination of the configuration of the many stereogenic centres, the problem is complex because of the waxy nature of these compounds. Comparison of the NMR spectra with those obtained for models with known configurations, allows determination of the relative configuration of the THF skeletons (18). For the absolute configuration it has been proposed to apply Yamaguchi's method (19), which consists in analysing the NMR spectra of the Mosher's esters of acetogenins at high fields and deducing absolute configurations of carbon atoms bearing the hydroxyl groups. However, even though this method has been used with success for several natural products, some exceptions have been observed (20). Therefore, the determination of absolute configurations made so far by this method (21) have to be confirmed by stereoselective synthesis. Another point to stress is that the absolute configuration of an isolated carbon atom bearing a hydroxyl group cannot be determined by this method because of the intrinsic limitations. Recently, the determination of configurations of isolated hydroxyl groups has been made possible through NMR analyses of formaldehyde acetal derivatives coupled with Mosher esters methodology (22). For the stereogenic centre of the y-methyl-y-lactone, it has recently been proposed that the absolute configuration is (S), because of the oxidative degradation studies made for uvaricin which have shown the presence of (5)-lactic acid (23). In fact it is unclear for most acetogenins if this configuration is correct or not, because of the lack of any degradation studies. However, circular dichroism (CD) spectra have been used in order to deteimine the absolute configuration of this stereogenic centre, and a negative Cotton effect is in accord with the {S) proposed configuration (24).

1.4 Biological Agtivife It is now evident, that all acetogenins isolated so far, possess, to varying degrees, in vitro cytotoxicity against a large variety of carcinogenic cell lines (25). These cytotoxicities, measured at ED50, range among 10"! to 10" 1^ ^lg/mL according to the nature of acetogenins and the cell line. Some acetogenins exhibit an antiparasitic activity, and preliminary studies have shown some structure-activity relationships, leading to compounds with good therapeutic index, which have been patended (26). Pesticidal activity has also been described for several acetogenins, confirming traditional uses in South America (25). Recently, an interesting immunosuppressive activity was shown on mixed lymphocytes reaction in mouse cell system (27). For example, annonacin afforded CI50 = 3nM on this model (compared to cyclosporin with 10 nM on the same model). The mechanism of action of these new compounds is unknown. It has been shown that annonacin improved extrusion of K+ from lymphocytes (28) through a possible mechanism similar to antibiotic ionophores. The strong activity recently observed against complex I in mitochondrias could explain the high cytotoxicity found for such compounds (29). It is to answer so many questions that different groups around the world are studying the total stereoselective synthesis of acetogenins of Annonaceae.

197 2.

SYNTHESIS OF ACETOGENINS OF ANNONACEAE OF TYPE A (MONO-THF) 2. 1 Introduction Total asymmetric syntheses of natural and un-natural acetogenins of type A and type B have

been recently reported in the literature. Most of them are dealing firstly with the preparation of the THF fragment bearing the right relative and/or absolute configurations of the stereogenic centres, secondly with the preparation of the lactone moiety and finally with the coupling of the two synthons. The asymmetric syntheses are based on two different approaches, namely : (i) stereospecific strategies using as starting material a compound from the chiral pool (a-amino acids, sugars) and (ii) asymmetric induction using homochiral catalysts (Sharpless' epoxidation, Sharpless' asymmetric dihydroxylation). Besides these pathways, numerous approaches have also been reported dealing with the preparation of models which can be used as building blocks in the total synthesis of natural acetogenins (e.g. 2,5-disubstituted tetrahydrofurans (30-35), contiguous THF rings (36-38), ymethyl y-lactones (39, 40), a-acetonyl-y-lactones (41), ...). However, these approaches will not be discussed in this presentation.

2. 2 Stereospecific synthesis from chiral pool 2. 2. 1 From q-amino agids 2. 2. 1. 1 Svnthesis of gwr-4-oxo-2.33-dihvdrosolamin 8 (42) a-Amino acids are very convenient starting materials for the stereospecific syntheses of natural products (43). Glutamic acid, one of the most inexpensive a-amino acids is commercially available as its (5) and (R) form, allowing access to both parts of the molecule (the THF moiety and the y-methyl-y-lactone) in either (R) or (5) series. When the total syntheses of solamin 1 and murisolin were undertaken the relative configurations of contiguous stereogenic centers were known but the absolute configurations were unknown. Therefore arbitrarily the (155, 165, 195, 205, 34R) isomer of solamin 1 and (45, 155, 165, 195, 205, 34/?) isomer of murisolin, which appeared in 1993 to be the unnatural enantiomers of both compounds (21), were synthesized. The retrosynthetic pathway used was based on a disconnection of the carbon-carbon bond, between C-6 and C-7, which could be formed by a radical coupling of an alkyl iodide and an enone. The carbonyl so obtained could then be either completely reduced to afford solamin 1 after introduction of the unsaturation, or partially reduced to afford murisolin. This required the preparation of the enone 16 bearing the requisite configuration at C-34, and the alkyl iodide 27 bearing the THF moiety with the desired relative and absolute configurations for the four contiguous stereocentres. The synthesis of the enone, summarized on figure 1, starts from pure (R) or (5)-y-methyl-ylactone 12 which can be prepared in 4 steps from L- or D- glutamic acid. Deamination of glutamic acid by NaN02 in acidic medium gave rise to the carboxylic lactone 9 with complete retention at the stereogenic centre. Reduction of the carboxylic acid 9 by BH3.SMe2 then afforded the corresponding

198 alcohol 10 which was tosylated in a straightforward manner (TsCl pyridine). Reduction of the tosylate 11 was then performed in THF under reflux in the presence of 1 eq. of sodium iodide and 1 eq. of tributyltin hydride and a catalytic amount of AIBN. The y-valerolactone 12 was obtained in 80 % yield for the last step (46 % overall yield from glutamic acid in 4 steps and > 99% ee). Alkylation of 12, by treatment with 1 eq. of LDA and allyl bromide, led to a diastereomeric mixture of cis and trans alkylated products 13. Oxidative cleavage of the double bond by a catalytic amount of osmium tetraoxide in the presence of sodium periodate in dioxan, gave the desired aldehyde 14 in 70 % yield which upon addition of vinylmagnesium bromide, followed by a Swem oxidation, led to the desired enone 16 in 49 % yield for the last two steps.

D-giutamic acid

13V

1 4\

15V

16

\

Reapents! 1) NaN02, H2SO4, 70 %; 2) BH3.SMe2, THF. 98 %; 3) TsCl, pyridine, 87 %; 4) Nal, n-BusSnH, AIBN cat., THF, 80 %; 5) (i) LDA, TMSCl, (ii) allyl bromide, THF, 90 %; 6) OSO4 cat., NaI04, dioxan, 70 %; 7) vinylmagnesium bromide, THF, 0 °C, 51 %; 8) (C0C1)2, DMSO, Et3N, 96 %.

Preparation of the alkyl iodide 27 also started from L-glutamic acid, through a deamination process as described above. Treatment of the carboxylic acid 9 with oxalyl chloride in dichloromethane with a catalytic amount of DMF, gave the desired carboxylic acid chloride 17 in 92 % yield. Acylation of dodecylmagnesium bromide at low temperature and concentration with the acid chloride 17 afforded the corresponding ketone 18 in 85 % yield. Reduction of this ketone with LSelectride^M gave rise to the syn compound 19, namely (+)-muricatacin (44), as the major product (syn/anti = 98:2). It is worth noting that the use of tri-n-butyltin hydride with silica gel in dichloromethane allowed the and compound to be prepared as the major product with a

IMlsynlanti

ratio (45). Muricatacin 19 was then protected as a silyl ether 20 in 92 % yield by treatment with tertbutyldimethylsilyl chloride in DMF in the presence of imidazole. Reduction of the latter by DIB AL in toluene at -78 °C afforded the desired hemiacetal, which upon addition of acetic anhydride led to a 1:1 mixture of the anomeric acetates 21. This mixture when treated with trimethylsilyl cyanide in Et20 in the presence of a catalytic amount of either trityl- or scandium perchlorate gave rise to a 1:1 mixture of cis and trans nitriles which were separated by flash chromatography. Treatment of the cis nitrile with sodium r^rr-butoxide at room temperature in r^rr-butanol for 24 h led to the trans product 22 in quantitative yield. DIBAL reduction of 22 then afforded the corresponding aldehyde 23 (46),

199 whereas direct treatment with a functionalized Grignard reagent in the presence of trimethylsilyl chloride gave rise to the expected ketone 24 in 87 % yield. Reduction of the latter with LSelectride""^ yielded the syn-trans-syn compound 25 (98:2 d.e. determined by NMR) with the (5) absolute configuration for all stereogenic centres. Deprotection of the silyl ethers with tetrabutylamonium fluoride (TBAF) led to the triol 26 which upon treatment with 1 eq. of tosyl chloride at 0 °C gave rise to the monotosylated compound in 61 % yield. Displacement of the tosyl group by sodium iodide afforded the desired iodo compound 27 in quantitative yield. The cross coupling of the enone 16 with the iodo derivative 27 was performed under radical conditions by treatment of a stoichiometric mixture of 16 and 27 with 2 eq. of tributyltin hydride and a catalytic amount of AIBN in toluene under reflux. The desired coupled compound 8, namely 4-oxo-2,33dihydrosolamin, was then obtained in 55 % yield (Fig. 2). The synthesis was therefore achieved in 14 steps and 6.4 % yield from L-glutamic acid. Two more steps, i.e. reduction of the carbonyl group and introduction of the unsaturation would lead to either solamin 1 or murisolin.

L-glutamic acid 1 8

COCizHzs

1 9>—C12H25

— C12H25

8 Reagents: 1) NaN02, H2SO4, 70 %; 2) (C0Cl)2, DMF cat., CH2CI2, 92 %; 3) dodecylmagnesium bromide, -78 °C, THE. 85 %; 4) L-SelectrideTM, .78 ^c, THE, 88 % {syn/anti= 98/2); 5) TBDMSCl, imidazole, DMF, 99 %; 6) (i) DIBAL, -78 X , toluene, 99 %; (ii) (Ac)20, Et3N, DMAP, 20 °C, 96 %; 7) (i) TMSCN, SCCIO4 cat., Et20, 0° °C, 96 %, (x/p= 1:1; (ii) tert-BuOK, tert-BuOH, 20 °C, 24 h, 100 %, a/p= 100:1; 8) rfrr-BuMe2SiO(CH2)8MgBr, toluene, TMSCl, -78 °C, 75 %; 9) L-SelectrideTM, -78 °C, THE, 71 %, isyn/anti=^ 98:2); 10) TBAE, 20 °C, THE, 91 %; 11) (i) TsCl, pyridine; (ii) Nal, acetone, 61 % (for the last two steps); 12) n-Bu3SnH, AIBN cat., 16, toluene, 55 %.

200 2. 2. 1. 2 Synthesis of gp/-corrossolin 41 (47) Wu reported the synthesis of an epimeric mixture of natural corrossolin, starting from L-glutamic acid, which does not bear either the correct relative or the correct absolute configurations of the stereogenic centres in the molecule. Corrossolin (48) was described having a threo-trans-threo configuration across the THF ring, which means that the absolute configuration must be either (155, 165, 195, 205) or (15i?, 16/?, 19/?, 20/?). The strategy used was based on the enantiocontrolled preparation of both parts of the molecule and coupling of the two synthons by addition of a lithium acetylide on an epoxide. The disconnection was envisaged between carbon atoms C-12 and C-U. The lactone fragment was synthesized from methyl undecenoate which upon treatment with 1 eq. of LDA and then (/?)-0-tetrahydropyranyl lactal gave the aldol type product which was protected as its methoxymethyl ether 28 (in 55 % overall yield) and then hydrolyzed by H2SO4 10 % in THF to yield the lactone 29 quantitatively. Epoxidation of the double bond by MCPBA led to the desired epoxide 30 as an epimeric mixture of (34 R) diastereomers at C-10, C-2 and C-33 (unsaturation will suppress the stereogenic centres at C-2 and C-33) (Fig. 3). Figure 3

Methyl undecenoate

2 9

MOMO

3 0

Rfiassma: 1) (i) LDA; (ii) (/?)-0-THPlactal, 65 % (for the last two steps); (iii) MOMCI, /-Pr2NEt, 85 %; 2) 10 % H2SO4, THE, 100 %; 3) MCPBA, 64 %.

The THF fragment was prepared from L-glutamic acid which upon deamination with NaN02 in acidic medium followed by reduction of the so formed carboxylic acid 9, led to the desired alcohol 10 which was protected as a benzyl ether 31 by the usual method. Reduction of the lactone by DIBAL-H at -78 °C then led to the corresponding hemiacetal which upon Wittig homologation with methylenetriphenylphosphorane gave the desired alkene 32. lodo-etherification of this y-hydroxy alkene led to a 5:1 trans/cis mixture of 2,5-disustituted THF 33. Iodine displacement by ammonium acetate followed by saponification, and subsequent oxidation of the resulting free alcohol 34, led to the desired aldehyde. Addition of dodecylmagnesium bromide to this aldehyde afforded a 3:1 mixture of the syn/anti alcohols, 35 and 36, respectively, in 67 % yield which was separated by flash chromatography. It is worth noting that the undesired anti isomer 36 can be oxidized into the corresponding ketone under Swem conditions and the latter reduced by L-Selecu-ide-"^ to give rise to the syn alcohol 35 in 51 % yield for the last two steps. Acetylation of the free hydroxyl group of 35

201 and hydrogenolysis of the benzyl ether function led to alcohol 37, which under Swem oxidation conditions led to the the expected aldehyde. Addition of propargylzinc bromide then gave the anti homopropargyl alcohol 38 as the major compound with a 8.3:1 anti/syn ratio. It is worth pointing out that the anti product is the undesired epimer (since in corrossolin, the relative configuration is syntrans-syn), but the synthesis was carried out with this compound. Therefore, after saponification of acetate 38 and protection of the free hydroxyl groups as tetrahydropyranyl acetal 39, n-butyl lithium was added followed by BF3.0Et2 and epoxide 30 to afford the coupled product 40 in 58 % yield. Hydrogenation of the triple bond, deprotection of hydroxyl groups with PPTS in methanol and dehydration of p-hydroxyl-y-methyl-y-lactone by treatment with DBU in THF at room temperature, then led to the title compound 41. The synthesis was achieved in 20 steps and in 1.14 % overall yield from L-glutamic acid (Fig. 4).

L-glutamic acid

C12H25

Reagents: 1) NaNOi, H2SO4, 70 %; 2) BH3.SMe2. 98 %; 3) Ag20, BnBr, 83 %; 4) (i) DIBAL; (ii) CH2PPh3, 49 % (for the last two steps); 5) I2, NaHCOs, 50 % {trans/cis^ 5:1); 6) (i) Et4N0Ac, 67 %; (ii) K2CO3, 100 %; 7) (i) (C0C1)2, DMSO, Et3N; (ii) dodecylmagnesium bromide, -20 °C, THF, 67 % (for the last two steps) {syn/anti= 3:1); 8) Jones* oxidation; 9) L-SelecUide^M, 51 % (for the last two steps); 10) (i) (Ac)20, pyridine, 98 %; (ii) H2, Pd-C, 96 %; 11) (i) (C0C1)2, DMSO, Et3N; (ii) Zn, propargyl bromide, DMF/Et20 (1:1), 65 % (for the last two steps) {anti/syn= 8.3:1); 12) (i) K2CO3, 100 %; (ii) DHP, PPTS, 98 %; 13) n-BuLi, BF3.0Et2, 30, -78 °C, 58 %; 14) (i) H2, Pd-C, 100 %; (ii) PPTS, MeOH, 76 %; (iii) 4 eq. DBU, THF, R.T., 4 h, 68 %.

202

2. 2. 2 Synthesis of aldehyde 23 Even though glycosides seem to be the starting materials of choice for the synthesis of monoTHF acetogenins of Annonaceae, yery few examples are known in the literature. Indeed only one approach has been proposed by Gesson (49, 50) starting with D-glucofuranose which has been protected as a bis-acetonide and a benzyl ether before oxidation at C-6 to giye the corresponding aldehyde 42. Addition of tetradec-3-ynyimagnesium bromide on 42 afforded a 4:1 mixture of syn/anti alcohols 43, 44 which could be separated by flash chromatography. Lindlar hydrogenation of the major compound 43 followed by oxidation with MCPBA gave rise to a 1:1 mixture of trans/cis THF compounds 46, 47, which were separated by HPLC. Treatment of the trans product 46 by acetic acid and then sodium periodate led to the desired trans aldehyde 23 (Fig. 5). The use of this aldehyde as a building block for the total synthesis of acetogenins has not been reported yet by these authors. Figure 5 C10H21

H

R2H

D-glucose RiO

H

H H

C10H2

47 (Ri= Bz) Bno^ +

BnO 4 2

'-.x

BJ

HO.

>

R10 = H H H - O I 5 .0

46 (R,= Bz)

BnO

43 : Ri= OH. R2= H 44 : Ri= H, R2= OH 5

ot

CioHj'

45 R,0 = H

'••X_ CioHat

^^^

H i f "

^

23 : Ri= H

Rgaggnts: 1) ref 51; 2) l-bromo-3-tetradecyne, Mg, THF, syn/anti= 4:1; 3) Lindlar hydrogenation; 4) (i) MCPBA. CH2CI2, cis/trans^ 1:1; (ii) BzCl; 5) (i) H2, Pd-C; (ii) ACOH/H2O (1:1). 50 °C; (iii) 3 eq. NaI04. H2O. 20 °C.

2. 3 Asymmetric synthesis of Solamin 1 2. 3. 1 Keinan's synthesis (52) Keinan prepared separately the two fragments (the THF moiety and the y-methyl-ylactone) and used, as a key step of his sequence, the asymmetric dihydroxylation (AD-mix.-p), the yery efficient Sharpless' procedure for the formation of a.p-diols. Then, the cross-coupling was performed by addition of an alkyne and a yinyl halide in the presence of palladium and copper catalysts (Fig. 6). Treatment of the unsaturated ester 48 (prepared in 4 steps from commercially ayailable starting material, and 65 % oyerall yield) with AD-mix.-p in êrr-butanol/water (1:1) with methanesulfonamide for 16 h at 0 T afforded the lactone 49 which possessed 3 carbon atoms out of the 4 with the desired absolute configuration. Inversion of the fourth stereocentre after acetonide

203 formation of the vicinal diol (2 steps: tosylation, and epoxidation) afforded the lactone-THF 53. DIBAL reduction of the latter, followed by Wittig homologation with dibromomethylene triphenylphosphorane gave rise to the bromo alkene 54. Alternatively the lactone fragment was prepared in a straightforward manner as an alkyne derivative 55 (in one step and 70 % yield from(2/?, 4S) and (25, 45)-4-methyl-2-phenylthio-Y-butyrolactone (53)), which upon reaction with the bromo-alkene 54 in the presence of palladium triphenyl tetrakis, copper iodide, EtsN in THF at 50 °C gave rise to the enyne 56 in 70 % yield. Hydrogenation of the enyne 56 afforded 57, which after oxidation-thermal elimination of the phenylsulfoxide led to the desired solamin 1. In conclusion, this synthesis was achieved in 14 steps and 7.7 % yield from commercially available starting material. It is worth noting that the total synthesis of reticulatacin 154 (54) (which differs from solamin only by the length of the alkyl chain which bears two extra carbon atoms) has also been realized by the same authors. Figure 6

O ^COOB

f^^

OH

1

C12H25

4 8

k^COOMe

°

51

°-^

^y. -,^^%^ 56

i^ o

solamin 1 : n= 1 reticulatacin 154: n= 3

.^

o k.^ \ uo

57

C12H25

Reagents: 1) AD-mix-P, 66 %; 2) DMP, acetone, TsOH, 98 %; 3) TsCl, EtsN, DMAP, CH2CI2, 97 %; 4) K2CO3, MeOH, 88 %; 5) BF3.0Et2, CH2CI2, 75 %; 6) (i) DIBAL, -50 °C, THF; (ii) BrCH2PPh3-^Br-, rerr-BuOK, THF, 60 % (for the last two steps); 7) 55, Pd(PPh3)4, Et3N, Cul, 70 %; 8) H2, RhCl(PPh3)3, 95 %; 9) (i) MCPBA; (ii) toluene reflux, 72 % (for the last two steps).

204 2. 3. 2 Tanaka's synthesis (55) The key steps in Tanaka's synthesis are on the one hand the very efficient asymmetric epoxidation of an allylic alcohol, known as Sharpless' epoxidation, and on the other hand crosscoupling of an alkyne with a vinyl halide catalyzed by palladium and copper. Alkylation of propargyl alcohol 59 with dodecyl bromide 58 in liq. ammonia by lithium amide, followed by Lindlar hydrogenation gave rise to the (Z) allylic alcohol 61. Asymmetric epoxidation by the improved Sharpless procedure afforded the epoxy alcohol 62 with 84 % ee. Tosylation of the free alcohol (TsCl, pyridine) and then displacement with iodine, gave the iodo epoxide 63 which was then reacted with lithium enolate of tert-huty\ acetate to yield the alkylated product 64. Acidic hydrolysis afforded the hydroxy lactone 19 (muricatacin) (56). The latter was protected as its methoxymethyl ether 65 before reduction with DIBAL, yielding the hemiacetal 66 which upon reaction with pent-4ynylidenetriphenylphosphorane gave the acyclic compound 67. Epoxidation with MCPBA, followed by acidic cyclization led to a 3:2 mixture of trans, cis products (detemiined later in the synthesis) with predominantly the desired trans compound, which was separated by thin-layer chromatography. Protection of the free hydroxyl as a benzoyl ester, and deprotection of both the hydroxyl groups led to the THF moiety 68 (Fig. 7).

Figure 7

n-C-j2H25Br •

5 8

^ ' 59

C12H25

3 0

C12H25

- \ / -

3 X * \ / V , ' 5 = ^ Ci2H25

C12H25

6 6

65

OMOM

C12H25,

C12H2S,

^^^ 69 : Ri=R2= H

solamin 1 : n= 1 reticulatacin 154 : n= 3

Rgaggnt?: 1) LiNH2, Et20, DMSO, 71 %; 2) Lindlar hydrogenation, 91 %; 3) rerr-BuOOH, Ti(0/-Pr)4, L-(+)-diethyl tartrate, M. S., CH2CI2, 76 %; 4) (i) TsCl, DMAP, Et3N; (ii) Nal, acetone. 97 % (for the last two steps); 5) tertbutylacetate, cyclohexylisopropylamine, n-BuLi, HMPA, 81 %; 6) camphosulfonic acid, CH2CI2, 70 %; 7) MOMCl, /-Pr2NEt, 94 %; 8) DIBAL, -78 T , CH2CI2; 9) Pent-4-yn-l-yliriphenylphosphoniuin iodide, NaOEt, 0 °C, DMF; 10) (i) MCPBA, CH2CI2, 56 % (for the last three steps); (ii) BzCl, pyridine, 0 °C; (iii) NaOH, MeOH, 79 % (for the last two steps); 11) 78, Pd(PPh3)4, Cul, Et3N, 61 %; 12) (i) H2, RhCl(PPh3)3, 60 %; (ii) MCPBA, (iii) toluene, reflux, 40 % (for the last two steps).

205

The lactone fragment 78 was prepared from ethyl lactate, which in a few steps gave the lactone 77, and from propargyl alcohol 71. Alkylation of the lactone 77 with the diiodo compound 76 (prepared in 5 steps from 71) gave the desired furanones 78 (Fig. 8). The cross-coupling reaction is based on the same palladium catalyzed reaction of a vinyl halide with an alkyne used by Keinan, but herein the alkyne bears the THF skeleton and not the lactone part, as in Keinan's strategy. Therefore the reaction of the two synthons 68 and 78 with Pd(PPh3)4 in the presence of Cul and Et3N gave the desired coupled product 79, which after hydrogenation followed by a two steps sequence (oxidation-thermal elimination) afforded the desired solamin 1 (Fig. 7). In conclusion, the synthesis was achieved in 16 steps and 1.5 % overall yield, using as key steps the Sharpless epoxidation and the palladium catalyzed cross-coupling of an alkyne with a vinyl halide. It is worth noting that the same authors succeeded in the total synthesis of reticulatacin 154.

Figure 8

70 ^^""^^^

59 OTBDMS

*^73o

^'^"

^ ^ *^

71

72

' ' ' ^ ^ ^ OTBDM9-

'^*

_,

75

^'n'

76

-DsPh

78 Reagents: 1) n-BuLi, 58 %; 2) KAPA, H2N(CH2)3NH2, 71 %; 3) TBDMSCl, imidazole, DMF, 92 %; 4) (i) nBu3SnH, AIBN cat.; (ii) I2, 70 % (for the last two steps) (£/2= 3:1); 5) TBAF, THF, 85 %; 6) (i) TsCl, pyridine; (ii) Nal, acetone, 81 % (for the last two steps) {E/Z=^ 3:1); 7) 77, NaHMDS, 51 %.

2. 3. 3 Trost's synthesis (57) The very elegant and original su*ategy used by Trost relied on key steps such as (i) asymmetric epoxidation of allylic alcohol, (ii) a new synthesis of 2,5-disubstituted THF via a Ramberg-Backlund olefination and (iii) a ruthenium catalyzed butenolide annelation to form a direct precursor of solamin (Fig. 9). The synthesis starts by treatment of propargyl alcohol 59 with n-butyl lithium in THF/HMPA, followed by addition of bromododecene 80 to afford the alkylated product 81, which upon Lindlar hydrogenation gave the (Z) allylic alcohol 82. Asymmetric epoxidation with tertBuOOH, Ti(0i-Pr)4 and L-(+)-tartrate gave the desired epoxide 83 in 82 % ee which after recrystallization gave >99 % ee. At this point, this intermediate was used to prepare the two halves of the THF skeleton. Firstly the hydroxy-epoxide 83 was convened into the corresponding iodide 84 in 93 % yield (I2, PPh3, Et3N, THF). Secondly hydrogenation of 83 led to the alkane 85 which after a Payne rearrangement, on treatment withy re/t-BuSH, afforded the expected sulfide. Removal of the

206 f^rr-butyl group was performed with Hg(0Ac)2, PhOMe and CF3COOH at 0 °C, to give the thiol 86. Coupling of the two halves was performed under basic conditions to give the 1,4-oxathiane 87 in good yield. The best protocol for the Ramberg-Backlund olefmation was then performed on the corresponding sulfone (MCPBA oxidation of 87) with the hydroxyl groups protected as their silyl ethers, by treatment with tert-BuOK in tert-BuOH in the presence of CCI4 at room temperature to afford the dihydrofuran 88 in 65 % yield. Ruthenium catalyzed butenolide annelation of diol 88 with the ynoate occured chemoselectively at the less sterically demanding double bond to give the bisdehydrosolamin 89. Chemoselective hydrogenation of the isolated double bonds was then performed with (Ph3)3PRhCl and H2 to yield solamin 1 in 95 % yield. In conclusion, the synthesis was achieved in 14 steps and 11.7 % overall yield using a new and very efficient method for the synthesis of 2,5-disubstituted THF rings as well as the atu*active ruthenium catalyzed butenolide annelation.

Figure 9

Esamils: 1) n-BuLi. -78 °C, THF, HMPA, 76 %; 2) Lindlar hydrogenation; 3) terhBuOOH, Ti(0/-Pr)4, L-(+)-diethyl tartrate, M. S. -20 '^C, CH2CI2, 90 %; 4) I2, PPhs, C3H4N2, Et3N, THF, 0 °C, 93 %; 5) H2, Pd-C, 98 %; 6) (i) tertBuSH, NaOH. tert-BuOK H2O, 81 %; (ii) Hg(0Ac)2, PhOMe, CF3COOH, 0 °C, 92 %; 7) CS2CO3, DMF, R.T., 92 %; (ii) KOH, H2O, tert-BuOH, 65 %; (iii) m-CPBA, PhH,-hexane, 0 T , 95 %; 8) (i) TMSCl, Et3N, CH2CI2, 0 °C, R.T., 94 %; (U) tert-BuOK, r^rr-BuOH, CCI4. R.T., 65 %; (iii) TsOH, H2O, EtOH, R.T., 95 %; 9) CPRU(C6D)CI! MeOH, EtOOCC98%e.e.e.) was converted to its iodide l.PPh3,12. iniidazole O" 0 " 2. < A ^ o M e N-0

N-0

3.CuI 127

H^ 137

I 136 °

59%

I

1. chromatography 2. L-Selectride ^ 3. chromatography 75%

i

2. PCC 75%

1. H2,Raney-Ni, ^"^^^

2.(C02H)2 77%

H^ H= 138 " (major:minor = 91:9)

QT, V I \ /kÔ-K/C02Me H H= 139

Scheme 18 using triphenylphosphine, iodine and imidazole, and the iodide treated with the dianion of methyl 2methylacetoacetate to give the p-ketoester 136. Reductive cleavage of the isoxazoline ring followed by oxalic acid-catalysed cyclisation gave the ketone 137 corresponding to Bartlett's intermediate. Rhodium catalysed hydrogenation followed by PCC oxidation provided the ketone 138 where the C-2, C-3 and C-6 centres are correctly established. The last centre at C-8 was regenerated by a stereoselective reduction of the ketone 138 with L-Selectride (14), which provided the methyl (-)-8epinonactate 139. Barrett and Sheth synthesised rerr-butyl (±)-8-0-rerr-butyldimethylsilylnonactate 145 by a stereoselective hydrogenation of 8-0-t-butyldimethylsilyldehydrononactate 144, the 8-epimer and 80-protected analogue of Bartlett's intermediate, and solved the C-6 to C-8 problem in a completely different way by another hydrogenation (Scheme 19) (31). 2,3,5-0-Triacetyl-D-ribonolactone produced the achiral diene 140 on treatment with l,8-diazabicyclo[5.4.0]undec-7-ene (DBU). Hydrogenation of the diene stereoselectively (>97:3) gave the lactone 141, establishing the relative stereochemistry at C-6 and C-8. Diisobutylaluminium hydride reduction of the lactone gave the lactol 142. Wittig reaction, followed by catalytic hydrogenation over rhodium on alumina gave a lactone,

243

and protection of the free hydroxy group gave its silyl ether 143. Claisen condensation and dehydration gave the dehydrononactate intermediate 144, and the usual catalytic hydrogenation over rhodium on alumina then provided the O-silyl derivative 145 of racemic ten-butyl nonactate. AcO

OAc

OAc

DBU

AcO

DIBALH

O 140 l.Ph3P=CHC02Et 2. H2, Rh/Al203 TBDMSO

jP"

-AX.

85%

X ^ Q - ^ OH 142

—'A Pd/CaCOs

94%

3. TBDMSCl 55%

90%

O 141 OLi

2. Ambcrlilc 120 h 75%

143

OTBS

_/^^''

OTBS CO2BU'

H2,Rh/Al203 89%

144

{cis:trans - ^5:\5)

145

Scheme 19 Sutherland and his co-workers developed a concise route for the synthesis of Barrett's intermediate 148, also using hydrogenation of a cyclic intermediate to establish the C-6 and C-8 OH

1. H2,Rh/Al203 O 2. PhCOCl, Py ,x^,>^^^

NaOMe, MeOH

MCPBA

^^

3. Cr03, H-'

93% OBz

/'^-S^Q'Ô

H

BzO

148 147

146

Scheme 20 relationship (Scheme 20) (32). Catalytic hydrogenation of methyl hydroquinone, benzoylation and chromic acid oxidation gave the c/5-2,4-disubstituted cyclohexanone 146. Baeyer-Villiger oxidation

^^.^ss^CHO

L-(+)-diisopropyl tariaratc

+ ClMg.

^. 149 (£:Z = 96:4) OH OH

45% conversion OAc OAc

150 RuCl3,NaI04

96% ee K2CO3, MeOH

9^' 9^'

x-^v^^>^0

CO2H

86%

H 154

153

Scheme 21 of this ketone took place regioselectively to give the lactone 147, which on methanolysis gave the racemic lactone 148. Sutherland and co-workers also developed an enantioselective route (Scheme

244 21) (32), based on Sharpless asymmetric epoxidation (33), for the synthesis of Barrett's intermediate 154. This synthesis of Barrett's intermediate also involves the synthesis of the diol 151, a diastereoisomer of the diol 70 in Bartlett's nonactic acid synthesis. 3-Butenylmagnesium bromide and crotonaldehyde gave the diene alcohol 149. The diol 151 was then made in 96% e.e. by kinetic resolution using Sharpless epoxidation, followed by reduction of the epoxide with red-Al. The diacetate 152 obtained from the diol was oxidised with ruthenium trichloride-sodium periodate (34) to give the acid 153, which was converted to Barrett's lactone 154, but now in homochiral form. Batmangherlich and Davidson developed an enantiodivergent route to both enantiomers of tenbutyl nonactate by way of the lactone 156, with C-6 obtained from the a-C of glutamic acid 155 (Scheme 22) (35). The hydroxylactone 156 was protected as its rerr-butyldimethylsilyl ether, which steps

silyl protection

HO2C—< CO2H NH2 155

I. hydrogenation ——1 ^ 2. desilylation 60% from 156

OLi 156

TBDMSO. 157

OBu'

,,^ / \ ^^ ^ , l.Swem HO^^^^COjBu^ ^ H H I 2.Ph3P=CHMe (-)-158 (95% one isomer) 50%

l.NBS.DMS0,H20

OH V"

2. Bu3onH 60%

/—V H^

CO2BU'

OH

,

V"

H H = 160

/ \ ^^ ^ , k^^.^C02Bu' H

H =

159

/"A H^H i 161

160:161 = 1:4

Scheme 22 condensed with the lithium enolate of r^rr-butyl propionate to give the dehydro intermediate 157. Hydrogenation gave the tetrahydrofuran 158 with the correct stereochemistry required for C-2, C-3 and C-6. The chiral centre at C-8 was introduced in the wrong sense by a stereocontrolled bromohydrin formation on a c/^-olefm 159, controlled by the alkoxy group at C-6, and the bromine removed reductively. This gave a 1:4 mixture of r^rr-butyl (-)-nonactate 160 and rerr-butyl (-)-8epinonactate 161. OLi J hydrogcnauon

TsO.

(+)-158 90% one isomer

165 Scheme 23

For the synthesis of the (+)-enantiomer 165, the configuration at C-6 in the common intermediate 156 was inverted (Scheme 23). The p-toluenesulfonate 162 of the lactone 156 was

245 treated with the lithium enolate of tert-butyl propionate to give, by way of the epoxide 163, the (E)alcohol 164, duly inverting the stereochemistry at C-6. Hydrogenation of the dehydro derivative 164 then gave the alcohol (+)-158 with the desired stereochemistry at C-2 and C-3 as well. The centre at C-8 was then set up in the same way as before. Batmangherlich

and Davidson

(35) also resolved racemic methyl nonactate by

chromatographically separating its esters 167 and 168 with (5)-0-acetylmandelic acid (36).

Ph

Ph OAc

C02Me

OAq

^X'^'^vxtÔ'W H ^H

167

C02Me

i

168

Honda and his co-workers synthesised methyl (+)-nonactate 179, setting up the C-6 to C-8 relationship by a chelation controlled allylsilane reaction on the aldehyde 169, and the C-3 centre by hydrogenation of the dehydro intermediate 176 carrying two methoxycarbonyl groups (37) (Scheme 24). The thiolactone 175 and dimethyl diazomalonate gave the dehydro intermediate 176 in the presence of dirhodium tetraacetate, by way of a sulfur-ylid rearrangement developed by these

OBn

SiMe3

1

TiCU

OBn OTHP

83%

OBn uan

l.HCl,H20

OBn OTHP

»-

3. Swem 4.NaC102, ^,,^^ KH2PO4, T 75%

2. DHP, PPTS 169

l.B2H6,THF 2.H202,NaOH

I—I

171 R=CH20H 172 R=CHO 173 R=C02H OBn

N2=C(C02Me)2, Rh2(OAc)4

2. Lawesson's reagent 83%

174 X=0 175 X=S

1. H2, Pd/C, 7 atm.

C02Me

» 2. HCl, H2O, MeOH 177

C02Me

l.TBDMSCl, imidazole

TBDMSO

2. KOBu\ Mel 100%

1. BU4NF

OH

CO2MC ' C02Me

/-^ C02Me

C02Me 2. NaCl, DMSO, H2O

H^H 1:1

i

180

Scheme 24 workers (38) and by Takano and his co-workers (39). The dehydro intermediate 176 was hydrogenated using different conditions (10% Pd-C, MeOH-5% HCl) from Bartlett's, and obtained a 4:1 mixture of cis and trans tetrahydrofurans 177. Protection of the hydroxyl group with ten-

246 butyldimethylsilyl chloride followed by methylation of the protected diester gave the ester 178. The stereochemical control for the C-8, C-6 and C-3 centres was good, but no control at C-2 was observed during Krapcho decarboxylation of the malonate derivative 178, which gave a 1:1 mixture of methyl (+)-nonactate 179 and its diastereoisomer at C-2 180.

2.3 Cyclisation of a 1,4-Diol Derivative. This approach requires both a method for setting up the C-3 and C-6 oxygen functions with the correct relative stereochemistry to give the c/5-2,5-disubstituted tetrahydrofuran ring, and their differenriation in order to use the selective displacement of one of them with inversion of configuration. Only two groups have used this approach. Takatori's group prepared both C-3 and C6 diastereoisomers, and made the overall synthesis convergent by separating them, and converting the wrong isomer into the right one with a Mitsunobu sequence. The Fleming group used two silicon-based approaches. In one, the C-3 and C-6 centres were set up with independent absolute control as C-Si bonds, which were later converted to C-0 bonds, and in the other they were set up by moving the chiral information along the chain. Takatori and his co-workers (40) started from the y-dithio-p-hydroxy ester 181 as a homochiral building block derived by yeast reduction of a ketone supplying C-3 ready resolved (Scheme 25). The ester 181 was methylated with stereocontrol at C-2 by the method of Frater (41), and the product converted by way of the C-1-reduced and protected intermediate 182 into the lithium OH l.LDA,THF TBDMSO ^ S ^ J k ^ C O ^ E t 2.McI,THF,HMPA ^S^^XlJ 3.LiAlH4 X^S

181

V^CHO BOMO i«^ bUMU 184 HMPA 73%

^^fpP'BOMO 82%

r

4:TBDMSC1

H" H = 187 "

OTBDMS

^ 182

TBDMSO OTBDMS I I . . o , ^X^ J l.separauon& recycle , OH ,«c 3. TsCl, DMAP, EtsN ^^MO

^

TBDMSO

^ ^ 8 2

49^^

, BOMO

OTBDMS l.Mel.CaCOs 2. PhgP, CBr4

l.ClO3,H2S04, OH^^Me^CO ^H 3.H2,Pd/C 43%

^

^^^

H^ H E 188 "

95%

TBDMSO I

OH

OTBDMS i

^ H^ H= 189

Scheme 25 acetylide 183 by way of the vinyldibromide. This acetylide was added to the known homochiral aldehyde 184 to give the alcohol 185 as a 1:1 mixture of diastereoisomers, but separation, and recycling of the undesired diastereoisomer by a Mitsunobu inversion-hydrolysis sequence (not illustrated, but taking place in 61% conversion yield), overcame the lack of selectivity. Catalytic hydrogenation provided the differentially protected 1,3,6,8-tetraol, which was tosylated to give the C-6 tosylated derivative 186. Deprotection of the C-1 and C-3 hydroxy groups gave the cis-

247

tetrahydrofuran derivative 187, the cyclisation taking place with complete inversion of configuration at C-6. Oxidation of the primary alcohol group gave a carboxylic acid, which was esterified with diazomethane and subjected to hydrogenolysis to give methyl (-)-nonactate 188, and hydrolysis gave (-)-nonactic acid 189. Fleming and his co-workers developed two independent methods for the synthesis of methyl nonactate by ring closure of 1,4-diol derivatives. The stereochemical control needed for the synthesis of the appropriately substituted 1,4-diol derivatives was based on their work on acyclic stereocontrol using organosilicon compounds, and their routes are unique, and in consequence uniquely long, in eschewing cyclic control almost completely. The three aspects of their method of stereocontrol are: the transposition of chiral information from C-1 to C-3 in the electrophilic substitution of allylsilanes (42), the setting up of stereogenic centres with a 1,3 relationship using the hydroboration of allylsilanes (43), and the setting up of stereogenic centres with a 1,2 relationship by alkylation of enolates having a p-silyl group (44). The hydroboration and enolate alkylations leave the phenyldimethylsilyl group in the molecule, and it is converted, with retention of configuration, into a hydroxy group at an appropriate stage (45). Perhaps the most striking feature of these methods of stereocontrol is the sense in which the word "control" really means control: with each method, it is possible to obtain relative stereochemistry in either sense, making the methods equally suitable for the synthesis of any diastereoisomer. In the first route (Scheme 26) (46), the 1,4 diol system was set up by independently introducing silyl groups with absolute stereochemical control, that at C-6 by a stereospecific allylsilane synthesis from a homochiral allylic alcohol derivative, and that at C-3 by conjugate addition of a silylcuprate to an a,p-unsaturated carboxylic acid attached to a chiral auxiliary. Formation of the c/5-2,5-disubstituted tetrahydrofuran was achieved by converting the phenyldimethylsilyl groups into hydroxy groups, and differentiating between them in order to ensure that inversion of configuration took place at the desired centre. The C-2 and C-3 relationship was estabUshed by anti-scltciiwt methylation of a p-silyl enolate, and the C-6 to C-8-relationship was set up by hydroboration-oxidation of a trans allylsilane. The (5)-propargylic alcohol 191 (70% e.e.) was prepared from the ketone 190 using (S)alpineborane following Brown's and Midland's procedures. The alcohol 191 was converted to its carbamate, semihydrogenation of the triple bond of which gave the c/5-alkene 192. Stereospecific silylcupration (47) then gave the (£')-allylsilane 193. Hydroboration with thexylborane followed by alkaline hydrogen peroxide oxidation gave the anti alcohol 194 with high selectivity {antr.syn =95:5). For the synthesis of the (-i-)-enantiomer, this alcohol was subjected to a Mitsunobu inversion to give the syn diastereoisomer, which was protected as its benzyl ether 195. The aldehyde group in 195 was unmasked, and a Wittig-Homer reaction using the phosphonate 196 carrying Koga's chiral auxiliary gave the a,p unsaturated imide 197. Silylcupration on this imide gave an inseparable mixture of diastereoisomeric bis-silyl derivatives 198 with poor selectivity (2:1) in favour of the isomer illustrated. Stereoselective methylation on the p-silyl ester gave the ester 199, conversion of the silyl groups to hydroxy using mercuric acetate and peracetic acid then gave the 1,4-diol derivative, which was hydrolysed to the acid 200. The only problem left to solve was to differentiate

248

the two hydroxy groups, which was achieved by treatment with an excess of benzenesulfonyl chloride. Two things happened: protection of the C-3 hydroxy group as the p-lactone 201 and benzenesulfonylation of the C-6 hydroxy group. The p-lactoneringopened in acidic methanol and ring closure promptly took place, with inversion of configuration at C-6 to give a mixture of 0OH

1. L i - s 2.BuLi,THF O ^

3.AC2O 61%

1. PhNCCEtsN

190

OCONHPh

^. 2.H2,Pd/CaC03 PbO, MnCl2 91% OH

5-alpine borane, THF

SiMe2Ph

9

) Ô

192

ij

h. 70% ee

l.BuUTHF 2.CuI,Ph3P

SiMe2Ph

2. H2O2, NaOH 82%

3. PhMe2SiLi 73%

1.4-02NC6H4C02H,Ph3P, Et02cN=NC02Et 2. NaOH, MeOH

1. thexylborane

OBn SiMe2Ph

l.TsOH,Me2CO,H20 Ph3C0-.,,

3.BnOC(CCl3)=NH,TfOH

rs

•(EiO)20P'^^^ 0 0 86% 196 Ph3C0—., l.McOMgBr 2. LiHMDS

194 Ph3C0—. OBn SiMe2Ph

^3. McI, DMPU 73%

197 OBn SiMe2Ph

PhS02Cl, Py C02H'

PhMe2Si

64%

199 OBn 0S02Ph

OH y — y

l.TsOH,MeOH

C02Me

C02Me O

2.H2,Pd/C 83%

H 202

HE 203

Scheme 26 benzyl methyl (-i-)-nonactate together with other diastereoisomers. Removal of the benzyl protection by hydrogenolysis gave methyl (+)-nonactate 202, which was separated from the other isomers with the major byproduct being its C-2 and C-3 diastereoisomer 203. Two successive reactions independently setting up stereogenic centres has an arithmetical advantage, at some expense in overall yield, with respect to the enantiomeric purity of the major product, as Horeau (48) and Eliel (49) have pointed out. Although the selectivity in the steps leading to 191 and 198 are only 85:15 and 67:33, respectively, the methyl (+)-nonactate 202 and its enantiomer were obtained at the end of the sequence in a ratio of 92:8. This is because the proportion of the major enantiomer 202 is obtained by multiplying 0.85 by 0.67, whereas the proportion of the minor enantiomer is obtained by multiplying 0.15 by 0.33. The enantiomeric purity of the

249 intermediate alcohol 191 could be raised to >97% e.e. by three recrystallisations of its 3,5dinitrobenzoate 204, which would make the whole synthesis capable of delivering methyl nonactate of >99% enantiomeric purity (50).

3,5-(02N)2C6H3COCl T'^v

Et3N,DMAP

^ ^

98%

In the second route (51), Fleming and Ghosh developed an enantiodivergent approach in order to synthesise both enantiomers. Two silyl groups were set up on adjacent centres, destined to become C-3 and C-4, with a known 1,2-relationship between them. The silyl group on C-4 was then made part of an allylsilane 212 so that the stereochemical information could be moved three atoms along the chain by epoxidation, leaving a 1,4 relationship between the remaining silyl group at C-3 and the incoming oxygen atom at C-6 in the alcohol 215. The C-6 to C-8 relationship could then be controlled in either sense by reduction of a p-hydroxyketone using Evans's and Narasaka's methods, and the C-2 to C-3 relationship could be set up reliably by enolate methylation. By a suitable choice of reactions, the common intermediate 215 was converted into both (+)- and (-)-nonactic acid derivatives. The synthesis of the first homochiral intermediate 209 is shown in Scheme 27. The dimethyl meso 3,4-bistolyldimethylsilyladipate 205 was prepared by a samarium(II) iodide induced coupling ^ O ^ ^ SiMe2Tol l.LiOH,MeOH„THF SiMe2ToI Sml2, THF, DMPU ,C02Me '^5^C02Me " i r n r T T T — ^ Me02C' 2.DCC CH2(C02Me)2 SiMe2Tol ToIMe2Si SiMc2Tol 72% 205 206

'U

5'^^^™ ,.Mc3Si(CH,),0H, CO2H ix:c, DMAP •

HO2C ^ r ^^

84% from 205

207

o

2. H2, Pd/C %:4

TolMe2Si CO2H

Me-iSi' O

SiMe2Tol 209

Scheme 27 of the methyl (Z)-2-tolyldimethylsilylacrylate in THF-DMPU in the presence of dimethyl malonate (52). The homochiral mono 2-trimethylsilylethyl ester 209 of the dicarboxylic acid was prepared from the dimethyl ester 205 in four steps. Lithium hydroxide gave the dicarboxylic acid, which was

250 converted into the meso anhydride 206 by treatment with dicyclohexylcarbodiimide. Diastereoselective opening of the me5o-anhydride with Heathcock*s /?-(+)-2-naphthylethanol (99.7% e.e.) (53), the enantiomeric purity of which was raised by Horeau's method (54), gave a 96:4 mixture of diastereoisomeric mono-esters 207 and 208. Esterification of the mixture with 2trimethylsilylethanol gave the mixture of diastereoisomeric diesters, which was hydrogenolysed to give the mono-ester 209 with an e.e. of 92%. The allylsilane 212 and the common intermediate 215 were made from this monoester (Scheme 28). The lithium dianion of the acid-ester 209 was treated with the aldehyde 210 and the mixture of four diastereoisomeric aldols 211 esterified with diazomethane. The four possible diastereoisomers, present in a ratio of 76:9:9:6 were separated and the 2-trimethylsilylethyl ester group removed by treatment with tetrabutylammonium fluoride. The individual diastereoisomeric 1.2LDA,THF,DMPU CHO McsSi 2. O J )

TolMe2S MesSi' O

SiMe2Tol

3. CH2N2 75%

209

211a and 211b: 3. PhS02Cl, Py 4. collidine, heat

SiMe2Tol

211c and 211d: O O 3. Me2NCH(OCH2Bu')2 ^—^ CHCI3, reflux 93% SiMe2Tol .^^^^Ji^^^CQ^H

KH,THF O

O

^\

OSiMe2Tol 214

SiMe2Tol 2. Bu4NfF, THF

P OH SiMe2Tol 4:5,5:6 211a synsyn 76% 211b anu,syn 9% 211c syn,anu 9% 21 Id anti.anti 6%

TolMe2Si SiMc2Tol ^ = Hi

l.KOH,THF,MeOH 2.MCPBA,Na2HP04 O \ CH2CI2 92%

O jC,^^ / / OH O - ^ 213

O

H2,Pt02,MeOH 87% from 213 O 215

Scheme 28 hydroxy acids were each converted to the required trans allylsilane 212, by syn stereospecific decarboxylative elimination by way of their p-lactones for the acids derived from the esters 211a and 211b, and by and stereospecific decarboxylative elimination for the acids derived from the esters 211c and 21 Id, following chemistry developed earlier (55). The methyl ester was hydrolysed to the acid, which was epoxidised using m-chloroperoxybenzoic acid. The epoxide must have been produced with high anti stereoselectivity (antr.syn = 97:3), but it rearranged to the 7-lactone 213 by a stereospecific 1,2-shift of the silyl group from C-4 to C-5, probably with retention of configuration at C-4 and inversion at C-5 (56). The alcohol 213 on treatment with potassium hydride under the conditions of standard Peterson olefmation underwent stereoselective eliminative rearrangement, well precedented in the work of Yamamoto (57), to give the unsaturated acid 214. Deprotection of the 0-

251 silyl ether and hydrogenation of the double bond gave the hydroxy acid 215 in 41% overall yield from the adipate ester 205. The hydroxy acid 215 was the common intermediate for the synthesis of both methyl (+)-nonactate 220 (Scheme 29) and benzyl (-)-nonactate 227 (Scheme 30). The ketal 215 was hydrolysed with pyridinium tosylate and the ketoalcohol reduced stereoselective^ to the and 1,3-diol 216 (antiisyn = 96:4) using Evans's method (58). The C-6 and C-8 hydroxyl groups were differentiated by formation of the seven-membered lactone 217 using Mukaiyama's method (59). The minor enantiomer of the lactone 217 was largely removed because the racemate crystallised, thereby improving the e.e. from 92% to >99%. The 8-hydroxy group was

SiMe2Tol ,C02H l.PPTS,Me2C0 O

O

u y OH 215

SiMe2Tol CO2H

l.TBDMSCl, imidazole 2. LDA, THF, DMPU SiMe2Tol • TBDMSO 3. Mel

O.

92%

TBDMSO

1^

2. separate from racemate 90%

2. Me4NBH(OAc)4 87%

SiMe2Tol

0 218

l.TsCl,DMAP,Py ^. 2. TsOH, MeOH

KBr, AcOOH ^> NaOAc, AcOH 73%

C02Me 220

91%

Scheme 29 protected as its rerr-butyldimethylsilyl ether, and the lithium enolate was methylated to give the lactone 218. Conversion of the tolyldimethylsilyl group into the hydroxyl group with retention of configuration at C-3 was achieved using potassium bromide in peroxyacetic acid, and the hydroxy group in 219 was converted into its tosylate. Methanolysis opened the lactone ring and allowed the free hydroxyl group to displace the tosylate, giving methyl (+)-nonactate 220. The overall yield of (+)-methyl nonactate from the common intermediate 215 was 47%. For the synthesis of benzyl (-)-nonactate (Scheme 30), the hydroxy acid 215 was esterified and deketalised to give the ketoester 221. Stereoselective reduction of the ketone group using Prasad's modification of Narasaka's method (23) gave the syn 1,3-diol (syn:anti = 90:10), which was converted to its acetonide 222. Stereoselective methylation of the open-chain p-silyl ester gave only the ester 223 with the anti relationship between the incoming methyl group on C-2 and the resident silyl group on C-3. Differentiation of the C-6 and C-8 hydroxyl groups was achieved by removing the acetonide, hydrolysing the ester group, and forming the seven-membered lactone 224 using Mukaiyama's procedure (59). As in the earlier sequence, this lactone was enantiomerically enriched (from 92% to >96% e.e.) by removal of the crystalline racemic lactone. The free hydroxyl group in the lactone 224 was protected with r^rr-butyldimethylsilyl chloride, and the lactone opened

252 with sodium benzyloxide to give the benzyl ester in quantitative yield. The C-6 hydroxy group was then converted to its tosylate 225, and the C-3 tolyldimethylsilyl group to hydroxyl, as before. The intramolecular displacement with inversion at C-6 226 then gave directly benzyl (-)-nonactate 227. The overall yield of benzyl (-)-nonactate from the intermediate 215 was 35%. l.Bu2BOMe,NaBH4 SiMe2Tol THF, MeOH C02Me

SiMe2ToI I.CH2N2 CO2H 2. PPTS, Me2C0 86% 215

—

^.

2. (MeO)2CMe2.PPTS

SiMe2Tol 1. PPTS, MeOH 2 cOiMe 2. KOH, THF, McOH

SiMe2Tol C02Me l . L D A , T H F , D M P U ^ 8 ^ ^ 6

r

2. Mel

3. CI-^N"^ Et3N

89% less ??%

4. separate from racemate 83% l.TBDMSCl, imidazole 2.NaOBn,BnOH,THF

"^

^V-/^^'^'^™3.TsaDMAP,Py g \ 85%

SiMe2Tol ^C02Bn

jj TBDMSO

OTs

225

224

KBr, AcOOH AcOH 78%

OH TsO ^ ^ i 226

H ^H

i

227

Scheme 30

2.4 Electrophilic Cydisation of y,8-Unsaturated Alcohols and Enols In their synthesis of racemic methyl nonactate 233 and its 8-epimer 234 (Scheme 31) (26), Baldwin and Mclver controlled the stereochemistry of C-2 and C-3 by conjugate addition of homoallylmagnesium bromide to 2,2-dimethyl-3(2H)-furanone 228 and methylation of the regenerated enolate, which took place with high selectivity (10:1) in favour of the trans dialkylfuranone 229. Conversion of the ketone to the oxime followed by fragmentation with thionyl chloride and protection gave the nitrile, and the now free alcohol group was protected as its 2,6dichlorobenzyl ether 230 {anti:syn = 32:1). Conversion to the corresponding aldehyde with diisobutylaluminium hydride^ followed by exposure to iodine in acetonitrile gave the cyclic iodoaldehyde, which was oxidised to the corresponding acid 231. The iodoetherification took place stereoselectively in favour of the desired stereochemistry at C-6 {cis'.trans = 50:1). Dithiane addition and esterification gave the masked aldehyde 232. After removal of the protecting group, the aldehyde was treated with dimethylzinc in the presence of titanium tetrachloride to give methyl

253

nonactate 233 and methyl 8-epinonactate 234 in a ratio of 24:1. The same reaction using lithium dimethyl cuprate took place selectively (4.5:1) in favour of methyl 8-epinonactate 234 1.

%,y-^MgBT

W

CuBr, MeaS

OCH2C6H3CI2 CN

. SCX:i2, CCI4 •!

^> 2. LDA, THF, Mel

228

l.NH20H,Py

229

56%

3.NaH,THF 4.2,6-Cl2C6H4CH2Br 59%

anti'.syn 10:1

l.DIBALH

C02Me

^2.12, MeCN 3.CrO3,H2S04 54% 1. HgO, BF3.0Et2

C02Me

^.

Me2Zn,TiCl4 Mc2CuLi

24:1 1:4.5

2. Me2Zn, T i C ^ 65% orMe2CiiLi 60%

Scheme 31 Walkup and Park synthesised not only methyl (±)-nonactate 240a but also (±)-homononactate 240b and (±)-bis- 240c and trishomononactate 240d (Scheme 32) (60) starting from hexa-4,5dienal and the appropriate lithium enolate 235 in each case. The relative stereochemistry of C-6 and OLi

Me4N-' (AcO)3BH!•

OHC^x--^^'

MeCN, AcOH, - 4 0 X 236a 236b 236c 236d

235

OH OH

R=Me R=Et R=Pr' R=Bu'

55% 58% 55% 55%

OTBDMS

TBDMSCl,

l.Hg(02CCF3)2

IN

imidazole 237a R=Me 237b R=El 237c R=Pr^ 237d R=Bu'

90% 90:10 80% 96:4 90% 99:1 84% >99:1

^. 2. PdCl2 cat., CuCl, 238a R=Me >98% CO, McOH 238b R=Et >98% 238c R=Pr' >98% 238d R=Bu' 25% + 6-silyloxy-8-ol 75% OH

H2, Rh/Al203 C02Me

y—X CO2MC

CO2MC H^H

cis: trans >98:2 239a R=Me 87% 239b R=Et 70% 239c R=Pr^ 80% 239d R=Bu' 80%

240a-d

1:1

:

241a-d

Scheme 32 C-8 was controlled by Evans' reduction (58) of the p-hydroxyketones 236 giving the anti 1,3-diols 237. The y-silyloxyallenes 238 were then subjected to a one-pot procedure already developed by

254 these workers involving oxymercuration coupled to a palladium-catalysed methoxycarbonylation (61), which gave the tetrahydrofurans 239 with high stereoselectivity (cis.trans >98:2). This short sequence of reactions established efficiendy the required stereochemistry at C-8, C-6 and C-3, but, unfortunately, the final stereogenic centre at C-2 was generated with no control, catalytic hydrogenation gave a 1:1 mixture of the desired products 240 and their C-2 diastereoisomers 241. Iqbal and his co-workers reported a synthesis of 2,5-disubstituted tetrahydrofurans from Y,6unsaturated alcohols (Scheme 33) (62). The stereochemistry of the C-2 and C-3 centres was set up with some selectivity by reduction of the p-ketoester 242. Epoxidation of the terminal double bond

u

I.NaH 2.BuLi

C02Me

NaBH4

C02Me

243 53% OH

CI

V^-N.X^C02Me

242

244

,C02Me

MCPBA ^>

m.^^^^^}^

+ Ho..,^,,X^

78% 243

245

89:11

246

Scheme 33 of the major alcohol 243 with w-chloroperoxybenzoic acid was surprisingly well controlled, with the epoxide undergoing cyclisation under the reaction conditions to give the cis and trans tetrahydrofurans 245 and 246 in a ratio of 89:11. The major product, the alcohol 245, is the racemic methyl ester corresponding to the intermediate 158 in the Davidson and Batmangherlich synthesis of rerr-butyl nonactate (Scheme 22).

o- o-

> T ^ - . XX OMe

TBDMSO.

247 1. base, Mel

OTBDMS

249

O

TBDMSO ^^v

38%

^^^^

251

/—V C02Me

^^ 75psi 86%

l.NPSP,Znl2 '^^™S0 n-^ • / \ ^ /\^C02Mc 2. separation » | jj O j

250

H2, Raney Ni

.TBDMSO,

248

C02Me 2. Lindlar 65%

CO2MC ^g^^

252

^'^\^n'i^-^

CO2H

253 Scheme 34

Ley also used the alkylation of a p-dicarbonyl dienolate 248 to assemble the precursor 250 for an electrophile-induced cyclisation (Scheme 34) (63). The enol of the p-ketoester 250 underwent

255 cyclisation with N-phenylselenophthalimide (NPSP) to give a separable mixture of two diastereoisomers, from which the selenide 251 with the correct C-6 to C-8 stereochemistry was isolated. Raney nickel induced hydrogenolysis of the now superfluous selenide as well as saturation of the C-2 to C-3 double bond, as in Bartlett's synthesis, and gave the 0-silyl protected methyl nonactate 252, which was converted to nonactic acid 253.

2.5 Intramolecular Conjugate Addition ofAlkoxides Gerlach and Wetter established the relative stereochemistry between C-6 and C-8 at the beginning of the synthesis, and made the tetrahydrofuran ring by an intramolecular conjugate addition of the C-6 alkoxide to an a,p-unsaturated ester (Scheme 35) (11). The 1,3-diketone 254, prepared from the dianion of acetylacetone, was reduced with sodium borohydride to give a mixture of the diols 255 and 256 (3:2), which were separated by chromatography. The undesired erythro diastereoisomer was converted to the desired three isomer by tosylation, displacement with acetate ion and hydrolysis, and the combined crops of threo diol 256 were acetylated. Ozonization of the diacetate followed by Wittig reaction of the aldehyde 257 with the carbanion of pmethoxycarbonylethyl diethyl phosphonate gave a mixture of (£") and (Z) isomers 258 {E:Z 85:15). Base catalysed cyclisation of the a,p-unsaturated ester 258 (E:Z = 7:3) gave a mixture in ratios of 100:68:56:71 in which methyl (±)-nonactate 259 was the major product, separated as its rerr-butyl ether and ester. OH OH O

o

O

lin

KNH2

o

255

NaBH4

l.TsCl 2. separate

70% "^ 255:256 3:2 OH OH

3. NaOAc

4. KOH 13%

256 OH OH

OAc OAc

1. AC2O ^> 2. O3, Me2S

256

(ElO)20P^ C02Me ^CHO

257

l.KOH,MeOH.MeCN :N C02Me 258

£;Z7:3

1

66%

2. CH2N2, H-' 97%

9«

r\

C02Me

H^Hi 259

Scheme 35 Sun and Fraser-Reid reported a synthesis of methyl (-)-nonactate starting from D-ribose, C-4 of which (sugar numbering) provided C-6 (nonactin numbering) of the tetrahydrofuran ring (Scheme 36) (64). The ribose-derived aldehyde 260, was converted to the ketone 261 by a Wittig reaction followed by hydrolysis of the enol ether. Raney nickel catalysed hydrogenation of the ketone 261

256 provided the (S)-alcohol 262a with the correct C-8 stereochemistry for methyl (-)-nonactate 265 with high selectivity (9:1), probably stemming from chelation of the nickel to the ring oxygen atoms. In addition, the minor isomer was converted into the major by displacement of its sulfonate with sodium benzoate. The alcohol 262a was hydrolysed and protected as its acetonide to give the aldose 263, which was treated with the phosphorane. Wittig reaction took place followed by intramolecular

OHC

OMe

yOy 6j0

'^'^

A

yOy

H2.Ni /^

o3o

A

260

"I—\ ^ COMe

y—V ^ OMe

2 steps /

\ ^

1. separate 262a 2. HsO""

oTo

3.Me,C(OMe),

A 261

262a R^=OH,R2=H90% 262b R ' = H , R 2 = 0 H 10%

HO

1. Ph3PYC02Me ^^

/\^0E ^

l.benzoylate

^ ^ 3 3. Me2NCH(OR)2 Xô^^^'"^' ^ ^ 4. Ac20,heat H H i 'V' "^ 5. H2, Pd 255 ^^ '^ ^ 6. NaOMe 263 264 78% Scheme 36 conjugate addition of the alkoxide on the unsaturated ester under kinetic control to give a 1:3 mixture ^

2. separate / 3.NaOMe 4. separate and recycle

of the two C-2 diastereoisomers, with the desired isomer 264 the minor component. Under kintetic control, the side-chain at C-3 (nonactin numbering) remains on the upper surface as illustrated, an observation of Moffatt and his co-workers (65). The ratio was improved to 3:2 by equilibration with sodium methoxide by a p-elimination-readdition pathway. After three cycles of equilibration and separation, 90% of the mixture had been converted into the diastereoisomer 264. The acetonide group in the benzoate of 2 6 4 was hydrolysed and the resulting diol subjected to Eastwood deoxygenation (66), which gave the corresponding dihydrofuran. Hydrogenation over palladium then gave methyl (-)-nonactate 265. Sun and Fraser-Reid also synthesised the (+)-enantiomer from the same starting material, which required that the configuration at C-4 be inverted (Scheme 37). The early intermediate 261 prepared from D-ribose was treated with base, which caused epimerisation to give the thermodynamically more stable isomer 266, with an equilibrium ratio of 9:1 as expected from Moffatt's precedents, but surprising at first sight, given that the side chain is endo in the bicyclic system. Nickel-catalysed hydrogenation, selective enough to give the alcohol 267 to the extent of 75%, deprotection of the acetal, and protection of the diol as its acetonide gave the aldose 268. The aldose was treated with the nitrile analogue of the same phosphorane as before to give an epimeric mixture of the nitriles 269. This mixture was epimerised in a few cycles, with separation after each cycle, finally providing the nitrile 270 in 84% yield. The nitrile was used in this sequence because it behaved better in the equilibration steps than the corresponding ester. Eastwood deoxygenation.

257 hydrogenation of the dihydrofuran, and conversion of the nitrile to the methyl ester gave methyl (+)nonactate 271. O

/

V^y Q

OMe

Q

NaOMe, MeOH MeOH NaOMe,

O

OH

OMe

^Jl^ ^ ^ ^^'^\.

90% epimerisation

Q

261

Q

H2. Ni

OMe

JC^^V"

2. separate

J

266

OH

PhsP^CN

267 l.NaOEt,ElOH

0^0

A

2. NaOMe 93%

268

l.benzoylate 2. HsO"^ 269 3. Me2NCH(OR)2 OH ^-AczO^heat JC4^.K^C02Me

OH CN

2. Me2C(OMe)2

OH CN

OvÔ

^

1.H30-'

2. separate 3. recycle and separate 84%

A

5.H2,Pd 6.H2O2 7. NOCl 8. CH2N2 9. NaOMe

0^0

A

270

Scheme 37 2.6 From Bicyclic Intermediates White and his co-workers were the first to use a bicyclic intermediate to control the relative stereochemistry (Scheme 38) (14). They set up the 8-oxa-bicyclo[3.2.1]octene 273 using O

vV Br

fl

q

2.""v^jÔ ""2.^^ "'^"^ CF3CO3H, Na2HP04

Zn-Cu 272

Br

° ^

273 .C02Me 220 T

1. NaOMe O 2.NaH,CS2,MeI 71%

274

92% C02Me

MeS2C0 " 1 " 275 OHC

1, (Sia)2BH

CrOs

C02Me

^.

»•

278:279= 1:1

95%

2. H2O2, HO"

^ OHC.,^ H

46% 1. separate 278 2. MeMgl

64%

HE 279

C02Me

C02Me 280

A ^

1:1

281

Scheme 38 Hoffmann's cycloaddition (67) of the oxyallyl cation 272, generated from 2,4-dibromopentan-2-one with LeGoff s zinc-copper couple (68). Hydrogenation followed by Baeyer-Villiger oxidation gave the lactone 274, with C-2, C-3 and C-6 correctly set up. Methanolysis gave a single hydroxyester,

258 which was converted into its xanthate 275. The xanthate on pyrolysis provided the terminal alkene 276,

which was subjected to hydroboration-oxidation to give the primary alcohol 277.

Unfortunately, the configurational identity at C-2 was lost during the hydroboration-oxidation, the alcohol proving to be a 1:1 mixture of C-2 epimers. These were separated after converting the alcohol 277 to the mixture of aldehydes 278 and 279, and treatment of the isomer 278 with methylmagnesium iodide gave methyl nonactate 280 and methyl 8-epinonactate 281 with no selectivity, a problem that was solved later by Baldwin and Mclver (Scheme 31). Warm and Vogel used 7-oxabicyclo[2.2.1]heptan-2-one 284 to control the relative stereochemistry of C-2 and C-3 of methyl nonactate. They also resolved it, and used the (+)enantiomer to synthesise methyl (+)-nonactate (Scheme 39), and the (--)-enantiomer to synthesise methyl (-)-nonactate (69). Zinc iodide-catalysed Diels-Alder reaction between furan and 1cyanovinyl acetate gave the adduct 282, which was saponified to give the racemic ketone. This was hydrogenated using palladium on charcoal, and the enantiomers (+)- and (-)284 were resolved by chromatography of their sulfoximides 285 and 286. Pyrolysis of each diastereoisomer gave the

Aco^cN

"ô^

^ ^ ÔAcC 282

^ C N

2.H2.P(1A:

^-^^

^ ^

4:1 O

42% each

OH

O

' ,

5-

285

(+)-284

l.KHMDS 2. Mel ^> 3. separate from dimethyl product 4. MCPBA, NaHCO^ 59%

,C02Me

C02Me 288:289 1:3 289 1 l.KOH 2. CH2N2 288:289 3 (36%): 4 (27%)

L-Seleciride ^. 82%

C02Me

CO2MC

4

290:291 10:1 l.PhC02H,Ph3P,DEAD

|

2. NaOMe, MeOH 85%

Scheme 39 enantiomerically pure ketones 284 (>99% e.e.). Methylation of the bicyclic ketone (+)-284 followed by Baeyer-Villiger oxidation gave the unstable oxoacetal 287. Addition of one equivalent of the silyl enol ether of acetone to a 1:1 mixture of the acetal 287 and titanium tetrachloride gave a 1:3 mixture of the ketone 288 and its trans isomer 289. However, the undesired isomer 289 could be equilibrated on treatment with potassium hydroxide, by p-elimination and readdition. Acidification

259 and esterification with diazomethane gave the ketones 288 and 289, as a 4:3 mixture. Reduction of the ketone 288 with L-Selectride gave a 10:1 mixture of methyl (+)-8-epinonactate 290 and methyl (+)-nonactate 291. Mitsunobu inversion of the major product and treatment with sodium methoxide gave methyl (+)-nonactate 291. The enantiomeric bicyclic ketone (-)-284 similarly provided methyl (-)-nonactate.

2.7 Ireland-Claisen Rearrangement Ireland and Vevret developed a route for the synthesis of both (+)- and (-)-nonactic acids, with the stereochemistry at C-6 derived from C-4 of D-gulonolactone and D-mannose, respectively (70). For the synthesis of (+)-nonactic acid 301 (Scheme 40), the furanoid glycal 295 was prepared in 10 steps from D-gulonolactone 292 by fairly straightforward functional group manipulations. The

HO

l.HCl,MeOH 2. Me2NCH(OMe)2 3. AC20,130°C

l.Me2C0,H-' 2. DIBALH

OH

•

0

0

3. NaH, BnCl, DMF

6H«

o^ô-^ I

293

HO V-^

l.BuLi 2.EtC0Cl

292

MOMO

\-4 H ^ 294

1. 25°C

MOMO

l.Li,NH3 2.CCl4,Ph3P MOMO OBn 3.Li,NH3

^. 2. CH2N2

C02Me H ^H 299 47% or 54% _

H2,Pt/C C02Me ^> 44% from 295

MOMO

3. LDA, THF, -78°C

H^ 295

11% from 292 r=^

1

OBn 4. 9-BBN 5. NaOH, H2O2 6. P2O5, CH2(OMe)2 OSiMcs"

4. MeaSiCl

H^ 296 l.HCl,MeOH 2. Swem

MOMO

K0H,H20

V—TV

H ^H 298 298:2-epi-298 86:14

^ 3. Mc2CuLi or McMgBr

CO2H

95%

C02Me

300 53%^r40%

Scheme 40 glycal 295 was converted into its propionate ester, which on treatment with lithium diisopropyl amide in THF and trimethylsilyl chloride gave the £-trimethylsilyl enol ether 296. The key step, the Ireland-Claisen rearrangement (71) setting up the stereochemistry at C-2 and C-3, took place at room temperature, and the product mixture was esterified to give a mixture of the C-2 diastereoisomeric

260 esters 297. Catalytic hydrogenation gave the corresponding mixture of esters 298 (86-89:14-11) in favour of the desired isomer. Evidently the Claisen rearrangement had taken place largely with the boat-like transition structure, and suprafacially on the dihydrofuran ring. Deprotection followed by Swem oxidation gave the aldehyde. No stereochemical control was observed in the dimethylcuprate addition to the aldehyde, which gave both diastereoisomers 299 and 300 in approximately equal amounts, in contrast to Baldwin's observation of good control, although in the wrong sense, in this reaction (Scheme 31). They observed somewhat better control when methylmagnesium bromide was used. The enantiomeric glycal 305 was prepared from the D-mannose 302 in 11 similar steps (Scheme 41), and (-)-nonactic acid 310 prepared from it exactly as described for the (+)-enantiomer 301 (Scheme 40).

V

OH l.Me2C0,H-' .OH 2. NaH, BnCI, DMF HO^ 'V-^^^^^*^" HO^'^^Ô-ÔH 302 Q' -Q

I.Li,NH3 2.CCl4.Ph3PMOMO

^ ^ O ' ^ O B n 3.Li,NH3 ^ H 36% 304 36% from from 302 302

H2, Rh/C

I

jiQ V-a

l.BuLi 2. EtCOCl

3.NaOH.H202 4. KH, MCOCH2CI

MOMO

^ V ^ >> 3.LDA,THF. H^ -78°C 3^5 305 4. Me.SiCL MesSiCl, 25°C 5. CH2N2

V

C02Me

49% from 305

I.AC20,130°C 2. 9-BBN

3.HCl,MeOH O ' ^ r ^ O ' ^ O cB n 4. Me2NCH(OMe)2, CH2CI2 V-n " 303 Me2N f ^ -i^-

MOMO

MOMO M UMU

O^Ô

H" H I 307

l.HCi,MeOH 2. Swem

k^^^tN^C02Me H H^ 306

x ' ^ . X h Q ' t v ^ CO2MC H "^ H =

3. Me2CuLi

308 40% OH

307:2-epi-307 89:11

r ^ H^ H= 309 45%

. - ^ X x t ^ Q ' t ^ C02Me

KOH, H2O

H "^ H 308

*- -'^"'^ô'^V' CO2H H ^ H= 310

Scheme 41

3.

SYNTHESES OF NONACTIN The synthesis of nonactin requires that the (+)- and (-)-nonactic acid units be joined together in

an alternating sequence, followed by closing the ends to give the macrocycle. There are two possibihries for ring-formation: (i) cyclodimerisation of a "dimer" and (ii) unimolecular cyclisation of

261 a "tetramer." Both strategies have been used, with several different ways to assemble the dimer and tetramer. In one strategy, the differentially protected nonactic acid enantiomers are coupled to give the protected dimer using standard esterification techniques that preserve the configuration at C-8. In the other strategy, the linear units are coupled by taking one enantiomer of nonactic acid, and using it as a carboxylate nucleophile to displace, with inversion of configuration, the 8-methanesulfonate or tosylate of the 8-epi-diastereoisomer of the other enantiomer, protected at the carbonyl group. 3.1 Synthesis ofNonactin by Unimolecular Cyclisation The first synthesis of nonactin was reported by Gerlach and his co-workers (Scheme 42) (72) in 1975 by the cyclisation of a linear tetraester, but the linear tetraester 317 was a mixture of diastereoisomers because it was made from racemic nonactic acid 311 (prepared in Scheme 35).

CO2H

[ I (+)-311 •^ Bu'OsCMe, MeSOsH 70%

NaH,BnBr ^ OBn

CO2BU'

Py, C i O z S - f J -

70%

COzBu^

H2, Pd/C CO2BU'

C02Bu^

I.CF3CO2H 2. H2, Pd/C

i

4. AgC104, MeCN, 10--*M 5. separate

nonactin (10%) + other diastereoisomers (30%)

Scheme 42 Appropriately protected monomers, the racemic benzyl ether 312 and the racemic rerr-butyl ester 313, were coupled using the mixed anhydride with 2,4,6-trimethylbenzenesulfonyl chloride to give

262 the protected dimer 314 as a mixture, inevitably, of four diastereoisomers, all racemic. Treatment with trifluoroacetic acid removed the rerr-butyl ester group from one portion of the dimer to give the acid 315, and catalytic hydrogenolysis removed the benzyl ether group from the other portion to give the alcohol 316. Activation of the acid 315 with 2,4,6-trimethylbenzenesulfonyl chloride and coupling with the alcohol 316 gave the linear tetramer 317, this time as a mixture of eight racemic diastereoisomers. Deprotection of the rerr-butyl ester with trifluoroacetic acid and of the benzyl ether by hydrogenolysis gave the free linear tetramer, which was cyclised by the Mukaiyama thioester method (73). At this stage the complexity of the mixture became rather less, since there are only four possible diastereoisomers, assuming that there is complete preservation of stereochemical integrity within each nonactic acid unit. Of these four diastereoisomers, three were isolated by chromatography in a ratio of 1:5:2, and the last proved to be nonactin 1. Schmidt and his co-workers (74) were the first to report a synthesis of nonactin from enantiomerically enriched components (Scheme 43). Potassium (-)-nonactate 321, prepared from

4îs^C02Bn H ^H =

H

Asx/o^C02H

H

H

(+)-319 i TsCl,Py 85% ,—V

OTs

H = (-)-320"

OH

An

OTs

C02Bn . A ^ ^ V ^ ^ ^ z "

OH

1

y

O

:

,

1

OH

I

V

C02Bn

O C02Bn

H ^ H = •

H ^H

H" H E

324

I l.H2,Pd/C 2.KHCO3 OH

r—.

O

=

J—.

OTs

-^ôWô-^-ôV^^^'

/-^

O C02Bn

/ ^ oH^ t H^ =- ^ o

327

OH

y

i

O

r

i

.

O

| 7 4 % from 324 I y 1

O

r

y

v ^C02Bn

H "^ H =

H^H

H "^ H =

H

H

328

I l.H2,Pd/C 2. (PyS)2,Ph3P I 3.AgC104,bei benzene 20% nonactin Scheme 43 the acid 318 (=38) was coupled, with inversion of configuration at C-8, with the 8-epi-tosylate 322, derived from benzyl (+)-8-epi-nonactate 319 (prepared from the methyl ester 40), to produce

263 benzyl (-)-nonactinoyl-(+)-nonactate 324. The diester 325 was prepared similarly using potassium (-)-8-epinonactate 323 and the same 8-epi-tosylate 322 used before. Hydrogenolysis of the "dimer" 324, and conversion to the potassium salt with potassium bicarbonate gave the left-hand component 326. Tosylation of the alcohol 325 gave the 8-tosylate 327, which gave the linear tetraester 328, again with inversion of configuration at C-8, on treatment with the potassium salt 326. Hydrogenolysis, activation and cyclisation following Gerlach produced nonactin in 20% yield together with C-2 and C-8 epinonactins in 12% yield. Fleming and Ghosh synthesised nonactin by cyclisation of the linear tetramer 335 assembled from methyl (+)-nonactate and benzyl (-)-nonactate (Scheme 44) (75). The (9-protected (+)-nonactic

C02Me (+)-329 I l.TBDMSCl, imidazole 2. KOH, THF, MeOH 98% TBDMSO

331 DCC, DMAP 93% TBDMSO C02Bn H ^ H = 100% H2,Pd/C TBDMSO

/—V

TsOH,AcOH 98%

O

Ovô^^

C02Bn

H

HÂ

H E

U H

^

U H

'

334 CI

DMAP, ClOC C ^ C l

95%

CI TBDMSO C02Bn H

H

H

H =

H ^ H =

335 l.H2,Pd/C 2TsOH, AcOH, H2O 3. CI

t nonactin

DMAP, ClOC • ^ - C I

69%

CI

Scheme 44 acid 330 was prepared from the hydroxyester 329 (=220), and coupled with benzyl (-)-nonactate 331 (=227), without inversion at C-8, to give the dimeric ester 332. A portion of this dimer was hydrogenolysed with palladium on charcoal to give the acid 333, while the other portion was

264 deprotected at the hydroxy group using acid to give the alcohol 334. The acid 333 was coupled to the alcohol 334 using the Yamaguchi mixed anhydride method (76) to give the protected linear tetramer 335. The protecting groups were removed to give the free tetramer, which was cyclised in high yield (73%), the best so far achieved, again using Yamaguchi's method. There was no improvement in yield when potassium tetrafluoroborate was present, indicating that coordination to potassium did not help the cyclisation. 3. 2 Synthesis ofNonactin by Cyclodimerisation Schmidt and his co-workers also reported the synthesis of nonactin (Scheme 45) (77) by cyclodimerisation of (-)-nonactinyl-(+)-nonactic acid 336 (=326), which was treated successively OH y i H "^ H

O

=

y y H "^ H

336

/V^V'-carbonyldiimidazole, DBU i

or (PyS)2,Ph3P

"poor yield"

nonactin

Scheme 45 either with carbonyldiimidazole and diazabicycloundecen or with the bispyridyl 2-disulphide and triphenylphosphine to give nonactin in "poor" yield in both cases.

H^ H I (-)-338 ^ MsCl, EI3N, DMAP 91% OMs / — ^ C02Me H

H = 340

C02Me H ^ H = 341 I LiSPr",HMPA 1 CO2H H ^ H l.(PhO)2POCl,Et3N I 2. heat, C6H6, DMAP 16% nonactin + cyclic "dimer" and "oligomers" and polymer

Scheme 46 Bartlett and his co-workers (20) synthesised nonactin by the cyclodimerisation approach (Scheme 46). The potassium salt 339 of (4-)-nonactic acid 337 (=87) and the mesylate 340 of the

265 (-)-8-epiester 338 (=86) gave the dimeric ester 341 with inversion of configuration at C-8, as in Schmidt's synthesis, but working with the enantiomer of each component. The methyl ester was cleaved by lithium n-propyl mercaptide with some difficulty, and with some (25%) epimerisation at the C-2 positions of the nonactic acid units to give the acid 342. Macrolactonisation using Masamune*s procedure (78), gave nonactin in 16% yield, accompanied by the cyclic "dimer" and "oligomers" and polymers. Fleming and Ghosh also synthesised nonactin by cyclodimerisation of the acid derived from the benzyl ester 343 (=334) using Yamaguchi conditions (Scheme 47) (75). Nonactin was isolated in lower yield than by the linear tetramer method, presumably because of the problem of "dimer" and "oligomer" formation.

Q^

o-^^-^/?k^^^^^" H H E 343

l.H2,Pd/C 2.

CI

aoc-0"Ci DMAP a

nonactin (52%) + cyclic "dimer" and "oligomers" and polymer Scheme 47

CONCLUSIONS The syntheses of nonactic acid and its derivatives illustrate many of the most popular methods of stereocontrol used in synthesis. There are examples of absolute control based on (a) resolution, (b) Sharpless asymmetric epoxidation and other methods of kinetic recognition, including an enzymatically controlled reduction, (c) the use of chiral auxiliaries, and (d) starting materials from the chiral pool such as sugars, and malic and glutamic acid. Relative stereochemical control has been achieved by such devices as (a) control on bicyclic frameworks, (b) the use of many different cyclic structures, especially five-membered rings with a predictable stereochemical bias, (c) the similar use of cyclic transition structures for hydride delivery, for enolate alkylations, the Ireland-Claisen rearrangement, and for ring-forming reactions, t>oth pericyclic and ionic, and (d) by the independent synthesis of separated stereogenic centres with absolute control. There are examples of such themes as (a) kinetic and thermodynamic control, (b) of convergent and linear synthesis, (c) of the recycling of unwanted disastereoisomers by Mitsunobu and other inversion processes, and by repeated equilibration and separation, and (d) of the problems of controlling distant stereogenic centres. And there are examples of a very wide range of the common reactions of organic chemistry, including those used in C-C bond-formation, functional group manipulation, and protecting group tactics. The syntheses illustrated here would, on their own, make a surprisingly good basis for an introductory course in organic synthesis.

266 REFERENCES 1

R. Corbaz, L. Ettlinger, E. Gaumann, W. Keller-Schierlein, F. Kradolfer, L. Neipp, V. Prelog and H. Zohner, Helv, Chim. Acta, 1955, 38, 1445.

2

(a) W. Keller-Schierlein, Fortschr. Chem. Org. Naturst., 1968, 26, 161; (b) W. Keller-Schierlein, Fortschr. Chem. Org. Naturst., 1973, 30, 313; (c) J. Dominguez, J. D. Dunitz, H. Gerlach and V. Prelog, Helv. Chim. Acta, 1962, 45, 129; (d) H. Gerlach and V. Prelog, Justus Liebigs Ann. Chem., 1963, 669, 121; (e) J. H. Prestegard and S. I. Chan, / . Am. Chem. Soc, 1970, 92, 4440; (f) M. Dobler, "lonophores and their Structure", Wiley, New York, 1981; (g) B. T. Kilbourn, J. D. Dunitz, L. A. R. Pioda and W. Simon, J. Mol. Biol, 1967, 30, 559.

3

(a) K. H. Wallhausser, G. Ruber, G. Nessenmann, P. Prave and K. Zept, Arzneimittelforschung, 1964, 14, 356; (b) K. Ando, H. Oishi, S. Hirano, T. Okutani, K. Suzuki, H. Okasaki, M. Sawada and T. Sagawa, J. Antibiot., 1971, 24, 347.

4

(a) S. N. Graven, H. A. Lardy and S. Estrada-0, Biochemistry, 1967, 6, 365; (b) S. N. Graven, H. A. Lardy, D. Johnson and A. Rutter, Biochemistry, 1966, 5, 1729.

5

D. H. Haynes and B. C. Pressman, J. Membr. Biol, 1974, 18, 1.

6

M. Dobler, J. D. Dunitz and B. T. Kilbourn, Helv. Chim. Acta, 1969, 52, 2573.

7

K. Mislow, Croat. Chem. Acta, 1985, 58, 353.

8

P. A. Bartlett, Tetrahedron, 1980, 36, 2. '

9

J.-C. Harmange, B. Figadere, Tetrahedron Asymmetry, 1993, 4, 1711.

10 G. Beck and E. Henseleit, Chem. Ber., 1971,104, 21. 11 H. Gerlach and H. Wetter, Helv. Chim. Acta, 1974, 57, 2306. 12 U. M. Kempe, T. K. Das Gupta, K. Blatt, P. Gygax, D. Felix and A. Eschenmoser, Helv. Chim. Acta, 1972,55,2187. 13 (a) J. Gombos, E. Haslinger, H. Zak and U. Schmidt, Monatsh. Chem., 1975, 106, 219; (b) U. Schmidt, J. Gombos, E. Haslinger and H. Zak, Chem. Ber., 1976, 109, 2628; (c) H. Zak and U. Schmidt, Angew. Chem. Int. Ed. Engl, 1975, 14, 432. 14 M. J. Arco, M. H. Trammell and J. D. White, /. Org. Chem., 1976, 41, 2075. 15 W. C. Still, L. J. MacPherson, T. Harada, J. F. Callahan and A. L. Rheingold, Tetrahedron, 1984, 40, 2275. 16 W. C. Still and I. Galynker, Tetrahedron, 1981, 37, 3981. 17 P. A. Bartlett and K. K. Jemstedt, J. Am. Chem. Soc, 1977, 99, 4829. 18 P. A. Bartlett and K. K. Jemstedt, Tetrahedron Lett., 1980, 21, 1607. 19 P. A. Bartlett, J. D. Meadows, E. G. Brown, A. Morimoto and K. K. Jemstedt, J. Org. Chem., 1982, 47, 4013. 20 P. A. Bartlett, J. D. Meadows and E. Ottow, J. Am. Chem. Soc, 1984, 106, 5304. 21 W. S. Johnson, C. Edington, J. D. Elliott and I. R. Silverman, / . Am. Chem. Soc, 1984, 106, 7588. 22 I. R. Silverman, C. Edington, J. D. Elliott and W. S. Johnson, J. Org. Chem., 1987, 52, 180. 23 K. Narasaka and F.-C. Pai, Tetrahedron,

1984, 40, 2233. See also, K.-M. Chen, G. E.

Hardtmann, K. Prasad, O. Repic and M. J. Shapiro, Tetrahedron Lett., 1987, 28, 155.

267 24 B. Lygo, N. O'Connor and P. R. Wilson, Tetrahedron, 1988, 44, 6881. 25 B. Lygo, Tetrahedron, 1988, 44, 6889. 26 S. W. Baldwin and J. M. Mclver, J. Org. Chem., 1987, 52, 320. 27 P.-F. Deschenaux and A. Jacot-Guillarmod, Helv. Chim. Acta, 1990, 73, 1861. 28 B. H. Kim and J. Y. Lee, Tetrahedron Utt., 1992, 33, 2557. 29 B. H. Kim, J. Y. Lee, K. Kim and D. Whang, Tetrahedron Asymmetry, 1991, 2, 27 and 1359. 30 B. H. Kim and J. Y. Lee, Tetrahedron Utt., 1993, 34, 1609. 31 A. G. M. Barrett and H. G. Sheth, J. Chem. Soc, Chem. Commun., 1982, 170. 32 P. C. B. Page, J. F. Carefull, L. H. Powell and L O. Sutherland, / . Chem. Soc, Chem. Commun., 1985, 822. 33 R. A. Johnson and K. B. Sharpless, in Comprehensive Organic Synthesis, ed. B. M. Trost and I. Fleming, Pergamon, Oxford 1991, Vol. 7, ed. S. V. Ley, Ch. 3.2, pp. 389-436. 34 H. J. Karlsen, T. Katsuki, V. S. Martin and K. B. Sharpless, J. Org. Chem., 1981, 46, 3936. 35 S. Batmangherlich and A. H. Davidson, J. Chem. Soc, Chem. Commun., 1985, 1399. 36 J. K. Whitesell and D. Reynolds, J. Org. Chem., 1983, 48, 3548. 37 T. Honda, H. Ishige, J. Araki, S. Akimoto, K. Hirayama and M. Tsubuki, Tetrahedron, 1992, 48, 79. 38 T. Kametani, K. Kawamura and T. Honda, J. Am. Chem. Soc, 1987, 109, 3010. 39 S. Takano, S. Tomita, M. Takahashi and K. Ogasawara, Synthesis, 1987, 1116. 40 K. Takatori, N. Tanaka, K. Tanaka, M. Kajiwara, Heterocycles, 1993, 36, 1489. 41 G. Frater, U. Muller and W. Gunther, Tetrahedron, 1984, 40, 1269. 42 M. J. C. Buckle, I. Fleming and S. Gil, Tetrahedron Lett., 1992, 33, 4479 and references therein. 43 I. Fleming and N. J. Lawrence, /. Chem. Soc, Perkin Trans. 1, 1992, 3309. 44 R. A. N. C. Crump, I. Fleming, J. H. M. Hill, D. Parker, N. L. Reddy and D. Waterson, J. Chem. Soc, Perkin Trans. 1, 1992, 3277. 45 I. Fleming and P. E. J. Sanderson, Tetrahedron Lett., 1987, 28, 4229. 46 M. Ahmar, C. Duyck and I. Fleming, Pure Appl. Chem., 1994, 66, 2049. 47 I. Fleming, D. Higgins, N. J. Lawrence and A. P. Thomas, J. Chem. Soc, Perkin Trans. I, 1992, 3331. 48 J. P. Vigneron, M. Dhaenens and A. Horeau, Tetrahedron,

1973, 29, 1055. See also V.

Rautenstrauch, Bull. Soc. Chim. Fr., 1994, 131, 515. 49 T. Kogure and E. L. Eliel, J. Org. Chem., 1984, 49, 576. See also, M. M. Midland and J. Gabriel, / . Org. Chem., 1985, 50, 1144; C. S. Poss and S. L. Schreiber, Ace Chem. Res., 1994, 27, 9. 50 U. Ghosh, unpublished work. 51 I. Fleming and S. K. Ghosh, /. Chem. Soc, Chem. Commun., 1994, 2285. 52 I. Fleming and S. K. Ghosh, /. Chem. Soc, Chem. Commun., 1992, 1775. 53 P. D. Theisen and C. H. Heathcock, J. Org. Chem., 1988, 53, 2374 and 1993, 58, 142. 54 I. Fleming and S. K. Ghosh, J. Chem. Soc, Chem. Commun., 1994, 99. 55 I. Fleming, S. Gil, A. K. Sarkar and T. Schmidlin, /. Chem. Soc, Perkin Trans. I, 1992, 3351. 56 I. Fleming and S. K. Ghosh, J. Chem. Soc, Chem. Commun., 1992, 1777.

268

57 K. Yamamoto, T. Kimura and Y. Tomo, Tetrahedron Lett., 1984, 25, 2155. 58 D. A. Evans, K. T. Chapman and E. M. Carreira, /. Am. Chem. Soc, 1988, 110, 3560. 59 T. Mukaiyama, M. Usui and K. Saigo, Chem. Lett., 1976, 49. 60 R. D. Walkup and G. Park, J. Am. Chem. Soc, 1990, 112, 1597. 61 R. D. Walkup and G. Park, Tetrahedron Utt., 1987, 28, 1023. 62 J. Iqbal, A. Pandey and B. P. S. Chauhan, Tetrahedron, 1991, 47, 4143. 63 S. V. Ley, Chem. Ind. (London), 1985, 101. 64 K. M. Sun and B. Fraser-Reid, Can. J. Chem., 1980, 58, 2732. 65 H. Ohnii, G. H. Jones, J. G. Moffatt, M. L. Maddox, A. T. Christensen and S. K. Byram, J. Am. Chem. Soc, 1975, 97, 4602. 66 F. W. Eastwood, K. J. Harrington, J. S. Josan and J. L. Pura, Tetrahedron Lett., 1970, 5223. 67 H. M. R. Hoffmann, K. E. Clemens and R. H. Smithers, J. Am. Chem. Soc, 1972, 94, 3940. See also R. Noyori, S. Makino and H. Takaya, J. Am. Chem. Soc, 1971, 93, 1272. 68 E. LeGoff, / . Org. Chem., 1964, 29, 2048. 69 A. Warm and P. Vogel, Helv. Chim. Acta, 1987, 70, 690. 70 R. E. Ireland and J.-P. Vevert, / . Org. Chem., 1980, 45, 4259. R. E. Ireland and J.-P. Vevert, Can. J. Chem., 1981,59,572. 71 R. E. Ireland, R. H. Mueller and A. K. Willard, / . Am. Chem. Soc, 1976, 98, 2868. 72 H. Gerlach, K. Oertle, A. Thalmann and S. Servi, Helv. Chim. Acta, 1975, 58, 2036. 73 T. Endo, S. Ikenaga and T. Mukaiyama, Bull. Chem. Soc Jpn., 1970, 43, 2632. See also E. J. Corey, K. C. Nicolaou and L. S. Melvin, J. Am. Chem. Soc, 1975, 97, 653. 74 J. Gombos , E. Haslinger, H. Zak and U. Schmidt, Tetrahedron Lett., 1975, 3391. U. Schmidt, J. Gombos, E. Haslinger and H. Zak, Chem. Ber., 1976, 109, 2628. 75 I. Fleming and S. K. Ghosh, J. Chem. Soc, Chem. Commun., 1994, 2287 76 J. Inanaga, K. Hirata, H. Saeki, T. Katsuki and M. Yamaguchi, Bull. Chem. Soc Jpn., 1979, 52, 1989. See also: M. Hikota, Y. Sakurai, K. Horita and O. Yonemitsu, Tetrahedron Lett., 1990, 31, 6367. 77 J. Gombos, E. Haslinger, A. Nikiforov, H. Zak and U. Schmidt, Monatsh. Chem., 1975, 106, 1043. 78 T. Kaiho, S. Masamune and T, Toyoda, J. Org. Chem., 1982, 47, 1612.

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol 18 © 1996 Elsevier Science B.V. All rights reserved.

269

Total Synthesis of Bioactive Natural Spiroethers, Tautomycin and Oscillatoxin D Akitami Ichihara, Hideaki Oikawa and H. Toshima 1. Introduction There are number of natural spiroethers which have cytotoxic and antitumor activities. Two such spiroethers, tautomycin and oscillatoxin Ds, were selected as target molecules of total synthesis, since it is known that tautomycin, which is one of okadaic acid class compounds, is the specific inhibitor of protein phosphatase and it could play an important role as a tumor promotor, while oscillatoxin Ds have antileukemic activity in the L-1210 cell line, and closely related analogs, aplysiatoxins, exhibit activity as tumor promotors. Therefore the study of the structure-activity relationships of these spiroethers can provide not only useful information on phosphorylation and dephosphorylation mechanisms at intraceller levels, but also about effective structural models for antitumor activity. In the total synthesis of tautomycin highly regio- and stereoselective reductions of the spiroketals have been developed in the synthesis of the spiroketal moiety. The construction of the spiroether units in oscillatoxin Ds has been achieved by a biomimetic pathway involving intramolecular aldol condensation and Michael type addition. The total synthesis provides a certain amount of oscillatoxin Ds which are difficult to obtain from natural sources. 2. Total Synthesis of Tautomycin In 1987, Isono and co-workers reported the isolation of tautomycin 1 from the culture filtrate of a soil fungus Streptomyces spiroverticillatus (1), an amorphous white powder exhibiting potent antifungal activity against Sclerotinia sclerotiolum (2) and inducing a morphological change (bleb formation) of human leukemia cells K562 (2). Since tautomycin enhances phosphorylation mediated by protein kinase C (PKC) in vitro (2), 1 was first assumed to be an activator of PKC, as phorbol dibutyrate. However, 1 does not compete with phorbol ester for binding to cells, and dose not activate PKC significantly in vivo (2). Later, Magae et al. (3) and MacKintosh et al. (4) independently found that enhanced phosphorylation caused by 1 due to the inhibition of protein phosphatase (PP), as found with the well-known tumor promoter, okadaic acid 2 (5, 6). Recently, the reversible phosphorylation of proteins has been recognized to be a major mechanism for the intracellular signal transductions in eukaryotic cells (7). The specific inhibitors of PP become a useful tool for studying such intracellular events. Tautomycin inhibited both type-1 and type-2A PP with IC50 of 22-32 nM (3, 4). Fujiki et al. proposed to classify 1 as belonging to the okadaic acid class of compounds which bind to okadaic acid receptors, PPl and PP2A, and inhibit their activities (8). Interestingly, all of those PP inhibitors (8) such as 2, dinophysistoxin-1, calyculin A, microcystins and nodularin are isolated from marine sponge and algae which are not available in large quantity. On the other hand, 1 is the only compound produced by a soil fungus which can be fermented in large scale. It therefore offers the great advantage that semi-synthesis of tautomycin analogues using a late

270

Tautomycin 1

Okadaic acid 2

Tautomycetin 3

intermediate derived from the degradation of natural 1 is possible once the total synthesis of 1 has been achieved. Isono's group also isolated the structurally related tautomycetin 3 from another soil fungus, Streptomyces griseochromogenes (9, 10). This antibiotic also shows antifungal activity and bleb formation in K562 cells (9). The similar biological activities of 3 to those of 1 strongly suggest that 3 is also a specific inhibitor of PP although this was not tested. The major structural difference in these antibiotics is present in the right hand portion of the molecules: tautomycin possesses a spiroketal moiety while this part of tautomycetin is replaced by a simple dienone. This suggests that the lefthalves of these antibiotics are essential for the inhibition of PP but the right-halves are variable. Furthermore, 1 and 2 show different affinities for PPl and PP2A (3, 4, 5) although the molecular size and partial structure of 1 are similar to those of 2. Thus, a systematic study of the structureactivity relationship of 1 would clarify the structural requirements for the inhibitory activity and enable us to design new specific inhibitors for PPl and PP2A. Our interests have therefore led us to develop an efficient synthesis of 1. In addition, the relative and absolute configurations of 1 (Id) have been determined mostly by NOE experiments and conformational analysis using MM2 calculations for the degradation products and their derivatives. Therefore, confirmation of the structure by total synthesis is necessary to secure its conclusion. In this article, we describe the first total synthesis of tautomycin 1(11, 12). 2-1. Chemical Property of Tautomycin Structurally, tautomycin 1 consists of a polyketide chain including a spiroketal moiety and a

271 unique acyl chain containing a dialkylmaleylanhydride unit. The name "tautomycin" came from the facile interconversion between the anhydride and the diacid (la, lb). Because of the steric congestion of the two alkyl substituents, the hydrolyzed diacid tends to regenerate the anhydride ring (eq. 1) (lb). Under nearly neutral conditions (pH 7.3), this moiety exists as two intereconverting anhydride and diacid forms in about 5:4 ratio (la, lb). Similar equilibration was found in an inhibitor of famesyltransferase chaetomellic acids (13) whose dialkylmaleylanhydride moiety is assumed to mimic a pyrophosphate group of famesylpyrophosphate. In analogy, one could speculate that the anhydride unit of 1 mimics the phosphate of a phosphorylated protein. Additionally, it was found that 1 and its derivatives with the anhydride unit were amenable to serious losses during their purifications on silica gel, possibly due to irreversible adsorption (14). This is one of the difficult problems rose during the handling of 1.

pH7.3

OH

O

H 0 2 C ^ ^

(eq.1) In the structure elucidation of 1, extensive degradations have been carried out by Ubukata et al. (Scheme 1) (la, lb). Alkaline hydrolysis of 1 with cesium carbonate at pH 9 gave the anhydride fragment 4 and anhydrodeacyltautomycin 5 which was further degraded with cesium carbonate at pH 10 to the enone 6 and the spiroketal 7 by retro-aldol cleavage. During the alkaline treatment at pH 9, C3 epimerization of 5 occurred (Id). These results indicated that dehydration of the C22 hydroxy group and epimerization at C3 are major problems and that operations at the final stage of the total synthesis must be carried out under neutral or acidic conditions. O

un

u

0

5°C, 70%.

Next, we investigated the protection of the three hydroxy groups in 1 (Scheme 2). Among the various protecting groups attempted, the silyl group was found to be suitable (12). Silylation of 1 with triethylsilyl triflate and t-butyldimethylsilyl triflate gave 11 and 12, respectively. While the former was deprotected with dilute HF to regenerate 1, the latter gave partially deprotected 12a whose C3'-t-butyldimethylsilyl group was resistant to acidic hydrolysis, and other conditions gave only degradation products. These results suggested that the Cs'-stereogenic center is sterically

273 hindered and triethylsilyl group or its equivalent is the protecting group of choice at the final stage. 2-2. Synthetic Plan Investigations on the chemical reactivity of 1 led us to focus on the following issues: (a) timing for the generation of the anhydride unit; (b) C3 epimerization; (c) C22-OH dehydration. For the first issue, we decided to construct the anhydride ring in the final reaction. Since introduction of the anhydride moiety by oxidation at a late stage while keeping other functionalities intact is expected to be difficult, we opted for hydrolysis of a differently protected maleyl diester, such as 13. The second problem can be solved by protection of the C2-carbonyl as an olefin (i.e.14) which is stable to various transformations and which can be converted to the corresponding methyl ketone by a Wacker type oxidation under neutral conditions. Finally, the most difficult third problem can be solved by direct assemblage of the C2J-C22 bond using two large subunits.

Tautomycin 1

DEIPSO s

V

O II

K

EtOzC

13

14 Scheme 3

Incorporating the considerations as shown above, retrosynthetic disconnection of the carbon backbone at the C21-C22 bond divides the target into two subunits, named the Left-wing 13 and the Right-wing 14 (Scheme 3). This rather bald disconnection also eliminates the difficulty of the O24acylation of the main chain and anhydride segment since facile hemiketal formation between C24hydroxy group and C20-carbonyl is expected to become a problem. The stereocontrolled aldol coupling of two key subunits is the key issue (synthetic challenge) of our synthesis since there is no established method for the stereochemically controlled coupling of these highly oxygenated segments. The Left-wing is further divided into dialkylmaleylanhydride segment 15 and the C22-C26 segment 16 (Scheme 4). On the other hand, retro-synthetically, the right wing is sectioned into C19-21 Cs-unit DEIPSO

O

'BUO2C, Et02C 13

15

Scheme 4

274

and C1-C18 spiroketal 17 which is further partitioned into the sulfone 18 and the aldehyde 19 (Scheme 5). The spiroketal 17 is synthetically equivalent to the degradation product 7. 0

^

f

OTES

0

1

"1

H

^Â^ 1

V^'%

^=^

^••H 14 17

{I SOaPh > r 5 OMOM : H

+

OHC

18

H RO^s

19 R = p-nitrobenzoyi

Scheme 5

For the right half of 1, we are aware that oxygenated carbons are present at exactly every five carbons and that the stereogenic centers are located near the oxygenated carbons as shown in partial structure A (Scheme 5). For synthesizing both segments 18 and 19, we planed to develop a new method for spiroketal reduction. At the beginning of our work, the stereochemistry at C15 was not settled; the possibility still remained to apply the method for the preparation of the Cu-Cig subunit. However, considering the eventually established structure of 1 possessing

\'h,\A-syn-\A,\'b-syn

stereochemistry which can be hardly prepared by our spiroketal reduction strategy, we decided to use another method for the synthesis of this segment. 2-3. Regie- and Diastereoselectivity of Spiroketal Reduction 2-3'L

Strategy of Spiroketal

Reduction

The chemistry of spiroketals, especially l,7-dioxaspiro[5,5]undecanes, is well-studied and reviewed in the hterature (15,16). Generally, the ratio of isomers of spiroketals may be controlled by several stabilizing factors such as stereoelectronic effects and 1,3-diaxial interaction of the substituents. Utilizing well-designed spiroketals, one can selectively prepare the most stable isomer by thermodynamic equilibration. Based on this idea, we planned to prepare a spiroketal represented by II which possesses several substituents with established (Cy and C5') and unestablished (€«) configurations. The spiroketal center and its a-position in II can isomerize to the more stable forms by steric effects caused by the established stereocenters on the ring. The prepared most stable isomer

275 is then subjected to reduction which gives another stereogenec center (Scheme 6). If the reduction proceeds regio- and diastereoselectively, this two-step process (thermodynamic equilibrationreduction) can be regarded as a formal remote chiral transfer. Using spiroketals as a template for manipulating functional groups on a ring, several excellent studies have been carried out in the total synthesis of natural products (17). To our knowledge, however, studies on the spiroketal reduction have been limited (18). In order to achieve this type of chiral transfer, we studied the reduction of spiroketals (19).

,''

\ Ri OH

H^

R2

reduction

illA

Mfol

H

IIIB

H

„

IIIC

HID

Scheme 6

reduction Y'

^ 20 21 22 23

Ri Ri Ri Ri

= H, R2 = H = CH3, R2 = H = CH3, R2 =CH20TBDPS = CH3, R2 = CHgOBn

24A 25A 26A 27A

Ri = H, R2 = H 24B R^ = H, Rj = H Ri = CH3, R2 = H 25B Ri = CH3, Rj = H R-, = CH3. R2 =CH20TBDPS 26B Ri = CH3, Rg =CH20TBDPS Ri = CH3, R2 = CH20Bn 27B R^ = CH3, R2 = CH20Bn

Scheme 7

From the examination of molecular models, we anticipated that the Ca methyl group of spiroketals 20, 21, 22 and 23 is large enough to interfere with the coordination of aluminum reagent

276 and Lewis acid at Oe' on these spiroketals. Since the reaction with alane reducing agent usually proceeds with retention of configuration (20), we expected type-A reduction products (24A, 25A, 26A and 27A) to be predominant in the reactions of DIBAH (eq 2). On the other hand, the several proposed mechanisms of silane-Lewis acid (SI-LA) reactions (21) suggested that we could change the stereochemical course by selecting a proper Lewis acid and an appropriate design of the substrates. For studying these reductions, the isomerically pure spiroketals 20, 21, 22 and 23 were synthesized (22,23) utilizing thermodynamic equilibration and subjected to reductions under several conditions (Scheme 7). The experimental results are summarized in Table 1. 2'3'2,

Spiroketal

Reduction

with DIBAH

As we expected, all cases (entries 1-4) in DIBAH reduction yielded type-A products predominantly with retention of configuration at the spiroketal center. In these reactions, selective cleavage of the C-Oe bond suggests that coordination of aluminum reagent predominantly occurred at the less crowded oxygen, Oe. As Yamamoto et al. proposed previously (20), the stereocontrol of DIBAH reduction may originate from the tight ion-pair complex, such as IM-1 (eq. 2), leading to rapid hydride transfer from the aluminum reagent to the oxocarbenium ion with retention. Yields of DIBAH reductions are normally satisfactory except for the sterically hindered ketal 22 (entry 3) which was less reactive and was converted to non-reduced products; the enol ethers 28, 29 were formed as major by-products under the reduction conditions. In the reduction of 23 (entry 4), the formation of a small amount of 27B suggested that chelation controlled reduction occurred to some extent (i.e. eq. 4) although such a reaction path was limited. Table 1. Reduction of a-methylspiroketals Entry

Spiroketal

Reagent

Yield (%)

DIBAH(a) DIBAH(a)

68

24A:24B:others

(94 : 2 : 4)(0

62 25°C, 94% (3 steps); (f) n-BuLi, 2-methyl-6-valerolactone, Et20-hexane, -65-^25°C, 70%; (g) TMSBr, CH2CI2, -30-»3°C, 89%; (h) Raney Ni (W-2), EtOH, reflux, 88%; (i) EtgSIH, SnCU, CHgClg, -78-^-60''C; AcOH, THF-HjO; (j) AcgO. Py, DMAP, CH2CI2, 78% {2 steps); (k) TBAF, THF; (I) MsCI, Py, DMAP, CHgClg; (m) LiBr, DMF, 70°C, 88% (3 steps); (n) Zn, EtOH-HgO, reflux, 87%; (o) p-nitrobenzoic acid, PPhg, DEAD, CQH^; (p) NaH, MeOH, 5°C, 54% (2 steps); (q) DMSO, (C0CI)2, CH2CI2, -78°C; EtgN, -78-^0°C, 88%; (r) K2CO3, EtOH, quant.; (s) AgsCOg-Celite, CgHe, reflux.

281 borane afforded the 5>'n-adduct 41 in high diastereo- and enantioselectivity. The alcohol 41 was protected as a methoxymethyl ether instead of an ethoxyethyl ether which was not stable under hydroboration conditions. Essentially the same procedure as shown above was used for the conversion of 42 to the sulfone 43. The coupling of lithiated 43 with 2-methyl-5-valerolactone was affected under similar conditions as those of the racemate to furnish the adduct 45 in 61% yield. In order to facilitate desulfurization, the sulfoxide 44 was also coupled with the lactone but this condensation under several conditions always gave low yields. The subsequent three-step conversion of 45 was carried out by treatment with bromotrimethylsilane to afford 39 as a single isomer in 88% yield (31). Deprotection of MOM group and spiroketalization proceeded below -30°C and then equilibration occurred at 0°C to room temperature. After desulfurization with a large excess of Raney Ni (W-2), the resultant spiroketal 22 was reduced by our newly developed procedure. Reduction of 22 with triethylsilane and tin(IV) chloride at -78 to -60°C was followed by acid hydrolysis of triethylsilyl ether to give chiral 26A in 98% yield. Completion of the synthesis of the Ci-Cio segment requires opening of the tetrahydropyran ring and inversion of the C5 configuration (Scheme 9). Acetylation of 26A was followed by desilylation with tetra-n-butylammonium fluoride to give the alcohol 47 which was converted to the bromide 48 by mesylation and substitution with lithium bromide in 88% yield. Reductive ring opening with zinc and acetic acid furnished 49 which was followed by Mitsunobu inversion (32) with p-nitrobenzoic acid and diethylazodicarboxylate and triphenylphosphine) to afford the diester 50. Selective hydrolysis of the ester 50 and Swern oxidation provided the aldehyde 19 in 42% overall yield in 4 steps. For the examination of coupling conditions, the lactones 52a and 52b were also prepared from the acetates 49 and 51, respectively. 2-4-3. Synthesis of Cij-Cjg

segment

We then moved to the synthesis of the Cn-Cig segment (Scheme 10) (29). The alcohol 53 (33) was converted to 55 by a standard C2 homologation procedure in 77% overall yield in 6 steps. Swern oxidation of 55 gave the aldehyde 56 which was submitted to Lewis acid catalyzed crotylstannylation (34). This reaction provided the expected eryr/zro-adducts 57a and 57b in a 3:1 ratio which were separated by medium-pressure silica gel chromatography. Protection of 57a as a MOM ether was followed by hydroboration with 9-borabicyclo[3.3.1]nonane to afford the alcohol 58 which was further converted to the sulfone 18 by the procedure shown above. In order to examine the coupling of the Ci-Cio and Cn-Cig segments and subsequent spiroketal formation, a model study was employed. The carbanion derived from 18 on treatment with n-butyl lithium was condensed with 5-valerolactone to give the adduct 59 in 59% yield. Desulfurization of the P-ketosulfone 59 with sodium amalgam gave 60 in modest yield. Spiroketal formation was then carried out. Treatment with bromotrimethylsilane gave the best result to furnish 61 as a single isomer in quantitative yield. At this stage, we could confirm the C13 and C14 configurations, which were installed by crotylstannation, using NMR data including NOE experiments. In addition, a good correlation of the ^H-NMR data of 61 with that of tautomycin 1 provided us further confirmation for assignment of the C15 configuration which was determined by

282 MM2 calculations of several C14-C15 rotamers in 1 and its C15 epimer (Id).

55 R = CH2OH 56 R = CHO

54

53 OH

I

• other isomers

57a

66:22

57b

OH

57a

, SOaPh

k, I

I.J

TBDPSO

58

'OMOM

18

X OMOM

59 X = SOaPh 60X = H

61

Scheme 10 (a) DMSO, (C0CI)2, CH2CI2, -70°C; EtgN, -70-^25X; (b) (EtO)2P(0)CH2C02Et, NaH. THF, -78-^25X, 98% (2 steps); (c) H2, Pd/C, EtOAc; (d) LiAIH4, Et20, 3-^25°C; (e) TBDPSCI, imidazole, DMF; (f) p-TsOH, MeOH, 79% (4 steps); (g) DMSO, (C0CI)2, CH2CI2. -78*C, EtgN, .78-^25°C; (h) trl-n-butylcrotylstannane, BF3*Et20, CH2CI2, -86->0°C, 92% (2 steps); (i) MOMCI, i-Pr2EtN, CH2CI2, quant.; (j) 9-BBN, THF, 3->25''C; 3M NaOH, H2O2, 3-425°C; (k) (PhS)2. n-BugP, Py; (I) mCPBA, NaHCOg, CH2CI2, 3->25'='C, 82% (3 steps); (m) n-BuLi, 5-valerolactone, Et20-liexane, -78->25°C, 59%; (n) Na(Hg), K2HPO4, MeOH, -20^25''C, 30%; (0) TMSBr, CH2CI2, -70^25°C, quant.

2-4-4. Coupling synthesis

of

of Ci'Cio

segment with Cn-Cig

segment and completion

of the

Right-wing

Coupling of the Ci-Cio segment with the Cn-Cig segment was next examined (Scheme 11). In the condensations of lactones 52a and 52b with lithiated sulfone derived from 18, no or low conversion to the product was observed in both solvent systems, non-polar toluene and ether-nhexane, which gave a satisfactory result in our similar coupling of 23 and 2-methyl-5-valerolactone. On the other hand, the reaction of aldehyde 19 with the sulfone carbanion proceeded smoothly to give adduct 62 which was inmiediately converted by Swem oxidation to 63 in 82% overall yield. The next reductive desulfonylation (36) proved to be difficult. Reduction of 63 with either aluminum amalgam (37) or n-tributyltin hydride (38) only gave uncharacterized reduction products in which the more sensitive nitro group were reduced. Finally, this problem was solved by use of samarium diiodide (39) affording the desired product 64 in 51 % yield. In this product, the nitro group was also reduced into the hydroxyamine (40). After hydrolysis of the benzoate group, cyclization to the spiroketal 65 was effected with bromotrimethylsilane as described above in 72% yield. Spiroketal 65 was then converted to the degradation product 5 in order to confirm the proposed structure of 5. The

283 silyl group of 65 was removed with tetra-n-butylammonium fluoride to give the alcohol 66 and subsequent oxidation with Dess-Martin periodinane (41) afforded aldehyde 17 in 80% yield. Wacker-type oxidation (42) of the alcohol 66 proceeded cleanly to give 67 which was finally converted to 5 by Swem oxidation in 94% overall yield. The synthetic material was identical with 5 derived from natural tautomycin in all respects. Thus, we unambiguously established the C3-C15 absolute configuration in tautomycin. , SOgPh

a, b TBDPSO

18

62 R = H, OH 63 R = 0

d, e, f, g

64 R = p-hydroxyaminobenzoyi

65 R = CH2OTBDPS 66 R = CH2OH 17 R = CHO

-^^

66

OHO,

67

Scheme 11 (a) n-BuLi, 19, Et20-hexane, -78-425°C; (b) DMSO, (C0CI)2, CH2CI2, -78X; EtaN, -78-> 0°C, 82% (2 steps); (c) Smig, THF-MeOH, -78°C, 51%; (d) K2CO3, MeOH, 60°C; (e) TMSBr, CH2CI2, -30->3X, 72% (2 steps); (f) TBAF. THF; (g) Dess-Martin periodinane, Py, CHjClg, 80% (2 steps); (h) O2, PdCl2, CuCI, DMF-H2O; (i) DMSO, (C0CI)2, CH2CI2, -78°C; EtgN, -78^0°C, 94% (2 steps).

0Ti(0'Pr)3

TBSO

17

'

(eq. 6)

68a

+

a I

OMgBr

TMSO

'

68b

For the introduction of the Cig and C19 stereocenters to the aldehyde 17, we first tested Heathcock's asymmetric anti-selective aldol reaction (eq. 6) (43). We anticipated that the stereocenters in 17 would have little effect on diastereoselectivity since the two methylene groups are

284 inserted between C15 and Cig. However, the reaction of 17 with the enolates 68a and 68b proceeded non-selectively to afford 69 as 3 isomeric mixtures in a 2:2:1 ratio. This is probably due to the other oxygen functionalities in 17 causing disorder of metal chelation structure in the transition states. Crotylboration (31,44) provided us a nice solution to this problem (Scheme 12). The reaction of 17 with Brown's (-)-(E)-crotyldiisopinocampheylborane afforded adducts 70a in very high diastereoselection but in modest yield ( ^

OH

O

> ^ ^ - ^

OR2

> ^ ^

n-CsH

118b R^ = MPM, Rg = TES 119b R, = Bz ^2 = TES 120b Ri = MPM, R2 = BOM

MP OTES n-CgHic

Scheme 20

121

Table 2. Model aldoI reaction of a,p-diaIkoxyaldehydes and methyl ketones entry aldehyde methyl ketone

conditions

syn/anti ratio

yield (%)

1 2

114 114

116 116

LHMDS, THF, -78°C LHMDS, ClTi(0'Pr)3, THF, -78->-30°C

0:100 0:100

33 45

3 4

114

116

LHMDS, MgBr2«Et20, THF, -78X

0:100

114

116

LHMDS, ZnCl2, THF, -78°C

22:42

47 64

5

114

116

TiCU, Tr2NEt, CH2CI2, -78°C

1:1

80

6

114

116

9-BBN triflate, 'Pr2NEt, CH2CI2, -78°C

33:37

70

7

115

116

LHMDS, THF, -78°C

80%) 15a

123

'BUO2C.

122a

EtOzC

124 (72%)

122b

a, b 125 (82%, 2 steps)

Scheme 22

(a) O2, PdCia, CuCI, DMF-HgO; (b) TESOTf, 2,6-lutidine, CH2CI2.

294 122a afforded the methyl ketone 124 (42). While deprotection of the t-butyl group with trifluoroacetic acid or aqueous HF was unsuccessful, even under Evans's conditions extensive decomposition occurred. Eventually, this problem was solved by use of milder Lewis acid triethylsilyl triflate. Thus, deprotection of the t-butyl group with triethylsilyl triflate and 2,6-lutidine and concomitant ring closure gave 1. In this reaction, the use of a limited amount of 2,6-lutidine (0.4 eq for triethylsilyl triflate) can avoid the silylation of other hydroxy groups. Our synthetic sample was identical in all respects with natural tautomycin. Under essntially the same conditions, 122b was converted to 22/?-epitautomycin 125 in 82% yield. In conclusion, the first total synthesis of tautomycin has been achieved via an efficient aldol coupling of two large subunits. Our synthetic route would provide an efficient way to prepare various analogues of 1 for biological evaluation. 3. Total Syntheses of Oscillatoxin D and 30-Methyloscillatoxin D Oscillatoxin D 126a and 30-methyloscillatoxin D 126b (75) are natural products derived from p-polyketides. They occur with aplysiatoxins 127a-127c (76), potent tumor promoters (77), in the marine blue-green algae belonging to the Oscillatoriaceae: Lyngbya majuscula,

Schizothrix

calcicola, and Oscillatoria nigroviridis. Their structures were mainly determined by spectral studies using EI-MS, IR, ^H-NMR, NOE, and CD spectra (75). They consist of a main spiroether moiety, corresponding to 1oxaspiro[5. 5]undec-4-ene-8-one ring system, (including seven asymmetric centers and possessing a (Z)-olefin, a mera-substituted phenol, and a P-ketocarboxylic acid functions) and a p-hydroxy-ylactone moiety. Oscillatoxin Ds are present as esters of both moieties. On the other hand, aplysiatoxins are present as macro bis-lactones including a spiroacetal. We are interested in the relationship between oscillatoxins and aplysiatoxins from a biosynthetic point of view.

OMe X

126a: Oscillatoxin D (R=H) 126b: 30-Methyloscillatoxin D (R=Me)

Fig. 1

127a: Aplysiatoxin (R=Me, X=Br, Y=OH) 127b: Debromoplysiatoxin (R=Me, X=H, Y=OH) 127c: Oscillatoxin A (R=X=H, Y=OH) 127d: 3-Deoxydebromoaplysiatoxin (R=Me, X=Y=H)

Structures of Oscillatoxin Ds and Aplysiato ins

The polyfunctional structures of aplysiatoxins and oscillatoxin Ds have received much attention as attractive synthetic targets and several synthetic studies have recently been reported (78). In connection with a series of our synthetic studies on aplysiatoxins, the total synthesis of 3-

295 deoxydebromoaplysiatoxin 127d has been completed (79). The synthetic 3-deoxy analogue 127d has been known to exhibit conGiparable activity as a tumor promoter (80). Debromoaplysiatoxin 127b and oscillatoxin A 127c show activity against P-388 leukemia in vivo (81). Oscillatoxin D 126a also displays antileukemic activity against the L-1210 cell line (82), however, its biological activity has been rarely examined because of the limitation of the natural sample for bioassay. Therefore, it is important to supply oscillatoxin Ds synthetically for biological investigations. In this article, we describe the first total syntheses of oscillatoxin D and 30-methyloscillatoxin D. The construction of the spiroether moiety has been achieved by biomimetic intramolecular aldol condensation and intramolecular Michael-type addition as key steps. Furthermore, some analogues of oscillatoxin D, which play an important role on the structure-activity relationship, can be prepared by our synthetic route. 3-1. Retrosynthetic Analysis

1. (i-elimination

^^ 2. Y-lactonization 27^0 OH

[A]

3. aldol-condensation 4. dehydration

Aplysiatoxins Oscillatoxin Ds

5. Michael-type addition

Q. Segment A -I-

Segment B + Segment C

"V

OP

•• ^

1 MeOO

Segment A

I

OMe

kJI QQ^ |

CHO

MeO

Segment B

Segment C

A Possible Transformation of Aplysiatoxins into Oscillatoxin Ds Retrosynthesis Scheme 23

We adopted a possible transformation of aplysiatoxins into oscillatoxin Ds in a retrosynthetic analysis (Scheme 23). Although aplysiatoxins exist as spiroacetals, they are equivalent to a

296 diketoalcohol [A] opening the acetal. The intermediate [A] may be converted into the intermediate [B] by two steps: p-elimination of the C27- carboxyUc acid to form the (Z)-olefin (step 1) and ylactonization between this carboxylic acid and the Cso-hydroxyl group (step 2). Aldol condensation between the C2 active methylene and the C7 ketone (step 3) and subsequent dehydration (step 4) may provide the intermediate [C]. Intramolecular Michael-type addition of the Cn-hydroxyl group to the C7 position (step 5) may complete the transfomation of aplysiatoxins into oscillatoxinDs. In the spiroetherification, it is important to control not only the configuration of the spiro center (C7) but also the conformation of the spiroether ring system, from a synthetic point of view. We expected to solve these problems by utilizing the thermodynamic equilibrium of the spiroether. Therefore, a practical retrosynthetic analysis was made for the intermediate [B], which was divided into three main segments, A (C8-C23), B (C1-C7, C24, C25, C26), and C (C27-C30 or C31). We chose an acetylenic compound as segment A, because its acetylide would act as a good nucleophile for an aldehyde (segment B), and the acetylenic bond would be selectively reduced to a (Z)-olefin at an appropriate step after coupling. The chiral pool method was applied during the synthesis of each segment. 3-2. Syntheses of Three Optically Active Segments 3-2-Jf. Synthesis

of Segment A

The C8-C23 segment of aplysiatoxins was synthesized as an alkyl iodide corresponding to 137 from D-glucose and 3-hydroxyacetophenone in the total synthesis of 3-deoxydebromoaplysiatoxin (79). The C1-C6 carbon unit of D-glucose was incorporated in the Cg-Cn skeleton of aplysiatoxins with sequential four asymmetric centers. The asymmetry of the C15 benzylic position was introduced by a diastereoselective reduction using a chiral reducing reagent. We modified this route and an acetylenic function was newly introduced to segment A. The first task involved the introduction of two methyl groups at the C3 and C5 positions of Dglucose with inversion of those configurations. A known tosylate 128 was readily synthesized in 6 steps from D-glucose (Scheme 24) (83) and possessed the required asynmietry at the C3 position. Treatment of 128 with sodium benzyl alcoholate gave the benzyl ether via an epoxide as an intermediate. The resulting secondary hydroxy 1 group was substituted with inversion by a chlorine atom to provide a chloro benzyl ether 129 in two steps. Acidic hydrolysis of 129 followed by sodium borohydride reduction gave an acycHc triol, which was again protected with 2, 2dimethoxypropane to give the 1, 3-acetonide 130 and the 1, 2-acetonide 131 in 50% and 46% yields, respectively. The desired 131 was further obtained by the acidic isomerization of 130 in almost the same yield and ratio as that obtained by the protection of the triol. Treatment of 131 with sodium methoxide gave an epoxide 132 in quantitative yield. Methylation of 132 with lithium dimethyl cuprate gave an alcohol opened at the C5 position as the main product in 11:1 regioselectivity. The undesired regioisomer was separated off by a silica gel column chromatography. Protection of the secondary hydroxy 1 group as a silyl ether followed by deprotection of the primary benzyl ether by hydrogenolysis gave an alcohol 133. At this stage, the first task was completed.

297

D-glucose

6 steps

a, b, c TsO

>^ I ' 0 '

BnO X t ^ O '

128

°)C

129

ci,e,f

134

OBn

135

OBn

Scheme 24 (a) BnONa / BnOH-THF; (b) MsCl, pyr. / CH2CI2; (c) n-Bu4NCl / PhH, 89%, 3 steps; (d) Amberlite IR-120 (H"^) / dioxane-H20; (e) NaBH4 / MeOH-H20, 88%, 2 steps; (f) 2,2-dimethoxypropane, p-TsOH I acetone , 130: 50%, 131: 46%; (g)p-TsOH / acetone; (h) NaOMe / MeOH, 100%; (i) MeoCuLi / Et20; (j) TBDMSOTf, EtsN / CH2CI2; (k) H2, 10 % Pd-C / MeOH, 81 %, 3 steps; (1) (COCl)2, DMSO, EtsN / CH2CI2, (m) Ph3P=CHC0N(0Me)Me / CH2CI2; (n) H2, 5 % Pd-C / EtOAc; (0) LiC6H4(w-OBn) / THF, 86%, 4 steps; (p) (-)-DIPCl / THE; (q) Mel, NaH / THE; (r) n-BuNF / THE, 74%, 3 steps.

The second task is the elongation of the carbon chain including the incorporation of an aromatic ring and introduction of asymmetry at the benzylic position. Swem oxidation of 133 gave an aldehyde, which was subjected to Wittig reaction with N-methoxy-N-methyl-2(triphenylphosphoranylidene) acetate (84). After catalytic hydrogenation, a saturated amide was obtained, which was treated with a 3-benzyloxyphenyl lithium (generated from the corresponding bromide with n-butyl lithium) to provide a ketone 134. For the diastereoselective reduction of the benzylic ketone 134, an asymmetric reducing reagent, Brown's (-)-diisopinocampheylchloroborane (this reagent provides 5 configuration) (85) was applied to give a sole diastereomer. Using ^HNMR (400MHz) spectroscopy, it was compared with the diastereomixture at the C15 position (1:1) prepared by the other method. The purity could be determined based on the methyl signals of the TBDMS group and the reduced product showed only two singlet methyl signals (86). After methylation of the resulted secondary hydroxyl group, in order to confirm the absolute stereochemistry at the C15 position, the benyl ether was converted to the corresponding phenol. The CD curve ([6]260 +404° in EtOH) of the phenol was quite similar to that of debromoaplysiatoxin (127b), thereby confirming the absolute stereochemistry. Deprotection of the

298 silyl ether gave a secondary alcohol 135 which isomerized to a primary alcohol 136 under the same acidic conditions (Scheme 25) as described in the recycle of 130 to 131. The third task was the introduction of an acetylenic function and the favorable choice of a protective group at the Ci i-hydroxyl group. Treatment of 136 with triphenylphosphine in refluxing carbon tetrachloride gave the chloride 137 in quantitative yield. Then 137 was subjected to the base induced elimination (87) to provide an acetylenic alcohol 138 in 72% yield. The Cn-hydroxyl group, which was involved in the construction of a spiroether, was protected as a silyl (segment Al), a methoxymethyl (segment A2), and a tetrahydropyranyl (segment A3) ethers. In practice, the protective group was exchanged successively for some synthetic operations as described hereinafter. In this way, the synthesis of segment A was completed (88).

O

OH

OMe

0

135

OBn

137

OBn

0

OMe

138

OBn

OMe Segment Al: P=TBDMS Segment A2: P=MOM Segment A3: P=THP

d,e, f OBn

Scheme 25 (a) p-TsOH / acetone, 135: 30%, 136: 70%; (b) PhsP / CCI4, 99%; (c) LDA / THF, 72%; (d) TBDMSOTf, Et3N / CH2CI2, 100%; (e) MOMCl, i-PrjNEt / CH2CI2, 87%, (f) DHP, PPTS / CH2CI2, 97%.

3'2'2.

Synthesis of Segment B The C4 position is the only asynmietric center in segment B. We used methyl (5)-3-hydroxy-

2-methylpropionate as the starting material, which was readily converted into the known alcohol 139 in two steps (Scheme 26) (89). Tosylation of 139 with tosyl chloride and pyridine followed by substitution with sodium iodide gave an alkyl iodide 140. Treatment with iodine, triphenylphosphine, and imidazole in benzene also gave 140 directly in good yield. The lithium enolate of ethyl isobutylate, as the four carbons unit, was alkylated with 140 to provide 141 in quantitative yield. Deprotection of the benzyloxymethyl ether by hydrogenolysis, and Swern oxidation of the resulting alcohol, followed by thioacetalization with 1, 3-propanedithiol gave a 1, 3-dithiane derivative 142. The ester group of 142 was reduced with lithium aluminum hydride to provide an alcohol, which was protected as a r-butyldiphenylsilyl (TBDPS) ether 143. Further

299 elongation by two carbons unit, including the generation of a masked aldehyde, was achieved by treatment of the lithiated 143 with l-bromo-2, 2-dimethoxyethane to afford 144 in quantitative yield. Deprotection of the TBDPS ether followed by oxidation using an activated dimethylsulfoxide method gave an aldehyde, segment B. This segment has one free aldehyde group required for coupling with segment A and two masked carbonyl synthons which can be employed to generate an activated methylene system for intramolecular aldol condensation. Methyl (5)-3-hydroxy2-methylpropionate

a, b

c,d (ore) 140

139

g, h, i

BOMi

COOEt 141

j,k

COOEt

142

OTBDMS

OTBDMS

143

Scheme 26 (a) BOMCl, /-Pr2NEt / CH2CI2, 100%; (b) LiAlH4/ Et20, 93%; (c) TsCl, pyr. / CH2CI2; (d) Nal / acetone, 86%, 2 steps; (e) I2, PhsP, imidazole / benzene, 82%; (f) LDA, ethyl isobutyrate / THF, 100%; (g) H2, Pd-black / EtOH; (h) (C0C1)2, DMSO, EtsN / CH2CI2; (i) 1,3-propanedithiol, BF30Et2 / CH2CI2, 85%, 3 steps; (j) LiAlH4, / Et20; (k) TBDPSCl, imidazole / DMF, 91%, 2 steps; (1) t-BuLi / HMPATHF, then l-bromo-2,2-dimethoxyethane, 99%; (m) n-Bu4NF/THF; (n) SOspyr., DMSO, Et3N/CH2Cl2, 86%, 2 steps.

3-2-5. Synthesis

of Segment C

p-Hydroxy-Y-lactone moieties, segment CI and segment C2, are known compounds whose syntheses have been reported. Therefore, we followed the reported methods and could readily synthesize both segment Cs. Segment CI was synthesized from (/?)-malic acid as a starting material in the following three steps (Scheme 27): (I) diesterification with acetyl chloride in methanol, (II) selective reduction of a-hydroxy ester with boran dimethylsulfide complex and sodium borohydride in THF, and (III) acid catalyzed lactonization of 145 (90). Segment C2 was synthesized from methyl (/?)-lactate as a starting material in the following three steps: (I) acetylation with acetic anhydride and pyridine, (II) intramolecular Claisen condensation of 146 using lithium bis-(trimethylsilyl) amide, and (III) hydrogenation of the enolic olefin 147 in the presence of rhodium catalyst. The required diastereomer (segment C2) was

300 produced in 9:1 diastereoselectivity and was readily separated by a silica gel column chromatography (91).

(/?)-Malic acid

DH

OH

a, b

J:S

145

Segment CI

OH

HoC COOMe

Methyl (/?)-lactate

-

0

^

146

0

.OH

^

147

Segment C2

Scheme 27 (a) AcCl / MeOH; (b) BHgSMes, NaBH4 / THF; (c) TFA / CH2CI2; (d) AC2O / pyridine; (e) LiN(TMS)2 / THF; (f) H2, 5% Rh-C / EtOAc.

3-3. Construction of Spiroethers Ring System 3'3'1,

Synthesis

of Cyclohexenone

Derivatives

by Intramolecular

Aldol

Condensation-Dehydration At first, segment Al (P=TBDMS) was applied for couping with segment B prior to subsequent manipulation. The lithium acetylide of segment Al generated with «-butyl lithium was successfully coupled with segment B to give the diastereomeric alcohol 148a (97% yield, diastereoselectivity: ca. 1:1), which possessed the required C1-C26 carbon skeleton of oscillatoxin Ds (Scheme 28). Oxidation of the secondary hydroxyl group followed by deprotection of the thioacetal with N-chlorosuccinimide gave a diketone 149a (68% yield, 2 steps). When the ^H-NMR of 149a was observed in chloroform-d, there was a slight singlet signal assigned to an aldehyde in addition to its own signals. We considered that the aldehyde signal might be due to the formation of a cyclohexenone derivative 150a. This indicates that intramolecular aldol condensation-dehydration takes place after hydrolysis of the dimethylacetal in a weak acidic medium. Therefore, we examined various acidic conditions in order to hydrolyze the dimethylacetal and further to cyclize to 150a. For example, transacetalization with a catalytic amount of /?-toluenesulfonic acid in acetone gave only a mixture of enols derived from the resulting 1, 3-dicarbonyl system. Fortunately, we could find the optimum condition, which was treating with 50% aq. trifluoroacetic acid / chloroform (1:5) (92), to give 150a in 67% yield. A deprotected product 150d was also obtained in 29% yield. This intramolecular cyclization is regarded as a biomimetic reaction corresponding to the conversion from the intermediate [B] to [C] in Scheme 23. Oxidation of the aldehyde 150a with sodium chlorite to the corresponding carboxylic acid, followed by treatment with diazomethane, provided a stable methyl ester. Partial hydrogenation of the acetylenic linkage in the presence of Lindlar catalyst in ethyl acetate then gave a (Z)-olefin 151a in 80% yield (3 steps from 150a). The TBDMS group was found to be an effective protective group through this sequence of reactions, however, it resisted deprotection since the C n hydroxyl group is sterically hindered by two neighboring methyl

301 groups, and furthermore, the TBDMS group itself is a rather bulky group. The approach of reagents would therefore be restricted, principally, due to steric hindrance of substrate 151a. We next chose an acetal-type protective group, among which methoxymethyl ether is the smallest one and can be deprotected under acidic conditons. The lithiated segment A2 (P=MOM) was coupled with segment B to give 148b in a 93% yield. Oxidation of 148b, followed by deprotection of the 1, 3-dithiane proceeded without difficulty to give 149b in 74% yield. Deprotection of the dimethylacetal and subsequent intramolecular aldol condensation-dehydration was achieved under the same acidic condition as the cyclization of 150a. The desired cyclohexenone 150b and the deprotected one 150d were hence obtained in 56% and 27% yields, respectively. The MOM group compared with the TBDMS group did not influence the yield appreciably until this step. Conversion into a methyl ester, followed by introduction of the (Z)olefin provided 151b in 49% yield (3 steps). The yields in these sequential steps were relatively lower than in the TBDMS series. Deprotection of the MOM ether and subsequent intramolecular Michael-type addition from 151b was possible in one pot.

CHO

Segment A1: P=TBDMS QBH Segment A2: P=MOM Segment A3: P=THP

Segment B

OBn MeO

U °" ^^

OP

OMe

148a: P=TBDMS (97%) 148b: P=MOM (93%) 148c: P=THP (93%)

OBn

149a: P=TBDMS (68%) 149b: P=MOM (74%) 149c: P=THP(72%) OBn

150a: P=TBDMS (67%) + 150d: P=H (29%) 150b: P=MOM(56%) + 150d(27%) 150c: P=THP (8%) + 150d (78%)

OBn 151a: P=TBDMS (80%, 3 steps from 150a) 151b: P=MOM (49%, 3 steps from 150b) 151d: P=H (54%, 3 steps from 150d)

Scheme 28 (a) «-BuLi / THF then Segment B; (b) SOspyr., DMSO, EtsN / CH2CI2; (c) NCS, AgNO^ I 10% aq. CH3CN; (d) 50% aq. TFA / CHCI3 (1:5); (e) NaC102, NaH2P04, 2-methyl-2-butene /t-BuGH - H2O (4:1); (f) CH2N2 / Et20; (g) H2, Lindlar cat. / EtOAc

The deprotected product 150d was also oxidized to the corresponding carboxylic acid without affecting its hydroxy! group. Treatment with diazomethane, followed by partial hydrogenation.

302

gave ISld in moderate yield. From this result, the dianion of the free alcohol 138 was used for coupling with segment B to give a diol. Selective oxidation of the propargyl alcohol with activated manganese dioxide instead of activated DMSO provided a ketone, which could be further converted into the cyclohexenone 150d. Since the overall yield of these four steps was only 13% yield, we examined the other protective group. It was proved that using a protected substrate gave a better result than using a non-protected one. Tetrahydropyranyl group, which might be readily deprotected under weakly acidic conditions, was next introduced to 138. Segment A3 (P=THP) was converted into the diketone 149c in good yield. By treating of 149c in 50% aq. TFA / CHCI3 (1:5), the desired compound 150d was obtained as the main product in 78% yield in one pot. Hydrolysis of the dimethylacetal, intramolecular aldol condensation-dehydration, and deprotection of the THP group proceeded successively. The THP group functioned as the most versatile protective group among the ones used. In this way, the direct precursor of intramolecular Michael-type addition was efficiently synthesized from segment A3 and segment B. 3-3-2, Synthesis

of Spiroethers

by Intramolecular

Michael-type

Addition

The cyclohexenone derivatives (151a and 151b) could be converted to the direct precursor 151d of intramolecular Michael-type (1, 4-conjugate) addition by deprotection. Various inter- and intramolecular 1, 4-conjugate additions by heteronucleophiles have been reported and achieved under not only basic but also acidic or nearly neutral conditions (93). Hence, on our substrates (151a and 151b), deprotection of the TBDMS or the MOM groups and the subsequent cyclization might proceed in one pot under appropriate deprotective conditions. At first, we attempted the deprotection on the TBDMS protected 151a under more than ten conditions. For example, tetra-n-butylanmionium fluoride in THF, the most popular condition which is nearly neutral, did not provide 15Id or a spiroether, but only led to the decomposition of 151a. The acidic conditions used for intramolecular aldol condensation gave a little 15Id with decomposition. No reaction took place under some conditions because of the steric hindrance. Hydrogen fluoride in aq. CH3CN gave 151d in 38% yield accompanied by spiroethers in 8% yield, as shown in Table 4, that was the best result among the attempted conditions on 151a. From this result, we obtained a significant information, namely, that the intramolecular conjugate addition would proceed under acidic conditions. We next explored the deprotection reactions on the MOM protected 151b. Among the several proton and Lewis acids employed in attempts for deprotection, treatment of 151b in cone. HCl / MeOH provided a mixture of four spiroethers (152,153, 154, and 155) which was not accompanied by the deprotected 151d. The stereoisomers were readily separated by preparative TLC on silica gel. We further applied the same acidic conditions on the non-protected 151d, and got a better result from the standpoint of the total yield of four stereoisomers as shown in Table 4. We considered that the four stereoisomers are under thermodynamic equilibrium under these conditions. The spiroether 152 had the same stereogenic centers as oscillatoxin Ds and the three steroisomers (153, 154, and 155) could isomerize to the natural form 152 under their cyclization conditions. The determination of

303 stereochemistries and conformations of the four spiroethers is discussed in the next section. On the other hand, treatment of 151d with potassium r-butoxide in THF gave only two spiroethers 153 and 155 in 43 and 24% yields, respectively. This results may reflect a kinetic controlled cyclization, so that the synthesis of the natural form must be achieved under thermodynamic equilibrium conditions. The mechanism of spiroetherification is also discussed in the next section. TABLE 4. Synthesis of Spiroether by Intramolecular Michael-Type Addition Substrate

Spiroether^ (Yield^

Condition 152

153

154

155

151a

47%HF/aq. ChsCN

8^^ (total yield of four isomers)

151b

conc.HCl/MeOH

10

15

5

14

151d

conc.HCl/MeOH

26

26

6

20

151d

f-BuOK/THF

0

43

0

24

a. The structures of spiroethers are shown below. b. Isolated yield after purification by preparative TLC c. The deprotected product 26d was obtained in a 38% yield.

MeO

MeO

152 (Natural form: 25, 4R, 7R) OBn

153 (7S-isomer)

154 (45-isomer)

155 (2R, 45-isomer)

In this way, the construction of the C1-C26 methyl ester 152, lacking the y-lactone moiety, has been accomplished by an intramolecular Michael-type addition, corresponding to the conversion from the intermediate [C] to oscillatoxin Ds as shown in scheme 23. This reaction may be regarded as a biomimetic reaction. 3-3-3, Stereochemistry

and Conformation

of

Spiroethers

Oscillatoxin Ds have eight or nine asynmietric centers, most of which are concentrated on the spiroether ring system. The relative sterochemistries, determined by ^H-^H NOE experiments, are 25*, 4/?*, 7/?*, 105*, 11/?*, and 125*. Since the cyclohexanone ring has three quaternary carbons (C3, Q , and C7), ^H-^H coupling dose not provide any information of significance about the relative stereochemistry. The absolute stereochemistry at C15 was determined to be 5 from the CD spectrum. In addition to NOE experiments, acid hydrolysis of aplysiatoxins produced the same ylactones which existed in oscillatoxin Ds and the absolute configurations at the corresponding centers in oscillatoxin Ds were most likely the same as those in aplysiatoxins. Therefore, the

304

absolute sterochemistries at all the centers of oscillatoxin Ds are 25", 4/?, 7^, 105, IIR, 125, 155, 29/?, and 30R (75). During the intramolecular Michael-type addition of 15Id, the stereochemistry at the two newly produced stereogenic centers (C2 and C7) must be controlled exactly. We obtained four stereoisomers of spiroethers whose stereochemistries were determined through the detailed analysis of NOE difference spectra as well as comparison to other natural products. The NOE results for the various stereoisomers are shown in Fig. 2. Successive irradiations of the C25 axial methyl proton (5 1.22) and then of the C24 equatorial methyl proton (5 0.88) gave positive NOEs for the H2, H4, H24 protons, and for the Hio, H25 protons, respectively. Furthermore, irradiation of the Hio proton gave positive NOEs for the C22, C23, C24 methyl protons. These results were closely similar to those of natural oscillatoxin Ds, and meant that the H2 and H4 protons had to be located in axial positions, and the ether oxygen had to be attached equatorially to the cyclohexanone ring. Therefore, 152 was assigned as the natural form, possessing 25, 4R, IR configurations.

H (irradiated)

152

153

154

155

NOE

H25

H(enhanced) H2, H4, H24

H24

Hio. H25

Hio

H22. H23, H24

H25 H2

H2, H4, Hgeq, Hg, H24

H4 H11

H2, H25, H26

H24 H11

H4, He, H25. COOMe

H25 H24

Hio. H24 H2, H4, He, H25

He

H2. H24

H4, He, H25 H24

H2.

Fig. 2 The NOE Data of Spiroethers The stereoisomer 153 exhibited NOE enhancements of the Hg olefinic proton in addition to the H2, H4, Hseq., H24 protons when the C25 axial methyl protons (6 1.20) were irradiated. Positive

305 NOEs for the Hg proton based on irradiation of the H2 proton and for the C24 methyl protons (5 0.89) based on irradiation of the H n proton indicated that the double bond had to be attached equatorially in contrast to 152. Hence, it was confirmed that the stereoisomer 153 was the corresponding C7 epimer of the natural form 152. Irradiation of the C24 axial methyl protons (5 1.12) in the stereoisomer 154 gave positive NOEs of H4, Hg, H25 (5 0.87), and ester methyl (on C2) protons. This meant that the H4, H24, and the methoxycarbonyl group had to be attached axially so that the stereochemistry at C4 was inverted to an S configuration. NOE enhancement of the H2 proton by irradiation of the Hn proton, indicated that the double bond existed equatorially while the ether oxygen was axial. Hence, the conformation of the cyclohexanone ring of the C4 epimer 154 flipped in contrast to those of 152 and 153. Irradiation of the C25 equatorial methyl protons (5 0.89) of the stereoisomer 155 gave positive NOEs for the Hio and H24 protons. In addition to the H4, Hg, H25 protons, the H2 proton was enhanced instead of the ester methyl proton, when the C24 axial methyl protons (6 1.23) were irradiated. Therefore, the methoxycarbonyl group of 155 had to be attached equatorially to the cyclohexanone ring, and the other stereochemistries were identical with those of 154. In all steroisomers, the axially oriented methyl group in the g^m-dimethyl group at C^ was observed at a lower magnetic field in the ^H-NMR spectra than the equatorially oriented one. This tendency was the same as that in natural oscillatoxin Ds. In 152 and 153, the C25 and C24 methyl groups occupy axial and equatorial orientations, respectively. In 154 and 155, the orientations of the C25 and C24 methyl groups were reversed as compared to 152 and 153. Except for the natural form, the ether oxygen at C7 was attached axially to the cyclohexanone ring. 3-3-4. Mechanism Addition

of spiroetherification

under acidic and basic

by Intramolecular

Michael-type

conditions

It is known that the 2-cyclohexenone system exists, principally, as two rapidly exchanging envelope (also called sofa) conformations (93, 94). Conjugate addition of a nucleophile can occur to either face of the 2-cyclohexenone. Parallel or anti-parallel (with respect to the axial substituent at C4) attack is possible in principle, however, a nucleophile must approach from an axial direction for satisfying the requirement of the stereoelectronic effect. Anti-parallel attack leads to a favorable chair-like intermediate, whereas parallel attack leads to an unfavorable boat-like intermediate in each case. In an anti-parallel attack, the newly introduced nucleophile forms a rraw^-diaxial arrangement found in a chair conformation. Conversely, parallel attack leads to a 5'>'?i-diaxial arrangement found in a boat conformation. Therefore, anti-parallel attack is favored as this leads to a lower energy intermediate. This description can be applied to our intramolecular conjugate addition in order to explain the reaction mechanism. The precursor 151d of spiroetherification exists as the two envelope conformers [Dl] and [D2], rapidly exchanging with each other (shown in Fig. 3.). The equilibrium between them would lie towards [Dl], because the 1, 3-diaxial interaction due to the C24 and C26 methyl groups contributes negatively in [D2], whereas in [Dl], such an interaction does not exist. In

306 the conformer [Dl], an anti-parallel attack (with respect to the C25 axial methyl group at €5) of the hydroxyl group to the C7 position of the enone leads to a chair-like enol intermediate [E] under acidic conditions. Then protonation of [E] from the axial (top face) direction provides the 75-isomer 153. On the other hand, when conjugate addition proceeds from the conformer [D2], a chair-like enol intermediate [Fl] is formed by anti-parallel attack (with respect to the C24 axial methyl group at C6). Furthermore, [Fl] would flip its conformation to the another chair-like one [F2], because of the 1, 3-diaxial interaction in the former. Protonation of [F2] from the axial (top face) direction provides the natural form, 152. Protonation of [Fl] from the axial (bottom face) direction should provide the 27?-isomer [G], however, in practice, it could not be isolated. Instead of [G], the further isomerized product 155 {2R, 45-isomer), which is more thermodynamically stable, because it lacks 1, 3-diaxial interaction, is produced. The 45-isomer 154 is also produced to a small extent by isomerization of 155. The 1, 3-diaxial interaction between the C24 methyl group and the methoxycarbonyl group would make a relatively less contribution to the thermodynamic stability than the interaction between the two methyl groups.

MeO

152

MeOOC

154

Fig. 3 Equilibrium of Spiroethers under Acidic Conditions

307

Under basic conditions, only two stereoisomers 153 and 155 were produced in ca. 2:1 ratio, respectively. It is considered that the result reflects a kinetic controlled cyclization. An enolate anion corresponding to [E] is produced via a half chair-like transition state [TSl] from [Dl], and then rapid protonation of the enolate from the top face provides 153 as a major product. Through a half chair-like transition state [TSl] from [D2], another enolate anion corresponding to [Fl] is produced. This enolate is rapidly protonated from the bottom face and then epimerization of the C4 stereochemistry leads to 155. In each transition state, it is also considered that the transition state [TSl] (having no 1, 3-diaxial interaction) is more stable than the transition state [TS2] (having a 1, 3-diaxial interaction of two methyl groups). 3-4. The First Total Syntheses of Oscillatoxm D and 30-Methyioscillatoxin D The methyl ester 152 seems to be an important precursor for the total syntheses of oscillatoxin Ds. However the carboxylic acid corresponding to 152 could not be obtained under various conditions. Under the usual alkaline hydrolysis, isomerization occurred to the other stereoisomers and further led to decomposition of the substrate. Some nucleophilic deprotections in aprotic solvents or using Lewis acids were not effective, and mainly resulted in no reaction. Therefore, this route via 152 was suspended, and an alternative biomimetic route via an intermediate [C], as shown in Scheme 23, was carried out according to the initial retrosynthesis. The carboxylic acid obtained from the aldehyde 150d was esterified with segment CI or segment C2 by the Yamaguchi's method (95) to give 156a in 29% yield or 156b in 15% yield. Significant amount of the pyrone derivative 156c was also obtained as a by product in this and in the other esterification. The cyclization seemed to be an intramolecular 1, 6-conjugate addition of the corresponding carboxylate, and followed the favorable "6-endo-Dig." mode of the Baldwin's rule (96). Efforts to increase the yield of esterification were made using a model system. Partial hydrogenation of 156a and 156b proceeded in moderate yields to give the (Z)-olefins 157a and 157b corresponding to [C], which were precursors for the intramolecular Michael-type addition. The conditions (cone. HCl / MeOH) used for the cyclization of the methyl esters 151b and 151d were first applied to 157a, but a fully cyclized product could not be obtained. A product possessing a methoxyl group was detected in ^H-NMR spectrum. It was considered that methanol used as solvent added to the C7 position of 157a, which was more sterically hindered due to the introduction of the y-lactone moiety than that of the methyl ester 151d. Hence intermolecular addition of methanol predominated over the intramolecular reaction on account of the sterically hindered secondary hydroxyl group. Methanol was therefore replaced to a non-alcoholic solvent, dichloromethane, and camphorsulfonic acid was used instead of hydrochloric acid. As a result of these changes, intramolecular cyclization of 157a was accomplished to provide the spiroethers as three stereoisomers: the natural form 158a, 12%, the 75-isomer 159a, 29%, and the 2R, 45-isomer 160a, 11%. The stereochemistries of these isomers were analyzed by NOE experiments similar to those applied for the determining the stereochemistries of the methyl ester series. A small long-range coupling (72,4=1 Hz) between the H2 and H4 axial protons, not explainable as a W-type coupling.

308 characteristically observed in oscillatoxin Ds, was also detected in the iR-NMR spectrum of 158a. The 45-isomer corresponding to 154 was not produced in this case because the contribution of the 1, 3-diaxial interaction between the C24 methyl group and the y-lactone ester would be greater in contrast to that in the methyl ester. The 7S-isomer 159a could be isomerized to the natural form 158a in 24% yield and the IR, 45-isomer 160a in 12% yield, with the recovery of the starting material in 38% yield, under the same acidic condition. It was also possible to isomerize 160a to provide the equilibrium mixture of 158a, 159a, and 160a (97).

R

OBn 156a: R=H (29%, 2 steps) 156b: R=H (15%, 2 steps)

158a: natural form (12%) 159a:7S-isomer(29%) 160a: 2R, 4S-isomer (11%) 158b: natural form (7%) 159b:7S-isomer(29%) 160b: 2R, 4S-isomer (12%)

OBn

158a 158b

e

^-

159a 159b

-£ _ _ ^

158a + 160a 158b +160b

160a 160b

^_ 1 ^ ^

158a + 159a 158b+159b

126a: Oscillatoxin D ( 3 6 % ) j 126b: 30-Methyloscillatoxin D (55%) j

Scheme 29 (a) NaC102, NaH2P04, 2-methyl-2-butene / f-BuOH - H2O (4 : 1); (b) 2, 4, 6trichlorobenzoyl chloride, EtaN / toluene, then Segment CI (R=H) or Segment C2 (R=Me), DMAP / toluene; (c) H2, Lindlar cat. /EtOAc; (d) CSA / CH2CI2; (e) Raney-Ni (W-2) / EtOH

Final deprotection of the benzyl ether (98) of 158a with Raney-Ni (W-2) (99) in ethanol provided oscillatoxin D in 36% yield. The yield was decreased by adsorption of the deprotected phenol on the nickel surface, but the double bond was not reduced under these conditions. Optimization of the deprotection was considered to be possible. The spectral data of the synthetic

309 oscillatoxin D were identical with those of the natural compound. In a similar manner, 157b was converted into 30-methyloscillatoxin D, whose spectral data were also identical with those of the natural compound. The yield of the final deprotection of 158b was improved to 55% yield (100). In conclusion, the first total syntheses of oscillatoxin D and 30-methyloscilIatoxin D, belonging to aplysiatoxins / oscillatoxins family, were accomplished according to a possible biomimetic pathway. The tricarbonyl intermediate, obtained by coupling between Segment A and the Segment B followed by convesion of functional groups, was subjected to intramolecular aldol condensation-dehydration (first cyclization) utilized the p-polyketide character, to afford the cyclohexenone derivative. Thermodynamically controlled intramolecular Michael-type addition (second cyclization) of the enone ester including Segment C provided the natural spiroether ring system of oscillatoxin Ds. The two stereoisomers could be isomerized to the natural form under the same conditions as their cyclization. The two intramolecular cyclizations were the key steps in our total syntheses. ACKNOWLEDGMENTS We are grateful to Dr. Masato Oikawa who actually carried out most of the tautomycin project, and Mr. Tohru Ueno who accomplished the difficult synthesis of the anhydride segment. Their enthusiasm and skillful technique were definitely essential to the compeletion of the tautomycin project. We also thank Kaken Pharmaceutical Co. for the crude sample of tautomycin, and to Dr. Ubukata at the Institute of Physical and Chemical Research and Prof. Isono at Tokai University for the spectra of the degradation products. The oscillatoxin D project was initiated under the direction of Prof. S. Yamamura at Keio University and we are grateful to him for his useful suggestions. We are also grateful to our coworker, Mr. Takashi Goto, who actually carried out the total syntheses of oscillatoxin Ds. These two projects were supported by a Grant-in-Aid from the Ministry of Education, Science, and Culture of Japan. REFERENCES AND NOTES 1.

2. 3. 4. 5. 6. 7.

(a) Cheng, X.-C; Kihara, T.; Kusakabe, H.; Magae, J.; Kobayashi, Y.; Fang, R.-P.; Ni, Z.F.; Shen, Y.-C; Ko, K.; Yamaguchi, I.; Isono, K. J. Antibiot. 1987,40, 907-909. (b) Cheng, X.-C; Ubukata, M.; Isono, K. / Antibiot. 1990, 43, 809-819. (c) Ubukata, M.; Cheng, X.-C; Isono, K. / Chem. Soc, Chem. Commun. 1990, 244-246. (d) Ubukata, M.; Cheng, X.-C; Isobe, M.; Isono, K. J. Chem. Soc, Perkin Trans. 1 1993, 617-624. (a) Magae, J.; Watanabe, C ; Osada, H.; Cheng, X.-C; Isono, K. / Antibiot. 1988, 41, 932937. (b) Magae, J.; Hino, A.; Isono, K.; Nagai, K. J. Antibiot. 1992, 45, 246-251. (c) Kurisaki, T.; Magae, J.; Isono, K.; Nagai, K.; Yamasaki, M. J. Antibiot. 1992, 45, 252-257. (a) Magae, J.; Osada, H.; Fujiki, H.; Saido, T. C ; Suzuki, K.; Nagai, K.; Yamasaki, M.; Isono, K. Proc. Jpn. Acad. Ser. B 1990, 66, 209-212. (b) Hori, M.; Magae, J.; Han, Y.-G.; Karaki, H.; Hartshome, D. J. FEES Lett. 1991, 285, 145-148. MacKintosh, C ; Klumpp, S. FEES Lett. 1990, 277, 137-140. Bialojan, C ; Takai, A. Biochem. J. 1988, 256, 283-290; Haystead, A. J.; Sim, A. T. R.; Carling, R. C ; Honnor, R. C ; Tsukitani, Y.; Cohen, P.; Hardie, D. G. Nature 1989, 337, 78-81. Suganuma, M.; Fujiki, H.; Suguri, H.; Yoshizawa, S.; Hirota, M.; Nakayasu, M.; Ojika, M.; Wakamatsu, K.; Yamada, K.; Sugimura, T. Pro. Nattl. Acad. Sci. USA 1988, 85, 17681771. Cohen, P. Annu. Rev. Biochem. 1989, 58, 453-508.

310

8.

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

16. 17.

18.

19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

Fujiki, H.; Suganuma, M.; Nishiwaki, S.; Yoshizawa, S.; Yatsunami, J.; Matsushima, R.; Furuya, H.; Okabe, S.; Matsunaga, S.; Sugimura, T. In Relevance of Animal Studies to the Evaluation of Human Cancer Risk; D. Amato, T. J. Slaga, W. Farland and C. Henry, Ed.; Wiley: New York, 1992; pp 337-350. Cheng, X.-C; Kihara, T.; Yinng, X.; Uramoto, M.; Osada, H.; Kusakabe, H.; B.-N., W.; Kobayashi, Y.; Ko, K.; Yamaguchi, L; Shen, Y.-C; Isono, K. 7. Antibiot. 1989,42, 141144. Cheng, X.-C; Ubukata, M.; Isono, K. J. Antibiot. 1990,43, 890-896. (a) Oikawa, M.; Oikawa, H.; Ichihara, A. Tetrahedron Lett. 1993, 34, 4797-4800. (b) Oikawa, H.; Oikawa, M.; Ueno, T.; Ichihara, A. Tetrahedron Lett. 1994,35,4809-4812. (c) Oikawa, M.; Oikawa, H.; Ueno, T.; Ichihara, A. / Org. Chem. 1995. m press. Other synthetic studies: (a) Ichikawa, Y.; Tsuboi, K.; Naganawa, A.; Isobe, M. SYNLETT 1993, 907-908. (b) Naganawa, A.; Ichikawa, Y.; Isobe, M. Tetrahedron 1994, 50, 89698982. (c) Nakamura, S.; Shibasaki, M. Tetrahedron Lett. 1994, 35,4145-4148. Singh, S. B.; Zink, D. L.; Liesch, J. M.; Goetz, M. A.; Jenkins, R. G.; Omstead, M. N.-. ; Silverman, K. C ; Bills, G. F.; Mosley, R. T.; Gibbs, J. B.; Schonberg, G. A.-. ; Lingham, R. B. Tetrahedron 1993, 49, 5917-5926. Oikawa, H.; Oikawa, M.; Ichihara, A.; Ubukata, M.; Isono, K. Biosci. Biotech. Biochem. 1994,55, 1933-1935. (a) Deslongchamps, P., In Stereoelectronic Effects in Organic Chemistry, Pergamon Press: New York, 1983; Vol. 1, pp 4-53. (b) Deslongchamps, P.; Rowan, D. D.; Pothier, N.; Sauve, T.; Saundares, J. K. Can. J. Chem. 1981, 59, 1105-1121. (d) Pothier, N.; Goldstein, S.; Deslongchamps, P. Helv. Chim. Acta 1992, 75, 604-620. (a) Perron, F.; Albizati, K. F. Chem. Rev. 1989, 89, 1617-1661. (b) Boivin, T. L. B. Tetrahedron 1987, 43, 3309-3362. Transformations of spiroketal templates for remote chiral transfer to linear chain subunits in natural products synthesis, see: (a) Totah, N. I.; Schreiber, S. L. 7. Org. Chem. 1991, 56, 6255-6256. (b) Bemet, B.; Bishop, P. M.; Caron, M.; Kawamata, T.; Roy, B. L.; Ruest, L.; Saube, G.; Saucy, P.; Deslongchamps, P. Can. J. Chem. 1985, 63, 2814-2818. (c) Ireland, R. E.; Daub, J. P.; Mandel, G. S.; Mandel, N. S. J. Org. Chem. 1983, 48, 1312-1325. (a) Pettit, G. R.; Albert, A. H.; Brown, P. J. Am. Chem. Soc. 1972, 94, 8095-8099. (b) Pettit, G. R.; Bower, W. J. J. Org. Chem. 1960,25, 84-86. (c) Deslongchamps, P.; Rowan, D. D.; Pothier, N. Can. J. Chem. 1981, 59, 2787-2791. (d) Zhao, Y.-b.; Albizati, K. F. Tetrahedron Lett. 1993, 34, 575-578. (e) Zhao, Y.-b.; Pratt, N. E.; Heeg, M. J.; Albizati, K. F. J. Org. Chem. 1993, 58, 1300-1301. (a) Oikawa, H.; Oikawa, M.; Ichihara, A.; Uramoto, M.; Kobayashi, K. Tetrahedron Lett. 1993, 34, 5303-5306. (b) Oikawa, M.; Oikawa, H.; Ichihara, A. Tetrahedron 1995, 51, 6237-6254. (a) Ishihara, K.; Mori, A.; Yamamoto, H. Tetrahedron 1990, 46, 4595-4612. (b) Ishihara, K.; Hanaki, N.; Yamamoto, H. J. Am. Chem. Soc. 1991,113, 7074-7075. (a) Denmark, S. E.; Almstead, N. G. J. Am. Chem. Soc. 1991,113, 8089-8110. (b) Mori, I.; Ishihara, K.; Flippin, L. A.; Nozaki, K.; Yamamoto, H.; Bartlett, P. A.; Heathcock, C. H. J. Org. Chem. 1990, 55, 6107-6115. The spiroketals 20 and 21 were prepared by the nucleophilic additions of Grignard reagent or dianion prepared by Cohen's procedure to the corresponding lactone, see ref. 19. The preparation of spiroketals 22 and 23 were described in the following section. Kirby, A. J. In The Anomeric Effect and Related Stereoelectronic Effects at Oxygen; SpringerVerlag: New York, 1983; ref. 15a, pp 209-221. When the SI-LA reduction of 21 at -20°C was quenched before completion, twenty six percent of Ca epimer of 21 was detected on GC-analysis of recovered spiroketals. In the similar reaction at -78°C, however, no isomer of 21 was detected. Hydoxy protecting group concerning non- and chelation of metal ions and Lewis acids, see: Chen, X.; Hortelano, E. R.; Eliel, E. L.; Frye, S. V. J. Am. Chem. Soc. 1990,112, 6130-6131 and references cited therein. Corcoran, R. C. Tetrahedron Lett., 1990, 31, 2101. (a) Bartlett, P. A.; Myerson, J. J. Am. Chem. Soc. 1978,100, 3950-3952. (b) Gonzalez, F B.; Bartlett, P. A. Org. Synth. 1985, 64, 175-181. Collum, D. B.; McDonald, J. H.; Still, W. C. J. Am. Chem. Soc. 1980,102, 2118-2120. The similar approarches for preparation of spiroketal with removable equilibration controller at Ca position were reported, see: WilHams, D. R.; Barner, B. A.; Nishitani, K.; Phillips, J. G.

311

31. 32. 33. 34. 35.

36. 37. 38. 39. 40. 41. 42. 43. 44.

45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

59. 60.

7. Am. Chem. Soc. 1982,104, 4708-4710; Iwata, C ; Hattori, K.; Uchida, S.; Imanishi, T. Tetrahedron Lett. 1984, 25, 2995-2998. (a) Brown, H. C ; Bhat, K. S. J. Am. Chem. Soc. 1986,108, 5919-5923. (b)Brown, H. C ; Bhat, K. S.; Randad, R. S. 7. Org. Chem. 1989, 54, 1570-1576. Hanessian, S.; Delorme, D.; Dufresune, Y. Tetrahedron Lett. 1984,25, 2515-2518. (a) Mitsunobu, O. Synthesis 1981, 1-28. (b) Martin, S. F.; Dodge, J. A. Tetrahedron Lett. 1991, 32, 3017-3020. (c) Hughes, D. L. Org. React. 1992,42, 335-656. Mori, K. Tetrahedron 1983, 39, 3107-3109. (a) Yamamoto, Y.; Yatagai, H.; Naruta, Y.; Maruyama, K. J. Am. Chem. Soc. 1980,102, 7107-7109. (b) Yamamoto, Y.; Yatagai, H.; Ishihara, Y.; Maeda, N.; Maruyama, K. Tetrahedron 1984, 40, 2239-2246. Review, see: (c) Yamamoto, Y. Aldrichimica Acta 1987, 20, 45-49. (d) Yamamoto, Y.; Asano, N. Chem. Rev. 1993, 93, 2207-2293. Simpkins, N. S. InSulphones in Organic Synthesis; Pergamon: Oxford, 1993. Corey, E. J.; Chaykovsky, M. J. Am. Chem. Soc. 1965, 87, 1341-1353. Smith, A. B., HI; Hale, K. J. Tetrahedron lett. 1989, 30, 1037-1040. Molander, G. A.; Hahn, G. J. Org. Chem. 1986, 51, 1135-1138. Kende, A. S.; Mendoza, J. S. Tetrahedron Lett. 1991, 32, 1699-1702. Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155-4156. Tsuji, J.; Nagashima, H.; Nemoto, H. Org. Synth. 1984, 62, 9-13. Review, see: Hosokawa, T.; Murahashi, S.-I. Ace. Chem. Res. 1990,1990,49-54. (a) Heathcock, C. H. Aldrichimica Acta 1990, 23, 99-111. (b) Van Draanen, N. A.; Arseniyadis, S.; Crimmins, M. T.; Heathcock, C. H. J. Org. Chem. 1991, 56, 2499-2506. (a) Roush, W. R.; Ando, K.; Powers, D. B.; Palkowitz, A. D.; Halterman, R. L. J. Am. Chem. Soc. 1990,112, 6339-6348. (b) Roush, W. R.; Palkowitz, A. D.; Ando, K. J. Am. Chem. Soc. 1990,112, 6348-6359. (c) Roush, W. R.; Bannister, T. D.; Wendt, M. D. Tetrahedron Lett. 1993, 34, 8387-8390. Mihelich, E. D.; Daniels, K.; Eickhoff, D. J. J. Am. Chem. Soc. 1981,103, 7690-7692. (a) Corey, E. J.; Schmidt, G. Tetrahedron Lett. 1979, 20, 399-402. (b) Herscovici, J.; Antonakis, K. J. Chem. Soc. Chem. Commun. 1980, 561-562. Scott, W. J.; Stille, J. K. J. Am. Chem. Soc. 1986,108, 3033-3040. Westley, J. W.; Evans, J., R. H.; Williams, T.; Stempel, A. J. Org. Chem. 1973, 38, 34313433. (a) Baumann, M. E.; Bosshard, H.; Breitenstein, W.; Gunter, R. Helv. Chim. Acta 1986, 69, 396-403. (b) Newman, M. S.; Stalick, W. M. J. Org. Chem. 1973, 38, 3386. (c) White, J. D.; Dillon, M. P.; Butlin, R. J. J. Am. Chem. Soc. 1992,1992, 9613-961 A. (a) Huff, R. K.; Moppett, C. E.; Sutherland, J. K. J. Chem. Soc. (C) 1968, 2725-2726. (b) Huff, R. K.; Moppett, C. E.; Sutherland, J. K. J. Chem. Soc, Perkin Trans. 1 1972, 25842590. Ubukata, M.; Cheng, X.-C; Isono, K. In Symposium Papers of 33th Symposium on the Chemistry of Natural Products; Osaka (Japan), 1991; pp 643-650. (a) Yang-Lan, S.; Mueller-Johnson, M.; Oehldrich, J.; Wichman, D.; Cook, J. M. J. Org. Chem. 1976, 41, 4053-4058. (b) Bertz, S. H.; Cook, J. M.; Gawish, A.; Weiss, U. Org. Synth. 1986, 64, 27-37. (c) Fu, X.-Y.; Cook, J. M. Aldrichimica Acta 1992, 25, 43-54. Ueno, T.; Oikawa, M.; Oikawa, H.; Ichihara, A., unpublished results. Ueno, T.; Oikawa, M.; Oikawa, H.; Ichihara, A., unpublished results. (a) Carruthers, W. In Cycloaddition Reactions in Organic Synthesis; Pergamon: Oxford, 1990. (b) Curran, D. P. The Cycloadditon Approach to b-Hydroxy Carbonyls; Wiley: New York, 1990. (a) Christ, M.; Huisgen, R. Chem. Ber. 1973,106, 3345-3367. (b) Martin, S. F.; Dupre, B. Tetrahedron Lett. 1983,2^, 1337-1340. Curran, D. P. J. Am. Chem. Soc. 1983,105, 5826-5833. Tanis, S. P.; Herrinton, P. M. J. Org. Chem. 1983, 48, 4572-4580. Related oxidation of furan, see: Sargent, M. V.; Cresp, T. M. In Comprehensive Organic Chemistry, E. Haslam, Ed.; Pergamon: Oxford, 1979; Vol. 4; pp 693-744; Lipshutz, B. H. Chem. Rev. 1986, 86, 795-819. Hughes, M. J.; Thomas, E. J.; Tumbull, M. D.; Jones, R. H.; Warner, R. E. /. Chem. Soc. Chem. Commun. 1985, 755-758. Hughes, M. J.; Thomas, E. J. J. Chem. Soc. Perkin Trans. 1 1993, 1493-1505. (a) Itsuno, S.; Ito, K.; Hirao, A.; Nakahama, S. J. Org. Chem. 1984, 49, 555-557. (b) Corey, E. J.; Bakushi, R. K.; Shibata, S. J. Am. Chem. Soc. 1987,109, 5551-5553. (c)

312

61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.

72. 73. 74. 75. 76. 77.

78.

79. 80. 81. 82. 83. 84.

Corey, E. J.; Bakushi, R. K.; Shibata, S.; Chen, C.-P.; Singh, V. K. J. Am. Chem. Soc. 1987, 709, 7925-7926. (d) Mathre, D. J.; Thompson, A. S.; Douglas, A. W.; Hoogsteen, K.; Carroll, J. D.; Corley, E. G.; Grabowski, E. J. J. / Org. Chem, 1993, 58, 2880-2888. Oikawa, H.; Ichihara, A., unpublished results. Shibasaki group also reported highly enantioselective reduction of the compound related with 95: Nakamura, S.; Shimizu, S.; Nakada, M.; Shibasaki, M. In Symposium Papers of 36th Symposium on the Chemistry of Natural Products', Hiroshima (Japan), 1994; pp 611-618. Parikh, J. R.; Doering, W. v. E. /. Am. Chem, Soc. 1967, 89, 5505-5507. Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartumg, J.; Joeng, K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-M.; Xu, D.; Zhang, X.-L. J. Org. Chem. 1992, 57, 2768-2771. (a) Oikawa, Y.; Nishi, T.; Yonemitsu, O. Tetrahedron Lett. 1983, 24, 4037-4040. (b) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O. Tetrahedron 1986, 42, 3021-3028. Toshima, K.; Mukaiyama, S.; Kinoshita, M.; Tatsuta, K. Tetrahedron Lett. 1989, 30, 64136416. Kraus, G. A.; Taschner, M. J. J. Org. Chem. 1980,45, 1175-1176. Gorthey, L. A.; Vairamani, M.; Djerassi, C. 7. Org. Chem. 1984,49, 1511-1517. Sharpless, K. B.; Caron, M. J. Org. Chem. 1985, 50, 1557-1560. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull Chem. Soc. Jpn. 1979, 52, 1989-1993. (a) Evans, D. A.; Ng, H. P.; Rieger, D. L. J. Am. Chem. Soc. 1993, 775, 11446-11459; (b) Evans, D. A.; Calter, M. A. Tetrahedron Lett. 1993, 34, 6871-6874; (c) Roush, W. R.; Bannister, T. D.; Wendt, M. D. Tetrahedron Lett. 1993, 34, 8387-8390; (d) Martin, S. F.; Lee, W.-C. Tetrahedron Lett. 1993, 34, 2711-2714; (e) Paterson, I.; Cumminng, J. G. Tetrahedron Lett. 1992, 33, 2847-2850. Evans, D. A.; Rieger, D. L.; Bilodeau, M. T.; Urpf, F. J. Am. Chem. Soc. 1991, 775, 10471049. Mukaiyama, T. Org. React. 1982,28, 203-331. (a) Reetz, M. T.; Kesseler, K. J. Org. Chem. 1985, 50, 5436-5438. (b) Reetz, M. T.; Jung, A. J. Am. Chem. Soc. 1983,105, 4833-4835. (c) Gennari, C ; Beretta, M. G.; Bemardi, A.; Moro, G.; Scolastico, C ; Todeschini, R. Tetrahedron 1986, 42, 893-909. Entzeroth, M.; Blackman, A. J.; Mynderse, J. S.; Moore, R. E. J. Org. Chem., 1985, 50, 1255-1259. Moore, R. E.; Blackman, A. J.; Cheuk, C ; Mynderse, J. S.; Matsumoto, G. K.; Clardy, J.; Woodard, R. W.; Craig, J. C. J. Org. Chem., 1984,49, 2484-2489. (a) Fujiki, H.; Suganuma, M.; Nakayasu, M.; Hoshino, H.; Moore, R. E.; Sugimura, T. Gann, 1982, 73, 495-496; (b) Suganuma, M.; Fujiki, H.; Tahira, T.; Cheuk, C ; Moore, R. E.; Sugimura, T. Carcinogenesis, 1984, 5, 315-318; (c) Jeffrey, A. M.; Liskamp, R. M. J. Proc. Nat. Acd. Sci. USA, 1986, 83, 241-245. (a) Park, P.; Broka, C. A.; Johnson, B. F.; Kishi, Y. J. Am. Chem. Soc, 1987,109, 6205-6207; (b) Ireland, R. E.; Tharisrivongs, S.; Dussalt, P. H. J. Am. Chem. Soc, 1988, 110, 5768-5779; (c) Walkup, R. D.; Cunningham, R. T. Tetrahedron Lett., 1987, 28, 4019-4022; (d) Walkup, R. D.; Kane, R. R.; Boatman, Jr., P. D.; Cunningham, R. T. Tetrahedron Lett., 1990, 31, 7587-7590; (e) Walkup, R. D.; Boatman, Jr., P. D.; Kane, R. R.; Cunningham, R. T. Tetrahedron Lett., 1991, 32, 3937-3940; (f) Okamura, H.; Kuroda, S.; Ikegami, S.; Tomita, K.; Sugimoto, Y.; Sakaguchi, S.; Ito, Y.; Katsuki, T.; Yamaguchi, M. Tetrahedron, 1993, 49, 10531-10554. (a) Toshima, H.; Yoshida, S.; Suzuki, T.; Nishiyama, S.; Yamamura, S. Tetrahedron Lett., 1989, 30, 6721-6724; (b) Toshima, H.; Suzuki, T.; Nishiyama, S.; Yamamura, S. Tetrahedron Lett., 1989, 30, 6725-6728. Nakamura, H.; Park, P.; Kishi, Y. 58th Annual Meeting of Chem. Soc Jpn., Kyoto, Japan, April 1989, Abstract II, pp 1189. (a) Mynderse, J. S.; Moore, R. E.; Kashiwagi, M.; Norton, T. R. Science, 1977, 796, 538540; (b) Moore, R. E. Pure & Appl Chem., 1982, 54, 1919-1934. Moore, R. E., Personal communication, The antileukemic activity was based on an assay using a limited amount of oscillatoxin D. Kinoshita, M.; Mariyama, S. Bull Chem. Soc Jpn., 1975,48, 2081-2083. (a) Preparation of the Wittig reagent: Evans, D. A.; Kalder, S. W.; Jones, T. K.; Clardy, J.; Stout, T. J. J. Am. Chem. Soc, 1990, 772, 7001-7031; (b) N-Methoxy-N-methylamide

313

(Weinreb's amide) as acylating reagent: Nahm, S; Weinreb, S. M. Tetrahedron Lett., mi, 22, 3815-3818. 85. (a) Brown, H. C ; Singaram, B. 7. Org. Chem., 1984, 49, 945-947; (b) Chandrasekharan, J.; Ramachandran, P. V.; Brown, H. C. J. Org. Chem., 1985, 50, 5446-5448; (c) (-)Diisopinocampheylchloroborane could be readily derived from (-i-)-a-pinene. Now, this reagent is commercially available as (-)-DIP-Chloride™ from Aldrich Chemical Company, Inc. 86. The desired diastereomer (5 configuration): 5 (ppm) -0.021 and 0.017. The undesired diastereomer {R configuration): 5 (ppm) -0.006 and 0.025. 87. (a) Yadav, J . S.; Chander, M. C ; Rao, C. S. Tetrahedron Lett., 1989, 30, 5455-5458; (b) Yadav, J . S.; Deshpande, P. K.; Sharma, G. V. M. Tetrahedron, 1990, 46, 7033-7046. 88. Toshima, H.; Goto, T.; Ichihara, A. Tetrahedron Lett., 1994, 35, 4361-4364. 89. Boekman, Jr., R. K.; Charette, A. B.; Asberom, T.; Johnston, B. H. /. Am. Chem. Soc, 1991,113, 5337-5353. 90. Saito, S.; Hsegawa, T.; Inaba, M.; Nishida, R.; Fujii, T.; Nomizu, S.; Mori wake, T. Chem. Lett, 19S4, 1389-1392. 91. Ortuno, R. M.; Bigorra, J.; Font, J. Tetrahedron, 1988,44, 5139-5144. 92. Elhson, R. A.; Lukenbach, E. R.; Chiu, C.-W. Tetrahedron Lett., 1975, 499-502. 93. Perlmutter, P. Conjugate Addition Reactions in Organic Chemistry, Pergamon Press, Oxford, 1992. 94. Chamberlain, P.; Whitham, G. H. J. Chem. Soc, Perkin Trans. 11,1972 130-135. 95. Inanaga, J.; Hirata, K.; Saeki, H.; Katuki, T.; Yamaguchi, M. Bull. Chem. Soc. Jpn., 1979, 52, 1989-1993. 96. Baldwin, J. E. J. Chem. Soc, Chem. Comm., 1976, 734-736. 97. The namral form and the 75-isomer were obtained slightly with the recovery of the starting 2R, 45-isomer mainly and decompositions decreased their yields. 98. Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yanemitus, O. Tetrahedron, 1986, 42, 3021-3028. 99. Mozingo, R. Org. Synth. Coll. Vol. 3,1955, 181-183. 100. Toshima, H.; Goto, T.; Ichihara, A. Tetrahedron Lett., 1995, 36, 3373-3374.



315

Aza-Annulation of Enamine Related Substrates with UybUnsaturated Carboxylate Derivatives as a Route to the Selective Synthesis of d-Lactams and Pyridones John R. Stille and Nancy S. Barta

1 . INTRODUCTION Six-membered nitrogen heterocycles are found in a wide variety of naturally occurring alkaloids, and are often incorporated into the design of biologically active pharmaceutical products. Of the many ways to construct these heterocycles, the aza-annulation of enamine substrates with a,p-unsaturated carboxylic acid derivatives has been a versatile and efficient method for the formation of dihydropyridone and pyridone products. The synthetic utility of this approach has led to incorporation of aza-annulation methodology as a key step in the synthesis of a number of interesting heterocychc molecules. A/-AcyJatJon

O

Condensation Conjugate Addition Figure 1.

There are many variants of the aza-annulation approach to heterocycle formation. However, the general process involves the combination of three fundamental components - an amine, an a,punsaturated carboxylic acid derivative, and a carbonyl compound (Figure 1). Although each variant of the aza-annulation reaction displays subtle component dependent differences that influence the interactions in the transition state, a detailed mechanistic analysis will not be presented. However, from a synthetic perspective, a general introduction to the characteristic types of reactions involved will provide a greater understanding of the effects that substrate and reagent properties have on the outcome of the reaction. Combination of the three annulation components results in three different bond-forming processes: (1) condensation of the nitrogen with the carbonyl derivative to generate an enamine, (2) conjugate addition of the carbonyl derivative to the a,p-unsaturated carboxylic acid derivative, which serves as a Michael acceptor, and (3) acylation of the nitrogen to form the amide bond. The two approaches used most often for aza-annulation are shown in Scheme 1. In each case, the initial reaction in the aza-annulation process involves incorporation the nitrogen atom.

316

Whether through initial A^-acylation, typically the use of readily available acrylamide or acrylonitrile, or through enamine formation, the stage is set for the important carbon-carbon bond formation step. Conjugate addition results in formation of the carbon-carbon bond to the acrylate derivative, and the cyclization is completed through either condensation or A^-acylation.

o O O "^ 1

'^'^N^ ^

Conjugate Addition

or

/^-Acylation I

p6

|| I

Condensation Conjugate Addition

^^^N'" 5

^

u R5

Scheme 1.

An alternative approach to aza-annulation again utilizes condensation to form an enamine, which is subsequently //-acylated (Scheme 2). The resultant acrylenamide is a stable, isolable intermediate which can be efficiently converted to the corresponding dihydropyridone through photochemical processes. However, application of this photochemical methodology has been the subject of previous reviews, and will not be discussed here.^ Alternative routes for conversion of the acrylenamide to the dihydropyridone, through the use of Lewis acids, protic acids, and heat, generally has been unsuccessful.

Rk^.H

^^^N'" Condensation

R S - ^

R^

A/-Acylation

R 6 ' ^

. ,

"Conjugate iiinntp Addition"

R^-V >.

Scheme 2.

The electronic and steric effects observed for the various carbonyl substrates and acrylate reagents used in this reaction clearly illustrate the pivotal role of the conjugate addition step in the azaannulation process (Scheme 3). The nature of the carbonyl derivative is highly dependent upon the

317 type of substituent present. Substituents R^ and R^ directly affect the imine-enamine tautomer equilibrium, the regiochemical formation of enamine tautomers, and the reactivity of the enamine. Due to the importance of the carbonyl derived substrate in determining the outcome of the reaction, this review is organized according to carbonyl substrate type.

Iv^ E = CN, CO2R Y = SR, OR, NR2

R5

Scheme 3.

The presence of electron withdrawing groups at the a position (R^) of the acrylate derivative increases the reactivity of the reagent toward conjugate addition, while substituents in the P position (R"^) tend to provide steric constraints that hinder carbon-carbon bond formation (Scheme 3). Of the various acrylate derivatives employed in these reactions, the most frequently used have electron withdrawing functionality such as a carboxylic acid, amide, ester, or nitrile group or a combination of these. Direct pyridone formation can be achieved primarily through the use of either a,p acetylenic esters or acrylate derivatives with P substituents (Y = SR, OR, NR2) that eliminate under the reaction conditions.

H2 Pd or Pt

DDQ or Mn02 0 "'"N-^

R.V B'

r"' "' ROH, H"'

Li, NH3 or NaBH4 or RgSiH, TFA

RO '^ R5

Scheme 4.

318 The dihydropyridone products formed through this aza-annulation process can also be modified to provide other important ring systems (Scheme 4). Pyridone species have been obtained by oxidation of the corresponding dihydropyridone with either DDQ or Mn02. Reduction can be performed to generate the corresponding tetrahydropyridones with either a cis or trans relationship of substituents R^ and R^, depending on the type of reducing agent employed. Acid catalyzed addition of nucleophiles to C-6 has also been reported. The scope and utility of the aza-annulation methodology is described herein, and ;s presented according to the different types of carbonyl derivatives used. Coverage of this subject encompasses the use of enamine substrates with the same oxidation state as aldehyde and ketone functionality, which leads to the formation of 6-lactam products. Substrates of the carboxylate oxidation state, such as ketene related enamine substrates in which R^ = SR, OR, or NR2, result in the formation of products with the oxidation state of glutarimide, and will not be covered in this review. For each class of carbonyl substrate, the effectiveness and selectivity of various a,p-unsaturated carboxylic acid annulation reagents is presented. Although special emphasis is placed on the use of the azaannulation in the synthesis of biologically active natural products, a perspective of the methodology developed for each substrate type is included. Recent use of this methodology for the highly diastereoselective generation of quaternary centers is also covered. 2 . ALKYL IMINE/ENAMINE SUBSTRATES 2.1

Acrylonitrile Reagents

1

NC,

1a: n = 0 lb: n = 1

H20/ACOH (Trace)

Cyclohexylamine (0.05 equiv.), AcOH (0.02 equiv.)

200 °C 2-4 h

^^

s^N

1

[Ref. 40]

100

(±)-Tubulosine

(±)-Dihydroprotoemetine

102

(±)-Emetine

Scheme 8.

Aza-annulation of 88 with an unsubstituted acrylate derivative, 37, followed by in situ reduction of the resultant dihydropyridone with NaBH4, provided 103 as a key intermediate in the preparation of ipecac alkaloids (eq. 24)."^^

331

7

MeO. MeO.

37

NaBH4 Benzene/MeOH (1:1)

^

MeO'

(Eq.24)

MeO'

47%

88

MeO.

103

[Ret. 41]

Aza-annulation of 104 with a variety of acrylate reagents has been utilized in the synthesis of indoloquinolizidine alkaloid skeletons (eq. 25). Aza-annulation of 104 was affected with acrylic acid (91%), acrylic acid/DPPA (95%), acryloyl chloride/DMAP (64, 63%), and methyl acrylate (37, 52%) to generate the pentacyclic eburnane skeleton 105.^^2 Carbonyl reduction gave Wenkert's enamine (106), which was carried on in the synthesis of (±)-apovincaniine (107) and the clinically active synthetic analog (±)-Cavinton (108).^2 (Eq.25)

y

5 Steps

HO-

[Ret. 42]

91%

H0-, DPPA

95%

CI-, DMAP

63%

MeO-

52%

^

R=Me: 27% R=Et: 23%

Yield 105: X = 0 106: X = H,H

RO,

d 107: R = Me (±)-Apovincamine 108: R = Et (±)-Cavinton

Imine 109 has been used as an important building block in the synthesis of several natural products. DPPA, in conjunction with acrylic acid derivatives 110, provided efficient annulation when substituents were present on the acrylic acid species, such as the aza-annulation of 109 with cinnamic acid (110b) to give 111b (eq. 26).^^

HO

O DPPA

R^ 109

[Ret. 43]

a b c d e

Me Ph (E)- Me-CH=CHH H

H H H Me Ph

yield 78% 65% 73% 87% 85%

(Eq. 26)

332 Aza-annulation with 97 provided an efficient route to intermediate 112, which was converted to (±)-corynantheal (113), and also constituted the formal total syntheses of (±)-corynantheine and (±)-ajmalicine (Scheme 9).^^ Through a different route, 112 was efficiently converted to (±)dihydrocorynantheol (114).^^

OMe 97

MeO

84%

109

[Ref. 44 and 45] 114 (±)-Dihydrocorynantheol

Scheme 9.

115

°X>

MeCN, 5 h

\=^

4 Steps

^-

84%

OH

109

46% 117 (±)-Deplancheine

MeO^^.0 MeOH/Benzene \ (1:1) 25 °C, 24 h

\^

Q

^^g Acetaldehyde

EtO OEt

NaBH4, MeOH 0 °C, 50 min 87%, (From 109) [Ref. 46]

Scheme 10.

333 The indoloquinolizidine alkaloid (±)-deplancheine (117) was prepared through two complementary aza-annulation procedures (Scheme 10)."^^ When treated with the a-methylene lactone 115,109 was converted to 116 in good yield. However, after a four step sequence, (±)deplancheine was generated as only a 60% component of a three compound mixture. In order to circumvent this problem, 109 was treated with 118 to give the Michael addition product 119, and reductive cyclization completed the annulation process to give 120. Wittig-Homer homologation selectively formed the alkene to give (±)-deplancheine (117). 121

COgMe

(-1:1)

rr'COsMe II tBu02C C02tBu

NaBH4 THF

80%

COaMe

109 tBu02C Rose Bengal 500-W Halogen Lamp 20-25 °C, 2 h

C02tBu

57%

3 Steps 37% C02Me tBu02C

C02tBu

C02Me tBu02C

C02tBu

125 O (±)-Camptothecin [Ret. 47]

Scheme 11.

The aza-annulation methods developed for conversion of 88 to 94 were extended to the synthesis of the antileukemic and antitumor natural product (±)-camptothecin (125, Scheme ll)."*^ Aza-annulation of 109 with 121 in the presence of NaBH4 resulted in heterocycle formation to give 122 without subsequent elimination of the malonate species. A dye sensitized photo-oxidation promoted the rearrangement of the indolo[a]quinolizinone ring to the indolizino[l,2-/?]quinolone ring 123. Compound 123 was converted to 124, which constituted a formal total synthesis of camptothecin (125). 3.2 p-Aryl Enamine Substrates Both acyclic and cyclic aza-annulation substrates with aryl substituents in the P-position can be used effectively in the construction of alkaloid skeletons. Equations 27 and 28 illustrate examples in which aza-annulation can be performed directly from methacrylamide (127) and a carbonyl

334 compound.'^^ In the first case, CsF and Si(0Me)4 were used to promote enolization of 126, which led to formation of 128 (eq. 27).^^ When regiochemical issues arose, as for the unsymmetrical ketone 129, regioselective annulation occurred. Compounds 130 and 132 were formed through conjugate addition at the aryl substituted a-carbon (eq. 28).'*^'^^ Again, reaction with acrylamide reagents required additional enhancement, and in the transformation of 129 to 130, KOtBu was employed in order to activate the carbonyl substrate.^^ An altemative approach, activation of the Michael acceptor through use of 131, also resulted in regioselective aza-annulation to generate 132 (eq. 28).50 u

CsF, Si(0Me)4 80 °C, 5 h

(Eq.27)

••

76% 127

u

131

127 f-BuOK Dioxane 130

57%

ÔEt

EtOH Reflux, 5 h 129

(Eq.28)

67%

[Ref. 49]

The use of enamine and imine derivatives of carbonyl substrates was also an effective means of performing aza-annulation. Treatment of either enamine 133 or imine 134 derivatives of (3tetralone with acrylamide resulted in the formation of 135 with (85% yield)^! or without^ solvent (eq. 29). In contrast, aza-annulation of the corresponding methyl enamine of P-tetralone with methyl methacrylate generated a mixture of products.^2

H2N

(Eq.29)

20 TsOH 80-130 °C

135

134

335 Aza-annulation of a number of dimethoxy-substituted p-tetralone derivatives, such as those represented by 136, with acrylamide was used to produce 137. In turn, 137 was an important intermediate in the synthesis of conformationally restricted congeners of dopamine (eq. BO).^^ 1)80°C, 3h 2)130°C,0.5h

H2N

(Eq.30)

> •

81%

20 136

137

OMe

MeO'

OMe

Control of ring fusion through post aza-annulation modification was also employed in the synthesis of benzoquinolinone steriod analogs, which have demonstrated selective and potent inhibition of human type I 5a-reductase enzyme (eq. 31).^^ Aza-annulation was performed with acrylamide (20) and tetralone 138 to regioselectively generate the quaternary carbon of 139, and ionic reduction led to formation of the trans fused product 140. Similarly, aza-annulation of enamine 141 with acrylamide generated enamide 142, and ionic reduction gave 143 as the trans fused ring system (eq. 32).^"^ Subsequent enantioselective syntheses of these molecules are discussed in Section 8.

) NaH, Mel 2) EtaSiH, TFA

,r±? =

(Eq.31) O^ N

139

1) NaH, Mel 2) EtaSiH, TFA O^ N

(Eq. 32)

[Ref. 54]

Formation of the methyl enamine of 144, followed by aza-annulation with methacryloyl chloride provided an 80:20 mixture of the desired tetracyclic system 145 to the N-acylation product

336 146 (Scheme 12).^^ Reductive modification of 146 was selectively performed to access either the trans or cis ring fusion for total synthesis of (±)-festuclavine (147) or (±)-costaclavine (149), respectively.

1) MeNH2 2)

O

CI

\ 66%

1)LiAIH4 2) Mn02 16P/o

[Ret. 55]

Scheme 12.

4 . VINYLOGOUS AMIDE DERIVATIVE SUBSTRATES

R2'

R 2 ^

R3 R3

Section 4.1 Vinylogous Amides p-Enamino Ketones

4.1

Section 4.2 Vinylogous Carbamates P-Enamino Esters

Section 4.3 Vinylogous Ureas p-Enamino Amides

Vinylogous Amide Substrates 4.1.1

Acyclic Dione Derivatives

p-Diketone substrates have been valuable in the aza-annulation reaction with a,P-unsaturated carboxylic acid derivatives, and both acyclic and cyclic P-diketone species have been investigated. The simplest acyclic p-diketone, 150, underwent condensation reaction with BnNH2 to generate the

337 corresponding (3-enamino ketone 151 (Scheme 13).56 Regioselective 5-lactam formation was affected through aza-annulation with this vinylogous amide, and solvent effects played an important role in this reaction. For example, reaction with acryloyl chloride in benzene at reflux generated 152 in 44% yield,56 while the same reaction generated an 94% yield when performed in THF.57,58 xê resultant vinylogous amide functionality of dihydropyridone 152 was catalytically reduced to give the predominantly cis substituted 6-lactam 153.^"^'58 Epimerization of the diastereomeric mixture, followed by Baeyer-Villiger oxidation established the trans stereochemistry of the oxygen substituent relative to the methyl substituent of intermediate 154, which gave 155 upon base catalyzed hydrolysis.

O

BnNHs, TsOH Benzene Reflux

O

N

AA

150

151

74%

OH

V

154

THF 94% (from 150)

1) DBU (cisrtrans, 24:76) 2) m-CPBA CF3CO2H

O

NaOH H2O

'")b 155

O

O 1 atm of H2 Pd/C , 81% NaaCOg p^ ^"^N

Cis:Trans (90:10)

45%

o

Scheme 13.

[Ref. 56-58]

Aza-annulation with unsymmetrical (3-diketone 156 resulted in regioselective generation of 157, which gave 158 upon reaction with acryloyl chloride in benzene (47%) or THF (96%) (eq. 33) 56,57,59 j h e reaction of cinnamoyl chloride with 157 in benzene gave 158 in 30% yield.^^ O O

BnNH2, TsOH Benzene Bn Reflux

O 156

H

(Eq.33)

CI

" v ^

[Ref. 56, 57, and 58] R H H Ph

Solvent Benzene THF Benzene

156-> 158 47% 96% 30%

338 Enamino ketoester 159 (in equilibrium with the corresponding imidizolidine) efficiently underwent aza-annulation with acryloyl chloride in the presence of pyridine and DMAP to give 160 (eq. 34).60 This dihydropyridone was then converted to 161, the pentacyclic skeleton of the 21epimer of the aspidospermine alkaloids.

(Eq.34) TiCU CICH2CH2CI 80 °C

/I.

63% 161

C02Me

Pyridone products were directly accessible through aza-annulation with p-heteroatom substituted acrylate derivatives. Aza-annulation of 150 with 86 in the presence of K2CO3 led to the formation of 162 upon elimination of MeSH (eq. 35).^ When facilitated by NaOEt, reaction of 150 with 163 resulted in carbon-carbon bond formation through conjugate addition and subsequent elimination of MeOH (eq. 36).^^ Intramolecular lactam formation generated bicyclic species 164.

O

O

-V"

MeS 150

SMe

K2CO3, DMSO 100°C,3h 6P/0

^

(Eq.35)

86

[Ref. 6]

XX

NaOEt

(Eq.36)

69%

150 [Ref. 61]

Pyridone products 166 were also generated through aza-annulation of (3-enamino ketones 165 with 57 (eq. 37).62 p-Enamino ketone 167 could be generated either by the condensation of 150 with ACONH4 or by hydrogenation of isoxazole 168 (eq. 38). Subsequent aza-annulation with 57 gave 169.^2 Application of this methodology to the synthesis of medorinone (170) was completed by the conversion of 169 to 170.

339 DMF 1)25°C,3.5h 2) Reflux, 24 h

MeO,^0 "^N IN - ^ O ^

I

165

57

[Ref. 62]

O

O

(Eq.37)

•

R = Et (43%) 4-F-C6H4 (52%) Thiophene (47%) CgHs (72%)

R^Ô

166

ACONH4

AA 150

^

N

^

O

57 DMF 1)25°C, 3.5h 2) Reflux, 24 h

(Eq- 38) u ""N

62%

1) (Me2N)2CHOtBu 2) ACONH4/DMF 59% 170 Medorinone

4.1.2

Cyclic Dione Derivatives

Cyclic p-diketones and their derivatives have also been the subject of aza-annulation studies. Methodology studies with enamine derivative 171 established the effectiveness of acryloyl chloride (64) for the formation of 172 (eq. 39).^^ Aza-annulation with a p-substituted aery late derivative was most efficiently accomplished through the use of diester Michael acceptors, as exemplified by reaction of 173 with 174 in the generation of 175 (eq. 40).^^ Benzene Reflux r ^

[Ref. 56]

64

(Eq. 39)

47%

171

(Eq.40) 71%

340 Aza-annulation of the benzyl enamine derivative of 179 was employed in the synthesis of (±)-5-epipuniiliotoxin (Scheme 14).^^ Condensation of 179 with BnNH2 followed by reaction with acryloyl chloride (64) gave the key bicyclic intermediate 180. Catalytic hydrogenation selectively established the cis ring fusion of 181. Addition of MeMgBr gave 182, and stereoselective dehydroxylation generated 183. Compound 183 was converted to (±)-5-epipumiliotoxin (184) by sequential deprotection, imidate formation, alkylation, and reduction procedures. 1)3atmofH2 Pd/C NaaCOa

1) BnNH2 Benzene Reflux

> •

& c

2)

179

2) (C0CI)2, DMSO, NEt3

THF CI

85%

64

75%

4 Steps t

52%

25%

182

184 (±)-5-Epipumiliotoxin Scheme 14.

[Ref. 64]

A

NC, 2

1

185

1)Mel 2) H2/Pt 3)LiAIH4 4) AcCI/Pyridine 60%

(±)-A/crMethyl-A/[3-Acetylphlegmarine (Mixture of 4 epimers) [Ref. 65]

1)NaH, CS2 2) Mel 3) BuaSnH AIBN

'"^N H.j

OH

341 Acrylonitrile (2) has also been used for heterocycle formation from P-diketone substrates. Conjugate addition of 185 to acrylonitrile produced 186 with previously reported conditions,^ and cyclization under hydrolytic conditions generated the corresponding lactam 187 (Scheme 15).65 Dissolving metal reduction provided the trans ring fused product 188, which was subsequently converted to 189. Reduction of 189 completed the synthesis of the Lycopodium alkaloid (±)-Namethyl-A^p-acetylphlegmarine (190) as a mixture of 4 diastereomers. Direct reaction of 191 with acryhc acid resulted in efficient formation of 192 (eq. 41).^^ Subsequent dehydrogenation at elevated temperatures provided the aromatic species 193, which was a key intermediate in the synthesis of 194, a compound that displays strong ^-blocking activity. Under similar conditions, reaction of 191 with crotonic acid, cinnamic acid, and ethyl acrylate did not generate the corresponding bicychc alkaloid skeletons.

A, 191

10%Pd/C 195°C, 3h Decaline 140 °C 3h 95%

OH

74%

[Ret. 66]

When treated with acrylic acid, bicyclic enamine 195 was converted to the tricyclic vinylogous imide 196, which was then incorporated into the synthesis of the Lycopodium alkaloid annotinine (197) as well as an annotinine degradation product (eq. 42).^'^ u 135°C

23 Steps

63%

98:2 cis stereoselectivity. Conversion to the corresponding methyl ketone followed by epimerization at C5 generated 251 with the stereochemistry desired for further elaborations. Baeyer-Villiger oxidation and protecting group manipulation gave 6-lactam 252. This key intermediate was transformed to the a-D-mannosidase inhibitors (±)-mannonolactam (253)^^ and (±)-deoxymannojirimycin (254),^^ as well as to the antibiotic and anesthetic agent (±)-prosopinine (255).^8''75 U

BnO'

P

246

BnNHs THF Reflux

O

THF Reflux OEt

OEt

247

62% (From 246)

249

EtQ-Ô

1 atm of H2 Pd/C Na2C03

80%

Eton

1) CF3CO2H m-CPBA 2) KOH, H2O 3) KOH, BnBr

1) MeMgBr NEt3 2) DBU, 25 °C

3nO^Xj EtOÔ

OH OH

OH OH

OH OH

253

(±)-f\/lannonolactam

254 (±)-Deoxymannojirimycin

255 (±)-Prosopinine

Scheme 19.

[Ref. 58 and 75]

Products related in structure to 249 and 250 could be accessed through condensation of benzyl amine with tetronic acid (256) followed by aza-annulation with acrylic anhydride to give 258 (eq. 52).58 Stereoselective generation of the cis ring fusion of 259 was accomplished by catalytic hydrogenation.

348 O

O (Eq.52)

BnNHs TsOH Benzene Reflux

OH ,

1 atm of H2 Pd/C Na2C03 EtOH

THF Reflux

257

83%

71% (From 256)

259

Incorporation of a P-phenyl substituent in the aza-annulation process, by the use of the appropriate acrylate reagent, was even more difficult than the reaction with crotyl derivatives and necessitated the use of doubly activated 260 to access 261 (eq. 53)7^

"'-H^" o

N

O

A^

75%

221

EtOÔ

260

o

261

[Ref. 76]

Reaction of 205 with benzotriazole enamine (262), generated through conjugate addition to the corresponding alkyne, led to pyridone 263 (eq. 54)7^ Formation of 263 was accompanied by a minor amount of a side product (264, 8%) in which the aromatic species on the nitrogen participated in the cyclization.

N-N

205 I

N-N

,H N' 262

(Eq. 54)

E

E = C02Me [Ref. 77]

Pyridone formation could also be accomplished by reaction of 208 with 86 in the presence of K2CO3 to yield 265 upon elimination of MeSH (eq. 55).^ Alternatively, reaction of 267 with Penamino ester 266 led to the formation of the pyridone 268 in good yield (eq. SG).^^

349

O

. "'•'Y" K2CO3 DMSO

O OEt

MeS

SMe

II

86

208

(Eq- 55) ^SMe

77%

EtO"^0

265

[Ref. 6]

o Pyridine

MeO" OEt 266

267 MeO.

[Ref. 78]

Y s;

s

II

(Eq.56)

81%

EtOÔ

268

0

The use of 1,3-dicarboethoxyallene (269) also provided a route to 5-lactam products through aza-annulation (Scheme 20)7^ Treatment of 221 with 269 resulted in formation of the pyridone derivative 270. Similar chemistry was performed with p-enamino ketones, but studies were far more extensive with the P-enamino ester substrates.^^ Elaboration of 270 led to hydrolysis of the ester to give 271 followed by construction of a lactone ring to give 272. O

1 eq. of NEt3 AcOHôluene(1:1) 100°C, 5h

EtO'

^^N'^ O

OEt

OEt 221

269

O

>

^ ^ N ^

OEt

70% EtO-Ô

270

NaOH, EtOH Reflux, 3 h MeO

01

AcOH

EtO ^ O [Ref. 79]

272

31% (From 270) Scheme 20.

Aza-annulation methodology that involved 269 was applied to the synthesis of (±)camptothecin (125, Scheme 21).^^ Combination of 205 and 273 generated 274, which was transformed to 275 under mild conditions by aza-annulation with 269. Intermediate 275 was then converted to 276, which was carried on to (±)-camptothecin (125).

350 O

EtO

1 eq. of NEta

Eto->.

\

N' H

EtO" 273

EtO EtgO

EtO^'^^^N'^ O

•

MeO O

MeOÂ^

O 205

269

MeOH 25 °C

V

OMe

OEt

EtO N

EtO"

OE OEt

^

45% ^

274

O'ÔMe 275

OMe

4 Steps

6 Steps

1% 125

(±)-Camptothecin Scheme 2 1 .

[Ret. 80]

An interesting acrylate derivative, 277/278, was also used for pyridone formation (eq. 57).81 Treatment of 221 with 277/278 resulted in formation of the corresponding a-acyl substituted pyridone 279.

OH DMF

^-N'^ O

OEt 221

68% 277

278

(Eq.57) EtO ^ O

279

[Ret. 81]

Malonic acid derivatives have also been effective as 1,3-dielectrophiles for the formation of 4hydroxypyridone products. Although not truly acrylate-type reagents, the tautomeric form of these species was similar in nature to their acrylate counterparts. Reaction of 280 with 281 led to the quantitative formation of 282 (eq. 58).^2 Examples of (J-enamino ketone substrates were also reported for reaction with 281, but their use was limited. ^^

351 O MeO,

MeO,

(Eq. 58)

281

100% 280 [Ret. 82]

The formation of pyridones by aza-annulation with aryl-substituted malonate derivatives was shown to be highly dependent upon the substituent pattern on the aryl ring (Scheme 22).^^ Ring formation was most efficient when the aryl group on the malonate reagent was either unsubstituted (phenyl, 283) or substituted in the 4-position (286). Product formation was significantly decreased when the malonate reagent differed from this substitution pattern. Examples of typical malonate azaannulation reactions are illustrated by the conversion of 221 to 284 by treatment with 283, and the analogous formation of 287 from the reaction of 221 and 286.^^ Subsequent reactions of 284 and 287 were highly dependent on nitrogen substitution. During hydrolysis of 287 with NaOH/H20, the intermediate carboxylic acid species decarboxylated rapidly to give 288. In contrast, hydrolysis of 284 under the same conditions gave the carboxylic acid as an isolable intermediate, and extreme conditions were required to produce 285 through decarboxylation. Both 285 and 288 were found to inhibit the growth of Mycobacterium tuberculosis.^^

1) NaOH/HaO Reflux 2) 220 °C 2,4,6-Trichlorophenol

MeO'

^^^

MeOÔ 285

R = Bn OEt 221: R = H 222: R = Bn

S.R = H

220 °C, 30 min

MeO' MeO ^ O [Ref. 83]

OH

80%

Bromobenzene Reflux, 1 h

Scheme 22.

352

^

°^^

OEt

Benzene 20 °C, 2 h

r=\

290 1) KOH, H2O Reflux 2) H C I / ^ ^ 85%

Brs

SOCI2, Benzene C

V.

50%

A very unique reagent (290) for aza-annulation with P-enamino ester 289 was reported (Scheme 23).^'^ Combination of 289 and 290 resulted in the formation of 291, which was further modified through a variety of pathways to produce ring-opening of the cyclopropane ring. The cyclopropyl ring could be opened to place the benzyhc fragment at C-3 (292), remove it entirely through hydrogenation (293), or situate the substituent at C-4 (294). 4.2.2

Cyclic Enamino Esters u

(Eq.59) 296 Ref. [85] [57, 59] [85]

Conditions EtO^Ô Conditions Pyridine, Toluene, Reflux THF, Reflux NaH, Et20, 25 °C

CI CI EtO

rield 72% 87% 75%

A variety of aza-annulation chemistry has focused on the conversion of cyclic enamino ester 296 to a variety of substituted indolizidinone products (eq. 59). The simplest aza-annulation process involved treatment of 296 with either acryloyl chloride (64) or ethyl acrylate, which generated the

353 bicyclic product 297.5'7'59,85 when acryloyl chloride was used, reaction occurred in either toluene or THF at reflux, and the reaction in THF produced a slightly higher yield. Deprotonation of 296 followed by treatment with ethyl acrylate, produced comparable results at ambient temperature.^^ Pyridone formation was achieved through a two step process, by sequential conjugate addition and cychzation (eq. 60).^^ Conjugate addition was accomplished through extensive heating of 296 with 57, and cyclization of 298 to 299 was facilitated by the subsequent addition of NaH.

MeO^Ô 57 Benzene Reflux, 4 d

.H OEt

(Eq.60)

NaH Benzene Reflux, 1 h OEt

60%

54%

296 MeCÔ

[Ret. 86]

The conversion of 296 to 297 was used as a key ring forming step in the synthesis of (±)tashiromine (301, Scheme 24).^9 Stereoselective introduction of the two vicinal stereogenic centers was accomplished through catalytic hydrogenation of 297, which resulted in >95:5 stereoselectivity for generation of 300. Further reduction of 300 gave (±)-5-epitashiromine (231), which was then efficiently converted to (±)-tashiromine (301).

THF Reflux 296

ffr/o

- c6

3 atm of H2 Pd/C Na2C03

Eton

297 E f O ^ O

OEt

95%

91%

CO

HE "ÔH 301 (±)-Tashlromine [Ref. 59]

1)LiAIH4 2) H2O, NaOH

1) (CIC0)2, DMSO NEt3 2) Piperidine, pTsOH 3) (C00H)2, H2O 4) LiAIH4 58% Scheme 24.

OH 231 (±)-5-Epitashiromine

354 O

A-N'

(Eq.61) ^R1

O OEt

Conditions

296 [Ret. 85] Conditions Pyridine, Toluene, Reflux Pyridine, Toluene, Reflux NaH, EtaO, 25 °C NaH, Et20, 25 °C

X CI CI EtO EtO

R^ Me Ph Me Ph

rield

Product a b a b

Yield 72% 53%

^ —

— —

63% 74%

O

[Ref. 87]

[Ref. 87]

[Ref. 86]

^ r-*Â^ NH; Etc " O 29%

305

97%

Benzene Reflux 1h

1) Benzene Reflux, 1 h 2) 205 °C, 3 h

HN 306 )

309

54% Benzene Reflux, 48 h

/^N'

296 a: R=Et

O OR

205 MeO > § s 1) Benzene Reflux, 2 h ^^Q 2)NEt3, MeOH 25 °C, 15nnin

O O

EtO OEt

(EtO)20P DP^^-^ÔEt 311

MeO^Ô 312 [Ref. 88]

74%

72% 312 54%

OEt MeC^'O Scheme 25.

[Ref. 88]

310 [Ref. 86]

355 The reaction of crotonic and cinnamic acid derivatives with 296 was very dependent upon the nature of the carboxylate derivative used in the aza-annulation and the conditions of the reaction (eq. 61). When the corresponding acid chloride was used in the presence of pyridine, only A^-acylation was observed to yield 303.^^ However, when 296 was deprotonated with NaH and then treated with the ethyl ester of either crotonic or cinnamic acid, aza-annulation occurred to give 302. The use of aza-annulation to generate dihydropyridone and pyridone products with substitution p to the lactam carbonyl was performed with a variety of other reagents (Scheme 25). The reaction of 296 with maleic anhydride (304) gave a high yield of the expected dihydropyridone annulation product 305 to generate a -CO2H substituent.^^ In contrast, the use of maleimide (306) under the same reaction conditions gave a good yield of the corresponding Michael addition product, but this species could only be cyclized by heating with NaH to give a low product yield of the amidesubstituted 307.^^ Fumarate derivative 308 was used the prepare the corresponding estersubstituted 309.86 Formation of pyridone products was accomplished in a number of ways (Scheme 25). The reaction of 205 with 296 gave the corresponding conjugate addition, and treatment with NEt3 facilitated cyclization to complete the aza-annulation process and the formation of 310.^6 Disubstituted pyridone 312 was prepared through aza-annulation with either 311 or allene 269 as part of a synthesis of camptothecin precursors.^^

/^N^^ O

EtaN OMe

296b

OMe

92%

313 86%

1) f-BuOK, EtI 2) (CH20)n f H2SO4, H2O

4 Steps 8%

125 (±)-Camptothecin [Ret. 89]

Scheme 26.

Aza-annulation of 296b with 313, the chloro analog of 311, led to formation of bicyclic pyridone 314 (Scheme 26).^^ Modification of this intermediate gave 315, which was then decarboxylated, oxidized, and utilized in the Friedlander quinoline synthesis to give (±)-camptothecin (125).

356 Formation of the corresponding a-substituted 6-lactams was accomplished by reaction of 296 with itaconic anhydride (316) to give dihydropyridone 317,^^ while pyridone formation was accomplished by aza-annulation with diester 47 to generate pyridone 318 (eq. 62).^'^

EtO" 316

"

Benzene Reflux, 1 h

,0H

92%

cu„„

(Eq.62)

OEt

47 OEt Benzene Reflux, 12 d

O

O OEt

60%

EtO'Ô

296

318 [Ref. 87]

Application of aza-annulation with cyclic enamino esters was reported in the synthesis of angiotensin converting enzyme inhibitor A58365A (323, Scheme 27).90 Aza-annulation of proline derivative 319, which was obtained in 4 steps from L-pyroglutamic acid, with a-methyleneglutaric anhydride (320) led to the formation of indolizidinone 321 as a mixture of diastereomers. Esterification followed by oxidation with DDQ gave 322, which was converted in 4 steps to the desired target 323. MeO.

Benzene Reflux

MeO. OH

95%

OBn 319

320

BnO ^ O 1)CH2N2 2) DDQ

HO

40%

MeO. 4 Steps

OH

[Ref. 90]

(67:33)

OMe

33% A58365A

BnOÔ Scheme 27.

Aza-annulation of 324, the six-membered ring analog of 296, with acryloyl chloride or ethyl acrylate led to the formation of a mixture of isomeric enamide products 325a and 326a (eq. 63).^^ The corresponding crotyl derivatives (b) were also successfully employed in these aza-annulation

357 studies. Regioselective formation was only observed when ethyl 3,3-dimethylacrylate, a derivative disubstituted (R,R = Me,Me) in the P-position, was used, but the yield was poor (34%). The lack of regioselectivity in this aza-annulation process has prevented the otherwise straight forward conversion of 325 and 326 to natural product targets such as (±)-lupinine (232). O (Eq. 63)

OLX„

Conditions

324

EtOÔ Conditions

EtOÔ

Product

325

326

Pyridine, Toluene, Reflux

CI

H

a

70

30

82%

NaH, Et20,25 °C

EtO

H

a

65

35

49%

Yield

Pyridine, Toluene, Reflux

CI

Me

b

60

40

68%

NaH, Et20,25 °C

EtO

Me

b

52

48

60%

[Ref. 85]

The seven-membered ring analog, 327, showed different properties than those observed for 324 (eq. 64).^^ Complete regioselective formation of 328 was observed for the acryloyl or crotyl derivatives. In these aza-annulation reactions, the use of acid chloride reagents resulted in higher yields than the corresponding ethyl esters.

(Eq. 64)

Conditions

^

327 Conditions Pyridine, Toluene, Reflux

CI

NaH, Et20, 25 °C Pyridine, Toluene, Reflux

EtO CI

Product H H

a a

Me

b

Etc " " O

328

Yield 95% 56% 68%

[Ref. 85]

An interesting variation on the aza-annulation of cyclic enamino ester substrates is illustrated in Scheme 28.^1 The amino acid derivative 329, formed by the reaction of the corresponding amino phenol with 205, underwent conjugate addition with a second equivalent of 205 to give 330. Treatment of 330 with a nucleophile such as EtOH or pyrrolidine resulted in the formation of 332. Formation of 332 was suggested to proceed through intermediate 331.^^ Evidence for this intermediate was acquired by isolation of the nitrogen analog 334, prepared in the same manner by

358 reaction of 333 with 205 (eq. 65).92 Several aromatic and aliphatic 1,2-diamine substrates, including unsymmetrical diamines, were successfully employed in this aza-annulation reaction. Overall, the reactions in Scheme 28 and eq. 65 illustrate an interesting class of conformationally restricted amino acid derivatives.

Me 0 , ^ 0

r^

OMe 329

Dioxane Heat

OMe

O'ÔMe 206

HY, CH2CI2 25 °C, 3 h Y = OEt; 89%

Y= N J

MeO

Scheme 28.

[Ret. 91]

MeO,^0

O [Ret. 92]

4.2.3

; 92%

OMe 333

1) Dioxane Reflux (54%) 2) DMSO Reflux, 1 h (64%)

(Eq. 65) OMe

O'ÔMe

MeOÔ ^

334

205

Tetrasubstituted Enamino Esters

Tetrasubstituted enamino esters have also been employed in the aza-annulation reaction with acryloyl chloride (64) (eq. 66-68).^^ As observed for the examples in which trisubstituted enamino esters were used, the ester functionality directed regioselective enamine formation, and the resulting carbon-carbon bond formation occurred at the more substituted site. In the case of tetrasubstituted enamino ester substrates, carbon-carbon bond formation resulted in the generation of a quaternary carbon with deprotonation to form the enamide functionality which occurred exocyclic to the 6-lactam ring. This process was performed with 335 to generate the fused bicyclic ring system 336, which was reduced by hydrogenation to a mixture of diastereomers 337 (eq. 66).59 p.Keto lactone 338 was exposed to the same reaction conditions to give the corresponding spirocyclic 5-lactam 339 (eq.

359 67).59 Hydrogenation of this annulation product resulted in stereoselective formation of 340. In each of these examples, formation of a stereogenic center occurred, and this methodology was used as the ground work for asymmetric induction in the aza-annulation reaction (see Section 8).

O

A^COsEt

1) BnNH2, TsOH Benzene, Reflux 2) AcryloyI Chloride THF, Reflux > •

89%

U

3 atm of H2 Pd/C, Na2C03 EtOH

r

TcogEt

335

^.

85%

I

336

rco2Et 337

(56:44)

[Ref. 59]

O

O

1) BnNHg, TsOH Benzene, Reflux 2) AcryloyI Chloride THF, Reflux

3 atm of H2 Pd/C, Na2C03 EtOH

^^

Bn^ N

(Eq.67)

83%

84% 339

338

340

[Ref. 59]

Aza-annulation of 341 with acryloyl chloride (64) provided 342, which was reduced to give the indolizidine-type ring skeleton 343 (eq. 68).^^

CI 64

/^NH \As^C02Et C02Et

THF, Reflux

^^ 75%

3 atm of H2 Pd/C, EtOH

-N Et02C C02Et

85%

342

341

^^

(Eq.68) "N Et02C C02Et 343

[Ref. 59]

4.2.4 a-Amido Aza-Annulation Reagents There has been increased use of acrylate reagents with an a-nitrogen substituent, and the corresponding aza-annulation reaction products were a-amido 6-lactams, which represented an interesting class of conformationally restricted peptide analogs. Oxidation of the dihydropyridones that resulted from aza-annulation led to the corresponding pyridones.

360

"°-VY

(Eq. 69)

NEt3, DPPA DMF, 0 °C

344

R = Et; 78% R = H; 88% 345: R = Et 346: R = H

104: R = Et 109: R = H [Ret. 42]

The first methods for aza-annulation with a-amido-derived acrylate reagents, such as 344, involved activation of the carboxylic acid toward acylation through the use of DPPA (eq. 69)."^^ Efficient aza-annulation of 104 and 109 generated the corresponding amino acid derivatives 345 and 346.

NEt3 Dioxane, 80 °C

"•N-P

(Eq.70)

OEt

O

221: R = H 222: R = Bn 266: R = Ph

EtOÔ

R = H; 64% R = Bn; 56% R = Ph; 77%

EtO*

348: R = H 349: R = Bn 350: R = Ph

[Ref. 93]

An alternative method for generation of the corresponding pyridone species was performed with 347, prepared from hippuric acid (eq. TO).^^ j ^ e reaction of enamino ester 221 with the novel acrylate derivative 347 gave a-amido pyridone 348 in a single step. Substituted derivatives 222 and 266 reacted in a similar manner to give 349 and 350, respectively.^^ (Eq.71) II

•

H0\V 344

NaH THF '"'^

EtO' 11

'

—^ NaoYr 351

THF

CI

1

EtO'Ô

O

H

Vr 352

[Ref. 94]

Aza-annulation has been efficiently performed with the mixed anhydride-type reagent represented by 352, which was generated by deprotonation of acid 344 with NaH followed by treatment with Et02CCl (eq. 71).94 Because attempts to isolate the active annulation species led to reagent decomposition, the reagent mixture was generated in situ, and the structure illustrated for

361 352 was proposed. In order to simplify the presentation of this chemistry, the active species will be referred to as 352.

O

1) BnNHa BF3*OEt2 Benzene Reflux

O

'Y^

2) 352, THF

DDQJoluene Reflux (73%) or Mn02, Xylenes,

Reflux'(90%)

O

H

^"^N^^^'^Y^

OEt 91%

208

0*^061

2) EtOgCCI

O

OEt

H.^,H

68%

H

O

78%

354

30% H2O2 KOH

KOH H2O

1)NaH 3)

O

353

H

O

'I

OÔH 357

355

356

H

OEt Scheme 29.

[Ref. 94]

When the enamine generated from 208 was treated with 352, formation of the corresponding dihydropyridone 353, which was similar to an Ala-Ala dipeptide, occurred in high yield (Scheme 29).^^ Oxidation of 353 with either DDQ or Mn02 gave the corresponding pyridone product 354. The dipeptide analog 354 could be selectively deprotected to generate 355, which was then converted to the tripeptide species 356 through standard peptide coupling techniques. Alternatively, hydrolysis of the protected carboxyl and amino termini could be affected in one step to transform 354 to 357. 1) BnNH2 BF3»OEt2 Benzene, Reflux

BnO^ 246

2) 352, THF

OEt

BnO.

(Eq. 72)

83P/o

[Ref. 94]

Conjugate addition of BnNH2 to an alkyne was also an effective method for generation of the enamine used in this class of aza-annulation reactions. Formation of the Ser-Ala dipeptide analog 358 was accomplished by conjugate addition of BnNHi to 246, followed by annulation with 352 (eq. 72).94 Interestingly, when the same methodology was used to access the Phe-Ala dipeptide

362 analog, the ester substituent controlled regioselective aza-annulation from 359, but kinetic deprotonation resulted in conjugation of the enamine with the phenyl substituent to give 360 (eq. 73).^"^ Enamine formation from 205 followed by aza-annulation generated the Asp-Ala analog 361, which was then oxidized to the corresponding pyridone 362 with DDQ (eq. 74).9'^ In the case of 358 and 360, treatment of the dihydropyridones with DDQ did not result in effective formation of the desired pyridone.

Ph'

1)BnNH2 BF3«OEt2 Benzene.Reflux 2) 352, THF

O OEt

359

(Eq.73)

61%

[Ref. 94]

1) BnNH2 BF3*OEt2 Benzene.Reflux 2) 352, THF

OMe

H

DDQ Toluene Reflux MeO,

P 206

O

71%

OMe

71% 361

Ô'ÔMe

C^

OMe

[Ref. 94]

/-N'^0

352, THF 77%

c6rY OÔEt

296

^ DDQ 78% I Toluene Reflux

KOH, H2O 25 °C

OÔH [Ref. 94]

82% Scheme 30.

O

H

•v-VV OÔEt

364

363 The dihydropyridone 363 with the features of a Pro-Ala dipeptide was prepared by azaannulation of 296 with 352 (Scheme 30).^^ Conversion to the aromatic ring system 364 was accompHshed by oxidation with DDQ, and hydrolysis of the substituent functionality gave the amino acid 365. 4.3

Vinylogous Urea Substrates Vinylogous urea substrates have also been used in the aza-annulation reaction to form 6-

lactam products. This process was illustrated by the condensation of P-keto amide 366 followed by aza-annulation with acryloyl chloride to give 367 (eq. 75).^'^'^^ Catalytic reduction of the tetrasubstituted double bond led to stereoselective formation of 368. The products formed in this reaction were p-enamino peptide units, and this chemistry can be extended to the preparation of triand tetrapeptide analogs.

1) BnNHs, TsOH Benzene, Reflux 2) Acryloyl Chloride THF, Reflux

^ ^ . - • ^

H2, Pd/C

(Eq. 75)

^^

53%

56%

366 367

H.

Ph

[Ref. 57 and 59]

1) BnNH2, BF3»OEt2 Benzene, Reflux 2) 352, THF

O

H

n

K>

N^'^Y Y ^

O

H

O II

H '

Reflux 76%

7^ Y? 7 EtO^ J^.,J^^

25 °C

421 ÔEt

HOEt

422

O

Scheme 34.

Aza-annulation was more efficient with pyridine substrates activated by the presence of ester functionality. As an example, the reaction of reagent 421 with ester stabilized pyridine substrate

370 423 resulted in a significantly higher yield than with the analogous ketone substrate 418 (eq. 86).102 O u

O u

Etc 421

OEt

25 °C NaOEt HOEt

O

O (Eq.86)

86%

[Ref. 102] O

O (Eq.87)

47 "ôPt 180°C

EtO^Ô

R = H; 52%, 67% R = Me; 80%

4 Steps

EtO^Ô 425: R = H 426: R = Me

423: R = H 424: R = Me

12% (From 425)

OH 232 (±)-Lupinine

[Ref. 103 and 104]

Treatment of 423 with 421, again at an elevated temperature (180 °C), led to the formation of 424 (eq. S1)A^^^^^^ In the case of the ester-stabilized pyridine substrates, substitution with an electron donating substituent such as a methyl group significantly enhanced the yield, as observed for the conversion of 424 to 426. Compound 425 was converted to (±)-lupinine by reduction and deprotection. 1) DMF, NaH 1 h, 60 °C 2) AcOH, Reflux 72h

(Eq.88)

MeO'

427: Y = OH 428: Y = N02

429: Y = OH 430: Y = N02

Doubly activated acrylonitrile reagents have also been utilized in the aza-annulation of pyridine substrates. Aza-annulation of 423 with 427 or 428 was performed at a mild temperature

371 (60 °C) by generation of the corresponding enolate with NaH (eq. 88).^^^ Through variation of the aromatic substituent Y, derivatives 429, 430, and others were prepared. Alternatively, 423 could be treated with 431 at 120 °C to give 432, which was converted to a number of heterocycle substituted derivatives related to 433 (eq. 89). 1^5 j ^ all of these cases, derivatives of 429, 430, and 433 exhibited selective inhibitory activity against IgE-antibody formation. As a result, these compounds have potential for treatment of diseases such as allergic rhinitis, atopic dermatitis, allergic bronchial asthma, and hypersensitiveness.^^^

||

SMe

MeCN Reflux 10 h

433

A nitro-substituted acrylate derivative 434 has also been used as an aza-annulation reagent. Treatment of 423 with 434 resulted in the formation of 435, which was reduced to the a-NH2 derivative 436 (eq. 90).^^^ When the pyridine substrate was substituted in the 6-position, reaction proceeded through an alternate pathway, and aza-annulation did not occur. ^^^ O NO2

EtO'

(Eq. 90) NH2

Zn, HCi

OEt

434

42%

45%

A more efficient approach to a-NHR substituted carbonyl derivatives was through the use of 347 (eq. 91). Aza-annulation of 423 with 347 produced 437 in good yield.107

(Eq.91) AcOH 79% EtO Ô 423 [Ret. 107]

EtO 347

437

372 Similar reactivity was observed for the reaction of 347 with nitrile activated substrates. Treatment of 438 with 347 generated the a-NHR substituted carbonyl derivative 439 (eq. 92). ^^'^

a 438

(Eq.92) AcOH 68%

0 ^ '

N

EtO

347

[Ref. 107]

(Eq.93) AcOH, 4 h

»

79% (From 438) 438

440

N

[Ref. 108]

An interesting reagent for aza-annulation was the acrylate-type reagent 440. When 438 was treated with 440, formation of 441 resulted (eq. 93).^^^ Subsequent reaction in AcOH resulted in cyclization to complete the aza-annulation process. An example of aza-annulation that involved the use of this reagent with a substrate activated by NO2 substitution, 443, was also reported (eq. 94) 108 Formation of 444 resulted, and cyclization gave 445 in low yield. (Eq. 94)

CI AcOH, 4 h

CI O 443

NO2

»

22% (From 443)

440

NO2

[Ref. 108]

EtO" 47 446: Y = OEt 447: Y = NH2 [Ref. 109]

0 M 'ÔEt OEt

NaOEt, HOEt 25°C, 20 °C^

^'vlJkJ^

Scheme 38.

The angular methyl derivative of 143 was prepared stereoselectively through a related approach (Scheme 38).^^^ In this reaction sequence, condensation and aza-annulation of the methylated substrate 138 led to diastereoseiective formation of 471 through asymmetric carboncarbon bond formation (25:1). Selective reduction of 471 generated 472 along with the cis isomer (6:1), and 472 was methylated to give 140.^ ^^ 8.2

Vinylogous Amide Substrates Enders and coworkers reported studies in which the RAMP and SAMP chiral auxiliaries were

employed in the aza-annulation process (Scheme 39).^ ^"^ Condensation of 179 with RAMP provided a route to the optically active enamino hydrazone 473, which was then metalated with nBuLi to generate the corresponding anion. Aza-annulation of 473 with 474 produced intermediate 475, which could be cyclized slowly (2 d) at 60 °C to give 476.

Alternatively, heterocycle

formation could be facilitated by an increase in reaction temperature (toluene, heat). Removal of the chiral auxiliary gave 477 in 50-52% overall yield from 179 in >99:1 enantiomeric purity. Substituents on the aromatic ring did not have a measureable effect on the yield of the aza-annulation reaction.

377

ri

•A„

1)n-BuLi,TMEDA THF, -78 °C 2) Ar OMe -78 °C

^^^''ÔMe

OMe r^-ô MeO ^ O 474 3) NH4CI, -30 °C

RAMP

*N'

Ar

OMe

^ - j s ^ 473

179

475 or Toluene Reflux

60 °C 2cl R

Ar

179-> 476

a

H

Ph

50%

b

Me

Ph

50%

c

Me

4-MeO-C6H4

52%

COgMe

R

OMe Zn, AcOH Reflux 6-12 h

C

V^^' O

O OMe

>99.i Enantioselectivity Scheme 39.

[Ret. 114]

8.3

476

Vinylogous Carbamate Substrates Stereoselective generation of quaternary carbon centers has been studied for a variety of

tetrasubstituted p-enamino ester substrates. Asymmetric enamines have been generated by condensation of either optically active 465 or an amino acid derivative with the corresponding p-keto ester. Formation of 479 from 478, followed by treatment with acryloyl chloride (64) gave 480 in good yield with highly diastereoselective formation of the 5-lactam (eq. 96).^ ^^ The high degree of stereoselectivity observed in product formation resulted from (1) the geometry of the enamine which was fixed by the constraints of the ring, and (2) the relative steric demands of the Me and Ph of the chiral auxiliary.

0

0

A 1 478

[Ref. 115]

H NH2 (R)-465 EtaO'BFs Benzene Reflux

Ph ^;?

H^N'^ 0

64 THF, Reflux

479 a b

0 1 2

O ^ " " ' ' 480 478-^480 76% 85%

Ratio 97:3 >97:3

378 Acyclic enamines were formed selectively as essentially a single geometric isomer through the stabilization that results from intramolecular hydrogen bonding. Aza-annulation of acyclic substrates such as 481a and 481b also resulted in the formation of 5-lactam products with high diastereomer ratios (eq. 97).^ ^^ Similar reactivity and stereoselectivity was observed with the analogous lactone derivative 484 in the formation of 486 (eq. 98). 11^ u

Ph ^^ H^NHs {R)-465 ^-^\ÔEt 481

(Eq.97) H^N'^ O

Et20»BF3 Benzene Reflux

64 OEt

Y 482

Afii -:.4a3 92% 58%

a Me b OBz

[Ret. 115]

OEt

THF, Reflux

Ratio

97:3 92:8

u

Ph ,^

V

H'^NH2


(Eq.98)

CI

(R)-465

64 L ^ 485

THF, Reflux 80% (From 484)

[Ref. 115]

1)

O

R^R2

487 V H NH2 Et20»BF3 Benzene, Reflux

O

(Eq. 99)

OEt 2)

O

478b THF, Reflux [Ref. 115]

a a b

Ri R2 C02Et Ph C02Et Ph iPr C02Me

Ratio (488:489) 79:21 98:2 43:57

Temp 66 °C -33 °C 66 °C

Yield 63% 77% 43P/o

Amino acid derivatives have also been explored as potential chiral auxiliaries in the asymmetric aza-annulation reaction. As reported for the Michael addition to acrylate derivatives, the reaction outcome has also shown sensitivity to the special balance of complementary steric demands of the methyl and phenyl substituents. The degree of diastereoselectivity in carbon-carbon bond

379 formation was dependent on the presence of a phenyl substituent. When the methyl group of phenethylamino auxiliary (465) was replaced with an ester group, as in the phenyl glycine derivative 487, stereoselectivity in the formation of 488a and 489a dropped considerably (79:21), and decreased reaction temperatures were required to achieve selective product formation (eq. 99).^^^ Further alterations in the source of asymmetry, through the use of the valine derivative b, led to minimal induction of asymmetry in the generation of 488b and 489b. Substituted acrylate derivatives have also been employed in the asymmetric aza-annulation reaction (Scheme 40). Aza-annulation of 479b with crotonyl chloride (490) demonstrated several important features of this reaction.l ^^ First, concomitant formation of two stereogenic centers gave 491 with high internal asymmetric induction, while high relative asymmetric induction resulted from the amine substituent. However, the presence of a methyl substituent at the p-position of the acrylate derivative slowed the reaction significantly, and resulted in a poor yield.

O

O

(^VÔEt k ^ 478b

H NH2 {R)-A65 Et20*BF3 Benzene Reflux

V

H^N'^0 479b 352, THF, -33 °C

492 THF, -33 °C 2)NaH

(73:27)

1) (52:48) 2) (83:17)

Substitution at the a-position of the acrylate derivative did not appear to significantly change the aza-annulation process, and formation of the quaternary center was highly stereoselective (Scheme 40).^^^ In contrast to the observations for the formation of 491, stereochemical control at the 3-position of the 5-lactam was only moderate. Annulation of 479b with methacryloyl chloride

380 (492) resulted in the formation of 493 with >98:2 stereoselective generation of the quaternary center, while a 52:48 kinetic ratio of isomers at C-3 resulted. Equilibration of C-3 by treatment with NaH provided an increased diastereomer ratio of 83:17. Slightly higher diastereoselectivity was obtained in the formation of 494. Modification of the products that resulted from the aza-annulation of tetrasubstituted enamine substrates with acrylate derivatives was very limited. The aza-annulation of benzyl ester 496 with the mixed anhydride, a mixture (497) preformed from Et02CCl and sodium acrylate, provided a route to 498 in >98:2 diastereoselectivity (eq. 100), which allowed access to the carboxylic acid derivative 499 through catalytic hydrogenation. ^ ^^ Further elaboration of either the ester or the acid derivative was unsuccessful, possibly due to the steric congestion around the reactive functionality. Extended hydrogenation did not reduce the enamine functionality, as observed in related substrates, and 498 was relatively stable to acidic hydrolysis conditions.

In addition, DCC (A^,A^'-

dicyclohexylcarbodiimide) coupling of acid 499 with either benzyl amine or glycine ethyl ester was unsuccessful. (Eq. 100)

^) % / (fl).465 H NH2 O

3 atm of H2 Pd/C EtOH

Et20»BF3 Benzene, Reflux

O

OBn

^

98%

EtOaCCI +

OH

497

496

THF, Reflux 70%

[Ref. 116]

(>98:2)

8.4

Vinylogous Urea Substrates Studies of aza-annulation reaction with tetrasubstituted P-keto amide substrates have also been performed. Investigations centered around those substrates that were analogous to the ester species described in Section 8.3. In general, the amide substrates were found to react 20-25% slower than their ester counterparts, and as a result, greater diastereoselectivity was observed.^^^ 1)

6" O

O

R^ R2

V

0

R\R2

H ^ N ^

H-^N

R\R2

H NH2 Et20«BF3 Benzene, Reflux

NHBn

NHBn 2)497 THF, Reflux

^ ^

500

[Ref. 116]

a b c d

0

R2 R1 Ph Me C02Et Ph iPr C02Me Bn C02Et

"*"

501 Ratio (501:502) >98:2 2:>98 >98:2 >95:5

0

(Eq.

^

NHBn ^

0

Yield 99% 96% 90% 46%

502

381 Condensation and aza-annulation of 500 provided 501 as a single observable isomer in excellent yield (eq. 101).î^ The (/?)- or (5)-stereoisomer of the quaternary carbon could be obtained depending on the chiral auxiliary used, and for each auxiliary, high selectivity was obtained. In the case of the phenylalanine derivative d, a lower yield was obtained, possibly due to the decreased steric protection of the enamide functionality which then allowed electrophilic reactions such as hydrolysis to occur. The analogous five-membered ring substrate 503 also gave highly stereoselective reaction upon condensation with (/?)-465 and subsequent aza-annulation to give 505 (>98:2), but the yield for this reaction was low due to the sensitivity of 505 to hydrolysis (eq. 102).

^

(f?)-465

o NHBn

503

H-^^N'^ O


\

I

(Eq.102)

487 THF, Reflux NHBn

50% (From 503)

505

^ - ^

504

(>98:2)

[Ret. 116]

Similarly, aza-annulation with an acyclic substrate resulted in a high degree of stereocontrol. These results suggested that intramolecular hydrogen bonding of the intermediate enamine controlled the enamine geometry and served to restrict rotation of the chiral auxihary (eq. 103).^ ^^ In this case, 507 was sensitive to hydrolysis, and isolation was performed after hydrolysis to 508.

1) H'^^NHs Ph ^^

(FJ)-465

u

Et20*BF3 Benzene, Reflux

u NHBn

2)497 THF, Reflux

506

H2O pTsOH

O

(Eq.103)

H-^N' NHBn

NHBn 82% (From 506)

508

(>98:2)

[Ref. 116]

Reaction of 500 with (/?)-465 followed by aza-annulation with a substituted acrylate derivative 352, gave 509 (eq. 104).^ ^^ Although the quaternary carbon was formed almost exclusively as the R isomer, an equimolar mixture of the substituent at C-3 of the 5-lactam resulted.

382

1) H ^ N H s {R)-A65 EtaO'BFs Benzene, Reflux

6" O

O

(50:50) (Eq. 104)

.Ph. . c Q / H

2) 352 THF. Reflux

tfr

NHBn

67% (From 500)

500 [Ret. 116]

O NHBn 509

8.5 p-Imino Sulfoxide Substrates The use of P-imino sulfoxide substrates has followed a different strategy than other asymmetric aza-annulation reactions. Instead of generating a stereogenic center during the azaannulation process, a chiral sulfoxide was used to modify the 5-lactam product in an asymmetric fashion after formation of the heterocyclic ring. The use of p-imino sulfoxide substrates led to a number of appUcations in natural product synthesis. For example, 510 was deprotonated to generate the corresponding a-sulfinyl ketimine anion, and addition of methyl acrylate resulted in the formation of 511 (eq. 105).^^"^'^^^ The next step involved stereoselective reduction of the enamide functionality with NaCNBHa to give 512 as the only diastereomer. Final reductive removal of the sulfoxide functionality and reduction of the carbonyl gave (/?)-(-)-indolizidine (513).

1)LDA 2) -30 °C-25 °C 2h, O MeO " \ rr II 37

510

NaCNBHa AcOH CF3CO2H

5 y ^ N - ^

1)25°C.2h 2)50°C.4h^

'if / ^ ^ ^

i7^

1) Raney-Ni EtOH 2) LiAIH4 83%

(Eq. 105)

CO H

513

[Ref. 117 and 118]

Further extension of this methodology demonstrated that cyclic acrylate derivatives could be used to construct tricyclic ring systems with the formation of stereogenic centers during azaannulation (Scheme 41).ll7,ll8 When treated with 514, the cis to trans ring fusion obtained for product formation was 70:2 for 515:516. Compound 515 was reduced to 517. Although azaannulation with 518 gave slightly lower selectivity in the formation of 519 and 520, formation of the cis ring fusion was still favored, and a good selectivity was obtained.

383

1) A?-BuLiorLDA 2) -78 °C- 25 °C

1) /7-BuLiorLDA 2) -78 °C- 25 °C

jy^-° xr^f'o 515

70%

71%

H

516

jy'-° jy^-f--o

2>/o

519

60%

520

1) NaCNBHa AcOH 2) Raney-Ni EtOH, Reflux

H

517

[Ret. 117 and 118]

Scheme 41.

The use of 521 led to the synthesis of a number of natural products through the azaannulation of this P-imino sulfoxide (Scheme 42).^ ^'^'^ ^^ Application of methyl aery late in this azaannulation process led to the formation of 522. The chiral auxiliary was then used to provide moderate stereochemical control in the reduction of the enamide alkene to 523 and 524 in a 1.9:1.0 ratio. Compound 523 was then reduced to remove both the sulfoxide and the lactam carbonyl to give (-)-l,2,3,4,6,7,12,12b-octahydroindolo[2,3-a]quinolizine (525).^^'^'^^8

384 1)LDA,THF -78 °C, 1 h 2) - 25 °C, 4 h 37

J JT/c

90%

NaCNBHs AcOH 25 °C, 3 h

1) Separation 2) Raney-Ni THF/EtOH 65 °C 3) LiAIH4, 25 °C 525 H-1,2,3,4,6,7,12,12b-0ctahydroindolo[2,3-a]quinolizine

75%

^'^ •• '-^^ r

\ 523

[Ret. 117 and 118]

ii 524

Scheme 42.

The reaction of 521 with 518 has been used to generate 526, a very versatile pentacychc intermediate in the synthesis of natural products (Scheme 43). ^ 17,118 Removal of the sulfoxide gave 528, which could be further reduced to give (-)-alloyohimban 529. Alternatively, 528 could be treated with LDA to cause epimerization of the stereogenic center a to the lactam carbonyl, and subsequent reduction gave (-)-yohimban 530. Initial reduction of 526 in the presence of the chiral auxiliary, was found to give 531 in slight preference to 532, and 531 could be reduced under standard protocol for these molecules to give (+)-3-ep/-alloyohimban (533).11'7'11^

385 1)LDA,THF 2) 25 °C, 1 h 60 °C, 14 h 518

35% (Recovered 521)

NaCNBHa X 80% A c O H / (From 526)

88% \ Raney-Ni (From 526) \ EtOH

Raney-Ni EtOH

^-

89% (From 532)

1)LDA 2) AcOH 3) LiAIH4

533 (+)-3-ep/-Alloyohimban [Ref. 117 and 118]

529 (-)-Alloyohimban

43%

530 (-)-Yohimban

Scheme 43.

Substituted acrylate derivatives have also been employed in the asymmetric synthesis of a natural product. In a model study, the deprotonation of 510 and aza-annulation with 534 led to a 60:40 mixture of 535 and 536 (eq. 106). ^ 1^

386 1)n-BuLi 2) O H MeO^"^

O

Y

O

H ÔtBu

OtBu

H ,N„.OtBu

To

534

34%

(Eq. 106) 535

510

536

(60:40)

[Ref. 119]

This methodology was applied to the substituted analog 537, which also gave a 60:40 ratio of diastereomeric products in 55% yield (Scheme 44). 1^^ In this case, isomers 538 and 539 could be separated and then carried through the same sequence of parallel steps to give (-)-slafraniine (540) and (-)-6-epislaframine (541). 119

1)/>BuLi 2) O H MeC

O

OtBu

H

O

H

M

ÔtBu

Y o

537

538

(60:40)

539

SiRs = Si(Me)2tBu 4 Steps

28% (From 538)

4 Steps

13P/o (From 539)

.»NH2 [Ref. 119]

^ AcO 540 (~)-Slaframine Scheme 44.

AcO 541 (-)-6-Epislaframine

OtBu

387

9. (1)

(2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38)

REFERENCES (a) Ninomiya, L; Naito, T. Heterocycles 1981,75,1433. (b) Ninomiya, L; Miyata, O. In Studies in Natural Products Chemistry, Atta-ur-Rahman, Ed.; Elsevier: New York, 1989; Vol. 3, p. 399. (c) Ninomiya, L; Naito, T. In The Alkaloids; Brossi, A., Ed.; Academic Press: New York, 1983; Vol. 22, p. 189. Vill, J. J.; Steadman, T. R.; Godfrey, J. J. / Org. Chem. 1964, 29, 2780. Kuehne, M. E.; Bommann, W. G.; Parsons, W. H.; Spitzer, T. D.; Blount, J. P.; Zubieta, J. J. Org. Chem. 1988, 55, 3439. Chelucci, G.; Cossu, S.; Scano, G.; Soccolini, F. Heterocycles 1990, 57, 1397. Murahashi, S.-L; Sasao, S.; Saito, E.; Naota, T. J. Org. Chem. 1992, 57, 2521. Tominaga, Y.; Kawabe, M.; Hosomi, A. / Heterocycl Chem. 1987, 24, 1325. Stork, G. Pure and Appl Chem. 1968,77,383. Ninomiya, L; Naito, T.; Higuchi, S.; Mori, T. J. Chem. Soc, Chem. Commun. 1971, 457. El-Barbar>% A. A.; Carlsson, S.; Lawesson, S.-O. Tetrahedron 1982, 38, 405. Hickmott, P. W.; Rae, B.; Pienaar, D. H. S. Afr. J. Chem. 1988, 41, 85. (a) Xia, Y.; Kozikowski, A. P. J. Am. Chem. Soc. 1989, 777, 4116. (b) Kozikowski, A. P.; Xia, Y.; Reddy, E. R.; Tuckmantel, W.; Hanin, L; Tang, X. C. J. Org. Chem. 1991, 56, 4636. Stork, G.; Kretchmer, R. A.; Schlessinger, R. H. J. Am. Chem. Soc. 1968, 90, 1647. Paronikyan, E. G.; Sirakanyan, S. N.; Lindeman, S. V.; Aleksanyan, M. S.; Karapetyan, A. A.; Noravyan, A. S.; Struchkov, Y. T. Chem. Heterocycl. Compd. (USSR) (Engl. Transl.) 1990,25,953 (Khim. Geterotsikl Soedin. 1989, 1137). (a) Sammour, A.; Alkady, M. Ind J. Chem. 1974, 72, 51. (b) El-Kady, M.; El-Hashash, M. A.; Sayed, M. A.; El-Sherif, M. Ind J. Chem., Sect. B 1981, 20, 491. Briet, P.; Berthelon, J.-J.; Depin, J.-C. European Patent 0 000 306, 1979. Chem. Abstr. 1979,97:20479v. Elgemeie, G. E. H.; Elghandour, A. H. H. Bull. Chem. Soc. Jpn. 1990, 63, 1230. Kambe, S.; Saito, K.; Sakurai, A.; Hayashi, T. Synthesis 1977, 841. Takahashi, K.; Miyake, A.; Hata, G. Bull. Chem. Soc. Jpn. 1972, 45, 2212. Norman, M. H.; Heathcock, C. H. J. Org. Chem. 1988, 53, 3370. Gardette, D.; Gramain, J.-C; Sinibaldi, M.-E. Heterocycles 1990, 57, 1439. Borne, R. P.; Fifer, E. K.; Waters, I. W. J. Med Chem. 1984, 27, 1271. Wenkert, E.; Chauncy, B.; Dave, K. G.; Jeffcoat, A. R.; Schell, F. M.; Schenk, H. P. J. Am. Chem. Soc. 1973, 95, 8427. Ito, K.; Yokokura, S.; Miyajima, S. J. Heterocycl. Chem. 1989, 26,111>. Kmetic, M.; Stanovnik, B.; Tisler, M.; Kappe, T. Heterocycles 1993, 35, 1331. (a) Meyers, A. I.; Reine, A. H.; Sircar, J. C ; Rao, K. B.; Singh, S.; Weidmann, H.; Fitzpatrick, M. J. Heterocycl. Chem. 1968,5,151. (b) Horii, Z.-i.; Iwata, C ; Ninomiya, I.; Imamura, N.; Ito, M.; Tamura, Y. Chem. Pharm. Bull. 1964, 72, 1405. Shabana, R.; Rasmussen, J. B.; Olesen, S. O.; Lawesson, S.-O. Tetrahedron 1980, 36, 3047. Kozikowski, A. P.; Reddy, E. R.; Miller, C. P. J. Chem. Soc, Perkin Trans. 1 1990, 195. Paulvannan, K.; Stille, J. R. J. Org. Chem. 1992, 57, 5319. Ninomiya, I.; Naito, T.; Higuchi, S. J. Chem. Soc, Chem. Commun. 1970, 1662. Hickmott, P. W.; Sheppard, G. J. Chem. Soc (C) 1971, 1358. (a) Rigby, J. H.; Balasubramanian, N. /. Org. Chem. 1984, 49, 4569. (b) Rigby, J. H.; Qabar, M. Synth. Commun. 1990, 20, 2699. (a) Dickman, D. A.; Heathcock, C. H. J. Am. Chem. Soc 1 9 8 9 , 7 7 7 , 1528. (b) Heathcock, C. H.; Norman, M. H.; Dickman, D. A. J. Org. Chem. 1990, 55, 798. (a) Aranda, V. G.; Barluenga, J.; Gotor, V. Tetrahedron Lett. 1973,2819. (b) Barluenga, J.; Muniz, L.; Palacios, F.; Gotor, V. J. Heterocycl. Chem. 1983, 20, 65. Aranda, V. G.; Barluenga, J.; Gotor, V. Tetrahedron Lett. 1974, 977. Janin, Y. L.; Bisagni, E.; Carrez, D. J. Heterocycl. Chem. 1993, 30, 1129. Kametani, T.; Terasawa, H.; Ihara, M. J. Chem. Soc, Perkin Trans. I 1976, 2547. Kametani, T.; Terasawa, H.; Ihara, M.; Fukumoto, K. Heterocycles 1977,6,37. Ihara, M.; Noguchi, K.; Ohsawa, T.; Fukumoto, K.; Kametani, T. Heterocycles 1982, 79, 1829.

388 (39) Kametani, T.; Suzuki, Y.; Terasawa, H.; Ihara, M. / Chem. Soc, Perkin Trans. I 1979, 1211. (40) Kametani, T.; Suzuki, Y.; Ihara, M. Can. J. Chem. 1979, 57, 1679. (41) Bhattacharjya, A.; Bhattacharya, P. K.; Pakrashi, S. C. Heterocydes 1983, 20, 2397. (42) (a) Danieli, B.; Lesma, G.; Palmisano, G. /. Chem. Soc, Chem. Commun. 1980, 109. (b) Danieli, B.; Lesma, G.; Palmisano, G. Gazz. Chim. Ital. 1981, 111, 257. (43) Danieli, B.; Lesma, G.; Palmisano, G.; Tollari, S. Synthesis 1984, 353. (44) (a) Kametani, T.; Kanaya, N.; Ihara, M. Heterocydes 1981, 76, 925. (b) Kametani, T.; Kanaya, N.; Hino, H.; Huang, S.-P.; Ihara, M. J. Chem. Soc, Perkin Trans. I 1981, 3168. (45) Kametani, T.; Kanaya, N.; Hino, H.; Huang, S.-P.; Ihara, M. Heterocydes 1980, 14, 1771. (46) Calabi, L.; Danieli, B.; Lesma, G.; Palmisano, G. Tetrahedron Lett. 1982,25,2139. (47) Kametani, T.; Ohsawa, T.; Ihara, M. J. Chem. Soc, Perkin Trans. I 1981, 1563. (48) Corriu, R. J. P.; Perz, R. Tetrahedron Lett. 1985,26,1311. (49) Singh, B. Synthesis 1985, 305. (50) Singh, B. U.S. Patent 4 347 363,1982. Chem. Abstr. 1982, 97:216018n. (51) Tanabe Seiyaku KK Japanese Patent J4-8 023 779, 1984. (52) Horii, Z.>I.; Iwata, C ; Tamura, Y.; Nelson, N. A.; Rasmusson, G. H. J. Org. Chem. 1964, 29, 2768. (53) (a) Cannon, J. G.; Hatheway, G. J.; Long, J. P.; Sharabi, F. M. J. Med. Chem. 1976,19, 987. (b) Cannon, J. G.; Suarez-Gutierrez, C ; Lee, T.; Long, J. P.; Costall, B.; Fortune, D. H.; Naylor, R. J. J. Med. Chem. 1919,22, 341. (c) Cannon, J. G.; Hamer, R. L.; Ilhan, M.; Bhatnagar, R. K.; Long, J. P. J. Med. Chem. 1984, 27, 190. (d) Cannon, J. G.; Chang, Y.; Amoo, V. E.; Walker, K. A. Synthesis 1986, 494. (54) Jones, C. D.; Audia, J. E.; Lawhom, D. E.; McQuaid, L. A.; Neubauer, B. L.; Pike, A. J.; Pennington, P. A.; Stamm, N. B.; Toomey, R. E.; Hirsch, K. S. J. Med. Chem. 1993, 36, 421. (55) Ninomiya, I.; Kiguchi, T. J. Chem. Soc, Chem. Commun. 1976, 624. (56) Hickmott, P. W.; Sheppard, G. / Chem. Soc (C) 1971, 2112. (57) Paulvannan, K.; Stille, J. R. Tetrahedron Lett. 1993,34,8197. (58) Cook, G. R.; Beholz, L. G.; Stille, J. R. J. Org. Chem. 1994, 59, 3575. (59) Paulvannan, K.; Stille, J. R. J. Org. Chem. 1994, 59, 1613. (60) Huizenga, R. H.; van Wiltenburg, J.; Pandit, U. K. Tetrahedron Lett. 1989,50,7105. (61) Augustin, M.; Frank, J.; Kohler, M. J. Prakt. Chem. 1984, 326, 594. (62) Singh, B.; Lesher, G. Y. J. Heterocyd. Chem. 1990, 27, 2085. (63) Lielbriedis, I. E.; Kampare, R. B.; Dubur, G. Y. Latv. PSR Zinat. Akad. Vestis., Kim. Ser. 1990,2, 212. (64) Paulvannan, K.; Stille, J. R. Tetrahedron Lett. 1993, 42, 6673. (65) (a) Leniewski, A.; Szychowski, J.; MacLean, D. B Can. J. Chem. 1981, 59, 2479. (b) Leniewski, A.; MacLean, D. B; Saunders, J. K. Can. J. Chem. 1981, 59, 2695. (66) Shono, T.; Matsumura, Y.; Kashimura, S. J. Org. Chem. 1981, 46, 3719. (67) (a) Wiesner, K.; Jirkovsky, I.; Fishman, M.; WiUiams, C. A. J. Tetrahedron Lett. 1967, 1523. (b) Wiesner, K.; Jirkovsky, I. Tetrahedron Lett. 1967, 2077. (c) Wiesner, K.; Poon, L.; Jirkovsky, I.; Fishman, M. Can. J. Chem. 1969, 47, 433. (68) (a) Sluyter, M. A. T.; Pandit, U. K.; Speckamp, W. N.; Huisman, H. O. Tetrahedron Lett. 1966, 87. (b) Dubas-Sluyter, M. A. T.; Speckamp, W. N.; Huisman, H. O. Rec Trav. Chim. Pays-Bas. 1972, 91, 157. (69) Wolf, U.; Sucrow, W.; Vetter, H.-J. Z. Naturforsch. 1979, 34b, 102. (70) Marcos, A.; Pedregal, C ; Avendano, C. Tetrahedron 1994, 50, 12941. (71) Barluenga, J.; Jardon, J.; Gotor, V. Synthesis 1988, 146. (72) Barluenga, J.; Iglesias, M. J.; Gotor, V. Synthesis 1987, 662. (73) SchroU, G.; Klemmensen, P.; Lawesson, S.-O. Ark. Kemi. 1967,26, 317. (74) Paulvannan, K.; Schwarz, J. B.; Stille, J. R. Tetrahedron Lett. 1993,54,215. (75) Cook, G. R.; Beholz, L. G.; Stille, J. R. Tetrahedron Lett. 1994, 35, 1669. (76) Knoevenagel, E.; Fries, A. Chem. Ber. 1989, 31, 761. (77) Sanna, P.; Nuvole, A.; Sequi, P. A.; Paglietti, G. Heterocydes 1993, 36, 259. (78) Capuano, L.; Boschat, P.; Miiller, I.; Zander, R.; Schranmi, V.; Hadicke, E. Chem. Ber. 1983,776, 2058.

389 (79) Danishefsky, S.; Etheredge, S. J.; Volkmann, R.; Eggler, J.; Quick, J. J. Am. Chem. Soc. 1971, 93, 5575. (80) Volkmann, R.; Danishefsky, S.; Eggler, J.; Solomon, D. M. J. Am. Chem. Soc. 1971, 95, 5576. (81) Heber, D. Arch. Pharm. 1987,520,445. (82) Ziegler, E.; Hradetzky, P.; Belegratis, K. Monatsh. Chemie 1965, 96, 1347. (83) Dannhardt, G.; Meindl, W.; Schober, B. D.; Kappe, T. Eur. J. Med. Chem. 1991, 26, 599. (84) Ried, W.; Batz, F. Liebigs Ann. Chem. 1972, 762, 1. (85) Brunerie, P.; Celerier, J.-P.; Huche, M.; Lhommet, G. Synthesis 1985, 735. (86) Nagasaka, T.; Inoue, H.; Hamaguchi, F. Heterocycles 1983, 20, 1099. (87) Nagasaka, T.; Inoue, H.; Ichimura, M.; Hamaguchi, F. Synthesis 1982, 848. (88) Danishefsky, S.; Etheredge, S. J. J. Org. Chem. 1974, 39, 3430. (89) Shen, W.; Cobum, C. A.; Bommann, W. G.; Danishefsky, S. J. J. Org. Chem. 1993, 58, 611. (90) Fang, F. G.; Danishefsky, S. J. Tetrahedron Lett. 1989, 30, 3621. (91) Kawahara, N.; Nakajima, T.; Itoh, T.; Ogura, H. Synthesis 1985, 644. (92) Kawahara, N.; Nakajima, T.; Itoh, T.; Ogura, H. Heterocycles 1983, 20, 1721. (93) Chiba, T.; Takahashi, T. Chem. Pharm. Bull. 1985, 33, 2731. (94) (a) Beholz, L. G.; Ph.D. Thesis, Michigan State University, 1994. (b) Barta, N. S.; Stille, J. R. Unpublished results. (95) Seidel, M. C. J. Org. Chem. 1972, 37, 600. (96) Sato, M.; Yoneda, N.; Kaneko, C. Chem. Pharm. Bull. 1986, 34, 621. (97) Singh, B.; Lesher, G. Y.; Brundage, R. P. Synthesis 1991, 894. (98) Kappe, C. O.; Kappe, T. Monatsh. Chemie 1989,120, 1095. (99) Yamada, Y.; Hatano, K.; Matsui, M. Agr. Biol. Chem. 1970, 34, 1536. (100) Openshaw, H. T.; Whittaker, N. J. Chem. Soc. 1961, 4939. (101) Thyagarajan, B. S.; Gopalakrishnan, P. V. Tetrahedron 1965, 21, 3305. (102) Thyagarajan, B. S.; Gopalakrishnan, P. V. Tetrahedron 1965,27,945. (103) Boekelheide, V.; Lodge, J. P., Jr. J. Am. Chem. Soc. 1951, 73, 3681. (104) Bohlmann, V. F.; Ottawa, N.; Keller, R. Liebigs Ann. Chem. 1954, 587, 162. (105) Kurashina, Y.; Miyata, H.; Momose, D.-I. European Pat 309 260, 1989. Chem. Abstr. 1989,777:153656. (106) Thyagarajan, B. S.; Gopalakrishnan, P. V. Tetrahedron 1964, 20, 1051. (107) Kolar, P.; Tisler, M. J. Heterocycl. Chem. 1993,50, 1253. (108) Forti, L.; Gelmi, M. L.; Pocar, D.; Varallo,M. Heterocycles 1986, 24, 1401. (109) Adams, R.; Reifschneider, W. J. Am. Chem. Soc. 1959, 81, 2537. (110) Tonetti, I.; Primofiore, G. 11 Farmaco 1980, 35, 1052. (111) (a) Sevin, A.; Masure, D.; Giessner-Prettre, C.; Pfau, M. Helv. Chim. Acta 1990, 73, 552. (b) d'Angelo, J.; Desmaele, D.; Dumas, F.; Guingant, A. Tetrahedron: Asymm. 1992, 5, 459. (c) Pfau, M.; Tomas, A.; Lim, S.; Revial, G. J. Org. Chem. 1995, 60, 1143. (112) d'Angelo, J.; Guingant, A.; Riche, C; Chiaroni, A. Tetrahedron Lett. 1988, 29, 2667. (113) Audia, J. E.; Lawhom, D. E.; Deeter, J. B. Tetrahedron Lett. 1993, 34, 7001. (114) Enders, D.; Demir, A. S.; Puff, H.; Franken, S. Tetrahedron Lett. 1987, 28, 3795. (115) Barta, N. S.; Brode, A.; Stille, J. R. /. Am. Chem. Soc. 1994, 116, 6201. (116) Benovsky, P.; Stille, J. R. Unpublished results. (117) Hua, D. H.; Bharathi, S. N.; Takusagawa, F.; Tsujimoto, A.; Panangadan, J. A. K.; Hung, M.-H.; Bravo, A. A.; Erpelding, A. M. J. Org. Chem. 1989, 54, 5659. (118) Hua, D. H.; Bharathi, S. N.; Panangadan, J. A. K.; Tsujimoto, A. J. Org. Chem. 1991, 56, 6998. (119) Hua, D. H.; Park, J.-G.; Katsuhira, T.; Bharathi, S. N. J. Org. Chem. 1993, 58, 2144.



391

Selective Reactions and Total Synthesis of Inositol Phosphates Yutaka Watanabe

1.

INTRODUCTION Biologically and structurally interesting natural products have

stimulated chemists to accomplish their total synthesis. For this purpose various useful synthetic and analytical methodologies have been devised. Such developments have resulted in the realization of efficient total syntheses of these molecules and their analogues. These achievements have greatly contributed in elucidating

their

physiological roles, understanding interactions of substrates with their recognized proteins

(enzymes) at the molecular

level, and

creating useful medicinal substances. Inositol derivatives present a good

representative

example

of

these

years. The structure of inyo-inositol configurational

isomers

of

developments

in

recent

(1), which is one of the nine

inositols

(1, 2 , 3 , 4 , 5,6-cyclohexane-

hexaols) and which is most widely distributed in its derivatives in nature (in animals, plants, and most bacteria), is illustrated below where three types of structures are given (Scheme 1-1) . HO OH

«oj"

l_lc J LOH OH

myo-lnositol

Scheme

1-1

Neurotransmitters, photons and hormones stimulate receptors on the surface of the target cells causing mobilization of the calcium ion

in

intracellular

stores,

thereby

triggering

physiological

responses to occur. The physiological process had been known but it was not clear how the outer information

(first messengers), which

cannot permeate the cell membrane, was transmitted to the calcium stores.

In

1983,

experimentally

this

that

question

was

D-myo-inos i t ol

resolved

by

showing

1 , 4 , 5 - t r i sphosphate

392 [ Ins (1,4 , 5 ) P3 / 2, transmittance

by

second

stimulating

messenger] cellular

mediates

calcium

the

signal

stores.^

Thus,

Ins(1,4,5) P3 is released to cytosol in cells by receptor-regulated hydrolysis of phosphatidyl-inyo-inositol 4,5-bisphosphate [PI(4,5)P2/ 3} in plasma membrane and stimulates a calcium (Ca^^) store by way of the receptor resulting in liberation of calcium ion which causes a

variety

of

biological

responses

(metabolism,

secretion,

contraction, neural activity, and cell proliferation).^ The other hydrolysis product, 1,2-di-O-acyl glycerol (DAG like 2-arachidonoyl1-stearoyl-sn-glycerol

shown in scheme 1-2) is also recognized to

act as a second messenger to stimulate protein kinase C

(PKC).^

These messengers of organic compounds were identified about 3 0 years after the discovery of cyclic AMP in 1957. ^ Disclosure of this signal

transduction

system has helped

to clarify

the processes

involved in various biological pathways. The present understanding (Scheme

of the metabolic cascade of PI(4,5)P2 is summarized below 1-2) .5 O II

*co'CO-i HO-

DAG

PLC

OH

•H03PO •H03PO. 5 OH O" PI(4,5)P2

•HO3PO •HO3PO 5 OH

OPOgH^

lns(1,4,5)P3 PI(4)P -

PI

1 Ins

I

>^^^/(l,4;

• (3.4)

I

insPg

lnsP2 '(1,4) >'^^(1.3)

' PI(4,5)P2

-^

lnsP4

(1.4,5) (1,4,5)

. * ^

> (1,3,4)

^ ^ ^

InsPc

(1.4,6) (3.4.6) InsPeP* (P*=pyrophosphate)

Scheme

1-2

Although inositol itself and many of its derivatives have been discovered over the years,^ the chemistry and biochemistry of inositol had been little investigated. However, in 1961, PI(4,5)P2

393 isolated from beef brain was already structurally characterized by Ballou'^ who described in the literature"^^ that phosphoinositides might be involved

"in the active transport of certain types of

molecules".^ In their work, the structure of Ins(1, 4, 5)P3 , which was obtained by the chemical hydrolysis of PI(4,5)P2, was confirmed. Ballou's

group

also

reported

phosphatidylinositols inositides

from

structure

elucidation

of

glycosyl

(GPIs), a series of mannosylated phospho(Scheme

mycobacteria

1-3) .^

recently, structurally similar glycosyl phosphatidylinositol

More (GPI)

anchors which hold membrane enzymes in the cell membrane through a covalent bond have been found although their physiological role is not clear yet

(Scheme

1-3) .^^

A GPI is also hydrolyzed by the

insulin action to the inositol phosphoglycan which seems to be a second messenger.

S3^

HO HO

HOHO ^ ^ —

•

HO

SliiJlTo

OH

HOHO-

-HS3^ HO

(R = fatty acid residue) Protein—C-N" H

^ P - 0 ^ OH HO ^ • HO HO HOHO HO ' O ^ ^ - ^

HO O^

-

0 Phosphatidylinositol pentamannoside

HO

O-^^HO-

HO.

HO / V n ' ^ e ^ ' " ' " ^

' V^/Cu° ^=^

^^-T?;A HOA.--^

HO

OH

o'^J^^^f-ô HO

Scheme

GPI anchor-protein

1-3

Furthermore, recent researches have shown that there is another type

of metabolic

tyrosine

pathway

kinase-linked

of

inositol

receptors

phospholipids

embedded

in the cell

where the membrane

394 participate. Thus, binding of growth factors and hormones such as insulin to the receptors causes activation of PLCyi or phosphatidylinositol 3-kinase (PI 3-kinase) which respectively hydrolyzes or phosphorylates PI(4,5)P2 resulting in the formation of Ins(1,4,5}P3 and DAG or PI{3,4,5)P3. In the trigger reaction of the old PI cycle described above, PLCpi is activated by the G-protein~ linked receptors in the plasma membrane resulting in the hydrolysis of PI(4,5)P2 (Scheme 1-4). Hormones G protein-linked receptors

acetylcholine histamine vasopressin

PLCft

Hormones (insulin) Growth factors

DAG

lns(1,4,5)P3

PI(4,5)P2

Tyrosine kinase-linked receptors

co-H •HOgPoi^'Ô

^mA

•HO3PO •HO3PO

P-0-

6-

OH

PK3,4.5)P3

Scheme

1-4

In organic chemistry, structurally characteristic features of inositol

phosphates

physiological chemists

to

and

related

compounds

characteristics as mentioned the

importance

of

as

well

above

inositol

have

as

their

awakened

chemistry

biochemistry. At present, the biological roles of many

and

inositol

derivatives are unclear. To disclose their functions, their chemical synthesis and analogues are quite useful. From the viewpoint of organic synthesis, they are structurally unique and challenging to synthetic chemists. These facts have directed researchers to prepare various inositol compounds. At earlier stages (around 1984) of the researches in the race to chemically synthesize Ins(1,4,5)P3 and related compounds, there were some problems to be solved: How to perform

multiple

including

phosphorylation

vicinally

of

situating

several

hydroxyl

polyhydroxyls;

groups

how

to

straightforwardly and conveniently protect inositol hydroxyls; how to

conveniently

derivative.

gain

access

to

an

optically

active

inositol

395 Problems

with

In the preparation of myo-

phosphorylation:

inositol phosphates and related compounds, the most crucial problem to be solved is the multiple phosphorylation of polyol derivatives. Especially vicinal diols 4 are very difficult to be transformed to the diphosphates 7, mainly because the monophosphorylation product 5 is prone to cyclization to the 5-membered cyclic phosphate 6 rather than undergoing facts

(Scheme

the second phosphorylation

stimulated

efforts

to

develop

a

new

1-5) . These

phosphorylation

methodology, and in 1987 two types of new phosphorylation methods employing P(III) and P(IV) reagents were successfully introduced for the synthesis of Ins(1,4,5)P3 and Ins(1,3,4,5)P4.

OH

K

OH

Schesne

As

well

as

the

exhaustive

1-5

phosphorylation

of

polyols,

a

regioselective partial phosphorylation of inositols is quite useful especially for introducing a phosphate function at the 1 position of a

1,2-diol

leading

to

the

phosphatidylinositol,

as

discussed

later. Such a methodology, however, was not known until the report on the phosphite-phosphonium approach in 1993.^^ Problems specified

with

For

protection:

free hydroxyl

groups, a

the

phosphorylation

short

access

to

a

of

the

properly

protected inositol is required. Although the protection technology has

developed

derivatives, hydroxyls

is

enormously,

in

the

case

of

the straightforward protection still

quite

difficult

due

inositol

and

its

of some of the six to

their

similar

reactivities. Some useful protecting methods have been

recently

reported in relation to the synthesis of inositol polyphosphates.

Problems

with

optically

active

inositol

derivatives:

Most

inositol phosphates are optically active. Therefore, a generally applicable procedure for getting a chiral derivative is required. This subject has been reinvestigated since starting the synthetic race for obtaining Ins (1, 4, 5) P3 . ^^ In most of the cases where myoinositol is chosen as a starting material, conventional

optical

396 resolution procedures which comprise the derivatization of a racemic compound

to

the

diastereomeric

mixture

and

their

subsequent

separation have appeared using a variety of chiral auxiliaries. Among

them,

camphanic

esters

are

most

frequently

used

as

diastereomeric derivatives resulting in the successful achievement of the optical resolution of various jnyo-inositols.^^ It should be noted

that although optical resolution is often mentioned

to be

cumbersome, there are available now a number of successful reports. A variety of chiral starting materials such as D-and L-quebrachitol, D-glucurono-6,3-lactone, D-glucose, galactinol, (-)-quinic acid, and D-pinitol has been also utilized. Although this methodology allows the avoidance of optical resolution, it does not always provide a concise synthesis. Another choice in obtaining a chiral compound is the asymmetric synthesis which includes enzymatic reaction. The last subject only is discussed in detail in the text. Several review articles and books concerning the synthesis of inositol phosphates and phospholipids are available.^^ This text does

not

cover

all

reports

on the

synthesis

of

inositols

but

principally deals with synthetic strategies and total synthesis of inositol

derivatives which

involve

selective

reactions

such as

stereoselective and regioselective ones. Abbreviations All Bn BOM Bz Cbz CSA DABCO DAST DCC DDQ DEAD DIBAL DMAP DMF DMPM DMSO EE GPI HMPTA LDA Lev mCPBA MEM Ment

Allyl Benzyl Benzyl oxyme thy 1 Benzoyl Carboxybenzyloxy D-10-Camphorsulfonic acid l,4-Diazabicyclo[2.2.2]-octane Diethylaminosulfur trifluoride Dicyclohexylcarbodiimide 2,3-Dichloro-5,6-dicyano-l,4benzoquinone Diethyl azodicarboxylate Diisobutylaluminium hydride 4-(Dimethylamino)pyridine iV, i\J-Diinethylformainide 3,4-Dimethoxyphenylmethyl Dimethylsulfoxide 1-(Ethoxy)ethyl Glycosyl phosphatidylinositol Hexamethylphosphoric triamide Lithium diisopropylamide Levulinyl m-Chloroperbenzoic acid Methoxyethoxymethyl Menthyl

MOM MPM NIS NMO PCC Ph PI PTC Py SEM SM TBDMS TBDPS TBPP TFA TfOH THF THP TIPS TMS Tr TsOH XEPA

Methoxymethyl p-Methoxyphenylmethyl iV-Iodosuccinimide i\7-Methylmorpholine-i\r-oxide Pyridinium chlorochromate Phenyl Phosphatidylinositol Phase transfer catalysis Pyridine 2-(Trimethylsilyl)ethoxymethyl Starting material t-Butyldimethylsilyl t-Butyldiphenylsilyl Tetrabenzyl pyrophosphate Trifluoroacetic acid Trifluoromethanesulfonic acid Tetrahydrofuran Tetrahydropyranyl 1,1,3,3-Tetraisopropyldisiloxanyl Trimethylsilyl Tri tyl (Triphenylmethyl) p-Toluenesulfonic acid o-Xylylene iV, iV-diethylphosphoramidite

397 2 . PHOSPHORYLATION 2.1 Exhaustive Phosphorylation To achieve

total

synthesis

of

inositol phosphates

derivatives, multiple phosphorylation most

crucial

step. Use

of

and

of polyol derivatives

dianilinophosphoric

chloride

only method of choice for this purpose. However, deprotection

of

several

the is

furthermore

dianilinophosphoric

esters

the same molecules is quite difficult, while phosphorylation 2,3,6-tribenzyl-jnyo-inositol

is the

8 was

its reactivity

not satisfactory for perphosphorylation of inositols and spontaneous

related

in

of D-

9 with the chloride 8 giving 11 in 60%

yield was accomplished after exploring proper reaction conditions in the first total synthesis of Ins(1,4,5)P3 the

phosphorylation

positional reported

isomer

product

12 was not

of 9 under

similar

(Scheme 2-1).^^^ formed

at all

conditions.

1987, it was

that the reaction of tetrabenzyl pyrophosphate

with alkoxides generated in polyphosphorylated

shown in Scheme

products

in good

smoothly gave

yields.

Examples

the are

2-1.1^^

BnO.

BnO^ÂÔBn

PO

PO^^'OBn OP OP 11 12 (/^Phosphoryl group introduced)

MoBn or

HO'* Sr"OBn OH 9 (R^=H, R2=Bn) 10 (R^=Bn, R 2 = H ) {PhNH)2P(0)-CI 8, Py [(BnO)2P(0)]20 13, n-BuLi (BnO)2PN/-Pr2 22, tetrazole then mCPBA

ItsX

60%

0%

70%

40%

88%^^^

^P-NEt2, tetrazole then mCPBA °

OP

OBn

OR'^

(fjf"

(TBPP, 13)

by the action of a strong base such

situ

as NaH,^^*^ KH,!^ or butyllithium^^^ on inositols desired

In

However,

from 1 0 , a

87%

97%

23

Scheme

2-1

It is interesting to note that when the monolithium salt of the vicinal

diol

diphosphate

14

was

treated

15 was produced

with

equimolar

in 2 6% yield

together with 46% of the starting material monophosphate was observed at all equiv

of

butyllithium

and

2.5

amounts

(based on

of

TBPP,

the

pyrophosphate)

(SM) and no corresponding

(Scheme 2-2).^^ equiv

of

TBPP

However, when

were

used,

15

2.4 was

obtained in 81%. A similar phosphorylation procedure for the simple 1, 2-trans-cyclohexanediol

16

gave

an

identical

result.

However,

398 product distribution was found to depend on electrophiles as shown in Scheme 2-2. when benzoic anhydride was used as an electrophile, the monobenzoate 18 (El=Bz) became the predominant product yield),

contrary

experiments

to the result

of the pyrophosphate

(92%

13. These

indicate that a salt exchange reaction between

the

lithium salt 19 and the monophosphate 18 [El= (BnO) 2? (0) ] formed first is much faster than the substitution reaction of 19 with TBPP while benzoic anhydride reacts faster with 19 than its salt exchange with the monobenzoate 18.

rvBuLi

BnO.

0 'C, 2 h

-78 'C, 10mln

^^^

{1 equiv each)

16

a:^'

rvBuLi

El OH

(BnO)2P{0)0

1.1 eq 2.5 eq

1.0 eq 2.4 eq

^"^^

Bn04

t(BnO)2P(0)]20

17

'OH

O-EI

18

(BnO)2P(0)-0(0)P(OBn)2 (BnO)2P(0)-F BnOC(0)-CI Ts-CI PhC(0)-OC(0)Ph

94: 6 79:21 45 : 55 34:66 8:92

a.

,0-EI 0-EI

17

Scheme The

second

phosphitylation

phosphorylation using

and

intermediate

subsequent 24

perphosphorylation reactivity the

of

synthesis

may of

provide

of a

with

and y i e l d

protected

which

21 such as

( 2 2 ) ^'^ a n d

oxidation

polyols

the reagent, of

methodology

phosphoramidites

diisopropylphosphoramidite (XEPA)IS

2-2

its

cyclic

the

more

general to

N,N-

analogue

resultant

respect

involves

dibenzyl

method

1,4,5-

for

applicability,

(Scheme 2-3) . In Scheme

myo-inositol

23

phosphite

and

2-1,

2,4,5-

399 trisphosphate from the corresponding p h o s p h o r y l a t i n g methods i s summarized.

ROH

(R'0)2PNR"2 21 tetrazole

•

[O]

ROP(OR')2 g^

Scheme

triols

•

using

various

ROP(0)(OR')2 ^5

2-3

2.2 Reaioselective Phosphorylation The methods described in the preceding section lead to smooth phosphorylation

of all hydroxyl groups

in inositol derivatives.

Contrary to this exhaustive phosphorylation, the regioselective one is also quite useful, especially for the selective formation of the phosphoric

diester

functions

at

the

C-1

position

in

1,2-diol

derivatives of myo-inositol in the synthesis of phosphoinositides. These 1,2-diol derivatives are easily accessible because 1,2-cisdihydroxyl

groups

of jnyo-inositol

ketalization in comparison with trans

are

easily protected

by

its

diols. The diol derivatives

thus obtained comprise one equatorial and the other axial hydroxyl groups. Since the former is generally more reactive than the latter, several electrophiles were selectively introduced at the 1 position of the diols as shown in Scheme 2-5. However, phosphorylation at C1 in a 1,2-diol was extraordinarily difficult when known procedures were

used

because

migration

of

the

phosphate

function

and

cyclization occured with ease.^^ Meek et al. reported a regioselective phosphitylation using a P(III) reagent, dimethyl phosphorochloridite Ins(l,4,5)P3

in the synthesis of

(Scheme 2-4).20 Thus, the dibenzoate 26 which can be

readily derived from myo-inositol in three steps, was treated with 3.3

molar

equivalent

of

the

ethyldiisopropylamine at -40

chloridite

in

the

presence

of

°C to afford 1,4,5-trisphosphite

in

high yield, the C-2 hydroxyl group remaining free. The resulting phosphite was acetylated and oxidized to give the fully protected form 27 of Ins(1,4,5)P3, which was finally deprotected. Treatment of the

1,4-dibenzoate

26

in refluxing aqueous pyridine afforded a

mixture of dibenzoates, from which the desired 2,4-dibenzoate 28 was isolated and transformed to Ins(1,3,4,5)P4, as above. In general, preparation of 1-phosphate

derivatives has been

accomplished according to a tedious sequence starting from 1,2-diols 29 as follows: temporary protection of the OH-1 in 29 with the allyl, MOM, silyls

[Ph2{t-Bu)Si, Me2(t-Bu)Si, EtsSi], and benzoyl

400 OAc BzO^ÂÔH OBz

HO'

(MeOgPCI

I.AcCI, DMAP

'-Pr2NEt DMF, -40 'C

2. H2O2

BzO^ÂÔP(0)(OMe)2

Py.HgO 100'C

ca. 88%,

LHBr.AcOH 2. LiOH

lns(1,4,5)P3

OBz HO.

OBz 0P(0)(0Me)2 27

{MeO)2{0)PO 94%

OH

(MeOgPCI

I.HBr.AcOH -^ 2. LiOH

/-PrgNEt DMF, -40 "C

lns(1,3,4,5)P4

Scheme 2-4 groups, protection of OH-2 with Bn, THP, ArCO, EE (used in 3-deoxy3-fluoro derivative), deprotection of OH-1, and phosphorylation of OH-1

in

32

(Scheme

regioselective

2-5) .17c, 21

alkylation,

;^s

acylation,

shown and

in

the

silylation

scheme, at

the

1

position in the 1,2-diol derivatives 29 have been documented with ease

by

conventional

procedures

but

attempts

at

selective

phosphorylation at C-1 of these molecules have failed. OH

OH

RO.

JvÔH

RO'

T^'''OR

RO.

+R1

+R2

^XÔR^

ÂsÔH 1 RO'^ Y^""OR

N

RO^' î""OR OR 31

RO'^ ^ I "'OR OR 30

OR 29

0R2

0R2

yKÔR^

OR 32

R^=AII, MOM, Ph2{f-Bu)Si, Me2(f-Bu)Si, EtgSi, PhCO R2=Bn, THP, ArCO, EE (for 3-deoxy-3-fluoro-mya-inositol)

Scheme

2-5

Jastorff et al. reported a novel approach to the preparation of inositol 1-phosphate via the selective ring opening of phosphate

derivatives

1,2-cyclic

34 with alcohols (Scheme 2-6).22 The best

selectivity was observed using t~BuOH at low temperature (0 'C) even though the chemical yield was low (28%) . A phosphorus derivative which has the reactivity between highly reactive P(III) and less reactive 0=P(V) is expected to be suitable for

the

selective

phosphorus generated in

P(IV) situ

of pyridinium

phosphorylation.

derivative,

Thus, a

phosphonium

salt

tetracoordinate 38

which

was

from the corresponding phosphite 37 by the action

bromide perbromide

reacted with various

alcohols

401 O II O—R^- OBn

YS-°

BnOP(NEt2)2 1 eirazcNe then mCPBA

BnO'

ROH 0 'C, 22 h

y "OBn OBn 34 0

OH BnO^ BnO^'

BnO,

»0-P-OBn OR 'OBn

+

7 25 >100

:

BnO^

OBn 35 ROH=MeOH hPrOH f-BuOH

49% yield 39 28

1 1 1

Scheme 2-6

resulting

in the

Arbusov-type

formation

of the phosphoric

decomposition

39

of

triesters

(Scheme

40 via

2-7).

This

phosphorylation methodology can be ideally applied to 1,2-dihydroxy inositol derivatives and 1-phosphates are thus selectively obtained. The

method

opens

a

convenient

way

to

synthesize

phosphatidylinositols. Phosphorylation of 1,2:4,5-dicyclohexylideneinyo-inositol 49a using glyceryl phosphites was also achieved at C-3 regioselectively by the phosphite-phosphonium

approach to afford

43b and 43c. The former product 43b was then subjected

to the

phosphorylation using XEPA giving the fully protected phosphatidylinyo-inositol 4-phosphate 44 give PI(4)P

quantitatively which was deprotected to

(Scheme 2-8). Similar

selective phosphorylation

was

observed in the reaction of 49b with glyceryl phenyl phosphate in the presence

of iV-mesitylenesulfonyltriazole

(Scheme

2-9) .^3

Phosphorylation of 49b with diphenyl phosphorochloridate was also recorded.^^ At present, methods for exhaustive and partial phosphorylation of

inositol

derivatives

have

been

established.

These

methods

facilitate greatly the syntheses of inositol phosphates, inositol phospholipids and their analogues.

402

PyHBra

37 (Ri=Bn, Me)

, +^0R2 (RÔ)2P: Br

^

(R^0)2P0R2

CH2CI2

R30H

rRÔUP^^'^

EtaN

•

'•Br

RÔ-P, 40

39

38

„ OR^

O17H35CO2

j

O17H35OO2 Product ?^ ".OR BnO,ÂÔ-P:

0R2

T OBn oV>c OBn 41

BnO

OH

(BnO)3P

(MeO)2PO—'

nupn—> (MeO)2PO

a

b

c

94% (-42 'C, 1.5 h)

85% (0 'C, 0.5 h)

93% (-42'C, 1.5 h)

94% (0 X , 0.5 h)*

^ OR^

BnO,

OP

42

Q JoR^ HO

I

95% (0 'C, 0.5 h)

61% (-18 'C, 1.25 h)* SM: 20% 4-P: 6% 1,4-P2: 6%

O

P=(3C^P(0) Scheme

85% (-20*C, 1.5 h)* SM: 11% 4-P: 4% 1,4-P2: 6%

*Py/CH2Cl2(1:10)

2-7

O )-P-OMe

XEPA tetrazole 43b

^-0' ^0-P

thenmCPBA

T

''o

U02CC17H35

0-y-\

'-02CC,7H35

PI(4)P

& ' Scheme

2-8

C15H31CO2-1 Ci5H3.C02-[^W_^^

/

ONa

C15H31CO2—I C,5H3iC02^

cîô-p-o-J

Me HO

H0''V''0 ' 45

49b I

(PhO)2P(0)-CI EtgN, DMAP r.t, 4 h

60% (4-P: ca. 5%, 1,4-P2: 20) PhCOCI, r.t. 47% (for 61) TBDPSCI, -10'C 75% (61/62=20)

9. HOV . . Z - 7 — - ^ O H OAc 63

9c

HO HO

OH 64 D-lns(1,4,6)P3

Scheme

3-6

1,2-Camphor k e t a l 65, which can be conveniently prepared by the r e a c t i o n of m y o - i n o s i t o l w i t h D-camphor dimethyl ketal r e g i o s e l e c t i v e l y (Scheme 4-4),^^''^'^ showed i d e n t i c a l r e a c t i v i t y

408 with that of 47. Thus, treatment of 65 with 1.1 equiv of TBDPS-Cl in the presence of imidazole in pyridine at -10 °C gave the 3-silyl ether 66 in 88% yield together with

'"^^^^'^ '"^^5)P

^

,^ ^ ^,« »ns(1,3,4)P3

^ "^^

PI(3A5)P3 PK3,4)P2 GPi

"^^

i: BzCI, Py; ii: BzCI, DMAP; iii: CH3CO(CH2)2C02H. DCC, DMAP

Scheme

3-9

The reaction of 1,2-cyclohexylidene-inyo-inositol dichloro-1,1,3,3-tetraisopropyldisiloxane a completely

47a with 1,3-

(TIPS-CI2) takes place in

regioselective manner to afford

the

3,4-disiloxane

derivative 76 in quantitative yield. ^-^ Compound 76 has been shown to be a very useful synthetic intermediate by the synthesis of various inositol phosphates and phospholipids

(Scheme 3-9). The usefulness

410 of 76 is based on suitable regulation of the reactivity of the two free

hydroxyl

groups

by

the

steric

bulkiness

of

the

diisopropylsilylene group at the side of C-4 as well as the easy availability of 76 as illustrated in the scheme. Furthermore, the two protecting

groups

in 76

can

be

distinguished

chemically,

consequently selective removal of the desired one is quite easy. In the

case

of

similarly

protected

1,2:3,4-diketals

48, a highly

regioselective reaction will not be expected and the function over the cis

trans-ketal

always has to be first removed at an earlier

stage of the synthesis. The completely regioselective acylation at the 6-position of 76 was achieved by benzoylation and levulinoylation. The monobenzoate 77 thus obtained was phosphorylated quantitatively by the method using PCI3 and the four-step deprotection procedure to give 5-phosphate

inositol

[Ins{5)P]

myo-

(Scheme 3-10). On the other hand,

phosphorylation of 77 using butyllithium and TBPP afforded the 6phosphate 81 resulting from the initial migration of the benzoyl group and subsequent phosphorylation of the less hindered OH-6. The phosphate

81

was

levulinic

ester 79

synthesis

of

converted

inositol

phosphatidylinositol which

are

to

is a pivotal

racemic

inositol

synthetic

phospholipids

and

GPIs,

to

play

an

important

The

for the

especially

3 , 4 , B-trisphosphate^^ g^^d

postulated

Ins(4)P.

intermediate

for

3 , 4-bisphosphate

role

in

a

new

intracellular signaling system connected with the tyrosine kinaselinked

receptors.

The

efficiency

of

their

synthesis

depends

largely on the usefulness of the disiloxanyl protecting group.

Olns(4)P -^ .0^....... 83x90x85%

TIPS5 II J J ^ o ' T "'OP ô^r"'nP ^^^

81

-* -* 4^/» 47%

77 77

•• TIPS II J J ^^^ o T "^" Puant •o'*N^"OBz

P=(BnO)2P(0)

• 89x85x61%

lns(5)P

^^

80

1:1. PCl3, 2. BnOH, 3. f-BOgH; ii: 1. (HOCH2)2 , TsOH, 2. EtgNHF, 3. Hg, Pd-C, 4. NH3, MeOH; iii: n-BuLi, TBPP

Scheme

3-10

Dibenzoylation of 76 can be also realized by using a combination of benzoyl chloride and DMAP in 88% yield. The resultant benzoate 78 was

converted

to

the

optically

active

menthoxyacetate

83 by

successive decyclohexylidenation, selective acylation at C-1, and optical resolution

(Scheme 3-11). A diastereomeric mixture of 83

411 and the other isomer can be separated effectively by a chiral column chromatography while the racemic diol 82 was difficult to separate by

the same

column.

Derivatization

of

a

racemate

to the

diastereomers bearing a proper chiral auxiliary is very useful for optical resolution by a chiral chromatography. Benzoylation of 83 followed by removal of the menthoxyacetyl and TIPS groups gave the 1,3,4-triol which was then phosphorylated by the method using XEPA.

V, 76%

78 97%

OBz OBz R^=R2=H R^=MntAc,

R2=H

R^=MntAc,

R2=B2

R^=H,

R2=:BZ

i: (HOCH2)2, TsOH; ii: MntAcCI; iii: BzCI, DMAP, EtN/'-Prg; iv: NHgNHg; v: EtgNHF; vi: XEPA, tetrazole then mCPBA

Scheme 3-11 Azidodeoxy-myo-inositol

87 was efficiently protected with 2.5

and 1.2 equiv of TIPS-CI2 under conditions shown in Scheme 3-12 respectively

giving the corresponding bis-disiloxane 88 and mono-

disiloxane 90 in high yields respectively. These were utilized to prepare the tritium-labeled jnyo-inositol analogue 89 and 3-azido-3deoxy-myo-inositol 2 , 4 , 5-trisphosphate 91.^-^

TlPS-Cig (2.5 eq) /

OH

/

TIPS

73"C,40h 80%

0'**' T ^ \ 1 TIPS-0

I.AC2O, DMSO 95%

1

TIPS

2. NaBT4

88

ÔH "OH

HO OH 87

\ \

\

TIPS-CI2 (1.2 eq) r.t, 10h 95%

OH

^

TIPS

OP03Na2

1.NaH, [(BnO)2P(0)]20 61% 2. a) TMS-Br, b) H2O ONaOH 68%

Scheme 3 - 1 2

OH _ . NaaOgPO

412 Bruzik and Tsai applied the protection strategy using the TIPS group to the synthesis of various optically active precursors for inositol

phosphates

and

inositol phospholipids

from

1,2-camphor

ketal 65 of niyo-inositol. ^^ The parent myo-inositol itself can be protected regioselectively by

the

TIPS

derivative

group

to

the

1, 6:3,4-bis (disiloxanylidene)

92 in 66% yield together with trace amounts of other

inositol derivatives inositol

give

(Scheme 3-13).^^ A similar type of protected

is difficult

to obtain by any known method. The novel

intermediate

92 was benzoylated followed by removal of the TIPS

group

phosphorylation

and

tetrakisphosphate 94, which Ins ( 1 , 3 , 4 , 6 ) P4 .

Thus,

a

to

form

was

finally

fairly

the

fully

protected

deprotected

concise

to

synthesis

give

of

the

tetrakisphosphate has been completed. OH lns(1,3,4,6)P4

93: R=Bz 97%

p= ( Q C ! ^^^^^

i: TlPS-Og, Py; ii: B2CI, Py, refl.; iii: aq. HF; iv. XEPA, tetrazde then mCPBA; v: Hg, Pd-C; vi: MeONa

Scheme

3-13

3.3 Reaction of mvo-Inositol Orthoformate The

orthoformate

95 of myo-inositol,

which

was

originally

reported by Lee and Kish,^^ has been utilized frequently

for the

synthesis of various inositol derivatives. It was first employed for the synthesis Ins (1, 3 , 4 , 5) P4 by two groups . ^^ ' ^'^^'55 rpj-^^ Merck group found that action of one molar equivalent each of NaH and several electrophiles on the orthoester 95 afforded the 4-0-monosubstituted products 96 in good yields, together with a trace of 4,6-dibenzyl ether

99

in the case

of benzylation

(Scheme

3-14) .

The

high

regioselectivity and monosubstitution may be rationalized in terms of

the

formation

of

a

thermodynamically

preferable

chelated

intermediate 97 and subsequent substitution at the fixed anion site. Introduction of the second benzyl group to 96 was not so selective while benzylation of 95 under catalysis

conditions

selectivities

gave

acid-catalyzed

the 2,4-dibenzyl

(Scheme 3-14). ^'^^

or phase ether

transfer

98 with good

413

,-/ô

, - ^ 0 1.NaH(1eq)

HO, HOI

OH

95 TBDMS-CI 2,6-lutidine

J^P HO.

2.RX

BnBr: 75% p-MeOPhCHsCI: 6 7 % AIIBr: 8 0 % [(BnO)2P(0)]20: 7 2 % BnOGHgCI: 6 7 %

HOI OR

96

97

o/ô •J3

58%

HO, BnO I OBn 99

B n O li

98 TBDMSO,

HOI OR 100: R=H - \ BnBr. r-BuOK 101: R=Bn-*^

I OH

Cl3CC{NH)0Bn, TfOH

81%

1 3 3 : R'=R^=H £ 5% ^ 1 3 4 : R^=R2=(BnO)2P(0)

. /- 130: R=H ' ' ^ 131: R=Bn 91%

i: 2,2-dimethoxypropane, TsOH; ii: BnBr, AggO; iii: MeONa; iv: PTS; v: (BnOlgPNf-Prg, tetrazole then mCPBA; vi: Hg, Pd-C

Scheme

3-19

Benzoylation of myo-inositol with 3.5 equiv of benzoyl chloride in pyridine at 90 ' C for 2 h gave

1,3,4,5-tetra-O-benzoyl-myo-

(135) as the main product

(37% yield), the quantity of

inositol

which was more than that statistically expected

(Scheme 3-20).^^

When the reaction was conducted by using 2.5 equiv of the chloride, 1, 3 , 5-tri-O-benzoate

137 was conveniently

isolated by a column

chromatography even though in low (15%) yield. ^-^ The benzoates thus

418 formed

which

interesting groups

can

be

separated

from

each

other^^

intermediates

because

replacement

phosphate

functions

provides

with

of

are

these

various

quite

benzoyl inositol

polyphosphates. For example, 135 was concisely converted to racemic Ins(1,3,4,5)P4 (Scheme 3-20) 62 OH OBz

BZO4

myo-inositol

OH

HO' OBz 137

lns(1,3,4,5)P4 OBn

Pd-C

OR

quant

XEPA. tetrazde / ' 1 0 2 : ' ^ = ^ ^ ^ Q thenmCPBA \.^4Q.

R^Cl

y{0)

90%

Scheme From

the

Ins(1,4,5}P3

symmetrical

3-20

tribenzoate 137,

optically

active

and Ins(1,3,4,5)P4 were efficiently prepared in the

shortest way, as shown in Scheme 4-2. 3.5 R^qjoselegtive Protection of 1,2-Diolg The 1,2-diol derivatives of iTiyo-inositol comprise the equatorial hydroxyl group at the 1-position and the axial one at C-2, so that selective modification of the less hindered former group is easier than

that

benzylation

of

the

employ

latter.

Old

conditions

procedures using

for

a hydroxide

allylation base

and

such

as

powdered NaOH or KOH and a hydrocarbon solvent such as benzene or toluene at refluxing temperature. More regioselective alkylation is achieved by using a stannylene intermediate^lb especially in the presence of CsF. The method was applied to get 1-0-MEM derivative 142 in high yield and selectivity.'^^ The MOM group was introduced to OH-1 of 141 in a completely selective manner without CsF even at higher temperature.^^ Some examples are presented in Scheme 3-21.

419 OBn AilO^^JsÔMEM

OH AIIO^^JsÔBn

Alio*' Y^^''OBn OAII 142

BnO^' X^^''OBn OBn 143

1. n-BugSnO 2. MEM-CI, CsF 3. BnBr, NaH 1. n-BugSnO

2. AIIBr, nSu^Nl 95 'C. PhMe

62%

85%

OBn

85%

1. n-BugSnO 2. MOM^NEtg-CI, 60 'C

OH

OH AIIO^^JsÔAII

BnO^J^ÔMOM

B n O * ' T "OAII OAII 144

BzO^* T ^ " ' O B n OBz 145

Scheme Various

OH

1. n-BugSnO 2. BnBr, PhMe, n-Bu4NBr, 2 h, refi.

regioselective

3-21

acylation,

silylation, ^'^^' 21ci, 66 ^^d

carbamoylation^'^ of 1,2-diol derivatives of inyo-inositol have been reported where the substituents were introduced at C-1 without

the

aid

of

the

tin

intermediate^^^ at

room

directly or

lower

temperatures because these electrophiles have enough reactivities toward alcohols

(Scheme 3-22) . Me,

r.u.n

OH

I

BnO^ J L ÔSiEtg (BnO)2P(0)0'

T OBn OP(0)(OBn)2 147

95% 'OBn

85%

EtaSiCI Py

NaOCN CF3CO2H

OH

BnO^ÂÔC(0)NH2

Scheme 3-22

420 The triethylsilyl ether 147 thus formed regiospecifically from the d i d 150 [R1=(BnO)2?(0), R2=Bn], which was optically resolved by a chiral column chromatography, was transformed to Ins(2,4,5)P3 and Ins(l,4,5)P3 (Scheme 3-23}.^^ At this stage, temporary protection of OH-1 with the silyl group is not necessary, i.e. 150 can be directly phosphorylated by the phosphite-phosphonium approach as described in the section on phosphorylation (Scheme 2-6).^^ The diol 150 was used furthermore as a versatile synthetic intermediate for the synthesis of myo-inositol 1,2-cyclic-4,5-trisphosphate 152 (Scheme 3-23},{8} 2-acyl analogues of Ins(1,4,5)P3, and inositol phospholipid.

o BnO,

O

Cl2P(0)0-

(BnO)2P(0)IO (BnO)2P(0)IO

OBn

Py quant.

y Phospholipids

(BnO)2P(0)IO' T " O ^ " (BnO)2P(0)IO 151

^^'^ -HO3PO''' T^^'"OH ^"^"^"HOgPO 152

0P(0)(0Bn)2 Bn04.î^^0SiEt3

I.PCI3 2. BnOH

^•^•^"^^^^^^ (BnO)2P(0).O^^V""OBn 84% (BnO)2P(0)IO 153 BnO, (BnO)2P(0)IO (BnO)2P(0)IO 154

0P(0)(0Bn)2

I.H2, Pd-C

OBn

2. NH4OH quant.

H2

lns(2,4,5)P3

^ci-C ^

lns(1,4,5)P3

i: 1. B2CI, DMAP, Py (97%); 2. aq. AcOH, TsOH (90%); 3. PCI3, BnOH, then /-BUO2H (85%) Scheme

3-23

As described above, the l-(or 3-) hydroxyl group of myo-inositol has higher reactivity over other equatorial hydroxyls. This tendency was observed in the following examples.^^ Phosphorylation of 155 with diphenyl phosphorochloridate produced the 1,3-diphosphate 156 predominantly in moderate yield together with 1,5-diphosphate 157 (Scheme 3-24).15,55 Benzylation of the triol 155 under PTC conditions led to the 1-benzyl ether 158 in 71% yield (overall yield from the orthoester 95 in three steps). The resultant tetrabenzyl ether 158 was again regioselectively acylated at C-3 with camphanic acid chloride for resolution (Scheme 3-24) A'^

421 OBn

OBn

HO.,AÔH Bno' T

"OBn

OH 155

(PhO)2P(o)ci

OBn

POyXôp

myX^• OF

DMAP. E\,H " BnO^ T " O B n "" BnO^^ T "OBn CH2CI2, 25 'C OH OP 58% 156 82 : 18 157 F=(PhO)2P(0)

71% BnBr,n-Bu4NI (from orthoester 95) i aq. NaOH, CH2CI2 OBn

OBn

BnO^ÂsÔH

(.)-camphanic acid chloride

BnO^ T " OBn OH

B n O ^ ^ A ^ O ^ ^

Py. 0 'C BnO 47 + 44% (for diastereomers)

158

Scheme

X " OBn OH

3-24

On the other hand, the sterically more hindered axial hydroxyl at the 2-position in 1 was ingeniously benzoylated without affecting OH-1

(Scheme 3-25). Thus, 1,4-di-O-benzoyl-inyo-inositol

(26) was

converted to the orthoester 160 by the regioselective reaction at the 2,3-cis-diol site in preference to the 5,6-trans

one. Subsequent

hydrolysis of the orthoester function of 160 in 80% aq. acetic acid afforded the 2-benzoylated product 161, which was transformed into racemic Ins(1,4,5)P3 as shown in Scheme 3-25.^^

PhC(0Me)3

OMe Ph-7^0 0^^^,x\ÔBz

TsOH, DMF Il0-C,12h

BzO'

OH

HO^yAÔBz D,n***k^'', BZO Y OH OH

OBz aq.AcOH

110-C

HO^

J^ÔBz 1. (EtO)2PCI 2. H2O2

B2O''

51%

26 OBz 1.TMS-Br

OBz

(EtO)2P(0)a

lns(1,4,5)P3 2. KOH, 60 'C

BzO"

0P(0)(0Et)2 OP(0)(OEt)2

162

Scheme

4.

3-2 5

ACCESS TO CHIRAL INOSITOLS An enantioselective reaction is a useful tool to obtain a chiral

compound.

However,

in a practical

sense, employment

of

such a

reaction in a total synthesis is limited mainly because it is not always

easy

to

get

the

product

with

high

optical

purity. A

422 b i o c a t a l y s i s system can a l s o be u t i l i z e d f o r t h i s p u r p o s e and s e v e r a l c h i r a l i n o s i t o l d e r i v a t i v e s with h i g h l y o p t i c a l p u r i t y have been a c c o r d i n g l y p r e p a r e d . In t h i s s e c t i o n , chemical asymmetric p r o c e s s e s and b i o c a t a l y s i s s y s t e m s a r e d i s c u s s e d . 4 • 1 K i n e t i c R e s o l u t i o n i n Chemical 4.1.1 Enantioselective OH BzO 1 ÔBz "^^-yX/

2 /—s MsCI, Me-N 0

HO

DMAP.THF.OC

,V-

I OH OBz

Processes

Tartarovlation OH BzO^^ÂÔBz O II , R*CO OBz

/Y*"0H

'

HO^V'"OIROBz

163a

137 163a /163b

R^COgH

\y^"-|ÔMe A Q ^ O H 0

OH Bz04,Â^0Bz T T ^

96:4

Yield. % chem. opt.

64

92

163b 163a /163b

R'COzH O •^y^ OMe ([^^^tvOH

l>164a

0

Yie\6, % chem. opt.

4 : 96

56

92

2 : 98

40

96

27:73

40

46

L-164b

g

'°' Scheme

4-2

The enantioselective tartaroylation of tribenzoate 137 has led to regioselective protection as well as optical resolution. The tetraacyl derivative 163a with high optical purity thus obtained was shown

to

be

a

suitable

material

Ins(1,3,4,5)P4 and D-Ins(1,4,5)P3

for

the

preparation

of

D-

(Scheme 4-2).^^ Thus, 163a was

silylated with triethylsilyl chloride to afford the disilyl ether 167 in 98% yield which became optically pure by recrystallization. Removal

of

the

four acyl groups

from 167

can

be

accomplished

successfully by the action of the ethyl Grignard reagent to give the tetrol

168

in high yield.'^^ Several

conventional

deprotecting

424 procedures for esters using nucleophiles such as ammonia, hydrazine, sodium methoxide, and DIBAL afforded various silyl group-migrated products. When the methyl Grignard reagent, which is less reactive than the ethyl one, was used the benzoyl group at C-3

in 167

remained intact resulting in the formation of 170. The tetrol 168 and the triol 170 were respectively

subjected

to a sequence of

reactions involving phosphorylation using XEPA and hydrogenolysis on Pd-C in aqueous MeOH for deprotection with accompanying removal of the silyl groups, and in the latter case additional methanolysis, giving rise to the target inositol phosphates. Both sequences are the shortest preparative

routes developed

to date

to optically

active D-Ins(1,3,4,5)P4 and D-Ins(1,4,5)P3. 4.1.2 Chiral Spiroketal Synthesis

I

jL0N>Me 173

B2O. OBz

CSA, CHCI3 refl.

70% 174 l.aq. NaOH, 96% 2. BnBr, NaH, n-Bu4NI, 74% 3. 95% TFA, 63%

177 CSA, CHCI3, refl.

BnO, OBn BzO OBz

0 2 X ^ ^ 8 " ' " " " ' ° " 91%^ 2. BnBr, NaH, ' n-Bu4NI. 87% 3. 95% TFA ,18%

176 Me

BnO OBn

D-175

H g S Z i h ^ g f " L-175

ph'^ôqf ] Me

177

^^-^

Scheme

178

4-3

Using a strategy similar to the asymmetric acylation described above (4.1.1), the symmetrical 2,5-dibenzoate 172 was transformed to the

optically

pure

enantioselectivity

by

diol the

174

in

reaction

70% of

yield one

with

pair

of

complete the

two

enantiotopic vicinal diols with the novel chiral pyranyl pyran 173 with a C2 (Scheme

axis

in the presence

of

camphorsulf onic

4-3). ^^2 This unique spiroketal, which

acid

is named

(CSA)

dispoke

425 174,

controls

regioselective

the

direction

pathways

of

the

resulting

in

enantioselective the

formation

of

and the

thermodynamically more stable adduct. This dispoke adduct 174 is anomerically

stabilized

tetrahydropyranyl

because

the oxygen

substituents

of

the

ring are disposed axially. Furthermore 174 has

equatorial methyl groups whereas the other unfavorable adduct would have

less

dispoke

stable

axial

adduct was

side

chain

converted

substituents. The

"matched"

to D-1, 2 , 5 , 6-tetra-O-benzyl-myo-

inositol D-175 as shown in Scheme 4-3. When the dipyran 177 was used instead of 173, the dispoke 176 was the

"matched" adduct,

which was converted to the opposite enantiomer L-175. The dipyran 178 was also shown to produce L-175 via an adduct similar to 176. 4.1.3 Camphor Ketal Formation Ketalization of jnyo-inositol substance,

was

carried

(1), which is also a symmetrical

out with

D-camphor

dimethyl

ketal

17 9

producing several products, from which after partial hydrolysis of the trans-ketal a diastereomeric mixture of four 1,2-cis-ketals 65a, b,

c, and d were obtained

initially

formed

(Scheme 4-4) .^^ When a mixture of the

ketalization

products

was

exposed

to

acidic

conditions in a solvent system of CHCI3, MeOH, and H2O, the major 65a

monoketal

precipitation

was

obtained

after

preferentially,

decomposition

of

resulting

the

from

trans-ketals

the and

equilibration. The latter procedure provides a practical method for the preparation of a chiral 1,2-protected jnyo-inositol which can be transformed into various optically active inositol derivatives.'^'^' ^^

OH

nor^-^ d h T ^ 1

/d^

1.H2SO4.DMSO

Meo'''''

-^z ^ n

2.TsOH.CHCl3-MeOH-H,0

179

^'^^

65a

OH 47%

65b

OH 65a

I.H2SO4, DMSO 2. partial hydrolysis

OH

^ ^ O ^ ^ ;_ ^ ,OH

&°» • ^ s : ^ - >s^r;^f^ OH

13%

65c

Scheme

OH

17%

4-4

65d

23%

426 4.2 Biocatalvsis Routes In organic synthesis, a variety of enzyme-catalyzed reactions have been examined recently with expectations of highly regio- and enantioselective reactions and functional group transformation under mild

conditions.'7^ In the field

of

inositol

chemistry,

such

selective reactions using isolated enzymes and microbes have been demonstrated providing useful chiral synthetic intermediates. 4.2.1 Enzvme-aided Enantioselective Hydrolysis Treatment

of

racemic

cyclohexylidene-myo-inositol

3 , 4-di-0-ac e t y 1 - 1 , 2 : 5 , 6 - d i - 0180 with bovine cholesterol esterase

(CE) yielded a mixture of the fully deacylated diol (-)-50a (51% chemical yield with 85% ee) and the monoacetate (+)-181 (38% yield with

86%

ee)

with

high

optical

purities

(Scheme

Enantioselective hydrolysis of racemic 4-butyrate or porcine pancreatic lipase

4-5) .^ ^

(±)-182 with CE

(PPL) was also demonstrated yielding

the diol (-)-50a and starting material (+)-182 with highly optical purities.^^^

Cholesterol esterase

O/'YÔAc

DMF 23*C,168h

H0'V\P ( ) >—^ (-)-50a

180

o\^"oAc 51% (85% ee)

( \ N—f

38% (86% (+)-i81

o^'Sr "OH >98% ee (after recrystallization) (+)-50a

OyAÔH

182

(+)-182 CE PPL

Scheme

86%ee 95%ee

4-5

HO,. A ..0

{.)-50a 93%ee 88%ee

427

The monoacetate ( + ) - 1 8 1 was h y d r o l y z e d followed by r e c r y s t a l l i z a t i o n t o p r o v i d e t h e d i o l (+)-50a w i t h o p t i c a l p u r i t y g r e a t e r t h a n 98%. The o p t i c a l p u r i t y of an i n o s i t o l d e r i v a t i v e contaminated with the other enantiomer often i n c r e a s e s by r e c r y s t a l l i z a t i o n . The d i o l (+)-50a was r e g i o s e l e c t i v e l y b e n z y l a t e d v i a a s t a n n y l e n e approach, followed by a c e t y l a t i o n t o p r e v e n t t h e m i g r a t i o n of t h e c i s - k e t a l during h y d r o l y s i s of t h e t r a n s - k e t a l i n t h e next s t e p . The f u l l y p r o t e c t e d d e r i v a t i v e 183 t h u s formed was c o n v e r t e d t o I n s ( 1 , 4 , 5 ) P 3 a c c o r d i n g t o t h e p r o c e d u r e s shown i n Scheme 4 - 6 . 3 4 a , 7 4

n-BugSnO Q O ^ Y ' ^ O H

W

BnBr.CsF

(+)-50a

OAc

AC2O, EtgN

O: T \_J

'OBn

"^"'

Q T OBn

t55 I.AcCl, MeOH.CHgCIs 72% 1 2 . KOH/MeOH

I.Pd-C.Hg

lns(1,4,5)P3

(BnO)2P-N/-Pr2

2. AcOH

o'\x^^ tetrazole R 0 T OBn thenmCPBA OR 97%

98%

, ^,^ . HO \ OBn OH

184

185 R=P(0)(0Bn)2

Scheme 4-6 The antipode diol

(-)-50a was used for the synthesis of D-

Ins(1,3,4,5)P4 and D-Ins(1,3,4)P3 which were obtained in multigram quantities according to conventional and convenient sequences shown in Scheme

4-7. 34a,39

-^

novel

phosphatidyl inos itol

3,4,5-

(-)-50a

(Scheme

trisphosphate was also synthesized starting from 6-3) .40 Porcine

liver

esterase

(PLE)

effected

enantioselective

transformation of the symmetric dibutyrate 196 of the orthoester 95 to the monoester 197 in 83% yield with >95% ee (Scheme 4-8). ^'^^ A diastereomeric chemically

pair

by

of chiral

resolution

intermediates

of

the

were

also

corresponding

obtained

iV- ( i^) - 1-

phenylethylcarbamates 198. These chiral orthoester derivatives were converted which

to D- and L-2,4-di-O-benzyl-jnyo-inositol

in turn,

were

transformed

D- and L-102

to D- and L-Ins ( 1 , 3 , 4 , 5 ) P4

428 respectively. Racemic Ins(1,3,4,5)P4 as well as the optically active compound described here were synthesized by using the orthoester strategy.

However,

sequence (Scheme

compared

the chiral with

synthesis

the synthesis

required

a much

of the racemic

longer

material

3-15) . OBn ^^yXÔ^

(-)-50a

OBn AIIQ^rÔAII

Viii

D-lns(1,3,4,5)P4 HO^^ I " OBn OH

^^'

AIIO*^ T '' OBn OAII

D-102

189

75% I

'% |\ 97% I Vii OH _ .

AiiQô

... ^ Aiio^y 0

i,v ^ AiioyyoH

••^x/* Alio I OH OH 186

80%

91%

Lipase CES AcgO, Dioxane OH (+)-201 49% (98% ee)

Scheme

4-9

OH (+)-47a

49%(100%ee)

cyclic

430 Contrary

to the above

results, dioxane

effective on a Lipase CES (from Pseudomonas

as

a

solvent

was

sp.)-catalyzed selective

acetylation of 1,2-cyclohexylidene-inyo-inositol 47a; furnishing the 3-acetate

(+)-201 and L-1,2-protected starting material (+)-47a in

theoretically

quantitative

selectivities

and

chemical

yields

(Scheme 4-9).^^ Lipase P gave a comparable result to that in the case of Lipase CES. It is interesting to note that the Lipase CES recovered by filtration after the reaction lost most of the activity but the reusable enzyme can be obtained by treating with water and drying, indicating the necessity of hydration of the enzyme for maintaining the activity. Both enantiomers (+)- and (-)-47a were employed to synthesize D-Ins(l,4,5)P3

(Scheme 3-6),^^ D- and L-Ins(1,4,6)P3

(Ref.48), D-

and L-Ins(l,4,5,6)P4,'76 and PI(3,4,5)P3 (Scheme 6-2).^7 A lipoprotein lipase from Pseudomonas

sp. effected not only

regioselective acetylation of the orthoester 95 without

observing

the acylation of the axial hydroxyl groups but also enantioselective butyroylation (>95% ee) of symmetrical 4,6-dibenzoate 203 which was derived from the acetate 202 by benzoylation and acidic methanolysis (Scheme

4-10) .^8 Regioselective

chemical

acylation

of

the

equatorial hydroxyl group in 95 was also reported using benzoyl (64% yield) and p-nitrobenzoyl chlorides (51%).^9

o

AcO"^

HO.

HOI 95

I OH

Lipoprotein lipase (from Pseudomonas sp.)

H O Il OH c 202 I.BzCi.Py 2. HCi, MeOH

OH HO^ ^ I ^

BzO

i OH 204

^02Cn-Pr

OBz

AvPrCOo^"^ Lipoprotein lipase (fronn Pseudomonas sp.)

(>95% ee)

Scheme

4-10

4 . 2 . 3 Microbial Oxidation of Arenes Mutant o x i d a t i o n {Pseudomonas putida) of a r e n e s 205 produces l , 2 - c i s - d i h y d r o x y c y c l o h e x a - 3 , 5 - d i e n e s 206. The chemical s y n t h e s i s of such n o v e l compounds i s v e r y d i f f i c u l t . Furthermore, when

431 monosubstituted benzenes such as toluene, fluoro-, chloro-, bromo-, and

iodobenzene

are used,

the c ojTiresponding' almost

cis-diols 206 can be obtained

optically

pure

(Scheme 4-11). These advantages

have

promoted their use as synthetic starting materials in a large number of total syntheses of various complex molecules. The diol derived

from benzene has been used

inositol

phosphates

by Ley's

206{R=H)

for the synthesis of some m y o (Scheme

group

5-3 and 5 - 4 ) . In this

section, the transformation of chiral diols is treated.

6

Pseudomonas putida

R=H, Me, F, CI, Br, I

205

Scheme

X 207

206

4-11

I.OSO4, MNO 85% 2. LiAIH4 85%

86% 86%

HO' > r ^ ^ ^ 0 OH

HO' 0 H O ' ^^^ ^ ^ ^ ^^^ O OH

208

209

mCPBA

,6: Br

O ' ^

1.MeOH,Al203 90% 2. HCI. H2O LMeOH.AIgOa 89%

MeO^SÔ

2..LiAIH4

2. HCI, H2O

HO^' Y ' ^ O H OH (-)-210 (+)-210

OH 212

211

OH

tô-

HO,, A^ÔH

'••rr' &^^

OH

HO^^^Y^ÔH OH (-)-213

(+)-213

Scheme Homochiral

OH MeO^ X ÔH

I.OSO4, NMO 63%

4-12

bromo-isopropylidene

207

ketal

derived

from

206(R=Br) has two double bonds, among which the olefin at C-3 is the more

electron

rich.

Therefore,

oxygenation

reactions

occur

preferably at C-3 and C-4 rather than at C-5 and C-6, as supported by calculation.SO Thus, osmylation and epoxidation of 207 proceeded in complete regio- and stereoselective manners to give the diol 208 and They

the epoxide were

then

211 in high converted

yields to

respectively

( + )-

and

(Scheme

(-)-pinitol

4-12) . 210

via

stereoselective oxygenation according to the procedures shown in the

432 scheme. By similar approaches, chloro- and bromo-cis-diols 206 (R=C1 and

Br)

were

transformed

into

both

enantiomeric

conduritol

E

epoxides (-)- and ( + )-213 respectively. ^^ Three stereoisomeric

inositols were prepared from the highly

functionalized derivative 215 which can be derived from 214 in one step by the reaction with KMn04 (Scheme 4-13).^2 Treatment of 215 with AI2O3

in an aqueous medium afforded

ketoalcohol

216 in 85% yield, which was converted efficiently to

ailo-inositol

stereoselectively

the

(217) essentially as a single product. The epoxy diol

215 was converted under basic conditions to D-chiro-inositol

(125)

with more than 95% selectivity by attack of the hydroxide ion on one side of the epoxide carbon atom while its treatment under acidic conditions furnished neo-inositol

(219) as a minor product along

with 125 (3:7) resulting from the attack of H2O from the other side. CI

KMn04

AI2O3 H2O, 80'C

HOvJCÔH

H2

HO.,JssÔH

H O ' S ^ OH "^^^

H O * ^ ^ ^ OH

OH

OH

216

217

OH HO^ > ^ ^ Q ^ r / \ HOs K> ^ ^0 OH

HO^ " T "OH OH 219

{neolchiroZ-1)

OH HO^ X ÔH resin(H*

un^^k^ HO V OH OH D-125

218 Scheme

4-13

4.2.4 Employment of Metabolic Enzymes In

a

living

system,

the

transformation

of

a

substance

is

catalyzed by its recognized enzyme. Therefore, all natural compounds and

their

analogues

can,

in principle, be prepared

by

a

sole

enzymatic reaction or by consecutive enzymatic reactions along the metabolic pathway in vitro

or in

vivo.

One of the metabolic enzymes for Ins(1,4,5)P3, Ins(1,4,5)P3 3kinase

isolated

from

rat

brain

cortex

was

Ins(1,3,4,5)P4 from Ins(1,4,5)P3.83 j^yo-Inositol directly without protection to fluorescent

used

to

prepare

(1) was converted

1-phosphatidylinositol

433 analogues by reaction with the corresponding cytidine diphosphate diacylglycerols 220 in the presence of PI synthase (rat liver microsomes used as the source of its activity.) (Scheme 4-14). ^"^ These analogues were used for studying the metabolism and intracellular transport of these lipids in living cells.^^ NH2

OH

0

N

•0-P-O-

RCO2 R'COa

PI synthase

NO2

n=5,11

Scheme Provided

that

(rat liver microsomes)

L-o-P=0 )—( 6" HO OH

OH

a reasonable

4-14 amount

of

the enzyme(s)

can be

obtained, the enzymatic preparation of the desired molecules would be practically substantial

realized.

In order

quantities, an over

to obtain

expression

a pure

of

enzyme

in

this protein

is

required. For this purpose, the purified protein is necessary but its purification is not always easy, at present. 4.3 Ferrier Reaction In the biogenesis of myo-inositol, D-jnyo-inositol 3-phosphate is known to be derived from glucose 6-phosphate by the action of myoinositol 3-phosphate

synthase. This transformation bears a close

resemblance to the Ferrier reaction.°" This biomimetic sequence is suitable

for the preparation

of some optically

derivatives. An exomethylenetetrahydropyrane

active

inositol

compound 222 derived

from D-glucose are transformed to the corresponding oxocarbocycles 223 by a Hg2+-assisted to

aldol-type

reaction

in an

(Scheme 4-15). The resultant hydroxyketone 223 was

aqueous solvent dehydrated

intramolecular

afford

the

enone

224

which was

in turn

reduced

stereoselectively with NaBH4 and CeCl3 yielding 225. The conduritol derivative

225

has

been

shown

to

be

a

versatile

synthetic

intermediate. Thus, the benzylation product 226 of 225 has a C2 axis, therefore its osmylation, which is conducted by the approach of OSO4

from both sides of the double bond, produced

a single

product. ^'^ The tin-mediated regioselective introduction of the MOM

434 group to the diol with the inyo-configuration followed by benzylation and hydrolysis furnished the triol 9 which is the phosphorylation substrate laminitol

leading

to Ins(1,4,5)P3 . In a similar

(229) and mytilitol

manner, {-)-

(230) were synthesized. ^'^^ The same

type of compounds as 226 with a C2 axis is derived from D-6,3glucronolactone, and osmylation and epoxidation on the double bond were similarly demonstrated as shown in Scheme 5-5.^^

R2O\...--VA R'O

acetone-H20 ^'^l^y^ ê«-

n OM M e.

222

R 2 O \ ^ - Y ^ O H —^ 77% "''

"RÔ ^

l.OsO„NMO

BnO Q ^

RÔ 224

Bno

MOM-CI.EtaN, r.t.

2 2 7 : R=H 2 2 8 : R=MOM 1. BnBr, NaH 2. aq. HCI

RôX.^

223

(Ri=Bn. R 2 = M 0 M )

BnO

^

' 100% '"^"'

^"0 226

^Z"

,

,

1

NaBH4 CeCl3

OJPK^

RÔ 89% (2 steps)

225

96% 90%

BnO

HO

HO

HO

OH^ ^"îsn 9

Me ^ ^ i H 229

Me ^ ^ 230

Scheme 4 - 1 5 Glucosaminyl-chiro-inositol phosphate 237 and its myo-homologue 241

have

been

proposed

as partial

structures

in unidentified

substances acting as second messengers of insulin action. These substances are postulated to be released resulting from hydrolysis of glycosyl phosphatidylinositols bound to the plasma membrane after the binding of insulin to target cells. ^^ Both compounds 237 and 241 were synthesized from the Ferrier product (Scheme 4-16).^^ The tribenzyl

ether

225 was subjected

to a hydroxyl

group-assisted

stereoselective epoxidation with mCPBA and its regioselective ring opening with

allyl alcohol

in the presence of borontrifluoride

etherate gave a chiro-inositol 232 in good yield, which was then benzylated at the equatorial site via the stannylene intermediate. The remaining axial hydroxyl of the product 233 was glycosylated by the Schmidt's method using glycosyl trichloracetimidate and TMS-OTf giving the a-glycosyl-chiro-inositol 234 and subsequently, according to the sequences shown in the scheme, 237 was obtained. The myo-

435

i n o s i t o l homologue 241 was prepared by using s i m i l a r p r o c e d u r e s . For t h e t r a n s f o r m a t i o n of t h e c h i r o - f o r m 238 t o t h e myo-one 239, t h e Mitsunobu r e a c t i o n , i n v e r s i o n of t r i f l a t e with b e n z o a t e , and some o t h e r a t t e m p t s were not s a t i s f a c t o r y but t h e o x i d a t i o n - r e d u c t i o n sequence proved s u c c e s s f u l . HO

mCPBA

HO

BnO,\,...^-^

BnOi^----^

BnO

93%

.„ ^ u

HO

BP

BnO

225

OH i

BnoX-'-vA 75%

BnO ^ ^ j j

231

232 n-Bu2SnO, A>-Bu4NBr

.OAc

BnO

Ô

^ ^

^S^-^°'^^^>^^'^

BnO^^î^

55/0 TMS-OTf

BnO r BnOj^,,

^p^^

OH

. r 234: R=AII > 235: R=H " W 236: R=(BnO)2P(0) ^•NH3

2. Hg. Pd-C rOH HO-^jA^^ Ô H V I uo ^ 0 H O - ^ ^ "ÔPOaH237

OAII j. [ir(COD)(Ph2MeP)2]PF6, Hg then Ig, THF. HgO • ^38 ii: 1. (BnO)2PNAPr2, tetrazole, 2. RUCI3. Nal04. 70% 1.PCC 2. (S)-Alpine hydride 580/, -90'C ^^^^ 1. (BnO)2PNAPr2 ^ J^îJPv ^ Setraiole BnO-^Y^-^QH 2. RUCI3, ^^\Q^^^ ^ " O ^ ' R O I MPMO ^.^.--^ ^"ÔAII BnO'T"'^-'-^ 83% 239 BnO-WÔP(0)(OBn), 240 \ . ^ O H 1.CF3C02H,84%\^ H O - ^ T - ^ O 2. glycosylation. 65% HO-^^--^^ rû 3NH3 "^ HoNlJdO. ? i i 4. deallylation ^ 0-^>^--r\OH 5. Hg, Pd-C 28% (overall) "HOaPO A-.-^-^^-V-OH 241

Scheme An

acetoxy-substituted

4-16

exomethylene

derivative

242 is a

promising substrate for the Ferrier reaction since the resultant product has all the oxygen functionalities of the inositol skeleton. Thus, vinyl acetate 242, which is derived from the corresponding protected glucopyranoses by oxidation and 0-acetylation, was treated with mercuric trifluoroacetate in aqueous acetone at 0 °C to form the oxymercuration intermediates 243 which cyclized by addition of the chloride

ion to give a mixture of diastereomers

(Scheme 4-

17).^^ The Ferrier reaction recorded fairly good stereoselectivities and

compound

245,

having

the

desired

configuration,

was

predominantly formed. Bender and R. J. Budhu found furthermore that

436 the cyclization of the silylated organomercurial intermediate 242c was promoted by various Lewis acids, among which extremely superior selectivity

SnCl4

showed

(Table 1) . In the case of the silyl

ether 242, two other products 247c and 248c were formed along with 245c and 246c, and interestingly the NMR analysis showed that both predominantly adopt the conformation in which the three silyloxy groups and the hydroxyl group are axial. The major product 245 was then reduced in a completely stereoselective manner to generate an equatorial hydroxyl group resulting in the formation of the myoinositol derivative 249. OAc

RO

HgsÔAc XHg

Hg(02CCF3)2 MeCOMe/H20(4:1) OMe 0 "C, 10 min

242

NaCI

RO RO. o\^..Bu4NBr, Proton sponge, 69%; ii: NaOH, MeOH, 95%;iii: (BnO)[CbzNH(CH2)30]PNAPr2, tetrazole then mCPBA, 85%; iv: DDQ, 78%;v: (BnO)2PN/-Pr2, tetrazole then mCPBA, 73%; vi: Hg, Pd-C, quant.;vii: 4-aziclosaiicylic acid A/-hydroxysuccinimido ester

Scheme

4-18

For the synthesis of glycosyl inositols, the disaccharides 252 were chosen and transformed to the target molecules based on the Ferrier reaction as shown in Scheme 4-19.^-^ The procedures before and after the Ferrier reaction were identical with those described above and the allyl alcohol 254 was thus obtained. The selectivity of osmylation on 254 was improved from 2;1

(256/255) to 5.6:1 by

using its acetate 254(R=Ac) as the substrate. BnO ÔBn UvS.^0

BnoXZ^rA

l.Nal

v-oTs

BnO o - T < ^

n-Bu,Ni

9,

2. DBU

BnOA-^-T-^

252

GAL

i

o

OR

B n O \ - - r A BnO^^ OH 256

60%

GAL "1

'7;:^-T^^v--q O'

BnO-^^^^"^ 253

^ " ° OMe

GAL .w

1.Hg(02CCF3)2 r.t., 12h,92% 2. MsCI. Py.81% 3. NaBH4, CeCls •78 'C, 87%

BnO ^ ^ ^

GAL OR

9*^

BnoV--A.OH BnO 255

Scheme

" ^ ~ ^•^920/0 (for R=Ac)

OR

B n O X . ^ aln °^^ 254

4-19

Contrary to the results described above

(Scheme 4-19), highly

diastereoselective osmylation of cyclohexene derivatives 258 with the more bulky substituent, TBDMS, on one side of the double bond, which were derived

from oxanorbornenone 257, was

reported. This

afforded the suitably protected chiral myo-inositol derivatives 259 (Scheme

4-20).^^

438

^:^Ôkc

Os04,Et3NO aq. acetone r.t.,24h

60% (diastereoselectivity=91:9) 88% (diastereoselectivity=92:8) R=TBDMS ^ Ph

O T^*OTBDMS OTBDMS

Scheme 4 - 2 0

5. NUCLEOPHILIC SUBSTITUTION Nucleophilic

substitution

with

inversion

or retention of

configuration of an inositol ring carbon is used to obtain inositol derivatives stereoselectively with the desired configuration. The reaction has been often employed when myo-inositol derivatives are derived from other starting materials such as natural products and arenes. 5•1 Transformation of Alcohols to Halides ^ ? MeO yjâ^T- OH y>r—/-OH f Ql_j HO

DAST \ 20 C, 45min 57%

260

MeO p . ^ - . - , J > 7 ^ OH

"HO r MMeO eO _ •OH F . . / ' - ' ^ ' " " - ^ OH OH

I OH HO

•I

261

262

BBr, P OH

OH 263 OH

C OH

OH 265

OH 264

9 O2CC15H31 0 - P - 0 > ^ x ^ O2CC15H31 OH

OMe H? MeO /-—^ „Mo>< OH

OH

292

293

76%

59%

(regioisomer: 12%)

(regioisomer: 15%)

Scheme

73% (regi

5-4

The functionalities of the six carbon atoms in D-glucurono-6,3lactone 294, which is commercially available, are properly protected except for the two hydroxyls at C-2 and -5. Epimerization at the C-5 carbon atom and cyclization between the terminal C-1 and C-6 carbons will lead the carbocycle with the myo-configuration.^8 Thus, 294 was

443

^303: R^=Tr, R^=H ^ 304: R^=Tr. R2=Bn or All ^305: R^=H, R2=Bn or All

OR

OR

307, 33 (myo) R=AII: R=Bn:

30% 25%

OR

308 (chiro)

309 {scyllo)

27% 25%

9% 11%

i: acetone, TsOH [64%]; ii: TsCI, Py [88%]; iii: DIBAL [88%]; iv: K2CO3, MeOH [87%], v: MeOH, HCI [00%]; vi: NaH, BnCI [73%]; vii: aq. H2SO4 [91%]; vlii: NaBH4 [75%]; Ix: TrCI. Py [80%]; x: NaH, AIIBr [93%] or BnCI, xi: aq. HCI [50% for 305 (R=Bn) from 303, 91% for 305 {R=AII)]; xli: (C0CI)2, DMSO, EtgN; xlii: TICI4, Zn(Cu)

Schezne 5-5 converted

to the substrate

297

for the epimerization

in Scheme 5-5. The tosylate exists as the cis

and trans

as

shown

forms, 296

and 297, which are readily interconvertible under basic conditions, therefore both isomers are converted with epimerization via the trans

isomer 297

to the epoxide 298, which

was

subjected

to

methanolysis to yield the methyl acetal 299 highly efficiently as a single product. Further transformation was carried out as shown in the scheme and the dialdehyde 306 was subjected

to a reductive

cyclization with low valent titanium complexes to produce the inositol chiro-

myo-

derivatives 307(R=All) and 33(R=Bn) accompanied with the 308 and scyllo-isomers 309. Similar titanium reagents are

known to transform simple dicarbonyl compounds into

1,2-cis-diols

stereoselectively. -^^^ A

stereoselective

intramolecular

pinacol

coupling

of

the

dialdehyde 312, which was derived from D-mannitol in 9 steps, was

444 recently accomplished by using samarium diiodide as the coupling reagent compound

giving

the

cis-diol

(Scheme

92:8)

313

5-6). l^*^

(cis vs

trans-isomer,

scyllo-

Similarly, D-1, 4 , 5 , 6 - tetra-0-

benzyl-myo-inositol was obtained by the reaction of the protected dialdehyde derived from L-iditol in 56% yield together with the two trans-isomers, scyllo-

and chiro-inositols (4% each).^^^

Ph2(f-Bu)Si04

SmU 0Si(f-Bu)Ph2 8 6 % (2 steps)

Scheme The

mixture

of

5-6

33, 308, and 309

can be

converted

to

the

cyclohexene derivative 310 with a C2 symmetry axis by reaction with triphenylphosphine and triiodoimidazole and by respective osmylation and

epoxidation

obtained

of

310, myo-

33 and chiro-inositols

OSO4 PhgP

33 + 308 + 309

were

BnQ

Y^

mCPBA

33 BnO,

'v > 337: R^=H, R2=Bn '^ * ^ 3 3 8 : R^=Ac, R2=Bn QJO/^

OBn

I OBn oVc

88%

jT " O B n OAll

OAc

O2OC15H31

^h 02CCi5H3i

342

PO'

4^''0Bn OP 341 P=(BnO)2P(0)

i: 1. n-BugSnO, 2. MEM-CI, CsF; ii: BnBr, NaH; Iii: ZnBrg; iv: AcgO, DMAP; v: Pd-C, TsOH, aq. MeOH; vi: (BnO)2PN/-Pr2, tetrazole, then mCPBA; vii: NaOH; viii: 1,2-di-O-palmitoyl-sn-glycerol 3-(benzyl A/,A/-diisopropylphosphoramidite), tetrazole then mCPBA; ix: Hg, Pd black (50 psi)

(BnO)2P(0)Q,)

OH

Jf

0-P-OBn

(BnO)2P(0)0'

•" Y'"OBn h O2CC17H35

(BnO)2P(0)0

O2CC17H35

343

Scheme

6-3

448 6.2 Synthesis of Glvcosvl Phosphatidvlinositols Total synthesis of the glycosyl phosphatidylinositol anchor of Trypanosoma brucei has been achieved by Murakata and Ogawa (Scheme 6.4)^114 rp]-ê hexasaccharide core 353 additionally involving the myo-inositol residue at the one terminal, in which all glycosidic linkages are a was assembled by stepwise stereoselective aglycosylation using glycosyl sulfide, fluorides, and chloride as glycosyl

donors.

Introduction

of

two phosphoryl

functions

was

accomplished efficiently by the H-phosphonate method which involved the

initial

formation

of

the

H-phosphonic

and

pivalic

mixed

anhydride 356 and subsequent oxidation with I2 as shown in Scheme 6-5.

BnO HO' MPMO-

OBn

BnO-X^^X"^

v

y

346 ^^^ 348

1

I

r.n^

AcO-^-'^^n

. ^

BnO

BnO

BnO-V^' HO-

347

N3

I BnO-7 3nO-, -, oBnO-i>—r~^ N3 n 349

^ n ^ BnOi n BnO-^^'' ' ^ ^ ^ BnO-

BnO r* T .

- 350 BnO _ _

AcO

BnO'^^'' BnOnO-^*-"^^

no-^-^-^n BnO-\ ?o

BnOBnO'' BnO-

y/^'^ ^"^

= 'O—^^ BnO 361 \

O H-P-O-r . OH h0C0Ci5H3i

A H 0 ^ 2 \ ?Q HOr ^ > L \ HO' HQ

-"isâi

362

POCOC15H31 O .••OCOC,5H3i n HO-P-O-J O OH OH

HO' HO HO 363

Scheme

O

6-6

In t h e following method, in c o n t r a s t t o t h e above one, all p r o c e s s e s s t a r t i n g from m y o - i n o s i t o l were performed i n completely

450 regiospecific manners by using only two protecting groups, cyclohexylidene and TIPS, for the inositol skeleton (Scheme 67)^120 Thus, 76 was glycosylated at C-6 with mannopyranosyl phosphite 364 followed by deprotection of the ketal to yield the aglycoside 366 which was separated into two diastereomers. The desired stereoisomeric triol 366 was phosphorylated only at C-1 by the phosphite-phosphonium method and the glycosylation of the resultant phosphate 368 took place regio- and stereoselectively at C-2 affording 369 together with no P-glycoside. The inert nature of the hydroxyl group at C-5 might be attributed to steric hindrance owing to the mannosyl moiety at C-6 as well as the TIPS group. Deprotection of the carbonate of 369 was efficiently carried out by using

the

ethyl

Grignard

reagent

and,

after

acylation

and

deprotection of 370, gave PIM2 in good yield.

BnO-vBnO 1.n-Bu4NF-3H20 PhCOOH 89% 2. PhSH, EtgN 89%

' Ci7H35C02~|

OBn 2. C17H35COCI OÔBn 73%

^^^f 0"

Scheme

7.

OBn

PiMo

quant.

6-7

CONCLUSION A variety of methodologies and strategies for the synthesis of

inositol derivatives has been reported in the recent decade. We can now prepare any of the desired inositols. Although development of

451 new methodologies is still interesting and important in the field of inositol chemistry, our efforts should be directed to create useful probes having a variety of functions which interact with receptors and metabolic enzymes in order to understand various cell signaling processes

including the metabolisms of inositols at the molecular

level.

8.

ACKNOWLEDGMENT The author is deeply grateful to Tomoko Nakamura for assistance

in the preparation of the manuscript.

9. 1 2 3 4 5

6 7

8 9 10 11 12 13

14

REFERENCES H. Streb, R. F. Irvine, M. J. Berridge, and I. Schulz, Nature, 306 (1983) 67-68. M. J. Berridge and R. F. Irvine, ibid., 312 (1984) 315-321. M. J. Berridge, Nature, 361 (1993) 315-325. Y. Nishizuka, Nature, 334 (1988) 661-665. T. W. Rail, E. W. Sutherland, and J. Berthet, J. Biol. Chem. , 224 (1957) 463-475. T. W. Rail and E. W. Sutherland, ibid., 232 (1958) 1065-1076. S. B. Shears, in: J. W. Putney, Jr. (Ed.), Advances in Second Messenger and Phosphoprotein research. Vol. 26. Chapter 4: Metabolism of Inositol Phosphates, Raven Press, New York, 1992, pp. 63-92. Newly discovered pathways were added by suggestion of Prof. Masato Hirata of Kyushu University. D. J. Cosgrove, Inositol Phosphates, Their Chemistry, Biochemistry and Physiology, Elsevier, Amsterdam, 1980. (a) C. Grado and C. E. Ballou, J. Biol. Chem., 236 (1961) 5460. (b) R. V. Tomlinson and C. E. Ballou, J. Biol. Chem., 236 (1961) 1902-1906. (c) H. Brockerhoff and C. E. Ballou, J. Biol. Chem., 236 (1961) 1907-1911. see also: L. E. Hokin and M. R. Hokin, Biochim. Biophys. Acta, 84 (1964) 563-575. Y. Chuan and C. E. Ballou, Biochem., 4 (1965) 1395-1404. M. A. J. Ferguson, S. W. Romans, R. A. Dwek, and T. W. Rademacher, Science, 239 (1988) 753-759. Y. Watanabe, E. Inada, M. Jinno, and S. Ozaki, Tetrahedron Lett., 34 (1993) 497-500. For pioneering work, see: V. I. Shvets, B. A. Klyashchitskii, A. E. Stepanov, and R. P. Evstigneeva, Tetrahedron, 29 (1973) 331-340. (a) D. C. Billington, R. Baker, J. J. Kulagowski, and I. M. Mawer, J. Chem. Soc., Chem. Commun. (1987) 314-316. (b) J. Gigg, R. Gigg, S. Payne, and R. Conant, J. Chem. Soc. Perkin Trans. I, (1987) 1757-1762. (c) J. P. Vacca, S. J. deSolms, and J. R. Huff, J. Am. Chem. Soc, 109 (1987) 3478-3479. V. I. Shvets, A. E. Stepanov, V. N. Krylova, P. V. Gulak, myoinositol and Phosphoinositides, Science Publishing House, Moscow, 1987. D. C. Billington, Chem. Soc. Rev., 18 (1989) 83122. B. V. L. Potter, Natural Product Reports, 7 (1990) 1-24.

452

15 16

17

18 19

20 21

22 23 24 25

26

A. B. Reitz (Ed), ACS Symposium Series 463, Inositol Phosphates and Derivatives, Synthesis, Biochemistry, and Therapeutic Potential, American Chemical Society, Washington, D.C., 1991. D. C. Billington, The Inositol Phosphates, Chemical Synthesis and Biological Significance, VCH, Weinheim, 1993. B. V. L. Potter and S. R. Nahorski, in: A. P. Kozikowski (Ed), Drug Design for Neuroscience. Chapter 14: Synthetic Inositol Polyphosphates and Analogues as Molecular Probes for Neuronal Second Messenger Receptors, Raven Press, NY, 1993, pp 3 83-416. S. Ozaki, Y. Watanabe, M. Hirata, T. Ogasawara, T. Kanematsu, and M. Yoshida, in: A. P. Kozikowski (Ed), Drug Design for Neuroscience. Chapter 15: Inositol 1,4,5-Trisphosphate Affinity Chromatography: Fishing out novel Ins(1,4,5)P3-recognizable Proteins, Raven Press, NY, 1993, 417-434. D. C. Billington and R. Baker, J. Chem. Soc. , Chem. Commun. , (1987) 1011-1013. (a) S. Ozaki, Y. Watanabe, T. Ogasawara, Y. Kondo, N. Shiotani, H. Nishii, and T. Matsuki, Tetrahedron Lett., 27 (1986) 31573160. (b) Y. Watanabe, H. Nakahira, M. Bunya, and S. Ozaki, Tetrahedron Lett., 28 (1987) 4179-4180. (c) S. Ozaki, Y. Kondo, N. Shiotani, T. Ogasawara, and Y. Watanabe, J. Chem. Soc. Perkin 1, (1992) 729-737. (a) K.-L. Yu and B. Fraser-Reid, Tetrahedron Lett., 29 (1988) 979-982. (b) G. Baudin, B. I. Glanzer, K. S. Swaminathan, and A. Vasella, Helv. Chim. Acta, 71 (1988) 1367-1378. (c) C. E. Dreef, C. J. J. Elie, P. Hoogerhout, G, A. van der Marel, and J. H. van Boom, Tetrahedron Lett., 29 (1988) 6513-6516. Y. Watanabe, Y. Komoda, K, Ebisuya, and S. Ozaki, Tetrahedron Lett., 31 (1990) 255-256. A. V. Luk'yanov, A. I. Lyutik, V. I. Shvets, and N. A. Preobrazhenskii, Zh. Obsh. Khim., 36, (1966) 1029-1031. V. N. Krylova, A. I. Lyutik, N. P. Gornaeva, and V. I. Shvets, Zh. Obsh. Khim., 51, (1981) 210-214. G. Lin, C. F. Bennett, andM.D. Tsai, Biochem., 29 (1990) 2747-2757. J. L. Meek, F. Davidson, and F. W. Hobbs, Jr., J. Am. Chem. S o c , 110 (1988) 2317-2318. (a) P. A. Gent, R. Gigg, and C. D. Warren, Tetrahedron Lett., (1970) 2575-2578. (b) P. A. Gent, R. Gigg, and C. D. Warren J. Chem. Soc. (C) , (1969) 2367-2371. (c) D. J. R. Massy and P. Wyss, Helv. Chim. Acta, 73 (1990) 1037-1057. (d) C. E. Dreef, R. J. Tuinman, A. W. M. Lefeber, C. J. J. Elie, G. A. van der Marel, and J. H. van Boom, Tetrahedron, 47 (1991) 4709-4722. (e) T. P. Zubkova, Z. Ya. Khrapkova, I. K. Sarycheva, and N. A. Preobrazhenskii, Zh. Org. Khim., 4 (1968) 2226-2228. C. Schultz, T. Metschies, B. Gerlach, C. Stadler, B. Jastorff, Synlett, (1990) 163-165. M. Jones, K. K. Rana, J. G. Ward, and R. C. Young, Tetrahedron Lett., 30 (1989) 5353-5356. S.-K. Chung and Y. Ryu, Carbohydr.Res., 258 (1994) 145-167. For cyclohexylidene ketal synthesis: S. J. Angyal, G. C. Irving, D. Rutherford, and M. E. Tate, J. Chem. S o c , (1965) 6662-6664. C. Jiang and D. C. Baker, J. Carbohydr. Chem., 5 (1986) 615-620. And Ref. 21c. For isopropylidene ketal synthesis: Ref. 21b. For the synthesis of dicyclohexylidene ketals: (a) P. J. Garegg, T. Iversen, R. Johansson, and B. Lindberg, Carbohydr. Res., 130 (1984) 322-326. And Ref. 21c. For the synthesis of diisopropylidene ketals: (b) J. Gigg, R. Gigg, S. Payne, and R. Conant, Carbohydr. Res., 142 (1985) 132-134. (c) J. Gigg, R. Gigg, S. Payne, and R. Conant, J. Chem. Soc. Perkin I, (1987),

453

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

45 46 47 48 49 50 51

2411-2414. (d) R. F. de la Pradilla, C. Jaramillo, J. JimenezBarber o, M. Martin-Lomas, S. Penades, and A. Zapata, Carbohydr. Res., 207 (1990) 249-257. AndRef. 24. J. P. Vacca, S. J. deSolms, J. R. Huff, D. C. Billington, R. Baker, J. J. Kulagowski, and I. M. Mawer, Tetrahedron, 45 (1989) 5679-5702. C. J. J. Elie, R. Verduyn, C. E. Dreef, D. M. Brounts, G. A. van der Marel, and J. H. van Boom, Tetrahedron, 46 (1990) 82438254. C. E. Dreef, C. J. J. Elie, G. A. van der Marel, and J. H. van Boom, Tetrahedron Lett., 32 (1991) 955-958. M. S. Shashidhar, J. F. W. Keana, J. J. Volwerk, and 0. H. Griffith, Chem. Phys. Lipids, 53 (1990) 103-113. J. J. Kulagowski, Tetrahedron Lett., 30 (1989) 3869-3872. K. M. Pietrusiewicz, G. M. Salamo' nczyk, K. S. Bruzik, and W. Wieczorek, Tetrahedron, 48 (1992) 5523-5542. S. Ozaki, Y. Kondo, H. Nakahira, S. Yamaoka, and Y. Watanabe, Tetrahedron Lett., 28 (1987) 4691-4694. F. Tagliaferri, S.-N. Wang, W. K. Berlin, R. A. Outten, and T. Y. Shen,Tetrahedron Lett., 31 (1990) 1105-1108. (a) D.-M. Gou, Y.-C. Liu, and C-S, Chen, Carbohydr. Res., 234 (1992) 51-64. (b) D.-M. Gou, W.-R. Shieh, P.-J. Lu, and C.-S. Chen, Bioorg. Med. Chem., 2 (1994) 7-13. P. J. Garegg, B. Lindberg, I. Kvarnstrom, and S. C. T. Svensson, Carbohydr. Res., 173 (1988) 205-216. C. Murakata and T. Ogawa, Tetrahedron Lett., 31 (1990) 24392442. S. Danishefsky and R. Hungate, J. Am. Chem. Soc. , 108 (1986) 2486-2487. S. Danishefsky, R. Hungate, and G. Schulte, J. Am. Chem. S o c , 110 (1988) 7434-7440. N. Nagashima and M. Ohno, Chem. Lett., (1987) 141-144. D.-M. Gou and C.-S. Chen, Tetrahedron Lett., 33 (1992) 721-724. D.-M. Gou and C.-S. Chen, J. Chem. S o c , Chem. Commun., (1994) 2125-2126. C. E. Dreef, W. Schiebler, G.A. van der Marel, and J. H. van Boom, Tetrahedron Lett., 32 (1991) 6021-6024. P. Westerduin, H. A. M. Willems, and C. A. A. van Boeckel, Tetrahedron Lett., 31 (1990) 6915-6918. C. E. Dreef, R. J. Tuinman, C. J, J. Elie, G.A. van der Marel, and J. H. van Boom,Reel. Trav. Chim. Pays-Bas, 107 (1988) 395397. R. Baker, J. J. Kulagowski, D. C. Billington, P. D. Leeson, I. C. Lennon, and N. Liverton, J. Chem. S o c , Chem. Commun., (1989) 1383-1385. See also: S. J. Angyal, M. E. Tate, and S. D. Gero, J. Chem. S o c , (1961) 4116-4122. K. S. Bruzik, J. Myers, and M.-D. Tsai, Tetrahedron Lett., 33 (1992) 1009-1012. L. Ling and S. Ozaki, Tetrahedron Lett., 34 (1993) 2501-2504. L. Ling and S. Ozaki, Carbohydr. Res., 256 (1994) 49-58. K. S. Bruzik and M.-D. Tsai, J. Am. Chem. S o c , 114 (1992) 6361-6374. Y. Watanabe, T. Ogasawara, S. Ozaki, and M. Hirata, Carbohydrate Research, 258 (1994) 87-92. G. M. Salamo'nczyk and K. M. Pietrusiewicz, Tetrahedron Lett., 32 (1991) 6167-6170. C. Schultz, G. Gebauer, T. Metschies, L. Rensing, and B. Jastorff, Biochem. Biophys. Res. Commun., 166 (1990) 1319-1327. Y. Watanabe, M. Mitani, T. Morita, and S. Ozaki, J. Chem. S o c , Chem. Commun., (1989) 482-483.

454 52 53 54 55 56 57 58 59 60 61 62 63 64

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

Y. Watanabe, H. Hirofuji, and S. Ozaki, Tetrahedron Lett., 3 5 (1994) 123-124. A. P. Kozikowski, A. H. Fauq, G. Powis, P. Kurian, and F. T. Crews, J. Chem. Soc. , Chem. Cornmun. , (1992) 362-364. H. W. Lee and Y. Kishi, J. Org. Chem., 50 (1985) 4402-4404. D. C. Billington, R. Baker, J. J. Kulagowski, I. M. Mawer, J. P. Vacca, S. J. deSolms, and J. R. Huff, J. Chem. Soc. Perkin 1, (1989) 1423-1429. T. Banerjee and S. M. Srikantian, Tetrahedron Lett., 35 (1994) 8053-8056. T. Desai, J. Gigg, R. Gigg, S. Payne, S. Penades, and H. J. Rogers, Carbohydr. Res., 216 (1991) 197-209. Gigg eq 1, 2; T. Desai, A. Fernandez-Mayoralas, J. Gigg, R. Gigg, and S. Payne, Carbohydr. Res., 205 (1990) 105-123. C. Liu, S. R. Nahorski, and B. V. L. Potter, J. Chem. S o c , Chem. Commun., (1991) 1014-1016. A. Aguilo, M. Martin-Lomas, and S. Penades, Tetrahedron Lett., 33 (1992) 401-404. A. Zapata, R. F. de la Pradilla, M. Martin-Lomas, and S. Penades, J. Org. Chem., 56 (1991) 444-447. Y. Watanabe, T. Shinohara, T. Fujimoto, and S. Ozaki, Chem. Pharm. Bull., 38 (1990) 562-563. Y. Watanabe, T. Fujimoto, T. Shinohara, and S. Ozaki, J. Chem. S o c , Chem. Commun., (1991) 428-429. S. Ozaki and Y. Watanabe, in: A. B. Reitz (Ed), ACS Symposium Series 463, Inositol Phosphates and Derivatives, Synthesis, Biochemistry, and Therapeutic Potential. Chapter 4: Synthesis of Inositol Polyphosphates and their Derivatives, American Chemical Society, Washington, D.C., 1991, pp. 43-65. Y. Watanabe and N.Sasaki, unpublished results. Y. Watanabe, T. Ogasawara, H. Nakahira, T. Matsuki, and S. Ozaki, Tetrahedron Lett., 29 (1988) 5259-5262. M. Brufani, M. C. Cesta, L. Donnarumma, L. Filocamo, G. Gostoli, S. Lappa, Carbohydr. Res., 228 (1992) 371-376. C.-Y. Yuan, H.-X. Zhai, and G.-H. Weng, Chinese J. Chem., 12 (1994) 174-178. S.-K. Chung, Y.-T. Chang, and K.-H. Sohn, Korean J. Med. Chem., 4 (1994) 57-65. Y. Watanabe, A. Oka, Y. Shimizu, and S. Ozaki, Tetrahedron Lett., 31 (1990) 2613-2616. Y. Watanabe, T. Fujimoto, and S. Ozaki, J. Chem. S o c , Chem. Commun., (1992) 681-683. P. J. Edwards, D. A. Entwistle, C. Genicot, K. S. Kim, and S. V. Ley, Tetrahedron Lett., 35 (1994) 7443-7446. C.-H. Wong and G. M. Whitesides, Tetrahedron Organic Chemistry Series Vol. 12, Enzymes in Synthetic Organic Chemistry, Pergamon, Oxford, 1994. Y.-C. Liu and C.-S. Chen, Tetrahedron Lett., 30 (1989) 16171620. L. Ling, X. Li, Y. Watanabe, T. Akiyama, and S. Ozaki, Tetrahedron Lett., 33 (1992) 1911-1914. L. Ling, Y. Watanabe, T. Akiyama, and S. Ozaki, Bioorg. Med. Chem., 1 (1993) 155-159. S. Ozaki, L. Ling, T. Ogasawara, Y. Watanabe, M. Hirata, Carbohydr. Res., 259 (1994) 307-310. Y. Watanabe, M. Tomioka, S. Ozaki, manuscript to be prepared. P. Andersch and M. P. Schneider, Tetrahedron Asymmetry, 4 (1993) 2135-2138. S. Ozaki, Y. Koga, L. Li, Y. Watanabe, Y. Kimura, M. Hirata, Bull. Chem. Soc. Jpn., 67 (1994) 1058-1063.

455 80

T. Hudlicky, J. D. Price, F. Rulin, and T. Tsunoda, J. Am. Chem. S o c , 112 (1990) 9439-9440. 81 H. A. J. earless, Tetrahedron Lett., 33 (1992) 6379-6382. 82 M. Mandel and T. Hudlicky, J. Chem. Soc. Perkin I, (1993) 741743. 83 S. Cerdan, C. A. Hansen, R. Johanson, T. Inubushi, and J. R. Williamson, J. Biol. Chem., 261 (1986) 14676-14680. 84 A. E. Ting and R. E. Pagano, Chem. Phys. Lipids, 60 (1991) 8391. 85 R. E. Pagan and R. G. Sleight, Science, 229 (1985) 1051-1057. 86 For a review article, see: R. J. Ferrier and S. Middleton, Chem. Rev., 93 (1993) 2779-2831. 87 (a) K. Sato, S. Sakuma, S. Muramatsu, and M. Bokura, Chem. Lett., (1991) 1473-1474. (b) K. Sato, M. Bokura, and M. Taniguchi, Bull. Chem. Soc. Jpn., 67 (1994) 1633-1640. 88 Y. Watanabe, M. Mitani, and S. Ozaki, Chem. Lett., (1987) 123126. 89 For review articles, see: M. P. Czech, J. K. Klarlund, K. A. Yagaloff, A. P. Bradford, and R. E. Lewis, J. Biol. Chem., 263 (1988) 11017-11020. A. R. Saltiel and P. Cuatrecasas, Am. J. Physiol., 255 (1988) Cl-Cll. 90 C. Jaramillo and M. Martin-Lomas, Tetrahedron Lett., 32 (1991) 2501-2504.C. Jaramillo, J.-L. Chiara, and M. Martin-Lomas, J. Org. Chem., 59 (1994) 3135-3141. 91 S. L. Bender and R. J. Budhu, J. Am. Chem. S o c , 113 (1991) 9883-9885. 92 V. A. Estevez and G. D. Prestwich, J. Am. Chem. S o c , 113 (1991) 9885-9887. 93 H. B. Mereyala and S. Guntha, J. Chem. Soc. Perkin 1, (1993) 841-844. 94 0. Arjona, A. de Dios, R. F. de la Pradilla, and J. Plumet, Tetrahedron Lett., 32 (1991) 7309-7312. 0. Arjona, A. Candilejo, A. de Dios, R. F. de la Pradilla, and J. Plumet, J. Org. Chem., 57 (1992) 6097-6099. 95 A. P. Kozikowski, A. H. Fauq, and J. M. Rusnak, Tetrahedron Lett., 30 (1989) 3365-3368. 96 A. P. Kozikowski, A. H. Fauq, G. Powis, D. C. Melder, J. Am. Chem. S o c , 112 (1990) 4528-4531. 97 A. P. Kozikowski, A. H. Fauq, J. Am. Chem. S o c , 112 (1990) 7403-7404. 98 A. P. Kozikowski, W. Tiickmantel, and G. Powis, Angew. Chem. Int. Ed. Engl., 31 (1992) 1379-1381. 99 A. H. Fauq, A. P. Kozikowski, A. Gallegos, and G. Powis, Med. Chem. Res., 3 (1993) 17-23. 100 S. C. Johnson, J. Dahl, T.-L. Shih, D. J. A. Schedler, L. Anderson, T. L. Benjamin, and D. C. Baker, J. Med. Chem., 3 6 (1993) 3628-3635. 101 T. Akiyama, N. Takechi, S. Ozaki, Bull. Chem. Soc. Jpn., 65 (1992) 366-372. 102 W. Tegge and C. E. Ballou, Proc Natl. Acad. Sci. USA, 86 (1989) 94-98. 103 S. V. Ley and L. L. Yeung, Spec. Publ. Royal Soc. Chem., Ill (Molecular Recognition: Chemical and Biochemical Problems II), (1992) 183-191. S. V. Ley, Pure Appl. Chem., 62 (1990) 20312034. 104 S. V. Ley, M. Parra, A. J. Redgrave, and F. Sternfeld, Tetrahedron, 46 (1990) 4995-5026. 105 S.V. Ley and L. L. Yeung, Synlett, (1992) 997-998. 106 E. J. Corey, R. L. Danheiser, and S. Chandrasekaran, J. Org. Chem., 41 (1976) 260-265. J. E. McMurry and J. G. Rico,

456

107 108 109 110 111

112 113 114 115 116 117 118 119 120

Tetrahedron Lett., 30 (1989) 1169-1172. J. E. McMurry and N. 0. Siemers, Tetrahedron Lett., 34 (1993) 7891-7894. J. L. Chiara and M. Martin-Lomas, Tetrahedron Lett., 35 (1994) 2969-2972. J. P. Guidot, T. L. Gall, and C. Mioskowski, Tetrahedron Lett., 35 (1994) 6671-6672. Other report: J. L. Chiara, W. Cabri, and S. Hanessian, Tetrahedron Lett., 32 (1991) 1125-1128. J. P. Guidot and T. L. Gall, Tetrahedron Lett., 34 (1993) 46474650. A. E. Stepanov and V. I. Shvets, Chem. Phys. Lipids, 25 (1979) 247-263. A. E. Stepanov and V. I. Shvets, ibid., 41 (1980) 151. J. R. Falck and A. Abdali, in: A. B. Reitz (Ed), ACS Symposium Series 463, Inositol Phosphates and Derivatives, Synthesis, Biochemistry, and Therapeutic Potential. Chapter 11: Enantiospecific Synthesis of Inositol Polyphosphates, Phosphatidylinositides, and Analogues from (-)-Quinic Acid, American Chemical Society, Washington, D.C., 1991, pp 145-154. J. R. Falck and A. Abdali, Bioorg. Med. Chem. Lett., 3 (1993) 717-720. Y. Watanabe and T. Ogasawara, unpublished results. C. Murakata and T. Ogawa, Tetrahedron Lett., 32 (1991) 671-674. C. Murakata a n d T . Ogawa, Carbohydr. Res., 235 (1992) 95-114. D. R. Mootoo, P. Konradsson, and B. Fraser-Reid, J. Am. Chem. S o c , 111 (1989) 8540-8542. U. E. Udodong, R. Madsen, C. Roberts, and B. Fraser-Reid, J. Am. Chem. S o c , 115 (1993) 7886-7887. A. E. Stepanov, V. I. Shvets, and R. P. Evstigneeva, Zh. Obs. Khim., 47 (1977) 1653-1656. C. J. J. Elie, C. E. Dreef, R. Verduyn, G. A. van der Marel, and J. H. van Boom, Tetrahedron, 45 (1989) 3477-3486. C. J. J. Elie, C. E. Dreef, G. A. van der Marel, and J. H. van Boom, J. Carbohydr. Chem., 11 (1992) 715-739. The results was presented at the 69th National Meeting of The Chemical Society of Japan, March, 1995. Y. Watanabe, T. Yamamoto, T. Okazaki, and S. Ozaki, manuscript to be prepared.


457

Synthesis of Phytosphingolipids Tadao Kamikawa

1.

INTRODUCTION Lipids occur throughout the living world in microorganisms, plants and animals. Recently, it has become well known that lipids not only contribute to the structure of cells and provide an energy store, they also participate in the transmission of chemical signals in living organisms. The majority of lipids are based on a glycerol backbone. Another important group of acyl lipids, however, have sphingosine-based structures. Sphingolipids have a rather specific distribution in cells where they are concentrated in the outer leaflet of the plasma membrane. Glycosphingolipids are recognized as mediators of cell-cell recognition and communication, cellgrowth regulators, cell immune response, and cell oncogenic transformation. The analysis and identification of sphingolipids have recently been substantially improved by the use of gas-liquid chromatography, fast-atom bombardment mass spectrometry, nuclear magnetic resonance spectroscopy and selective enzymatic cleavage. Despite these efforts, however, many important members of this class of biomolecules remain relatively inaccessible. Isolation of pure compounds is still difficult because of the diversity and heterogeneity of lipids. Effective synthetic routes to these compounds are, therefore, extremely important to investigate their chemical and biological properties. The chemistry of sphingolipids is rather well documented,! so the recent developments in the synthesis of phytosphingolipids will be presented in this review. 2.

BASIC STRUCTURE OF SPHINGOLIPIDS Sphingosine is a long-chain amino alcohol. Several long-chain amino alcohols which occur in nature are shown in Figure 1.

NH2 sphinganine, dihydrosphingosine, D-e/yf/)ro-2-amino-1,3-dihydroxyoctadecane

458

4-sphingenine, sphingosine, D-eAyf/7ro-2-amino-1,3-clJhydroxy-frans-4-octaclecene OH

OH

NH2

4-hydroxysphlnganine, phytosphingosine, D-f/bo-2-amino-1,3,4-trlhyclroxyoctaclecane OH

OH

NH2

4-hyclroxy-8-sphJngenine, dehydrophytosphingosine, D-rtoa-2-amino-1,3,4-trihydroxy-c/s-8-octadecene

Fig. 1. Long-Chain Amino Alcohols

Both the amino and the alcohol moieties of sphingosine can be substituted to produce the various sphingolipids {Fig. 2).

y Cerebroside

Ceramide < Sugar

^ Natural Sphingolipid

Sugar—Sugar

Sphingosine (Long-Chain Base)

Fig. 2. Sphingolipids

Ceramides are N-acylated sphingosines. This acyl linkage is resistant to alkaline hydrolysis and therefore can be easily distinguished from the 0-esters found in glycerol-based acyl lipids. Attachment of hexosides to ceramides yields cerebrosides. The naming "cerebroside" was originally used for the galactosyl ceramide of the brain but is now widely used for monoglycosyl ceramides. Further attachment of hexosides to cerebrosides yields natural glycosyl sphingolipids. These are conveniently written in abbreviated form.

459 e.g., Galal-44Galpl-^lCer (Gal: galactopyranosyl; Cer: Ceramide). Some glycosphingolipids contain one or more molecules of sialic acid in the sugar residues of ceramide oligosaccharide. These lipids are called gangliosides. Esterification of the alcohol moiety of the sphingosine base with phosphocholine yields sphingomyelin. 3.

BIOLOGICAL ACnVITIES OF SPHINGOLIPIDS Monogalactosyl ceramides are the largest single component of the myelin sheath of nerves. They are also found in the lung, kidney, liver, spleen, serum and almost all tissues although in trace amounts. Essentially all the glycosphingolipids are immunologically active, either in heptenic reactivity in vitro or in antibody-producing potency. Glycosphingolipids play key roles in many biological processes. They have been shown to be cell-surface receptors for viral and bacterial toxins, ^ regulators of cell proliferation by interacting with transmembrane signal transducers and to be mediators of cell-cell recognition events.^ Gangliosides (e.g. I) at the surface of leukocytes, which carry the Sialyl Lewis^ or Sialyl Lewis^ epitopes, are likely to play an important part in inflammatory responses.^ LÔH

CO.H

o,

.OH

OH

OH ^Ci3H2

HO

unÔH

I

NHAc

OH

BuLi

29

30

j^

OBn

BnO>.

C14H29

C14H21

95%

OBn

72%

N3

NaNa

OBn

1. HS(CH2)3SH EtaN 2. C23H47COOH

CL,

32

CH3

NH

/BU3N

r

72%

H2/10%Pd-C

OBn

BnO.

^.

2. TsNHNHs AcONa

OBn

65%

OMs OBn BnO.

C 12^21

94%

C14H29

C14H29

Scheme 4

An alternative versatile starting material is 2,4-Obenzylidene-D-threose 35, which is readily obtained from D-galactose. Retrosynthetic analysis indicates that the 2,4-disubstituted D-threose derivative and carbanions would offer a shorter route to D-ribophytosphingosines as well as D-ribodehydrophytosphingosines (Scheme 5). NH2 OH

N3

OH

OH

O

RO, OR

Scheme 5

Schmidt's group ^^ reported the first synthesis of the lactosyl ceramides 44 (Scheme 6). Reaction of excess n-tetradecylmagnesium bromide with 35 in THF at 60°C afforded a 1:1 mixture of D-arabino- and L-xyZo-octadecanetetrol derivatives, 36a and 36b, respectively. Reactions at lower temperature or in different solvent did not result in a dramatic change in the ratio of 36a and 36b. It is interesting, however, that the addition of salts, such as the copper bromide-dimethyl sulfide complex or titanium tetrachloride, as a catsdyst led

466

to preferential or exclusive formation of 36b. This preferential formation of 36b is probably due to si-face attack of the nucleophile on the metal complex with three oxygen functions as in A.

s/face A

After separation of the diastereomers, the required C(2)-0-activation of 36a was simply done by treatment of methanesulfonyl chloride in pyridine to give 3 7 . This favorable chemoselectivity is attributed to the oxygen atoms of the 1,3-dioxane ring causing increased nucleophilicity of C(2)-OH through accumulation of lone-pair orbitals and/or through higher acidity because of hydrogen bonding. Treatment of 37 with sodium azide afforded the azide 3 8 which was deprotected by acid treatment to give 39, For the synthesis of glycosphingolipid, the azide 3 9 was first partially protected with 2,2-dimethoxypropane and p-toluenesulfonic acid to give the 3,4-O-isopropylidene derivative 40. Glycosylation of 4 0 with the lactosyl donor 41 using Schmidt's protocol^^ gave the 1-O(p-Iactosyl)phytosphingosine derivative 42. Reduction of the azide 42 followed by acylation of the resulting amine with palmitoyl chloride gave 43 which was deprotected to furnish the desired glycophytosphingolipid 44. OH O^ J^ Ph^ÛôCZjLy^ 35

+

BrMg(CH2)i3CH3

OH

O^f

Ph-..Ô^X^l^(QH2)^3CH3 H 36a 35%

THF/60°C

OH

+

0^»i*

^^^--^^•^--^T^(CH2)i3CH3 HO 36b 36%

MsCI/py OMs lOH

NaN3/DMF

Ph^0CZf.(CH2)i3CH3 H ^ ^"" "

63% ""'°

O

37

^

OH

HCI

P^-^0CZ^^0H2).3CH3 H ' - ' — -

65%

38

467

39

41 OAc (

OAc

^^^^

ISJ3

Aco-r;:^ôr^-q o,A_(CH,,,3CH3

^^"^^

ArON

69%

^ ^

OAc

^'^^

OAc

..

o

1. LiAIH4/NiCl2

^'^^

2. CH3(CH2)i4COCI

X

Ci2H25 BnO^x^-^X/^'/^

1.Na/liqNH3 2. Ac20/py/DMAP 62%

Ac 1

H ^ y v^Ci2H25 ACO-^N*'Ay/^/^

OBn

OAc

83

04a

Boc

Ac 1

on

OU

+

BnO'ô*'%/^'/u

£

T

1.(Boc)20/NaOH 2. MsCI/py/DMAP 3. NaH/DMF 72%

H ^ v v>Ci2H25 Ac0^x^"^\/^''l_j

>H25

OAc

1.TFA 2. H2/Pd-C 3. Ac20/py/DMAP 50%

OBn 85 or

04^

1.TFA 2. AC20/py/DMAP 3. H2/Pd-C/H* 4. Ac20/py/DMAP 23%

Scheme 13

" Ac I

"

^

fteCÔi Ac

...„. _

R' -R^ H" O A C

AcO H

Ac I

R^

O^R'

Acd H Scheme 14

Ac

=^ R^ O^^' AcO H

475

(f) From L-Ascorbic Acid New ceramide digalactosides were isolated by Hayashi's group from the marine sponge Halichondria japonica.'^^ To the major glycosphingolipid was assigned the structure 86, except for the stereochemistry, using FAB/MS, IR and iH NMR spectroscopy. The first synthesis of the ceramide 8 7 and therefore the structure determination were described by us.25

"T

^cT/OH OH

OH

0=

NH OH OH 86

The synthetic strategy (Scheme 15) for assembling the phytosphingosine 88 was the stereoselective ring opening of the epoxide 90 with 2-alkyl-2-lithio1,3-dithiane A, followed by functional transformation to give a precursor B. The requisite consecutive stereochemistry of the target intermediate C could be obtained by means of Dondoni's protocol^G using 2-trimethylsilylthiazole 89 (see p. 2?). (2R)-Hydroxydocosanoic acid 91 could also be derived from the epoxide 90. The synthesis began with the treatment of 3 , 4 - a n h y d r o - l , 2 - 0 isopropylidene-D-erythritol 9 0 with 2-alkyl-2-lithio-l,3-dithiane 92 to give 93 (Scheme 16). Reductive desulfurization of 93 and transacetalization of the resulting 94 by the following reaction sequence (1. acidic hydrolysis; 2. protection of the primary hydroxyl group; 3. ketalization; and 4. basic hydrolysis) afforded the primary alcohol 95. The Swern oxidation of 9 5 yielded the aldehyde 96. To create a new chiral center at C(2) and to introduce a hydroxymethylene group simultaneously, Dondoni's method^S was used. Treatment of the aldehyde 96 with 2-(trimethylsilyl)thiazole 89 afforded the highly diastereoselective adduct 97, but it had an undesired configuration at the new chiral center. The high diastereoselectivity is attributed to the preferred transition state D. Attempts to invert the configuration at C(l) were fruitless, owing to severe steric hindrance. So a two-step oxidation-reduction sequence was used to obtain the C(l)-C(2) syn product. Reduction of the ketone 98 with NaBH4 in the presence of CeCls afforded the best result (99:97 = 85:15). The

476 preferred formation of 99 may be rationalized based on the transition state E by assuming that complexation of the cerium ion occurs between the O atom of the isopropylidene group and the N atom of the thiazole ring and that the hydride ion attacks from the less hindered side. A series of protection, methylation, reduction, hydrolysis and reduction steps provided the alcohol 100, which was subjected to transformation into the alcohol 101. Final conversion of the alcohol 101 into phytosphingosine 102 was carried out in the usual manner. The EDC-mediated condensation of (2K)-benzoyloxydocosanoic acid^^ and the amine 102 followed by debenzylation gave t h e ceramide monobenzoate 103. OH

NHR O H '^{CH2)iiCHMe2

HOOC''^C2oH4i 91

87

R = (2R)-C2oH4iCH(OH)CO

88

R=H

OH OP^ PÔ.

(CH2)iiCHMe2 OP^

HOOy^C2oH4i OH

C

OP"*

H

>-SiMe3

89

+

OHC.

"Y'^(CH2)iiCHMe2 p3 OP^

V-0

(CHgjgCHMea

90

B

OH

Scheme 15

A

- ^

OH

L-ascorbic acid

477

Raney Ni (CH2)9CHMe2

HQ (CH2)9CHMe2

I.PTSA/MeOH 2. PivCI/py 11—^

HO-^ (CH2)iiCHMe2 )—( ^ ^

3. Me2C(OMe)2/PPTS 4. LiOH

(CH2)iiCHMe2 W

o

^

o

1.lL/>-SiMe3 " 75 2.TBAF

97%

(CH2)iiCHMe2

O

O

O

99%

NaBH4 CeClgyHgO

O

4. CuCl2/CuO

X

^ ^

1.PTSA/MeOH 2. BnBr/NaH

1. PMBCI/NaOH 2. Mel 3. NaBH4

^S OH || / ) — ( (CH2)iiCHMe2 o

3. DDQ

^-^^^^^ 9d

61%

OH OBn o n 1 ^ ^""-^ ^ ^ " ^ ( C H 2 ) i i C H M e 2 "^g^

81%

^MsCI ^' '"'^^ 3. LiAIH4 56%

100

101 QBz

NH2 OBn BnO A J^ ^^•-^^Y"^(CH2)iiCHMe2 OBn

^

gj

90

/ ^

(C0CI)2/DMS0 Et3N

^

96%

P'^'^^ H2)iiC H 0 H 2^C- -- ^^ ( ( C (CH2)iiCHMe2 0

o o

MeOH

X

^ ^

(CH2)nCHMe2

86%

95

^S p [I / ^ - ^

95

| ^ W ^N W

^

(C0CI)2/DMS0 EtgN 1 ^

V / \

64%

\0

89%

1. (2R)-benzoyloxydocosanoic acid ^^^ ^ 2.H2/Pd-C 84%

Ov^Â^ 7 (CH2)9Me NH OH HO X A ^^^^^^Y'^(^^2)iiCHMe2 OH 103

Scheme 16

478

SiMea

(g) From (S)-Malic Acid Guanti's group^s has developed a new "electrophilic amination" method (Scheme 17) for the P-hydroxyester 107, derived from dimethyl(S)-malate 105, with di-tert-butylazodicarboxylate 108. By performing the reaction at -50 °C a moderate selectivity {erythro:threo = 67:33) was observed. The two isomers 109e and 109t were easily separated and 109e was converted into the N , 0 isopropylidene acetal 110. Reduction of 1 1 0 with calcium borohydride and successive protection and deprotection gave the primary alcohol 112, The Swem oxidation of 1 1 2 followed by treatment with lithium tetradecyne in the presence of HMPA preferentially yielded the anti adduct {113a:113s = 85:15). The alkyne 113a was converted by the u s u a l reaction sequence into tetraacetyl D-ribo-Cis-phytosphingosine 104. OH

A^CO, Me

BH3/NaBH4{cat)

Me02C

OH HO^^^Âs^COgMe

83%

105

OH TBDRSO^^Â^COaMe 107

''•LDAn-HF

109B

71%

^.

71%

106

OH

OH T B D P S O ^ ^ ^ ^ C O J Me

TBDPSOs^^^x^s^COaMe

Boc'' "NHBoc

Boc'^^'NHBoc

2. BocN=NBoc 108 62%

MeOC(CH3)=CH2 PTSA

f-BuPhgSiCI imidazole

109t

109e

NaBHVCaClg

TBDPSO, COaMe 110

0^N'^'"= TBDPSO.

N.

90%

Boc

OH 111

479

1. MEMCI/EtsN 2. n-Bu4NF

V

86%

i

xBoc

T "°^ ÔMEM

rr2

X.

Î

, ^ ,Boc Î

A..

I

1. (C0CI)2/DMS0 EtgN

?

2. n-Ci2H25CECLi THF-HMPA 71%

LAcOH/HCI 2. ^. H2/Pt02 n2/riU2

T T OH

7

^Boc

^°^ K^^^^

113a : anti 113s : syn

_ . .,L,. ^'^'^ 1 1^^"^

3.AC2O

^^ ^^

rrja

A 104

Scheme 17

(h) From (S)-Serine Diastereoisomeric D-ribo, D-Iyxo-, D-arabino-, and D-xyZo-Cie-phyto sphingosine tetraacetates were S5aithesized from the oxazoline 114, derived from (S)-serine, by Komori's group29 (Scheme 18). Treatment of the alkenylalane 115, prepared from n-tridecyne and DIBAH, with the oxazohnecarbaldehyde 114 gave a mixture of 116a and 116b. After chromatographic separation, the allylic alcohol 116a was then oxidized with vanadyl acetylacetonate and tert-butylhydroperoxide to give a mixture of the diastereomeric epoxy alcohols 117a and 117b in a 3:2 ratio. The epoxy alcohol 117a was reduced with DIBAH to give an inseparable mixture of 1 1 8 a and 118b. Debenzylation and acetylation gave D-ribo-Ciephytosphingosine tetraacetate 119a and its isomerl 1 9 b . Other diastereomers of phytosphingosine were synthesized by a similar method.

r^ HO. L-serine

^

'^V > =MN ^^^

115

480

(CH2)ioCH3 OH 116b

17%

V0(acac)2 TBHP Ph.

Ph.

OH 117b

33%

DIBAH

BnHN

BnHN H O \ . A s ^ - - v . ^ (CH2)ioCH3

OH ^(CH2)iiCH3 OH

OH

118a

118b

OH

1. Pd-C/cyclohexene/HCI 2. Ac20/py

AcHN AcO.

AcHN

OAc (CH2)iiCH3

+

AcO.,Â.,^-s,^(CH2)ioCH3 OAc OAc 119b 24% from 117a

Scheme 18

4.3 Chiral Induction Obtaining optically active compounds by chiral induction represents a more refined solution to organic chemists. The kinetic resolution developed by Katsuki-SharplessÔ for sdlylic alcohols is superior in enantiotopic face differentiation and in versatility. Another interesting chiral induction method has been developed by Dondoniî using 2-(trimethylsilyl)thiazole as a masking formyl group. These methods are

481 more efficient because stereoselectivity.

the desired isomer is obtained

with

high

(a) Katsuki-Sharpless Epoxidation The first asymmetric synthesis of a phytosphingosine was accomplished by Komori's group^^ (Scheme 19). Racemic 120 was kinetically resolved by asymmetric epoxidation using (+)-diisopropyl tartrate as a chiral auxiliary to give the (4R)-allylic alcohol 121 and the (4S)-epoxy alcohol 122. Protection of 121 followed by ozonolysis gave the aldehyde 123. The aldehyde 123 was then converted into the epoxy alcohol 126 by means of 1. Homer-Emmons reaction, 2. DIBAH reduction and 3. Katsuki-Sharpless epoxidation. After conversion of 126 into the benzyl urethane 227, it was treated with sodium hydride to give the 2-oxazolidinone 128 via intramolecular base-catalyzed epoxide opening. Subsequent hydrolysis of 128 followed by cleavage of the benzyl and the MOM ether group gave the Cie-phytosphingosine 129. They also reported the synthesis of acanthacerebroside A^s 134^ which was isolated from the starfish Acanthaster planci.^^ DCC-mediated condensation of 129 and (R)-2-acetoxytetradocosanoic acid 130 afforded the ceramide monoacetate 131. The glycosidation of 131 v/ith 2,3,4,6-tetra-Oacetyl-p-D-glucopyranosyl bromide 132 in the presence of silver triflate and molecular sieves gave a mixture of the p-monoglycoside 133 and a diglycoside in 37 and 15% yield, respectively. Hydrolysis of 133 yielded acanthacerebroside A 134. Ti(0Pr^4/(+)-DIPT TBHP n-Ci2H25

^-012^25

+

n-Ci2H25

120

1.MOMCI 2. O3; Me2S 121

OMOM n-Ci2H25 ^

"CHO

OMOM

(EtO)2P(0)CHC02Et ggo/^ ^^Q^ ^21

^-î2H25 124

123

DIBAH 82%

OMOM

^.

ÔH ^-C 12^25 125

C02Et

Ti(0PrV(-)-DIPT TBHP ^ 63%

OMOM n-Ci2H25 126

482 OMOM NaH 72%

97% r27

OMOM, n-Ci2H25

MOMO

-'/AN ^

[I 0

Bn

n-C,2H25

A

Bn.

P

N--\

f ^ OH

^

t25 LNaOH 2. Pd-C/cyclohexene HCI

OH n-Oi2H25

3. HCI

NH3 CI

y

^^

OH

59%

t29

Scheme 19 OAc HoAy.C22H46 CI HgN

OAc 130

OH

HO.

(CH2)iiCH3 OH

NH

73%

AcO-

131

] ^ (CH2)2lCH3 ^,^ Q,^

AcO-j

/-O

ÂcV AcO^^ OAc

AgOTf 37%

o X

A

^^-'"^Y'^^^"2)iiCH3 OH 133

OH "^f

1 ^

^(CH2)iiCH3 OH

OAc

t32

(CH2)2iCH3

VO-^x^Y^(CH2)iiCH3

HO^'—f OH

OH

H0>.

DCC/HOBT

129 -OBr DAcS AcO , OAc

Y'^^^^2)2lCH3

OH 134 Scheme 20

K2CO3

483

(b) Dondoni Carbon Chain Extension Method Chain elongation of aldehydes via 2-(trimethylsilyl)thiazole (2-TST) 135 involves two key steps: A, construction of a chiral hydroxyalkyl chain at C(2) of the thiazole ring (functionalization); B, the liberation of the aldehyde by cleavage of the thiazole ring (unmasking). The mechanism of the functionalization step is shown in Scheme 21.30 R

R

r-N'ÔSiMes

RCHO

RCHO

,. - N

OSiMes

^S^SiMea 135

136

137

i

/ R

r\

\

R N-^0SiMe3

- RCHO

^g-^^^ÔSiMea

R

R

p

138

140

139

Scheme 21

The initial reaction is 1,2-addition of the carbon-silicon bond of 1 3 5 to an aldehyde giving the thiazolium 2-ylide 136. The 2-ylide 136 may then react with a second aldehyde molecule to give a 2:1 adduct 137 which in turn react with a third aldehyde molecule to give a 3:1 adduct 139. Silyl migration of 1 3 7 followed by removal of an aldehyde molecule would yield the 2substituted thiazole 140. Dondoni's group^S described the synthetic utility of 2-TST for various Nprotected a-amino aldehydes. Application for the synthesis of phytosphingosine is outlined in Scheme 22.

y.„-Boc OHO 141

BU4NF ^S^SiMea 135

85% OSiMea

484 \ /

^Boc

\ /

,Boc

OSiMes

OSiMea

142a

1^2b 1.Mel 2. NaBH4 3.HgCl2

I

65% 1. BnBr/NaH

-hn:

Boc

BU4NI

0^>v,.CHO ^^^ y OH

Boc

Ov

'

1. BnBr 2. Mel

143

^1 ^ ^ N

2.2-TST 64%

^

S-N

QBn CHO

Boc

OH

+

N. i V // OBn S -

0

144a

n-Ci3H27PPh3 Br

>J .Boc 7-N OBn

Raney Ni 'C12H21

n-BuLi

70%

66%

-N

Boc

OH C14H2

I.CF3CO2H/H2O 2. Ac20/py

AcHN _

OAc _ C14H29

57%

Scheme 22

The reaction between equimolar a m o u n t s of 2-TST and N-tertbutoxycarbonyl-L-serinol acetonide 141 occurred smoothly at room temperature to give, after desilylation of the resulting adducts with tetrabutylammonium fluoride, a separable mixture of amino alcohols 142a and 142b in high diastereoselectivity [142a:142b = 92:8). This high anti diastereoselectivity may be attributed to the Felkin-Anh open-chain model for as3mimetric induction,^^ and the preferred transition state is presented in A.

485

A

The unmasking protocol consists of three sequential operations: Nmethylation, reduction, and hydrolysis. Thus 142a was converted into the aldehyde 143, which after protection was subjected to a further one-carbon homologation. The addition of 2-TST in dichloromethane at room temperature was rather unselective (ds = 60%) but became quite diastereoselective by using tetrahydrofuran at 0 °C giving adducts 144a and 144b in an 85:15 ratio. The anti configuration of the major isomer 144a is again consistent with the non-chelate Felkin-Anh model for diastereoselection. Protection and unmasking provided the aldehyde 145. The synthesis of D-ribo-Cis-phytosphingosine 104 was achieved by a series of Wittig olefination, reduction, hydrolysis and acetylation steps. (c) Catalytic Asymmetric Aldol Reaction Kobayashi's group^^ developed a new enantioselective synthesis of Cis phytosphingosine using catalytic asymmetric aldol reactions as a key step (Scheme 23). The key catalytic aldol reaction of acrolein with the ketene silyl acetal 148 derived from phenyl a-benzyloxyacetate was carried out by using tin(II) triflate, chiral diamine 149, and tin(II) oxide. The desired aldol product 150 was obtained in high diastereo- and enantio-selectivities [aywanti = >98:

3. H2NNH2

OBn

63%

OAc

160

1. ACOCH2CO2NSU 2. CH2N2 3. Ac20/py 79%

AcO AcO" ACOCH2COHN

OAc

C02Me

1.H2/Pd-black 2. CI3CCN/DBU

^~

AcO 162

AcO !,^^Si^^^.^:LÔBn AcO OAc

93%

488 AcO AcO«ACOCH2COHN

OAc

COgMe r57/BF3-OEt2

P: AcO

AcO AcO

163

28% AcOA O^NH CCI3

AcO AcO ACOCH2COHN

OAc

C02Me

AcO

0-A AcOAcO

164

0;^C2iH43 N NH AcO

CO2H H0CH2C0HN--^^T'*^ O HO HO HO 156

OBz C14H2

1. NaOMe 2. NaOH

•

85%

OBz

0 1111111111111111111 M11111111111 n 11 T< n f 1111111111111 R T

7.000

7.200

7.400

7.600

7.800

8.000

8.200

8.400

8.600

Fig. 3 GC/MS ion profile of bismethaneboronates of 6-deoxo-24-epicastasterone (33) and 6-deoxo-28-norcastasterone (34) isolated from shoots of Orniihopus sativus Brot.

8.800

9.000

505 100

1273

%

Me-B

50

79

273 ^"-H

484 67

97 205

108 121

50

100

m

213 8288

[145

200

150

469

1313 319 ^^^ \ / 343 367

228

414 / 427

tAfc/lHlMli/"\i^t M v'/«-r . . M l ^ ' . ' k - V /' , , I •.'II 250

300

400

350

450

500

Fig. 4 EI mass spectrum of the bismethaneboronate of 6-deoxo-28-norcastasterone (34)

100

156

%

156

376

Me

/

404 (-H)

••.

o'\

490 ^

195

TMSO""

50 i

• 332 211 Ar-'-H

545 85 75 95 121 H07I139

490 96

177

ilii# m

50

100

150

200

211 I

332

287 316 I

375^04

531

476 470 I 440 I,

60 V'T^'^'^

250

300

350

400

450

500

_L

550 m/z

Fig. 5 EI mass spectrum of the methaneboronate/trimethylsilyether of 2-deoxybrassinolide (35)

m/z

506 156

100 %

TMSO'"'"

50 J

195 95 85il 121 73 I I1071 139

50

II

100

545 476 A90

177

m

150

332

211

375 404

269 287 316 I

l;ll.ltlLji.i^.r,ili.î,.^ .^. yil.

200

250

y-t

300

470 440 ^, ,

531 560

4^

"H f

350

400

450

500

550 m/z

Fig. 6 EI mass spectrum of the methaneboronate/trimethylsilyether of 2-deoxy-24-epibrassinolide (42)

545

TMSO'

213 227

/ 150

200

269 297 250

300

391 350

400

470 450

531 500

550 m/z

Fig. 7 EI mass spectum of the methaneboronate/trimethylsilyether of 2-deoxy-3,24-diepibrassinolide (43)

507

288, 273, 205 and 141 (Figs. 3 and 4). Prominent ions at m/z 288, 273 (base peak) and 205 are characteristic of 6-deoxobrassinosteroids with a 2,3-diol moiety in ring A.^^ j ^ g jon at m/z 141 indicates that 5 exhibits hydroxyls at C-22 and C-23 but no methyl at C-24 in the side chain. Both the molecular ion and other key ions in 34 showed a mass shift of 14 amu compared with 10 and 33. Therefore, compound 34 is proposed to be 6-deoxo-28-norcastasterone. We found another new brassinosteroid 35 in the seeds of Apium graveolens. The mass spectrum of the methaneboronate/trimethylsilyl derivative of this compound shows a molecular ion at m/z 560. Both the M+-ion and the key ions at m/z 545, 531 and 470 (loss of methyl, ethyl and trimethylsilanol, respectively) appeared with a mass shift of 16 amu compared with typhasterol (7) and teasterone (8), indicating an additional oxygen function. The key ions at m/z 404 and 156 characterizing the side chain are complementary ions arising by cleavage of the bond C-20/C-22 (Fig. 5). The ion at m/z 332, also appearing in the EI mass spectrum of the bismethaneboronate of brassinolide (1), represents a key ion for B-homo-6a-oxa lactonetype brassinosteroids with hydroxyls at C-22 and C-23 as well as a methyl at C-24.39 Further important key ions appear at m/z 375 / 376 (cleavage C-17/C-20), 211 (ring B-cleavage), 195 and 177. The GC retention data and the EI mass spectrum of the methaneboronate/silyl derivative of compound 35 were compared with those of synthesized 2-deoxy-24-epibrassinolide (42) and 2-deoxy-3,24-diepibrassinolide (43) (see, Section 3). Compound 35 is eluted earlier than compounds 42 and 43. The difference in the GC retention data of 35 and the 24/?-conrigurated 42 (relative retention times RR^) is typical for other 245/24/?-epimeric brassinosteroids.^^'^3''^5,48 -r^g gj xn2i^^ spectrum of the methaneboronate/trimethylsilyl derivative of 35 is consistent with that of 42, but quite different from 43 (Figs. 5-7). Therefore, the new brassinosteroid can be regarded as 2-deoxybrassinolide (35). A similar methodology was used for the identification of the new member homoteasterone (30) from seeds of Raphanus sativus.^^ The distribution of brassinosteroids in plants investigated since 1991 is summarized in Table 2. Among them six new members along with teasterone myristate (9) from the anthers of Lilium longifoliurrfi, and a new type of brassinosteroid conjugates could be identified. The occurrence of brassinosteroids in species of several plant families hitherto not investigated has been verified. Our investigations also showed that 24epicastasterone (13), firstly found only in the green alga Hydrodictyon reticulaturn}^, is widely distributed in higher plants. It represents the only brassinosteroid in Phoenix dactylifera.^^

3.

SYNTHESIS OF NEW BRASSINOSTEROIDS The original structure of brassinosteroids and the requirement for reference compounds and

sufficient amounts of brassinosteroids for biological studies has tremendously stimulated the synthesis of such phytohormones and their analogues up till now.2.4»60-62 Much efforts have especially been focussed to developing convenient and effective methods for constructing the brassinosteroid side chain with (22/?,237?)-diol function, which is essential for a high bioactivity.^^ Starting from suitable phytosterol precursors with a A^^ double bond the alkyl

508 Table 2 Distribution of brassinosteroids detected since 1991

Plant family

Brassinosteroids

Apium graveolens (seed)

Umbelliferae

2-Deoxybrassinolide (35)

52

Beta vulgaris (seed)

Chenopodiaceae

Castasterone (3) 24-Epicastasterone (13)

53

Cassia torn (seed)

Leguminosae

Brassinolide (1), Castasterone (3), Typhasterol (7), Teasterone (8), Brassinone (16)

54

Daucus carota ssp. sativus (seed)

Umbelliferae

Brassinolide (1), Castasterone (3), 24-Epicastasterone (13)

55

Distylium racemosum (leaves)

Hamamelidaceae 3-Dehydroteasterone (31)

44

Lilium elegans (pollen)

Liliaceae

Brassinolide (1), Castasterone (3), Typhasterol (7), Teasterone (8)

56

Lilium longiflorum (anthers, pollen)

Liliaceae

Brassinolide (1), Castasterone (3), Typhasterol (7), Teasterone-myristate (9) 3-Dehydroteasterone (31)

9,44, 57

Omithopus sativus (seed, shoots)

Leguminosae

Castasterone (3), 6-Deoxocastasterone (10), 24-Epicastasterone (13), 6-Deoxo-24-epicastasterone (33), 6-Deoxo-28-norcastasterone (34)

47,48

Phoenix dactylifera (pollen)

Palmae

24-Epicastasterone (13)

Raphanus sativus (seed)

Cruciferae

Brassinolide (1), Castasterone (3), Teasterone (8), Homoteasterone (30)

Rheum rhabarhanim (panicles)

Polygonaceae

Brassinolide (1), Castasterone (3), 24-Epicastasterone (13)

59

Secale cereale (seed)

Gramineae

Homobrassinolide (2), Castasterone (3), Typhasterol (7), Teasterone (8), 6-Deoxocastasterone (10), Brassinone (16), Secasterone (32)

43

Triticum aestivum (grain)

Gramineae

Castasterone (3), Typhasterol (7), Teasterone (8), 6-Deoxocastasterone (10), 3-Dehydroteasterone (31)

45

Plant species

Reference

(plant part)

51 50,58

509 substituent at C-24 has a significant influence on the stereochemistry of the hydroxylation to 22,23 diols with osmium tetroxide. Thus, the (245)-alkyl substituent directs the hydroxylation prefentially to the undesired (22»S,23iS')-configuration and also starting with (24/?)-alkylated A22precursor more of the corresponding (225,23iS^-diol is formed.^2 Since stereoisomers with the unnatural (22iS',235)-diol function are inactive or less potent growth stimulators, an improved method for obtaining the natural (227?,23/?)-stereoisomers was required. In 1990 an enantioselective method for the osmium-catalyzed asymmetric dihydroxylation of olefins using potassium ferricyanide (K3Fe(CN)6) as the co-oxidant was reported.^ Applying this method with the chiral ligand dihydroquinidine p-chlorobenzoate (DHQD) for the dihydroxylation of the (22£',247?)-24-methyl substituted steroidal side chain a 8:1 ratio of (22R,23R) and (22iS',23»S')-stereoisomers was formed.^^ The reaction was carried out at room temperature in tert. butanol-water 1:1 (v/v) using 1.0-2.0 mmol DHQD, 6.0 mmol K3Fe(CN)6, 6.0 mmol K2CO3, 0.1 mmol Os04 for 1.0 mmol olefin for 4 - 6 days with stirring. Based on this procedure an improved sequence for converting cheap ergosterol (44) to 24epibrassinolide (12), which is the most important brassinosteroid for biological large scale experiments^^, was published (Scheme 4).^'^ Thus, ergosterol mesylate (45) was transformed to isoergosterol (46) which was oxidized with chromium trioxide in pyridine to the corresponding enone 47 . Reduction with lithium in liquid ammonia afforded a cyclopropyl ketone 48 which was isomerized to the A2-6-ketone 49 by heating with pyridinium hydrochloride and lithium bromide in dimethylacetamide at 160° C. Upon hydroxylation of 49, employing the method of asymmetric dihydroxylation with the chiral ligand DHQD, the yield of the desired (22/?,23/?)-stereoisomer 24-epicastasterone (13) was raised to 80 %, besides obtaining 8 % of the (225,235)-epimeric trisepic as taster one (50). The reaction rate was increased substantially in the presence of methane sulfonamide as an additive. Direct conversion of 13 to 24-epibrassinolide (12) was accomplished by Baeyer-Villiger oxidation with trifluoroperoxyacetic acid (CF3CO3H). The overall yield of 24epibrassinolide (12) starting from ergosterol (44) was 26 % in seven steps. To study the biosynthesis and metabolic pathways of brassinosteroid, labeled precursors are required. For such purpose we have developed an effective procedure for labeling 24epibrassinolide (12) with deuterium or tritium (Scheme 5).^^ Thus, reaction of the tetraacetoxy derivative 51 of 24-epicastasterone (13) with 2H2O in dimethylformamide in the presence of triethylamine afforded smoothly the corresponding tris-deuterated compound 52 (R=2H) as the main product. The position of the introduced deuterium is confirmed in the NMR spectrum which lacks the signals for the 5a- and 7a,7p-protons at (6) 2.57 and 2.33, respectively. Baeyer-Villiger oxidation of 52 (R=2H) with CF3CO3H in dichl(6ro)iiethane afforded the 6-oxo-6a-oxa-lactone 53 ( R = 2 H ) besides traces of the isomeric 5a-oxa-6-oxo-lactone, which were separated by Si02chromatography. Treatment of the deuterated 53 with K2CO3 in methanol/water followed by acidification with HCl in tetrahydrofuran yielded the [5,7,7-2H3]24-epibrassinolide (12, R=2H, Scheme 5). This procedure can be used also as a simple method for the introduction of tritium using 3H2O as a labeling reagent. Thus, starting from 51 without isolation of the intermediates the

510

44 R = H 45 R = Ms

OSO4, K2CO3, K3Fe(CN)6 CH3SO2NH2, DHQD t-BuOH/H20

OH

=

24-Epicastasterone (13) (22R, 23R)

Trisepicastasterone (50) (22S, 23S)

H2O2 / (TFA)20 CHCI3

OH

=

H3C0.

24-Epibrassinolide (12)

Scheme 4 Synthesis of 24-epibrassinolide (12) from ergosterol (44)

511 desired [5,7 J-3H3]24-epibrassinolide (12, MBq/mmol was obtained.

R=3H,

Scheme 5) with a specific radioactivity of 222

OAc =

OAc =

AcO. AcO'

AcOi, AcO' 53 R = 2H R = 3H

12 R = 2H R = 3H

o -

HOii,

Oifi./

XoX

HO«''

54 R = H R = 2H R = 3H

13 R = 2H R = 3H

Scheme 5 Synthesis of labeled 24-epibrassinolide (12) In an improved procedure the diisopropylidene derivative 54 (R=H) was used for the tritiation to afford the corresponding 5,7,7-tris-labeIed intermediate 54 (R=3H) which was oxidized under simultaneous deprotection directly to the desired tritiated 24-epibrassinolide (12, R = 3 H ) with a specific radioactivity of 232 MBq/mmol. Therefore, starting from 13 this modification represents a smooth and simple pathway for labeling the biologically important phytohormone 12 in only three steps. On the other hand acid treatment of the diisopropylidene derivative 54 with R = ^H or ^H afforded the corresponding labeled 24-epicastasterone 13 (R = 2H or 3H, Scheme 5).

512 3.1

Synthesis of secasterone and further epimeric 2,3-epoxy brassinosteroids The structural determination of endogenous brassinosteroids, present only in minute

amounts in plant material, requires the availability of corresponding reference standards. Thus, for the final identification of the new brassinosteroid secasterone (32) isolated from Secale cereale (see. Section 2), the four epimeric brassinosteroids with 2,3-epoxy function derived from castasterone (3) and 24-epicastasterone (13), respectively, were synthesized (Scheme 6).^^ For synthesis of both (24/?)-configurated 2,3-epoxides 39 and 40 the 3a,5-cyclo-A22.5. ketone 48 was used as key intermediate. The enantioselective modification of the osmiumcatalyzed dihydroxylation of (22E)-olefm 48 using K3Fe(CN)6 as the co-oxidant and DHQD as the chiral ligand gave 73 % of the desired diol 55 with (22/?,23/?)-configuration. However, direct isomerization of the unprotected diol 55 with pyridinium hydrochloride and lithium bromide in dimethylacetamide led to a ring A saturated 3-chloro derivative. The same reaction starting from the isopropylidenedioxy derivative 56 smoothly afforded the A2-6-keto acetonide 57, which was deprotected with 2 N HCl to give 22,23-diol 58. Epoxidation of 58 with m-chloroperbenzoic acid (MCPBA) afforded, via attack from the less hindered a-side, stereoselectively (22/?,23/?,24/?)22,23-dihydroxy-2a,3a-epoxy-24-methyl-5a-cholestan-6-one (39). For synthesis of the (24^)-configurated 2a,3a-epoxy compound 38 the known*^^ diacetyl ketone 59 was used. Hydrolysis to the (22/?,237?)-diol 60 followed by epoxidation with MCPBA gave 38. To prepare the (24/?)-2P,3p-epoxide 40 the A2-6-keto acetonide 57 was transformed with N-bromosuccinimide (NBS) in dimethoxyethane (DME) to the bromohydrin 62. Acid deprotection to 63 followed by hydrogen bromide elimination with sodium methoxide led to the desired compound 40. Using a procedure similar to the one described for the preparation of the 2p,3p-epoxy compound 40, the known (245)-configurated A^-G-keto acetonide 61^^ was transformed via the bromohydrin 64, deprotection to 65 and HBr elimination, to the (245')-2p,3p-epoxide 32, which was found to be identical with the native secasterone from Secale cereale (see, Section 2). The spectral data of the new compounds are in agreement with the given structures. The observed low field shifts (A 5 + 0.09) of the 19-methyl singlet in comparison to 39 confirm the pconfiguration of the 2,3-epoxy function in compound 40. The same shift was found also for both (24^)-epimers 38 and 32, respectively. 3.2.

Synthesis of 3-dehydroteasterone, 3-dehydro-24-epiteasterone and 6-deoxo-24-epicastasterone 3-Dehydroteasterone (31), the first naturally occurring 3,6-diketo brassinosteroid from

Distylium racemosum and Triticum aestivum, respectively, was synthesized from typhasterol (7)'*5 or teasterone ( 8 ) ^ by oxidation of their con-esponding isopropylidenedioxy derivatives 66 and 67, respectively, with pyridinium chlorochromate and subsequent deprotection (Scheme 7). For the synthesis of the 24-epimer 41, expected also as a native brassinosteroid, the 3,5cycloketone 48 was directly solvolyzed with aqueous H2SO4 to give the 3p-hydroxy-6-ketone

513

48

OR 1.HCI, MeOH

'^H 57 58 59 60 61

(24R), (24R). (24S), (24S). (24S),

62 63 64 65

(24R), (24R), (24S), (24S).

2. MCPBA

R = MesCC R= H R = Ac R=H R = Me2CC[

R = MezCC R= H R = MesCC R=H

OH

°''L

'H 38 (24S) 39 (24R)

32 (24S), secasterone 40 (24R)

Scheme 6 Synthesis of secasterone (32) and further epimeric 2,3-epoxy brassinosteroids

514

68J2 Subsequent Jones oxidation led to the 3-dehydro derivative 69, which afforded, upon asymmetric dihydroxylation, the 3-dehydro-24-epiteasterone (41).69

31

66 3a-OH 67 3P-0H

69

68

41

Scheme 7 Synthesis of 3-dehydroteasterone (31) and 3-dehydro-24-epiteasterone (41) Two new brassinosteroids could be detected from the shoots of Omithopus sativus (see, Section 2). One of them, 6-deoxo-24-epicastasterone (33), was synthesized by us from 24epicastasterone (13) via the corresponding thioacetal 70 and subsequent reductive elimination of the thioketal group by reaction with tri-n-butyltin hydride (Bu3SnH) in the presence of 2,2'azabis-2'-methylpropionitrile (AIBN) (Scheme %).^^ OH ^

13

Scheme 8 Synthesis of 6-deoxo-24-epicastasterone (33)

70

33

515 3.3.

Synthesis of 24-epiteasterone, 24-epityphasterol, 2-deoxy-3,24-diepibrassinolide, and 2-deoxy-24-epibrassinolide To investigate their possible occuirence in plants, we have developed convenient methods

for the synthesis of 24-epiteasterone (71) and 24-epityphasterol (75) as well as their corresponding B-homo lactones 43 and 42, respectively (Scheme 9). For the synthesis of compound 71 the (24/?)-3P-hydroxy-6-ketone

68 was used. Asymmetric catalytic

dihydroxylation of the A^^ double bond of 68 gave the (22/?,23/?)-diol 71 as the main product, besides U'aces of its (225',23iS')-epimer. Baeyer-Villiger oxidation of 71 with CF3CO3H led to a 1 : 0.6 mixture of 2-deoxy-3,24-diepi brassinolide (43) and its 5a-oxa-6-oxo isomer 72, which were separated by preparative HPLC. The corresponding 3a-hydroxy lactone 42 was synthesized from 68 using the Mitsunobu procedure (diethyl azodicarboxylate/triphenylphosphine/formic acid) for inversion of the hydroxy function in position 3J'^ The resulting 3a-formyloxy ester 73, upon hydrolysis afforded the 3aalcohol 74. Asymmetric dihydroxylation of 74 yielded 24-epityphasterol (75) as the main product. Baeyer-Villiger oxidation of 75 led to 2-deoxy-24-epibrassinolide (42) and its isomeric lactone 76 in a 1 : 0.6 ratio.^"^ The spectral data of all new compounds are in agreement with the given structures. Especially the ^H NMR spectra confirm the lactone/isolactone stmctures of 43 and 72 as well as of 42 and 76, respectively. Both isomeric B-homolactone series are easy to differentiate by their characteristic H-5 and H-7 chemical shifts (Fig. 8), whereas in the case of 6-oxo-6a-oxa-lactones the H-5 signal appears as a double doublet at 5 2.86 (43) or 3.18 (42) and that of H-7 (2 H) at 6 4.06 (43) or 4.10 (42), while in the isomeric 5a-oxa-6-oxo-lactones H-5 resonates at 5 4.26 (72) or 4.62 (76) and H-7 at 6 2.48 (72) or 2.49 (76). Also the observed opposite circular dichroism allows a clear differentiation between both the isomeric lactone series (Fig. 9; 6a-oxa-lactones 43 and 42: A£ -h0.201 and +0.265, respectively, at 215 nm; 5a-oxa-lactones 72 and 76: As -0.143 and -0.123, respectively, at 212 nm). 3.4.

Synthesis of A^-T-oxygenated and A^''7-unsaturated brassinosteroids Investigations of the plant extract of celei^ (Apium graveolens) suggest the occun'ence also

of A^-7-oxygenated brassinosteroids.^^ Therefore, we have developed a strategy for the synthesis of such compounds (Scheme 10).''^ Starting from stigmasterol (77) via isostigmasterol (78) the catalytic dihydroxylation of the A^^ double bond led to the (22^,23/?)-diol 79 as the main product, which was isomerized'^^ to the new (22/?,23/?)-22,23-dihydroxystigmasterol (80). Subsequent acetylation of the three hydroxyl groups and allylic oxidation with chromic acid in dichloromethane led to the enone derivative 81, which was hydrolyzed to the enone triol, 82. Reduction of 82 to the 7p-hydroxylated compound 84 was achieved with sodium borohydride in the presence of cerium trichloride in tetrahydrofuran/methanol. On the other hand, the reduction of 82 with L-Selectride in tetrahydrofuran at -78° C^'^ gave the corresponding 7a-hydroxylated allylic alcohol, 83.

516

O

O

68

71 CF3CO3H CHCI3

N—COsEt

II

N—C02Et, PhaP, HCO2H, benzene

OH

=

HO O

75

73 R = HCO 74 R = H

CF3CO3H CHCI3

OH

=

HO''

76

Scheme 9 Synthesis of 24-epiteasterone (71), 24-epityphasterol (75), 2-deoxy-3,24-diepibrassinolide (43), and 2-deoxy-24-epibrassinolide (42)

517

M|]tMjii»Mii»»|iiii|Uiijiiii)iiii|iiii|nnjini|Mii|ini[nii|MM|iiiijMn|Mi»jiiu|iiiijniiiiii>}iiiiiiiii|»Mi|n

4.6

4.2,

3.8

3.4

3.0

2.6

ppm

zll_tizt

M|r»>jNn||i>ii|iiii|iinjHii|iiiijiniiiiii|iiii|iiM}iiii|ii

4.6

4.2

3.8

3.4

3.0

2.6

ppm

I L-j^^->j** t»[iiiijiiii|ini}!ii>inii}iiiipMi|ini|iiiijiiii|tiii{Mii]iiMjini|wiijiiii|iiu[uri|iiii{iiii|iiiijii[i|iwijini|n

4.6

4.2

3.8

3.4

3.0

2.6

ppm

HO^

^^ £ O

76 i»inimHi|i»iijnii|iiii|iiMiiiMpir>|iiiijiiMpiM|iiii|iiM{iin|iiii[ini|iiii|iMi|Mii|iin|iiii|Mii|iiii|Uii|ii

4.6

4.2

3.8

3.4

3.0

2.6

ppm

Fig. 8 Low-field region of ^H NMR spectra of the isomeric lactones 76, 72, 42 and 43 (500 MHz, solvent: CDCI3)

518 3.000E+01

I

I

» '

I

'

I

'

•

'

'

I

'

'

'

'

I

I

111 I 1 11 I 1 1 1 I I I I I 11 ' '

'J

OH

CD [mdegj

-3.000E+01 3.000E+01

CD [mdeg]

I I I I I I I I I I I I

300.0

Fig. 9 Circular dichroism spectra of the isomeric lactones 43 and 72 as well as 42 and 76 (in trifluoroethanol)

519

HC^

^NNHS02^^^^_V-CH: 85

86

SchemelO Synthesis of A5-7-oxygenated and A5,7-unsaturated brassinosteroid analogues 82, 83, 84 and 86

520 For the synthesis of the corresponding A^.^-unsaturated brassinosteroid analogue 86 the A^-7-keto derivative 82 was reacted with toluene-4-sulfonohydrazine in dry tetrahydrofuran under anaerobic conditions at 75° C to give the corresponding tosylhydrazone, 85. Reductive elimination of compound 85 with lithium hydride in toluene at 100° C^^ yielded (22/?,23;?,245)-22,23dihydroxy-28-homoergosterol (86), the structure of which was confirmed by spectral data7^

4.

METABOLISM OF BRASSINOSTEROIDS

4.1.

Current status of brassinosteroid biosynthesis 79

In 1991, Yokota et al., published a hypothetical pathway of brassinosteroid biosynthesis.

The authors suggested phytosterols, e. g. campesterol, as biogenetic precursors of brassinosteroids. Teasterone (8) and typhasterol (7) may be regarded as the first compounds of the biosynthetic sequence bearing some of the major characteristics of brassinosteroids: trans-fustd A/B-ring system, vicinal hydroxyl groups in the side chain at C-22 and C-23, and oxygenation at C-6. Following this hypothesis, which is confirmed also by the occurrence of 6deoxobrassinosteroids, e.g. compounds 10, 33, and 34 in Ornithopus sativus (see. Section 2), bis-hydroxylation of the side chain should occur prior to 2a-hydroxylation. The final step of this hypothetical pathway is the Baeyer-Villiger type oxidation of castasterone (3) to yield brassinolide (1). This sequence was based on common occurrence and mechanistic considerations rather than on experimental results. The bioactivity, measured by means of the rice lamina inclination bioassay, increased with each biosynthetic step, and was used as a supporting argument for the suggested pathway. In the mean time several of the proposed biosynthetic steps have been supported by experimental results. Feeding experiments using [26,28-^H]labeled precursors and GC-MS analysis have established the biosynthetic sequence teasterone (8) —> typhasterol (7) —> 80 castasterone (3) —> brassinolide (1) in Catharanthus roseus. Teasterone (8) was demonstrated to serve as a biosynthetic precursor of typhasterol (7) in crown-gall and non-transformed cells of Catharanthus roseus in which both compounds are endogenous. This conversion probably 81

proceeds via 3-dehydroteasterone (31) as an intermediate naturally occurring brassinosteroid in Triticum aestivum

which has recently been identified as a and Distylium racemosum.

This result

is in analogy with ecdysteroids where epimerization has been demonstrated to occur through a 3dehydro type compound.

After feeding of [26,28-2H]3-dehydroteasterone (31), labeled

typhasterol (7) was detected as a major product and labeled teasterone (8) as a minor one. This reversibility of inversion suggested an 3-epimerase system, similar as shown for ecdysteroids.^^ When [26,28-2H]typhasterol (7) was administered to cultured cells of Catharanthus roseus, GCMS revealed castasterone (3) in extracts obtained after 24 h and 48 h, respectively. Because of the co-occurrence of teasterone (8), typhasterol (7), castasterone (3), and brassinolide (1) in several plant species, the proposed biosynthetic sequence might be ubiquitous in higher plants. The transformation of castasterone (3) to brassinolide (1) via a Baeyer-Villiger type reaction previously

521 has been demonstrated for crown gall cells of Catharanthus roseus using ^H labeled castasterone 84

2

(3). This was later confirmed by H-labeling experiments in cell suspension cultures of the same 85 species. The complete pathway (Scheme 11) between teasterone (8) and brassinolide (1) was 80

established by the same authors after feeding of deuterated precursors.

Phytosterol

Teasterone (8)

Typhasterol (7)

Castasterone (3)

3,6-Dehydroteasterone (31)

Brassinolide (1)

Scheme 11 Biosynthesis of brassinosteroids in Catharanthus roseus However, there are a lot of open questions in the biosynthesis of brassinosteroids. An interesting aspect which remains to be studied is the mechanism of the Baeyer-Villiger reaction in a biological system. Furthermore, the early steps of the brassinosteroid biosynthesis between phytosterols and teasterone (8) are an open field because there exist no labeling studies. It would be quite interesting to clarify if there are several parallel pathways between the phytosterols and brassinosteroids or if there is only one route, for instance the pathway outlined in Scheme 11. The co-occurrence of phytosterols and brassinosteroids with corresponding side chains implicate the former variant. However, it was also argued that major sterols do not account for the transformation to brassinosteroids in a proportional ratio, indicating rather selective transformation of 24-methyl and 24-methylene sterols. Investigations on the enzymatic level are not yet known. 4.2.

Brassinosteroid conjugates In contrast to the biosynthesis, aspects of interconversion and metabolism of brassino-

steroids have been poorly investigated until now. As assumed for classical phytohormones, also in the case of brassinosteroids, different types of conjugates may be involved in the biosynthesis,

522 transport, compartmentation and storage processes. However, only a few brassinosteroid conjugates are hitherto described. Whereas both the glycosides 23-0-p-D-glucopyranosyl-25methyldolicosterone (36) and 23-0-p-D-glucopyranosyl-2-epi-25-methyldolicosterone (37) occur native in Phaseolus viilgaris^^ 23-0-P-D-glucopyranosyl-brassinolide (87) was identified as a 79 87

metabolite of brassinolide in Vigna radiata. '

OH OH

23-0-p-D-Glucopyranosyl-brassJnollde(87)

Remarkably, all these glycosides bear the sugar moiety at the 23-OH. Until now no 3-0glucoside of any brassinosteroid has been found, although this is a commonly glucosylated position in other steroidal compounds.

88

Based on reports on fractions with significant activity in

89

87 90

the rice lamina inclination test and corresponding fractions in metabolic studies * a wide spread occurrence of hitherto unknown glucosides and hydrophilic non-glucosidic brassinosteroid conjugates is likely. Before starting our investigations, 23-0-p-D-glucopyranosyl-brassinolide (87) was the only known metabolite of plant origin derived from an exogenously applied 91

brassinosteroid. Acyl type conjugates were very frequently described for various phytosterols.

The first fatty acid conjugate, teasterone-3-myristate (9), was isolated recently from the anthers of Lilium longiflorwn. Acyl glucosyl conjugates, another common type of phytosterol conjugates in plants,

88

are expected to exist for brassinosteroids too. The lack of knowledge on the interconversion and the fate of brassinosteroids in plant

systems prompted us to study the metabolism of selected compounds of this type in detail. At the beginning of the metabolic studies on brassinosteroids we had to choose the compounds to be investigated as well as suitable plant systems. 4.3.

Prerequisites for metabolic studies 24-Epicastasterone (13) and 24-epibrassinolide (12) are naturally occurring brassino-

steroids with significant bioactivity in the rice lamina inclination test as well as in other bioassays. Both compounds are readily available

and a procedure for radioactive labeling with tritium was

established^^ (see, Section 3). [5,7,7-3H]24-Epicastasterone (13 in Scheme 5: R = 3 H ) and [5,7,7^H]24-epibrassinolide (12 in Scheme 5: R=^H) were used in our experiments. Following the biosynthetic sequence shown in Scheme 11 and experimental results, '

the 6-keto- and oxa-

lactone series are closely related biosynthetically. However, in Oryza saliva, castasterone (3), as an example for the 6-kelo compounds, may not serve as a precursor for brassinolide (1)^^ but may either be physiologically active per 5e or act as a precursor for unknown active brassinosteroids.

523 Cell suspension cultures possess several advantages over whole plants as objects in metabolic studies.

92

Thus, hitherto known experiments on the biosynthesis of brassinosteroids

were mainly performed using cell cultures of various plant species. In our experiments cell suspension cultures of Lycopersicon esculentum and Ornithopus sativus were used. Both cell lines are fast growing, well characterized and easy to handle. Preceeding experiments indicated that the cell growth of both species were not significantly influenced by concentrations up to 10'^ M of exogenously applied compounds 12 and 13. 4.4.

Metabolism of 24-epibrassinolide in cell cultures oi Lycopersicon esculentum The use of radiolabeled precursors allowed the measurement of the distribution of both the

parent compound and the metabolites in cell suspension cultures. In tomato cell cultures [5,7,7^H]24-epibrassinolide (12) was rapidly taken up by the suspended cells. As shown by TLC of ceU extracts obtained at day 4 after administration of the labeled compound, 24-epibrassinolide (12) has been converted to several hydrophilic metabolites. The structures of the major metabolites, clearly separated by TLC (see, insertion c and d in Fig. 10) were determined by MS and NMR analysis as 25-P-D-glucopyranosyloxy-24-epibrassinolide (88) and 26-P-D-glucopyranosyloxy93 94

24-epibrassinolide (89). '

_L

The FAB-MS spectra of both compounds (m/z 659 [M-t-H]"*", m/z

681 [M-i-Na]"^) exhibited nearly identical fragmentation patterns and relative intensities with negligible differences between the corresponding peaks (Fig. 10 a, b). The fragmentation patterns (e. g. m/z 409) indicated an additional fifth hydroxyl group located in the terminal part of the side chain beyond C-23. The positions of the new hydroxyl groups and of the glucosyl moieties at these newly functionalized carbon atoms in both compounds were unambiguously established by detailed NMR investigations (see, Section 5). Thus, the metabolic conversion of 24epibrassinolide (12) to 25-p-D-glucopyranosyloxy-24-epibrassinolide (88) and 26-p-Dglucopyranosyloxy-24-epibrassinolide (89), respectively, is a two-step process including hydroxylation to 25-hydroxy-24-epibrassinolide (90) and 26-hydroxy-24-epibrassinolide (91), respectively, and subsequent glucosylation (Scheme 12). These compounds represent the first brassinosteroids of plant origin with a hydroxyl group at the C-25 or C-26 position. Furthermore, the glucosides 88 and 89 are the first brassinosteroid glucosides are not to have the glucose moiety at C-23. The compounds 90 and 91 were not found in a non-glucosidated state in the cultured cells but were obtained by acid or enzymatic hydrolysis of the glucosides 88 and 89, respectively. The results of the rice lamina inclination test (RLIT) indicated an extraordinary high activity of 25-hydroxy-24-epibrassinolide (90). This compound is about ten times more active than 24epibrassinolide (12), indicating that the hydroxylation at C-25 is an activating step in the brassinosteroid metabolism. Therefore, 25-hydroxy-24-epibrassinolide (90) is, next to brassinolide (1), one of the most active brassinosteroids known until now. In comparison with 25hydroxy-24-epibrassinolide (90), the 26-hydroxylated metabolite (91) was clearly less active. As in other groups of steroidal hoimones, for instance vitamine D metabolites, hydroxylation at C-25 seems to be essential for high activity.

524

100

409

% 90

•Qc

60 50

479

497

lit.lll^l[^ll , t l l , t . t l , . , . . . ^ ! kf.hlfk Âf

500

100

.11,11 I l ^ l i l j t l l

700

m/2

l4lH|>lltl^l.l^.tl^L î.ijil.^1 lvL,i^.t.jL*^|L,Jltiii^4l', "J 500 700 600

m/z

300

600

409

%

90

60

479

40

497

20

461

349 379

10

300

400

yMiw

Fig. 10 FAB mass spectra of (a) 25-P-D-glucopyranosyloxy-24-epibrassinolide (88) and (b) 26-P-D-gIuco-pyranosyIoxy-24-epibrassinolide (89). Insertions (c) and (d): Radio-TLC profiles of extracts of Lycopersicon esculentum cultured cells

525 12

/

25-Hydroxy-24-epibrassinollde (90)

OH

\

26-Hydroxy-24-eplbrasslnollde (91)

OH

rr

25-p-D-Glucopyranosyloxy-24-eplbrasslnolJde (88)

=

26-p-D-Glucopyranosyloxy-24-epJbrassinolide (89)

Scheme 12 Metabolism of 24-epibrassinolide in cell suspension cultures of Lycopersicon esculentum In comparison with the aglycones, the glucosides 88 and 89 exhibited less but also significant activity in the RLIT which may be due to hydrolysis within the test system. These findings suggest that 25-hydroxy-24-epibrassinolide (90) and its 25-0-glucoside (88) are not detoxification products of exogenously applied 24-epibrassinolide (12) but could be regarded as final members of the biosynthetic chain of brassinosteroids. The hydroxylation at C-25 and C-26 of 24-epibrassinolide (12) found in tomato cell cultures provided the opportunity to study these reactions in more detail. Hydroxylations in general are expected to be catalyzed by cytochrome P-450 dependent monooxygenases which are commonly characterized by their sensitivity to carbon monoxide and specific inhibitors. Thus, treatment of tomato cell cultures with various monooxygenase inhibitors simultanously with administration of 24-epibrassinolide (12) was supposed to influence the pathway of hydroxylation. From this approach infomiations on the specificity and the character of the enzymes involved were expected. The ratio of 89 : 88 within extracts of inhibitor non-treated cells was about 1 : 1. Tetcyclasis changed this ratio in favour of

26-p-D-glucopyranosyloxy-24-

epibrassinolide (89). This finding is quite opposit to the effect of cytochrome c which inhibited the formation of 89. Consequently, the concentration of 25-p-D-glucopyranosyloxy-24-

526

OH

=

1.0

Rf

Fig. 11 Radio TLC profiles of brassinosteroid glycosylation in cell cultures of Lycopersicum esculentum: (a) regiospecific glucosylation of 25-hydroxy-24-epibrassinolide (90), (b) non-specific glycosylation of 26-hydroxy-24-epibrassinolide (91)

527

epibrassinolide (88) was significantly increased. The different yields of 88 and 89, respectively, after ti*eatment with these inhibitors suggested that the hydroxylation of 24-epibrassinolide (12) at C-25 and C-26, respectively, is catalyzed by two regiospecific enzymes of different types.

This

was confirmed by CO poisoning, which is a principal criterium for a cytochrome P-450 96 involvement in an oxidation reaction. Following exposure to carbon monoxide (CO : O2 9:1), hydroxylation at C-26 was drastically reduced. The ratio of 89 : 88 was found to be 8 : 92 in this experiment. The CO poisoning effect was partially reversible by light. In contrast to the C-26 hydroxylation sensitive to carbon monoxide, the C-25 hydroxylation, non-typically for a cytochrome P-450 enzyme, was completely resistant. To examine the regiospecificity of the glucosyltransferases involved in the metabolism of 24-epibrassinolide (12) in cell cultures of Lycopersicon esculentum, tritium labeled compounds 90 and 91 obtained by enzymatic hydrolysis of biosynthetically prepared 88 and 89, respectively, were applied. After 4 days of incubation, radio TLC and reversed phase HPLC indicated that only 25-p-D-glucopyranosyloxy-24-epibrassinolide (88) was formed from 90 (Fig. 11 a). FAB-MS of 25-P-D-glucopyranosyloxy-24-epibrassinolide (88), obtained from re-application experiments confirmed the position of the glucose moiety at the terminal part of the side chain. These results strongly suggested that 25-hydroxy-24-epibrassinolide (90) did not undergo conversion except to 25-p-D-glucopyranosyloxy-24-epibrassinolide (88) nor was it subjected to any remarkable catabolism. In contrast, H labeled 26-hydroxy-24-epibrassinolide (91) under the same conditions afforded at least four glucosyl conjugates (Fig. 11 b). Among them the major compound was identical with 26-P-D-glucopyranosyloxy-24-epibrassinolide (89). Thus, it was concluded that 91 may be glucosylated at various hydroxyl groups. Comparing the behaviour of 90 and 91 in reapplication experiments it has to be stated that the glucosylation of the 25-hydroxy compound (90) proceeds in a highly regiospecific manner at this position while the glucosylation of the 26hydroxy compound (91) was less regiospecific. This implied that the aglycone (90) and its 25-0glucoside (88) could be involved in the mode of action of 24-epibrassinolide (12) in plants. 4.5.

Metabolism of 24-epicastasterone in cell cultures of Lycopersicon esculentum 24-Epicastasterone (13) is assumed to be the immediate biogenetic precursor of 24-

epibrassinolide (12), analogously with the related couple castasterone (3) and brassinolide (1) in the (245")-series.

In further experiments H labeled 24-epicastasterone (13) was fed to cell

suspension cultures of Lycopersicon esculentum and the extract obtained four days after administration was shared into two halfs. One part was subjected to enzymatic hydrolysis (aglycone fraction) while the other one was further used in the glucosidic state. Four brassinosteroid glucosides were identified as major metabolites within the nonhydrolyzed fraction. As described for the formation of 88 and 89 after administration of 12, hydroxylation and glucosylation also occurred at C-25 and C-26 of the side chain of 13 yielding 25-P-D-glucopyranosyloxy-24-epicastasterone (92) and 26-P-D-glucopyranosyloxy-24-epicastasterone (93) (Scheme 13). The structures of both glucosides were established by FAB-MS (92: m/z 665 [M]"^) and (93: m/z 665 [M]"*") and NMR. The fragmentation patterns of both compounds

528 OH

OH

=

=

OH

26-p-D-Glucopyranosyloxy24-epicastasterone (93)

25-p-D-Glucopyranosyloxy24-epicastasterone (92)

t

OH = OH

\ - \

OH

^Ky HO""^- ' ' ' ' ^ N /

13 I—

I-

25-Hydroxy-24-epicastasterone

26-Hydroxy-24-epicastasterone

-J

(95)

(94) OH

=r

3-Dehyclro-24-eplcastasterone (96)

i

14

25-Hydroxy-3,24-diepicastasterone (97) OH

=

HO—V

OH 2-O-p-D-Glucopyranosyl3,24-diepicastasterone (98)

HO-

HOHO.

.S^:^' 3-0-p-D-Glucopyranosyl3,24-diepicastasterone (99)

Scheme 13 Metabolism of 24-epicastasterone in cell suspension cultures of Lycopersicon esculentum

529 (m/z 393, bond fission between C-23 and C-24) confirmed the position of the glucosyloxy moieties at the terminal part of the side chain beyond C-23. Obviously, the metabolic pathway which was established for the 24-epibrassinolide (12) is also operating for 24-epicastasterone (13), namely the regiospecific hydroxylation of 13 and glucosylation of the newly formed hydroxyl groups in the intermediates, 25-hydroxy-24-epicastasterone (94) and 26-hydroxy-24epicastasterone (95). Furthermore, 2-0-p-D-glucopyranosyl-3,24-diepicastasterone (98) and 3-03-D-glucopyranosyl-3,24-diepicastasterone (99) were detected and the structures were elucidated by NMR analysis of the non-separated mixture of both compounds (see. Section 5). These compounds represent the first brassinosteroid glucosides bearing the sugar moiety at the ring A hydroxyl group. The 3p-conriguration in 99 suggested epimerization at C-3 prior to glucosylation. This was confirmed by the isolation of related aglycones from the hydrolyzed parts of the extract. These metabolites belong to a metabolic sequence starting with the oxidation of the 3a-0H. The first compound of this sequence is 3-dehydro-24-epicastasterone (96), a new 3,6-diketobrassinosteroid. It has to be regarded as an intermediate in the epimerization to 3,24-diepicastasterone (14), 2 which is known as a naturally occurring compound. Compound 14 is a branching point in this metabolic sequence. It is either glucosylated at 3P-0H or at 2a-0H yielding 99 and 98, respectively, or it can be hydroxylated at C-25 to give 25-hydroxy-3,24-diepicastasterone (97). For MS fragmentation, see Scheme 16. Alternatively, compound 97 may be derived also from the intermediate pentahydroxylated 25-hydroxy-24-epicastasterone (94). The latter compound 94 and 26-hydroxy-24-epicastasterone (95) were not detectable in a non-glucosylated state, probably due to very small endogenous pool sizes. Analogously, the intermediate pentahydroxylated metabolites 90 and 91 were also not detectable in the cell culture medium. Scheme 13 shows the metabolic pathways of 24-epicastasterone (13) in cell suspension cultures of Lycopersicon esculentum so far currently known.

97

All these compounds, both glucosides and aglycones, were exclusively

isolated from the suspended cells of Lycopersicon esculentiim. The medium of this cell culture did not contain significant amounts of any brassinosteroid metabolite, 4.6.

Metabolism of 24-epicastasterone and 24-epibrassinolide in cell cultures of Ornithopus sativus A quite different distribution of metabolites was found in cell suspension cultures of

Ornithopus sativus. Surprisingly, from 1 h after the beginning of the experiments over the remaining incubation time of several days, the distribution of the radioactivity between the cells and the medium did not change significantly. About 40% of the radioactivity after feeding of 24epibrassinolide (12), and 25% after feeding of 24-epicastasterone (13), were found in the medium and the remainder was present within the cells. There was also a clear compartmentation of the different types of metabolites between the cells and the culture medium. While in the medium the non-conjugated metabolites were almost solely found, the cells did contain both hydrophilic and lipophilic conjugates. From the medium brassinosteroid-derived pregnan-like compounds were 98 isolated and their structures were elucidated by MS and NMR analysis. Following apphcation of

530 24-epibrassinolide (12), 2a,3P-dihydroxy-B-homo-6a-oxa-5a-pregnane-6,20-dione (102, m/z 364 [M]"^), and after exogenous application of 24-epicastasterone (13), 2a,3P-dihydroxy-5apregnane-6,20-dione (108, m/z 348 [M]"*") and 2a,3P.6P-trihydroxy-5a-pregnane-20-one (109, m/z 350 [M]"^) were found. '

These compounds are the first side chain degradation products of

brassinosteroid origin described in plant material. Compounds 102 and 109 seem to be the final products of catabolism of 12 and 13, respectively, in Ornithopus sativus cell cultures and two completely elucidated metabolic sequences revealing these compounds were established (Scheme 14 and 15).^^^ 12 OH

3,24-Dlepibrasslnolide (100)

=

25-Hydroxy-3,24-diepibrassinollde (103)

OH

(20R)-Hydroxy-3,24-diepibrassinoflde(101)

3,24-Diepibrassinollde-

3p-laurate (104)

o

3p-myrlstate (105) 2a,3p-Dlhydroxy-B-homo6a-oxa-5a-pregnane-6,20-dione (102)

3p-palmitate (106)

Scheme 14 Metabolism of 24-epibrassinolide in cell suspension cultures of Ornithopus sativus Starting from 24-epibrassinolide (12) and 24-epicastasterone (13), in both cases the metabolism involves epimerization of the 3a-hydroxyl group to the equatorial 3p-0H, leading to compounds 3,24-diepibrassinoIide (100) and 3,24-diepicastasterone (14), respectively. The mass

531 spectra resemble those of the parent compounds 12 and 13, respectively. However, the H,^H coupling constants in the H NMR spectra indicated axial position of H-3 (see, Section 5). In the next step, hydroxylation at C-20 takes place. The mass spectra of the resulting (20/?)-hydroxy3,24-diepibrassinolide (101, m/z 497 [M]"^) and (20/?)-hydroxy-3,24-diepicastasterone (107, m/z 481 [M] ), respectively, are characterized by molecular ions of low intensities (about 1%) and very strong fragment ions of m/z 365 [M]"^ (79) for 101, and m/z 349 [M]"^ (100) for 107, indicative for bond fission between C-20 and C-22 (fragment b in Scheme 16, R2 = OH, R3 = R4 = H). Obviously, this bond between C-20 and C-22 is destabilized by C-20 hydroxylation and

13

14 3,24-Dlepicastasterone-

3(3-)aurate (110)

0

(20R)-Hydroxy-3,24-dlepicastasterone (107) ap-palmitate (112)

2a,3n-Dihydroxy5a-pregnane-6,20-dJone (108)

2a,3p,6{^-Trihydroxy5a-pregnane-20-one (109)

Scheme 15 Metabolism of 24-epicastasterone in cell suspension cultures of Omithopus sativus it is hence accessible to enzymatic attack. This assumption was confirmed by the very small concentrations of metabolites 101 and 107, respectively, suggesting rapid side chain cleavage between C-22 and C-20. The 20,22,23-trihydroxy structural feature, even more then the general structure of brassinosteroids, resembles the ecdysteroids which frequently bear a 20-hydroxyl

532 group.101,102 Within this metabolic sequence of 24-epibrassinolide (12) (Scheme 14), the pregnane-like compound 102 is the final product. The seven-membered lactone ring structure obviously prevents further conversions, which in the metabolic sequence of 24-epicastasterone (13) (Scheme 15) via reduction of the 6-keto group led to 2a,3p,6P-trir ydroxy-5a-pregnane-20one (109),^^ The structure elucidation of another minor metabolite of 24-epibrassinolide (12) from the cell culture medium of Omithopus

sativus revealed the presence of 25-hydroxy-3,24-

diepibrassinolide (103, m/z 497 [M+H]*). In contrast to 25- and 26-hydroxy-24-epibrassinolides (90 and 91) which were found in the glucosidic state in Lycopersicon esculentuniy this compound occurred as an aglycone. In the mass spectra, the fragments m/z 379 of 103 (fragment a in Scheme 16, R p -COO, R2 = R4 = H, R3 = OH) and m/z 365 of 101 (fragment b in Scheme 16, Rj = COO, R2 = OH, R3 = R4 = H),

C4•,'i'i...!

OH HQ,,,

R^-j-O^

J3

^ R 1 ^

R

! . - • ( ]1

R2

1R1

^co ^co ^co ^co ^co -co-co-co-co-co-

R3

R4

H

OH

H

25-Hydroxy-3,24-diepicastasterone

(97)

OH

H

H

(20R)-Hydroxy-3,24-clieplcastasterone

(107)

H

H

lauryl

3,24-Dieplcastasterone-3p-laurate

(110)

H

H

myristyl

3,24-Dieplcastasterone-3n-myristate

(111)

H

H

palmltyl

3,24-Diepicastasterone-3n -palmltate

(112)

-0- •

OH

H

H

(20R)-Hydroxy-3,24-dlepibrassinollde

(101)

-0-

H

OH

H

25-Hydroxy-3,24-diepibrassinolJde

(103)

H

H

lauryl

3,24-Dleplbrasslnollde-3p -laurate

(104)

-0-0-0-

•

H

H

myristyl

3,24-DleplbrassmolJde-3p -myristate

(105)

H

H

palmltyl

3,24-Dleplbrassmolide-3(3 -palmltate

(106)

Scheme 16 EI-MS fragmentation of brassinosteroid metabolites of the 3,24-diepi series respectively, represent diagnostic ions of pentahydroxylated brassinosteroids derived from 12 (for 25-hydroxylation in the first case and for 20-hydroxylation in the second one). As found for 3,24-diepicastasterone (14) in L esculentum

cell cultures, 3,24-

diepibrassinolide (100) is a branching point in the metabolism of 24-epibrassinolide (12) in

533 Ornithopus sativus. Besic'e 25-hydroxylation as a minor metabolic reaction in 0. sativus, the fatty acid esters were mainly formed from 3,24-diepibrassinolide (100). However, these lipohilic 99

metabolites were not present in the medium but only inside the cells.

After purification of the

lipophilic fraction of the cell extract by TLC and separation by HPLC (RP-8), the structures of these metabolites were elucidated by spectroscopic methods. Three fatty acyl esters were derived from both 100 and 14, all in nearly the same quantity. 3,24-Diepibrassinolide-3P-laurate (104, m/z 662), -3p-myristate (105, m/z 690), -3P-palmitate (106, m/z 718) were found as metabolites of 24-epibrassinolide (12), and 3,24-diepicastasterone-3P-laurate (110, m/z 646), -3P-myristate (111, m/z 675), -3P-palmitate (112, m/z 702) were metabolites of 24-epicastasterone (13). The position of the fatty acid residue at ring A can be deduced from the fragment ions which appear after fission a, b, or c (Scheme 16). The base peaks of the acyl-conjugated 3,24diepibrassinolides (104 - 106) appear at m/z 361 (a-RCOOH) and of the acyl-conjugated 3,24diepicastasterones (110 - 111) at m/z 346 (a+H-RCOOH), respectively. The NMR spectra of the fatty acyl conjugates are very similar to each other. In comparison to the spectrum of the nonconjugated compounds 100 and 14, respectively, H-3p exhibits a downfield shift of about 1.2 ppm due to an ester bond at this position. The H, H coupling constants establish that epimerization has occurred at C-3. In addition to the signals of the genines, the H NMR spectra exhibit signals of the fatty acid methylene protons (5 1.25) and the terminal methyl groups (5 0.88). Our results represent the first report of fatty acid conjugates as metabolites of exogenously applied brassinosteroids. The function of these fatty acyl ester derivatives of brassinosteroids still remains unknown. However, they may be compartmentalised within membrane structures as generally described for phytosterol acyl esters. 103 Several minor hydrophilic compounds were detected in cultured cells of O. sativus which may be glucosides, but the major part of radioactivity was associated with the acyl ester fraction. Summarizing die results of our studies on die metabolism, it can be stated that there are two principle pathways of brassinosteroid conversion in plants: First, hydroxylation in the terminal part of the side chain followed by glucosylation of the newly formed hydroxyl group. This pathway, at least hydroxylation at C-25 of 24-epibrassinolide (12), significantly increases the bioactivity of the 25-hydroxylated compound 88 compared with the parent substance and therefore can be regarded as an activation reaction. Second, catabolic side chain removal was found. This pathway starts with epimerization at C-3, followed by hydroxylation at C-20 and bond fission between the vicinal hydroxyl groups at C-20 and C-22. Conjugation at C-2 and C-3, respectively, with glucose or at C-3 with fatty acids seems to require equatorial position of the corresponding hydroxyl group which is a result of the preceeding epimerization. Surprisingly, the expected biosynthetic Baeyer-Villiger oxidation of 24-epicastasterone (13) to yield 24-epibrassinolide (12) was not observed either in cell cultures of Lycopersicon esculentum or in Ornithopus sativus in our experiments. Altogether, 26 metabolites of exogenously applied brassinosteroids, (except compound 14) described for the first time, have been found in our studies until now.

534 5.

NMR SPECTROSCOPY OF BRASSINOSTEROE) METABOLITES NMR spectroscopy is a powerful tool for structural elucidation of brassinolid metabolites.

Based on complete and unambiguous assignments of proton and carbon NMR signals of the main brassinosteroids^^"*'^^^ the structures of metabolites can be determined on the basis of changes in chemical shifts and coupling constants as well as by correlations found in two-dimensional NMR experiments. Modem NMR spectrometers with cryomagnets and special designed probeheads have made it possible to record proton detected one- and two-dimensional NMR spectra of very small amounts of natural compounds. Nowadays, direct ^^C measurements can be done with amounts of brassinosteroid metabolites down to about 1 -2 fimol, whereas proton detected spectra can be recorded with amounts down to about 0.1 - 0.2 jimol of metabolites. So-called inverse-detected ^H-^^C chemical shift correlation spectra allow the assignments of carbon signals even in cases in which no direct ^^C NMR spectra can be recorded because of poor signal-to-noise ratios. Since in these inverse heteronuclear shift correlation experiments magnetization instead of ^-^C magnetization is detected, the sensitivity is significantly better than in conventional 2D experiments using ^^C detection. Important hints regarding structural changes during metabolic processes can often be achieved by inspection of the relevant parts of the proton NMR spectrum from the metabolite in comparison with the parent brassinosteroid. Metabolic hydroxylation of the methine carbons in the brassinosteroid side chain results in a change of the coupling patterns and in a low-field shift of the adjacent side chain methyl group proton signals (Fig. 12). For example, the ^H NMR spectrum of 24-epibrassinolide (12) shows four side chain methyl group doublets at 5 0.97 (Me-21), 0.91 and 0.86 (Me-26 and Me-27) and 0.83 (Me-28), whereas the proton signals of Me-26 and Me-27 in 25-P-D-glucopyranosyloxy-24-epibrassinoHde (88) appear as singlets at 5 1.36 and 1.30. Hydroxylation of a side chain methyl group gives rise to disappearance of the corresponding methyl doublet in the high-field region and to the appearance of two new proton multiplets in the low-field region. The absence of several methyl group ^H NMR signals indicates degradation of the side chain. Thus, the ^H NMR spectrum of 2a,3P-dihydroxy-5a-pregnane-6,20-dione (108) exhibits only two methyl singlets in the high-field region (5 0.81 and 0.62; Me-19 and Me18, respectively). The methyl singlet at 5 2.13 of 108 is assigned to Me-21 in a 20-ketopregnane side chain moiety. An unchanged side chain is proved by the occurrence of four methyl doublets with the same chemical shifts as found for the feeded brassinosteroid. Even if the methyl region of the proton NMR spectra is superimposed by signals of impurities or by incompletely separated minor metabolites, which is not unusual if the amounts of metabolites are very small, it is possible to recognize an unchanged side chain by its fingerprint region in the ^H-^H chemical shift correlated 2D NMR (COSY) spectrum (Fig. 13). The H-^H COSY spectra correlate proton chemical shifts through homonuclear coupHngs. Starting from a separated, readily assigned signal the appearance of cross peaks allow identification and assignment of the complete spin system. For instance, H-5a shows correlation peaks with H-4a and H-4p, which correlate with H-3p (or H-3a), the latter showing further cor-

535

ppm

Fig. 12 ^H NMR high-field region of 24-epicastasterone (13)^ and 25-P-D-glucopyranosyloxy-24epicastasterone (92)^ (500 MHz, solvent: 0.16 ml CD3 OD, ^ 2.0 mg, ^ 1.9 mg) * May be reversed relation with H-2p. Finally, H-2P correlates with H-la and H-lp. Providing that H-5a is assigned by its chemical shift and coupling pattern (doublet of doublet), simply mapping the ^H-^H couplings by a COSY spectrum is sufficient to assign all signals of ring A protons. Unfortunately, assignment of all the side chain protons starting from a methyl signal suffer from the frequent overlapping of H-20 and H-24. Chemical shifts and coupling constants (or linewidth) of signals in the low-field region of the proton NMR spectrum are significant for changes at rings A and B or in the side chain of the brassinosteroid. Thus, epimerisation at C-3 (3a-0H —> 3p-0H) results in a high-field shift of H-5a due to the absence of the deshielding 1,3 diaxial interaction with 3a-0H. On the other hand a dramatic enlargement of the linewidth of the H-3 multiplet (Fig. 14) is observed, since H-3p is an equatorial oriented proton, whereas the axial H-3a shows a different coupling pattern because of the large axial-axial vicinal coupling constants. Esterification of 3-OH results in a downfield shift of H-3 of about 1 ppm. Finally, glycosylation leads to new spin systems in the low-field proton region, which can be recognized in a ^H-^H COSY 2D NMR spectrum.

536 OH

JQAÎ/ 3" J &

J

(ppmll 1.0

I^ ' i.1.2i

D

*-H

i.6i

i.8i

e

2.0 2.2 2.4i 2.6 2.8' 3.0-

i

3.2i

^

3.41 3.6

r[iiiiinii|iiii|iiiiniii|riii|im|iiir]iiii|ini|mi)irn

3.4

3.0

2.6

2.2 Fl

ri|mi[iiii|iin|iiii|iiii|iiii|iiii|iin|iiii[iiii|iiii[ifiniiii

i.B

1.4

1.0

0.6

(ppm)

Fig. 13 %-^H 2D COSY spectrum of 3,24-diepicastasterone (14) (500 MHz, solvent: 0.16 ml CDCI3, 0.05 mg) Marked: fingerprint region of an unchanged 24-epicastasterone (or 24-epibrassinolide) side chain

537

CO.

d

|Hi>|iin|ini[iiii)iiii|iiii|inijiiin"ii|iiii|iiii[iiiniiiijini|iiiijiiir|iiiijii 4.0

3.B

3.6

3.4

3.2

3.0

2.8

2.6

2.4

ppm

Fig. 14 Low-field region of ^H NMR spectra of 24-epicastasterone (13)^ and (20R)-hydroxy3,24-diepicastasterone (107)^ (500 MHz, solvent: 0.16 ml CDCI3, ^ 2.0 mg, ^ 0.2 mg) After assignment of proton signals in the low-field and the methyl region, this assignments can be transferred to carbon signals by a ^H-^^C shift correlation via one bond (Fig. 15), which should be carried out as the proton detected experiment (HMQC: heteronuclear multiple quantum correlation)^^ for sensitivity reasons. Such two-dimensional spectra show correlation peaks at the ^-^C chemical shift in one dimension and at the ^H chemical shifts (of those protons which are bound directly to the carbon) in the other dimension. Mostly, complete sU'uctural elucidation requires information from ^H-^^C shift correlations via two or three bonds (so-called "long-range" correlations). Again, for sensitivity reasons the proton detected version (HMBC: heteronuclear multiple bond correlation)^^^ is the experiment of choice. Each angular methyl group gives four correlations via ^JQ J^ and ^J^^ ^ with carbon signals, whereas each side chain methyl group gives three correlations (Table 3). While methyl groups always cause strong correlation peaks in the HMBC spectrum, correlations between methine or methylene protons and carbons via two and three bonds may be weak or even undetectable, depending on the ^H-^^C coupling constant over two or three bonds and on the proton signal multiplet splitting. Because Me-18 (proton singlet) and Me-21 (proton doublet) have a mutual HMBC correlation to C-17, these signals are easy to assign. In the same way Me-26 and Me-27 can be as-

538

-

"1

(ppm)I

f

20-

_

0.0

II

30-]

-

11

i»«

•

40-

_ -

•II

• 10

60-

70-

i*

III

c-

50-

•

1

* ' "•*

0

-

# • ' ' 1 ~T 1 1 r i • ' ' '

4.0

3.5

1' ' ' ' 1' • 3.0 2.5 F2

T—r"i—1—r—J J J p—1 J 1 J 1 1 p J

2.0

1.5

1.0

(ppm)

Fig. 15 ^H-^^C one-bond shift correlated 2D NMR spectrum (HMQC) of 3-dehydro-24epicastasterone (96) (500 MHz, solvent: 0.16 ml CDCI3, 0.4 mg)

539 Table 3 Expected correlations between ^H NMR methyl group signals and ^^C NMR signals in ^H-'^^C shift correlation 2D NMR spectra via two and three bonds for brassinosteroids

c

1 5 9 10 12 13 14 17 20 22 23 24 25 26 27

Me-18

Me-19

Me-21

Me-26

Me-27

X X

X X X

Me-28

X X X X X X X X

X X X X X X

X

signed due to their mutual correlation with C-25 and C-24. C-22 and C-23 can be assigned by the correlation with Me-21 and Me-28, respectively. Considering the correlations found in the HMQC spectrum, the proton NMR signals of H-22 and H-23 can be readily assigned, too. In a similar manner, the assignment of the methyl group proton signals and a lot of carbon signals is possible in the case of the metabolic side chain hydroxylation. For example (Fig. 16), two ^H methyl group singlets of (20/?)-hydroxy-3,24-diepicastasterone (107) (5 1.30, side chain methyl group because of three HMBC correlations; 5 0.88, angular methyl group because of four HMBC correlations) show a mutual HMBC correlation to a carbon signal at 5 56.7, which therefore has to be assigned to C-17. Consequently, the side chain methyl group (5 1.30) is assigned to Me-21 and the angular methyl group (5 0.88) to Me-18. Me-21 exhibits two further HMBC correlations to carbon signals at 5 78.0 and 74.7, respectively. The former carbon signal gives no correlation in the HMQC experiment and it has to be therefore a quartemary carbon. The carbon signal at 5 77.4 is known to belong to a methine carbon from the HMQC spectrum. Considering the ^^C chemical shifts, both carbons must be hydroxylated. Thus, hydroxylation at C-20 has taken place in the course of the metabolic conversion. The HMBC correlation between the anomeric proton of the sugar unit and the three bond distant steroid carbon in glycosylated metabolites is very important for elucidation of the glycosylation site. Thus, the ^H NMR signal of the anomeric proton of glucose in 25-|3-D-glucopyranosyloxy-24-epibrassinolide (88) exhibits, apart from coirelations with glucose carbons, one correlation to a genin carbon (Fig. 17) which can be assigned to C-25 by its HMBC correlations with the two ^H methyl group singlets at 5 1.36 and 1.30 (Me-26 and Me-27). Unfortunately, such HMBC correlations involving anomeric proton signals may be weak or superimposed by

540

25-d C-25

30-d 35H

AoA

C-12

- > OOo C-lOv

C-24

45-§ 50-j 55

0®0

-4-

C-17 - • dQO

60-:] 65 70H

C-23

- > oOOo

C-22

75H

C-20 I I j I I t I I I I I I I I I i I I I I I I I I I I I I I I I I I I I I I I I 1 I "I I 1 I I I I I 1 I I r I I r 1 1 I I I I T

1.30

1.20

1.10

1.00 F2

0.90

0.80

0.70

(ppm)

Fig. 16 Part of the ^H-^^C long-range shift correlated 2D NMR spectrum (HMBC) of (20R)-hydroxy-3,24-diepicastasterone(107) (500 MHz, solvent: 0.16 ml CD3OD, 0.2 mg); for the shown spectrum CD3OD was used as solvent instead of CDCI3 because the cross-peaks are more separated

541

_^

Me.28

F2 (ppm)

Me-26/ Me-27

1.5i 26

2 . OH

OH ^^^

'

CH3 '^CH3

2.5H

3.0"

3.5H

4.0H

glc-Hl*

4.511 r {I u 1 j I M 1111111n11111111111111 n 1}! 111 j n 111

86 85

84 83 82 8 1 80 79 78 Fl

77

(ppm)

Fig. 17 Part of the ^H-^^C long-range shift correlated 2D NMR spectrum (HMBC) of 25-P-D-glucopyranosyloxy-24-epibrassinolide (88) (500 MHz, solvent: 0.5 ml CD3OD, 1.8 mg) other correlations as in the case of 26-p-D-glucopyranosyloxy-24-epibrassinolide (89). The determination of the glycosylation position is also possible based on ^H acetylation shifts after complete acetylation of the metabolites. Protons at the acetoxylated carbons resonate about 1 ppm downfield in comparison with the non-derivatized metabohte, whereas those at the glycosylation sites are only slightly shifted. For example, the non-separated mixture of the two metabolites 2-0-p-D-glucopyranosyl-3,24-diepicastasterone (98) and 3-0-p-D-glucopyranosyl3,24-diepicastasterone (99) show after acetylation in the ^H-^H COSY spectrum two sets of H-2p/H-3a signals at 5 4.92/3.61 (major component) and 3.76/4.74 (minor component), respec-

542 lively. Consequently, 99 represents the major and 98 the minor glucosylated metabolite. Nuclear Overhauser effects (NOE) contain information about the spatial proximity of protons (or hetero nuclei). Detailed NOE investigations can allow estimation of configuration and/or conformation. NOE measurements can be carried out one-dimensional as NOE difference spectrum or as a two-dimensional NOESY or ROESY experiments. In case of brassinosteroid metabolites it can be advantageous to perform the NOE difference experiment because of its better overall sensitivity, especially if only some NOE interactions will yield the desired information. In such a manner NOE enhancements found for (20R)-hydroxy-3,24-diepicastasterone (107) (Fig. 18) by irradiation of methyl group and side chain proton signals suggest (20R) configuration. This is in agreement with the Me-21 ^H downfield shift of 0.36 ppm compared with 24epicastasterone (13). A similar downfield shift has been found also for corresponding 20-hydroxylated cholesterols^^^, pregnanes^^^ and withanolides^^^.

X I

' 1

mt »W»i.>»iiiMHi», iiitli^^Wm^mnyU*! n n » » » n W , » > w y A L,gW^M

X I

7

\

»J\^

i

X I 0'^-îmAml>*frm^mt^ * « t i H M M « i » * « i i ^

OH ^21 OHf OH

U^

CH3 I '

'**»t,Jft*^',%,

27

CH3 26

l_ÂâJ^4>V^

xJ^^

^»ffyU

LJJIL

M|iiiiiiiiijMii|iiii|Miit.iii|iiii|inijiMi|nii|iiii{iin)iiM|[MijiMi|nii|iiii|iiii[iiii|inijiiiitnii|iiinMiijniu

3-6

3.2

2.8

2.4

2.0

1.6

1.2

ppm

Fig. 18 ^H NMR NOE difference spectra of (20R)-hydroxy-3,24-diepicastasterone (107) (500 MHz, solvent: 0.16 ml CDCI3,0.2 mg)

543 ^H and ^^C chemical shifts of brassinosteroid metabolites are shown in Tables 4 and 5, respectively.

Table 4 ^H chemical shifts of parent brassinosteroids 12 and 13 and metabolites 14, 88, 89, 92, 93,96, 97 and 99 -107; chemical shifts are obtained from the ^H NMR (500 MHz), 2D COSY or 2D HMQC spectra; * may be reversed; ^ solvent: CDCI3; ^ solvent: CD3OD; ^ solvent: 95 vol. % CDCI3 + 5 vol. % CD3OD; n. d.: not detected because of poor signal-to-noise ratio and/or overlapping with other signals

H

12a

13a

14a

88^

89^

92b

93^

96^

1 1.86/1.55 1.74/1.55 2.06/1.24 1.82/1.60 1.82/1.60 1.68/1.57 1.68/1.57 2.54/1.44 2 3.60 3.65 3.77 3.60 3.64 3.60 3.65 4.25 3.98 3.39 3.92 3 4.05 3.92 3.94 3.96 4 2.10/1.93 1.92/1.72 1.96/1.60 2.07/1.80 2.05/1.80 1.76/1.66 1.78/1.67 2.72/2.52 3.12 2.69 2.32 3.21 5 3.20 2.72 2.64 2.73 7 ca.4.10 2.30/2.00 n.d. 4.18/4.08 4.19/4.08 2.20/2.11 2.21/2.11 2.39/2.00 n. d. 1.72 1.76 1.69 1.69 8 1.80 1.79 1.85 1.34 1.40 n.d. 1.42 9 1.30 n.d. 1.37 1.43 1.81/1.44 1.68/1.34 1.68/1.41 1.68/1.42 n.d. 11 1.80/1.40 1.65/1.34 n.d. 12 1.99/1.22 2.02/1.28 n.d. 2.04/1.27 n.d. 2.09/1.32 2.07/1.31 2.06/1.30 n.d. 14 1.18 1.31 1.25 n.d. 1.37 1.37 1.33 1.73/1.24 1.59/1.14 1.59/1.15 1.60/1.15 n.d. 15 1.68/1.22 1.58/1.11 n.d. n.d. 1.99/1.35 2.00/1.37 2.02/1.34 16 1.99/1.25 1.98/1.30 n.d. 2.00/1.39 1.60 1.57 n.d. 1.59 1.56 n.d. 1.57 17 1.56 0.74 0.68 0.77 0.78 0.70 0.71 0.68 0.73 18 1.04 0.76 0.92 0.81 0.90 0.90 0.76 0.76 19 1.54 1.59 1.48 1.46 1.58 1.60 1.46 1.47 20 0.98 0.97 0.98 0.98 1.06 0.97 0.98 1.05 21 3.66 3.69 3.69 3.65 3.67 3.70 3.69 3.66 22 3.41 3.54 3.54 3.35 3.41 3.41 3.35 3.37 23 1.50 1.99 1.93 1.50 1.98 1.93 1.48 1.51 24 1.89 2.20 2.19 1.90 1.90 1.90 25 3.71/3.41 0.92 0.92 0.92 3.72/3.40 1.35 0.92 1.36 26* 0.87 0.87 1.29 0.87 0.87 1.30 27* 0.87 0.87 0.85 0.85 0.89 0.85 0.85 0.85 0.90 0.85 28 1' 2' 3' 4' 5' 6'

4.23 4.22 4.56 4.56 3.17 3.17 3.12 3.12 3.36 3.332 n.d. 3.33 3.24 n.d. 3.28 n.d. 3.24 3.24 n.d. 3.25 3.81/3.60 3.86/3.65 3.80/3.61 3.86/3.64

544 Table 4 Continued H 1 2 3 4 5 7 8 9 11 12 14 15 16 17 18 19 20 21 22 23 24 25 26* 27* 28 1' T y 4' 5' 6* -CH2-CH3

97b

99b

100^

101^

102^

103^

104^

105^

n.d. n.d. 2.22/1.20 2.24/1.21 n.d. 1.99/1.22 2.05/1.23 iTd 3.72 3.54 3.72 3.51 3.52 3.46 3.47 3.63 4.55 4.56 3.39 3.34 3.40 3.40 3.48 3.60 n.d. n.d. n.d. n.d. 2.08/1.93 2.09/1.96 1.83/1.50 2.01/1.52 2.97 2.91 2.97 2.91 2.90 2.46 2.91 2.45 4.12/4.03 4.14/4.01 4.12/4.03 4.12/4.01 4.13/4.02 4.12/4.02 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n,d. n.d. n.d. n.d. n.d. n.d. 1.42 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 1.34 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n. d. n.d. n.d. n.d. n.d. n.d. n.d. 1.54 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.72 0.87 0.70 0.71 0.66 0.71 0.71 0.73 0.79 0.78 0.97 0.96 0.97 0.97 0.97 0.97 1.49 n.d. 1.49 1.43 n.d. n.d. 1.02 0.98 0.96 1.33 2.13 0.97 1.00 0.97 3.66 3.66 3.65 3.43 3.68 3.63 3.68 3.48 3.38 3.33 3.75 3.41 3.41 3.46 1.72 1.47 n.d. 1.49 1.77 n.d. n. d. 1.95 n.d. 2.01 n.d. n.d. 0.91 1.22 0.92 0.95 1.30 0.92 0.92 1.20 0.86 0.87 0.82 1.28 0.87 0.87 0.83 0.84 0.83 0.76 0.82 0.85 0.85 4.36 n.d. n.d. n.d. n.d. 3.84/3.65 1.26 0.88

1.25 0.88

545 Table 4 Continued

H 1 2 3 4 5 6 7 8 9 11 12 14 15 16 17 18 19 20 21 22 23 24 25 26* 27* 28 -CH2-CH3

106^

107a

108^

109^

2.05/1.23 2.08/1.26 1.94/0.98 3.60 3.61 3.63 3.39 3.40 3.45 1.96/1.61 1.98/1.60 1.81/1.70 2.34 2.32 1.28 3.88 4.13/4.02 2.32/n. d. n.d. 1.82/1.19 n.d. n.d. n.d. n.d. 1.34 n.d. n.d. n.d. 1.66/n. d. n.d. n.d. n.d. n.d. n.d. n.d. 2.08/1.31 1.31 n.d. n.d. n.d. 1.84/n. d.* n.d. n.d. n.d. 1.59/n. d.* n.d. n.d. n.d. 1.90 n.d. n.d. n.d. 0.64 0.84 0.62 0.71 0.80 0.81 1.07 0.97 2.13 2.12 n.d. 1.34 0.97 3.35 3.68 3.77 3.41 1.60 n.d. 2.02 n.d. 0.95 0.92 0.83 0.87 0.77 0.85 n.d. 3.68 4.55 n.d. 2.97

1.25 0.88

110^

IIP

112^

n.d. 3.80 4.58 n.d. 2.38

n.d. 3.79 4.57 n.d. 2.38

n.d. 3.80 4.58 n.d. 2.38

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.68 0.81 n.d. 0.98 3.69 3.41 n.d. n.d. 0.92 0.87 0.85



1.26 0.88

1.25 0.88

1.25 0.88

546 Table 5 ^^C chemical shifts of parent brassinosteroids 12 and 13 and metabolites 88, 89, 92, 93, 96, 99 and 107 ^^C chemical shifts of 12, 13 and 88 are obtained directly from the ^^C{^H} NMR (126 MHz) spectra; ^^C chemical shifts of 89, 92, 93, 96^ 99 and 107 are obtained from the ^H (500 MHz) detected HMQC and HMBC spectra; * may be reversed; ^ solvent: CDCI3; ^ solvent: CD3OD; n. d.: not detected because of poor signal-to-noise ratio and/or overlapping with solvent signals

c

12^

13^

88^

89^

92^

93^

96^

99b

107^

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26* 27* 28

41.2 67.9 68.0 31.1 40.9 176.8 70.5 39.1 58.0 38.2 22.2 39.6 42.2 51.2 24.7 27.6 52.5 11.5 15.3 40.1 12.3 72.4 76.0 41.4 26.9 22.1 17.2 10.8

39.9 68.0 68.1 26.2 50.7 212.9 46.6 37.7 53.6 42.5 21.1 39.3 42.7 56.4 23.8 27.6 52.5 11.7 13.4 40.1 12.3 72.4 75.9 41.4 26.9 22.0 17.1 10.7

42.3 69.0 69.2 32.8 42.2 179.3 71.8 40.5 59.4 39.2 23.3 41.0 43.7 52.5 25.8 29.0 54.7 12.1 15.8 43.4 13.8 73.1 79.0 45.6 84.1 20.9 27.0 13.9

42.2 69.0 69.2 n.d. 42.2 n.d. 71.8 n.d. 59.4 39.1 n.d. 41.0 43.6 52.4 n.d. n.d. 54.1 12.2 15.8 41.1 13.0 73.7 77.8 36.3 33.4 75.2 12.1 11.6

40.9 69.1 69.4 27.8 52.0 214.5 47.4 39.0 55.0 43.6 22.3 40.8 44.0 57.8 24.9 29.0 54.7 12.2 13.7 43.2 13.7 73.2 79.0 45.5 84.1 20.9 26.9 13.8

40.9 69.1 69.4 27.7 52.0 n.d. 47.4 39.1 55.0 43.6 22.4 40.8 44.0 57.8 24.8 28.6 54.1 12.3 13.8 41.0 13.0 73.8 77.8 36.2 33.4 75.1 12.0 11.5

47.8 72.0 211.4 35.0 58.6 208.2 46.3 37.8 53.4 42.1 21.8 39.2 42.8 56.3 23.8 27.6 52.6 11.8 13.8 40.2 12.4 72.6 76.4 41.4 27.0 22.0 17.2 10.8

45.8 n.d. n.d. 28.9 57.0 n.d. n.d. n.d. 54.8 39.8 n.d. 40.7 43.9 57.8 n.d. n.d. 54.1 12.2 15.6 41.6 13.0 73.4 77.3 42.7 27.9 22.5 17.4 11.1

44.3 72.1 75.8 27.8 56.7 209.8 46.4 n.d. 53.9 42.5 21.5 39.8 43.5 56.4 22.8* 23.6* 54.9 13.8 14.3 77.8 23.2 75.4 71.4 42.2 26.7 21.4 16.0

98.1 75.1 78.1 71.6 77.9 62.8

105.0 n.d. n.d. n.d. n.d. 62.9

98.1 75.1 n.d. n.d. n.d. 62.7

104.8 75.1 67.3 71.7 77.9 62.8

r T 3' 4' 5' 6'

102.8 74.9 n.d. n.d. n.d. 62.5

9.9

547 REFERENCES 1

H. G. Cutler, T. Yokota and G. Adam (Eds), Brassinosteroids - Chemistry, Bioactivity, Applications, ACS Symp. Ser. No. 474, American Chemical Society, Washington DC, 1991. 2 V. Marquardt and G. Adam, in: W. Ebing (Ed.-in-Chief), Chemistry of Plant Protection. Vol. 7, Springer, Berlin, 1991, pp 103-139. 3 A. Sakurai and S. Fujioka, Plant Growth ReguL, 13 (1993) 147. 4 V. A. Khripach, F. A. Lavich and V. N. Zhabinskii, Brassinosteroids (in Russian), Science and Technics, Minsk, 1993. 5 M. D. Grove, G. F. Spencer, W. K. Rohwedder, W. K. Mandava, J. F. Worley, Jr., J. D. Warthen, G. L. Steffens, J. L. Flippen-Anderson and Jr. J. C. Cook, Nature 281 (1979) 216. 6 T. Yokota, S.-K. Kim and N. Takahashi, in 13th Int. Conf. on Plant Growth Substances. Abstr. No. 168, Calgary, 1988. 7 S.-K.Kim, in ref. 1, pp. 26-35. 8 W. Schliemann, Naturwissenschaften 78 (1991) 392. 9 S. Asakawa, H. Abe, Y. Kyokawa, S. Nakamura and M. Natsume, Biosci. Biotech. Biochem. 58 (1) (1994) 219. 10 T. Yokota, S.-K. Kim, Y. Fukui, N. Takahashi, Y. Takeuchi and T. Takamatsu, Phytochemistry 26 (1987) 503. 11 S. Takasuto, H. Abe and K. Gamoh, Agric. Biol. Chem. 54 (1990) 1057. 12 G. Adam and V. Marquardt, Phytochemistry 25 (1986) 1787. 13 N. Ikekawa, S. Takatsuto, T. Kitsuwa, H. Saito, T. Morishita and H. Abe, J. Chromatogr. 290 (1984) 289. 14 K. Richter and J. Koolman, in ref. 1, pp. 265-278. 15 K. Richter and G. Adam, Naturwissenschaften 78 (1991) 138. 16 N. B. Mandava, J. M. Sasse and J. H. Yopp, Physiol. Plant. 53 (1981) 453. 17 W. J. Meudt, Plant Physiol. 83 (1987) 195. 18 J. M. Sasse, Physiol. Plant. 80 (1990) 401. 19 T. Iwasaki and H. Shibaoka, Plant Cell Physiol. 32 (1991) 1007. 20 J. Gartz, G. Adam and H.-M. Vorbrodt, Naturwissenschaften 77 (1990) 388. 21 J. G. Roddick, A. L. Rijuenberg and N. Ikekawa, Physiol. Plant. 87 (1993) 453. 22 J. G. Roddick, Phytochemistry 37 (1995) 1277. 23 H. Ronsch, G. Adam, J. Matschke and G. Schachler, Tree Physiology 12 (1993) 71. 24 K. Hammada, FFTC Book Ser. 1, 34 (1986) 188. 25 G. Adam and U. Petzold, Naturwissenschaften 89 (1994) 210. 26 P. Allevi, M. Anastasia, R. Cerana and P. Ciuffreda, Phytochemistry 27 (1988) 1309. 27 T. Yokota, K. Higuchi, Y. Kosaka and N. Takahashi, in: Progress in Plant Growth Regulation, (C. M. Karssen, L. C. Van Loon, D. Vreugdenhil, Eds.). Dordrecht: Kluwer 1992, pp. 298-305. 28 J. M. Sasse, T. Yokota, P. E. Taylor, P. G. Griffiths, Q. N. Porter and D. W. Cameron, ibid. pp. 319-325. 29 O. N. Kulajewa, E. A. Burkhanowa, A. B. Fedina, N. V. Danilowa, G. Adam, H.-M. Vorbrodt and V. A. Khripach, Dokl. Akad. Nauk SSSR 305 (1989) 1277. 30 S. D. Clouse and D. Zurek, in ref 1, p. 122-140. 31 J. M. Sasse, in ref. 1, pp. 158-166 32 D. M. Zurek and S. D. Clouse, Plant Physiol. 104 (1994) 161. 33 D. M. Zurek, D. L. Rayle, T. C. McMorris and S. D. Clouse, Plant Physiol. 104 (1994) 505. 34 K. Gamoh, I. Yamaguchi and S. Takatsuto, Anal. Sciences 10 (1994) 913. 35 K. Wada, S. Marumo, N. Ikekawa, M. Morisaki and K. Mori, Plant Cell Physiol. 22 (1981)323. 36 K. Wada, H. Kondo and S. Marumo, Agric. Biol. Chem. 49 (1985) 2249. 37 T. Yokota, S. Watanabe, Y. Ogino, I. Yamaguchi and N. Takahashi, J. Plant Growth Regul. 9 (1990) 151. 38 S. Takatsuto, B. Ying, M. Morisaki and N. Ikekawa, J. Chromatogr. 239 (1982) 233. 39 N. Ikekawa and S. Takatsuto, Mass Spectroscopy (Japan) 32 (1984) 55. 40 H. Abe, T. Morishita, M. Uchiyama, S. Takatsuto and N. Ikekawa, Agric. Biol. Chem. 48 (1984) 2171.

548 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82

R. D. Plattner, S. L. Taylor and M. D. Grove, J. Nat. Prod. 49 (1986) 540. K. Gamoh, K. Omote, N. Okamoto and S. Takatsuto, J. Chromatogr. 469 (1989) 424. J. Schmidt, B. Spengler, T. Yokota, M. Nakayama, S. Takatsuto, B. Voigt and G. Adam, Phytochemistry 38 (1995) 1095. H. Abe, C. Honjo, Y. Kyokawa, S. Asakawa, M. Natsume and M. Narushima, Biosci. Biotech. Biochem. 58 (1994) 986. T. Yokota, M. Nakayama, T. Wakisaka, J. Schmidt and G. Adam, Biosci. Biotech. Biochem. 58 (1994) 1183. F. J. Hammerschmidt and G. Spiteller, Tetrahedron 29 (1973) 3995. J. Schmidt, B. Spengler, T. Yokota and G. Adam, Phytochemistry 32 (1993) 1614. B. Spengler, J. Schmidt, B. Voigt and G. Adam, Phytochemistry (1995) in press. T. Yokota, M. Morita and N. Takahashi, Agric. Biol. Chem. 47 (1983) 2149. J. Schmidt, T. Yokota, B. Spengler and G. Adam, Phytochemistry 34 (1993) 391. A. K. Zaki, J. Schmidt, F. M. Hammouda and G. Adam, Planta Medica 59 (1993) A 613. J. Schmidt, B. Voigt and G. Adam, Phytochemisu-y (1995) in press. J. Schmidt, C. Kuhnt and G. Adam, Phytochemistry 36 (1994) 175. K.-H. Park, J.-D. Park, K.-H. Hyun, M. Nakayama and T. Yokota, Biosci. Biotech. Biochem. 58 (1994) 1343. J. Schmidt and G. Adam, in preparation. H. Suzuki, S. Fujioka, T. Yokota, N. Murofushi and A. Sakurai, Biosci. Biotech. Biochem. 58 (1994) 2075. H. Abe: in ref. 1, pp. 200-207. J. Schmidt, T. Yokota, G. Adam and N. Takahashi, Phytochemistry 30 (1991) 364. J. Schmidt, U. Himmelreich and G. Adam, Phytochemistry (1995) in press. K. Mikami and S. Sakuda, J. Chem. Soc, Chem. Commun. 1993, 710. T. G. Back, P. G. Blazecka and M. V. Krishna, Can. J. Chem. 71 (1993) 156 . B. G. Hazra, P. L. Joshi, B. B. Bahule, N. P. Argade, V. S. Pore and M. D. Chordia, Tetrahedron 50 (1994) 2523 . T. Yokota and K. Mori, Mol. Struct. Biol. Act. Steroids 317 (1992) (M. Bohl and W. L. Duax, Eds., CRC Boca Raton, Florida). H. L. Kwong, C. Sorato, Y. Ogino, H. Chen and K. B. Sharpless, Tetrahedron Letters 31 (1990) 2999. L.-F. Huang, W.-S. Zhou, L.-Q. Sun and X.-F. Pan, J. Chem. Soc, Perkin Trans. 1 1993 1683. H. G. Cutler, in: Natural and Engineered Pest Management Agents, (P. A. Hedin, J. J. Menn, R. M. Hollingworth, Eds.), ACS Symp. Ser. No. 551 (1994) 85. T. C. McMorris and P. A. Patil, J. Org. Chem. 58 (1993) 2338. A. Kolbe, V. Marquardt and G. Adam, J. Labeled Comp. Radiopharm. 31 (1992) 801. B. Voigt, S. Takatsuto, T. Yokota and G. Adam, J. Chem. Soc, Perkin Trans. 1 1995 in press. M. Aburatani, T. Takeuchi and K. Mori, Agric Biol. Chem. 51 (1987) 1909. S. Takatsuto, N. Yazawa, M. Ishiguro, M. Morisaki and N. Ikekara, J. Chem. Soc, Perkin Trans. 1, 1984 139. M. J. Thompson, N. Mandawa, J. L. Flippen-Anderson, J. F. Worley, S. R. Dutky, W. E. Robbins and W. Lusby, J. Org. Chem. 44 (1979) 5002. O. Mitsunobu, Synthesis 1 (1981) 1. B. Voigt, J. Schmidt and G. Adam, in preparation. B. Hellrung, B. Voigt, J. Schmidt and G. Adam, in preparation. K. Wada and S. Marumo, Agric Biol. Chem. 45 (1981) 2579. A. Amann, G. Ourisson and B. Luu, Synthesis 11 (1987) 1002. W. G. Dauben and D. S. Fullerton, J. Org. Chem. 36 (1971) 3277. T. Yokota, Y. Ogino, H. Suzuki, N. Takahashi, H. Saimoto, S. Fujioka and A. Sakurai, in ref. 1, pp. 86-96. H. Suzuki, S. Fujioka, S. Takatsuto, T. Yokota, N. Murofushi and A. Sakurai, J. Plant Growth Regul. 13(1994)21. H. Suzuki, T. Inoue, S. Fujioka, S. Takatsuto, T. Yanagisawa, T. Yokota, N. Murofushi, A. Sakurai, Biosci. Biotech. Biochem. 58 (1994) 1186. N. P. Milner and H. H. Rees, Biochem. J. 231 (1985) 369.

549 83 84 85 86

87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110

G. F. Weirich, M. J. Thompson, J. A. Svoboda, Arch. Insect Biochem. Physiol. 12 (1989) 201. T. Yokota, Y. Ogino, N. Takahashi, H. Saimoto, S. Fujioka and A. Sakurai, Agric. Biol. Chem. 54 (1990) 1107. H. Suzuki, S. Fujioka, S. Takatsuto, T. Yokota, N. Murofushi and A. Sakurai, J. Plant Growth. Regul. 12 (1993) 101. T. Yokota, S.-K. Kim, Y. Kosaka, Y. Ogino and N. Takahashi, in: K. Schreiber, H. R. Schutte, G. Sembdner, (Eds.), Conjugated Plant Hormones, Structure, Metabolism and Function, Proceedings of the Int. Symp. Conj. Plant Horm., 3-7 Nov., 1987, Gera, p. 288-296. H. Suzuki, S.-K. Kim, N. Takahashi and T. Yokota, Phytochemistry 33 (1993) 1361. Z. A. Wojciechowski, in: G. W. Patterson and W. D. Nes (Eds.), Physiology and Biochemistiy of Sterols, Am. Oil Chem. Soc, Champain, Illinois, 1991, pp. 361-395. J. Schmidt and G. Adam, (unpublished results). C. D. Schlagnhaufer and R. N. Arteca, J. Plant Physiol 138 (1992) 191. L. Dyas and L. J. Goad, Phytochemistry 34 (1993) 17. M. H. Zenk Phytochemistry 30 (1991) '3861. B. Schneider, A. Kolbe, A. Porzel and G. Adam, Phytochemistry 36 (1994) 319. T. Hai, B. Schneider and G. Adam, Phytochemistry (1995) in press. N. Ikekawa, Bioorg. Med. Chem. Lett. 3 (1993) 1789. 0 . Kappler, C. Hetru, F. Durst, J. Hoffmann, in: J. Koolman (Ed.), Ecdysone - From Chemistry to Mode of Action, Thieme, Stuttgart, 1989, pp. 161-166. T. Hai, B. Schneider, A. Porzel, G. Adam, Phytochemistiy (1995) in press. A. Kolbe, B. Schneider, A. Porzel, B. Voigt, G. Krauss and G. Adam, Phytochemistry 36 (1994) 6-71. A. Kolbe, B. Schneider, A. Porzel, J. Schmidt and G. Adam, Phytochemistry (1995) (in press). A. Kolbe, B. Schneider, A. Porzel and G. Adam, Phytochemistry (1995) in press. H. H. Rees, in: J. Koolman (Ed.), Ecdysone - From Chemistry to Mode of Action, Thieme, Stuttgart, 1989, pp. 28-38. R. Lafont, D. H. S. Horn, ibid. pp. 39-64. L. J. Goad, in: B. V. Charlwood, D. V. Banthorpe (Eds.), Terpenoids, Vol. 7 of "Methods in Plant Biochemistry", Academic Press, 1991, pp. 369-434. A. Porzel, V. Marquardt, G. Adam, G. Massiot and D. Zeigan, Magn. Reson. Chem. 30 (1992) 651. T. Ando, M. Aburatani, N. Koseki, S. Asakawa, T. Mouri and H. Abe, Magn. Reson. Chem. 31 (1993) 94. A. Bax and S. Subramanian, J. Magn. Reson. 67 (1986) 565. A. Bax and F. M. Summers, J. Am. Chem. Soc. 108 (1986) 2093. A. Mijares, D. I. Cargill, J. A. Glasel and S. Liebermann, J. Org. Chem. 32 (1967) 810. C. H. Robinson and P. Hofer, Chem. Ind. (London) (1966) 377. G. Adam and M. Hesse, Tetrahedron 28 (1972) 3527.



551

Structure Elucidation and Synthesis of the Lignans from the Seeds of Hernandia ovigera L. Masao Arimoto, H. Yamaguchi and S. Nishibe

1 Introduction The lignans are groups of natural products whose carbon skeltons are constructed by the linking of C^Cj-units (1), which are formed biogenetically through the shikimate pathway. The term "lignan", reflecting the woody tissue from which many examples derive, was introduced by Haworth (ref. 1), and it is applied to structures that are composed of two C6C3-units, linked / ? - / ? ' (8-8'). The frequent occurrence of this linkage can be ascribed to ^ - /?' coupling of radicals (2), derived by oxidation of, for example, p-hydroxycinnamyl precursor.

etc.

Lignans, which are main area of our concern in this article, are distributed widely in the plant kingdom. Many biologically active lignans are isolated from medicinal plants. A number of reviews about total syntheses and biological activities have been reported (ref. 2). The plants of Hernandiaceae

comprise about sixty-five species in four generic

kinds. The plants of Hernandia ovigera L. are widespread in the tropics and the subtropics. In Japan, these plants grow on the beaches of Okinawa and Ogasawara islands. In the Okinawa region, the plants have been used as therapeutic agents. A number of groups have been involved in the isolation and structure elucidation of aporphine-type and isoquinoline-type alkaloids from these plant materials as well

552 as in the determination of their biological activities. On the other hand, studies of the non-alkaloidal constituents have been scarcely reported. In 1942, Hata has isolated hernandion (3) and isohernandion (4) from the seeds of Hernandia ovigera L. collected in Taiwan (ref. 3a). In 1972, Furukawa et al. have isolated desoxypicropodophyllin (4) from the barks of the roots of Hernandia ovigera L. collected in the Bonin Islands (ref. 3b). In 1973, Nishino and Mitsui have isolated epiaschantin (5) and epimagnolin (6), which are dioxabicyclooctane-type lignans, from the leaves of Hernandia ovigera and carried out their structure elucidation (ref. 4).

0CH3

610

Fig. 3. The NOEs detected for enones 3 and 9.

Acid treatment of 8 resulted in deacetalization and intramolecular aldol cyclization to afford three products, in which the enone 3 was the major. The stereochemistries of the enones 3 and 9, were determined by NOE experiments. On irradiation of methyl groups at C-4 (5 1.03) and C-5 (5 1.08) of the enone 3, a NOE at H-6 (8 3.03) was observed in both cases (Fig.3). While in the case of the enone 9, when H-6 (5 3.06) was irradiated, a NOE at H-4 (6 1.91) was observed. Therefore the cis arrangement of the 1,2-dimethyl groups introduced by 1,4-addition was established at this stage as well as the configuration of the methoxycarbonyl group. Since the alcohol 10 has hydrogen bonding with the carbonyl group at C-1 and acid treatment under similar conditions leads to the enone 3, the stereochemistry was determined as shown in the formula. Hydrogenation of the enone 3 afforded a cis dihydro derivative 11, which was further reduced to give two isomeric alcohols, 4 and 12 (Scheme 2) (6).

611

b, c, d, e, b

COzMe

/—\

CHO COzMe

+ COzMe

1

\

? +

I " COjMe

COjMe

10

C02Me

11

COaMe

Scheme 2. a) CH2=CHMgBr,CuBr«SMe2AHF; b) HOCH2CH2OH, TsOH; c) BH3'THF; H202/NaOH; d) Jones; e) CH2N2; f) LDA, CH2=CHCH2Br; g) O3; Zn-AcOH; h) TsOH/Me2CO-H20; i) H2, Pd-C;j)NaBH4/MeO

612

CamO HO

"9H

H COaMe (+)-4

13 COjMe

f^-

HO

CamO

COoMe

COzMe (-)-4

14

Cam

HO

CamO

H

HO COjMe 15 COoMe 12

COzMe (+)-12

"?H

CamO „ : H

= COzMe 16

COzMe (-)-12

0 (+)-4 or (+)-12

(-)-4 or COaMe (-)-ll

(-)-12 J

=

COaMe

(+)-ll

Scheme 3. a) l(5)-(-)-camphanic chloride, DMAP/CHjClj-Py; b)KOH/MeOH;c) Jones.

613

Both alcohols were treated with l(5)-(-)-camphanic chloride in the presence of DMAP in pyridine to afford a mixture of diatereoisomers, which were separated by HPLC. In the case of alcohol 4, diastereoisomers 13 and 14 were obtained and these were then treated with KOH/MeOH to yield optically active alcohols, (+)-4 and (-)-4. Jones reagent was then used to oxidise these and afforded the ketones, (-)-ll and (+)-ll, respectively. The CD spectrum of (-)-ll showed a negative Cotton effect, and that of (+)-ll a positive. Using the Octant rule, the absolute configurations were assigned as shown in Scheme 3 (7). These results were confirmed by the X-ray analysis of ester 14 (8). The alcohol 12 afforded the ketones (-)-ll and (+)-ll through 15, 16 and 12. The absolute configurations of 15,16 and 12 were determined as shown in Scheme 3.

"9H COiMe (+)-4

(+)-chiloscypholone (18)

(-)-chiloscyphone (1)

isochiloscyphone (19)

Scheme 4. a) POCl3/Py/100°C; b) LiAlH4; c) Swem; d) CH2=C(Me)MgBr; e) Jones

614

One of the alcohols, 12, afforded the disubstituted olefin as a sole product on dehydration, while the other, 4, gave only the trisubstituted olefin 17. The optically active chiloscyphone, (-)-l, [a]^ -15.r (Ht. -24.4') (3), was synthesized in 4 steps from (+)-17, which was obtained from (+)-4 (Scheme 4). The absolute configuration of the alcohol (+)-4 is as shown, therefore the absolute configuration of the natural product was determined as depicted in (-)-l. The absolute value of the specific rotation was a little bit smaller in the synthetic one which is presumably due to the minute quantity of the samples. However we have further confirmed the absolute configuration by the independent synthesis of the antipode (+)-l, [o]^ +24.2% starting from (-)-4 (7, 8). (+)-Chiloscypholone (18) was also isolated as a natural product (3), and it yielded (-)-chiloscyphone (1) and isochiloscyphone (19) on dehydration. Thus the absolute configuration of (+)-18 was also elucidated (Scheme 4).

3.

Tamariscol The liverwort sesquiterpene, (-)-tamariscol (20), isolated from Frullania

tamarisci is a very unstable compound under acidic conditions and has been found in the European species but not in the Japanese species. This is a very strange phenomenon and at first we thought that these two species are different. However a detailed inspection of the Japanese species revealed that those collected at higher altitude or in the northem area do contain tamariscol, but those collected in other parts of Japan do not (9). Connolly and his co-workers investigated the structure of this compound using 2D-NMR techniques including a 2DINADEQUATE and established the structure formulated as 20. However the absolute configuration remained undetermined (10). Tamariscol belongs to a pacifigorgiane skeleton, in which pacifigorgiol (21), isolated from the soft coral Pacifigorgia admsii, is included (Fig. 4) (11). The absolute configuration of this compound is also not known.

615

(-)-tamariscol (20) Frullania tamarisci

pacifigorgiol (21) Pacifigorgia adamsii

Fig. 4. Pacifigorgiane-type sesquiterpenes.

We are interested in compound (-)-20, which has a pleasant fragrance. Later the quality of the fragrance was recognized by companies world wide, although the original paper just mentioned a "pungent taste" (10). Thus we are interested in not only the structure but also the absolute configuration of tamariscol. It is important to know the absolute configuration for developing this compound as a perfume. We started the synthesis to confirm the relative configurations at C-6 and C-7 first. This was previously assigned by consideration of the ^^C NMR chemical shifts and by comparing with known compounds (10). From the literature, compound 23 can be obtained from 22 (12). The relative structure of the enone 23 was verified by independent synthesis of pacifigorgiol (21) (12). It was thought that reductive acylation of the enone 23, through the epoxide 25, may lead to the natural product, but this assumption was not very good. However compound 20 was prepared by alkylation of 27, but the yield was very poor (Scheme 5).

616

c=:> I H COOMe

^

CH2OH

25

H I

>

3

27 tamariscol (20) Scheme 5. Synthetic plan of tamariscol.

At first, the enone 23 was catalytically reduced to afford almost exclusively a cis fused ketone which was expected from the preceding examples. Birch reduction gave cis and trans products in a ratio of ca. 5:4 (13). In general six-membered ring ketones produce trans products preferentially under Birch reduction conditions. In this case the trans product was not the preferred product and the ratio did not change when several different conditions were tried. Thus Birch reduction followed by carboxylation with carbon dioxide of the enolate and methylation afforded the ester, whose ratio of the cis and trans products was ca.

617

3:2. The stereochemistries were estabhshed at the stage of the ketones 38 - 40. The enolizable P-ketoester was cw-fused compound 28, and the non-enolizable one (keto-form) was trans-fustd compound 24. Reduction of the carbonyl group, mesylation or benzoylation, and then base treatment yielded the corresponding a,P-unsaturated ester in each case. Further reduction afforded the allyl alcohols 29 and 30 (Scheme 6).

c, d, e, f a,b

'O^ÔMe

23 g, h, i, f

30

Scheme 6. a) NHsCliq.), ^BuOH; C02/rHF/-78^C; b) CH2N2; c) NaBH4AleOH; d) MsCl/Py; e) DBU/PhH; f) LiAlH4; g) L-Selectride; h) BzCl/Py; i) LDA.

The cis and trans allyl alcohols, 29 and 30, were epoxidized and acetylated to afford the acetates 31, 32, 35, and 25. The stereochemistries were revealed at a later stage. The epoxide 31 was treated with Me^CuLi to give a diol 33 in good yield (Scheme 7a). The other epoxides, 32 and 35 were treated similarly to give diols 34 and 36, respectively, however the epoxide 25 did not react at all. The reason for this is not yet clear, but one reason may be that the epoxides tend to open in the direction to yield diaxial alcohols. In order to obtain an a-hydroxy aldehyde, each diol, 33, 34, and 36, was treated with NCS/Me^S to yield an aldehyde with a methylthiomethyloxy group (for example 37). However these

618

aldehydes decomposed under Wittig reaction conditions (Scheme 7b). A route along this pathway was therefore abandoned and instead we turned our attention to the synthesis of ketone 27.

a- c

29

31

31

33

38

32 CH2OH

34

39

Scheme 7a. a) LiAlH4; b) mCPBA; c) Ac20/Py; d) Me2CuLi; e) PCC

The diols, 33, 34, and 36, were oxidized with PCC to afford in good yield the ketones 38 - 40 respectively. The stereochemistries were determined by a detailed inspection of the NMR spectra of these compounds. The hydrogen at C-1, a to the carbonyl group in 38 and 39, will show the same coupling pattern if these are the trans fused compounds. However, if these are cis fused compounds they have different coupling patterns as they adopt different conformations due to

619

O'^l H CH2OAC

35

35

Hcri H CH20H

36

40

O'î H

\ ^ J CHO 37

Scheme 7b. a) LiAlH4; b) mCPBA; c) AcsO/Py; d) Me2CuLi; e) PCC; f) NCS/Me2S the difference in the configuration of the methyl group at C-3. In fact, compound 38 displays H-1 at 5 2.70 (br t, J=6.4 Hz) and compound 39 at 5 2.79 (q, J=8.5 Hz). The coupling patterns of these protons are quite different from each other and thus show that these are the cis fused compounds. Therefore compound 38 adopts a non-steroidal conformation having an equatorial methyl group at C-3, while in the case of 39, the methyl group is equatorial and adopts a steroidal conformation. Thus ketone 40 is a trans fused compound and the stereochemistry of the methyl group at C-3 is axial, because a NOE between the hydrogen at C-1 and the methyl group at C-3 was observed. So the methyl group at C-3 of 40 is a-axial (Fig. 5).

620

M

"IT 'fi

v " ^ ^ ^

V'

(brt,y=6.4Hz)

2.79 (q, 7=8.5 Hz)

38

39

5HI2.70

8H,

H

'P^ ^•"^NOE 5H,2.13

(ddd, 7=12.7,10.3,7.3 Hz) 40

Fig. 5. Conformations and coupling patterns of ketones 38 - 40.

As ketone 27 was not obtained directly through 25 it was thought that isomerization of the other ketones may give this compound. The MM2 calculation predicts that the trans isomer 27 is the most stable among the four possible isomers. Thus 38 was treated with KfiOjfMeOH under isomerization conditions. Unexpectedly compound 27 was not the most abundant formed, but compound 38 was the major one and compounds 39 and 40 were formed only in minute amounts as expected. The reaction mixture was refluxed overnight, but the ratio did not change, suggesting that the equilibrium was complete and the energy difference between 38 and 27 must be very small. In order to obtain the desired ketone, the enone 23 was converted into 41, followed by 1,4-reduction to afford 26 as shown in Scheme 8. Hydroboration-oxidation gave the alcohol 42, presumably by attack from the less hindered p face. In fact, oxidation of 42 with PDC afforded the ketone 38 . Isomerization of the ketone 38 under K^COg/MeOH conditions yielded 27 . The ratio did not change even if it was started from 39 or 40 (Scheme 8).

621

23

c

41 H

,.-s^p i H HO 4:2

26 H

d • ' • ^ ^ "

.g>

Q,

^

nH

0

38

H

O

H

39

40

Scheme 8. a) Ph3P=CH2; b) Li/NH3(liq.); c) BH3.THF; H202/NaOH; d) PDC; e) K2C03/MeOH/reflux.

The stereochemistry of alkylation was next investigated.

2-

Methylcyclohexanone (43) was treated with MeLi to afford predominantly the trans methylated product 45 . However, when the aluminium compounds MAD or MAT were present, they coordinate to the carbonyl group and cause the configuration to be reversed (14). Very complex cases using other sophisticated alkylating reagents are not reported. When 2-methyl-l-propenyl magnesium bromide was reacted with the ketone 27, the isomer of natural product 46 was the

622 sole product. However the lithium salt of this bromide afforded natural product 20, but the yield was very poor. Several trials did not change the yield. In these cases alkylation with groups other than a methyl group did not result in a reversed stereochemistry, and the methodology obtaining the reversed stereochemistry has not been developed yet (Scheme 9). Therefore the relative stereochemistry of tamariscol was confirmed to be completely correct.

MeLi

44 none MAD

8 93

Me2C=CHX

X=MgBr X=Li

tamariscol (20)

46

0 1

100 30

Scheme 9. Stereochemistry of alkylation

623

As the total synthesis was completed we turned our attention to the absolute configuration next. The yield of alkylation was so bad that we decided to degrade the natural product. Collection and extraction of Frullania tamarisci afforded (-)-tamariscol (20), which was then epoxidized and treated with LiAlH/Etp. The reaction did not proceed in ether but, in benzene solution the allyl alcohol 48 was produced. Oxidation with NalO^ gave the ketone (+)-27, which was completely identical with the racemic synthetic product. The CD spectrum of (+)-27 showed the negative Cotton effect at 294 nm and the absolute configuration was suggested by the Octant rule as depicted in the formula. Independently (-)-27 was prepared from (-)-carvone (49) using a 12 step reaction (15, 16). Thus (-)-carvone (49) was reduced with NaTeH to give the trans-dihyAxo derivative 50, which was reduced and treated with dihydropyrane to afford compound 51. Ozonolysis and WittigHomer reaction provided the methyl ester 52, which was converted into the diol 53 in three steps. Swem oxidation and aldol condensation afforded the enones 54 and 55 after separation. The ketone (-)-39 was obtained by catalytic reduction of 55. As already mentioned, (-)-39 was isomerized into the optically active ketone (-)-27, whose CD spectrum showed the positive Cotton effect at 294 nm, which is opposite to that obtained for the natural one. Therefore the absolute configuration of the natural compound, (-)-20, should be assigned as depicted in the formula (Scheme 10). Each ketone, 27, 38, 39 or 40 has a pleasant fragrance. The fragrance of (-)-tamariscol (20) is woody, earthy floral, a little bit different from these ketones mentioned. A minute quantity of such a ketone as an impurity may alter the quaUty of fragrance. An easy method for qualitative evaluation of the quality of fragrance is being waited.

624

(-)-20

g.h

1, m \^»**'

HH

O

(-)-38

OH

OTHP

OH

52

53

>+ L J ^> %^»'+* o (-)-27

(-)-39

40

CD[ (+)-27]: Ae-1.33(CHCl3)

Scheme 10. a) mCPBA; b) LiAlH4/PhH; c) NaI04; d) NaTeH/EtOH; e) NaBH4; f) DHP/PPTS; g) O3; PPhj; h) (MeO)2P(0)CH2C02Me/ NaH; i) Hj/Pd-C; j) LiAlH4; k) PPTS/MeOH; 1) Swem; m) PhCOjH/ EtjN; n) Hj/PtOj; 0) KzCOj/MeOH

625

4.

Conocephalenol Conocephalenol (56) has been isolated from the European liverwort

Conocephalum conicum and has a similar structure to tamariscol (20) (17). The first compound of this class was brasilenol (57), which was isolated from the alga Laurencia obtusa oi ihtrnxdihrdxiohAplysia brasiliana (Fig. 6) (18). The planar structure of conocephalenol (56) was revealed by Connolly and his group by using a 2D-INADEQUATE spectrum and is shown in Fig. 6 (17). However the ^H NMR spectrum of this compound is very congested in an upfield region and has no characteristic signals. Hence it was difficult to determine the stereochemistry by only spectroscopic methods. It is very interesting from the evolution point of view that similar terpenoids with the same skeleton have been isolated from both liverworts and algae. Since the absolute configuration of no terpenoids in this class has been elucidated, we have started to synthesize conocephalenol (56), whose relative and absolute configuration has not been determined.

OH

brasilenol (57) Laurencia obtusa Aplysia brasiliana

(-)-conocephalenol (56) Conocephalum conicum

Fig. 6. Brasilane-type sesquiterpenes

626

COOMe

H

63

0

O 23

"^ 'O'ÔMe

COOMe

59 +

10 |ig/ml. The same resuh was obtained by the latter authors for 10. The plakorin stereochemical configuration seems to confer the highest cytotoxicity in the P388 assay since 3, 4, 5, 6, and 11 exhibited IC5Q values of 0.05, 0.1, 0.05, 0.1, 0.3 ^g/ml and 12 3 îg/ml respectively. In addition xestin A [11] was strongly active at 5 |ig/ml against lung, colon, and mammary tumors where xestin B [12] was inactive in the same concentration. Plakorin itself is a potent activator of sarcoplasmic reticulum Ca^"ÂTPase (165). Plakortis lita collected off Truk Island and Okinawa had only one compound in common, namely chondrillin [1]. The amounts were very different since the Okinawan sample had 1 as the main compound (0.19% of dry weight) while the Truk sample had 1 as the least abundant peroxyester (0.0008% of dry weight). The Okinawan C^g derived compounds 3, 4, 5, and C20 derived 6 are homologous to the Truk C22 derived compounds 2, 9, and 10, respectively, which are accompanied by the stereoisomeric 7 and 8. In the case of Xestospongia the C24 derived compound 11 is homologous to the Truk C22 derived 10 and the Okinawan Cjg and C20 derived 5 and 6, while 12 is homologous to 8 from the Truk sample. Furthermore, the activity in the P388 assay shows some astounding variations apart from the discrepancy concerning chondrillin [1], namely the fact that C22 derived 10 and the homologous Cjg derived 5 and C20 derived 6 exhibit a variation of at least a factor 100 in IC5Q (>10, 0.05, and 0.1 ng/ml respectively). Of course this apparent inconsistency may be coincidental and merely reflects that the effect is not evaluated against the receptor at which the products are aimed. The latter notion may fmd support in the fact that the main metabolite in the Okinawan sample is chondrillin [1], the least active compound in the P388 assay. Curiously enough these compounds are apparently not chemotaxonomically related since the plakinid sponges belong to the Homosclerophorida, Chondrilla to Hadromerida, and Xestospongia to Nepheliospongia. It is noteworthy that other plakinid sponges such as Plakortis halichondroides and Plakortis angulospiculatus (170) as well as a hadromerid sponge, Chondrosia collectrix, have cycHc peroxides derived from branched chain acids. Some of these compounds exhibit antimicrobial activity, ichthyotoxicity and cytotoxicity (166). Other examples of unique secondary metabolites or metabolite classes with wide taxonomic occurrence are discussed in Appendix 11. The common crust of bread sponge, Halichondria panicea, was found to contain large amounts of sulfur compounds, notably dimethyl trisulfide (171), to such an extent that the organism was considered hazardous to handle by the fishermen in the area. Samples of this species collected at other localities were found to contain trace amounts of the sulfur compounds.

Actually the literature contains several examples comparable to those described

720

above as pointed out by Christophersen and Jacobsen (172).

Appendix 9 G Proteins Some G-proteins (Gg) relay receptor activation to adenylate cyclase and thereby activate cAMP mediated reactions. In bacteria receptors and adenylate cyclase interact directly. Gg proteins consist of a heterotrimer of three polypeptides where one, the a chain (Gg^j^), binds and hydrolyses GTP and activates adenylate cyclase. Tight p chain and y chain complexes anchor Gg to the plasma membrane. It is believed that the complex releases the active a chain on activation. Another G-protein (Gj) contains the same py chains but a different a chain, G^^, This factor on receptor activation inhibits adenylate cyclase. Still another G-protein (Gp) serves to activate phospholipase C and thereby generates inositol triphosphate (InsP3) and diacylglycerol (DAG) from phosphatidylinositol-biphosphate (PIP2). InsP3 releases Ca^"*" from the calcium sequestering compartment thereby making this multipurpose ion available in the cytosol. The DAG produced in the hydrolysis of PIP2 can act as precursor for arachidonic acid, giving rise to the production of prostaglandins and other lipid signalling molecules, and it can activate protein kinase C (PKC, C because it is Ca^"^ dependent). This activation is transient since DAG is rapidly transformed, either by cleavage to arachidonic acid or by phosphorylation to phosphatidate. PKC transfers the terminal phosphate group from ATP to specific serine or threonine residues on target proteins to produce a variety of physiological alterations.

Appendix 10 Evolution of Receptors It is by now well documented that many key biochemical structures are surprisingly highly preserved during evolution. The two- subunit structure of ribosomes is universal and the stem-loop structures of rRNAs are extremely similar in all RNAs. Also the G-proteins (Appendix 9) seem to be evolutionary related containing similar subunit structure and amino acid sequence. The G-protein transducing from the vertebrate eye couples the reception of a photon by the rhodopsin molecule to the activation of a phosphodiesterase enzyme hydrolysing cyclic GMP. The transducing a subunit is about 65% identical in amino acid sequence to the Gj a subunit. Furthermore the receptors linked to G-proteins also seem to constitute a family of evolutionarily closely related receptors. They all consist of seven-pass transmembrane proteins and include such important receptors as the p-adrenergic receptors, the muscarinic acetylcholine receptor, many neuropeptide receptors, rhodopsin, and the cannabinoid receptor. Even bacteriorhodopsin and receptor proteins used by yeast belong in this

721

family which presumably arose early in evolution. A very important component in eucaryotic cells is the transcription factors. It is well established that the primary control of gene expression lies at the level of gene transcription (173). The transcription factors recognize and bind to regulatory sites in DNA. A limited number of families of site specific protems exist. To start transcription of protein-coding genes at an acceptable rate a transcription factor TFIID and RNA polymerase II with accessory proteins are necessary. TFIID specifically binds to the TATA box, usually located about 30 nucleotides upstream of the RNA start site of protein encoding genes. In addition specific activators are normally involved to accelerate transcription (174). Human functional TFIID has been cloned (175-177) and consists of 339 amino acids. The carboxy terminal 181 amino acids share 80% identity with the analogous protein in Saccharomyces cerevisiae, which is functionally replaceable in vitro (175). The budding yeast Saccharomyces cerevisiae and the fission yeast Schizosaccharomyces pombe have diverged several hundreds of millions of years ago (178) and yet the S. pombe TFIID is 93% identical to S. cerevisiae TFIID in a region consisting of a direct repeat (179). This evolutionary conserved core is found in all TFIID's known imtil now. Relative to human TFIID that of Drosophila is 88% conserved, Arabidopsis-l 83%, Arabidopsis-l 82% (177). The conserved structural motif is sufficient for binding as well as transcription promotion. In contrast, the A^-terminal is qualitatively and quantitatively different in different species. Even in the two TFIID's from the plant Arabidopsis thaliana, where about 200 amino acids share 93.5% identity, the A^-terminal sequences are only 75% conserved (180). There is circumstantial evidence that the TFIID itself is subjected to sophisticated regulatory systems (181). It is very likely that this general and important system is target for many secondary metabolites since interference with this system must have serious implications for the organism affected. Furthermore, even apart from the highly conserved core structural motif, the species specific variant A^terminal section lends itself to biochemical interference by reaction with endogenic as well as exogenic secondary metabolites. In the same way nAChR seems to represent an evolutionary highly conserved structure. This is demonstrated by the observation that nAChR's from so dissimilar organisms as the ray, Torpedo californica, the electric eel Electrophorus electricus, and mammalian muscle exhibit similar antigenic properties (182). Strong sequence homology exists in subunits from material from a mouse cell line (BC3H-1), Torpedo, and skeletal muscle nAChR although the mouse system apparently has a function closer to neonatal muscle nAChR (183). The latter observation indicates that the evolution of the nAChR's is reflected in the embryogenesis. The subunits exhibit a conspicuous homology among themselves, and may have diverged from the same primordial gene and accordingly show highly conserved sequences across species. Thus the mouse a subunit is 80% homologous with Torpedo, 86% with chicken, 95 with calf, and 96% with human a subunits (184). Also nAChR from vertebrates, although different, shows significant molecular similarities to nAChR from insects

722

(185). Organisms like the ciliated protozoan Paramecium^ baker's yeast Saccharomyces cerevisiae and the Gram-negative bacterium Escherichia coli have nonselective, mechanically gated cation-channels. The channel from yeast has conductance comparable to nAChR (186). The voltage sensitive sodium channels from eel electroplax have an amino acid composition remarkably similar to that of nAChR and the repeats from Drosophila sequences are closely comparable to the ones from eel and rat (187). The presence of acetylcholine has been well established in all mayor taxonomic groups of the Plant Kingdom and even though the AChR has not yet been isolated from plants strong acetylcholine binding activity has been demonstrated (188) and evidence for the presence of nicotinic-Iike as well as muscarinic-like AGiRs has been presented (189). In plants as in animals acetylcholine is synthesized by choline acetyltransferase and hydrolysed by cholinesterase (189-190). Although the physiology of acetylcholine in plants will have to await further studies, there is circumstantial evidence that it mediates phytochrome action (188). Acetylcholine has been shown to induce Na"*^ and K"*^ output from intact chloroplasts (191) and to stimulate gemiination of photoblastically positive seeds and inhibit gennination of negative ones in continuous white light (192).

Other ion-channel proteins and other basic receptors as well have long evolutionary histories. In a recent compilation of data on synaptic membrane proteins Betz (193) concludes that although the functional diversity of synaptic protein superfamilies largely arose by divergent evolution of common ancestral building blocks, the families themselves share significant sequence homology and have common ancestors early in phylogeny. Furthermore it was noted that there is a certain heterogeneity in channel properties depending on variable subunit composition giving rise to embryonic forms and adult isoforms. These variations offer different targets for secondary metabolites allowing even more subtle effects to ensue. Naturally the added aspect of species dependent variations in receptor structure allows a rigorous definition of target. The voltage sensitive calcium channels have recently been studied using conotoxins from marine cone snails. The o)-conotoxin GVIA is lethal to fish, amphibians, and birds but not to mammals, while MVIIA is only lethal to fish (194).

Evolution of Secondary Metabolites Phylogenetic evolution The present hypothesis assumes interaction between receptors and secondary metabolites. As a consequence receptor evolution reflects an evolutionary selection in secondary metabolites. The main lines in the chemical evolution of living systems, the RNA-world developing to a breakthrough organism, the ancestor of the progenote from which the three kingdoms, archaebacteria, eubacteria, and eucaryotes evolved (195) are established. Within each kingdom the biochemical evolution is still almost

723

terra incognita although much circumstantial evidence is apparent from contemporary chemotaxonomy. In the case of receptors, as discussed above, and other proteins, especially enzymes, the pedigrees are in the process of evaluation (see e.g. 196), but the low molecular weight substrates of these systems still await the emergence of usable working hypotheses to comprehend the available data. In order to reconstruct the paleontology of secondary metabolites information on past families of metabolites is needed. Although information on biochemical fossils is slowly emerging (197), the area is still largely dominated by speculations. The field where most work has been done and from which the most successful results have been gained so far is the lipophilic chemical fossils (198). In connection with crude oil investigations the pattems of sterols and derivatives have been carefully scrutinized. It has been possible to correlate types of geochemical intermediates with source organisms and to devise detailed pathways leading from secondary metabolites to geochemicals. In a few cases these pathways have been established in detail comparable with well investigated biogenetic pathways. Naturally the correlations are the easier and more reliable the less geochemical changes have occurred. However, even very ancient material may still yield interpretable information. Sterols are relatively stable organic structures, configurational isomerization reactions can in the absence of deep burial take longer than the Phanerozoic, viz. about 6(X) million years, to reach equilibrium. Furthermore, their presence in undegraded crude oil attests to the gentle physical conditions prevailing during their geological lives. In conclusion the sterols are presumably nearly optimal structures for geological preservation. Many other geolipids present in petroleum bear witness to the activity of living organisms in earlier geological eras. A complicating factor in crude oil investigations is the very low abundance of polar compounds. Oil deposits are never found at the location of generation. The oil has imdergone primary and secondary migration from the source rock to the reservoir (199-200). During these migrations, which can be of substantial distance, the geological formations act as a huge chromatographic column retaining the polar constituents (201). This obstacle can be overcome by investigating organic matter in sediments. As the sediments mature the kerogen (insoluble part of the organic material) assumes exceedingly complex hetero-oligomeric/polymeric structures. Although it is possible to degrade such structures, for example pyrolytically, the structural information gained in this process requires elaborate knowledge of the geological as well as the pyrolytic reactions employed. At present this knowledge is not detailed enough to allow unambiguous deductions. Geolipids that have been subjected to diagenetic and catagenetic transformations can be manipulated to yield useful information, while metagenesis degrades organic material to methane and graphite with total loss of structural information. In conclusion, even if ancient molecular structures can be elucidated, and have been, the processes of geological diagenesis, catagenesis and related mechanisms are still too incompletely known to allow definite conclusions. Consequently the deductions regarding the original molecular structures are at best tentative at present. However, as knowledge accumulates and

724

techniques advance it is inevitable that molecular paleontology will establish itself as an integral part of contemporary paleontology. At this stage studies of secondary products from extinct taxons and their evolution and dispersion in the living world may be achieved scientifically. The rule of parsimony Another method of gaining information about the evolution of extinct taxa is the use of the principle of parsimony which demands that no more causes or forces should be assumed than are necessary to account for the facts. The method involves the construction of a model of the extinct organism based on known or suspected descendants. Common traits displayed in each of these descendants are then assigned to the model progenitor. Successful use of the principle demands that the trait considered is not expressed as a result of lateral information transfer or convergent evolution. It must represent vertical inheritance and originate in the organism studied (rather than in symbionts or being accumulated from food etc.). Presumably this principle can be used on the pattern of metabolites from contemporary organisms to predict the metabolic set-up of extinct and extant progenitors.

Appendix 11 Taxonomic dispersion of unique structures A vexing problem in the dispersion of specific secondary metabolites between taxa is the appearance of unique compounds in very distantly related organisms.

Scheme 15. Structure of tetrodotoxin, the cause of human poisonings from ingestion of tetrodetoxic fishes "Fugu".

An example is the alkaloid tetrodotoxin which has been detected in several species of puffer fish (family Tetraodontidae), eggs of the Califomian newt, Taricha torosa, a goby (Gobius criniger), an octopus (Hapalochlaena maculosa), a frog (Atelopus chiriquiensis), five marine gastropods, namely, an ivory shell {Babylonia japonica), a trumpet shell (Charonia sauliae), Tutufa lissostoma, Zeuxis siquizorensis, and the lined moon shell Natica lineata (202), an arrow worm (phylum Chaeto-

725

gnatha), a starfish (Astropecten polyacanthus), ribbon worms (Cephalothrix linearis, Lineus fuscoviridis), a nemertean (Tubulanus punctatus, 203) and several other systematically unrelated groups. Tetrodotoxin synthesis was recently traced to bacteria of the genera Listonella {Vibrio), Alteromonas and Shewanella (31, 204). Even if the presence of this potent toxin in all these diverse taxons is the result of bacterial symbiosis, the genes coding for the biosynthetic pathway leading to tetrodotoxin production must be quite widespread in procaryotes. Owing to the absence of biosynthetic studies, the hypothesis that tetrodotoxin diversity is a result of convergent evolution cannot be completely ruled out but the hypothesis seems far fetched. Other comparable examples are known. The main alkaloid in a Nova Scotia collection of Flustra foliacea, dihydroflustramine C, is related to pseudophrynaminol from the skin of the myobatrachid burrowing frog Pseudophryne coriacea. Both classes of alkaloids are based on the 3a-prenyl pyrrolo[2,3-fe]indole skeleton (31).

OMe

R^R

R = H , R' = OH R = H , R' = OMe R = O H . R ' = OMe R = R' = OMe

Scheme 16. Pseudophrynamines from frogs of the genus Pseudophryne.

726

The pseudophrynamines are only known from seven species of the genus Pseudophryne in the family Myobatrachidae (205). However, almost all Pseudophryne species investigated also contained the well known dendrobatid alkaloids pumilotoxins in some populations they could not be detected. Pseudophrynamines were the dominant alkaloids in two species {P. guentheri and P. occidentalis) from Western Australia, while all five eastern species contained significant amounts of pumilotoxins as well. It is interesting to note that the pseudophrynamines vary like the bryozoan alkaloids and that they were the main alkaloids only in some populations of the same species. The debromoanalog of 6-bromo-A^jj-methyl-N^j-formyltryptamine (40) has been identified from the bark of the hallucinogenic plant Virola sebifera (Myristicaceae) (206).

R= -Me R=-Et

L: R =-CH =CH2 5 : R = -COMe

R=-CHOHMe

Scheme 17. Harnian [1], 1-ethyl-P-carboline [2], 5-l-(r-hydroxyethyl)-P-carboline [3], andpavettine [4] from the bryozoan Costaticella hastata. 1-Acetyl-P-carboline [5] are known from Arenaria kansuensis.

The p-carboline alkaloids harman [1] and pavettine [4] from the bryozoan Costaticella hastata (207) are known from numerous terrestrial plants. Still another bryozoan compound (S)-l-(r-hydroxyethyl)-p-carboline [3] is new but closely related to 1-ethyl-P-carboline [2] present in the bryozoan. It is also known from the roots of Hannoa klaineana (Simaroubaceae) (208) and 1-acetyl-p-carboline [5] from Arenaria kansuensis (Caryophyllaceae) (209). Incidentally, pavettine [4] was detected in only one collection of the bryozoan. In these examples, and many others not addressed here, nothing is known about the biosynthesis of the secondary metabolites. Accordingly it can not a priori be refuted that the biosynthetic pathways may differ and hence the chemical resemblances of the structures are only coincidental. Neither can it be excluded that the compounds are microbial metabolites. However, especially in the cases of the terrestrial plants mentioned above, this explanation seems unlikely. But even if the latter possibility is taken into account we are left with the need to explain how presumably

727

distantly related microorganisms have been selected to express nearly identical compounds. If convergent evolution is disregarded the genes coding for the enzymes necessary to effect these pathways must either be very old or very efficiently dispersed by lateral information transfer. In this connection it may be mentioned that migration of genetic information from mitochondria to the nucleus in Saccharomyces cerevisiae has been found to be a high frequency event (around 2x10'^ per cell per generation), while transfer in the opposite direction is at least 100,000 times less frequent (210). In addition other processes could lead to analogous results. In the major class of red alga, Floridophyceae, more than 15% of the genera are parasitic on other red algae, often closely related. Host specificity is in some cases so narrow that the parasite discriminates between different populations of the same species. In the case of the parasite Choreocolax and the distantly related host Polysiphonia transfer of nuclei from parasite to host has been demonstrated (211). This transfer confers new morphology and histology to the host cell attesting to the activity of the parasite genetic material. There is a possibility of an additional explanation for this wide taxonomic distribution of unique structures. In the case of tetrodotoxin the taxonomic diversity is remarkable ranging from highly developed teleosts through most of the animal kingdom to procaryotic bacteria. It is possible that the capacity for tetrodotoxin production is an extraordinarily ancient trait. As all the eucaryotic organisms in which tetrodotoxin has been detected have mitochondria, it cannot be ruled out that the genetic machinery for synthesis of this molecule originates from the procaryotic organisms that entered into symbiosis with an ancient eucaryotic organism many million years ago. Thus it is conceivable that tetrodotoxin is produced by the mitochondria or that the genes involved have been transferred to the host nucleus and now function as an integral part of the host genome. Obviously some of the contemporary procaryotes still retain this ancient genetic machinery. Tetrodotoxin has been detected in a variety of marine bacteria and even in Escherichia coli (212). This explanation may also account for the other examples discussed above. To further resolve considerations of this type details about the enzymology connected with secondary metabolite synthesis are essential. 9.

ACKNOWLEDGMENTS

I am indepted to professor P.J. Scheuer for introducing me to marine natural product chemistry (213-214), Dra. R. Encamacion for introducing me to traditional medicine (112, 215), Dr. U.W. Smitt to biologically active products from higher plants (216-217), Dr. S. Wium-Andersen to freshwater plants (218), Dr. O.S. Tendal to sponge systematics (102). Professor M. G. Ettlinger has contributed significantly to the final version of this paper by way of discussions and a very careful scrutiny of the whole manuscript. However most of all I am indebted to my colleagues Drs. U. Anthoni and P. H. Nielsen for their willingness to consider and reconsider the

728

reasoning resulting from my preoccupation with subjects with seemingly no relation to organic or natural products chemistry or to chemistry at all. Last but not least I am grateful to Ms Bente Karberg for idiomatic corrections and for preparing the manuscript for printing. 10.

REFERENCES

1

U. Anthoni, P.H. Nielsen, L. Smith-Hansen, S. Wium-Andersen, C. Christophersen, J. Org. Chem., 52 (1987) 694-695. M. Gajhede, U. Anthoni, C. Christophersen, P.H. Nielsen, Acta Cryst., B45 (1989) 562-566. U. Anthoni, C. Christophersen, P.H. Nielsen, D.H. Christensen, O.F. Nielsen, M. Gajhede, Spectrochim. Acta, 45A (1989) 1157-1164. U. Anthoni, D.H. Christensen, C. Christophersen, M. Gajhede, L. Henriksen, O.F. Nielsen, P.H. Nielsen, J. Molecular Structure, 220 (1990) 43-54. K. Hagen, H.V. Volden, U. Anthoni, C. Christophersen, M. Gajhede, P.H. Nielsen, J. Phys. Chem., 95 (1991) 1597-1600. U. Anthoni, T. Borresen, C. Christophersen, L. Gram, P.H. Nielsen, Comp. Biochem. Physiol., 97B (1990) 569-571. U. Anthoni, C. Christophersen, M. Gajhede, P.H. Nielsen, Structural Chemistry, 3 (1992) 121-128. A. Guerriero, M. D'Ambrosio, P. Traldi, F. Pietra, Naturwissenschaften, 71 (1984) 425-426. M. D'Ambrosio, A. Guerriero, F. Pietra, Helv. Chim. Acta, 67 (1984) 14841493. K.-A. Karlson and B.E. Samuelson, Biochem. Biophys. Acta, 337 (1974) 204213. H.B. Mortensen and C. Christophersen, Biochem. Biophys. Acta, 707 (1982) 154-163. H.B. Mortensen and C. Christophersen, Clin. Chim. Acta, 134 (1983) 317326. H.B. Mortensen and C. Christophersen, Diabete & Metabolisme (Paris), 9 (1983) 232-234. H.B. Mortensen, Aa. Volund, C. Christophersen, Clin. Chim. Acta, 136 (1984) 75-81. C. Christophersen, H.B. Mortensen, Acta Paediatr. Scand., 73 (1984) 855-856. P. Keil, H.B. Mortensen, C. Christophersen, Acta Chem. Scand., B39 (1985) 191-193. E. Haslam, Nat. Prod. Rep., 3 (1986) 217-249. D.H. WiUiams, M.J. Stone, P.R. Hauck, S.K. Rahman, J. Nat. Prod., 52 (1989) 1189-1208. C. Mitter and J.A. Klun, J. Chem. EcoL, 13 (1987) 1823-1831.

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

729

20 21 22 23 24 25 26 27 28 29 30 31 32

33 34 35 36 37 38 39 40 41 42 43 44 45

J.B. Harbome, Nat. Prod. Rep., 6 (1989) 85-109. R. Nishida, T. Ohsugi, S. Kokubo, H. Fukami, Experientia, 43 (1987) 342344. C. Christophersen, Comp. Biochem. Physiol., 98B (1991) 427-432. U.W. Smitt, C. Comett, A. Andersen, S.B. Christensen, P. Avato, J. Nat. Prod., 53 (1990) 1479-1484. H. Fliick, in: T. Swain (Ed.), Chemical Plant Taxonomy, Academic Press, London, 1963, pp. 167-186. I.S. Homsey and D. Hide, Br. phycol. J., 11 (1976) 175-181. D.W. Phillips and G.H.N. Towers, J. Exp. Mar. Biol. EcoL, 58 (1982) 285293. S. Ghosal, S.K. Singh, S.G. Unnikrishnan, Phytochemistry, 29 (1989) 805-811. I.S. Homsey and D. Hide, Br. Phycol. J., 11 (1976) 63-67. V.J. Paul and W. Fenical, in: P.J. Scheuer (Ed.), Bioorganic Marine Chemistry 1, Springer-Verlag, Berlin Heidelberg, 1987, pp. 1-29. P.B. Hoist, U. Anthoni, K. Bock, C. Christophersen, P.H. Nielsen, Acta Chem. Scand., 48 (1994) 765-768. U. Anthoni, P.H. Nielsen, M. Pereira, C. Christophersen, Comp. Biochem. Physiol, 96B (1990) 431-437. C. Christophersen in: P.J. Scheuer (Ed.), Marine Natural Products. Chemical and Biological Perspectives Vol. 5. Chapter 5, Academic Press, New York, 1983, pp. 259-285. C. Christophersen, Acta Chem. Scand., B39 (1985) 517-529. C. Christophersen in: A. Brossi (Ed.), The Alkaloids. Chemistry and Physiology, Vol. 24, Chapter 2, Academic Press, New York, 1985, pp. 25-111. P. Keil, E.G. Nielsen, U. Anthoni, C. Christophersen, Acta Chem. Scand., B40 (1986) 555-559. P.B. Hoist, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 57 (1994) 997-1000. J.S. Carle, C. Christophersen, J. Am. Chem. Soc, 101 (1979) 4012-4013. J.S. Carle, C. Christophersen, J. Org. Chem., 45 (1980) 1586-1589. J.S. Carle, C. Christophersen, J. Org. Chem., 46 (1981) 3440-3443. P. Wulff, J.S. Carle, C. Christophersen, Comp. Biochem. Physiol., 7IB (1982) 523-524. P. Wulff, J.S. Carle, C. Christophersen, J. Chem. Soc. Perkin Trans. I (1981) 2895-2898. P. Wulff, J.S. Carle, C. Christophersen, Comp. Biochem. Physiol, 7 IB (1982) 525-526. J.L.C. Wright, J. Nat. Prod., 47 (1984) 893-895. M.V. Laycock, L.C. Wright, J.A. Findlay, A.D. Patil, Can. J. Chem., 64 (1986) 1312-1316. P.B. Hoist, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 57

730

46 47 48 49 50 51 52 53 54 55 56

57 58 59 60 61 62

63 64 65 66 67 68 69 70

(1994) 1310-1312. Q.-S. Yu, HJ.C. Yeh, A. Brossi, J. Nat. Prod., 52 (1989) 332-336. T. Sjoblom, L. Bohlin, C. Christophersen, Acta Pharm. Suec, 20 (1983) 415419. L. Andersson, G. Lidgren, L. Bohlin, L. Magni, S. Ogren, L. Afzelius, Acta Pharm. Suec, 20 (1983) 401-414. C. Christophersen, J.S. Carle, Naturwissenschaften, 65 (1978) 440. L. Chevolot, A.-M. Chevolot, M. Gajhede, C. Larsen, U. Anthoni, C. Christophersen, J. Am. Chem. Soc., 107 (1985) 4542-4543. U. Anthoni, L. Chevolot, C. Larsen, P.H. Nielsen, C. Christophersen, J. Org. Chem., 52 (1987) 4709-4712. P.H. Nielsen, U. Anthoni, C. Christophersen, Acta Chem. Scand., B42 (1988) 489-491. U. Anthoni, K. Bock, L. Chevolot, C. Larsen, P.H. Nielsen, C. Christophersen, J. Org. Chem., 52 (1987) 5638-39. G. Guella, A. Guerriero, L Mancini, F. Pietra, unpublished results mentioned in ref. 55. F. Pietra, Gazz. Chim. Ital., 115 (1985) 443-485. E. Bignetti, A. Cavaggioni, R. Tirindelli in: A. Borsellino, L. Cervetto, V. Torre (Eds.), Sensory Transduction, Plenum Press, New York, 1990, pp. 6780. M. Gajhede, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Cryst. Spectrosc. Res., 20 (1990) 165-171. U. Anthoni, C. Larsen, P.H. Nielsen, C. Christophersen, Biochem. Syst. Ecol., 18 (1990) 377-379. U. Anthoni, P.H. Nielsen, C. Christophersen, Dansk Vet. Tidsskr., 73 (1990) 495-499. K.-E. Kaissling in: A. Borsellino, L. Cervetto, V. Torre (Eds.), Sensory Transduction, Plenum Press, New York, 1990, pp. 81-97. L.A. Matsuda, S.J. Lolait, M.J. Brownstein, A.C. Young, T.L Bonner, Nature, 346 (1990) 561-564. W.S. Agnew, T. Claudio, F.J. Sigworth, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. xiii-xvii. E.M. Kosower, Biochem. Biophys. Res. Commun., I l l (1983) 1022-1026. E.M. Kosower, FEBS Letters, 155 (1983) 245-247. J.R. Smythies, G. Kemp, J. Receptor Res., 9 (1989) 199-201. J.S. Carie, C. Christophersen, J. Am. Chem. Soc., 102 (1980) 5107. J.S. Carie, C. Christophersen, Bull. Soc. Chim. Belg., 89 (1980) 1087-1091. J.S. Carie, C. Christophersen, Toxicon, 20 (1982) 307-310. J.S. Carie, H. Thybo, C. Christophersen, Contact Dermatiris, 8 (1982) 43-47. T. Rasmussen, C. Christophersen, P.H. Nielsen, R. Rajagopal, J. Mar.

731

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86

87 88 89 90

91

Biotech., (in press). T. Rasmussen, J. Jensen, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 56 (1993). 1553-1558. M.H.G. Munro, R.T. Luibrand, J.W. Blunt, in: PJ. Scheuer (Ed.), Bioorganic Marine Chemistry 1, Springer Verlag, Berlin Heidelberg, 1987, pp. 93-176. A. Gescher and I.L. Dale, Anti-Cancer Drug Design, 4 (1989) 93-105. R.L. Huganir, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 147-164. G.R. Pettit, D.L. Herald, F. Gao, D. Sengupta, C.L. Herald, J. Org. Chem., 56 (1991) 1337-1340. J.S. Ramsdell, G.R. Pettit, A.H. Tashjian, Jr., J. Biol. Chem., 261 (1986) 17073-17080. B.S. Warren, Y. Kamano, G.R. Pettit, P.M. Blumberg, Cancer Research, 48 (1988) 5984-5988. W.S. May, S.J. Sharkis, A.H. Esa, V. Gebbia, A.S. Kraft, G.R. Pettit, L.L. Sensenbrenner, Proc. Natl. Acad. Sci. USA, 84 (1987) 8483-8487. H.G. Drexler, S.M. Gignac, R.A. Jones, C.S. Scott, G.R. Pettit, A.V. Hoffbrand, Blood, 74 (1989) 1747-1757. H.G. Drexler, S.M. Gignac, G.R. Pettit, A.V. Hoffbrand, Eur. J. Immunol., 20 (1990) 119-127. H. Hennings, P.M. Blumberg, G.R. Pettit, C.L. Herald, R. Shores, S.H. Yuspa, Carcinogenesis, 8 (1987) 1343-1346. A.D. Hess, M.K. Silanskis, A.H. Esa, G.R. Pettit, W.Stratfor May, J. Immunology, 141 (1988) 3263-3269. A.S. Kraft, F. William, G.R. Pettit, M.B. Lilly, Cancer Research, 49 (1989) 1287-1293. H. Mohr, G.R. Pettit, A. Plessing-Menze, Immunobiol., 175 (1987) 420-430. G. Trenn, G.R. Pettit, H. Takayama, J. Hu-Li, M.V. Sitkovsky, J. Immunology, 140 (1988) 433-439. P. Taylor, P. Culver, S. Abramson, L. Wasserman, T. Kline, W. Fenical, Biomedical Importance of Marine Organisms, Fautin, D. G., Califomia Academy of Sciences, 1988, pp. 109-114. U. Anthoni, L. Bohlin, C. Larsen, P. Nielsen, N.H. Nielsen, C. Christophersen, Toxicon, 27 (1989) 717-723. U. Anthoni, L. Bohlin, C. Larsen, P. Nielsen, N.H. Nielsen, C. Christophersen, Toxicon, 27 (1989) 707-716. U. Anthoni, C. Christophersen, P.H. Nielsen, J. Agric. Food Chem., 37 (1989) 705-707. B. Witkop and A. Brossi in: P. Krogsgaard-Larsen, S.B. Christensen, H. Kofod, (Eds.), Natural Products and Drug Development, Munksgaard, Copenhagen, 1984, pp. 283-300. U. Anthoni, C. Christophersen, J. 0gard Madsen, S. Wium-Andersen, N,

732

92 93

94 95

96 97 98 99

100

101 102 103 104 105

106 107 108 109 110 111 112 113 114

Jacobsen, Phytochemistry, 19 (1980) 1228-1229. C. Christophersen, Arch. Pharm. Chem., 94 (1987) 383-395. C. Christophersen, C.T. Pedersen, J. Becher (Eds.), Developments in the Organic Chemistry of Sulfur, Gordon and Breach, Science Publishers, New York, 1989, pp. 155-163. C. Christophersen, Phosphorus, Sulfur and Silica, 43 (1989) 155-163. M.E. Eldefrawi, I.M. Abalis, S.M. Sherby, A.T. Eldefrawi in: M.G. Ford, G.G. Lunt, R.C. Reay, P.N.R. Usherwood (Eds.), Neuropharmacology and Pesticide Action, Verlag Chemie and EUis Horwood, 1986, pp. 154-173. N. Jacobsen and L.-E.K. Pedersen, Pestic. Sci., 14 (1983) 90-97. S. Wium-Andersen, U. Anthoni, C. Christophersen, G. Houen, Oikos, 39 (1982) 187-190. J.F. Martin and A. Demain, Microbiol. Rew., 44 (1980) 230-251. A.S. Khokhlov, I.T. Tovarova, L.N. Borisova, S.A. Pliner, L.A. Shevchenko, E.Y. Komitskaya, N.S. Ivkina, LA. Rapoport, Dokl. Acad. Nauk SSSR, 177 (1967) 232-235. I.S. Hunter and S. Baumberg, in: S. Baumberg, I. Hunter, M. Rhodes (Eds.), Microbial Products: New Approaches, Cambridge University Press, Cambridge, 1989, pp. 121-162. P.R. Bergquist, W. Hofheinz, G. Oesterhelt, Biochem. Syst. Ecol., 8 (1980) 423-435. C. Christophersen, P.G. Nielsen, P.H. Nielsen, O.S. Tendal, Comp. Biochem. Physiol., to be pubHshed. S. Berking, Roux's Arch. Dev. Biol., 195 (1986) 33-38. S. Berking, Development, 99 (1987) 211-220. P.R. Bergquist, R.J. Wells in: P.J. Scheuer (Ed), Marine Natural Products, Chemical and Biological Perspectives., Academic Press, New York, 1983, vol. V, pp. 1-50. D.M. Anderson, D.M. Kulis, J.J. Sullivan, S. Hall, Toxicon, 28 (1990) 885893. H. Ravn, C.U. Schmidt, H. Sten, U. Anthoni, C. Christophersen, P.H. Nielsen, Comp. Biochem. Physiol., (1995), in press. M. Kodama, T. Ogata, S. Sakamoto, S. Sato, T. Honda, T. Miwatani, Toxicon, 28 (1990) 707-714. U. Anthoni, C. Christophersen, N. Jacobsen, A. Svendsen, Tetrahedron, 38 (1982) 2425-2427. L. Teuber and C, Christophersen, Acta Chem. Scand., B42 (1988) 629-634. L. Teuber. and C. Christophersen, Acta Chem. Scand., B42 (1988) 620-622. C. Christophersen, C, Larsen, R.D. Encamacion, The H. C. 0rsted Institute, Copenhagen, 1991, 1-41. C. Christophersen, C. Larsen, R.D. Encamacion, Danida Report (1989) 1-43. M. Gajhede, R.D. Encamacion, G.C. Leal, J.C. Patino, C. Christophersen, P.H.

733

115 116 117 118 119 120 121 122 123

124 125 126 127 128 129 130 131 132

133 134 135 136 137 138 139 140

Nielsen, Acta Cryst., C45 (1989) 2012-2014. R.D. Encamacion, S.G. Keer, P.H. Nielsen, C. Christophersen, J. Ethnopharmacology, 31 (1991) 43-48. R.D. Encamacion, N.A. Ochoa, U. Anthoni, C. Christophersen, P.H. Nielsen, J. Nat. Prod., 57 (1994) 1307-1309. S.E. Nielsen, U. Anthoni, C. Christophersen, C. Comett, Phytochemistry (1995), in press. G. Green, Marine Biology, 40 (1977) 207-215. P.R. Bergquist, J.J. Bedford, Marine Biology, 46 (1978) 215-221. S.J. Giovannoni, T.B. Britschgi, C.L. Moyer, K.G. Field, Nature, 345 (1990) 60-65. D.M. Ward, R. Weller, M.M. Bateson, Nature, 345 (1990) 63-65. C.A. Suttle, A.M. Chan, M.T. Cottrell, Nature, 347 (1990) 467-469. C.A. West, in: D.T. Dennis and D.H. Turpin (Eds.), Plant Physiology, Biochemistry and Molecular Biology. Longman Scientific and Technical, England, 1990, p. 353-369. M. Wink, J. Exp. Botany, 44, Supplement, (1993) 231-246. K.C. Engvild, Physiologia Plantarum, 77 (1989) 282-285. J. Theodor, Nature, 227 (1970) 690-692. O. Von Fiirth, Vergleichende chemische Physiologic der niederen Tiere Verlag von Gustav Fischer, Jena, 1903, pp. 453-457. J.B. Harbome, Introduction to Ecological Biochemistry, Third Edition, Academic Press, New York, 1988. M. Aubert, Mar. Poll. Bull, 21 (1990) 24-29. F.E. Round, The Ecology of Algae, Cambridge University Press, Cambridge, (1981). D.J. Futuyma, Evolutionary Biology, Second Edition, Sinauer Associates, Inc., Sunderland, Massachussetts, 1986. H. VanEtten, D. Matthews, P. Matthews, V. Miao, A. Maloney, D. Straney in: B.J.J. Lugtenberg (Ed.), Signal Molecules in Plants and Plant-Microbe Interactions, NATO ASI Series H: Cell Biology Vol. 36. Springer-Veriag, Beriin, 1989. S. Adhya and S. Garges, J. Biol. Chem., 265 (1990) 10797-10800. H.J. Zeringue Jr. and S.P. McCormick, Toxicon, 28 (1990) 445-448. H.J. Zeringue Jr., Phytochemistry, 29 (1990) 1789-1791. R. Baute, M.-A. Baute, G. Deffieux, M.-J. Filleau, Phytochemistry, 15 (1976) 1753-1755. M. Piattelli, New J. Chem., 14 (1990) 777-782. G. Combaut, L. Codomier, J. Teste, Phytochemistry, 20 (1981) 2036-2037. J.F. Biard, J.F., Verbiest, Y. Letoumeux, R. Floch, Planta Med., 40 (1980) 288-294. J.F. Biard, J.F. Verbiest, R. Floch, Y. Letoumeux, Tetrahedron Letters, 21

734

141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167

(1980) 1849-1852. G. Combaut, L. Piovetti, Phytochemistry, 22 (1983) 1787-1789. L. Hougaard, U. Anthoni, C. Christophersen, P.H. Nielsen, Phytochemistry, 30 (1991) 3049-3051. L. Hougaard, U. Anthoni, C. Christophersen, P.H. Nielsen, Tetrahedron Letters, 32 (1991) 3577-3578. R. Vails, B. Banaigs, C. Francisco, L. Codomier, A. Cave, Phytochemistry, 25 (1986) 751-752. L. Semmak, A. Zerzouf, R. Vails, B. Banaigs, G. Jeanty, C. Francisco, Phytochemistry, 27 (1988) 2347-2349. A. San-Martin and J. Rovirosa, Biochem. Syst. EcoL, 14 (1986) 459-461. M.S. Doty, G. Aguilar-Santos, Nature, 211 (1966) 990. M.S. Doty, G. Aguilar-Santos, Pac. Sci., 24 (1970) 351-355. P.G. Nielsen, J.S. Carle, C. Christophersen, Phytochemistry, 21 (1982) 16431645. M. Mahendran, S. Somasundaram, R.H. Thomson, Phytochemistry, 18 (1979) 1885-1886. O.R. Gottlieb, Micromolecular Evolution, Systematic and Ecology, An Essay into a Novel Botanical Discipline, Springer-Verlag, Berlin, 1982. A.J. Blackman, R.D. Green, Aust. J. Chem., 40 (1987) 1655-1662. R. Clauser, A. De Guerriero, F. Pietra, impublished results mentioned in ref. 55. M. Suffness, D.N. Newman, K. Snader, in: P.J. Scheuer (Ed.), Bioorganic Marine Chemistry 3, Springer-Verlag, Berlin Heidelberg, 1989, pp. 131-168. L. Gunthorpe and A.M. Cameron, Toxicon, 28 (1990) 1199-1219. L. Gunthorpe and A.M. Cameron, Toxicon, 28 (1990) 1221-1227. J. W. Blunt, M.H.G. Munroe, C.N. Battershill, B.R. Coop, J.D. Combs, N.B. Perry, M. Prinsep, A.M. Thompson, New J. Chem., 14 (1990) 761-775. M.R. Keman, D.J. Faulkner, R.S. Jacobs, J. Org. Chem., 52 (1987) 3081-3083. E.D. de Silva, P.J. Scheuer, Tetrahedron Letters, 21 (1980) 1611-1614. E.D. de Silva, P.J. Scheuer, Tetrahedron Letters, 22 (1981) 3147-3150. R.C Cambie, P.A. Craw, P.R. Bergquist, P. Karuso, J. Nat. Prod., 51 (1988) 331-334. M.R. Keman, R.C. Cambie, P.R. Bergquist, J. Nat. Prod., 53 (1990) 724-727. M.R. Keman, R.C. Cambie, P.R. Bergquist, J. Nat. Prod., 53 (1990) 13531356. R.J. Wells, Tetrahedron Letters, (1976) 2637-2638. T. Murayama, Y. Ohizumi, H. Nakamura, T. Sasaki, J. Kobayashi, Experientia, 45 (1989) 898-899. S. Sakemi, T. Higa, U. Anthoni, C. Christophersen, Tetrahedron, 43 (1987) 263-268. T. Higa, C. Christophersen, S. Sakemi, U. S. Patent 7, 731, (1988) 377, Mar.

735

168 169 170 171 172 173 174 175 176 177 178 179 180 181 182

183

184 185 186

187

188

15, 1988. F.S. De Guzman, F.J. Schmitz, J. Nat. Prod., 53 (1990) 926-931. E. Quinoa, E. Kho, L.V. Manes, P. Crews, G.J. Bakus, J. Org. Chem., 51 (1986) 4260-4264. S.P. Gunasekera, M. Gunasekera, G.P. Gunawardana, P. McCarthy, N. Burres, J. Nat. Prod., 53 (1990) 669-674. C. Christophersen, U. Anthoni, P.H. Nielsen, N. Jacobsen, O. Tendal, Biochem. Syst. EcoL, 17 (1989) 459-461. C. Christophersen, N. Jacobsen, R. S. C. Annual Reports, B (1979) 433-447. D.S. Latchman, Biochem. J., 270 (1990) 281-289. J. Darnell, H. Lodish, D. Baltimore, Molecular Cell Biology. Second Edition, Scientific American Books, W. H. Freeman and Company, New York, 1990. C.C. Kao, P.M. Lieberman, M.C. Schmidt, Q. Zhou, R. Pei, A.J. Berk, Science, 248 (1990) 1646-1650. M.G. Peterson, N. Tanese, B.F. Pugh, R. Tijan, Science, 248 (1990) 16251630. A. Hoffmann, E. Sinn, T. Yamamoto, J. Wang, A. Roy, M. Horikoshi, R.G. Roeder, Nature, 346 (1990) 387-390. B. Alberts, D. Bray, J. Lewis, M. Raff, K. Roberts, J.D. Watson, Molecular Biology of the Cell. Second Edition, Garland Publ., Inc., New York, 1990. J.D. Fikes, D.M. Becker, F. Winston, L. Guarente, Nature, 346 (1990) 291294. A. Gasch, A. Hoffmann, M. Horikoshi, R.G. Roeder, N.-H. Chua, Nature, 346 (1990) 390-394. M. Ptashne and A.A.F. Gann, Nature, 346 (1990) 329-331. E. Hawrot, K.L. Colson, T.L. Lenz, P.T. Wilson in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.) Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 165-196. S.M. Sine and J.H. Steinbach, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 134-146. T.L. Lentz and P.T. Wilson, Intemat. Rev. Neurobiol., 29 (1988) 117-160. H. Breer, D.B. Sattelle, J. Insect Physiol., 33 (1987) 771-790. Y. Saimi and C. Kung, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 1-14. W.S. Agnew, E.C. Cooper, W.M. James, S.A. Tomiko, R.L. Rosenberg, M.C. Emerick, A.M. Correa, J.Y. Zhou, in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 329-369. E. Hartmann and R. Gupta in: W.F. Boss, D.J. Morre (Eds.), Plant Biology 6, Second Messengers in Plant Growth and Development, Alan R. Liss, Inc.,

736

189

190 191 192 193 194

195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212

New York, 1989, pp. 257-287. A. Tretyn, M.E. Bossen, R.E. Kendrick, in: CM. Kassen, L.C. Loon, D. van Vreugdenhil (Eds.), Progress in Plant Growth Regulation 306, Kluwer Academic Publishers. The Netherlands, (1992). G.A. Miura and T.M. Shih, Physiol. Plant., 61 (1984) 417-421. V.V. Roshchina, Photosynthetica, 23 (1989) 197-206. A. Tretyn, J. Kopcewicz, E. Slesak, Biol. Plant., 30 (1988) 338-342. H. Betz, Biochemistry, 29 (1990) 3591-3599. L.J. Cruz, D.S. Johnson, J.S. Imperial, D. Griffin, G.W. LeCheminant, G.P. Miljanich, B.M. Olivera in: W.S. Agnew, T. Claudio, F.J. Sigworth (Eds.), Molecular Biology of Ionic Channels, Academic Press, San Diego, 1988, pp. 417-430. S.A. Benner, A.D. Ellington in: H. Dugas (Ed.), Bioorganic Chemistry Frontiers, 1, Springer-Verlag, Berlin, 1990, pp. 3-70. S.A. Benner, A. Glasfeld, J.A. Piccirilli, Topics in Stereochemistry, 19 (1989) 127-207. B. Nagy, Naturwissenschaften, 69 (1982) 301-310. A.S. Mackenzie, S.C. Brassell, G. Eglinton, J.R. Maxwell, Science, 217 (1982) 491-504. J.M. Hunt, Petroleum Geochemistry and Geology. W. H. Freeman and Company, San Francisco, 1979. B.P. Tissot and D.H. Welte, Petroleum Formation and Occurrence, A New Approach to Oil and Gas Exploration. Springer-Verlag, Berlin, 1978. C. Christophersen, U. Anthoni, Sulfur Reports, 4 (1986) 365-456. D.F. Hwang, C.H. Chueh, S.S. Jeng, Toxicon, 28 (1990) 1133-1136. A.E. Ali, O. Arakawa, T, Noguchi, K. Miyazawa, Y. Shida, K. Hashimoto, Toxicon, 28 (1990) 1083-1093. U. Simidu, K. Kita-Tsukamoto, T. Yasumoto, M. Yotsu, Intemat. J. Syst. Bacteriol., 40 (1990) 331-336. J. W. Daly, H.M. Garraffo, L.K. Pannell, T.F. Spande, C. Severini, V. Erspamer, J. Nat. Prod., 53 (1990) 407-421. K. Kawanishi, Y. Uhara, Y. Hashimoto, Phytochemistry, 24 (1985) 1373-1375. A.J. Blackman, D. Matthews, C.K. Narkowicz, J. Nat. Prod., 50 (1987) 494496. L. Lumonadio and M. Vanhaelen, Phytochemistry, 23 (1984) 453-455. F.-E. Wu, K. Koike, T. Nikaido, Y. Sakamoto, T. Ohmoto, K. Ikeda, Chem. Pharm. Bull., 37 (1989) 1808-1809. P.E. Thorsness and T.D. Fox, Nature, 346 (1990) 376-379. L.J. Goff and A.W. Coleman, Proc. Natl. Acad. Sci. U. S. A., 81 (1984) 54205424. M.L. Tamplin, in: S. Hall and G. Strichartz (Eds.), Marine Toxins, Origin, Structure, and Molecular Pharmacology, American Chemical Society,

737

213 214 215 216 217 218

Washington, D. C, 1990, pp. 78-86. B.J. Burreson, C. Christophersen, P.J. Scheuer, J. Am. Chem. Soc., 97 (1975) 201-202. B.J. Burreson, C. Christophersen, P.J. Scheuer, Tetrahedron, 31 (1975) 20152018. R.D. Encamacion, G. Carrasco, M. Espinoza, U. Anthoni, P.H. Nielsen, C. Christophersen, J. Nat. Prod., 52 (1989) 248-251. S.B. Christensen, U. Rasmussen, C. Christophersen, Tetrahedron Letters, 21 (1980) 3829-3830. S.B. Christensen, I.K. Larsen, U. Rasmussen, C. Christophersen, J. Org. Chem., 47 (1982) 649-652. S. Wium-Andersen, K.H. Jorgensen, C. Christophersen, U. Anthoni, Arch. Hydrobiol., I l l (1987) 317-320.



739

The Celastraceae from Latin America Chemistry and Biological Activity

O. Muhoz, A. Penaloza, A.G. Gonzalez, A.G. Ravelo, LL. Bazzocchi and N X . Alvarenga 'Universidad de Chile, Facultad de Ciencias, Casilla 653-Santiago, Chile. C.P.N.O. Antonio Gonzalez, Universidad de La Laguna, Carretera La Esperanza 2, La LagunaTenerife. Espafia.

1. INTRODUCTION The CELASTRACEAE family was last reviewed in 1978 (Brtining & Wagner) [1] and since then a great deal of new chemical and phannacological information has accumulated. Celastraceae species have a long tradition of use in medicine and folk agriculture, especially in Asia and Latin America but also in other continents and undoubtedly took on a new lease of life in the seventies when the MAYTANSINOIDS, compounds with exceptional antitumoral properties, were discovered [I]. Nonetheless, the maytansinoids have not been made into a useful drug form as they cause serious gastro-intestinal damage when applied to rats [2]. For some time now several research laboratories have been intensively researching this fiamily, inspired by its broad and varied botanical distribution, the interesting chemical nature of its secondary metabolites, the complexity of the biogenetic processes which produce them, and most of all by the different types of pharmacological action displayed by preparations of its constituents. In Latin America (Mexico, Central America, the Caribbean and South America) this study is particularly momentous due in tfie main to a socio-economic and cultural climate which has not in the past lent itself either to sound development or the rational exploitation of the resources of the various countries involved. In the course of these research programmes, cytotoxic quinones, polyester sesquiterj)enes and pyridine-sesquiterpene alkaloids with antifeedant and/or insecticidal properties have been isolated from Latin American species, m particular those of the Maytenus genus which is extensively used by rural communities and tribes in the Andes and the Amazon basin. Recently some sesquiterpene alkaloids with immunosuppressive activity and sesquiterpenes with antitumoral activity have also been described [3].

740 As a general rule, the biosynthesis of skeletons belonging to the Celastraceae family is extremely specific, the triterpene-quinones and P-dihydro-agaroftiran type skeleton sesquiterpenes from these species having a notably high degree of oxidation. The presence of triterpene-quinones indicates the biosynthetic specificity' of the Celastraceae family since these compounds are synthesized in the roots and are virtually exclusive to the family. The next few pages are an update on the state of the Celastraceae family in Latin America, detailing the different chemical structures found and the results of the studies of biological activity carried out since the last publication on the subject [1].

2. THE SYSTEMATICS OF THE CELASTRACEAE FAMILY AND THE LATIN AMERICAN GENERA

THE CELASTRACEAE FAMILY

The Celastraceae family consists of about 55 genera and 850 species. According to Takhtajan [4], Hippocrateaceae would be subordinate to the Celastraceae. Systematically the Celastraceae family is arranged hierarchically as follows: Division: Spermatophyta; Subdivision: Magnoliophyta (Angiosperms); Class: Magnolitae (Dicotyledons); Order: Celastrales; Family: Celastraceae The Celastraceae family is probably related phylogenetically with the Aquifoliaceae; the presence of glandular discs around the ovary and the bright coloured aril in the Celastraceae are the principal differences between the two families. The Celastraceae family is pantropically distributed with radiation towards temperate or temperate-cold climates. In other words, the Celastraceae are principally concentrated in the tropical and subtropical regions and to a lesser extent in the temperate zones of the world (Figure la). The family is better represented in Central America and the West Indies than m South America except for the Maytenus genus [5] (15 species in Peru and 15 in Venezuela). They are found growing as upright trees, bushes and lianas and almost invariably have resin ducts or cells in the bast of the stems and leaves. The leaves are simple, usually alternate and opposite; stipules small and deciduous or missing; the flowers are small, fasciculate, actinomorphic, forming cymes, very occasionally racemes or inflorescence. The flowers are in general bisexual and rarely polygamodioecious plants. The petals are freestanding or coneshaped. The fruits are berries, capsules, drupes or samaras. The seeds generally have a coloured aril and contain embryos with large cotyledons and a relatively oleaginous endosperm. The chromosomic numbers described are X = 8,12,14.

741 The Celastraceae genera are very diversified; some of die taxa widi most species are: Maytenus (225; tropical), Salacia (200; tropical), Euonomynus (176; Himalayas, China and Japan), Hippocratea (120; tropical. South America, Mexico and the South of the USA), Cassine (40; South Africa, Madagascar, tropical Asia and the Pacific), Celastrus (30; principally Asia, some in Australia and in tropical and temperate zones of America), Elaodendrum (16; tropical and subtropical), Pachystima (5; North America) and Gyminda (3; Central America, Mexico and Florida). Gentry [5] compared the woody taxa found in Africa and America and commented that tiae genus Maytenus was in taxonomic terms closer to other genera of the same family found on the same continent than to other species of Maytenus from the other side of Ae Atlantic; this resemblance pattern is observed also at the level of secondary metabolites. Although there is no clear explanation for the foregoing, paleobotanic studies have shown that the paleoflora of South America and Africa were very similar in the late Paleozoic era (340-240 My) [6]. Bearing in mind that the early stages of Continental Drift between Africa and South America occurred about 135 My ago it can be argued diat many populations widely distributed in Gondwana were separated as the resuh of the Continental Drift (Figure lb) wliich meant that the split populations would develop independently Aereafter. Although few examples of Ae Celastraceae family have been found and only recently, fossilized remains of Celastrus show that during the Tertiary Age this genus was unrestrictedly distributed throughout America [7] and Europe [8]. It has even been suggested that the present distribution pattern of some species of American Celastrus corresponds better to a dispersion centre in Asia than to (Mie in Central America [7]. The phylogenetic history of the Celastraceae family, - and of other widely distributed families,appears to run parallel with that of the Continental Drift. Modem chemomolecular studies may help to determine the phylogenetic relationships of its different members, as well as with other families of the Celastrales order. If the theory of the Continental Drift is true, it is not surprising that the phylogenetic relationships of the Celastraceae with odier plant families should date back a long time. New chemotaxonomical studies are constantly appearing which relate the Celastraceae with tfie Lamiaceae, for instance. The presence in both femilies of sesquiterpenes, diterpcnes and triterpenes of high and specificfrmctionalityconfirms such a relationship [9].

742

Figure la.- Present-day distribution of the species of the Celastraceae family. As can be seen, this family has developed for preference in die tropics in the New and Old World, and is extending towards temperate or temperate/cold climates in the southern hemisphere.

Figure lb.- In overall terms, the Continental Drift is theorized as shown in this scheme. Some 200 My ago, all Ae continents formed a single mass, known as Pangea (a), which later split into two, Laurasia (NorA America, Europe and Asia) to Ae north and Gondwana (the Antarctic, South America, India, Australia and New Zealand) to the south (b). Still later Gondwana disintegrated, with India breaking off first and fmally colliding with Asia, then South America litting off from Africa, New Zealand from Antarctica and last of all, Australia from Antarctica (c &d).

743 3. SESQUITERPENES The salient feature of the family has been its wealth of sesquiterpenes with abnost a hundred of these compounds being isolated and characterized chemically. This group of metabolites very common in Latin American species have the eudesmane basic skeleton, and originate the wellknown p-dihydro-agarofuran skeleton polyester sesquiterpenes which are esterified by a series of common organic acids - acetic, benzoic, 3-furoic, trans-cinnamic acid, etc. (Table I).

A. POLYESTER SESQUITERPENES Sesquiterpene esters based on the dihydro-agarofuran moiety occur mainly within the Celastraceae family.

The basic polyhydroxy skeleton vary according to the position, number and

configuration of the esterresiduesin the dihydro-p-agarofuran sesquiterpene. The interest generated by polyester sesquiterpenes from the Celastraceae has increased in line with the complexity of the substances isolated and the possibility of their being applied to combat insect plagues instead of synthetic insecticides. The complexity and increasing numbers of these sesquiterpenes makes it difficult to arrange them systematically. They can however, be treated as derivatives of a basic polyhydroxy skeleton and thereafter organized in sunpler series. Accordingly, 37 series of P-dihydro-agarofuran type sesquiterpenes have been proposed ranging from a skeleton with two hydroxyls (boariol) to one with nine (euonyminol and isoueonyminol series) (Figure 2).

Sesquiterpenes with two hydroxy groups

K

HO

Bovbt

Sesquiterpenes with three hydroxy groups OH

OH

OH

OH

OH

OH

HOi

^ - < Isocekxbicol

4{J-Hyd'«y-6-deo)y-oekxt3icol

744 Sesquiterpenes with four hydroxy groups CH2OH OH

Maikangunbi

OH

OH

OH 4 (J-Hyd-ocy-cebrbicol

OH

2p.4p-Dihydroy^<Jecjymthe Celastraceae characterized by the presence of a tetrahydro-oxepine nucleus. It would seem that these new types of skeleton are only biosynthesized by species of the MORTONIA genus, which consists of just four species, endemic to Mexico and the southern Unites States. The chemical study of three of these four species led to the isolation and characterization of eight new sesquiterpenes [3133](Figure3). OBz

OBz OBz

082 OBz

COOH 64 Mortonin A, R=H

66 Mortonin C

HOî 67 Mortonin D

68 Mortonol A, R=H 69 Mortonol B, R= OAc

65 Mortonin B, R=OAc

Figure 3. Sesquiterpenes isolated fix)m the genus Mortonia

The structures proposed

for MORTONINS A and B are the first recorded example of a

natural product in which ring B of the eudesmane skeleton undergoes oxidative cleavage to the the y-lactone. The subsequent isolation and characterization of the di-ester ketone MORTONOL (68) from M. greggi suggests that this sesquiterpene might be the biogenetic precursor of the whole MORTONIN series (Figure 4).

752

64.66

67

Figure 4. Possible formation of Mortonins sesquiterpenes

Boariol [18,23] is another new sesquiterpene isolated from the Chilean species M boaria Mol. which does not conform to the classic model of the sesquiterpenes previously described, and is in fact the simplest of all the compounds recorded from the Celastraceae. 'H and ^^C nmr studies showed the presence of a secondary and a tertiary OH, the latter at C-4 but with the opposite configuration to the customary p-hydroxyl at this position. The application of the Horeau method and an X-ray diffraction study confirmed the absolute configuration of the compound [18,23]. The absence of substituents at C-1, another notable feature of this structure, casts doubis on the biogenetic theory for the P-dihydro-agarofuran sesquiterpenes from this family which presumes that such substituents are present in nature. The possibility that boariol (34) might be an artifact was ruled out on the basis of two data: several sesquiterpenes with the classic C-4 d-OH configuration have been isolated from M. boaria Mol., even some with C-3 substitution; no products were obtained with carbonyl groups at C-3 and without hydroxy groups at C-4 which could have been hydrated non-stereospecifically via enol formation [18,23]. Fig. 5.

753 B. SESQUITERPENE ALKALOIDS Sesquiterpene alkaloids have similar structures to polyester sesquiterpenes except that the hydroxy groups of the eudesmane basic skeleton are esterified by nicotinic acid and/or its derivatives.

Little has been published about sesquiterpene alkaloids from American species

which tend to be found in the roots of the plants (Table II).

TABLE 11. N4AYT0LIN-TYPE SESQUITERPENE ALKALOIDS 15 10

f. M>" *13

Form

C-1

C-2

C-3

C-4

C-6

70

aOBz

2H

2H

pOH

pONic

71

aOCinn

POH

2H

pOH

pONic

72

aOBz

2H

2H

POH

pONic

73

aOBz

2H

2H

pOH

pONic

74

aOAc

aONic

2H

POH

I 75

aONic

aONic

2H

pOH

76

aOAc

aONic

2H

77

aOAc

aONic

78

aONic

79

aOAc

C-9

j C-8

C-15

Ref

pOAc

H

1 34

2H

POAc

H

34

POAc

aOAc

H

34

pOH

aOAc

H

34

pONic

aONic

pOBz

OAc

35

pOBz

aONic

POBz

OAc

35

pOH

pOBz

aONic

pOBz

OAc

35

2H

pOH

pOAc

aONic

POBz

OAc

35

aONic

2H

pOH

pOAc

aOBz

pOBz

OAc

35

aOBz

2H

H

pOAc

2H

PONic

OH

36

2H

'

80

aOAc

aOBz

2H

H

pOAc

2H

PONic

OAc

36

81

aOAc

aONic

2H

(30H

POH

aONic

pOBz

OAc

37

82

aONic

aONic

2H

pOH

pOH

aONic

pOBz

OAc

37

83

aONic

aONic

2H

pOH

1 pOAc

aONic

\ pOBz

OAc

37

1 ^"^ 1 85

aONic

aOAc

pOH

POH

aONic

POBz

OAc

: 37

aOAc

1 aONic

L ^QAc

1 aONic

2H

1 2H

L_H

1 j

1 pOBz 1 OAc 1 37 __..

C. N4ACROCYCLIC SESQUITERPENE ALKALOIDS Celastraceae also elaborate other, more complex, alkaloids, also polyester sesquiterpenes, incorporating a macrocycle derived from an evonic, wilfordic, cassinic or other type p>Tidine dicarboxylic acid with an additional alkyl chain of the basic eudesmane cycle at C-3 and C-7 (Table III). Celastraceae alkaloids are well-documented for the European and Asian genera, particularly Catha, Celastrus, Euonymus and Trypterigium but are relatively rare among the Latin

754 American species. Except for a few from the Hippocratea, Peritassa and Orthosphenia genera, most new Celastraceae alkaloids have been obtained from species of Maytenus. As in the case of the polyester sesquiterpenes, structural elucidation has been based on 1H13C nmr correlations (HETCOR) and long range inverse detection (HMBC and HMQC). Relative configurations have been determined by the combined use of NOESY experiments. The absolute configuration of almost all the compounds was established by circular dichroism applications using the exciton chirality method in 1,2-dibenzoate systems. TABLE III. MACROCYCLIC ALKALOIDS FROM AMERICAN CELASTRACEAE (Wilfordate type)

!

Form

R'

R2

R3

COMPUESTO

Rcf|

86

OBz

OBz

OAc

EbcnifolincW-l

35

87

OBz

OBz

OH

Ebenifoline W.2

35

88

OBz

OAc

OAc

Euojaponine F

35

89

OAc

OAc

OAc

Euonine

35

1

90

OAc

OBz

Cangorinine W-I

36

1

91

OAc

OBz OBz

ONic

CingminiDe W-II

36

TABLE IV. MACROCYCLIC ALKALOIDS FROM AMERICAN CELASTRACEAE (Evoninoate type)

755

Rl

R^

R^

R4

R5

R7

R«

COMPUESTO

i ^^

OBz

OH

OH

OAc

OAc

OAc

H

OAc

EbcnifolincE-l

Rcf. 1 38 1

I 93

OB7

OAc

OH

OBz

OAc

OAc

H

OAc

Ebenifolinc E-2

38

OAc

OBz

OAc

H

OAc

Ebcnifolinc E-3

38

OAc

OAc

H

OAc

Ebenifolinc E-4

38

R6

94

OBz

OAc

OH

95

OBz

OAc

H

OAc

1^

OBz

OAc

OH

OH

OBz

OAc

H

OAc

Ebenifbline E-5

38

97

OBz

OH

OH

OBz

OAc

OAc

H

OAc

EuojaponineC

38

98

OBz

OAc

OH

OAc

OAc

OAc

H

OAc

Mayteine

38

99

OAc

OAc

OH

OAc

OAc

OAc

H

OAc

Euooymine

38

100

OAc OAc

OH

OBz

OAc

OAc

H

OAc

CangoriniE-I

39

101

OAc

OAc

OH

OAc

OBz

H

OAc OAc

Horridtne

40

I 102 OAc OAc

OH

OH

OAc

H

OH

OAc

Acanthotfaamine

41

103

OBz

CNMP

OH

OAc

OAc

OAc

H

OAc

Hippocrateine I

[AQ4_

OAc

CNMP

OH

OH

OAc

OAc

H

Mb

Hippocratcine n

42 J2

j

CNMP= 5 Carboxy-N-methylpiridonyl Mb= 2-Methylbutyroyl Orthosphenin (105) breaks the classical mould of the Celastraceae macrocyclic alkaloids described to date and is the only example of an evoninol nucleus with an oxo group at C-8 and residual cassinic acid. Its structure was ascertained by the spectroscopic methods mentioned above, hydrolysis and the preparation of derivatives [43]. Two new evoninate-type alkaloids have recently been described, peritassin A and B, obtained from species of the genus Peritassa. These structures are distinguished by the macrocyclic unit which consists of an evoninic acid isomer in which the pyridine ring of the dibasic acid is substituted at 4'-5' instead of the more usual substitution of evoninic acid at 2'-3' [44].

R«OAc PcritassineA R = OBz PeritassincB Figure 6. The Structures of Orthosphenin and Peritassin A and B

756 IV. DITERPENES In general, very little has been written about diterpenes from the Latin American Celastraceaeas these structures are not often found. Abietriene type diterpenes have been the general rule in the Celastraceae although the chemical study of the minor constituents of Orthosphenia mexicana and Rzedowskia tolantonguensis did enable pimarane type diterpenes to be isolated and chemically characterized [43,45] and the second of these species afforded a series of new diterpenes with an isopimarane skeleton, described for the first time in the Celastraceae. The structure of the diterpene 107 (C20H30O3) was established by spectroscopic methods and confirmed by x ray diffraction studies while, under nmr, the nor-diterpenes 109-113 proved to be structures with an exocyclic methylene and no carboxylic groups at C-4 and are assumed to be the result of an oxidative decarboxylation process as has occurred elsewhere. Orthosphenia mexicana yielded another new diterpene of the nor-isopimaradiene type (C19H28O3) related to the abovementioned products [43]; spectroscopic analysis and chemical trans-formations established its structure with a tertiary hydroxy 1, an a,p-unsaturated keto group and the presence of a typical vinyl system of the ABX type.

CH2OH

W^' Fig.

106 107 108

Rl (M COCHi CH2OH

R2

0 2H 2H

Ref 43 45 45

CHzOH

R2.„

^

Fig. 1 109

110 111 112 113

Rl 0 BOH aOH 0 POH

R2

H H H

(m OH

Ref

757 V. ALKALOIDS M. loesner Urb. and M buxifolia (A. Rich) Griseb collected on the island of Cuba have been extensively studied by H. Ripperger et al. [46-47] who isolated a series of new macrocyclic alkaloids of the spermidin type, commonly found in the Celastraceae family; the new alkaloids could be related to others akeady obtained by Kupchan's group. feHs o H

Fig

R

COMPOUND

Ref

114

OH

Mayfoline

46

115

OAc

N( 1 )-acetyl-N{ 1 )-deoxymayfoline

46

"T^ ^

^3^^..^^' OAc

Fig.

R

COMPOUND

Ref

116

C^rl] jK^H^^H^H-

Loesenerine

47

Q')T{^-Qr\\^-Q\\-(Z')r{^-

17.18-Didehydroloesenerine

47

CnHs-CHOH-C4H4-

16,17-Didehydroloesenerin-18-ol

47

i 117

118

VI. TRITERPENES A. TRITERPENES FROM THE AMERICAN CELASTRACEAE The triterpenoids hitherto described for the Celastraceae almost invariably belonged to the FRIEDO-OLEANE series (including methylene quinones and phenolic compounds), LUPANE, OLEANE, GLUTINANE AND TARAXERANE series. Characteristic of the family are the triterpene methylene quinones synthesized in the roots of the plant and considered as taxonomic indicators and the same holds true for the American species. To date about 12 different endemic species belonging to eight different genera have been studied and 26 new triterpenes have been described as well as new triterpenoid dimeric structures. As is usual, all the species studied have a broad range of biological activity probably due to the presence in most, of triterpene methylene quinones of known biological effect such as pristimerin, celastrol, tingenone, iguesterin [48] etc.

758 Particularly interesting has been the case of Orthosphenia mexicam which yielded five new triterpene methylene quinones with a new carbon skeleton, a greater degree of conjugation than hitherto reported, an extra 14-15 double bond and a rearranged methyl at C-15. Its structure was elucidated by a succession of chemical transformations, spectroscopic methods and X ray diffraction which determined the absolute configuration of this compound [49,50]. TABLE V. METHYLENE QUINONE TRITERPENES

Rl

R2

CO^Me

OH

CO^Me

OH

H

C07Me

OH

H

CO^Me

OH

H

CO^Me

H

H

R3

H

R4

R5

COMPOUND

Ref

0

Me

Netzahualcoyone

49

H? OH

Me

Netzahualcoyonol

50

Me

Netzahualcoyondiol

50

H? H-,

CO^Me

Netzahualcoyol

50

Me

Netzahualcoyene

50

Rl

R2

R3

COMPOUND

CO^^CH:,

H.

Pristimerin

50

H

H? 0

50

0

H7 OH

Tingenone

H

22-6-Hydroxy-tingenone

51

OH

0

H.

20-a-Hydroxy-tingenone

51

Celastrol

50

i COOH

JH2_

ik^

Ref

!

759

R2

R» CO,CH,

H

COMPOUND

Rcf,

Hz

Isopristtmerm in

52 1

0

Isotingcoooc m

. _52 . ]

TABLE VI. FRIEDELANE TYPE SKELETON TRITERPENES

[RJ

R3

R^

o o

H:

0

0

H:

H2 oOH H2 aOH 0

[CH,

0

H2

H?

CH^

0

H7

|CH,

o o o

CH, CH^ CH,

|ca>H

rco,cH^ [COjH

R2

H?

R5

R6

COMPOUND

Rcf

CH,

FricdcUmc-3,15-l-olida (119) y 3-oxo-25(9->8) abeo-J5nedoolean-(4)(23)-en-24-al (120) transposedfriedo-oleanetype skeletons isolated from Schaefferia cuneifolia

TABLE IX. OLEANE SKELETON TRITERPENES

762

R»

R2

=o

R3

R4

F5

R6

R7

R8

COMPOUND

ReJ

CH7OH

CH2OH

H

H

H

H

3-Oxo-28,29-1 y;ir

0^

y ^

diketo radical

122

Fig. 10. Probable Formation of Dimers 121 and 122 by Radical Coi^ling

767

121 Ri=a-Me; R2=CCX)Me; R3=R4=H

Umbellatina

122 Ri=p-Me; R2=COOMe; R3=R4=H

Umbellatin p

124 Ri=a.Me; R2=H; R3=R4=0 125 Ri=a-Me; R2=H; R3=R4K) Fig. 11. Some Dimeric Triterpenesfromthe Celastraceae C00CH3

CO2CH3

C00CH3

Cangorosin B

Cangorosin A Atropcangorosin A Dihydroatropocangorosin A CO2CH3

CO2CH3

Mageilanin Fig. 12. DimersfromM. ilicifolia and M. magellanica

768 TABLE XV. ^^c NMR ( 50 MHz ) Data ( 6, CDCb, Chemical Shifts in ppm Relative to Me4Si) of Pristimerin and Ethers 121 and 122 Pristimerin

122*

121'

C

jC

c

C

119.0(d)

! 110.8(d)

114.9(d)

11.3(d)

2

178.4(s)

179.2(8)

173.4(8)

188.0(8)

115.3(d) 174.0(8)

i3

146. l(s)

171.3(8)

145.3(8)

171.5(8)

144.7(8)

1

u

117.0(8)

j 91.2(8)

j 124.0(8)

92.1(8)

|5

127.5(s)

128.5(8)

132.0(8)

127.7(8)

6

133.9(d)

129.0(d)

189.6(8)

126.7(d)

7

118.1(d)

117.4(d)

126.3(d)

116.2(d)

8

169.9(8)

164.5(8)

151.3(8)

161.4(8)

9

42.9(8)

38.8(8)

44.0(8)

38.2(8)

124.0(8) 130.1(8) 189.0(8) 126.2(d) 150.5(8) 41.9(8)

lio

164.7(8)

137.7(8)

151.3(8)

137.7(8)

151.0(8)

11

33.6(t)

33.0{t)

34.1(t)

33.0(t)

34.2(t)

[l2

29.7(t)

29.5(t)

29.7(t)

29.7(t)

|l3

39.4(8)

39.1(8)

39.3(8)

39.0(8)

[l4 lis

45.0(8)

44.5(s)

44.5(s)

44.7(8)

28.7(t)

28.7(t)

29.4(t)

28.7(t)

|l6

36.4{t)

36.5(t)

36.5(t)

36.4(t)

29.8(t) 39.9(8) 44.2(8) 28.6(t) 36.5(t)

|l7

30.6(8)

30.7(s)

30.7(s)

30.6(s)

30.6(s)

|l8

44.4(d)

44.8(d)

44.8(d)

44.5(d)

44.7(d)

|l9

30.9(t)

30.9(t)

31.0(t)

30.9(t)

31.0(t)

bo

40.4(s)

40.6(8)

40.7(s)

40.5(8)

40.5(8)

21

29.9 (t)

29.8(t)

30.0(t)

29.9(t)

29.9(t)

|22

34.8(t)

34.9(t)

35.1(t)

34.9(t)

34.9(t)

[23

10.2(q)

24.7(q)

13.3(q)

22.5(q)

12.8(q)

[25

38.3(q)

37.8(q)

40.2(q)

37.6(q)

37.7(q)

|26

1 21.6(q)

21.0(q)

22.5(q)

1 20.9(q)

S 22.2(q)

|27

18.3(q)

18.3{q)

18.6(q)

18.3(q)

18-6(q)

|28

31.6(q)

31.7(q)

31.7(q)

31.6(q)

! 31.6(q)

|29

178.7(8)

179.0(8)

1179.0(s)

179.0(s)

179.0(8)

[30

32.7(q)

33.0(q)

j 33.0(q)

32.7(q)

32.8(q)

[31

1 31.6(q)

L5L6(q)

! 51.8(q)

1 51.5(q)

1 51.5(q)

^ The values of the pairs C and C may be interchanged.

1 j 1 1 1 j 1 1 1

1 1 1 1 1

769 TABLE XVI. ^H NMR ( 200 MHz ) Data (5, CDCI3, for The Methyls. Zeylasterone Pristimerin

121 H*

H

122

2,3-DmiethyI ether

H

H*

23.Me

2.21

1.37 2.72

1.41

2.73

25-Me

1.48

1.48

1.57

1.47

1.58

1.60

26-Me

1.26

1.26

1.26

1.27

1.25

1.32

2.66

27-Me

1.10

1.05

1.08

1.06

1.08

1.12

|28.Me

1.18

1.16

1.16

1.16

1.16

1.18

30-Me

0.53

0.52 0.54

0.53

0.54

0.60

1

TABLE XVU. C NMR. (100 MHz) (6, CDCI3) Pristimerin

Magellanin

Pristimerin

Magellanin

C

c 1

C.28

31.6

31.5

31.8

C-1

119.0

115.6

108.0 1

C-29

178.7

178.9

179.3

C-2

178.4

191.1

140.0 1

C-30

32.7

32.8

32.2

C-3

146.1

91.8

C.31

51.6

51.6

51.6

C-4

117.0

78.7

C.5

127,5

130.7

C.6

133.9

126.3

C-7

118.1

116.3

C-8

169.9

160.5

137.6 122.4 125.0 124.0 129.1 45.5

C-9

42.9

41.6

38.2

1 1 1 1 1 1

TABLE XVm. H NMR. (200 MHz) Magellanin H-1

CUCD

C.D.

6.06 d

6.07 d

J(Hz) (1.16)

(1.44) 6.42dd

C-10

164.7

173.7

143.7

H-6

C-11

33.6

32.9

31.2

J(Hz) (1.16,6.30)

C-12

29.7

29.7

29.5

C-13

39.4

39.0

38.9

C-14

45.0

44.5

44.3

H-y

6.70 brs

7.04 brs

C-15

28.7

28J

28^

H-6'

6,63 dd

6.60 dd

|c-16

36.4

36.4

36.3

J(Hz) (2.66,1020)

(2.85,9.88) 1

C-17

30.6

30.6

30.4

H-T

5.48 dd

C-18

44.4

44.4

44.2

\^&) (2.48,10.20)

IH-7

6.32 dd 5.92 d*

J(Hz) (6.30)

5.90 dd*

(1.44,6.46) 5.32 d (6.46)

(2.58,9.88)

[c-19

30.9

31.0

30.5

|c-20

40.4

40.4

40.5

|c-21

29.9

30.0

29.8 !

TABLE XIX. Three-bond^H-^^C

|c-22

34.8

34.8

36.0

coupling (HMBC) in Magellanin

|c-23

10.2

22.5

10.9

H-I

C-3,C-5.C-9

|c-25

38.3

34.9

22.2

H-6

C-4,C-8,C-10

|c-26

21.6

22.2

17.0

H-7

C-5,C-9,C-14

|c-27

18.3

18.3

17.5

H-r C-3',C-15'

•overlapping signals

1

770 VII. MISCELLANEOUS A number of heterogeneous natural products have been isolated from American species including aromatic and phenolic compounds, flavonoids, catechins etc. The following table indicates the main studies on the subject.

TABLE XX. SOME NON-CLASSIFIED PRODUCTS FROM THE CELASTRACEAE COMPOUND

COMPOUND

Ref.

2,6-Diacetoxy-7-hidroxy-8-metoxychromone

61

4,5-Dihydroblumenol A

72

1

Blumenol A

72

(-H'-O-methyl-epigallocatechin

65

Ouratea proanthocyanidin A

65

Dulcitol

65

j

Epicatechin

18

1

5'-0-Methylgallocatechin

18

1

4-Hydroxybenzalddiyde

18

Femiginol

9,73-75]

Sakuranctin

9, 73-75 1

Vni. TRANSFORMATIONS The Tenerife group which is responsible for about 70% of all the research published on the Latin American Celastraceae has concentrated on the isolation and structural characterization of secondary metabolites; ahnost incidentally they have also developed various transformations and partial syntheses to test biogenetic theories in vitro and prove structural correlations by means of chemical transformations [76]. Thus, a succession of transformations showed fiiedoleane triterpenes with hemiacetal 24hydroxy-3-keto grouping to be possible key intermediates in the biogaietic pathway of the Celastraceae triterpene quinones, and a triterpene with a hemiacetal group in the remote C-24 position was synthesized from fiiedelin, as shown in the scheme [76] (Fig. 13).

771

FrtocMki

(20% yield based on lactone) i) Na BH4 in ether ii) IBDA/12, py/CH2Cl2, 100 W tunsgten filament iii) t-butyl chromate/ ether iv) LiAlH4 V) K2O + n-bu4N^Cr/THF + (NH4)6Mo7024H2/K2C03, H2O2

Fig. 13. Synthesis of a Friedelane Triterpene with a 24-Hydroxy-3-oxo-hemiacetal Group

XI.

BIOLOGICAL ACTIVITY

A. ANTIFEEDANT ASSAYS It has been known for some time that some polyester sesquiterpenes of the p-dihydroagaroftiran type such as those IBrom the Celastraceae deter various msects from feeding. In China the powdered root bark of Celastrus angulatus [77] has been sprayed on crops to protect them against insect attack. Chemical and biological analysis has shown that the powder is active against various species of insects including the cucumber beetle (Aulacohora femoralis chinensis\ the crucifer beetle {Colaphellus lowringi), the cabbage work (Pieris rapae) and migrant locusts {Locusta migratoria migratorioides and Locusta migratoria manilensis). Wilfordin, tryptofordin and celangulin (Fig. 14) are antifeedant compounds obtained from extracts of the Celastraceae species Maytentis rigida [78], Trypterigium wilfordii [79] and Celastrus angulatus [77, 80], respectively, and as some products isolated from South American species have similar structural characteristics, they too have been assayed.

772

The insects used for assay were fifth-instar larvae of Spodoptera littoralis Boisduval {Lepidoptera, Noctuidae) and the methodology used to determined the FR50 was that described in references [18, 81]. Antifeedant activity has been discerned in 16 sesquiterpenes obtained from five endemic Latin American species. The results are set out in Table XXII. All compounds were active at a dosage of lO^g/cm^ with 72 the most active with FR5o !•

m/k Figure 4- FAB-MS of native glycosphingolipid from Leishmania amazonensis

1164

795 The anomerjc configuration and linkage sequence of GSL were analyzed by ^H-NMR spectroscopy in dimethyl sulfoxide-d6/D20. The spectrum (Figure 5) of GSL displayed four anomehc resonances, one with a configurations (^Ji.2=3.4 Hz) and three with p configurations {^Ji.2= 7-8 Hz). The spectrum also showed the presence frans-vinyl signals of sphingosine at 5.36 and 5.54 ppm (R4 and R5 respectively) and a very weak triplet at 5.32 ppm representing the c/s-vinyl protons from unsaturated fatty acids. To confirm the glycan sequence, enzymatic degradations, using a-and p-galactosidases from bovine testis, were performed and monitored by HPTLC and ' ' H - N M R spectroscopy.The combination of data confirmed the complete structure of GSL as a novel globosehes structure: Gal P ( 1 ^ 3)Gal a(1 -^)Gal p(1 ->4) Glc 1 ^Cer, which has P(1 -^ 3)Gal substituting for the p (1-> 3) GalNac of globoside. These data reveal the existence of a high concentration of neutral GSLs in Lamazonensis amastigotes, in contrast v^th their virtual absence in promastigotes. It indicates clearly that a remarkable change in cell membrane glycoconjugate composition is associated with cell phase differentiation in Lamazonensis. Modulation of GSL biosynthesis and expression in amastigotes may be an important step correlated with their survival and proliferation in host macrophages.

QaV»1-»3Gat.cV4Gal|»V4Qlc/»V1Cer IV

R-5

R-4 ds

rnn

fiY) /

lU

il

I

R

I

M

-It-^llV5A0

5.20

'*'• ^-^^J^ 5J00 4.80

4J60

4.40

4^0

4X)0

PPM Figure 5 - ^H-NMR of a novel globoside from L amazonensis. The assignment of each resonance Is indicated by arable numbers for the positions of protons and roman numerals for sugar residues. R=frans-vinylprotons of sphingosine backbone. cis= c/s-vinyl protons of unsaturated fatty acid chains.

796 GiycosphingoUpids of Trypanosoma cruzi Trypanosoma cruzi, the causative agent of Chaga's disease (American trypanosomiasis) exists in three morphologically different forms related to the three different environments in which it lives. These three forms comprise amastigote, a dividing form found intracellularly in mammalian hosts, epimastigote, a multiplying form found in the vector's digestive tract and in culture, and the trypomastlgote a non-multiplying infective form, that occurs In the lumen of the rectum of the reduviid bug and is infective to the mammal. T cruzi multiplies discontinuously In the vertebrate host; the amastigote intracellular stages multiply by binary fission and they transform into nondividing trypomastigotes which emerge from tissues into the bloodstream, where they circulate for a certain period before penetrating cells and resuming their complex life cycle (45). Infection with T cruzi, results in the formation of chronic lesions In many tissues, Including muscle and the nervous system (46). The chronic pathological phase of Chaga's disease Is thought to be of autoimmune origin, due to the presence of crossreactive antigens betvy/een the parasite and the mammalian host or due to the failure of self-non self discrimination in the infected host (47,48,49). GiycosphingoUpids have been isolated from epi and trypomastlgote forms of Trypanosoma cruzi. The major neutral glycosphingolipids from 7. cruzi ceramide monoand dihexosides (CMH and CDH, respectively), have been purified and their structures elucidated using a combination of techniques I.e. column chromatography on latrobeads RS 2060, HPTLC and GC together with FAB-MS spectrometry and 500 MHz ' ' H N M R spectroscopy (50).The ceramide monohexoside fraction (CMH) which chromatographed as a double band, arising from the fatty acid heterogeneity, contains either glucose or galactose,

sphingosine

and

as

fatty

acyl

groups

mainly

C-24

saturated,

monounsaturated or 2-hydroxy fatty acids as fatty acyl groups (Table 2). The FAB-MS spectrum of the peracetylated CMH and CDH fraction from 7. cruzi contained

major

molecular ions [M+Na]+ at m/z 1220 and 1248 (Figure 6). Ions at m/z 331 and 289 together with m/z 619 are derived from the hexose-hexose resldue.The pattern of fragmentation of CDH is depicted in Scheme 1. The distribution of fatty acid chains was calculated from the relative intensities of the most intense [M+H-60]+ ion of the CDH fraction. The peracetylated CDH fraction was further analyzed by 1- and 2-dimenslonal NMR spectroscopy. The results shown in Table 3 are in complete agreement with the proposed structure of a lactosylceramide.

Table 2 - Relative distribution of fatty acid chain lengths of the ceramide monohexoside fractions CMH-CoHand CMH-Cn from T. cruzi as calculated from the (M + H 60)’ ions of the peracetylated compounds.

-

-

[M + H 601’ 848 850 8 76 878 904 906 908 918 920 932 934 936 946 948 960 962 964 974 976 978 990 992 1004 1006 1018 1020 1032 1034 1046

(Nz)

Chain length of n-fatty acid 16:l 16:O 18:l 18:O 20:l 20:o

CMH-Cn (X)

21 :1 21:o 22:l 22:o

2 5 3 10

23:l 23:O 24:l 24:O

3 8 6 16

251 25:O

7 7

3 15 3 6 2 4

Chain length of a-hydroxy

CMH-Co” (%)

16:l 16:O

8

18:l 18:O

5 5

20:l 20:o

3 3

21:l 21 :o 22: 1 22:o 23: 1 23:O 24: 1

4 5 1 8 tr. 9 8 20 5 8

24:O

251 25:O 26: 1

7

Based on the MS data only sphingosine is present in the ceramide moiety CMH-C,, ceramide monohexosides from T. cruzi with n fatty acids; CMH-Con, CMH with a-hydroxy fatty acids from T. cruzi; tf, trace

4 4

798

MASS PC* CHARCC

MASS PCR CHARGE

Figure 6 - Molecular ion region of the FAB-Mass spectra of peracetylated C M H - C Q H (A) and CDH (B) showing the [M + H - 60]* ions

-CMj-CO 289331

^

''^ ,

•H OAc , •V CH

Ac

Acorv p

Scheme 1

H ^ ^ \

HN OC M-^Na'^

-1248

M+H*

«1226

[Hd+H-eor -1166

<x

i^°;^^yV^^/^ OC

H-1

ppm 4.49

J1.2

H-2

5.11

J2.3

H-3

4.96

^J3.4

H-4

5.35

^J4.5

H-5

3.88

J3.6; J5.6

H-6;6'

4.09;4.14

[Hz] [8.0] [10.5] [3.0] [1.0] [6.0;6.9] [11.4]

ppm 4.44

III

[Hz] [7.5]

4.87

[10.0]

5.19

[10.0]

3.80 3.59

CH2—R"

[10.0] [2.1;4.5]

4.08;4.51

[12.0]

ppm 3.90;3.52

H-1;1' •^1.1; Jl.2; J1.2

[Hz] [10.0;3.6;4.2]

4.30

H-2

m

J2.3

H-3

5.25

^J3.4

5.35

H-4 J4.5; J4.6

H.5

5.77

H-6;6»

2.02

H-N

5.62

JNH.2

H-aliph H3C(R';R"I

1.28-1.35 0.88

[7.5]t [15.0;0.9] [6.0]t

[9.0]

m, multiple!; t, triplet

The role of the monohexosylceramide isolated from epimastigote forms of T. cruzi, on the interaction of T. cmz/with heart muscle cells was studied by Vermelho et a! (51). The results show a large

increase in the number of infected cells when the highly

infective Dm 28c clone of T. cruzi and heart muscle cells were preincubated with the glycolipid before the interaction (Figure 7). This finding may be due to the uptake of the lipid by both parasites and host cells, as shown previously in fibroblasts (52) and tumor cell lines (53). The simultaneous addition of glycolipid and metacyclic trypomastigotes results in a decrease in the penetration of the parasite.

Competition

between

the

800

(%)

-I—Ti

Time (18 h)

Figure 7 - Effect of glycosphingolipid on the infectivlty of the Dm 28c clone of T. cruzi with heart muscle cells. Control (open column), heart muscle cells infected with T. cruzi, Dm 28c clone. Experiment 1 (suppled column), metacyclic fomfis, preincubated with glycoHpid for 30 min, following addition to the cell culture. Experiment 2 (solid column), heart muscle cells incubated with glycolipid for 30 mIn, before metacyclic addition. Experiment 3 (cross-hatched column), heart muscle cells incubated simultaneously with metacyclic fonms and glycolipid.

glycolipid and the protozoan for the receptor connobining site of the heart muscle cells or saturation of the receptors may be responsible. As glycosphingolipids were shown to be immunogenic (54) and parasite glycolipids as well as other glycoconjugates stimulate the host immune response (55), the reactivities of Chagasic patients sera and sera from rabbits hyperimmunized with epimastigote cells, were assessed using highly purified glycosphingolipid fractions from 7. cruzi epimastigote (56). A strong reactivity with GSL was obtained with T. cmzHmmunlzed rabbit. Reactivity with GSL was also obtained with human Chagasic sera.Compared to a group of normal individuals, the reactions of antibodies directed against lipid antigens were considerably increased in sera of patients with Chaga's disease. Recent studies by Avila and Rojas (55) showed the presence of elevated cerebroside antibody levels in chronic Chaga's disease. Chagasic sera did not differentiate between glycolipids with terminal p glucosyl or p galactosyl non-reducing units. They however discriminated between glucosylceramides with differences in the ceramide structure. These

results

suggest

that

antibody

801 recognition of cerebrosides involves either a complex epitope formed by the terminal sugar and elements from the ceramide moiety or, alternatively, the presence of the ceramide lipid allows an increased affinity of the antibody for a single carbohydrate unit. Although the glycolipids are recognized by Chagasic patients sera, the reactive antibodies are not specific since they

also reacted quite well with murine laminin

(Figure 8). Most of them seem polyreactive, strongly binding to murine laminin. The homologous reactivity with the parasite glycolipid v^s much less Intensive involving mainly IgM. These results and others with different antigens (57) give support to polyclonal B cell activation in Chaga's disease.

UMMMrfPURFEDAbs

la-PURFED Ate

Figure 8 - Reactivity of Chagasic semm antibodies purified on crude glycolipid on laminin Immunosorbents with both antigens. Antigen on ELISA plate: A and C, murine laminin at 100 ng/well; B and D, crude glycolipid at 10 |ig/we!l. IgG • ; IgM i .

Cross-reactive lipid antigens were isolated from epimastigote forms of 7. cruzi and from the mammalian brain with the monoclonal antibody VESP 6.2 (58) which had been raised against T. cruzi-re\aied trypanosomes T. dionisii and T. verpertillionis.Chemical reactions indicated that the sulphate group of the lipids is an important part of the epitope recognized by the Mab. The specificity of VESP 6.2 for these isolated lipid antigens was demonstrated by three different methods: a) high-performance thin layer chromatografy, b) solid phase radioimmunoassay and lysis of artificial liposomes (59).

802 A partially characterized sulfoglycosphingolipid

is also present

in 7. cruzi

trypomastigotes (60). Besides neutral and sulfoglycosphingolipids, sialoglycolipids have been characterized In trypomastlgote forms of 7. cruzi (61). Glycosphingolipids of Trypanosoma dionisii A monohexosylceramide from a 7 cruzi related trypanosome, 7 dionisii, has been Isolated and analyzed by HPTLC, gas liquid chromatography and FAB-MS (62). As sho\A^ in Figure 9, the GSL developed as a doublet band by TLC with the same migration as galactosylceramide from calf brain.

Figure 9 - High-performance thin-layer chromatography (HPTLC) of neutral glycosphlngollpld from 7 dionisii. Lane 1: Neutral glycosphingolipids from human erytrocytes and bovine brain. Lane 2: Monohexosylceramide from 7 dionisii. Runnlno solvent: CHCIa/MeOH/water, 60:25:4 v/v. Detection: orcinol-H2S04 reagent.

This GSL was methanolyzed and the methyl glycosides converted to their trifluoroacetyl derivatives for GLC analysis. The GLC chromatogram shows the carbohydrate composition of CMH, with galactose and glucose in the molar ratio of 1:1 (FIgurelO). The FAB-mass spectrum in positive mode provided molecular weight information and reflected the sample heterogeneity. The molecular ion region of monohexosylceramide of 7 dionisii is shown in Figure 11. Intense signals between 906 and 1020 derived from the molecular ion [M+ H-60]+ are shown.

803

.Oal

QIC

Retention time (min) Figure 10 - Gas chromatogram of trifluoroacetyl derivatives of the sugars from T. dionisii monohexosilceramides (CMH).

MH*-HOAc

Figure 11 - Molecular Ion region of the FAB* mass spectrum of peracetylated CMH from 7. dionisii, showing the [MH*-HOAc] Ions. The relative amounts of fatty acids calculated from the molecular Ion peak Intensities of CMH are shown in Table 4. The presence of sphingosine in the ceramide moiety was confirmed by a fragment at m/z 264.

804 Table 4 - Relative amounts of fatty acid chain lengths in the CMH from T. dionisii calculated to the intensities of the [MH - + - HOAc] ions in the FAB -i- mass spectrum of the peracetylated species. Fatty add 20:0 22:0 24:0 14:0-OH 16:0-OH 18:0-OH 20:0-OH

MH - HOAc m/z 906 934 962 936 964 992 1020

I. rel [%] 117 13.5 22.6 9.1 15.0 12.6 15.5

A similar monohexosylceramide has been previously described for 7. cruzi. Using a monoclonal antibody directed against 7. dionisii and 7 vespertilionis, Retry et al (58) demonstrated cross-reactivity between these parasites and 7 cruzi (59). Previous results on the mechanism of entry of 7 dionisii into non-phagocytic cells (63,64) and its subsequent intracelular development provided evidence for the close similarity between 7 dionisii and 7 cruzi. Consequently, 7 dionisii is useful as a nonpathogenic model for the study of Chaga's disease. Glycosphingolipid of Trypanosoma meaa 7 mega is a trypanosomatid of cold-blooded vertebrates isolated from the African toad, Bufo regularis. Like the trypanosomatids in general, 7 mega has two principal stages in its biological cycle. In its vertebrate host, the trypanosome multiplies in the trypomastigote form while the epimastigote (crithidia) form is present in the invertebrate host (65,66). A glycosphingolipid fraction from 7

mega (67) was isolated after

acetylation and further purified on a silicic acid column, as shown in Scheme 2. The carbohydrate components of the glycolipid were fucose and galactose in approximately equimolecular amounts. Fatty acids, forming amide group with the sphingosine bases, were a mixture of normal and a-hydroxy fatty acids. Normal C16:0,C18:0 and 2-hydroxy C18:0 were the predominant fatty acids (Figure 12, Table 5).

805 T. mega cells

I

Extraction with 200 vol. of chlorofonm-methanol (2.1 and 1.2 v/v] filtration lipid extract

residue

Acetylation Florisil chromatography 1) Hexane-dichloroethane (1:4) discarded 2)Dichloroethane-discarded 3)Dichloroethane-acetone (1:1) - glycolipids 4)Dichloroethane-methanol-water (2:8:1) - discarded Chromatography on silicic acid 1)Chloroform-methanol (1:1) - discarded 2)Chloroform-hexane (3:2) - discarded 3)Chlorofonfn-discarded 4)Chloroform-methanol (95:5) - glycolipid

Scheme 2 - Isolation and purification of the neutral glycolipid from T. mega.

loo

11

2

Uu

10 h2

XJJ4J

yiLudj

ul

1516

Figure 12 - Reconstructive ion chromatogram from GLC-MS experiment of fatty acid from T. mega glycolipid. Peak identification listed in Table 5.

806 Table 5 - Composition of fatty acids of the giycolipid of Trypanosoma mega.'

Peak mumber 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Fatty acid C14:0 CI 5:0 C16:0 CI7:0 CI8:0 CI8:0 C20:0 C22:0 C23:0 C22:0 C24:0 C23:0 C25:0 C24:0 C26:0 C25:0

RRT'^ Normal 0.81 0.90 1.00 1.07 1.16 1.25 1.30 1.44 1.51 1.53 1.57 1.59 1.63 1.65 1.69 1.71

a-hydroxy

5.0 1.5 49.0 1.0 33.0

-

8.0

-

1.0

-

1.0

-

3.0

-

0.5

1.0 2.5 1.5 4.0 0.5

0.5

Principal fragments (m/e) 74, 87,101, 129,143, 185, 199. 211, 213,242 74,87,101.129,143,157,185,225,227,256 74,87,101,129,143.157,185,239,241,270 74,87.101,129,143,185,253.255,284 74,87,101.129,143.185.267,269,298 73,89,103,129,159.327,371,386 74,87,101.129,143.185,295,297.326 74,87.101.129,143.185,323.325,354 74,87.101.129,143,185,337,339,368 73,89.103.129,159,389,399,427 74,87,101,129,143,185,351,353,382 73,89,103,129.159.397.441,456 74,87,101.129,143,185,365,367,396 73,89,103,129.159,411,455,470 74.87,101,129,143,185,379,381.410 73.89,103.129.159,425,469,484

^ Nomial fatty acids were analyzed as their methyl esters on an SE-54 column. Hydroxylated fatty acids were analyzed as their 0-TMS derivatives. Values are derived from GC-MS analysis. ^ Retention times relative to that of n CI6:0.

Glycosphingoliplds of fungi The nitrogen-containing lipid was described by Zellner (1911) and called "fungus cerebrin". The structure of the cerebrin was determined by Oda (68) as 2-amino-1,3,4tri-hydroxyoctadecane. Because of the close similarity with the animal sphingosine, it was called "phytosphingosine". Although glycosphingoliplds of animal tissue have been extensively studied (69,70), the structure and function of glycolipids of fungi are less well known. Glycosphingolipids are present in fungi of the most primitive class, i.e., Phycomycetes, as well as in the most complex class,namely, Basidiomycetes. Glycosphingolipids from Zygomycetes Cerebrosides have been isolated from Phycomyces blakeslearus (3), a fungus often found on animals dung, by extraction of mycelia with acetone and chloroform/methanol mixtures and purified on a silicic acid column, followed by a Florisll column. The bases obtained after hydrolysis were all phytosphingosine homologs ranging in length from C17 to C22.Palmitic,steahc,oleic, llnoleic and hydroxy palmitic acids were the major fatty

807 acids present in this glycolipid. The cerebrosides contained equal amounts of galactose and glucose. Gfycosphingolipids from Deuteromycetes Fungi apparently lacking a sexual phase (perfect stage) are commonly called imperfect fungi or Fungi Imperfecti (71). Monohexosylceramides have been isolated from Sporothrix schenckn,Fusicoccum amygdali, Fusarium lini, Fusarium solani, Aspergillus oryzae, A. fumigatus and A. versicolor (72,73,3,74,75,76). A glucocerebroside from the yeast form of the dimorphic human pathogen S. schenckii has been isolated and its components characterized by thin layer and gas liquid chromatography

and

mass

spectrometry

(72).

It

was

found

to

contain

glucose.sphingosine and a-hydroxy stearic acid (1:1:1). No role has been attributed to this compound In association with infection and cell surface reactivity. Balllo et al (73) isolated and characterized a N-2-hydroxy-3-trans-octadecenoyl-1-0-Dglucosyl-9-methyl-cis-4,8-sphlngadienine from Fusicoccum amygdali Del., a fungus responsible for almond and peach "canker". A similar structure was seen In Aspergillus oryzae, the Japanese yellow mold (75). Trans-unsaturated hydroxy fatty acids ( 2-hydroxyoctadec-3-enoic acid) found in F. amygdali and A, oryzae cerebrosides have not been detected in sphlngollpids of animals and plants. The structural characterization of the glycosphingoliplds (Figure 13) isolated from two pathogenic species of Aspergillus, the etiological agents of a number of different lung diseases including asperglHoma ( fungus ball) and Invasive aspergillosis, were carried out using high-resolution "^ D,2D-''H-NMR and ''^C-NMR spectroscopy and fast atom bombardment mass spectrometry (FAB-MS) (78). Thin layer chromatography of native and peracetylated glycosphingoliplds Isolated and purified from Aspergillus on silica gel and a latrobead column afforded fractions with a mobility corresponding to a monohexosylceramide, Figure 14 (A,B). The FAB-mass spectra of the peracetylated glycolipid from Aspergillus in the presence and in the absence of sodium acetate are shown in FigurelS. In the molecular ion region, intense signals were seen at m/z 946 [M+H-HOAc] and m/z 886 [M+H-2H0Ac]. Ions indicating a terminal hexose were present at m/z 331 [HexAc4] and m/z 229 and 169 (secondary Ions). The ceramlde

moiety

808

4 i . »

f

B M

m

n

Figure 13 - High-performance thin-layer chromatography of Aspergillus neutral glycosphlngolipids isolated by latrot)eads column cromatography. Lane 1: Folch lower phase. Lane 2: Glycosphingolipid fraction otJtained by silica gel chromatography. Lane 3: Chlorofomi fractions. Lane 4 Chloroform/methanol (95:5) fraction. Lanes 5-11: Chloroform/methanol (9:1) fractions. Lane 12-13 Chloroform/methanol (8:2) fractions. Lane 14: Methanol fraction. Solvent system chlorofomn/methanol/water (65:25:4 vA/). Detection: orcinol-H2S04 spray-reagent. Purified neutral glycosphingolipid fractions (Lanes 7-11) were combined.

B CMH CDH

CMH

i

CTH QbO

Li

a

Figure 14 - High-perfomiance thin-layer chromatography (HPTLC) of native (A) and peracetylated (B) neutral glycosphlngolipids from Aspergillus. (A) Lane 1: Neutral glycosphlngolipids from human erythrocytes and bovine brain. Lane 2: Glycosphingolipid from Aspergillus. Running solvent: chloroform/methanol/water (65:25:4 vA^). (B) Lane 1: Galactosylceramide from bovine brain. Lane 2: Glycosphingolipid from Aspergillus. Running solvent: 1,2 dichloroetha/acetone (8:2 v/v). Detection: orcinol-H2S04 reagent.

809 was represented by peaks at m/z 658 [CerAc2], m/z 598 [CerAc2-H0Ac] and m/z538 [CerAc2-2HOAc].

NMt H I CMMW

Figure 15 - FAB*-MS of the peracetylated glycosphjngolipid from A. fumigatus 2140. (A) (M+H*). (B) (M.Na').

The structure of the long chain base was deduced from the 1D and 2D-NMR spectra. The HH-COSY spectrum (Figure16) revealed the complete connectivity of the long chain base starting from proton H-1 until H-11, from proton H-2' until H-11, and from proton H2' until H-5' of the fatty acid. The chemical shift of 5.30 ppm (long chain base H-3) and 5.50 ppm (fatty acid H-2') Is typical for protons wnth acetylated OH-groups adjacent to the olefinic bonds. For the protons H-4 and H-5 of the long chain base, a coupling constant of 15.3 Hz and for protons H-3' and H-4' of the fatty acid 14.3 Hz, respectively, were observed. These are typical for frans-double bonds. The CH-COSY also confirmed the double bonds (Figure 17). The E-configuration of the double bond in position 8 of the long chain base was evident from the NOE experiment by irradiating

proton H-10.

The methyl group at position 9 of the base was determined by COLOC, as shown in Figure18. The sugar component of CMH from both A. fumigatus and A. vers/co/or was glucose. The (3 configuration of the sugar was evident from the coupling constant of H-1" (7.9 Hz) (Table 4).

no

Vj'w

* * .^^ 1

^x'\

• 4^5

r2

I I I I I I I I I I I I I I t I I » TI t I I I I I I I I 1 t I > I I I I

ppm Figure 16 - 500 MHz HH-COSY of the peracetylated glucosylceramlde from A. fumigatus 2140 in CDCI3.

811

mmmmmmmm

ML

m

^ ^0|

^!

6^5'

16-

t^"

•2

-\'

^

^

CO

E

01

g^

00

ou

CO

^

a

CO

i 1

OH

s

^ a

^

/^

Ui c;

9

§'*•'

c -S

! i-i a>

tio

tiX)

^2

i^ ii112

a

1 ^^^^ f

O

^

(N

P4

a; X3 Cli

-M

2

ffi ffi ffi

a; X\ 0.

^4->

S

^

1 ^

01

0

^>

ffi

CO

>.

CO

>>

0 K" ûp

9!_,

CO

a ^tdD

>^

a;

CO

OH

4I.

CO

a ^yio

^

cu ^

jt

s i

-

S

I

CD

CO

-s

.Si

CO

3

CO

•S

I

s;

§

o b o c o o o o 2

95 cb

12;

a

XI O,

^

•+->

a;

^' M

s

a;

a;

§

& ^

S 2

>.

(D

^< M

^

o

1 -J-J

a

^

s s s s s g i Oi s0

1 4-»

>. >. O

g-g-g-g^^^^

^ 1

9 o o ^

a;

-M

CO

s

E O

1

?

^

\ij

CO

o CO

^ a ^

i i ^

s CO1

^

.$S

^

ffi

CO

ffi

836 the spatial array of the side chains, but more importantly in the peptide backbone, too. Recently we succeeded to determine a preferred conformation of the fully active CCK-analog [Thr,Nle]-CCK-9 (see table 2) by ^H-NMR analysis in the cryomixture dimethylsulfoxide (DMSO)/water at 278 K . As shown in fig. 9, it consists of a ytum centered on the threonine and separated by the glycine residue from an ahelix involving the C-terminus, whilst the N-terminus is flexible and salt-bridged between the tyrosine-O-sulfate and the arginine guanido function (78-79). Again the space-filling model demonstrates a rather compact structure particularly in the C-terminus. The spatial structure of the CCK-9 analog strongly reminds in the C-terminal tetrapeptide portion, i.e. in the active site, the 3D-structure of gastrin. These findings fully agree with the pharmacological properties of CCK which is capable of interacting with the gastrin/CCK-B receptor with high binding affinity. But then its selective recognition by the CCK-A receptor has to be attributed exclusively to a specific electrostatic interaction of the tyrosine-Osulfate moiety with a receptor counterpart. Thereby the exact location of the hemi-sulfate ester moiety is apparently playing a decisive role since sulfated gastrin with the tyrosine-O-sulfate shifted by one position in the sequence is poorly recognized by this type A-receptor (48). As discussed in section 2.2., N-terminal extensions of the active site portion of the gastrin molecule leads to a remarkable stabilization of its potential bioactive conformation with concomitant increase of hormonal potency. This aspect has been analyzed in the case of CCK too, by using the series of peptides listed in table 2 (64). CD measurements revealed that N-terminal extensions of the CCK mrlecule in sequence mode are affecting only marginally the conformational St: s of the bioactive core. As expected from what was known for gastrin, the CD sp ra of the CCK-peptides in aqueous solution were consistent in shape and in isity with predominantly random coiled structures. However, titration of the aqueous solutions of the CCK-peptides with TFE was found to induce ordered conformation which is apparently stabilized at least to some extent by N-terminal elongation up to the undecapeptide. On the other hand, the biological properties of these CCK peptides, as determined in different assay systems, were insensitive towards an increase of the CCK-peptide size, thus definitely confirming that CCK8 and acylated CCK-7 represent the smallest fully active sizes of the CCK hormone regarding CCK-A receptor-mediated signal transduction (64). It seems, therefore, reasonable to conclude that, differently from gastrin, the biologically relevant conformational states of the bioactive core, i.e. of CCK-7, at least reirarding the receptor A-mediated activities, are not affected by the peptide size, bi "lat stabilization of this conformation in the larger size CCK-peptides derives m :y by weakened endgroup effects. The CD spectra recorded for this set of

837

B Fig. 9.

The preferred conformation of IThr.NlehCCK in DMSO/water. A) ribbon drawing of the peptide backbone; B) space fRing model

838

CCK-peptides at high TFE concentrations are very similar to those of gastrin peptides of similar size both regarding the overall pattern and the intensities of the maxima. The latter spectra were attributed to y-tums, and a folding of the Cterminal portion of gastrin into a helical structure in TFE had been confirmed by NMR analysis (67). The presence of the identical C-terminal pentapeptide in the CCK molecule makes it most reasonable to assume that the observed CD properties in aqueous TFE reflect a similar conformational state of this sequence portion in CCK, too. This would also compare well with the 3D structure determined for (Thr,Nle)-CCK-9 by NMR in aqueous DMSO since the relatively small conformational shift from a SIQ- to an a-helix could easily result from the strong a-helix inducing effect of TFE as solvent (80,81). In presence of negatively (SDS) and positively charged (CTAH) surfactant micelles a significant induction of ordered structure for the (Thr.Nle)-CCK-9 analog was detected by CD measurements (64), but the spectra were different from those in aqueous TFE. The location of the negative maxima are more consistent with ptype structures. In presence of the neutral octyl-P-D-glucop5n:anoside micelles again a transition from unorderd to ordered structure is induced confirming strong interactions with the uncharged micelles, and the resulting CD pattern compares well with that obtained in aqueous TFE. This would suggest that differently from what has been found for the homologous peptide gastrin, the CCK-peptides exhibit a pronounced tendency to assume various ordered conformations depending upon the physicochemical environment.

Antisera

CCK-12

CCK-10

gastrin

INlelSj-fiastrin-17

lNlelSj-gastnn-17

Anti-CCK-10

1.6x10-^2

1.7 X 10-^2

2.3 X 10'^2

1.5 X 10-12

3.0 X 10-12

6.0 X 10*12

1.8x10-^2

2.0 X 10"^2

2.6xl0'12

3.0X10-12

2.4 X 10*12

5.0 > 10-12

fiastim-(14-171

(2725) Anti-CCK-13 (2795)

Tables.

Specificity of anti-CCK antisera raised with CCK-10 and CCK-13 iso-lcytochrome c corgugates in guinea pigs as determined by competition assays using EUSA techniques. Crossreactivities of CCK- and gastrin peptides are expressed by the respective IC^Q values using CCK-12 as coated antigen.

Additional support for a folding of the CCK-peptides into a conformation similar to that of the C-terminal portion of gastrin with the N-terminal tail as flexible arm for a decisive electrostatic interaction with a receptor counterpart was obtained with immunological experiments (82,83). CCK-10, CCK-12 and CCK-13 were linked covalently to the cysteine residue 107 of iso-l-cytochrome c via the

839 maleimide/thiol reaction principle. These conjugates were used In immunization experiments in order to examine the effect of an increased spacing of the bioactive core of the CCK-molecule from the carrier molecule on the specificity of the antisera raised in guinea pigs. Antisera were obtained which were uncapable of discriminating CCK-peptides from the gastrin peptides. Irrespective of whether the C-terminus is recognized by the immune system in its preferred conformation or not, it represents the identical continuous epitope as present in gastrin (table 3).

2 A. Interaction of Gastrin and CCK with Lipid Bilayers

Both aqueous organic solvent mixtures and differently charged micelles can mimic only roughly the environment of natural cell membranes. In order to analyze in more appropriate model systems possible interactions of gastrin and CCK with cell membranes and to determine their conformational states in lipid bilayers, we have recently investigated in detailed manner this aspect using liposomes. The similarity betwen liposomes and natural membranes is extensively exploited both in vitro and in vivo because of the ability of liposomes to mimic the behaviour of natural membranes. Moreover, the value of liposomes as model membrane systems derives from the fact that they can be constructed with natural constituents. In our approach, we selected as model membranes those formed with the zwitterionic lipids di-myristoylphosphatidylcholine (DMPC) and di-palmitoylphosphatidylcholine (DPPC) as these lipids constitute the major components of most cell membranes. Moreover, in order to operate with a simple system, small unilamellar vesicles (SUVs) were used, i.e. with a diameter between 25 and 250 nm as resulting by rod-type sonication or by extrusion (51). A possible interaction of gastrin with the model membranes at peptide/lipid ratios of up to 1:100 has been investigated by monitoring conformational changes via CD spectroscopy, insertion of the aromatic chromophores of the peptides into more hydrophobic compartments of the bilayer via fluorescence measurements and by analyzing changes of the thermotropic behaviour of the lipid bilayer. All assays indicated the absence of a detectable interaction of the negatively charged gastrin molecule with the zwitterionic bilayer (84). This result was not unexpected in view of the lack of insertion of gastrin into negatively charged or neutral micelles at neutral pH values as reported in section 2.2, but it contrasts the findings of Schwyzer et al. (85) obtained by ATR-FTIR on lipid films with incorporated gastrin. The forced interaction of gastrin in lipid films which led to propose a perpendicular insertion of the C-terminal tail of gastrin in a-helical conformation into lipid bilayers, may be too artificial, since in our more natural two-phase model no interaction with the bilayer accompanied by conformational

840

transitions could be detected. Our findings disagree also with those obtained with pentagastrin, i.e. Boc-p-Ala-Trp-Met-Asp-Phe-NH2 (86-88). An insertion of this molecule into lipid bilayers has been well documented by various spectroscopic methods related to the tryptophan moiety, but it cannot be excluded that the observed lipid affinity is mainly dictated by the hydrophobic "non-gastrin" portion adjacent to this aromatic residue. A significantly different behaviour was observed for (Thr.Nle)-CCK-9 which interacts transiently with the lipid bilayer of DMPC SUVs according to highsensitivity differential scanning calorimetry (hs-DSC) measurements. It causes fusion of the vesicles and is then expelled because of the higher degree of order of the bilayer in larger vesicles (84). These findings raise the question of whether a similar physical phenomenon is implicated in the known fusion of neurotransmitter vesicles with the membrane of the synaptic junctions with concomitant release of the peptides (89.90), since CCK peptides are known to act as neurotransmitters. The observed fast expulsion of the CCK from the lipid bilayers and the absence of detectable insertion of gastrin into DMPC bilayers should not exclude interactions of these peptides with natural membranes, as these are known to be structured in domains of different lipid composition and of differentiated lipid bilayer packing. In order to overcome the problem of negligible interaction of the peptides with model bilayers and to force an insertion of the peptide hormones into lipid bilayers, lipo-derivatives of the peptides were synthesized. 3 . Lipophilic Derivatives of Gastrin and CCK

3.1, Design and

Synthesis

Nature is anchoring biomolecules to cell membranes with single transmembrane helices, helix bundles or sticky fingers. Among the lipophilic derivatives the most widely occurring forms are fatty-acylated amino groups, hydroxyl groups or cysteine thiol functions, isoprenylated cysteine thiol functions as well glypiated carboxyl functions. In order to assure optimal interdigitation of artificial sti-^ky fingers with membrane bilayers and tight entrapment of lipo-peptide derivai es in bilayer structures we have proposed the use of di-fattyacyl-glycerol moieties (91). These handles are very similar to the structure of the natmal glycosylphosphatidylinositol anchor shown in fig. 10. As the chirality of the glycerol moiety is only marginally affecting the packing of bilayers (92-94), the synthetic efforts were confined to the preparation of roc-di-fattyacyl-glycerol congeners suitable for a selective conjugation to peptides and proteins (95).

841

,-v.

Protein

. ?

o-p=o

0 1 6Mana1

±Qalo1-

±Gaia1

F\Q, 10.

^ 2Mana1

^2Gaia1

y

2Gata1

X .6 Mana1

4GicNH^1

6myt>-4no8itol1

I

o 1 o=p-o' 1

?

, ,

Structure of the glycosylinosttolphosphate structure used by nature to anchor proteins at the C-terminus to cell membranes.

Similarly to what nature is using, the thiol function has become a main target in conjugate chemistry as it allows selective crosslinking of molecules via disulfide or sulfide bonds exploiting mild thiol disulfide interchange or thiol addition reactions. For latter reaction type, the maleimide group as thiol acceptor (96) has found widespread applications. It was shown that this group is sufficiently stable under certain conditions of peptide synthesis to enable its incorporation at preselected peptide chain positions and thus, to represent an ideal anchor for subsequent covalent linkage of thiol-functionalized molecules (65,82,97-101). Correspondingly, symmetric and unsymmetric rac-1.2-di-fattyacyl-3mercaptoglycerol derivatives were synthesized in good overall yields following scheme 1 and 2. Both in the case of gastrin (53,97) and CCK (64,102) it was known that Nterminal modifications do not affect their bioactivity profile. Correspondingly, the N-terminus of gastrin and CCK was used for grafting the lipid moiety via the thiol/maleimide approach. For this purpose p-alanine was chosen as spacer of the maleimide group since the methylene moiety allows for sufficient flexibility without displacing too much the peptide chain from the double-tailed lipid. This fact was expected to allow more appropriate mimicry of natural lipids and thus, a better interdigitation with lipid bilayers.

842

r-\

HN-CQz-N

O

f-Bu-S-N -COfe-N

O O

,—OH I—OH

-

RC02H/DMAP/DCC r.L. 12 h

HO—1

1NNaOHA3ioxane r.L, 12h

HOH

92%

85-08%

I—S—S-^-Bu

-SH

L—S—S-f-Bu

O

X

BU3P/CF3CH2OH r.L. 12 h >95%

^SH

Scheme 1. Synthesis of symmetric l,2'di-fattyacyl'3'mercaptoglycerol derivatives

HO-n HO—

Trt-CIAoluene Py, 60'C. 20h

R'CO,H DMAP/DCC r.t, 12h

Trt-

HOH

69% •—S~S-f-Bu

95% L—S-S~f-Bu

RÔ,H/DMAP DCC.r.t., 12h

ZnBr^CH,CU MeOH, r.t., 6min 85%

Trt~0-

*"v°i

85-90% -S-S~f-Bu

O

- S - S - • -Bu

Rl

V

R'^0^ L—S~S-f-Bu

BUjP/CFjCHjOH/ MeOf-Bu/HjO r.t., 12h >95% O

Scheme 2. Synthesis of derivatives.

^SH

imsymmetric

l,2-di-fattyacyl-3'mercaptoglycerol

843

AoH

DMF.r.t. 1.5h

H2N^-ÔH

o

*

b HOSU/DCC 0°Ctor.t.4-24h

°"tir'^'p O

OH

OSu

p p

i

+

H-peptide

-*- [I

bsu

b

O

FTÔ

R*^ " O H

+

N-(CH2)2-CO R^

DMF, r.L, 30 min 75-85%

,

• N-(CH2)2-COR,

-SH

Scheme 3. Synthesis reaction.

N-(CH2)2-COR,

of lipo-peptide derivatives via the maleimide thiol addition

O CH3-(CH2),TC-O-6H

O

CHz-S

v-/

-(CH2)2-ii-Peptide

Peptide: Arg-Asp-TyrCSOaH^Thr-Gly-Trp-Nle-Asp-Phe-NHj Gly-Pro-Trp-Leu-(Glu)5-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH2

Fig. 11.

Chemical structure ofDM-CCK and DM-gastrin.

[Thr.Nle]-CCK-9 [Nle^VHG-[2-17]

844 The maleoyl-P-alanine N-hydroxysuccinimide ester was conveniently prepared following the procedure of scheme 3 (101,102), and was then used to convert [Nlei5]-gastrin-[2-17] and [Thr,Nle]-CCK-9 into the reactive maleimido-derivatives for the final coupling of the l,2-diacyl-3-mercaptoglycerols to produce the lipopeptide derivatives DM-gastrin and DM-CCK shown in fig. 11 (51,104).

3.2. Physical Properties ofLipo-Gastrin and Lipo-CCK

By covalent attachment of the double-tailed lipid moieties to the N-termlni of [Nlel5].gastrin-[2-17] and [Thr,Nle]-CCK-9 the highly charged hydrophiUc peptides were transformed into the amphiphilic compounds DM-gastrin and DMCCK which were expected to aggregate in aqueous media into micelles or vesicles. Whilst sonication of DM-gastrin produces a polydispersed system of vesicles, extrusion generates a monodispersed population of vesicles of the diameter size of the filter used and of surprisingly high stability as determined by light scattering measurements (51). Moreover, by freeze-fracture electron microscopy only unilamellar bilayers were detected. Although this does not exclude the presence of a few multilamellar vesicles, formation of monolamellar vesicles should be preferred in view of the bulkiness of the headgroup and of the large hydration shell of the hydrophilic peptide headgroup. Interestingly, the surface of the lipogastrin vesicles appeared rough if compared to that of DMPC liposomes (105). In the case of DM-CCK sonication and extrusion, but surprisingly even simple vortexing leads to clearing of the aqueous solution indicating a significantly differentiated behaviour of the two homologous Upo-peptides (104). Taking into account the size of the headgroups in terms of peptide length an opposite effect was expected. Moreover, even after extrusion the lipo-CCK system rearranges into a polydispersed population of vesicles, the light scattering of which does not exclude the presence of micelles, too (104). The different behaviour of the gastrin and CCK derivatives can only be explained on the basis of a different interference of the two headgroups with the packing of the lipid tails into bilayer structures. Fluorescence quenching experiments with iodide were expected to give first indications in this context as both peptides contain in the C-terminal portion of the molecule a tryptophan residue, and gastrin an additional one near the lipid-grafting position. As shown in table 4, a blue shift of the fluorescence emission maximum was observed for both lipopeptide vesicles. This indicates an enhanced hydrophobic environment of the tryptophan residues referred to that of the unmodified peptides in aqueous solution (51,84). Similarly, the Stem-Volmer quenching constant [k^] was found to be remarkably lower for DM-gastrin than for the unmodified peptide. The experimental value was lower than that expected for monolamellar

845 Peptide [Nlel5]-gastrin-17 DM-gastrin [Thr,Nle]-CCK-9 1DM-CCK (vortexed) DM-CCK (sonicated)

A;ijnax(nin) 0 4-5 0 5-6 5-6

*^v

5.89 1.78 2.97 1.20 1.62

i

Table 4. Fluorescence properties and accessibility of the tryptophan residues of [Nle^^]-gastrin-17 and the lipo-gastrin derivative (DM-gastrin), of[Nle,Thr]CCK-9 and DM-CCK to iodide quenching; ^) the relative error on kû amounts to 3%.

vesicles where the gastrin moiety should be statistically distributed on the inner and outer bllayer surface and thus, be accessible to the iodide quencher only to 50%. One tryptophan residue, most probably that located near the N-terminus. i.e. near the lipid moiety, must therefore be less accessible. As reported in table 4, similar results were obtained in the case of DM-CCK. The tryptophan residue of this molecule is located in the C-terminal portion, and an insertion of this residue into more hydrophobic compartments of the bilayer can occur only via chain reversal of the peptide headgroup. This differentiated behaviour of the two lipo-peptides is further confirmed by their dichroic properties in aqueous buffer. The CD spectra of [Nle^^l-gastrin-17] and DM-gastrin are shown in fig. 12. Whilst at low ionic strength interaction of the gastrin moiety with the bilayer structure allows for partial folding into ordered conformations, an increase of the ionic strength to more physiological values leads to exposure of the gastrin moiety to the bulk water in mainly random coiled structure (51). Thereby the shoulder around 215 nm could derive from P-type conformations involving aggregation, e.g. of the N-termini. Conversely, the CD spectrum of DM-CCK (fig. 13) even at high ionic strength reflects considerable contents of ordered structure of 7- and a-type turns (104) as present in the NMRstructure of [Thr,Nle]-CCK-9 in DMSO/water (79). Taking into account these indications as well as the ^H-NMR experiments on DM-CCK vesicles which allowed for an unambiguous assignment of an NOE between the indole group of the peptide moiety and the myristoyl-alkane chain of the lipid portion (84), MD simulations were performed in a CC^/water cell which mimics the cristalline L^phase of a phospholipid membrane (106). The results of these MD simulations showed that the y-tum at the threonine is maintaiined and that the C-terminus adopts an a-helical conformation with a superimposed p-sheet hydrogen bonding

846

JS

-10 h

196

206

ai6

225

Xtnm)

Fig, 12,

CD spectra of DM-gastrin (curve 1) cmd [Nle^^J-gastriri'l? (curve 2) in 5 mM phosphate buffer containing 100 rnM NaCl at pH 7,0.

Fig. 13.

CD spectra ofDM-CCK in 5 mM phosphate, 100 mM NaCl (pH 7.0 ) at a peptide concentration of 1.4-10-'^ M.

847 pattern (fig. 14). The helix axis of the C-terminal portion lies parallel to the interphase separating hydrophobic from hydrophilic residues, the only exception beeing the norleucine residue which points into the bulk water. The tryptophan is buried into the hydrophobic phase shielded by the phenylalanine

7]

/

^

Y

^

""^ vCV^^

^

A

%

\y

[y

Fig. 14. Stereoview of the energy-mintmized conformation of DM-CCK tn its averaged position in the biphasic water/CCU ceVL The tipper phase is water and the lower CCI4 .

residue. This hydrophobic clustering compares weU with the observed strong blue shift of the fluorescence emission maximum and with the reduced Stem-Volmer iodide-quenching constant described above. It also agrees with the NMR experiments where a distinct NOE was detected between the indole side chain of the tryptophan residue and the alkane chain of myrlstlc acid. Regarding the whole molecule, i.e. including the fatty acid chains, the chain reversal in the peptide moiety produces a conical shape which in view of the mismatch of the cross-areas between headgroup and fatty acid chains, if compared to those of DMPC. is expected to greaUy affect the order of the bilayers. This would explain the fast rearrangement of extruded DM-CCK vesicles into a polydispersed system. By analyzing the thermotropic behaviour of the two lipo-peptides with hs-DSC no chain melting transition could be detected above 5* C indicating that the vesicles

848 of DM-gastrin and DM-CCK are in the liquid state despite the double-tailing of the lipid moiety (104,107). This fluidification of the bilayers can be attributed to the presence of the hydrophilic and relatively large peptide chains that function as polar headgroups. It should not result from the racemic di-fattyacylthioglyceryl moiety as it is known that the configuration of the phospholipid is affecting only marginally the phase transition temperatures (92-94). The higher fluidity of the DM-CCK than that of DM-gastrin, as deduced from the easier clearing of the vesicle preparation, should derive from its conical shape, i.e. from the chain reversal with insertion of the C-terminus into the lipid bilayer. Conversely, in the case of DM-gastrin the peptide moiety is mostly exposed to the bulk water phase, thus allowing for a more ordered packing of the lipo-tails.

3.3. Interaction of Lipo-Gastrin and Lipo-CCK with Phospholipid Bilayers

A lipid transfer process from one vesicle population to another is related to an equilibrium that involves the redistribution of lipids between donor lipid bilayers and acceptor structures of non-equivalent chemical potential due to their differing lipid compositions. The mechanism by which spontaneous lipid transfer occurs between membranes has been extensively investigated in various independent laboratories (108-111). In general it has been well established that in intervesicular lipid transfer processes the relative fluidity of the donor vesicles is much more important than that of the acceptor bilayers (112). Therefore, the low phase transition temperature of the lipo-peptides was expected to facilitate their transfer to phosphatidylcholine vesicles as model of cell membranes. In fact, DMgastrin transfer to DPPC vesicles was found to proceed at high rates (107). Replacement of the myristoyl chain with the palmitoyl chain in the lipo-gastrin (51) reduces the rate of transfer in agreement with previous findings (110,112,113). Moreover, the experiments confirmed that the lipid transfer process is strongly favored by the fluid state of the acceptor vesicles as previously reported for similar experiments by Martin and MacDonald (114). Upon insertion of the DM-gastrin into DPPC SUVs the sharp peak at 40.5" C broadens to give a large peak of low intensity with unchanged phase transition temperature indicating a statistical insertion of the lipo-gastrin molecules into the phospholipid bilayer (107). The transfer of DM-CCK to DMPC SUVs occurs rapidly and quantitatively even below the phase transition temperature of DMPC (104). Differently from what was observed in the case of lipo-gastrin, the endotherm of the system DM-CCK/DMPC exhibits a broad peak at 24.97° C corresponding to the phase transition temperature of the DMPC bilayer with statistically inserted lipo-CCK molecules

849 and two additional peaks at 20.24" C and 18.42° C, respectively. These could correspond to differently enriched CCK-domains. In order to confirm this working assumption the effect of Ca^^ ions on the phase transition temperatures of this merged system was analyzed (104). Reduction of the electrostatic charge of the lipid head groups, as a result of Ca^* binding, is known to induce bilayers to condense increasing the packing density in the gel phase and. thus, to raise the phase transition temperature. Upon addition of Ca2+ to the system the overall pattern of the endotherm was retained, but the turbidity of the solution increased visibly as a result of charge neutralization and thus, liposome aggregation. By depleting the bilayer surface of water, Ca2+ was found to cause a parallel shift of the transition temperatures of the three peaks to higher values. In particular, an increase of the transition temperatures of 5" C was observed for the two peaks at lower temperature whereas only a 3 ' C temperature increase was observed for the DMPC-rich domain. This confirms the presence domain structures in the merged vesicles. It has been clear for many years from phase diagrams of simple lipid mixtures that lipids with different head groups or acyl chains mix non-randomily and form clusters (115,116); this has been recently confirmed by Ca2+ binding measurements (117). Taking into account the strong mismatch of the cross-areas of the headgroup and of the lipid portion of DM-CCK, its non-ideal mixing with DMPC, i.e. formation of differently enriched DM-CCK domains, becomes highly favored. The fact that a similar phenomenon was not observed for the lipo-gastrin, has to be attributed to a different structure of its peptide portion, i.e. to its protrusion into the bulk water without interference with the interdigitation of its lipid portion with the acceptor bilayer. A conformational study was performed on the lipo-peptides inserted into lipid bilayers by CD in order to prove this working assumption (104,107). To assure quantitative transfer of the lipo-peptides to the DMPC SUVs and thus, to operate in an homogeneous population of merged vesicles, dye leakage experiments were performed on DM-gastrin vesicles with entrapped carboxyfluorescein by adding increasing amounts of DPPC vesicles (107). At a lipo-gastrin/DPPC molar ratio of 1:50 quantitative dissolution of the DM-gastrin vesicles has occurred. As the lipopeptide transfer to DPPC vesicles was found to proceed at significantly lower rates than in the case of DMPC, this ratio as well as incubation above the phase transition temperature of DMPC were used to prepare the samples for the CD measurements. As shown in fig. 15, the overall CD pattern of DM-gastrin inserted into DMPC vesicles excludes the presence of ordered conformation at significant extents. By comparing the spectrum of DM-gastrin in DMPC vesicles with that of DM-gastrin (fig. 12) the negative shoulder around 215 nm is reduced to a weak shoulder. This spectrum reminds that of the parent gastrin molecule in aqueous buffer, thus suggesting that the gastrin moiety is exposed to the aqueous environment in mainly random coiled structure with minimal interactions with

850

Fig, 15. CD spectra ofDM-gastrin in 50 mM Tris adjusted to pH 7.0 with H3PO4 at a peptide/DMPC molar ratio of 1:100.

o

196

205

215

225

235

245

Xiran)

Fig. 16. CD spectra of a 1:25 molar mixture of vortexed DM-CCK vesicles and DMPC in 5 mM phosphate buffer, 100 mM NaCl (pH 7.0) after 12 h incubation at 30'C.

851 the phosphatidylcholine headgroups. Conversely, the CD spectrum of DM-CCK transferred to DMPC bilayers (fig. 16) is different from that of the CCK peptide headgroups in DM-CCK vesicles (see fig. 13). The red-shift of the negative maximum to 214 nm and the crossover point at 202 nm is consistent with peptide-peptide interactions in p-sheet type aggregates. This would fully agree with the formation of CCK-rich domains in which the CCKpeptide portion retains its chain reversal with insertion of the C-terminus into the bilayer structure and thus, the tryptophan residue into more hydrophobic compartments of the bilayer as confirmed by the additional blue shift of the fluorescence emission maximum. The chain reversal would lead to full exposure of the highly charged peptide portion Arg-Asp-Tyr(S04H) as head group to the water phase. This picture of the CCK-headgroup would also agree with that obtained by MD simulations in the CCl4/water box.

4 . Ca^-*- Binding of Gastrin and CCK in Membrane Environments

The different efi'ect of Ca2+ ions on the thermotropic behaviour of the CCK- and DMPC-rich domains in the fused DM-CCK/DMPC vesicles (see section 3.3) suggests a higher Ca2+ affinity of the CCK head groups than that of the phosphatidylcholine groups. Several peptide hormones have been shown to exhibit high Ca2+ affinity in membrane-mimetic conditions (118-122). This led to propose the Ca^+Zpeptide complexes as the bioactive states of the hormones (123). Previous findings that gastrin is capable of complexing up to three Ca2+ ions in TFE as monitored by CD changes in the near and far uv (118) led us to investigate into more details Ca^"^ binding to gastrin and CCK peptides in various media as well as their ionophoretic activities (124, 125). The affinity of peptides for cadcium can be determined by monitoring possible conformational changes induced by complexation of the metal ion or by measuring the energy transfer phosphorescence of Tb^"*". Lantanide ions are known to replace Ca2+ without causing structural modifications in proteins as both metal ions exhibit a strong propensity for oxygen donor groups, very similar ionic radii, lack of directionality in binding donor groups and an apparent variability in the coordination number (126). As shown in fig. 17, addition of Tb3+ to CCK leads to a decrease of the fluorescence emission maximum of the tryptophan residue with concomitant increase of the Tb^"*" phosphorescence which allows for titration of the peptide metal ion binding sites under various conditions. The results of these experiments are summarized in table 5. For [Thr,Nlel-CCK-9 neither Ca2+ nor Tb3+ binding could be detected in aqueous solution, whereas for gastrin a biphasic titration curve was obtained which

852

0.00

350

Fig. 17.

450

550

Fluorescence emission spectra ofDM-CCK (6 juM, 10 mM MOPS, pH 7,0, 30° C] in absence (curve o) and presence (curve b) ofTb^'^, Spectra were recorded with 284 nm exitation. Terbium emission was monitored at 549 nrru

exhibited a first plateau at a Tb^+/peptide molar ratio of 2 followed by a continuous increase to a maximum of six metal ions corresponding to the six carboxylate functions present in the gastrin molecule. By measuring CD changes in the near uv a Ca^+Zgastrin ratio of 1.5 was obtained indicating that Ca2+ binding is enhancing the rigidity of one of the two tryptophan side chains. In TFE the CCK-peptide shows affinity for one metal ion, whereas in the case of gastrin conformational changes are indicating a maximum of three and energy transfer phosphorescence a maximum of two metal ions bound to the peptide.

Peptide

[Nlel5]-gastrin-17 1 |Thr,Nlel-CCK-9

lOmM MOPS.

lOmM MOPS, pH 7.0; Ca2+

98% TFE Tb3+

98% TFE

pH 7.0; Tb3+ 2 (max. 6)

1.5

2 (or 3with CD)

3.0

-.-

-.-

1.0

1.0

Ca2+

DM-gastrin (SUVs, 0. IM

2.0

2.0

n.d.

n.d.

NaCl)

1.75

n.d

n.d.

n.d.

DM-CCK (SUVs: O.IM NaCl)

Table, 5. Metal ion binding affinities of gastrin and CCK and their lipo-derivatives as determined by CD (Ca^'^ binding) and luminescence (Tb^'^ binding).

853

B

Fig. 18, Low-energy calcium binding sites of [Nle^^hgastrin'15-17] (A) and IThr,Nle]-CCK'9 (B) as calculated with the GRID programme using the peptide coriformations determined by A/MR analysis in aqueous organic media.

854 As discussed in sections 2.2 and 2.3, for both gastrin (62) and CCK (79) the preferred conformational states in aqueous organic media have been determined by iH-NMR analysis. Using these 3D structures the potential Ca2+ binding sites have been calculated with the GRID program, and the results are shown in fig. 18. Both peptides show a high preference for a C-terminal binding site involving the side chain carboxyl function of the aspartic acid and the carbonyl group of glycine. In the case of gastrin a second low energy metal binding site was detected in the penta-glutamic acid sequence. In fact, gastrin has a preferrence for binding more than one metal ion whereas CCK tends to form a 1:1 complex. As shown in fig. 19, binding of one Ca2+ ion to the CCK peptide leads to a significant enhancement of the dichroic intensities and thus, to an increased content of ordered structure. This observation allows for a modelling of the 3D structure of the [Thr,Nlel-CCK/Ca2+ complex which is reported in fig. 20. Conversely, binding of Ca2+ to gastrin in TFE provokes an opposite effect (118), i.e. a collapse of the ordered conformation, as weU assessed by the changes in the dichroic properties in the far uv. In the aggregated states of the lipo-peptides. i.e. in the corresponding SUVs, the peptide headgroups of the inner layer should not be accessible for Ca^* ions and correspondingly metal ion/peptide molar ratios of 1.5 (or 1):1 for DM-gastrin and 0.5:1 for DM-CCK were theoretically expected. The experimental ratios were for both lipo-peptides significantly higher as reported in table 5. This can be

Fig. 19.

CD-spectra of IThr,Nle]-CCK-9 in aqueus TFE (—), in presence of 1 eq.

855 attributed to the fact that the vesicles are leaking or that transbilayer flip-flop is taking place at high rates. Both processes are possible because of the high fluidity of the vesicles. Additionally, in the case of DM-CCK intermolecular ion complexation involving the aspartic acid residue located in the N-termlnal portion has to occur in order to explain the high molar ratio of 1.75. Intermolecular Ca2+ complexation and thus aggregation of aspartic acid-containing peptides in lipid bilayers has been reported (127). Binding of Ca2+ to DM-CCK is again stabilizing the conformation of the peptide head group, whereas an opposite eflfect was observed in the case of DM-gastrin. Finally, to simulate the conditions of peptide hormones in a cell membrane-bound state, the effect of Ca2+ on the merged lipopeptide/DMPC vesicles was examined. At 1 to 2 mM Ca2+ concentration the CD patterns of both lipo-peptides were very similar to those of the peptide headgroups in the pure lipo-peptide vesicles, but the intensities of the dichroic bands were lower. Lower dichroic intensities have been attributed, e.g. in the case of a-helices, to the formation of bundles (128). Similarly, the lower intensities revealed in the present case can reasonably be attributed to a clustering of the lipo-peptides into domains as favored by intermolecular Ca^* complexation. These findings, besides confirming the experimental results of the microcalorimetric measurements, at least in the case of CCK, could be of biological relevance

Fig. 20,

Model of the energetically most favored Ca^^/IThr,Nlel'CCK'9 complex using for calcium a van der Waals radius of 1.95 A.

856 in terms of facilitating an accumulation of peptide (neuro)hormones at the cell membrane surface. Recent studies of Ananthanarayanan (122) have shown that various peptide hormones are capable of inducing Ca2+ influxes into phosphatidylcholine vesicles. The observed affinity of gastrin and CCK as well as of their lipoderivatives for Ca^^ in membrane mimicking environments led u s to examine rates of Ca2+ influxes induced by these hormones. Calcium ion fluxes mediated by a variety of channels and ionophores into liposomes and cells have been studied by loading the vesicles with indicator dyes like arsenazo III (129) or quin-2 (130). The significantly higher calcium affinity of the fluorescence indicator fura-2 (131) was a major advancement in the detection of Ca2+ concentrations in small cells and liposomes (132-134).

- |

1

j -

-J

I

I

r 3

I L Ca**

20

Fig. 21.

I

L_

40 60 Incubation time (mln)

80

Dependence of the Jluorescence ratio on incubation time after the addition of lThr,Nle]-CCK-9 (0), [Nle^^l-gastnn (A), DM-CCK (x ) and DM-gastrin (O). The buffer is used as blank (+).

Although [Nlel5]-gastrin-17 and [Thr,Nlel-CCK-9 are capable of binding calcium ions in TFE, they were unable to induce Ca2+ influxes into DMPC vesicles as shown in fig. 2 1 . This fully agrees with the findings from microcalorimetric and CD measurements which excluded major interactions of these peptides with the DMPC bilayer. Conversely, the induced lipid interaction of their lipophilic derivatives DM-gastrin and DM-CCK provoked ion influxes with full equilibration

857 of the system after more than Ih. The rate is similar to that observed for other peptides (122), but significantly lower than that of ionophores. It has recently been reported that interaction of peptides with lipid bilayers leads to strong perturbation of the fatty acid chain packing which markedly increases the rates of transbilayer flip-flop of lipid monomers (135). Therefore, the observed ion influxes should derive mainly from relatively fast flip-flop processes of the lipopeptides with bound csilcium ions, even if calcium ions are known to increase the packing of lipid bilayers. Both the rate of ion influxes and the most probable mechanism exclude that the observed effect is of physiological relevsmce. Moreover, such ionophoretic activity of bioactive peptides not mediated and restricted by the receptor recognition would represent an unspecific activity, irrespective of the cell encountered by the peptide hormone on its endocrine pathway. Thus, hormone-stimulated Ca2+ release from membrane pools has to occur mainly as a result of increased cytosolic inositol(tris)phosphate concentrations as induced by the specific hormone receptor interaction at the target cell (136,137).

5. Biological Properties of Lipo-Gastrin and Lipo-CCK

CCK and gastrin peptides are known to exert their physiological function via two receptor subtypes, i.e. the CCK-A receptor, mainly located in the pancreas and selective for the sulfated forms of CCK peptides, and the CCK-B receptor, widely distributed in the central nervous system and in the gastrointestinal tract, which recognizes CCK and gastrin peptides with similar affinities independently of their state of sulfation. These two receptor subt3T>es have been cloned and expressed in COS-7 cells (138,139). A comparison of the sequences of the CCK-A and CCK-B receptors showed 48% identity as expected for receptors within the same family (139). In analogy to other members of the GPCR superfamily seven putative transmembrane segments were identified which are connected by extra- and intracellular loops. Cysteines in the first and second extracellular domains are conserved in both receptors and may be involved in disulfide bridges as required for the stabilization of these domains. Moreover three potential asparagine-linked glycosylation sites are identified in the N-terminal domain. Great efforts are presently paid to identify the CCK and gastrin binding sites of these two receptors. The biological functions of the lipo-gastrin and lipo-CCK peptides were analyzed on these two receptors of known sequence using rat pancreatic acinar cells and the tumoral rat pancreatic acinar cell line AR42J. The CCK-A receptor has been thoroughly characterized in the rat pancreatic acinar cells (140-142) and in its

858 recombinant form expressed on COS-7 cells (139.142). Most of the receptors (80%) exist in the veiy-low-affmity state and only a small percentage (1%) in the high-affinity-state (142). The ability of the receptor to exist in three different states is an intrinsic property of this CCK-A receptor. The CCK-B receptor has been well characterized pharmacologically in gastric mucosal cells (143-146), in the recombinant form of transfected COS-7 cells (139) as well as in the AR42J cell line (147). In contrast to the normal rat pancreas the AR42J cells contain a majority (at least 80%) of high-affmity CCK-B/gastrin receptors and a minority of CCK-A receptors. Binding of DM-gastrin to CCK-B receptors on AR42J cells was determined in competition assays using as radioligand 125i-BH-[NlelS]-gastrin-[2-17] and compared to the parent [Nle 15]-gastrin-17 (51). As shown in fig. 22, under standard conditions the lipo-gastrin exhibits a 7-fold lower binding affinity than the immodified gastrin. Binding of DM-CCK to isolated rat acinar cells using l25i_ BH-[Thr,Nle]-CCK-9 as radioligand (fig. 23) led to very similar results, i.e. 5-fold lower affinity (84). Although N-terminal modification of gastrin and CCK should not affect the receptor recognition and binding capability, incorporation of the highly lipophilic di-fattyacyl moity leads to self-aggregations of the lipo-peptides in aqueous solutions at ionic strengths similar to those of the binding assay media (51,104). The mode of presentation of the lipo-hormones to the membranebound receptors is, therefore, completely different from that of the parent peptides. In fact, even if the characterization of the aggregational states of DMgastrin and DM-CCK has been performed at dilutions up to lO'^M, the stability of

100

-11

-10

-9

concentration. lQg(M)

Fig. 22. Receptor hinding affinities of [Nle^^J-gastrin-l? (•) and DM-gastrin (O) using AR4-2J membrane preparations and ^^Î-BH'[Nle^^]-gastrin-[2'17] as radioligand.

859 the vesicles suggests the persistence of aggregated states in form of vesicles or micelles even at the dilutions of the binding assays (lO''^ - I O - ^ ^ M ) . Radioligand displacement by DM-gastrin and DM-CCK occurs in parallel mode to that obtained with the parent gastrin and CCK peptide, respectively. This indicates that the binding of the lipo-peptides is directly proportional to their concentrations, irrespective of their aggregational state; moreover, even the binding mode is apparently not affected by the modification. Within the incubation time of the binding assay the transfer of the lipo-peptides to the cell membrane should be quantitative according to the results of the model experiments discussed in section 3.3. Therefore the higher IC50 values obtained for the lipo-peptides can either derive from an intrinsic lower affmity for the receptor binding sites or from the induced lipid interaction and thus, restricted two-dimensional migration to the receptor, since an escape of the lipo-peptides into the extracellular water phase is energetically highly unfavored. Support for the second hypothesis derives from following observations. By strengthening the interdigitation of the lipid tail with the membrane bilayer and thus, lowering the rate of diffusion, e.g. with a di-palmitoyl tail, the IC50 value is increased (51). As shown in the model experiments discussed in section 3.3, the interaction of DMCCK with the membrane bilayer leads to strong perturbations and thus, the

-12

Fig. 23.

-11

-10 -9 concentxatlon,

-8 log(M)

Receptor binding affinities of IThr,Nle]-CCK-9 (O) and DM-CCK (•) after 45 min incubation with isolated rat pancreatic acini using ^^Î-BHIThr,Nle]-CCK-9 as tracer.

860 interdigitation of its lipid tail with the bilayer should be weaker than in the case of DM-gastrin. In fact, the receptor affinitiy of DM-CCK is higher than that of DMgastrin. Finally, the estimated rates of two-dimensional diffusion of lipids in bilayers are about one order of magnitude lower than those of a threedimensional diffusion of peptides of this size in water (84); this compares well with the experimental values. Correspondingly, longer incubation periods of the lipo-peptides with the cells should anneal the observed differences. The data of fig. 24 show that this is really the case as almost identical binding affinities were obtained for DM-CCK and its parent peptide confirming that induced lipid interaction is lowering significantly the association rate. Analysis of the data of fig. 24 showed that at binding equilibriimi both high and low affinity receptors were occupied by the lipo-CCK as it is known to occur for CCK peptides (148-151). There was, however, a significant 3-fold lower affmity of DM-CCK as compared to [Thr,Nle]-CCK-9 for the low affinity binding sites. Regarding functional binding, i.e. final biological response, a good con: i.-îon between binding and potency was observed for DM-gastrin with a siigi/ J ^ enhanced efficacy of the lipo-derivative compared to that of the parent gastrin hormone (107). In the case of DM-CCK again the CCK-typical up and down strokes of the dose-response curve related to amylase hypersecretion were obtained. Although the lipo-CCK showed the same efficacy as [Thr,Nle]-CCK-9 its potency was 100 times lower. Conversely, the potency of DM-CCK in stimulating increase of the IP3 concentration was reduced only by a factor 4 compared to the [Thr,Nlel-CCK-9, but again the efficacy of both ligands were identical.

-11

Fig. 24.

-10 -9 coxicentratlozi.

-8 log(M}

Receptor binding affinities of IThr,Nle]-CCK'9 (O) and DM-CCK (•) after 3 h incubation with isolated rat pancreatic acini using ^^Î-BH-VThr^Nle]CCK'9 as tracer.

861 CCK is generally believed to stimulate secretion of digestive enzymes from rat pancreatic acini by activating phospholipase C which hydroljês the membrane lipid phosphatidylinositol biphosphate with release of inositol(tris)phosphate (IP3) and diacylglycerol, increase of C5ôsolic calcitim and activation of kinase C (148,150,152-154). These CCK actions are mediated by three states of the CCK receptor. Of these three CCK receptor states the high- and low-affinity states are involved in the competitive binding experiments as the very-low-affinity state cannot be identified by binding with the CCK-radioligand (141). The low-affinity state mediates the upstroke of the amylase dose-response curve and the veiy-lowafflnity state its downstroke. The biological data obtained with DM-CCK and [Thr,Nle]-CCK-9 indicate a similar amplification factor between binding to the functional low-affinity receptors and IP3 production for both peptides suggesting that the coupling of the G-protein (Gq) between receptor and phospholipase C was equivalent. However, the amplification factor between IP3 formation and amylase secretion was 27 times lower with DM-CCK than with IThr,Nle]-CCK-9. Therefore the difference in amplification between IP3 formation and the final biological response was responsible for the difference in amplification between binding and amylase secretion (a 35-fold factor between DM-CCK and unmodified CCK peptide). Stimulus secretion coupling in pancreatic acini involves several G-proteins (152). It is conceivable that DM-CCK occupies low-affinity receptors in such a way that they are poorly coupled to G-proteins different from the Gq, but contribute to the normal response. The experimental data concerning receptor binding affinities and final biological responses obtained with the lipo-derivatized gastrin and CCK aUow for interesting conclusions to be drawn. Despite the tight anchorage of the lipo-peptides to the membrane bilayers, binding to the receptor is occurring. This binding is timedependent and equilibration of the systems is obtained after longer periods of time than with the parent peptide hormones. This phenomenon can be correlated to the lower rates of two-dimensional migration of the lipo-peptides in cell membrane bilayers than those of the underivatized peptides. These are free to diffuse in three-dimensions to the receptor in the extracellular space or to float on the membrane surface with more or less pronounced interactions with the lipid headgroups or with more hydrophobic compartments. The generally accepted model of the spatial structure of the receptor, however, raises immediately the question of how a membrane-bound ligand can find its way to the receptor binding site. As an escape of the lipo-peptides into the extracellular water phase is energetically highly unfavored, the binding site (or sites) are reached by lateral penetration at the lipid/water interphase where the spectroscopic measurements in model systems are locating the lipo-CCK and

862 lipo-gastrin in a more or less conformationally ordered form. This would imply a remarkable flexibility and mobility of the boundary domains of the receptor, particularly, in view of the large size of the peptide ligand. A lateral penetration of the entire lipo-peptide molecule, i.e. including the lipid tail, is difficult to rationalize as the helix boundle of the receptor represents a tight assembly which precludes diffusion of membrane lipids into its core structure in order to maintain its three-dimensional assembly Consequently, it should therefore preclude also a penetration of the lipo-tail of DM-gastrin and DM-CCK, and the lipo-peptides should approach the receptor with the tail inserted into the lipid bilayer and then protrude into the binding cleft across the extracellular loops.

Peptide

PepUde:

Fig, 25.

^

^

,

Arg-Asp-TyT(S03H)-Thr-Gfy-Trp-Nle-Asp-Phe-NH2

rrhr.Nlel-CCK-9

Gly-Pro-Trp-Lcu-Glu-{Glu)4-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NH2

INle ^ ^l-gastrln-17

Schematic representation of DM-CCK, showing the spacer (bold) and the lipo moiety inserted in the membrane.

In the case of the extensively studied CCK-A receptor, besides heterogeneity in its affinity states (139,142), heterogeneity in binding sites regarding agonists and antagonists have been detected (156,157). This fact could suggest a ligand binding process of a cascade-like dynamic for an optimal signal transduction, i.e. G-protein coupling. If these dynamics are hindered by the pivot-like restriction of the lipid tail, the binding can still occur at full extents, but its functionality may be impaired. In the present case, the different results obtained with DM-gastrin and DM-CCK in the functional binding could then be rationally explained by the different length of the spacer between site of anchorage to the bilayer and

863 bioactive core of the hormones. As shown in fig. 25, the fully active sequence of CCK is spaced from the lipid-tail by a dipeptide, whilst in DM-gastrin a pentapeptide chain is doing this job. This working assumption could give the tools for investigating an important open question of the hormone-mediated signal transduction, i.e. whether binding of the hormone to a first recognition site is followed by a switch to a second functional binding site located in inner compartments of the helix boundle, as proposed for the acetylcholine receptor (16, 159, 160).

6. Peptide and Non-Peptide Antagonists

The wide range of physiological responses which have been attributed to CCK-A and gastrin/CCK-B receptor-mediated hormonal messages have stimulated the search for agents which mimic or block the action of gastrin and CCK. As there were opportunities for drug discovery in the areas of analgesia (CCK-B antagonist), anxiety (CCK-B antagonist), drug dependency (CCK-B antagonist), memory (CCK-A agonist), parkinsonism (CCK-B antagonist) and psychosis (CCKA agonist), high affinity peptide and non-peptide antagonists which are capable of distinguishing between these two receptor subtypes have been developed (for recent reviews see ref. 33, 161-164). The availability of antagonists has offered new tools with which to explore the role of CCK and gastrin in periphery and brain. Among the antagonists only few peptide structures were found with deletion sequences, C-terminal variations of the parent peptide sequences or by backbone modifications. More successful were non-peptide antagonists based on benzodiazepine and peptoid structures (see fig. 26 for selected examples of CCK-A and CCK-B antagonists). High binding potencies and receptor selectivities have been realized with nonpeptide ligands which differ distinctly in their chemical structure from the natural ligands. Moreover, they differ structurally from one another, too. This fact is difficult to rationalize in terms of identical binding sites at the receptor level and of mimicry of the natural ligands in their bioactive conformation. Nevertheless based on the working assumption that agonists and antagonists should exhibit a correspondence among the pharmacophoric groups, attempts have been made to identify the bioactive conformations of gastrin and CCK by conformational energy analysis and selection of the low-energy conformers of acetylated CCK-7 and gastrin-(14-17) by their structural similarity with the two most studied antagonists L-364,718 (CCK-A antagonist) and L-365,260 (CCK-B antagonist) (165). This led to propose for CCK-B agonists an a-helical array in the C-terminus as determined experimentally in the studies discussed in sections 2.2 and 3.2 and a p-bend structure for the CCK-A receptor agonists. In the latter case

864 the p-bend induces the tyrosine and phenylalanine side chains to appoach one another with the tryptophan side chain pointing away from these two aromatic rings. This conformation differs significantly from that discussed in section 2.3 and 3.2. However, in the computed low-energy conformation of acetyl-CCK-7 with methionine-2 replaced by threonine as in the case of the fully active [Thr,Nle]CCK-9 analog, a sharp reverse turn between glycine and tryptophan is found, and the tyrosine and phenylalanine side chains lie far apart from one another in a manner more similar to the structure of the CCK analog in DMSO/water reported in section 2.3 (165). In this case a superimposition with the non-peptide antagonist L-364.718 does not allow to identify correspondence in the array of potential pharmacophoric groups.

CCK-A

CCK-B O

O

HO'

B

K^W^^ D

Fig. 26.

Structures of the CCK-A antagonists L364J18 (A) and lorglumide (B) and of the CCK-B antagonists L365,260 (C) and CI-988 (D).

In view of the most recent results which indicate distinct binding sites for agonists and antagonists on CCK-A and CCK-B receptors (156-158,166), this lack

865 of structural similarity is not surprising. However, at the same time an important method for verification of proposed bioactive conformations is lost. Surprising in this context was the finding that even peptide antagonists like H-Tyr(S03H)-NleGly-D-Trp-Nle-Asp-2-phenylethylester (167) occupy receptor binding sites which could not directly bind agonist ligands (157,158).

Conclusion

Double-tailed lipo-derivatization of the homologous peptide hormones allowed to bypass the problem of a partition of these peptides between the lipid and water phase of model lipid bilayers and of cell membranes, by shifting the equilibrium in great favour of the membrane-bound state. The main drawback of such an approach derives from the predetermined site of lipid anchorage. Therefore, the lipid moieties were grafted to the N-termini of two fully active gastrin and CCK analogs, as this portion of the peptide molecules is known not to be involved in the receptor recognition process. As expected from the resulting chemical constitution of the lipo-peptides this derivatization induces a spontaneous aggregation into (unilamellar) vesicles. By comparing the aggregational properties of the two lipo-peptides, the most striking observation was the pronounced effect of the peptide headgroups. In fact, replacing the sequence Gly-Pro-Trp-Leu(Glu)5-Ala-Tyr of [Nlel5]-gastrin-[2-171 with Arg-Asp-Tyr(S03H)-Thr of [NlcThrJCCK-9, but retaining the identical C-terminus Gly-Trp-Nle-Asp-Phe-NH2 substantial differences in the packing of the fatty acid chains and in the display of the peptide headgroups were revealed. Despite the sequence homology of the lipo-peptides a remarkably differentiated behaviour was also revealed in their mode of insertion and interaction with model bilayers and of their display at the water/lipid interphase. This would indicate that already at the level of the collisional events with the target cell membrane sequence-dependent characteristic properties are initiating the differentiation process for the selective receptor recognition process. Most striking were the findings that the CCK headgroup interacts with the lipid bilayer in an amphipathic helical array which strongly reminds that determined by NMR in aqueous DMSO. Thereby a pronounced tendency to cluster into domains was detected which was further enhanced in presence of calcium ions. This phenomen could be of physiological relevance as it should facilitate accumulation of the CCK hormone on the cell membrane with a preorientation and prefolding into bioactive conformations. Conversely, for the gastrin peptide even upon an induced lipid interaction a preferred ordered structure could not be detected. In all experimental models the

866 gastrin headgroup is exposed to the bulk water phase in randomly coiled structure, whereby even the presence of calcium is not enhancing the interaction of the peptide head group with the bilayer. All the data suggest that this peptide hormone may not be accumulated on the cell membrane since the zwitterionic lipids used in the model experiments represent the main constituent of natural membranes and acidic lipids present in cell membranes, too, are expected to prevent even more by electrostatic repulsion an interaction of the negatively charged peptide with the membrane. The effect of this repulsion was well assessed by comparing the behaviour of the gastrin in negatively charged micelles as mimicry of acidic cell membrane domains. Correspondingly, the preferred Ushaped conformation of gastrin as determined by NMR in TFE does certainly not represent its structure in membrane environments. The model experiments, however, do not exclude that this hairpin structure determined in aqueous organic media may be assumed by the gastrin peptide in the receptor-bound state. The very similar conformations of gastrin and CCK in the active site portion of the molecule correlates well with the biochemical properties in terms of identical recognition of both hormones by the CCK-B/gastrin receptors where the negative charges accumulated in the N-terminal, more or less flexible tail of CCK and gastrin may play a secondary role to increase the binding afBnity. For this purpose a specific conformation of the N-terminus seems not to be required, since in the case of the equally potent CCK peptide no preferred ordered structure could be detected in this portion of the molecule. The a-helical structure in the Nterminal portion of gastrin could therefore result from the TFE which is known t j be a strong a-helix inducing solvent (80,81). Then the increasing pc v 'ies of gastrin peptides in function of chain length would result from add lioixal electrostatic ligand/receptor interactions and the parallelism observed for the onset of a-helical conformation could be a fortuitous coincidence. On the other hand, the fully flexible N-terminus of CCK with an exact location of the tyrosineO-sulfate moiety plays a crucial role for the selective recognition by the CCK-A receptor. The whole body of data of these series of studies clearly reveals a surprisingly differentiated behaviour of the two homologous peptide hormones, with CCK showing affinity for membrane bilayers and gastrin not. However, independently of wether the peptide hormones are accumulated on the cell surface in their target cells, with the lipo-peptides it was definitely confirmed that a membranebound pathway in the mechanism of the hormone-receptor binding process as proposed by Schwyzer (23,24) is indeed possible. Thereby with the helical array of the CCK peptide parallel to the water/lipid interphase, the aromatic side chain ring of the phenylalanine residue is heading the structured molecule and could play a decisive role in the lateral penetration of the receptor assembly.

867 References 1.

2. 3. 4. 5. 6. 7. 8. 9.

10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

24. 25. 26. 27.

Lemieux, R.H., Delbaere, L.T.J., Beverbeek, H., Spohr, U. (1991) in: HostGuest Molecular Interaction from Chemistry to Biology, Ciba Foundation SjTiiposium, pp. 231-248, Wiley, Chichester. Fischer, E. (1894) Chem Bar. 27, 2985-2993. Richards, F.M. and Vithayathil, P.J. (1959) J. Biol Chem 234, 1459-1465. Tamburro, A.M., Scatturin, A., Rocchi, R., Marchiori, F., Borin, G. and Scoffone, E. (1968) FEBSLett. 1, 298-300. Rocchi, R. Borin, G., Marchiori, F., Moroder, L., Peggion, E., Scoffone, E., Crescenzi, V. and Quadrifoglio, F. (1972) Biochemistry 11, 50-57. Kartha, G., BeUo, J. and Harker, D. (1967) Nature 213, 862-865. Wyckoff, H.W., Hardmann, K.D., Allewell, N.M., Inagami, T., Johnson, I.N. and Richards, F.M. (1967) J. Biol Chem 242, 3984-3988. Moroder, L., Borin, G., Marchiori, F., Rocchi, R. and Scoffone, E. (1973) in: Peptides 1971 (Nesvadba, H., ed.) pp. 367-372, North-Holland, Amsterdam. Scoffone, E., Marchiori, F., Moroder, L., Rocchi, R. and Borin, G. (1973) in: Medicinal Chemistry III, Special Contributions (Pratesi, P., ed.) pp. 83-104, Butterworth, London. Garcia, K.C., Ronco, P.M., Verroust, P.J.. Brunger, A.T. and Amzel, L.M. (1992) Science 257, 502-507. Stem, L.J., Brown, J.H., Jardetzky T.S., Gorgia, J . C , Urban, R.G., Strominger, J.L. and Wiley, D.C. (1994) Nature 368, 215-221. Brown, J.E. and Klee, W.A. (1971) Biochemistry 10, 470-476. Bierzynski, A., Kim, P.S. and Baldwin, R.L. (1982) Proc, Natl Acad. Set USA 79, 2470-2474. Kim, P.S. and Baldwin, R.L. (1982) Nature 307. 329-334. OUveira, L., Paiva, A.C.M., Sander, C. and Vriend, G. (1994) TIPS 15, 170172. Wess, J. (1993) TIPS 14, 308-313. Hargrave, P.A. (1991) Curr. Opin, Struct Biol 1, 575-581. Burt, S.K., Hutchins, C.W. and Greer, J. (1991) Curr. Opin. Struct. Biol 1, 213-218. MerriU Jr., A.H. and Liotta, D.C. (1991) Curr. Opin. Struct. Biol 1, 516-521. Marsh, D. (1992) Curr. Opin. Struct. Biol 2, 497-502. Glaser, M. (1993) Curr. Opin. Struct. Biol 3, 475-481. Deveaux, P.F. (1993) Curr. Opin. Stmct. Biol 3, 489-494. Schwyzer, R.(1986) in: Natural Products and Biological Activities (Imura, H., Goto, T., Murachi, T. and Nakajima, T., eds.) pp. 197-207, Tokyo Press, Elsevier, Tokyo. Schwyzer, R. (1991) Bvopolymers 31, 785-792. Henderson, R., Baldwin, J.M., Ceska, T.A., Zemlin, F., Beckmann, E. and Downing, K.H. (1990) J. Mol Biol 213, 899-929. Schertler, G.F.X., Villa, C. and Henderson, R. (1993) Nature 362, 770-772. Baldwin, J. B. (1993) EMBOJ. 12, 1693-1703.

868

28. OUviera, L., Paiva, A.C.M. and Vriend, G. (1993) J. Comput. Aided Mol Design 7, 649-658. 29. Hibert, M.F. Trumpp-Kallmeyer,, S., Bruinvels, A. and Hoflack, J. (1991) Mol Pharmacol 40. 8-15. 30. Cronet, P., Sander, C. and Vriend, G. (1993) Prot Eng. 6, 59-64. 31. Dockray, G.L, Dimaline, R., Pauwels, S. & Varro. A. (1980) in: Peptide Hormones as Prohormones - Processing, Biological Activiies, Pharmacology (Martinez, J. ed.) pp. 245-284, Ellis Horword, Chichester. 32. Martinez, J. and Poitier, P. (1986) I7PS, 139-147. 33. Martinez, J. (1990) in: Comprehensive Medicinal Chemistry (Hansch, C , Sammer, P.G. & Taylor, J.B., eds.) Vol. 3, pp. 925-959, Pergamon Press, New York. 34. Gregory, R.A. and Tracy, H.J. (1964) Gut 5, 103-117. 35. Eysselin, V.E., MaxweU, V., Reedy, T., Wunsch, E. and Walsh, J.H. (1984) J. Clin. Invest 73, 1284-1290. 36. Carter, D.C., Taylor, I.L., Elashoff, J. and Grossman, M.I. (1979) Gut 20, 705-708. 37. Moran, T.H., Robinson, P.H., Goldrich, M.S. and McHugh, P.R. (1986) Brain Res. 362, 175-189. 38. Jorpes, J.E., Mutt, V. and Toczko, R. (1964) Acta Chem. Scand. 18, 24082410. 39. Mutt, V. (1980) in: Gastrointestinal Hormones (Glass, J.B.J., ed.) pp. 169221, Raven Press, New York. 40. Tracy, H.J. and Gregory, R.A. (1964) Nature 204, 935-938. 41. Morley, J.S. (1968) Proc. R. Soc. Ser, B 170, 97-111. 42. Jensen, R.T., Lemp, G.F. and Gardner, J.D. (1982) J. Biol Chem. 257, 55545559. 43. Gillessen, D., Trzeciak, A., Muller, R.K.M. and Studer, R. (1979) Int. J. Peptide Protein Res. 13, 130-136. 44. Wunsch, E., Moroder, L., Gillessen, D., Soerensen, U.B. and Bali, J.-P. (1982) Hoppe-Seyler's Z. Physiol Chem. 363, 665-669. 45. Kenner, G.W., Mendive, J.J. and Sheppard, R.C. (1968) J. Chem. Soc (C) 761764. 46. L. Moroder, L. Wilschowitz, M. Gemeiner, W. Gohring, S. Knof, R. Scharf, P. Thamm, J.D. Gardner, T.E. Solomon and E. Wunsch, E. (1981) HoppeSeylefs Z. Physiol Chem. 362, 929-942. 47. Moroder, L., Gohring, W., Nyfeler, R., Scharf, R., Thamm, P. and Wendlberger, G. (1983) Hoppe-SeylefsZ. Physiol Chem. 364, 157-171. 48. Huang, S.C, Yu, D.H., Wank, S.A., Manley, S., Gardner, J.D. and Jensen, R.T. (1989) Peptides 10, 785-789. 49. Wunsch, E., Moroder, L., Gohring, W., Borin, G., Calderan, A. and Bali, J.-P. (1986) FEBSLett. 206, 203-207. 50. Borin, G., Calderan, A., Ruzza, P., Moroder, L., Gohring, W., Bovermann, G. and Wunsch, E. (1987) Biol Chem. Hoppe-Seyler 368, 1363-1373. 51. Romano, R., Musiol, H.-J., Dufresne, M., Weyher, E. & Moroder, L. (1992) Biopolymers 32, 1545-1558.

869 52. Bodansky, M., Martinez J., Walker, M.D. Gardner, J.D., and Mutt, V. (1980) J. Med, CheiTL 23, 82-85. 53. G5hring, W., Moroder, L., Borin, G., Lobbia, A., Bali, J.-P. and Wunsch, E. (1984) Hoppe-Seylefs Z. Physiol 365, 83-94. 54. Peggion, E., Jaeger, E., Knof. S., Moroder, L. and Wunsch, E. (1981) Biopolymers 20, 633-652. 55. Peggion, E., Foffani, M.T., Wunsch, E., Moroder, L., Borin, G., Goodman, M. and Mammi, S. (1985) Biopolymers 24, 647-666. 56. Wu, C.C., Hachimori, A. and Yang, J.T. (1982) Biochemistry 21,4556-4562. 57. Torda, A.E., Baldwin, G.S. and Norton, R.S. (1985) Biochemistry 24, 17201727. 58. Goodman, M. and Listowsky, I. (1962) J. Am, Chem, Soc. 84, 3770-3771. 59. Urry, D:W. (1972) Biochim. Biophys, Acta 265, 115-168. 60. Pham van Chuong, P., Penke, B., de Castiglione, R. and Fromageot, P. (1979) in: Hormonal Receptors in Digestion and Nutrition (Rosselin, P., Fromageot, P. and Bonfils, S., eds.) pp. 33-44, Elsevier, North-Holland, Amsterdam. 61. Mammi, S., Goodman, M., Peggion, E., Foffani, M.T., Moroder, L. and Wunsch, E (1986) Int, J. Pept Protein Res. 27, 145-152. 62. Mammi, S., Mammi, N.J. and Peggion, E. (1988) Biochemistry 27, 13741379. 63. Mammi, S., Foffani, M.T., Peggion, E., Galleyrand, J.C., Bali, J.-P., Simonetti, M.. Gohring, W., Moroder, L. and Wunsch, E. (1989) Bkx:hemistry 28, 7182-7188. 64. Moroder, L., Romano, R., Weyher, E., Svoboda, M. and Christophe, J. (1993) Z. Naturforsch.4Sh. 1419-1430. 65. Moroder, L., Mourier, G., Dufresne, M., Bovermann, G., Gohring, W., Gemeiner, M. and Wunsch, E. (1985) BioL Chem. Hoppe-Seyler 368, 839848. 66. Mamimi, S., Mammi, N.J., Foffani, M.T., Peggion, E., Moroder, L. and Wunsch, E. (1987) Biopolymers 26, Sl-SlO. 67. Mammi, S. and Peggion, E. (1990) Biochemistry 29, 5265-5269. 68. Moroder, L., Bali, J.-P. and Kobayashi, Y. (1991) Biopolymers 3 1 , 595-604. 69. Moroder, L., Musiol, H.-J., Kocher, K., BaU, J.-P., Schneider, C.H., Guba, W., MuUer, G., Mierke, D. and Kessler, H. (1993) Eur. J. Biochem, 212, 325-333. 70. Moroder, L. and Wunsch, E. (1993) in: Gastrin (Walsh, J., ed.) pp. 187-194, Raven Press. New York. 71. Durieux, C., Belleney, J., Lallemand, J.Y., Roques, B.P. and Foumie-Zaluski, M.C. (1983) Biochem. Bvophys. Res. Commun. 114, 705-712. 72. Koizuka, I., Watari, H., Yanaihara, N., Nishina, Y., Shiga K. and Nagayama, K. (1984) Biomed. Res. Suppl 5, 161-168. 73. Foumie-Zaluski, M.C., Durieux, C., Lux, B., Belleney, J., Pham, P., Gerard D. and Roques, B.P. (1985) Biopolymers 24, 1663-1681. 74. Foumie-Zaluski, M.C., Belleney, J., Lux, B., Durieux, C., Gerard, D., Gacel, G., Maigret, B. and Roques, B.P. (1986) Biochemistry 25, 3778-3787. 75. Loomis, R.E., Lee, P.O. and Tseng, C.C. (1987) Biochem. Bvophys. Acta 911, 168-179.

870

76. Pattou. D., Maigret B.. Foumie-Zaluski. M.C. and Roques. B.P. (1991) Int. J. Peptide Protein Res. 37, 440-450. 77. Fang. S., Nikiforovich. G.V., Knapp, R.J., Jiao, D., Yamamura, H.I. and Hruby, V.J. (1992) in: Peptides, Chemistry and Biology (Smith, J.A. and Rivier, J.E. (eds.) pp. 142-143, ESCOM, Leiden. 78. Moroder, L., Weyher, E., D'Ursi, A., Bcone, D. and Temussi, P.A. (1992) ReguL Peptides 40, 213. 79. Moroder, L., D'Ursi. A., Bcone. D.. Amodeo, P. and Temussi, P.A. (1993) Biochem. Biophys. Res. Commun. 190, 741-746. 80. Jackson. M. and Mantsch. H.H. (1992) Biochim. Biophys. Acta 1118. 139143. 81. Sonnichsen, F.D.. van Eyk. J.E., Hodges. R.S., Sykes. B.D. (1992) Biochemistry 31, 8790-8798. 82. Moroder. L., Papini. A., Siglmuller, G.. Kocher. K.. Dorrer, E. and Schneider, C.H. (1992) Biol Chem. Hoppe-Seyler 373, 315-321. 83. Moroder, L. and Temussi, P.A. (1993) in: Peptides 1992 (Schneider, C.H. and Eberle, A.N., eds.) pp. 813-814, ESCOM, Leiden. 84. Moroder, L.. Romano. R.. Guba. W.. Mierke. D.F., Kessler, H.,Delporte. C , Winand, J. and Christophe, J. (1993) Biochemistry 32, 13551-13559. 85. Schwyzer, R., Kimura, S. and Erne, D. (1992) in: Peptides, Chemistry cmd Biology (Smith, J.A. and Rivier, J.E., eds.) pp. 168-170, ESCOM, Leiden. 86. Surewicz, W.K. & Epand, R.M. (1984) Biochemistry 23, 6072-6077. 87. Surewicz, W.K. & Epand, R:M: (1985) Biochemistry 24, 3135-3144. 88. Epand, R.M., Surewicz, W.K. & Yeagle, P. (1988) Chem. Phys. Lipids 49, 105110. 89. Brown, A.g. (1991) in: Nerve Cells and Neuronal Systems. Introduction in Newoscience (Brown, a.G., ed.) pp. 61-71, Springer Verlag. London. 90. Crawley, J.N. (1991) Trervds Pharmacol Set 12, 232-236. 91. Moroder, L. and Musiol, H.-J.(1990) in: Peptides, Chemistry, Structure and Biology (Rivier, J.E. and Marshall, G.R., eds.) pp. 811-812, ESCOM. Leiden. 92. Boyanov. A.I., Tenchov. B.G.. Koynova, R.D. and Koumanov, K.S. (1983) Biochim. Biophys. Acta 732, 711-713. 93. Eklund, K.K., Virtanen, J.A. and Kinnunen, P.K.J. (1984) Biochim. Biophys. Acta 793, 310-312. 94. Tenchov, E.G., Boyanov, A.I. and Koynova, R.D. (1984) Biochemistry 23. 3553-3558. 95. Moroder, L.. Musiol, J.-H. and Siglmuller, G. (1990) Synthesis 889-892. 96. Keller, O. and Rudinger, J. (1975) Helv. Chinh Acta 58, 531-541. 97. Wunsch, E., Moroder, L., Nyfeler, R., Kalbacher, H. and Gemeiner, M. (1985) Biol Chem. Hoppe-Seyler 366, 53-61. 98. von Grunigen, R., Siglmuller, G.. Papini, A., Kocher, K., Traving, B., G6hring, W. and Moroder, L. (1991) Biol Chem. Hoppe-Seyler 372, 163-172. 99. Papini. A., Rudolph, S., Siglmuller, G.. Musiol, H.-J., Gohring, W. and Moroder, L. (1992) Int. J. Peptide Protein Res. 39, 348-355. 100. Moroder. L.. Papini. A.. Siglmuller, G.. Kocher, K., Dorrer, E. and Schneider, C.H. (1992) Biol Chem Hoppe-Seyler 373, 315-321.

871 lOl.Gemeiner, M., Leidinger, E., Muller, I. and Moroder, L. (1992) Biol Cherru Hoppe-Seyler373. 1085-1094. 102.Winand, J., Poloczek, P., Delporte, C , Moroder, L., Svoboda, M. and Christophe. J. (1991) Biochim. Biophys, Acta 1080, 181-190. 103. Nielsen, O. and Buchardt, O. (1991) Synthesis 819-821. 104. Romano, R., Bayerl,, T.M. and Moroder, L. (1993) BiochiirL Biophys. Acta 1151, 111-119. 105. Bayerl, T.M., Werner, G.D. and Sackmann, E. (1989) Biochinh Biophys. Acta 984, 214-224. 106. Guba, W. and Kessler, H. (1994) J. Phys, Chem. 98, 23-27. 107. Romano, R., Dufresne, M., Frost, M.-C, Bali, J.-P., Bayerl, T.M. and Moroder, L. (1993) BiochiirL Biophys. Acta 1145, 235-242. 108. McLean, L.R. and Phillips, M.C. (1981) Biochemistry 20. 2893-2900. 109. Nichols, J.W. and Pagano, R.E. (1981) Biochemisiry 20, 2783-2789. llO.Thilo, L. (1977) Biochim. Biophys. Acta 469, 326-334. lll.Papahadjopoulos, D., Hui, S., Vail, W.J. and Poste, G. (1976) Biochim. Biophys. Acta 448, 245-264. 112.De Cuyper, M., Jordan, M. and Dangreau, H. (1983) Biochemistry 22, 415419. 113.Lentz, B.R., Carpenter, T.J. and Alford, D.R. (1987) Biochemistry 26, 53895397. 114. Martin, F.J. and MacDonald, R.C. (1976) Biochemistry 15, 321-327. 115.Mabrey, S. and Sturtevant, J.M. (1976) Proc. Natl. Acad. Set USA 73, 38623866. 116.Tenchov, B.G. (1985) Prog. Surf. Membr. Set 20, 273-340. 117. Huang, J., Swanson, J.E., Dibble, A.R.G., Hinderliter, A.K. and Feigenson, G.W. (1993) Biophys. J. 64, 413-425. 118.Peggion, E., Mammi, S., Palumbo, M., Moroder, L. and Wunsch, E. (1983) Biopolymers 22, 2443-2457. 119.Ananthanarayanan, V.S. and Orlicky, S. (1992) Biopolymers 32, 1765-1773. 120. Brimble, K.S. and Ananthanarayanan, V.S. (1992) Biochim. Biophys. Acta 1105,319-327. 121. Brimble, K.S. and Ananthanarayanan, V.S. (1993) Biochemistry 32, 16321640. 122. Ananthanarayanan, V.S. (1994) in: Peptides, Design, Synthesis and Biological Activity (Basava, C and Anantharamaiah, G.M., eds) pp. 223-234, Birkhauser Boston. 123. Ananthanarayanan, V.S. (1991) Biochem Cell Biol 69, 93-95. 124. Moroder, L., Lutz, J., Romano, R., Grams, F., Golbik, R. and Weyher, E. in: Peptides 1994 (Maia, H.S., ed.) ESCOM, Leiden, in press 125. Moroder, L. and Romano, R. (1994) Pure andAppl Chem 66, 2111-2114. 126. Brittain, H.G., Richardson, F.S. and Martin, R.B. (1976) J. Am Chem Soc. 98, 8255-8260. 127.0toda, K., Kimura, S. and Imanishi, Y. (1993) Bull Chem Soc. Jpn. 66, 1466-1471.

872 128.Woolley, G.A., Epand, R.M., Kerr, J.D., Sanson, M.S.P. and Wallace. B.A. (1994) Biochemistry 33, 6850-6858. 129.Blau, L., Stem, R.B. and Bittman, R. (1984) Biochinh Biophys. Acta 778, 219-223. 130.Ohki, K., Shunjl, N., Sagami, M. and Nozawa, Y. (1986) Chem. Phys. Lipids 39, 237-249. 131. Grienkiewicz, G., Poenie, M. and Tsien. R.Y. (1985) J. Biol Cheiru 260, 34403450. 132.Tsien, R.Y., Rink, T.J. and Poenie, M. (1985) CeU CalciumG, 145-157. 133. Blau, L. and Weissmann, G. (1988) Biochemistry 27, 5661-5666. 134.Berendes, R., Burger, A., Voges. D., Demanche, P. and Huber, R. (1993) FEBSLett. 317, 131-134. 135.Fattal, E., Nir, S., Parente, R.A., Szoka, F.C.Jr. (1994) Biochemistry 33, 6721-6731. 136. Irvine, R.F. (1992) FASEBJ. 6, 3085-3091. 137. Irvine, R.F. (1992) Curr. Opin. Cell Biol 4, 212-219. 138. Wank, S.A., Harkins, R., Jensen, R.T., Shapira, H., de Weerth, A. and Slattery, T. (1992) Proc. Natl Acad. Set USA 89, 3125-3129. 139. Wank, S.A., Pisegna, J.R. and de Weerth, A. (1992) Proc. Natl Acad. Set USA 89, 8691-8695. 140. Steigerwalt, R.W. and Williams, J.A. (1981) Endocrinol 109, 1746-1753. 141.Vinayek, R., Patto, R.J., Menozzi, D., Gregory, J., Mrozinski, J.E., Jensen, R.T. and Gardner, J.D. (1993) Biochim. Biophys. Acta 1176, 183-191. 142.Talkad, V.D., Fortune, K.P., PoUo, D.A., Shah, G.N., Wank, S.A. and Gardner, J.D. (1994) Proc. Natl Acad. Set USA 91, 1868-1872. 143.Takeuchi, K., Speir, G.R. and Johnson, L.R. (1979) Am. J. Physiol 237, E284-E294. 144.Takeuchi, K., Speir, G.R. and Johnson, L.R. (1980) Am. J. Physiol 239, G395-G399. 145. Magous, R. and BaU, J.-P. (1982) Eur. J. Pharmacol 82, 47-54. 146.Magous, R., Galle5n:and, J.-C, Leonard, A , Choquet. A. and Bali, J.-P. (1987) in: Gastrin and Cholecystoktntn. Chemistry, Physiology and Pharmacology (Bali, J.-P. and Martinez, J., eds.) pp. 153-158, Elsevier, Amsterdam. 147. Christophe, J. (1994) Am. J. Physiol 266, G963-G971. 148. Sankaran, H., Goldfme, I.D., Deveney, C.W., Wong, K.-Y. and Williams, J.A (1980) J. Biol Chem. 255, 1849-1853. 149.Wank, S.A., Jensen, R.T. and Gardner, J.D. (1988) Am. J. Physiol 255, G106-G112. 150. Menozzi, D., Vinayek, R., Jensen, R.T. and Gardner, J.D. (1991) J. Biol Chem. 266, 10385-10391. 151.Winand, J., Poloczek, P., Delporte, C., Moroder, L., Svoboda, M. and Christophe, J. (1991) Biochim. Biophys. Acta 1080, 181-190. 152. Hootman, S.R. and Williams, J.A. (1987) in: Physiology of the Gastrointestinal Tract (Johnson, L.R., ed.) 2nd ed., pp. 1129-1146, Raven Press, Nev^ York. 153. Gardner. J.D. and Jensen, R.T. (1987) in: Physiology of the Gastrointestinal Tract (Johnson, L.R., ed.) 2nd Ed. pp. 1109-1127, Raven Press, New York.

873

154.Matozaki, T. and WiUiams, J.A. (1989) J. Biol Chem, 264, 14729-14734. 155.Winand, J.. Delporte, C , Poloczek, P., Cantraine, F., Dehaye, J.-P. and Christophe. J. (1992) Second Messengers and Phosphoproteins 13, 173-186. 156.Beinbom, M., Lee, Y.-M., McBride, E.W., Quinn, S.M. and Kopin, A.S. (1993) Nature 362, 348-350. 157. Silvente Poirot, S., Hadjiivanova, C , Escrieut, C , Dufresne, M.. Martinez, J., Vaysse, N. and Fourmy, D. (1993) Ew. J. Biochem. 212, 529-538. 158. Silvente Poirot, S., Escrieut, C , Dufresne, M., Martinez, J., Bouisson, M., Vaysse, N. and Fourmy, D. (1994) Mol Pharmacol 45, 599-607. 159.Trumpp-Kallmeyer, Hoflack, J., Bruinvels, A. and Hibert, M., F. (1992) J. Med, Ghent 35, 3448-3462. 160. Hibert, M., F., Trumpp-Kallmeyer, Hoflack, J . and Bruinvels, A. (1993) TIPS 14, 7-12. 161.Freidinger, R.M. (1989) Med. Res, Rev, 9, 271-290. 162. Bock, M.G. (1991) Drugs of the Future 16, 631-640. 163. Woodruff, G.N. and Hughes, J. (1991) Anna. Rev, Pharmacol Toxicol 31, 469-501. 164.Harro, J., Vasar, E. and Bradwejn, J. (1993) TIPS 14, 244-249. 165.Pincus, M.R., Carty, R.P., Chen, J., Lubowsky, J., Avitable, M., Shah, D., Scheraga, H.A. and Murphy, R.B. (1987) Proc, Natl Acad, Set USA 84, 48214825. 166. Chang, R.S.L., Lotti, V.J., Chen, T.B. and Kunkel, K.A. (1986) Mol Pharmacol 30, 212-217. 167.Lignon, M.F., Galas, M.C., Rodriguez, M., Laur, J., Aumelas, A. and Martinez, J. (1987) J, Biol Chem. 262, 7226-7231.



875

When Two Steroids are Better than One : The Dimeric Steroid-Pyrazine Marine Alkaloids A. Ganesan INTRODUCTION This chapter reviews the history of a remarkable family of natural products first isolated from the marine tube-inhabiting invertebrate worm Cephalodiscus gilchristi. C. gilchristi, found in the temperate Southern Hemisphere, is often attached to bryozoans and sponges. This tiny worm (- 5 mm in tube colonies) can exist independently of the coenicium (worm tube), and exposure to predators during such moments may have aided the evolution of chemical defence mechanisms. In 1972, Pettit's group collected a sample of C. gilchristi at a depth of approximately 20 m in the Indian Ocean off southeast Africa. In 1974, methanol and aqueous extracts of C. gilchristi were confirmed active in the American National Cancer Institute's primary antitumour assay at that time, the murine lymphocytic leukemia P388 (PS system), with 32-41 % life extension at 25-37.5 mg/kg. Identification of the active components was a slow process hampered by the limited material, and required over ten years effort with recollections of fresh sample. Success was achieved with a 1981 batch of 166 kg (wet weight, including worm tubes) of C. gilchristi, from which PS bioassay-guided solvent partitioning^ (Scheme 1) yielded active dichloromethane and carbon tetrachloride fractions. 166 kg wet weight

CH2Cl2-MeOH extraction

H2O fraction

C H2CI2 fraction

hexane 9:lMeOH-H20 fraction fraction

CCI4 fraction PS active, 42 g

4:1 MeOH-H20 fraction

CH2CI2 fraction PS active, 28 g

3:2 MeOH-H.O fraction

Scheme 1.. Solvent partitioning scheme for the extraction of cephalostatins. The active fractions were further purified by extensive column chromatographic and HPLC separations (for details of the protocol, see next section) to obtain 138.8 mg (8x10'^^ % yield) of a pure compound, cephalostatin 1, mp 326 °C dec. Evidence from TLC staining, NMR, and mass

876 spectrometry indicated a steroidal alkaloid but the complete structural elucidation required X-ray analysis of crystals carefully grown from a pyridine solvate. As with many other compounds of marine origin, the structure of cephalostatin 1 (Figure 1) is unprecedented and has no analogy with terrestrial natural products. The compound is an unsymmetrical steroid dimer linked at the C-2,3 positions (for ease of reference, steroid numbering is used throughout; dimers are numbered from C-1 for the "right" half and C-T for the "left" half) by an aromatic pyrazine ring. Oxygenation of the side chain results in a spiroketal ring system as in many saponin aglycones. Two other noteworthy features are the presence of a C-14,15 alkene in both halves, and the oxidation of the C-18' angular methyl to a hydroxy methyl group.

HO

O1 P

Figure 1. Cephalostatin 1.

ISOLATION OF OTHER CEPHALOSTATINS Cephalostatin 1 has an ED50 of 10"^-10"^ M-g/ml against the PS cell line and is in fact among the most potent compounds ever recorded in the NCI's antitumour screening program. Over the next few years, the Pettit group solved the structures of several other cephalostatins purified from the 1981 batch of C. gilchristi. Compared to the parent compound cephalostatin 1, cephalostatins 2-4^ (Figure 2) are hydroxylated at the C-9' ring juncture of the left half. Furthermore, cephalostatin 3 contains an additional methyl group in the side chain, while in cephalostatin 4 the C-14'-15' alkene is epoxidized. These three cephalostatins have similar PS activity to cephalostatin 1. The isolation"^ of cephalostatins 5 and 6 (Scheme 2) is typical of the careful and lengthy separations performed for these compounds by the Pettit group. In these two cephalostatins, ring C of the left half is aromatized. The PS activity of cephalostatins 5 and 6 is dramatically reduced to an ED50 of lO'"^ |Xg/ml, possibly due to the flattening out of the steroid and the loss of C-D ring stereochemistry. Cephalostatin 6 can be derived from cephalostatin 4 by a plausible sequence of events. First, dehydration at C-9' would generate an enone (a similar enone was later isolated from C gilchristi, see cephalostatin 14, vide infra). The epoxide ring can then undergo nucleophilic ring opening at C14' by participation of the enone to yield a dienone. Finally, a retro-aldol reaction at C-18' would complete the aromatization. Alternatively, the angular hydroxymethyl group may be lost via oxidative fragmentation similar to that observed in the biosynthesis of cholesterol from lanosterol. Cephalostatin 5 is presumably related to the epoxy derivative of cephalostatin 3 by the same reaction sequence.

877 CH2CI2 fraction 28 g, from Scheme 1 Silica gel, 1:1 hexane-EtOAc to 4.5:4.5:1 EtOAc-MeOH-HjO

5.8 g (>10)

9.3 g

«io-')

9.8 g (0.3)

Sephadex LH-60 10:10:1 hexane-CH2Cl2-MeOH 6.53 g (10-') Silica gel, gradient 10:10:0 to 10:10:4 hexane-EtOAc-MeOH

1.15 g in two fractions (10-')

402 mg (~ 10-') Silica gel, gradient 10:10:1 to 10:10:4 hexane-EtOAc-MeOH

212 mg in three fractions

171 mg

(lo-'-io-^)

1. C-18 reverse phase HPLC, 1:1 MeOH-HjO to MeOH 2. Silica normal phase HPLC, 30:70:0 to 30:70:10 hexane-EtOAc-MeOH

127 mg in nine fractions (10-^-10-^)

10 mg (10-") 1.Sephadex LH-20 4:5:1 hexane-CH2Cl2-MeOH 2. HPLC, C-18, 1:1 MeOH-HzO to MeOH 3. Sephadex LH-20 MeOH

Cephalostalin 5, 5.5 mg (4 x 10-^)

Cephalostatin 6, 3.0 mg (2 x 10-')

Scheme 2. Detailed chromatographic separation scheme for cephalostatins 5 and 6. Numbers in brackets refer to PS bioassay activity, in )Lig/ml.

878

*OH

'^

Cephalostatin 2 R = H Cephalostatin 3 R = Me

R

HO

O ^

O^

OH O 1

O-^

Figure 2. Cephalostatins 2-6. Cephalostatins 7-9 (Figure 3) exhibit further structural variation in the left half. In cephalostatin 7, the C-18' angular methyl is no longer oxidized, and the side chain now forms a 6/5 spiroketal. This spiroketal can be derived from the right half's 5/5 spiroketal by ring opening followed by recyclization and removal of the C-23 alcohol. Thus, the two halves of cephalostatin 7 are almost identical. Cephalostatins 8 and 9 retain the C-18' hydroxymethyl group. Compared to the parent cephalostatin 1, cephalostatin 8 has one more methyl group and is less oxygenated in the side chain. Meanwhile, cephalostatin 9 is really the hemiketal form of cephalostatin 1. By now, the NCI had switched from the PS bioassay to a panel of 60 human solid tumour cell lines. Against this panel, cephalostatins 7-9 are of similar potency to cephalostatin 1-4, with TI50 values of 10"^-10'^° molar against a number of the cell lines. The cephalostatins also display a characteristic panel graph, and exhibit one of the most extreme cases of differential cytotoxicity encountered in the NCI assay. Cephalostatin 1, for example, has GI50 values in these cell lines ranging from 6x10'^ to 2.5x10''' molar.

879

HO /

OH O \

O^

H^ OH O ^

oJ^^'ÔH O

Cephalostatin 8

He ^4^^^^0H OH 0 1 9 ^

Cephalostatin 9

Figure 3. Cephalostatins 7-9. In the last two years, Pettit's group has reported the structure of eight more cephalostatins isolated from a 450 kg collection of C gilchristi made in 1990. Cephalostatins 10 and 11^ (Figure 4) are the C-1 and C-l' methoxy derivatives of cephalostatin 2 respectively. Cephalostatin 10 thus represents the first member with a structural change in the right half, which has remained constant in cephalostatins 1 through 9. In their cytotoxic effects, both these cephalostatins are of similar activity to cephalostatin 1. The disymmetry of the cephalostatin halves can be considered a biosynthetic puzzle - was the enzymatic machinery selectively fusing two non-identical steroids only, or was it differentiating the two halves of a homodimer? The latter possibility sounds more likely, and is supported by the similarity between the two halves of cephalostatin 7 {vide supra). The structures^ of cephalostatins 12 and 13 (Figure 4) provide further corroborating evidence. For the first time, the two halves of cephalostatin 12 are identical and correspond to the right half of earlier cephalostatins. Cephalostatin 13 differs from the symmetrical dimer only by C-l' hydroxylation. Interestingly, these compounds were found in the more polar n-butanol fraction during solvent partitioning, while previous cephalostatins were retained in the dichloromethane layer. The symmetrical compounds are also much less active against the NCI panel; while cephalostatin 1 has a mean panel GI50 of 1 nmolar, cephalostatin 12 and 13 were 400 nmolar and >1 fimolar respectively. The implications of this are unclear. Possibly, the increased polarity of the left half in these compounds is responsible for the decreased activity.

Cephalostatin 10 R = OMe, R' = H Cephalostatin 11 R = H, R' = OMe OH O 1 V^

)^0 HO-K\^'

i OH OH Cephalostatin 12 R = H Cephalostatin 13 R = OH

Figure 4. Cephalostatins 10-13 Cephalostatins 14 and 15 (Figure 5) are related to cephalostatins 2 and 3 respectively by aepoxidation at C-14'-15', dehydration of the C-9' alcohol, and hydroxylation at C-8'. Cephalostatins 14 and 15 display reduced activity, with mean panel GI50 of 100 and 68 nmolar respectively, perhaps due to the epoxide orientation - with steroidal bufadienolides, the p-epoxides are more cytotoxic.

OH O 1

Cephalostatin 14 Cephalostatin 15

p-^

R =H R = Me

Figures. Cephalostatins 14 and 15. The most recent cephalostatins to be reported are cephalostatin 16 and 17^ (Figure 6). Cephalostatin 16 is composed of the left half of cephalostatin 2 coupled to the left half of cephalostatin 7. Cephalostatin 17 also contains the left half of cephalostatin 2, while the other half is identical to the typical right half of cephalostatins 1-11 except for one less hydroxyl group. The

mean panel values for cephalostatins 16 and 17, at 1 nmolar and 4 nmolar respectively, are comparable to cephalostatin 1.

n- o ^ ^ f ' M l l

OH O c

Jij^3c

11 ^X-/

' ii

THT

N

H

J

X''///

•0

Cephalostatin 16

OH 0 1

pOH

Cephalostatin 17 Figure 6. Cephalostatins 16 and 17. In general, the cephalostatins isolated recently are in relatively minor abundance compared to cephalostatin 1 (for example, only 3.8 mg, 8x10*^ % yield, of cephalostatin 17 was obtained), and of lower activity except for cephalostatins 16 and 17. The Pettit group has also detected*^ other new cephalostatins in very small quantities (approximate yield of 10"^ %) with promising activity against brain cancer xenografts.

THE RITTERAZINES It remains to be seen what structural surprises are in store with future compounds isolated from C gilchristi, and also if similar compounds are produced by Antarctic members of the Cephalodiscus genus. Recently, exciting developments have been disclosed^ ^ from an unexpected quarter by Fusetani's group, working with the tunicate Ritterella tokioka Kott 1992 collected off the coast of Japan. The lipophilic extract of the tunicate showed promising cytotoxicity. PS bioassaydirected fractionation of 5.5 kg of tunicate yielded 2.9 mg of ritterazine A, with an ED50 of 10'^ |lg/ml in the PS assay. The structure of ritterazine A (Figure 7) was solved based on ^H and ^^C NMR data, and it bears an obvious resemblance to the cephalostatins. Prior to the extraction of the tunicate, colonies were first washed free of macroepibionts and sands. No attached hemichordates were observed.

882

HOI Figure?. RitterazineA. The left half of ritterazine A is identical to that of cephalostatin 7, except for an additional C7' hydroxyl group. The right half comprises a rearranged steroid nucleus. A reasonable biogenetic pathway for the right half is shown in Scheme 3 (In this and later schemes, hydrogens at steroid trans ring-junctures and portions of the steroid nucleus that are not participating in the reaction are usually omitted). Protonation of the C-14,15 alkene is followed by a 1,2-Wagner-Meerwein shift and trapping of the resulting carbocation by water to give the observed skeleton. The intermediate carbocation can also be derived by an alternative mechanism via protonation at C-15, pinacol-like rearrangement, and Prins reaction between the aldehyde and alkene.

ritterazine A

Scheme 3. Possible biogenesis of the ritterazine A right half. Subsequently, two other ritterazines with an unrearranged steroid nucleus as in the cephalostatins were isolated^^ from the same collection of R. tokioka (Figure 8). Ritterazine B has the same left half as ritterazine A. The right half is comparable to that of cephalostatin 1, except for the absence of hydroxylation at C-17, C-23, and C-26. Ritterazine B is also the first of these dimeric steroids where C-14 is not part of an alkene or epoxide. Instead, a p-hydrogen giving rise to cis C-D ring fusion is present. In ritterazine C, one half is identical to the right half of ritterazine B, while the other half is identical to the right half of cephalostatin 1 (this half is chosen as the "right" half in Figure 8 to emphasize this relationship) except for additional C-7 hydroxylation. Ritterazines B and C have an IC50 in the PS assay of 1.8x10"'* |ig/ml and 9.4x10"^ |ig/ml respectively.

883

\ OH OH

Ritterazine B

OH O 1

^/"s^r-

H

^\/N^^,.-\

THT

1

i^Sî ^ ^

0 V

Hj •

0

OH

i

O^

N

^ ^

Ritterazine C

Figure 8. Ritterazines B and C. Further work with an 8.2 kg collection of the tunicate led to the characterization'^ of ten new ritterazines (Figure 9). Ritterazine D is the C-22 epimer of ritterazine A. Ritterazine E has one additional methyl group compared to ritterazine D. The biological activity of these two compounds is similar to that of ritterazine A. In ritterazines F and G, C-22 is again epimeric to the configuration observed in ritterazine B; ritterazine G has the additional modification of C-14,15 unsaturation. In the PS bioassay, ritterazines F and G have IC50 values of T.SxlO'"^ M^g/ml - the highest activity seen among the ritterazines apart from ritterazine B (note that this is still much lower than the activity of cephalostatin 1). Ritterazines H and I form another pair of C-22 epimers, with C-12 now at the ketone oxidation state and biological activity reduced by approximately twenty-fold. The presence of both 5/5 spiroketal diastereomers for a number of ritterazines implies that epimerization at C-22 is not energetically prohibitive. Interestingly, such spiroketal epimers have yet to be observed among the cephalostatins. Ritterazines J-M are most closely related to the left half of cephalostatin 7. Ritterazine K is the symmetrical dimer formed from this half, while ritterazine J has one additional hydroxyl group. Ritterazines L and M are a pair of C-22 spiroketal epimers in which the C-17 hydroxyl group is lost. These ritterazines all have IC50 values around 10'^ fig/ml in the PS assay. The biological profile of the ritterazines, with ritterazines B, F, and G the most potent, seem to indicate the importance of the 5/5 and 6/5 spiroketals, while the CI4,15 alkene is not crucial. Oxidation at C-12 to a ketone results in decreased activity, as does the rearrangement of the steroid nucleus as in ritterazines A, D, and E. However, further data on related compounds will be needed before these hypotheses can be confirmed. Meanwhile, the difference in nomenclature between the cephalostatin and ritterazine families is rather confusing. Regardless of their origin, these compounds clearly belong together and perhaps they should be reclassified under a single family of steroidal alkaloids.

884

HO'

i OH OH

RitterazineD R = H Ritterazine E R = Me

Ritterazine F Rj = OH, R2 = R3 = H HOI

\ OH OH Ritterazine G Rj = OH, R2 = H, C14,15 alkene Ritterazine H Rj = R2 = O, R3 = H Ritterazine I

Rj = R2 = O, R3 = H, C-22 epimer OH Rol

f' OH

H R i = Ro

HOL

^ OH OH

Ritterazine K R j = H R2 = Ritterazine L Rj = R2 =H Ritterazine M R^ = R2 = H, C-22 epimer

Figure 9. Ritterazines D-M. The isolation of the ritterazines from a phylum unrelated to the hemichordates leaves the true source of these dimers unclear. With certain natural products isolated from marine macroorganisn it is now established'"^ that they are actually produced by symbiotic microflora, and this may be case here. If a microorganism that produces these dimers can be identified and grown in laborai. conditions, it raises the exciting prospect of obtaining these highly potent steroids in greater quantities by large-scale fermentation. The biosynthesis of these compounds seems to occur in two phases: (1) coupling of two steroids via a pyrazine linker, and (2) relatively unselective oxidation at various positions. Some of these compounds are related to others by simple processes: the hydration of cephalostatin 1 to its hemiketal form cephalostatin 9; the dehydration of cephalostatin 2 to an enone, which in turn may be an intermediate to cephalostatin 6; the skeletal rearrangement of ritterazine B to ritterazine A, which may be acid catalyzed; and the pairs of ritterazines epimeric at C-22. One can speculate whether such reactions are non-enzymatic transformations occurring in the organism or even during the isolation procedure.

885

SYNTHETIC STEROID-PYRAZINE DIMERS VIA a-AMINO KETONES The intriguing structure of these steroids coupled with their potent biological activity and limited availability makes them an attractive challenge for the synthetic organic chemist. One of the key features of any attempted synthesis is the central heterocyclic ring. The classical method of pyrazine synthesis involves the dimerization of a-amino ketones. A steroidal example relevant to the cephalostatin problem was reported^^ as early as 1968 (Scheme 4): androstanolone (1) was converted to a-oximino ketone 2, which was hydrogenated to afford the hydrochloride salt of a-amino ketone 3. The salt was neutralized to the free base, and condensed in situ to yield symmetrical dimer 4 in modest overall yields.

l.H2,Pd-C,HCl 2. aq. Na2C03

Scheme 4. First preparation of a pyrazine linked steroid dimer. Subsequent to the isolation of the cephalostatins, there was renewed interest in such steroid dimers. An improved procedure for the preparation of 4 was reported^^ by the Fuchs group (Scheme 5), in which the intermediate a-amino ketone was produced in higher yield by the sequence of bromination with phenyltrimethylammonium bromide, displacement by azide to give a-azido ketone 5, and hydrogenation (an identical route was independently developed^^ in the Heathcock group). OH

OH l.PTAB 2. [(Me2N)2CNH2] N3 j ^ •

O

71 %

H2, Pd-C - ^ Dimer 4 41 %

Scheme 5. Synthesis of 4 via azido ketone 5. The Fuchs group also prepared dimers 6-11 (Figure 10) by analogous procedures. These compounds were tested by the MTT cytotoxicity method in a panel of five human tumour cell lines. The dimer 6 with the cholestanyl side chain had little activity, with an ED50 > 100 |ig/ml. However, all the other dimers displayed some degree of cytotoxicity. For example, dimer 4 had an ED50 of 7 |ig/ml against the colon adenocarcinoma cell line HT-29, and values around 30 jig/ml against the others.

886 Although the activity is low, it is interesting given the huge simphfication in structure compared to the natural products. The results suggest that more sophisticated cephalostatin analogues with improved biological activity can be prepared synthetically. Such compounds would be more accessible than the scarce natural products; moreover, their cytotoxicity can be modulated, whereas the natural products may be too toxic for direct therapeutic use. Dimers 4 and 10 were also tested in two murine epithelial tumour xenografts transformed by mutations in the low-molecular weight guanine nucleotide binding protein Ras. Ras is an important protein in cellular signal transduction, cycling between an inactive GDP-bound state and an active GTP-bound form. The protein also has intrinsic GTPase enzymatic capability, thus preventing permanent activation. In the mutated Ras, GTP hydrolysis is greatly diminished, leading to signals that cause cell proliferation. At the maximally tolerated dose of 150 mg/kg/day, 4 inhibited the tumours by 28 and 59 %, without any deaths due to cytotoxicity.

Common pyrazine and A-B ring core of dimers 6-11.

HO

9

10

11

Figure 10. Symmetrical dimers prepared from azido ketones by Fuchs and coworkers. While preparing large quantities of 4 for animal testing, the Fuchs group isolated 5-10 % of a byproduct, identified as its CI-azido derivative. Control studies determined that this compound was produced by the presence of excess azide during the displacement reaction, and the suggested mechanism is shown in Scheme 6. Thus, azide acting as a base^^ enolizes a-azido ketone 5, which then loses nitrogen to yield a-amino enone 12. This process can also be recreated by treating 5 with DBU. Dimerization of 12 followed by reaction with azide completes the process. It is possible that similar generation of an electrophilic centre adjacent to the pyrazine occurs during cephalostatin biosynthesis, as some of the natural products also show C-1 substitution. Interestingly, dimer 13 has increased cytotoxicity compared to 4 and 6-11, with an ED50 in the 0.2-0.4 M-g/ml range against the same tumour cell lines. Presumably, this increase is a nonspecific effect caused by the presence of the toxic azide functional group.

887

"Xt H,N

I

azide

dimerization

:

^""m^

OH

*

13

Scheme 6. Proposed mechanism for formation of the C-1 azido dimer 13. Further improvement was achieved^^ in the a-amino ketone dimerization process by Smith and Heathcock. Cholestanone (14) was converted to its 2a-azido derivative 15. The azide (Scheme 7) was reacted with aqueous triphenylphosphine, rather than reduction by hydrogenation, and the crude dihydropyrazine dimers aromatized by air oxidation in the presence of p-toluenesulfonic acid. The product (6) was then isolated in high yield by simple filtration. The triphenylphosphine reaction proceeds via an imino phosphorane, which in principle can undergo dimerization by a Staudingerlike reaction. However, no dimerization was observed under anhydrous conditions, implying that hydrolysis of the phosophorane ylide to the a-amino ketone occurs first. The rate of phosphorane hydrolysis in NMR experiments was relatively slow (several hours), and it may be possible to produce unsymmetrical dimers by trapping the a-amino ketone as it is being formed by a more reactive keto steroid.

PPh,

""H C oXX

PhgP

x^:^

.o

[/.• "2°. H,N,,^^4/

' • ^

i;4^ ^ rvt^^^x) ^ N - ^ ^

87% overall

:

*

.

^

.

Dimer 6

Scheme 7. Preparation of pyrazine dimers using triphenylphosphine for azide reduction.

UNSYMMETRICAL PYRAZINES An obvious disadvantage of a-amino ketone dimerizations is their unsuitability for crosscoupling. The alternative of condensing a 1,2-dicarbonyl component with a 1,2-amine is also inappropriate for most cephalostatins and ritterazines, which are unsymmetrical both from right to left and top to bottom. It is possible that unsymmetrical condensation could be carried out in a stepwise manner using two a-amino ketones protected in different ways. However, this would

decrease the efficiency of pyrazine formation, which would be a late step in a total synthesis with highly functionalized and precious monomers. In the Heathcock group, a number of solutions to this problem were explored. One discovery^^ was the hetero-Diels-Alder-like reaction between oxadiazinones and enamines (Scheme 8). Although this reaction has the required regiospecificity and takes place under very mild conditions, it proved impossible to prepare an oxadiazinone fused to a cycloalkane, as required in a cephalostatin synthesis. -80 ^C to RT H3C -CO2, -pyrrolidine y-^Y>^

H3C N

I

I

+

• K' \

O

Ph

80%

N Ph

"3

I \ --"3

Scheme 8. Synthesis of an unsymmetrical pyrazine via a Diels-Alder like reaction. Later, ^^ Smith and Heathcock studied dimerizations with a-amino oximes as one of the components. Amino oxime 16, derived from 15 in two steps (Scheme 9), was reacted with epoxy acetate 17. The product mixture contained "trans" pyrazine dimer 6 and its A^-oxide derivative 18 in low yield, together with 2a-acetoxycholestanone (19). However, the expected "cis" dimer 20 was produced in too small a quantity for isolation. Heating 16 alone yielded 6 together with a trace of Aôxide 18. Epoxy acetate 17 was also unstable to the reaction conditions, slowly rearranging to 2pacetoxycholestanone (21) which then epimerized to the equatorial acetate 19. Heating amino oxime 16 together with 2p-acetoxycholestanone gave dimers 6 and 18 with the latter predominating, as observed in the initial experiment. Furthermore, the ratio of 6 to 18 increased with higher initial concentrations of 21, suggesting that dimer 18 was the product of cross-coupling. I.NH2OH 2. Ph3P, H2O O

15

90%

":xt OH

1. AC2O O

16 + 17

2. dimethyldioxirane ^ O , 14

74%

O Ac

17

toluene, 85 ^C. 24 h • •

N 6,5

«

tt txxi: N 18, 10 %

Ac

O

19

20, not detected Scheme 9. Dimerization of the cholestanyl amino oxime 16 and epoxy acetate 17.

For improved yields, an analogue of amino oxime 16 less prone to self-dimerization was needed. Towards this end, the 0-methyl derivative 22 was prepared, which undergoes complete dimerization only at 140 ^C. Heating acetoxy ketone 21 with 22 at 85 °C gave dimer 6 in 3.5 % yield after 1 day, together with unidentified compounds which may be intermediates in its formation. Extending the reaction time to 14 days increased the yield to 23 % (Scheme 10). A protocol was worked out involving initial heating at 85-90 ^C during which the acetoxy ketone and amino oxime ether preferentially react with each other, followed by heating at 140 °C to complete the process. Acetoxy ketone 19 was also isomerized to the 3(3-acetoxy-2-one 23, which then dimerizes with 22 to give "cis" pyrazine 20.

140 ^C, 24 h OMe

22

87%

Dimer 6

Ac 22, 85 ^C, 14 d >. Dimer 6 21

23%

Ac

:rt; 19

Me4N0Ac

•:ii

35%

Ac

22, 90 ^C, 5 d 20%

23

tt

N Dimer 20

Scheme 10. Dimerizations involving amino oxime ether 22. A complication in these reactions is the similar properties for "trans" dimer 6 and the "cis" dimer 20. The only noticeable difference in the NMR spectrum is the shift of one ^H resonance by 0.01 ppm, and a shift in the pyrazine ^^C resonances by about 0.1 ppm, while the optical rotations are identical ([ajo = +82^). However, the compounds have very different solubilities - trituration of the mixture with ethanol yields the trans dimer upon filtration, while the crude cis dimer is obtained by evaporation of the filtrate. This separation procedure enabled purification of the mixture obtained from reaction of 2a,3a-diaminocholestane (obtained by borane reduction of 22) and 2,3diketocholestane (obtained by oxidation of 3-cholestanone with potassium t-butoxide and oxygen) (Figure 11). H2N,, H2N

'tt O'

110QC,24h^ Trans Dimer 6, 28 % + Cis Dimer 20, 31 %

Figure 11. Pyrazine dimers via reaction between 1,2-diamines and 1,2-diketones. The utility of the amino oxime ether - acetoxy ketone combination was shown using two different steroid monomers. Androstanolone was converted to the 2p,17p-diacetate 24 (Scheme 11), which was then reacted with amino oxime ether 22 to afford unsymmetrical dimer 25 in reasonable yield. In this dimerization, replacement of 24 by its 2a-acetoxy epimer gave similar results, while the 2a-bromoketone instead led to messy reaction mixtures. Reactions between 22 and 2,3-epoxy acetates were also tried, which produced an approximately 1:1 ratio of cis and trans pyrazine dimers.

890 Compound 25 was also hydrolyzed to give the corresponding dimer with the free C-17 hydroxy! group.

r ' j A 2.dimethyldioxirane Ac ^ [ \ amino ^"^' oxime ether 22 ^^^xvslÂs^^x-^^ 3. refluxing toluene ^ O^^x-Vsfxx^.x^'^^ ^^^''C, 24 h; 140 ^C 24 h O^^^--^^

58%

O ^ ^ - ^ ^

.

^

^^^

^

Scheme 11. Smith and Heathcock's synthesis of an unsymmetrical pyrazine. Starting from hecogenin acetate (26), an inexpensive commercially available sapogenin whose side chain resembles that of the cephalostatins, keto alcohol 27 was prepared (Scheme 12). This was carried forward to acetoxy ketone 28, the recrystallization step in this sequence being necessary as the MoOPD oxidation yields an inseparable mixture of the 2-hydroxy-3-one, 3-hydroxy2-one, and 4-hydroxy-3-one. Reaction of 28 with amino oxime ether 22 under the usual conditions provided 29 % of unsymmetrical dimer 29. The acetate group in 29 was also hydrolyzed to give a more crystalline dimer. Dimers 6, 25, and 29, together with the deacetoxy derivatives of 25 and 29, were all submitted for testing in the NCI's solid tumour panel. None of the dimers were sufficiently active to warrant further investigation. l.NaBH.

1. IDA; MoOPD 2. AC2O 3. recrystallization 38%

28

Scheme 12. Synthesis of a unsymmetrical dimer with a spiroketal side chain. Another route to unsymmetrical steroidal pyrazines is the likely biosynthetic process of differentiation of a symmetrical dimer, as accomplished^^ by Winterfeldt's group. Starting from

891 hecogenin acetate, compound 30 containing the C-14,15 alkene present in the cephalostatins was synthesized by a literature procedure^^ involving the interesting sequence of photochemical isomerization and Prins reaction (Scheme 13). O

"'-. Q^>^'''

-^

0 I]

\*o

•

0

.,r^

^ 0'

HT*

r '^ .y^ J ^ OH OH

HO /i

Jk>>

"^ N'

OH O 1

I lO ^

7' OH Ritterazine K, 18 '

Scheme 31. Total synthesis of cephalostatin 7, cephalostatin 12, and ritterazine K.

902 Samples of the synthetic compounds were provided to Professor Pettit's group, who confirmed the identity of cephalostatins 7 and 12 based on NMR and chromatographic comparison. If this mechanism for pyrazine formation occurs in C. gilchristi, one would expect it to also produce ritterazine K. A search among the currently unidentified residual Cephalodiscus extracts revealed a substance with identical chromatographic profile to ritterazine K. However, it was present in only microgram quantities, and its identity could not be confirmed by NMR.

THE FUCHS APPROACH TO DIHYRDOCEPHALOSTATIN 1 The Fuchs group has also prepared a monomer for the left half of dihydrocephalostatin 1. The dihydro compound was chosen in order to investigate the spiroketalization process and also the importance of the C-14,15 alkene for biological activity. The route begins^^ with keto aldehyde 69 (Scheme 32), available in 60 % yield from Marker degradation of hecogenin. This was reduced to keto alcohol 70, and the C-18 methyl group functionalized by Meystre's hypoiodite method,^^ after which Jones oxidation provided lactone 71. The C-3 acetate was hydrolyzed, and the free alcohol reprotected as a silyl ether, followed by reduction to give triol 72. Regioselective carbenoid insertion into a neopentyl alcohol set the stage an intramolecular Wadsworth-Emmons reaction (see Scheme 25, vide supra), after which the oxidation state was adjusted to yield 73. l.HOCH2CH20H,PPTS 2. NaBH4, ^^^^3 3.H2,Pt02

d^

OH

4. PPTS 59%

70

l.Pb(OAc)4,l2 2. H2Cr04 56%

l.Et02CC(N2)PO(OEt)2, Rh2(OAc)4 2. H2Cr04 3.NaH 4. LiAlH4 5. TFAA, DMSO, Et3N

I.KHCO3 2. TBDPSCl 3. LiAlH.

—

68%

72

73%

CHO

TBDSO Scheme 32. Preparation of a keto aldehyde precursor for dihydrocephalostatin 1.

•

903 Compound 73 was reacted with methallylstannane to give a separable pair of diastereomeric alcohols (Scheme 33) in a 1:2.7 ratio favouring the desired product. The unwanted diastereomer was recycled in 79 % yield by Mitsunobu inversion. The alcohol was then benzylated to afford intermediate 74, after which reduction of the C-12 ketone yielded a 1:9 ratio of a - and p-C-12 epimeric alcohols. The alcohols were osmylated and subjected to periodate cleavage to provide 75. This was reacted with methyl Grignard, followed by acid catalyzed cyclization with (+)camphorsulfonic acid. Three spiroketal products 76, 77, and 78 were isolated in a 1:15:1 ratio.

OBn CHO y\

1. methallyl stannane, BF-^.EtÔ 2. BnBr, NaH

90%

l.MeMgBr 2. CSA

Q HO

78% 76

''^ 77

Scheme 33. Elaboration of 5/5 spiroketal of dihydrocephalostatin 1. Attempts at producing X-ray quality crystals of the spiroketals were unsuccessful. The compounds were treated with fluoride to effect C-3 deprotection, and nOe effects used to assign C-12 stereochemistry. Oxidation of the diols gave ketones 79 and 80 (Scheme 34), whose structures were determined by 2D-NMR experiments.

l.TBAF 2. H2Cr04 76 + 77 + 78

Scheme 34. Equilibration of the 5/5 spiroketal.

904 According to molecular mechanics calculations, the C-23 benzyl group favours equatorial attack of the alcohol on the oxonium ion intermediate during spiroketalization, hence explaining the kinetic preference for product 77. The calculations also revealed that 79 and 80 should be less stable than their diastereomeric spiroketals by approximately 2 kcal/mol. Indeed, heating 79 with camphorsulfonic acid gave a new ketone 81, while similar treatment of 80 produced 82 (Scheme 34). Upon extended reaction times, 81 was converted to 30 % of 82 (estimated to be approximately 3 kcal/mol lower in energy) along with decomposition products. This last reaction requires epimerization at C-20 through an oxonium ion-enol ether equilibration. Deprotection of the benzyl group in diketone 82 yielded 83 (Scheme 35), a dihydrocephalostatin 1 left half intermediate suitable for coupling with the right half.

BnO

H2, Pd-C 100 % O 82 83 Scheme 35. Preparation of a dihydrocephalostatin 1 left half monomer.

SUMMARY AND FUTURE PROSPECTS The family of dimeric steroid-pyrazine alkaloids isolated from Cephalodiscus and Ritteria now stands at thirty members. There are undoubtedly other examples of this group of steroidal alkaloids that have yet to be discovered, and it is probable that members common to both sources will be found. On the synthetic front, the discovery of these alkaloids has sparked interest in the construction of unsymmetrical pyrazines, and the methods developed will be useful in other settings as well. The Fuchs group has successfully accomplished landmark syntheses of tetrahydrocephalostatin 12, cephalostatin 7, cephalostatin 12, and ritterazine K and is clearly close to a synthesis of dihydrocephalostatin 1. These efforts have added significantly to the areas of steroid and spiroketal chemistry. The ability to make unnatural cephalostatins will greatly aid our understanding of the biological potency of these compounds. For example, one half could be kept identical to a natural product, while varying the other with synthetic steroids prepared from commercially available materials. In this sense, the cephalostatins provide a unique opportunity for such experiments, as the steroid skeleton is a readily accessible and well understood scaffold in which the effects of particular substituents can be determined. The situation here is in stark contrast to certain other potent biologically active compounds such as taxol (which promotes microtubule assembly) and bryostatin (a protein kinase C inhibitor). In the latter cases, construction of the skeletal framework is a formidable enterprise, hindering further dissection of the biological activity. The availability of larger quantities of biologically active material, whether from natural or synthetic sources, will unravel the site of action of these steroids, about which nothing is known at present. These alkaloids do not contain functional groups commonly assisted with antitumour agents e.g. alkylation and Michael acceptor sites, intercalators, redox-active quinonoid groups, and radical generators. The range of biological activity among the various natural products contains some tantalizing structure/activity clues which are difficult to fully decipher.

905 Soon after the 1988 communication on cephalostatin 1, it was predicted"^^ that the compound acts on the cell membrane. Steroids are components of eukaryotic cell membranes, where they incorporate into one half of the phospholipid bilayer and provide rigidity. Taking into account the dimeric nature of the cephalostatins, these steroids may now traverse the full length of the bilayer (for example, cephalostatin 1 is 30 A x 9 A x 5 A) and adversely affect its properties. A number of other highly oxygenated marine natural products (e.g. brevetoxin, palytoxin) are also membrane active agents. Fuchs has made two suggestions on the cephalostatins' mechanism of action. First,^^ the oxygenated functional groups of these compounds may form a network of hydrogen bond donors and acceptors that interacts with a specific enzyme target. More recently,"'^ Fuchs has implicated the C14,15 alkene as being important in biological activity, and there is some evidence"^^ supporting this hypothesis. He points out that the C-14,15 alkene may be susceptible to electrophilic attack in vivo, either by protonation or epoxidation, followed by a rearrangement similar to that postulated in Scheme 3 (vide supra) which would generate a number of reactive centres in the molecule. It will be interesting to see if the CD-ring skeletal rearrangement can be achieved in the laboratory, and if the reactive intermediates can alkylate DNA, for example. Another potentially reactive site in these molecules is the spiroketal ring system. Perhaps this undergoes ring opening to generate reactive carbonyl and free alcohol groups. Finally, returning to the title of this chapter, the necessity for these compounds to be dimeric remains unclear. It is not known if any of the advanced synthetic mono-steroid intermediates prepared also exhibit high cytotoxicity. Furthermore, whether the central pyrazine ring is simply a linker or serves some additional function is also a mystery. This could be tested by examining the biological activity of unnatural dimers with other linkages e.g. benzene rings, other heterocyclic systems, or even acyclic tethers.

ACKNOWLEDGEMENTS I wish to thank Professors George Pettit and Philip Fuchs for kindly providing preprints of references 9, 36, 38, and 40, and for their comments on the manuscript.

REFERENCES ' Pettit, G. R.; Kamano, Y.; Aoyagi, R.; Herald, C. L.; Doubek, D. L.; Schmidt, J. M.; Rudloe, J. J. Tetrahedron 1985, 41, 985-994. ^Pettit, G. R.; Inoue, M.; Kamano, Y.; Herald, D. L.; Arm, C; Dufresne, C ; Christie, N. D.; Schmidt, J. M.; Doubek, D. L.; Krupa, T. S. J. Am. Chem. Soc. 1988,110, 2006-2007. ^Pettit, G. R.; Inoue, M.; Kamano, Y.; Dufresne, C ; Christie, N.; Niven, M. L.; Herald, D. L.; /. Chem. Soc, Chem. Commun. 1988, 865-867. Also see J. Chem. Soc, Chem. Commun. 1988, 1440 for correction of a minor typographical error in the structures. '^Pettit, G. R.; Kamano, Y.; Dufresne, C ; Inoue, M.; Christie, N.; Schmidt, J. M.; Doubek, D. L.; Krupa, T. S. Can. J. Chem. 1989, 67, 1509-1513. ^Pettit, G. R.; Kamano, Y.; Inoue, M.; Dufresne, C ; Boyd, M. R.; Herald, C. L.; Schmidt, J. M.; Doubek, D. L.; Christie, N. D. J. Org. Chem. 1992, 57, 429-431. ^Pettit, G. R.; Xu, J.; Williams, M. D.; Christie, N. D.; Doubek, D. L.; Schmidt, J. M.; Boyd, M. R. J. Nat. Prod. 1994, 57, 52-63. ^Pettit, G. R.; Ichihara, Y.; Xu, J.; Boyd, M. R.; Williams, M. D. Bioorg. & Med. Chem. Lett. 1994, ^,1507-1512. ^Pettit, G. R.; Xu, J.; Ichihara, Y.; Williams, M. D.; Boyd, M. R. Can. J. Chem. 1994, 72, 22602267.

906 ^Pettit, G. R.; Xu, J.-P.; Schmidt, J. M. Bioorg. & Med. Chem. Lett. 1995, 5, 2027-2032. '^Pettit, G. R. Pure&Appl. Chem. 1994, 66, 2271-2281. ^' Fukuzawa, S.; Matsunaga, S.; Fusetani, N. J. Org. Chem. 1994, 59, 6164-6166. ^^Fukuzawa, S.; Matsunaga, S.; Fusetani, N. 7. Or^. Chem. 1995, 60, 608-614. ^^Fukuzawa, S.; Matsunaga, S.; Fusetani, N. Tetrahedron 1995, 57, 6707-6716. '"^For example, see: Kobayashi, J.; Ishibashi, M. Chem. Rev. 1993, 95, 1753-1769. ^^(a) Ohta, G.; Koshi, K.; Obata, K. Chem. Pharm. Bull. 1968, 7(5, 1487-1497. (b) Smith, H. E.; Hicks, A. A. J. Org. Chem. 1971, 36, 3659-3668. ^^Pan, Y.; Merriman, R. L.; Tanzer, L. R.; Fuchs, P. L. Bioorg. & Med. Chem. Lett. 1992, 2, 967972. '^Ganesan, A. Ph.D. dissertation. University of California-Berkeley, 1992. 'Êdwards, O. E.; Purushothaman, K. K. Can. J. Chem. 1964, 42, 712-716. ^^ (a) Smith, S. C ; Heathcock, C. H. J. Org. Chem. 1992, 57, 6379-6380. (b) Full paper: Heathcock, C. H.; Smith, S. C. 7. Org. Chem. 1994, 59, 6828-6839. ^^Ganesan, A.; Heathcock, C. H. J. Org. Chem. 1993, 58, 6155-6157. ^' Kramer, A.; Ullmann, U.; Winterfeldt, E. J. Chem. Soc, Perkin Trans. 11993, 2865-2867. ^^Welzel, P.; Janssen, B.; Duddeck, H. LeibigsAnn. Chem. 1981, 546-564. ^^ Jeong, J. U.; Fuchs, P. L. J. Am. Chem. Soc. 1994,116, 773-774. ^^Marker, R. E.; Wagner, R. B.; Ulshafer, P. R.; Wittbecker, E. L.; Goldsmith, D. P. J.; Ruof, C. H. J. Am. Chem. Soc. 1947, 69, 2167-2230. For modifications, see: (a) Dauben, W. G.; Fonken, G. J. J. Am. Chem. Soc. 1954, 76, 4618-4619. (b) Micovic, I. V.; Ivanovic, M. D.; Piatak, D. M. Synthesis 1990,591-592. ^^(a) Suryawanshi, S. N.; Fuchs, P. L. J. Org. Chem. 1986, 57, 902-921. (b) Jain, S.; Shukla, K.; Mukhopadhyay, A.; Suryawanshi, S. N.; Bhakuni, D. S. Synth. Commun. 1990, 20, 1315-1320. ^^ Corey, E. J.; Da Silva Jardine, P.; Virgil, S.; Yuen, P.-W.; Connell, R. D. J. Am. Chem. Soc. 1989, 777,9243-9244. ^^ Jeong, J. U.; Fuchs, P. L. Tetrahedron Lett. 1994, 35, 5385-5388. ^^Kim, S.; Fuchs, P. L. Tetrahedron Lett. 1994, 35, 7163-7166. ^^Sharpless, K. B.; Gao, Y. J. Am. Chem. Soc. 1988, 770, 7538-7539. ^° Reich, H. J.; Peake, S. L. J. Am. Chem. Soc. 1978,100,4888-4889. ^' Henry, Jr. K. J.; Grieco, P. A.; Jagoe, C. T. Tetrahedron Lett. 1992, 33, 1817-1820. ^^ Kim, S.; Sutton, S. C ; Fuchs, P. L. Tetrahedron Lett. 1995, 36, 2427-2430. "Barton, D. H. R.; Basu, N. K.; Hesse, R. H.; Morehouse, F. S.; Pechet, M. M. J. Am. Chem. Soc. 1966,88,3016-3021. ^^ Jeong, J. U.; Fuchs, P. L. Tetrahedron Lett. 1995, 36, 2431-2434. ^^ Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartung, J.; Jeong, K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-M.; Xu, D.; Zhang, X.-L. J. Org. Chem. 1992, 57, 2768-2771 36 ^^ Jeong, J. U.; Sutton, S. C ; Kim, S.; Fuchs, P. L. J. Am. Chem. Soc. 1995, 777, 10157-10158. 37 Li, C ; Shih, T.-L.; Jeong, J. U.; Arasappan, A.; Fuchs, P. L. Tetrahedron Lett. 1994, 35, 26452646. ^^ Bhandaru, S.; Fuchs, P. L. Tetrahedron Lett. 1995,46, 8347-8350. Heusler, K.; Wieland, P.; Meystre, C. H. Organic Synthesis 1965,45, 57-63. "^^ Bhandaru, S.; Fuchs, P. L. Tetrahedron Lett. 1995, 46, 8351-8354. "^^ Ganesan, A.; Heathcock, C. H. Chemtracts-Org. Chem. 1988, 7,311-312. Unpublished results from E. Winterfeldt's group, quoted in reference 19(b).

Atta-ur-Rahman (Ed.) Studies in Natural Products Chemistry, Vol. 18 O 1996 Elsevier Science B.V. All rights reserved.

907

Human IgGl Hinge-Fragment as a Core Structure for Immunogens Luis Moroder, Gerd Hiibener and Manfred Gemeiner 1) Max-Planck-Institut fur Blochemie, 8 2 1 5 2 Martinsried. Germany; 2) Veterinar- Medizinische Universitat, Wien, Austria

Abstract: After the early discovery that synthetic fragments of proteins if suitably presented to the immune system, are capable of eliciting antibody responses cross-reactive with the parent proteins, the concept of synthetic vaccines has become a main target in immunology. Today there are wellestablished tenets that generation of the humoral response requires cooperation and communication between B and T cells and that small-sized antigens induce immunity only when appropriately presented for recognition by both the B-lymphocytes and T-helper cells. As immunological responsiveness is dependent on the HLA genotype of an individual, genetic restriction presents an additional burden for the design of fully synthetic immunogens as mimetics of surface regions of pathogens, but which nonetheless give raise to antibodies capable of neutralizing the microorganism and abolishing its infectivity. In the context of knowledge in the field our contribution is reviewed by comparing the intensity and specificity of the immune responses to a self-antigen, i.e. to the hormone gastrin, in various approaches as antigen/protein conjugates, built-in immunoadjuvanticity, liposomal preparations and finally with well defined constructs in which the hinge segment of human IgGl with its peculiar structural and functional properties was exploited. The conservative character of the hinge as pivot in the dynamics of the immunoglobulins, and its apparently universal role as recognition site in the interplay of immunologically relevant molecules was found to induce surprisingly high immunogenicity even in the case of selfantigens indicating that such hinge-peptide/antigen constructs as interesting molecules for studying the processes of immunity at cellular level.

1. Introduction With t h e development of vaccines medicine gained the ability to control, eliminate, or even eradicate selected diseases. The classical vaccines consisting mostly of killed or live a t t e n u a t e d microbial agents or their isolated c o m p o n e n t s have been highly successful. Smallpox h a s been eradicated, and viral d i s e a s e s like measles, mumps,

poliomyelitis,

rubella

and

yellow fever

rarely

occur

in

developed

countries. Similar s u c c e s s h a s been achieved with bacterial d i s e a s e s s u c h as diphteria, t e t a n u s , tubercolosis and whooping cough, a n d equally successful were vaccines in veterinary medicine. However, m a n y d i s e a s e s r e m a i n for which no

908 vaccines exist, e.g. malaria, herpes and t h e autoimmunodeficiency

syndrome

(aids). Despite t h e success, c u r r e n t p r o c e d u r e s of vaccine p r e p a r a t i o n p r e s e n t serious s h o r t c o m i n g s s u c h a s whether a particular viral vaccine p r e p a r a t i o n is completely killed or sufficiently a t t e n u a t e d , t h e genetic variation of v i r u s e s , or the difficulty in p r e p a r i n g enough material for vaccine production a s well a s the h a z a r d to p e r s o n n e l and environment w h e n working with large a m o u n t s of p a t h o g e n s , b u t also the low t e m p e r a t u r e s required for storage a n d t r a n s p o r t (1,2). With t h e existing vaccines immunization p r o g r a m m e s t h a t c a n r e a c h every child at c o s t s acceptable for every country are very difficult to be realized. The C h i l d r e n ' s Vaccine Initiative envisages the ideal vaccine of t h e future to b e safe, heat-stable, and advances

in

technology,

effective w h e n administered orally early in life (3). Recent

the

field

peptide

of molecular synthesis,

and

cellular

polysaccharide

immunology, chemistry

recombinant

and

microbial

p a t h o g e n e s i s m a y provide t h e clues for the rational design of vaccines of the future t h a t generate a d e q u a t e a n d s u s t a i n e d protective i m m u n e r e s p o n s e s even w h e n n a t u r a l disease fails to do so (1). Although the molecular b a s i s of immunological recognition r e m a i n s unclear,

a wealth

immunocompetent

of information

has

recently

been

accumulated

molecules and their interactions in p r o d u c i n g

largely on

the

successful

antibody r e s p o n s e s . With t h i s improved knowledge considerable efforts are now exerted t o w a r d s t h e development of new vaccines (1,4,5). One of t h e strategies used for this purpose is r e c o m b i n a n t DNA technology for the p r o d u c t i o n of pathogen s u b u n i t s , i.e. proteins, in either bacterial, yeast or animal cells, or even live vaccines by introducing relevant genes into t h e genome of vaccinia virus. In vitro expression of immunogenic proteins of p a t h o g e n s a s a m e t h o d for producing new vaccines h a s been often a n d unexpectedly u n s u c c e s s f u l (1). This indicates t h a t a protein isolated from a micro-organism is rarely a s i m m u n o g e n i c as t h e same protein w h e n it is a c o n s t i t u e n t p a r t of the Abbreviations: IgGl, immunoglobulin Gl; KLH, keyhole limpet hemocyanin; BSA, bovine serum albumin; RNase A. ribonuclease A; LDH, lactate dehydrogenase; MHC, major histocompatibility complex; LHRH, luteneizing hormone releasing hormone; CCK, cholecystokinin; VIP, vasoactive intestinal peptide; TASP, template-assembled synthetic proteins; MAPS, multiple antigen presenting system; SUV. small unilamellar vesicle; DCC, dicyclohexylcarbodiimide; HOSu, N-hydroxysuccinimide; Mal>, maleimido; TFA, trifluoroacetic acid; TFE, trifluoroethanol; DMF. dimethylformamide; DMSO, dimethylsulfoxide; CD, circular dichroism; NMR nuclear magnetic resonance spectroscopy; ELISA, enzyme-linked immunosorbent assay; MD, molecular dynamics; Mox, methoxinine (oxa-analogue of methionine); P-Ala. 3-aminopropionic acid; all amino acids are of L-configuration unless stated otherwise; abbreviations for amino acids and derivatives are according to the lUPAC-IUB nomenclature.

909 parent organism probably as a result of the altered environment. Moreover, the subunit approach, where success was more fortuitous than rational, does not provide much insight into the mechanism of immune responses at molecular level which, however, could be obtained from the use of synthetic peptides as antigenic determinants of microbial proteins (1). This alternative approach towards the development of potential vaccines is based on rather simple fully synthetic structures which are mimicking small regions in the pathogen, but can nonetheless give raise to antibodies capable of neutralizing the microorganism and abolishing its infectivity (1,5). The concept of synthetic vaccines based on relatively small peptides has been proposed already in the early 1970s (6,7). Actually, Anderer (8,9) made the very important discovery that a short tryptic fragment of the protein of tobacco mosaic virus and its synthetic replicate elicit antibodies capable of neutralizing the infectivity of the virus. But it was not until Sela and coworkers (10) demonstrated that a synthetic fragment of the coat protein of Ms2 bacteriophage leads to antibodies that cross-react with the virus particle, that the concept of synthetic vaccines received the due attention. Since these early pioneering studies great progresses were achieved in the better understanding of the structural requirements for an effective immunization, i.e. for the ability to provoke antibody responses and sensitized cells of appropriate specificity that lead to protection, for built-in immunoadjuvanticity and possibly for a bypass to the genetic restriction and antigenic competition. Our contributions to the field were exclusively devoted to the chemical approach, and are reviewed here in the context of some central issues in the field as induction of immunogenicity into fully synthetic constructs and of immunological memory. The the gastrointestinal hormone gastrin was selected as antigen throughout the study to allow for a more precise comparative analysis of the various approaches used to induce immune responses to this self-antigen.

2. Immunogens on Peptide/Carrier Basis In terms of immunological properties proteins or related fragments exhibit essentially two distinct characteristics which do not necessarily coincide. The first, immunogenicity. is the capacity to elicit an immune response, manifested either by antibody production (humoral immune response) or by cell-mediated immunity. The second. (B-cell) antigenicity, refers to the capacity to be recognized in a specific manner by immunocompetent cells or by an antibody. Macromolecular antigens, mostly proteins and glycoproteins, usually express in their structure a multitude of possible antigenic determinants, or epitopes, that

910 dictate the specificity of the immune response. However, only a limited number of potential epitopes are involved in the induction of immunogenicity. It is well established that immunizations with short peptides generally result in low levels of antibodies and only a few natural (11) and synthetic peptides (12-20) were sufficiently immunogenic to fulfil the requirements of a fully competent immunogen. Therefore conjugation of peptides to synthetic or natural (protein) carriers for macromolecularization is required to elicit strong immune responses. The carriers are typically proteins like keyhole limpet hemocyanin (KLH), tetanus toxin, bovine serum albumin (BSA), ovalbumin, chicken immunoglobulin and thyroglobulin, all being potent immunogens which apart from providing polyvalent presentation of the antigenic peptide as B-cell epitope, also function to provide Tcell specific epitopes. Today there are well-established tenets in immunology that generation of the humoral immune response requires cooperation and communication between B and T cells (21) and that small-sized antigens induce immunity only when covalently attached to proteins that can be recognized upon processing by T-cells (22-24). Moreover, immunological responsiveness is dependent on the HLA genotype of an individual (25), which has now been correlated with the ability of the individual's MHC molecules to bind fragments of the immunogen for presentation to the T-cells (26). The rules that govern the binding of peptides (fragments of the immunogen) to class I and class II MHC (27-29), the 3Dstructures of the two classes of MHC molecules (30-36) and the sequence motifs characteristic for recognition of T-cell epitopes by the MHC molecules of T-helper and cytotoxic T-cells have been disclosed (37-41).

2.1. Conventional Peptide/Protein

Conjugates

Generally, peptides are coupled to the carriers by homobifunctional or heterobifunctional reagents (45-46) among which the most commonly used are carbodiimides, particularly in the water soluble form, glutaraldehyde, isocyanates and diimidoesters. This ill-defined chemistry leads to extensive inter- and intramolecular crosslinking of the component parts of the conjugate as a result of lack of selectivity of the reactions exploited with the reagents. The presence of various reactive functional groups both in the antigenic peptide and in the protein carrier generates chemical modifications and crosslinkings at different positions of the peptide chain and carrier surface. Correspondingly, the peptide antigen may be presented to the immune system in heterogeneous form regarding its chemical structure, its carrier-surface environment and thus, possibly also its spatial array in the conjugate. Heterogeneous epitopes both in terms of chemical and spatial

911 structure are formed which generally induce heterogeneous responses (47,48). Additionally, epitope suppression and destruction, and generation of new dominant epitopes related to the cross-linking reagents are obvious consequences of this methodology (49-51). Using peptides of the gastrin hormone family as model antigens these uncontrolled effects were fully confirmed in our laboratory (48,52). In fact, gastrin/BSA conjugates prepared in conventional manner by the water-soluble carbodiimide led in animal-dependent manner to antibodies of differentiated specificity in full agreement with results from other laboratories where immunization experiments were usually performed on a larger number of animals to obtain in a "trial and error" manner antisera of the desired specificity (53-55).

2.2. Selective Polyvalent Peptide/Protein

Conjugates

Aware of these serious shortcomings resulting from the poor selectivity of the conjugation procedures, we have proposed the use of a more defined chemistry based on a reactive anchor group introduced at well defined positions of the synthetic antigen and suitable for a selective conjugation to a protein (47,48,56,57). Similarly to what nature is using in some of the posttranslational processings of proteins, thiol functions accessible on the surface of proteins lend themselves as ideal targets for conjugate chemistry since these reactive groups allow for highly selective crosslinking of molecules via disulfide- or thioether-bond formation exploiting mild thiol interchange or thiol addition reactions. For latter type reaction Keller and Rudinger (58) introduced the maleimide group as highly reactive thiol acceptor on the basis of previous studies on the reactivity of thiols with maleimide (59,60). Whilst other laboratories contemporarily proposed maleimidated carrier proteins as reaction partners for cysteine-containing peptides (61), our efforts in the field were focussed on the synthetic accessibility of peptide derivatives containing the reactive maleimide handle at positions of the peptide chain preselected in view of a maximum retention of the antigenic structure (Fig. 1). The maleimido-protein approach which today represents the most frequently used selective conjugation technique is based on maleimido-benzoylated proteins and synthetic peptides containing at their N- or C-termini an additional cysteine residue (61). This method, however, bears two main difficulties: i) The maleimido-function which is introduced via reaction of the carrier protein, generally KLH, with m-maleimidobenzoyl-N-hydroxysuccinimide ester, leads to a bulky aromatic anchor group on the protein surface with can generate strong

912 new artificial immunodeterminants in the conjugation site of the antigenic peptide. ii) The solid phase synthesis of peptides with C-terminal cysteine residues is accompanied by tedious side reactions (62,63) and peptides with N-terminal cysteine exhibit high tendency for oxidative dimerization (64). As bypass to these side reactions at least regarding N-terminal cysteine the use of Niso-glutajninyU^^ palmitoyV'lysyl'S'tert'butyl-cysteamtne, N-hydro3Qrsuccinimide as additive; this procedure greatiy facilitated the synthetic accessibility of the N-acetylmuramyl compounds. The adjuvant molecules were then deprotected at the thiol function by reduction with tii-butylphosphlne and

928 reacted with the malelmldo-gastrin derivative to yield the well defined adjuvant/gastrin conjugates shown in Fig. 12 (175). Upon immunization of rabbits with these conjugates no immime response could be observed at least in terms of detectable anti-gastrin antibodies. The question arises of whether the immunological castration induced with the covalent N-acetylmuramyl-dipeptide/LHRH adduct derives from an immune response to the hormone or fi-om occupancy of the LHRH receptor containing cells with immunoadjuvant moieties and correspondingly, firom an immune response to these cells.

A)

CH3CH ^fHAc C0-Ala-D-Glu-NH2

O

NH-(CH2)2-S-t--\ I N-(CH2)2-CO.[Moxl5].HG-[2.l7]

B) CH2OH )H HO I CH3CH

NHAc

C0-Ala-D-Glu.NH2 /9 LLys-NH-(CH2)2-S-r^ I N-(CH2)2-CO-[Moxl5]-HG-[2-17] CH3-(CH2)i4-CO "-^ O

Fig, 12, Chemical structures of the (A) N'Ocetymuramyl'dipeptide'Cystecanine/ and IB) lipo-N'aceytlrrairamyl'tripeptide'Cysteamine/[MQx^^]-gc^ 12-17] adduct

929 4,2. T- and B-Cell Epitope Chimeras One of the main disadvantages of the peptlde/carrier conjugate approach for antipeptlde immunizations Is the possible carrier-induced B-cell epitope suppression (176-178). the expression of new antlgen-imspecific immime epitopes related to the crossllnklng structures (49) and the reduced conformational space of the antigenic peptides on the carrier surface (177.179.180). Moreover, in view of the development of synthetic vaccines the choice of the carrier is extremely difficult. Taking into account the current knowledge about the mechanisms of Immune responses which require for both humoral and cell-mediated immimity the cooperation and communication between B and T-cells. it was suggested that Band T-cell epitopes need to be linked in order for T-cells to provide cognate help for B-cell activation and antibody production. On the basis of these concepts there have been numerous reports showing the successful construction of synthetic immunogens by the simple combination of well-defined T-cell determinants and B-cell epitopes. For this purpose different approaches have been used as linear polymerization of peptides (181-183). copol)niierization of B-cell and T-cell determinants by bifunctlonal crossllnklng reagents (184). or by collnear synthesis of B-cell epitopes with "natural" or with "foreign" T-cell epitopes, whereby a surprisingly strong effect on the specificity of the antibodies was found to derive from the orientation of the epitopes in the synthetic constructs (185-191). Moreover, it has recently been demonstrated that covalent linkage of the B- and Tcell epitopes does not represent an essential prerequisite for a good humoral response (192.193). The conformational preferences of the B-cell determinants in these chimeric linear peptides may play a critical role in the process of Induction of peptide antibodies of predetermined specificity, particularly if crossreactlvities with the parent protein is the goal of the immunization. These aspects have found only recently the due attention (194-199). The experiences gained in this field and from the de novo design of miniproteins opened new rational synthetic approaches for the synthesis of conformationally stabilized immunogens. In a long-term study Kaumaya (200) investigated this conformational aspect via the synthesis of topographic detennlnants by preserving maximal sequence homology to the native protein antigenic site in order to retain the functional specificity, and by introducing artificial mutations to facilitate a folding of the synthetic construct into a conformation mimicking the antigenic protein surface. For the conformational stabilization of the constructs secondary structural motifs known for proteins such as a-helices. p-sheets. p-tums and loops of the antigenic sites of the C4 isoenzyme of LDH as model protein were engineered to fit into known stable supersecondary structural motifs as ap. pap, papa or 4-a-helical bundles.

930 The Immunogeniclty of these constructs and the specificity of the antibody responses confirmed the validity of this promising approach (199.201.202). Besides the reconstruction of these natural subdomains of proteins as Immunogens it has been proposed to use synthetic templates for the design of conformationally stabilized fuUy synthetic miniproteins as potential immunogens has also been proposed. This concept foresees built-in folding devices for the induction of protein-like folding units, i.e. the employment of polyfimctionalized synthetic scaffolds on which peptides of specific conformational preferences are linked in a manner to stabilize via interchain interactions three-dimensional arrays of well-defined functionality (203). This template-assembled synthetic protein fFASP) approach, pionieered by Mutter (204). allowed for interesting progresses even in immunology (205). As alternative to these fully, more or less rigid synthetic templates used in the de novo design of proteins, we have introduced the concept of a natural template and for this purpose we have chosen the hinge segment of immunoglobulins (206,207). 5. Hinge-Peptide Based Fully Synthetic Immunogens 5.1. Hinge Segment oflgGl As known from x-ray crystallographlc studies on several antibody molecules and related fragments the two heavy and two light chains of immunoglobulins are

switch peptides Paratop

N-Termini

C-Tennini Fig. 13. Schematic drawing of the IgGl structure.

931 folded into domains which are arranged in pairs interacting by non-covalent forces except the CH2 domain of the Fc portion (208): interchain disulfide bridges provide further stability to these complex molecules (Fig. 13). In human IgGl the two heavy chains are linked by two proximal disulfide bridges in a portion of the molecule which connects the two Fab arms with the Fc segment. Because of the unique spatial structure and functional properties of this portion of the molecule it has been named the hinge segment. This segment can be divided into three regions: the upper hinge, the core and the lower hinge as shown in Hg. 14 (209210).

F(ab22

^

^ fak

.FcgJ ^^

E&

H-Glu-Ppo-Lys-Ser-Cys-Asp-Lys-Thi^Hb-Thr-Cys-Pro-Pi^Cyf-PitHAJa-Pro-Glu-Leu-Leu-Gly-Gly-Pn^ H-Glu-Pro-Lys-Ser-Cys-Asp-Lys-Thr-Hb-Thr-Cyi-Pi^PnHCys-Pro-Ala-Pro-Glu-Leu-Leu-Gly-Gly-Pro-OH 216 225 v-1 , ' 23« core

^

^^ upper hinge

^.^ middle hinge

.

^

lower hinge

genetic hinge

Fig. 14. The amino acid sequence of the human IgGl hinge segment with indication of the enzymatic cleavage sites and the resulting fragments.

The upper hinge includes the amino acids from the end of the CHI domain to the first residue in the hinge that restricts free motion (the first cysteine that forms an interheavy-chain bridge). In the x-ray structure this upper hinge folds into a oneturn helix with little inherent stability and full solvent accessibility (Fig. 15). In fact, the crystal structure of the trypsin generated Fc(t) (211) and comparative NMR analysis of various enzymatic hinge fragments of IgGl (212.213) as well as of i3C-labelled mouse IgG2a (214) indicate a high degree of flexibility for this portion of the heavy chain which correlates well with the known segmental flexibility responsible for the functional movements of the Fab arms. Interactions between the CHI domains and the upper hinge are absent in the SD-structure and thus, the spatial structure of the Fab fragments is fully retained upon enzymatic removal of the hinge segment. The rigid core hinge contains the two interheavychain disulfide bridges and it is folded into two parallel disulflde-linked poly-(Pro)II heUces (Fig. 15).

932

F^. 15. Stereoview of the SD-stmcture of the upper and core hinge as determined by x-ray analysis. The lower hinge connects the rigid core to the CH2 domain. The crystal structure of Fc(t) has shown that the two CH3 domains pair tightly in lateral contacts whereas no interaction takes place between the two CH2 domains (211). This would suggest an extended and to some degree rigid structure for the lower hinge. NMR analysis of the enzymatic IgGl fragments (213) led to the conclusion that the structure of the lower hinge, i.e. of the segment Pro230-Leu^^. in the doublestranded hinge fragment Lys222-Leu234^ m F(ab')2 and Fc(t) is essentially the same as that in the intact IgGl; thereby an extension of the poly-(Pro)-II helix from the core us. the lower hinge was suggested. More recent ^^c-NMR analysis are more supportive for high degree of flexibility of this lower hinge at least in mouse IgG2a (214). From these data it was concluded that the cyclic portion of the hinge segment although located in an overall flexible region, acts like a pivot for the functional movements of the Fab arms and Fc portion (215). Because of these peculiar properties a synthetic replicate could function as as suitable template in synthetic more or less constrained miniproteins as protein surface mimetics and thus, as potential immunogens. Essential premise for such an application was sufficient stability of the parallel disulfide linked cyclic peptide towards disproportion ation into monomers, antiparallel dimers as well as oligomers under the acid or basic conditions required in vairlous synthetic steps as well as under the physiological conditions of in vivo biological and immunological experiments. On the other hand, sufiiciently fast enzymatic processing is also required in the antigen presenting cells too, if an immunogenic construct is the goal.

933 5.1. Synthesis and Conformation of the IgGl hinge fragment 225-232/225'-232' Taking into account the results of the NMR analysis on enzymatic hinge fragments by Ito and Arata (212) and the fact that preliminary synthetic studies on the cyclic bis-cystinyl-octapeptide [H-Cys-Pro-Pro-Cys-OH]2 were not as satisfying as expected, an IgGl hinge-fragment was selected with N- and particularly. C-terminal exocyclic extensions as shown in Fig. 16 (216) in order to avoid in first instance the known proximal endgroup effects on the stability of biscystinyl structures (217). H-Thr-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH

I

I

H-Thr-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH

Fig. 16. Hinge fragment 225'232/225''232'ofhumanIgGl.

The synthesis of the bis-cystinyl-octapeptide dimer in parallel alignment as present in the native human IgGl was achieved by the use of a selective cysteineNps-Thr(tBu)-Cys(Acni>Pro-Pn)-Cys(StBu)-Pn>-Ala-Pn>-OH 1 tributylphosphine H-Thr(tBu)-Cys(Acin)-Pn>-PrQ-Cy$-PrtHAIa-Pro-OH

I

j Boc-N=N-Boc H-Thr-OH Boc-N-^fH-Boc +1

H-Thr(tBu)-Cy$(Acm)-Pro-Pro-Cys-PrcHAIa-Pro-OH H-Thr(tBu)-Cy$(Acm)-Pro-Pro-Cys-Pr-Ala-Pro-OH 1) I2 / 80 % acetic acid 2)TFA H-Thr-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH H-Thr-Cys-Pro-Pro-Cys-Pro-AIa-Pro-OH

+ H-Thr(tBu)-Cys-Pro-Pro-Cys(Acin)-Pro-Ala-Pr(>-OH

H-Thr(tBu)-Cys(Acm)-Pro-Pro-Cys-Pro-AIa-Pro-OH HO-Pn>-Ala-PrQ-Cys(Acm>-Pro-Pro-Cy$-Thr(tBu)-H 1)12/80% acetic acid 2)TFA H-Thr-Cy»-Pro-Pro-Cys-Pro-Ala-Pro-OH HO-Pro-Ala-Pro-Cys-Pn>-Pn>-Cys-Thr-H

Fig. 17. Synthesis of the hinge-peptide 225-232/225''232' in parallel and antiparallel alignment by selectiDe disulfide bridging methods.

934 pairing strategy based on the two S-tert-butylthio and the S-acetamido cysteine protecting groups. As outlined in fig. 17, the choice of this protection scheme allowed for a selective two-step disulfide bridging of the bis-cysteinyl-monomer to the cyclic structure (216). Upon reductive cleavage of the disulfide-type protecting group with tributylphosphine (218) and mild oxidative interchain disulfide formation with azo-dicarboxylic acid di-tert-butyl ester (219). the second disulfide bridge was generated by the iodine method in acetic acid. Exploiting the concept of high dilution in order to avoid oligomerization almost exclusively the desired parallel dimer was obtained. A similar route was applied for the preparation of the antiparallel dimer. In the iodine-oxidation step a surprisingly high percentage of parallel dimer was obtained, the formation of which could only be explained by an attack of the intermediate sulfenyl-iodide on the existing disulfide. Although this type of reaction is known from sulfur chemistry (220), to our knowledge it has not yet been observed as side reaction in peptide chemistry. Its occurrence in the present case at well-detectable extents can rationally be explained only if the hinge-peptide exhibits a high inherent preference for a parallel alignment. In order to analyze this aspect we have studied into details the oxidation of the monomeric bis-cysteinyl-peptide as well as the disproportionation of the antiparallel into the parallel dimer. Object was to determine the thermodynamically most favored product distribution and thus, the stability of the parallel hinge peptide in view of its use as a core structure for the assembly of multichain constructs (221). Studies on air oxidation of the bis-cysteinyl-peptides H-Cys-(Gly)n-Cys-OH with n = 0 to 15, have clearly shown that the nature of the oxidation products is largely dictated by the probability of collision of the thiol groups when n > 4 (222-225). Thus, under conditions of high dilution (10"3 to lO'^ M) intrachain disulfide bridging with formation of > 20-membered rings was in all cases predominant. Conversely, for n < 3 the peptide chain is not sufficiently flexible and simple statistical theory cannot be applied anymore. Besides intrachain-bridged monomers, formation of dimers and oligomers is expected to occur. Thereby sequence-dependent conformational preferences can play a decise role, particularly if oxidation is performed in aqueous media where thiol disulfide exchange reactions is shifting the product distribution towards the thermodynamically most favored structure. This has recently been fully confirmed in a detailed study on peptides containing the Cys-X-Y-Cys sequence and related to the active sites of thiol-protein oxidoreductases (226.227). In the heavy-chain hinge fragment 225-232 the intervening sequence between the two cysteine residues is Pro-Pro and the native structure of this portion of the IgGl molecule consists of a parallel dimer. By estimating for the Cys-Pro-Pro-Cys portion the propensity for chain reversal, i.e. for p-tum formation, according to

935 H-Thr(tBu)-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH CD oxidatian

H H

J '

(2)

j 1

OH OH

HO—I H '

(3)

1 '

H OH

H-

(4)

-OH

+ oligomers

Fig. 18, Air-oxidation of the bis-cysteinyl-octapeptide expected products.

225-232 of human IgGl and

Chou and Fasman (228) a p-bend potential of = 1.9x10*4 was derived which is significantly higher than the average potential of 5.5x10"^. Nevertheless airoxidation of the bis-cysteinyl-octapeptide 1 of Fig. 18 at pH 6.8 under exclusion of heavy metal catalysis and at a concentration of 3x10"^ M led to the surprising product distribution of 90:8:2 molar ratios for the compoimds 2:3:4 (221). Only in absence of water and thus, by preventing thiol disulfide exchange reactions, oxidation with azodicarboxylic acid di-tert~butyl ester was found to generate the statistically more or less expected ratios of 12:10:78 for 2:3:4. Addition of thioredoxin to the oxidation experiment in aqueous solution was found to significantly enhance the rate of air-oxidation, but without any effect on the final product distribution. These results as well as the observation that the antiparallel hinge dimer is converted again to a product distribution of 90:8:2 for 2:3:4 when incubated with the bis-cysteinyl-peptide 1, clearly confirmed the thermodynamic preference for the parallel form of the hinge-peptide (221). Formation of the correct disulfide pattern in proteins occurs concomitantly with acquisition of the correct folded form and it is driven by the thermodynamic stability of the native 3D structure. In the initial stages of protein folding processes thermodynamically stable local structures may play an important role (229-232). Thereby short range interactions are essentially implicated to promote stable core structures around which the rest of the protein chain will fold. These sequence-specific short range interactions may suffice for folding of isolated protein fragments into stable native-like structures as weU demonstrated with the bovine pancreatic trypsin inhibitor mono-cystinyl fragment 20-33/43-58 (233).

936 Similarly, sequence-specific information must be the driving force for the observed predominant parallel alignment of the hinge-peptide 225-232/225'-232' In aqueous solution. Despite the relatively small size of this proteinfiragment.the parallel alignment corresponds to an energetically highly favored structure and may represent a stable subdomain which could play an important role as

Ftg. 19. CD spectra of [H'ThrltBu)-CyS'PrO'Pr(>Cys-Pro-Ala'PrO'OH]2 (curve 1), [HThr(tBu}'Cys(Acw)'Pro-Pro-CyS'PrO'Ala'Pro-OH]2 (curve 2) and H-TtuitBu}' CyS'ProPrcyC^S'PrO'Ala'Pro-OH (curve 3). The CD spectrum of the unprotected hinge-peptide 225-232/225''232', Le. [H-Thr-Cys-Pro-Pro-CysPro-Ala-PrO'OH]2. is identical to curve 1. "chain folding initiation structure" (234) in the assembly of the immunoglobulins. This assumption is supported by the observation that the initial step in covalent interchain crossllnking of immunoglobulins is the disulfide formation between a nascent heavy chsdn and a completed one (235). In order to confirm the protein subdomaln-llke character of the hinge-peptide 225-232/225'-232' its preferred conformation was analyzed by CD and vl spectroscopy (88.236) and compared to its 3D-structure in the native proteiii. a«? mentioned above formation of the correct disulfide pattern in this proteLu fragment was expected to occur as consequence of a thermodynamically highly favored conformation. The CD spectra of the monomeric bis-cysteinyl-compound H-Thr(tBu)-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH. of the dlmerlc mono-cystinyl-

937 compound [H-Thr(tBu)-Cys(Acm)-Pro-Pro-Cys-Pro-Ala-Pro-Pro-OH]2 as intermediate of the cysteine pairing process and of the parallel dimers [HThr(tBu)-Cys-Pro-Pro-Cys-Pro-Ala-Pro-OH]2 and [H-Thr-Cys-Pro-Pro-O^s-Pro-AlaPro-OH]2 in aqueous solution at pH 7.0 are reported in Fig. 19. A comparison of these CD spectra indicate a gradual blue shift of the negative maximum with an abrupt increase of the CD intensity by a factor of 2 upon ring closure. Similarly, the weak positive maximum around 225-230 nm appears only upon formation of the second disulfide bridge. The CD spectrum of the parallel dimer with its strong negative maximum at 198 nm and the weak broad positive extremum at 228 nm is strongly supportive of a poly-(Pro)-II helical conformation (237). From comparison of the CD spectra of the synthetic precursors with that of the blscystinyl-compounds no clear evidence was obtained for the existence at the conformational equilibria of correctly prefolded conformers of the linear or intermediate mono-cystlnyl compound as driving force for the preferred parallel alignment in the oxidative folding of this protein subdomaln. However. CD spectroscopy is not sufiBclently sensitive to detect small populations of conformers suitable for favoring this process. Nevertheless, the observation that the Thr sidechain protecting group was without any detectable effect on the CD pattern was useful, since for the construction of complex molecules on the hinge scaffold this protection is necessary. These findings were further confirmed by reoxldatlon experiments on the deprotected llnesir bis-cystelnyl-octapeptlde which led to the identical results as described above for the threonlne-protected derivative (unpublished results). 2D-NMR measurements on the synthetic hinge peptide 225-232/225'-232' in DMSO combined with MD calculations allowed to define more precisely the preferred conformation of this double-stranded protein fragment (236). As shown in Fig. 20. the core portion consists of two parallel poly-(Pro)-II helices connected by the disulfide bridges of identical chirallty as in the x-ray structure of the Kolprotein (215), i.e. left-handed for Cys226/Cys226' {^^^ = -100.5') and right-handed for Cys229/cys229' (5^gg = 91-), These findings fully confirm the interpretation of the CD spectra. Superimposition of the NMR and x-ray structure of the cyclic portion gives a rms value of 1.07 A which indicates that the 3D-structure of the core hinge-peptide in solution is practically Identical to that of the parent protein in crystals. The two strands are well separated without interchain interactions due to the extended side chain conformations of the cystines which act as spacers. Nonetheless the mobility of the core structure is strongly limited as a result of the restricted conformational space of the prolines. The occurrence of short range interactions are generally believed to be responsible for the onset of local structure. These Interactions which could favor

938 the observed correct oxidative folding of the hinge-peptide are absent and therefore, this role is possibly exerted by the C-terminal extensions. The flexibility of the exocyclic portion in N-terminal direction towards the Fab. which is responsible for the movements of the Fab arms in the intact protein, cannot be

Fig, 20. Stereoview of the core hinge SD-structure as determined by NMR analysis of the synthetic hinge-peptide 225-232/225''232' in DMSO sobxtion. analyzed in the synthetic fragment due to the shortness of the extension. Although in direction Fc the mobility is increased, the preference for an extended conformation is retained; an extension of the double helical fold beyond the disulfide bridge could not be observed: Similarly, interchain interactions in this exocyclic portion were not detected in the NMR experiments. Because of the high stability of the hinge-peptide structure and its rigidity In the core portion, but flexibility of the exocyclic parts, this structural unit fulfils the requirements of a template for the construction of synthetic mlnlproteins as potential inmiunogens (206.207).

5.2. Synthesis and Conformation of Hinge-Peptide/Gastrin Chimeras Gastrin as endogeneous regulatory peptide is a self-antigen against which induction of Immune responses should be more difficult than against foreign antigens. It Is. however, well established that conjugation of self-antigens to immunogenic proteins generates the desired antibody responses. In the case of

939 gastrin this has been well documente using both conventional polyvalent (42.53) and the selective monovalent iso-1-cytochrome c/gastrin conjugate (67). Less successful were the attempts to Induce an immune response against this hormone by simple coupling of the gastrin molecule to lipophilic adjuvant molecules (175) or by its lipo-derivatization (unpublished results). Among mammalians the known gastrin sequences exhibit a highly conservative character as shown in Rg. 10. with an identical C-termlnal portion in all known sequences. Thus, interspecies sequence vsiriatlons should not be responsible for the observed immune responses in the experimental animals whilst such mutations were found to represent the main immunodeterminants in the highly conservative family of C3ôchrome c (238). Despite this expected immunological inertness of endogenous hormones, gastrin was selected as model antigen in our studies on fully s)nithetic inmiunogens for the following reasons: i) As a hormone gastrin has the ability to fold into a defined bioactive conformation at receptor level for its specific signal transduction. Conformational studies on gastrin in membrane and receptor mimicking environments led to the proposal of a potential bioactive structure of this hormone and to identify its biological relevance (82.83,239). ii) Immunological studies on gastrin/carrier conjugates clearly confirmed expression of a conformational B-cell epitope with antibody responses of gastrin receptor-like specificity, if sufficiently free conformational space is retained by the antigen in the conjugate to assume its preferred secondary structure (68). The main task of our model studies was to assess the biological and conformational properties of various hinge-peptide/gastrin constructs and to correlate these with the specificity of the antibody responses if an immune response ccould be elicited by such simple chimeric compounds. The presence of tryptophan residues in the gastrin sequence severely impairs selective disulfide bridging of linear bis-cysteinyl-hinge-octapeptide/gastrin chimeras by the approach outlined in Fig. 17 for the hinge-peptide 225-232/225'232'. since the intermediately formed sulfenyl-iodide is known to react with the indole function of tryptophan to produce thio-indole derivatives as noxious side products (240). On the other hand, synthetic steps on double-stranded cystine peptides are difficult to control analytically. Therefore, air-oxidation was chosen for the production of parallel hinge-peptide/gastrin constructs in view of the excellent results obtained with the hinge-peptide itself (221) and on the assumption that the subdomain character is retained in the colinear

940

hlnge/gastrln peptides despite the strong sequence dlflferences if compared to those adjacent to the hinge in the IgGl molecule. As outlined in Fig. 21. the hlnge-peptlde compounds extended N-termlnally with

2xa-OH+

H-Thr(tBu)-CY$ uhCYS 1 H-Thr(tBi

Cvs cvs I

OH

H-Thr(tBu)-Cvs

OH

HO

— CCyk yi

Cy,s

-OH

Cys-Thr(tBu)-H ^-I

M.A. xa-Thr(tBu)-C);s

Cvs

2xa-0H +

OH

xa-Thr(tBiu ) - c y s — C y s — OH

M.A.

xa-Thr(tBu)-CyjS

Cys

HO

Cyi.Thr(tBu).xa

Cyl

l)HCI/2-meihyUndole 2) TFA/2-methylinclole/anisolc 3) chromatography

I 1 '

OH

DHCl^-mclhylindolc 2) TFA/2-mcthylindolc/ anisole 3) chromatography

xb-

Cys

Cys

OH

xb-

-Cvs

Cys-

-OH

xb-

Cys

Cys

OH

HO-

-Cys—Cys-

-xb

J

I main product

minor side product

air oxidation xb-

-Cys-

Cys-

-OH

t tributylphosphinc xb-

- Cys(StBu)

Cys(StBu)

OH

4 1) HCl/2-mcthylindolc 2) TFA/2-methylindolc/anisole 3) chromatography xa-Thr(tBu)-Cys(StBu)

Cys(StBu)

OH

Cys(StBu)

OH

A T M.A. xa-OH +

H.Thr(tBu)-Cys(StBu)

xa = Nps-Lcu-[GIu(OtBu)l 5-Ala-Tyr(tBu)-Gly-Trp-NIe-Asp{0tBu)-Phcxb = H-Lcu-lGlulj-Ala-Tyr-Gly-Trp-NIe-Asp-Phc-

Fig.21. Scheme for the selective synthesis of the lN\e^^]-gastrinrl5'17]/hingepeptide compound in parallel and antiparallel alignment and of the parallel compound by air-oxidation.

941

gastrin in parallel and antiparallel alignment were synthesized by direct acylatlon of the side-chain protected double-chain hinge-peptlde 225-232/225'-232' followed by deprotectlon and purification (241). EfBlclent scavengers like mercaptanes were avoided in the deprotectlon step because of a possible scrambling of the disulfide bridges; nevertheless the compounds could be isolated as structurally well defined reference substances for the analytical control of the air-oxidation experiments. For latter reaction the fully deprotected colinear gastrln/bls-cysteinyl-hinge-peptlde was incubated at pH 6.8 and 10'^ M concentration under air-ojgrgen. With the help of the parallel and antiparallel reference compounds hplc allowed to establish a surprisingly high preference (>90%) for formation of the desired parallel dlmer which could then be isolated in good yields as structurally well characterized compound.

Fmoc-Thr(tBu)-Cy$—Cys—OH

I

I

+ 2H-ya

Finoc-Thr(tBu)-Cys— Cys—OH

I M.A. Fmoc-Thr(tBu)-Cys--Cvs—ya Finoc-Thr(tBu)-Cys--Cys—ya 1) TFA/2-methylindoIe/anisole 2) pipcridinc 3) chromatography H

Cys—Cys—yb

H

Cys—Cys—yb

f

H

l) air oxidation 2) chromatography Cys—Cys—yb 1) TFA/anisole/l,2-€thandithioIe 2) pipcridinc 3) iribuiylphosphine

Fmoc-Thr(tBu)-Cys(StBu)-Cys(StBu)—ya A DCC/HOSu Fmoc-Thr(tBu)-Cys(StBu)-Cys(StBu)—OH

+ H-ya

ya = -Gly-Pro-Trp-Lcu-(Glu(OtBu)] 5.AIa-Tyr(tBu)-Gly-Trp-Nlc-Asp(OtBu)-Phe-NH2 yb = -Gly-Pro-Trp-Leu-IGIu] 5-Ala-Tyr-Gly-Trp-Nlc-Asp-Phc-NH2

Fig. 22. Scheme for the synthesis of hinge-peptide/[Nle^^]-gcistrin'[2'17] in paraUel alignment.

942 Following a similar route the effect of C-termlnal extensions of the hlnge-peptlde with (Nlei5]-gastrln-I5-17] on the product distribution in air-oxidation experiments was also analyzed, and again full retention of the high preference for the parallel alignment was observed. Thus, a relatively facile synthetic access to the hlnge-peptlde/gastrln constructs was uncovered. As potential Immunogens for gastrin besides the hlnge-peptide/[Nlel5]-gastrln-[517] compound also the related INlel5j-gastrtn-[2-17J adduct was synthesized (79) In order to analyze the effect of a larger spacing between bloactlve core of the hormone and scaffold in terms of free conformational space and thus, of biological and immunological properties. For the synthesis of this chimeric compound the N-terminal threonine residue was protected as Fmoc derivative as outlined in Fig. 22 for two specific reasons: I) Acidolytic deprotectlon of the side chain-protected hlnge-peptlde 225232/225'-232' was found to be accompanied by formation of a side product at extents up to 6% (216). Detailed studies on this side product and its orlgine indicated that the N-terminal threonine residues, unexpectedly, undergo N«trifluoroacetylation by exposure to trifluoroacetlc acid even in presence of water to hydrolyse trifluoroacetc anhydride as potential contaminant, via intra- and/or intermolecular 0->N shift of the Intermediately formed Otrifluoroacetyl-threonine derivatives (242). II) It was known from previous studies that llpophillzation of antigens enhances their immunogenlcity (160). The bulky and highly hydrophobic fluorenylmethyl group at two adjacent positions of the molecule was expected to possibly play this role in the case immunizations with the underivatized compound would fail. Previous immunological studies on big-gastrin/BSA conjugates (big gastrin = gastrin-34 which contains as C-terminal portion gastrin-17) have shown that besides the dominant antigenic site located in the C-terminal part of the molecule, a second antigenic determinant is centered around the Pro-Pro-His sequence in the N-terminal portion of the molecule (243). In order to analyze the feasibility of bivalent immunogens with the hinge-template, the sequence 1-14 of big-gastrin was incorporated at the N-termini and the sequence 2-17 of gastrin (= sequence 19-34 of big-gastrin as Nle^^-analogue) at the C-termini of the hinge-peptide (79). Thereby the synthetic route via the double-ch2Lin hinge-peptide served again to produce the reference substance in parallel alignment (Fig. 23). The linear fully protected octatriacontapeptide was assembled by successive fragment condensation steps. Subsequent removal of the acid-labile side-chain protecting groups was followed by reductive cleavage of the

943

Fmoc-Thr(tBu)-Cys— Cys—ya Fmoc-Thr(tBu)-Cys— Cys—ya 1 piperidine H-Thr(tBu)-Cys—Cys—ya 2 xa-OH

+

I

I

H-Thr(tBu)-Cy$—Cys—ya j M.A.

xa-Thr(tBu)-Cys—Cys —ya xa-Thr(tBu)-Cys—Cys—ya I l)TFA/2-mcthylindole/amsoIc ^ 2) chromatography xb

Cys—Cys—yb

xb

Cys—Cys—yb k 1) air oxidation I 2) chromatography

xb

Cys— Cys—yb A 1) TFA/anisolc/l,2-cthandithiol I 2) tribuiylphosphinc

xa-Thr(tBu)-Cys(StBu)—Cys(StBu) — ya I DCC/HOSumOOBt xa-OH + H-Thr(tBu)-Cys(StBu)—Cys(StBu) —ya T piperidine Fmoc-Thr(tBu)-Cys(StBu)—Cys(StBu) — ya

xa = Pyr-Lcu-Gly-Pro.Gln-Gly-Pro-Pro-His-Lcu-Val-AIa-Asp(OtBu)-Proxb == Pyr-Leu-Gly-Pro-Gln-Gly-Pro-Pro-His-Leu-Val-AIa-Asp-Proya = -Gly-Pro-Trp-Leu-(Glu(OtBu)) 5.Ala-Tyr(tBu)-Gly-Trp-Nlc-Asp(OtBu)-Phe-NH2 yb = -Gly-Pro-Trp-Lcu-[GluI ^-Ala-Tyr-Gly-Trp-Nle-Asp-Phe-NHj

Fig. 23, Synthetic route for big-gastrin-ll'MJ/hinge-peptide/ir^^^J-gc^ S-tert-butylthlo groups, and the resulting bis-cysteinyl-peptide was then exposed to air-oxidation. As expected from the model studies, the parallel dimer was formed almost exclusively, despite the large size of the compound which as 76-

944 membered construct represents a de novo designed synthetic protein of remarkable size.

I

I

« ^

260 X(nm)

Fig, 24. CD spectra of the htnge-peptide (curve 1), [Nle^^ygastnn-[2'17] (curve 2) and hinge'peptide/[Nle^5]'gastnn-l2'17] (curve 3) tn aqueous solutiorh The computed spectrum (curve 4) was obtained by addition of the CD curves 1 and 2. The CD spectrum of hinge-peptide/lNle^^J-gastrin-lS-l?] is almost superimposable to that of curve 3.

The gastrin receptor-like specificities of the antibodies raised with the monovalent iso-l-cytochrome-c/gastrin conjugate (68) raised the question of whether the gastrin molecule folds on the surface of the carrier protein into its preferred bioactive structure as a result of the protein environment and thus is recognized as a preformed conformational epitope by the B-cell surface immunoglobulins or whether the folding of the gastrin moiety occurs upon interaction with the B-cell receptors. This important question could not be answered definitely by the CD measurements on the cytochrome conjugate (88). However, more precise information was expected from the low mass and structurally defined hingepeptide/gastrin chimeras. The extension of the double-chain hinge-peptide 225-232/225'-232' at the two Ctermini with the INle 15].gastrin sequences 5-17 and 2-17 leads to CD spectra (Fig. 24) in aqueous solution which are practically identical to those computed by

945 linearly combining the dichroic contributions of the two components (88). Similarly, an additional extension of the hinge-peptide/INle^5j.gastrin-[2-171 at the two N-termini with the big-gastrln sequence 1-14 leads to a CD spectrum in water that corresponds to the computed spectrum. In these spectra clearly the dominant contribution to the dichrolsm results from the double-stranded poly(Pro)-II conformation of the hinge-peptide core. From extensive conformational analysis of gastrin-17. big-gastrin and their related fragments by CD techniques it was concluded that in aqueous solution these peptides are essentially unordered (80). According to the dichroic properties of the chimeric compounds the gastrin-related components and the hinge portion retain their own conformational properties in water without reciprocal interferences and conformational restrictions. This fact could accoimt for the surprising observation

o

0

Fig. 25. Comparison of the CD spectra of hinge'peptide/lNle^^l'gastrin'12'17] (curve 1) and hinge-peptide/INle^^hgastrin-lS'l 7] (curve 2) in 95% TFE. that both N- and C-terminal extensions of the hinge sequence 225-232 with IgGlunrelated peptide sequences does not affect the intrinsic ability of the hingepeptide portion to fold correctly by alr-oxldation in aqueous solution into the parallel dimeric alignment.

946 Changes in solvent polarity, as obtained with aqueous TFE. have often been used to mimic the water-limited environment of biologically active peptides while interacting with target molecules such as their receptors. CD and NMR studies on gastrin and gastrin-related peptides and the corresponding biological data led to the assumption that the structure assumed by this hormone in TFE is of biological relevance (83,84). A conformational model was proposed consisting of two helical segments at the chain ends stabilized by mutual interactions in a Ushape mode. The CD spectrum of the hinge-peptide/[Nlei5]-gastrln-[5-171 in 95% TFE was found to be identical to that computed by the sum of the spectra of the constituent parts (Fig. 25). The CD spectrum of the hinge-peptide/[Nlei5]-gastrin[2-17]» however, dlifers from the computed spectrum by a significantly weaker dichroism in the 210-230 nm range with appearance of a broad negative band of weak intensity around 230-235 nm. The dichroism at 208 nm is retained suggesting helical structures as present in the gastrin molecule in TFE. The difference spectrum between the experimental and computed spectrum reminds the classical B-spectrum for type II p-tums (244). The sequence difference between the gastrin-[5-171 and gastrin-[2-17] chimeras is the extension of the exocyclic hinge portion Pro-Ala-Pro by the gastrin tripeptide Gly-Pro-Trp which leads to a sequence capable of folding into a poly-(Pro)-II helix, but also of high ptum potential. The CD spectra do not allow for a definite differentiation of these two possible conformational states of the spacer, although the difference CD spectrum is more indicative of a p-tum local conformation. The bivalent construct containing both the gastrin and the big-gastrin antigenic determinants exhibits dichroic properties in TFE which reflect in the N-terminal extensions the presence of p-tum type ordered structures (88) as already suggested by the CD studies performed on the synthetic big-gastrin fragment 1-20 (80). The results of the CD measurements strongly suggest that in the three hingepeptide/gastrin chimeras the gastrin components retain largely their ability to fold into gastrin-characteristic conformational states both in water and TFE, although quantitative evaluations cannot be drawn because of the limited accuracy of this analytical technique. Therefore more precise NMR analysis was performed and related data were used for molecular dynamics calculations (245). The final 3D-structures resulting from the molecular dynamics simulations are reported in Figs. 26 and 27. The most significant and surprising result was the loss of symmetry in both the gastrin-[5-171 and gastrin-l2-171 constructs regarding the gastrin moieties whereas the hinge-peptide core structure remaines fully constrained in its cyclic portion as already observed for the hinge-peptide

947 225-232/225'-232' itself (235). As shown In Fig. 26, in the gastrin-l2-17] dlmer at least one chain remains largely accessible in Its hairpin structure, whereas the second chain folds more tightly over its N-terminus and the spacer between gastrin and the cyclic portion of the hlnge-peptlde. This indicates that in the dimeric construct only one gastrin chain retains sufficiently ifree space to fold into its preferred conformational state and thus, to be recognized by the gastrin receptors or B-cell surface immunoglobulins.

C'terminus

N'terminus

Fig. 26. Stereoview of the starting conformation of hing€-peptide/[Nle^^]'gastTin-l2' l 7] (top) as modelledfromthe conformational states of the component parts in TFE and after 50 ps simulation (bottom). In the case of the shortened gastrin construct, i.e. hlnge-peptide/lNlel^J-gastrln[5-17], strong interferences of the peptide chains, particularly in their central portion, are observed which allow Just the C-termlnal part of one gastrin molecule to protrude into free space (Fig. 27). As a result of the packing of the two gastrin chains the accessibility and the conformational space is greatly restricted and should allow recognition of just the C-termlnus of one of the two gastrin moieties by target receptor molecules. For the molecular d3niamlcs simulations the

948 conformation of the single components In TFE was used to buUd the starting structures since according to the CD spectra these conformations are more or less retained in the chimeric compounds in TFE. The partial coUaps of these structures in the simulations confirm once more that TFE is a too strong a-helix inducer (246.247). The results of these conformational studies clearly reveal the disadvantages of template-assembled synthetic constructs in terms of interchain interferences. If defined secondary structural elements like ap, aa. PaP have to be stabilized by direct anchorage on the template, successful application of this approach has been repetedly reported (203). More globular ordered structures as in the present case need more conformational space which can be created by larger spacers. In view of this result it is not surprising that multiple attachment of protein fragments on a branched lysine carrier may trigger immune responses uncapable of recognizing the parent proteins (123).

C-terminus

N-terminus

i

Fig, 27. Stereoview of the starting conformation of hinge-peptide/lNle^^l-gastrin-fd' 17] (top) as modeUedfrom the conformational states of the conqx>nent parts in TFE and after 50 ps stmidation (bottom)

949 5.3, Bioacttvities of the Hinge-Peptide/Gastrin Chimeras The hlnge-peptide/gastrln chimeras contain the fully active [Nlei5]-gastrln-[2-17] and [Nlei5j-gastrln-[5-17] sequences (85) dimerlzed at their N-termini via the bridging sequence Pro-Ala-Pro-Cys Cys-Pro-Ala-Pro which according to the NMR studies on the hinge-peptide is restricted in its mobility, but in an extended secondary structure that should assure more conformational free space than present in most of the literature known dimeric hormone preparations (248-253). The gastric acid stimulatory potency of the dimeric gastrin constructs were compared with those of the parent gastrin analogs [Nlei^j-gastrin and [Nle^S]gastrin-[5-17] (245). As shown in Table 2, on molar basis the gastrin-[2-171 dimer was found to be as potent as the parent hormone: however, the hingepeptlde/gastrin-[5-171 dimer exhibited only a 7-fold reduced potency which is comparable to those of C-tenninal gastrin peptides, i.e. of the tetra- and

Gastrin compound

ED50 (pmol/kg)

IC50 (nm)

[Nlel5]-gastrin or [Nle 15]-gastrin-[5-17]

26.5 ± 10.0

0.4

hinge-peptide/[Nle 15].gastrin-[2-17]

23.5 ± 4.50

0.5

hinge-peptide/ [Nle 15] -gastrin-[5-17]

179.5 ± 10.0

30

hinge-peptide/des-amido-[Nle 15]-gastrin-[5-17]

-.-

-.-

[Nle 15]-gastrin-[5-17] /hinge-peptide

-.-

-.-

1

Table. 2, ED^gfa^ ^ ^^^ stimulation of gastric acid secretion in rats by gastrin and hinge-peptide/gastrin chimeras and their receptor binding affinities in parietal cells isolated from rabbit gastric fundus.

heptagastrin (254). The dimers lacking the C-terminal amide were inactive as expected from previous structure function studies on gastrin peptides since this structural element is known to represent a crucial factor for the hormonal activity of gastrin (254). These des-amidated compounds were also devoid of any antagonistic activity. This was further confirmed by their inability to displace the receptor-bound radiolabelled ligand from isolated parietal cells (see Table 2). In this receptor binding assay the gastrin-15-17) dimer exhibited again a remarkably lower affinity than the parent hormone, i.e. 1%, whereas the gastrin-[2-17] dimer

950 was recognized on a molar base with Identical affinity as the monomerlc form. In terms of gastrin equivalents the dlmerlc construct exhibits half the potency and receptor affinity of the monomerlc gastrin. This raises the question of whether both gastrin components are half as potent as the monomer due to mutual Interferences, e.g. sterlcal hindrance or reduced conformational space for optimal receptor Interaction, or whether In the dlmerlc construct only one gastrin moiety remains fully accessible with the second one restricted In Its conformational space, thus preventing a flip-flop-type receptor Interaction as one of the possible mechanisms responsible for the enhanced activities observed for several peptide hormone dlmers (253). Latter working hypothesis Is supported by the conformational studies discussed above which could also explain the low potency and receptor affinity of the gastrln-[5-17] dlmer. If the restricted conformational space Is responsible for the observed biological properties a similar pattern was expected from Immunization experiments where again the dlmerlc compounds have to be recognized by the B-cell surface receptors In a more or less accessible form.

5,3 Immunogenicities of the Hinge-Peptide/Gastrin Chimeras Standsird protocols were used for Immunizing guinea pigs with the hingepeptide/gastrin compounds and with the monovalent lso-1-cytochrome-c/gastrin conjugate as reference (245). The titers of anti-gastrln antisera were determined by standard ELISA procedures and are listed in Table 3. Significant anti-gastrin antisera titers were obtained which were comparable to those raised by gastrin/protein conjugates indicating that the synthetic constructs behave as fully competent Immunogens. Interestingly the immune response against the hinge-peptide portion was found to be very weak or none. In order to avoid misleading results deriving from nonadsorption of the hinge-peptide or from inadequate exposure of this peptide on the polystyrene surface a hinge-peptideA^P-[4-18] construct was used as antigen in the ELISA. With the chimeras containing N-terminally the antigenic peptides and the sole threonine residue as spacer very weak crossreactivities with the hingepeptide antigen were detected possibly resulting from the generation of an overlapping, although weak new B-cell epitope. This was confirmed by the antisera raised against the [Nlel5]-gastrin-[5-17]/hinge-peptide which did not crossreact with intact human IgGl although weakly recognizing the hinge-peptide antigen. Unmodified gastrin was found to be non-immunogenic in guinea pigs as previously already observed in rabbits and mice where even conjugation of gastrin-l2-17] with the muramyl-dipeptide or a lipophilic muramyl-tripeptlde

951

Immunogen

Anti-gastxin antibody titers* i

hlnge-peptide/[Nlel5]-gastrin-[2-17]

3.64

hinge-peptlde/[Nlel5)-gastrin-[5-17]

4.02

hlnge-peptide/des-amldo-INle 15].gastrln-[5-17)

3.45

[Nle 15j .gastrln-[5-17)/hlnge-peptlde

3.20

big-gastrin-11-14] /hinge-peptide/ [Nle ^ ^j -gastrln-I2-17]

3.55

[Nle^Sj.gastrin iso-l-cytochrome-c/gastrin-[2-17J

-.3.37

Table 3. Immune responses in guinea pigs to hinge-peptide/gastrin chimeras and to iso-l'Cytochrome-c/gastrin for comparison; *) the titers are averages of the antisera of three animals.

derivative failed to induce immunogenicity (175). Consequently, the observed immunogenlcity of the hinge-peptide/gastrin chimeras has to arise primarily from the hinge-peptide portion, although the function of this moiety at cellular level is not yet clear. A possible pathway could derive from reaction of the disulfide bonds of the synthetic chimeras with cysteine-containing endogenous proteins which would present the hinge-peptide/gastrin as macromolecularized antigen to the immune competent cells. The stability of the hinge disulfides which should be in the order of that of intact IgGl, makes this hypothetical pathway rather unlikely. Moreover, pulse/chase experiments on human Epstein-Barr virus-transformed B-cells have shown that the hinge-peptide/[Nlel5]-gastrin-[2-17J compound is internalized and bound to MHC class II from which it is released in acidic media in its intact form as monitored by a comparative hplc analysis with a fluorescent probe of the hinge-peptide/gastrin compound (245). The recovery of the intact dimer upon internalization indicates a remarkable stability of the bis-cystinyl structure in the in vitro experiments, although quantitative estimations are difficult in such pulse/chase experiments. On the other hand, the high degree of paralleUsm between the receptor binding assay in vitro and the gastric acid stimulation potency determined in vivo excludes that reduction, disproportionation or thiol

952 disulfide exchange reactions with serum or cellular proteins is occurring at extents capable of sensibly affecting the results.

Fig. 28, Stereoview of the MHC class U protein with the bound hinge-peptide 225232/225''232'as resulting from modelling experiments and energy mintmization of the complex. A) Lateral view; B) view along the binding cleft. The MHC class II molecules bind T-cell epitopes apparently with less defined restriction regarding the peptide size than the MHC class I molecules and as known from the x-ray crystallographic studies preferentially in an extended

953 conformation (43,44). In this context the hlnge-peptlde portion In the gastrin constructs Is folded in an extended poly-(Pro)-II conformation, but as a dlsulfidellnked dlmer. The pulse/chase experiments and the observed immunoglobulin Isotype switch to IgG would suggest that the hlnge-peptlde portion Is playing the role of an efficient T-cell epitope. Modelling experiments performed with the use of the x-ray structure of the MHC class II protein DRB1*0101 (30.31) showed an excellent fit of the double-stranded hlnge-peptlde with a surprisingly good overlapping of the peptide backbone of one strand with that of the bound and cocrystalllzed Influenza virus peptide (Fig. 28). Additionally, the second strand of the hlnge-peptlde is only minimally protruding out of the binding pocket whereby the double stranded molecule apparently occupies most of the binding pocket with contacts to the bottom and lateral helices in Van der Waals dlstancles. In crystals of the IgGl Kol-protein the hinge region is in an extensive close contact with the hypervariable segments of a second molecule leading to well-defined and strong densities for all the residues Involved. As the loss of surface area of the hypervariable segments in this contact is 1314 A^ the observed Interaction in the crystals has been classified as "antigen-like" binding of the hinge segment to the antigen-binding site of the IgGl (215). Moreover, it is well known that natural IgGanti-F(ab')2 antibodies belong to the immune repertoire of healthy individuals and represent potent immunoregulatory molecules (255-257). It has been suggested that some of the £intl-F(ab')2 antibodies recognize conserved domains of the F(ab')2 portion of IgGl (255.258). Recently, with the help of the synthetic hlngepeptide 225-237/225'-237' which contains an extended lower hinge in comparison to the hinge peptide used for the construction of the gastrin chimeras, it has been demonstrated that these IgG-antl-F(ab')2 antibodies bind strongly to this double-chain hinge fragment as a conformational epitope expressed or stabilized by extension of the lower hinge sequence to position 237 (259). CD spectroscopy allowed to detected in this extended hinge fragment a stabilized exocyclic P-tum which could possibly contribute to the expression of the epitope. Since all molecules Involved in Immune regulation, e.g. surface B-cell receptors, MHC molecules, Fcy-receptors, rheuma factors, are characterized by a similar structural assembly, the hinge segment could possibly represent an universal type epitope and thus, be Involved in various pathways as recently demonstrated in the case of the Fey receptors (260) and rheuma factors (259). In cell culture experiments the lgG-anti-F(ab')2 antibodies are known to suppress B-cell proliferation (257). Correspondingly, treatment with the hingepeptide/gastrin chimeras could block these IgG-anti-(F(ab')2 antibodies in the serum and as a consequence, the B-cell Immune response would be stimulated. Alternatively, as reported by Roosnek and Lanzavecchia (261). rheuma factors-

954 producing cells exert a particular antigen-presenting function. Immunization with an antigen leads to production of antibodies which upon complexing the circulating antigen are recognized by the rheuma factor producing B-cells. The immune complex is internalized, processed and fragments of the antigen are presented to the T-cells which then exert their immunostimulating function. As the anti-F(ab')2 antibodies which belong to the class of rheuma factors, recognize the hinge-peptide there are anti-hinge B-cells capable of recognizing and internalizing the hinge-peptide/gastrin chimeras and thus, of presenting these molecules or parts of them to T-cells with subsequent immunostlmulation. Finally, injected hinge-peptide/gastrin constructs are recognized by the anti-hinge antibodies, the Fc of which binds to the Fcy-receptor of macrophages. The complex is internalized, processed and the antigen presented to the T-cells.

216

225

Human yl E P K S C D K T H T Mouse Y l V P R D C G Guinea p i g y l Q S W G H T Rabbit y A P S T C S K P M

232

CPP CKPCI C P P C I P C

238

C P A P E L L G G P CTVP B V S C G A P Z L L G G P P P P E L L G G P

Fig. 29. Comparison of the hinge sequence of human IgGl with those of guinea pig, rabbit and mouse. The hinge-peptide sequence used for the design and synthesis of the gastrin chimeras is related to human IgGl and inmiunizations were performed in guinea pigs. An alignment of the genetic hinge from different species is reported in Fig. 29. The lower hinge of human, guinea pig and rabbit IgGl shows the same size and very similar sequences, whereas in the mouse IgGl this hinge portion is significantly shorter. The high degree of sequence homology between the human and guinea pig hinge may possibly be responsible for the strong immune responses in guinea pigs if this is induced by one of the mechanisms discussed above. Supportive for this hypothesis is the observation that in Balb/c and C57BL/6 mice the hinge-peptide/gastrin chimeras were found to trigger no or very weak immune responses (Gemeiner, M. and Moroder, L., unpublls ;c:d results). Specificity of the Immune Response to Hinge-Peptide/Gastrin Chimeras The specificity of the antisera raised against the various hinge-peptide gastrin constructs was determined using a whole set of gastrin-related peptides (245).

955 Both the hlnge-peptlde/[Nlei5]-gastrln-[2-171 and the big-gastrin-[l-141/hlngepeptlde/INlei5]-gastrln-[2-17I generated antibodies of monoclonal-type character In all animals regarding specificity vs. the gastrin antigen. As shown in Fig. 30, the C-termlnal gastrin peptide amides 14-17 and 13-17 are recognized to negligible extents. Similarly, the complementary sequence 1-13 and its constituent subfragments 1-5 and 4-13 are not competitive ligands. Expression of an immunodeterminant sequential epitope in the antigen gastrin can therefore be excluded. Extension of the C-termlnal gastrin peptides vs. the characteristic pentaglutamic acid sequence produces a sharp transition of the binding affinities upon incorporation of four glutamic acid residues to reach a maximum value at the level of the biologically fully active [Nlei5]-gastrin-[5-171 and [Nlei^j-gastrin

IC«,[%]

Pyr Gly Pro Trp Leu Glu Glu Glu Glu Glu Ala lyr Gljr Trp Met Asp Phe-NH,

1 2 3 4 5 6 7 8 9 10 11 1213 14 1516 17

Fig. 30. Crossreactivities of gastrin related peptides as determined by 50% inhibition of binding of gastrin to antigastrin antibodies raised against hinge'peptide/[Nle^^]-gastrin'l2-17] in guinea pigs, and expressed as percentage of the IC50 of gastrin. The gastrin peptides analyzed are: gastrin-[14-17] (^), lPyr^0^j^l5Ugastnn-ll0'17] (0), IPyT^^Nle^^j-gastrin-lQ17] (•). [Pyi^,Nle^5lgastnn-[8'17] (O), [Pyr7,Nle^5].gastnn-l7-17] (A), [Pyf^.N]e^^]-gastrin-[6-17] (A). [Nle^5J-gastrin'[5'17] (•) and [Nle^^l-gastrin (or gastrin) (Q).

sequences. Since the biologically inactive des-amido-[Nlel5]-gastrin-l5-17) behaves as a very weak llgand, the experimental data indicate that the gastrin moiety in the synthetic construct is recognized by the B-cell surface immunoglobulins as a

956 conformational epitope similar to the bioactive structure of the hormone at receptor level. This hypothesis is further supported by a comparison of the parallelism between bioactivity. gastrin receptor affinity and onset of conformational order shown in Fig. 31 and by the competitive affinity of the gastrin peptides for the antl-gastrln antibodies. Moreover, the correlation of the crossreactivles of these antibodies with gastrin analogs of differentiated blopotencies is surprisingly good. This allows to conclude that with the hlngepeptide/gastrin constructs in which the bioactive core sequence of the hormone, i.e. 6-17, is sufficiently spaced from the template to allow for at least one gastrin moiety to fold into its preferred structure, antibodies are obtained of gastrin receptor-like specificity.

5 10 15 Number of residues In the chain

Fig, 31. Relative variation of molar ellipticity values (Al]/[Q]ioi) at 216 (0) and 192 run (A) as a fimction of chain length. Data of hormonal potency of the various gastrin peptides are reported for comparison (•).

Conversely, Immunizations performed with the hinge-peptide/[Nlel5]-gastrin-l517] compound clearly revealed a restricted accessibility of the shortened gastrin moieties. In fact, the antibodies raised against this immunogen completely differ in their specificity from those described above. All crossreactivity data obtained with the gastrin related peptides indicate expression of an epitope related to the C-terminal portion of the molecule and most probably of sequential character, as amino acid replacements not seen by the gastrin receptor in this portion of the molecule are strongly affecting the binding affinity for the antisera. These data are in full agreement with the conformational properties of this chimeric construct which clearly indicated strong interchain interferences with restriction of the

957 conformational space and with free accessibility limited to the C-terminal portion of the gastrin moiety. Conformational studies on shortened gastrin peptides lead to the conclusion that their extension to the critical length of gastrin 7-17 is necessary to allow for the hairpin conformation to be assumed. Therefore free presentation of just the C-terminal portion of gastrin to the B-cell surface iipmunoglobulins allows for recognition of this part only as sequential epitope. Very similar antibodies were obtained with the hinge-peptide/des-amido-INle^^j. gastrin-[5-17] dimer. The dimeric gastrin compound [Nlel5l-gastrin-[5-171/hinge-peptide in which the gastrin molecule is linked C-terminally to the hinge-peptide and thus, in a biologically inactive form as confirmed in the bioassays, generates an antibody population capable of recognizing only the N-terminus of the gastrin-[5-17] antigen (245). The antibodies crossreact with gastrin peptides containing the Nterminal portion Leu-Glu-Glu independently of whether the amino function is free or acylated. Gastrin-peptides lacking this portion are not anymore recognized indicating again the expression of a sequential epitope restricted to this short portion which in analogy to what was observed for the hinge-peptide/[Nlei^)gastrin-[5-17] probably represents the only segment fully accessible whereas the rest seems to be shielded by interchain interferences. Interestingly, immunizations with the bivalent immunogen big-gastrin-[l14]/hinge-peptide/[NlelSl-gastrin-[2-17I led to immune responses containing high titers of antigastrin antibodies of gastrin receptor-like specificity and low titers of antibodies directed against the N-terminal antigen. These antibodies are unable to recognize fragments of the big-gastrin-[l-141 sequence like the peptides 1-6 and 714 (245). The peptide 7-14 contains at its N-terminus the peculiar sequence ProPro-His which according to the conformational studies discussed above is probably involved in a P-tum type conformation. Endgroup effects can prevent the onset of this secondary structure element in the fragment 7-14 and thus, the crossreactivity with the antibodies unless this lack of binding affinity results from charge effects. The results again confirm that even the second antigenic site retained sufficient conformational free space to express its sequence specific conformational epitope. The results of the immunization experiments clearly underline the difficulty of presenting antigens to the immune system in a manner that intrinsic conformational properties can fully evolve and trigger the immune responses. Conjugation of peptides to proteins is known to suppress expression of epitopes and to provoke immune responses capable of recognizing the peptide but not its intrinsic conformational epitope as present on the protein surface. Bundled antigens on multiple antigen presenting systems are characterized by similar

958 shortcomings since interchain interactions may lead to restricted conformational space as well documented in the model studies with the hinge-peptide constructs. Therefore in the de novo design of synthetic immunogens particular attention should paid to the conformational analysis if better insights into the mechanism of immune recognition is the goal.

Conclusion Cell-surface receptors can be divided into three classes, depending on whether information is transmitted by allosteric conformational changes, by receptor dimerlzation or by receptor aggregation (262). The signal transduction of the hormone gastrin occurs upon interaction with a membrane-bound receptor with seven putative transmembrane helices (263) where allosteric conformational changes are most probably transmitting the signal, although clustering of hormone receptors may not be excluded and could possibly be responsible for the higher potencies observed for some hormone dimers. Recently a detailed investigation of bradykinin antagonist dimers has clearly demonstrated a decisive role of the size of the spacer used in the dimerlzation (253), a fact which could support contemporary interaction with two receptors. Dimerization of regulatory peptides has repeatedly been used in the search for superactive agonists and antagonists, but with contradictory results in terms of the resulting hormonal potencies. Detailed studies on the conformational effects of these dimerizations as potential factors responsible for the unpredictable biopotencies of dimers of peptide hormones have not been reported so far. In the case of the hingepeptide/gastrin dimers the conformational analysis clearly revealed a strong correlation of the differentiated biopotencies with different degrees of interchain interactions of the gastrin moieties and correspondingly, with their accessibility for recognition by the receptors. Aggregation-activated receptors are frequently encountered in the immune system (264-266). The members of this receptor class bear short cytoplasmatic domains which act to bind and recruit other cellular factors following the aggregation of their extracellular domains. Aggregation of these surface receptors, e.g. surfaceimmunoglobulins of B-cells. could be favored by multiple presentation of the antigens (142). Clustering of these receptors would lead to differentiation of the Bcells into plasma cells and to their proliferation (267.268). The dimerization of the antigen gastrin in the hinge-peptlde constructs hardly allows for such concomitant capping of two B-cell receptors as the differentiated immune response to the hinge-peptlde/[Nlei5]-gastrin-[5-171 and hinge-peptide/lNlel^jgastrin-l2-17] is more in agreement with the conformational preferences resulting

959 from the strong intrachain interferences which allow for recognition and binding of solely one gastrin chain. Nonetheless strong immune responses were obtained with these dimerlc constructs, and these have then to be attributed to the peculiar properties of the hinge-peptlde portion. In view of the multiple functions of the hinge segment in the immunoglobulin molecules it seems proper to interpret the experimentally observed strong immune responses in terms of specific recognition of the hinge portion by immimoglobulln-related molecules involved in the mechanisms of immtmlty. In this sense the hlnge-peptide could play the role of an universal recognition site, since molecular modelling indicates a surprisingly good fit of this double-chain peptide even into the binding cleft of MHC class II proteins. For this reason constructs with the hlnge-peptide could represent Interesting tools for Investigating at molecular level the intricated mechanisms Involved in Immunology. References 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17)

Brown, F. (1994) Phil Trans. R Soc. Lond. B 344, 213-219. Rabinovich, N.R., Mclnnes, P., Klein, D.L. and HaU, B.F. (1994) Science 265, 1401-1404. La Montagne, J.R. and Rabinovich. N.R (1994) Int. J. Technol Assess. Hecdth Care 10, 7. Amon, R. (1986) TIBS H , 521-524. Sela, M. and Amon, R. (1992) Vaccine 10, 991-999. Amon, R. (1972) in: Immunity and Viral and Rickettsial Diseases (Kohn, A. and Klingberg, A.M., eds.) pp. 209-222, Plenum Press, New York. Sela, M. (1974) BuIL Inst. Pasteur 72. 73-86. Anderer, F.A. (1963) Z. Naturforsch. 18, 1010-1014. Anderer, F.A. (1963) Biochim. Bvophys. Acta 7 1 , 246-248. Langbeheim. H., Amon, R. and Sela, M. (1976) Prcx:. Natl Acad. Set USA 7 3 , 4636-4640. Senyk. G., Williams, E.B., Niteckl, D.E. and Goodman, J.W. (1971) J. Exp. Med. 133, 1294-1308. MiUch. D.R., McLachlan, A., Chisari, F.V. and Thornton, G.B. (1987) Nature 329, 547-549. DlMarchi, R., Brooke. G., Gale. C . CrackneU. V.. Doel. T. and Mowat. N. (1986) Science 232. 639-641. Francis. M.J.. Fry. CM.. Rowlands. D.J.. Brown. F.. Bittle. J.L.. Houghten. R.A. and Lemer. R.A. (1985) J. GerL Virol 66. 2347-2354. Francis. M.J.. Fry. CM.. Rowlands. D.J.. Bittle. L.. Houghten, R.A.. Lemer. R.A. and Brown, F. (1987) Immunology, 61.1-6. Antonl. G.. Mariani, M.. Presentlni. R.. Lafata. M.. Neri. P.. Bracci. L. and CianfrlgUa (1985) Mol Immunol 28. 1237-1241. Emlni. E.A., Bradford, A.J. and Wimmer, E. (1983) Nature 304, 699-703.

960 18) Shapira. M,. Jibson. M., Muller. G. and Amon. R. (1984) Proc, Natl Acad. Set l/SA 8 1 . 2461-2465. 19) Young. C.R and Atassi. M.Z. (1982) iTramm. CommuTL 11. 9-16. 20) Mariani. M.. Bracci. L., Presentlni. R . Nucci. D.. Nerl. P. and Antonl. G. (1987) MoL JmmunoL 24. 297-303. 21) Claman. H.. Chaperon, E. and Trlplett. R. (1966) Soc. Exp. Biol Med. 122. 1167-1171. 22) Landsteiner. K. (1946) in: The Specificity of Iwimmological Reactions, Harvard University Press. 1946. 23) Mitchison. NJL (1971) Eur. J. Immimol 1. 18-27. 24) Katz. D.H. (1980) Adv. Immun. 29. 137-207. 25) Benacerraf. B. and McDevltt. H. (1972) Science 175. 273-279. 26) Guillet. J.-G.. Lai. M.-Z.. Brlner. T.J.. Buus. S.. Sette. A.. Grey. H.M.. Smith. J.A. and Gefter. M.L. (1987) Science 235, 865-870. 27) Rosenthal. A.S. (1978) Immunol Rev. 40. 136-152. 28) Berzofsky. J.A.. Cease. K.B.. Comette. J.L.. Spouge. J.L.. Margalit. H.. Berkower. I.J.. Good. M.F.. Hiller. L.H. and DeUsi. C. (1987) Immunol Rev. 98. 9-52. 29) Benacerraf. B. (1978) J. Immunol 120.1809-1812. 30) Brown. J.. Jardesky. T.. Gorga. J.. Stem. L.. Urban. R . Strominger. J. and Wiley, D. (1993) Nature 364, 33-39. 31) Stem. L., Brown, J.. Jardesky. T., Gorga, J., Urban. R , Strominger. J. and Wiley, D. (1994) Nature 368, 215-221. 32) Bjorkman, P., Saper, M., Samraoui, B., Bennett. W., Strominger, J. and Wiley. D. (1987) Nature 329. 506-511. 33) Bjorkman. P., Saper, M., Samraoui, B., Bennett, W., Strominger, J. and Wiley, D. (1987) Nature 332, 512-518. 34) Saper, M., Bjorkman, P. and Wiley D. (1991) J. Mol Biol 219, 277-319. 35) Fremont, D.. Matsumura, M., Stura, E., Peterson, P. and Wilson, I. (1992) Science 257, 919-927. 36) Matsumura, M., Fremont, D., Peterson, P. and Wilson, I. (1992) Science 257, 927-934. 37) Rammensee, H.-G., Falk, K. and Rotzschke, O. (1993) CUJT. Immunol 5, 3544. 38) Rotzschke, O. and Falk, K. (1991) Immunol Today 12, 447-455. 39) Engelhard. V.H. (1994) Curr. Immunol 6. 13-23. 40) Chicz, RM. and Urban, R G . (1994) Immunol Today 15, 155-160. 41) EUiott, T. (1991) Immunol Today 12, 386-388. 42) SinigagUa, F. and Hammer, J. (1994) Curr. Immunol 6, 52-56. 43) Rothbard, J.B. (994) Curr. Biol 4, 653-655. 44) Cresswell, P. (1994) Curr. Biol 4, 541-543. 45) Das, M. and Fox, F. (1979) Rev. Biophys. Bioeng. 8, 165-193. 46) Han, K.-K., Richard, C. and Delacourte, A. (1984) Int. J. Biochem 16, 129145. 47) Moroder, L., Gemeiner, M., Kalbacher, H., Nyfeler, R and Wiinsch, E. (1983) in: Peptides 1982 (Blaha, K. and Malon. P., eds.) pp. 667-672. Walter de Gruyter, Berlin, New York.

961

48) Moroder. L.. Nyfeler, R.. Gemeiner, M.. Kalbacher, H. and Wunsch, E. (1983) Biopolymers 22, 481-486. 49) Brland. J.P.. Muller, S. and Van Regenmortel. M.H.V. (1985) J. Immunol Meth, 78, 59-69. 50) Van Regenmortel. M.H.V., Briand, J.P., MuUer, S. and Plaue, S. (1988) in: Synthetic Polypeptides as Antigens (Burdon, R H . and van Knlttenberg, P.H., eds.) pp. 95-216, Elsevier, Amsterdam. 51) Reichlin. M. (1980) MetK Enzymol 70, 159-165. 52) Geiger, R., Moroder, L., and Wunsch, E. (1984) in: Peptides 1984 (Ragnarsson. U., ed.) pp. 451-456. Almqvist and Wicksell Int.. Stockholm. 53) Rehfeld. J.F. (1978) J. Biol Chem 253. 4016-4021. 54) Yanalhara, Ch., Yanahaira, N., Shimlzu, F., Sato, H., Uehata, S. and Imagawa, K. (1980) Biomed, Res. 1, 242-247. 55) Turkelson, CM., Dale, W.E., Reidelberger, R. and Solomon. T.E. (1986) ReguL Peptides 15. 205-217. 56) Wunsch, E., Moroder, L., Nyfeler, R.. Kalbacher. H. and Gemeiner. M. (1985) Biol Chenh Hpppe-Seyler 368. 855-861. 57) Moroder, L.. Tzougraki. Ch.. Gohrlng. W.. Mourier. G.. Musiol. H.-J. and Wunsch. E. (1987) Biol Chenh Hoppe-Seyler 368, 855-861. 58) KeUer, O. and Rudinger, J. (1975) Helv. Chtm Acta 58, 531-541. 59) Marrian, D.H. (1949) J. Chenh Soc. 1515-1516. 60) Wold, F. (1967) Meth. Enzymol 11, 617-640. 61) Green, N., Alexander, H. Olson, A., Alexander, S., Shinnick, T.M, Sutcliffe, J.G. and Lemer, R.A. (1982) CeU28, 477-487. 62) Eritja, R., Ziehler-Martin, J.P.. Walker. P.A., Lee, T.D., Legesse, K., Albericcio, F. and Kaplan, B.E. (1987) Tetrahedron 43, 2675-2680. 63) Batz, H.-G., Hubner. C , Kerscher. L.. Klein. C. and Kunz. H. (1989) in: Peptides 1988 (Jung, G. and Bayer. E., eds.) pp. 754-756, Walter de Gruyter, Berlin. New York. 64) Routhe. M.R., Bowman, G.J., Gangitano, J.M., Lambing, J.L., Pratt, S.M. and Hopp, T.P. (1994) in: Peptides - Chemistry, Structure and Biology (Hodges, R.S. and Smith, J.A., eds.) pp. 159-160. ESCOM, Leiden. 65) Munson, M.C., Lebl, M., Slaninova, J. and Barany, G. (1993) Peptide, Res. 6, 155-156. 66) motz, I.M. and Heiney. R.E. (1962) Arch. Biochenh Biophys. 96. 605-612. 67) Moroder, L., Mourier, G., Bovermann, G., Dufresne, M., Gohring, W., Gemeiner. M. and Wunsch. E. (1987) Biol Chem. Hoppe-Seyler 368, 849-853. 68) Moroder, L., Mourier, G., Dufresne. M.. Bovermann. G.. Gohring. W., Gemeiner, M. and Wunsch, E. (1987) Biol Chenh Hoppe-Seyler 368, 839-848. 69) Wunsch. E.. Moroder. L.. Gillessen. D. and Soerensen. U.B. (1982) HoppeSeyler's Z. Physiol Chenh 363. 665-669. 70) Krois, D.. Dufresne, M.. Neumann, H., Moroder, L. and Wunsch, E. (1990) Biol Chenh Hoppe-Seyler 371, 43-48. 71) Dickerson, R.E. and Timkovich, R.T. (1975) in: The Enzymes (Boyer, P.D., ed.) pp. 397-547, Academic, Press, New York. 72) Reichlin. M., NisonofT, A. and MargoUash, E. (1970) J. Biol Chenh 245, 947954.

962 73) Schwartz. R.H., Fox. B.S.. FVaga. E.. Chen. C. and Singh B. (1985) J. Immunol 135, 3028-3033. 74) Shell. J.M.. Schell. T.D.. Shepherd. S.E.. Kllmo. G.F.. Kioschos. J.M. and Paterson. Y. (1994) Eur. J. Immunol 24. 2141-2149. 75) Narita. K. and Tltanl. K. (1969) J. Biochem. 65. 259-267. 76) Motonaga. K.. Misaka, E.. Nakajlma. E.. Ueda. S. and Nakanishl. K. (1965) J. Biochem. 57. 22-28. 77] Dickerson. R.E., Takano. T.. Elsenberg. D.. Kallal. O.B., Samson. L.. Cooper. A. and MargoUash. E. (1971) J. Biol Chem. 246. 1511-1535. 78) Salemme. F.R.. Freer. S.T.. Xuong. N.H.. Alden. RA. and Kraut. J . (1973) J. Biol Chem. 218. 3910-3921. 79) Wunsch. E.. Moroder, L.. Hubener. G.. Musiol. H.-J., von Grunigen. R.. Gohrlng. W.. Scharf. R. and Schneider. C.H. (1991) Int. J. Peptide Protein Res, 37. 90-102. 80) Peggion. E.. Jaeger. E.. Knof. S.. Moroder. L. and Wunsch. E. (198 ) Biopolymers 20, 633-652. 81) Peggion. E., Foffani. M.T., Wunsch, E., Moroder. L.. Borin. G.. Goodma . vf and Mammi, S. (1985) Biopolymers 24, 647-666. 82) Mammi, S., Mammi, N.J. and Peggion, E. (1988) Biochemistry 27, 1374-i .79. 83) Mammi, S. and Peggion, E. (1990) Biochemistry 29, 5265-5269. 84) Mammi, S., Foffani, M.T., Peggion. E.. Galleyrand. J . C . Bali. J.P.. Simonetti. M., Gohring. W.. Moroder, L. and Wunsch, E. (1989) Biochemistry 28, 71827188. 85) Gohring, W., Moroder, L., Borin, G., Lobbia, A., Bali, J.-P. and Wunsch, E. (1984) Hoppe-Seyler's Z. Physiol Chem. 365, 83-94. 86) Moroder, L., Bovermann. G., Mourier, G., G6hring, W., Gemeiner, M. and Wunsch, E. (1987) BioL Chem. Hoppe-Seyler 368, 831-838. 87) von Grunigen, R., Siglmuller, G., Papini, A., K6cher, K.. Traving, B., Gohring. W. and Moroder, L. (1991) BioL Chem. Hoppe-Seyler 372, 163-172. 88) Moroder, L., Bali, J.-P. and Kobayashi, Y. (1991) Biopolymers 31, 595-604. 89) Moroder, L., Papini, A., Siglmuller, G., Kocher, K., Dorrer, E. and Schneider, C.H. (1993) BioL Chem. Hoppe-Seyler 373, 315-321. 90) Moroder, L. and Temussi, P.-A. (1993) in: Peptides 1992 (Schneider, C.H. and Eberle, A.N., eds.) pp. 813-814, ESCOM, Leiden. 91) Martinez, J. (1990) in: Comprehensive Medicinal Chemistry (Hansch, C , Sammer, P.G. and Taylor, J.B., eds.) Vol. 3. pp. 925-959, Pergamon Press, New York. 92) Moroder, L., Romano, R., Weyher, E., Svoboda, M. and Christophe, J . (1993) Z. Naturforsch. 48b. 1419-1430. 93) Moroder, L., D'Ursi, A , Picone, D., Amodeo, P. and Temussi, P.A. (1993) Biochem. Biophys. Res. Commun. 190, 741-746. 94) Shinnik, T.M., SutcUffe, J . C , Green, N. and Lemer, R.A. (1983) Ann. Rev. Microbiol 37, 425-426. 95) Barlow, D.J., Edwards, M.S. and Thornton, J.M. (1986) Nature 322, 747-748. 96) Lemer, R.A. (1984) Adv. Immunol 36, 1-4. 97) Van Regenmortel, M.H.V. (1987) 77BS 12, 237-240. 98) Van Regenmortel, M.H.V. (1989) Immunol Today 10, 266-272.

963

99) Lowell, G.H., Smith, L.F., Seid, R.C. and Zollinger, W.D. (1988) J. Exp, Med. 167, 658-663. 100) Lowell,G.H., Ballou, W.R., Smith, L.F., Wirtz, RA., Zollinger, W.D. and Hockmeyer, W.T. (1988) Science 240, 800-802. lODPrasch. C.E. and Peppier, M.S. (1982) Infect. Immun. 37, 271-280. 102)Gotschllch, E.G., Eraser. B«A.. Nlshlmura, O., Robbins, J.B. and Liu, T.-Y. (1981) J. Biol Chem. 256, 8915-8921. 103) Wichek, M. and Bayer, E.A. (1988) Anal Biochem. 174, 1-32. 104) Wichek, M. and Bayer, E.A. (1984) Immunol Today 5, 39-43. 105) Weber, P.C., Ohlendorf, D.H.. Wendoloski. J . J . and Salenmie, F.R. (1989) Science 243, 85-88. 106)Hofmann, K., Titus. G., MontibeUer. JJL and Finn. M.F. (1982) Biochemistry 21, 978-984. 107) Shapira. M.. Gibson. M.. MuUer. G.. Amon. R and Sela. M. (1985) Proc. Natl Acad. Set USA 81. 2461-2465. 108) Audibert. F.. JoUvet. M.. Chedid. L.. Amon. R and Sela. M. (1982) Proc. Natl Acad. Set USA 79. 5042-5046. 109) Jacob. C O . . Amon. R. and Sela, M. (1986) Immunol Lett. 14. 43-48. 110)Hudecz. F.. Hilbert. A., Mezo. G. et al. (1993) Peptide Res. 6. 263-271. l l D H u d e c z . F.. Rajnavolgyi. E.. Price. M.R and Szekerke. M. (1992) in: Peptides 1992 (Schneider. C.H. and Eberle. A.N.. eds.) pp. 869-870. ESCOM Leiden. 112)Jaffe. B.M.. Newton. W.T. and McGuigan. J.E. (1970) Immunochemistry 7. 715-725. 113)Posnett. D.N.. McGrath. H. and Tarn. J.P. (1988) J. Biol Chem 263. 17191725. 114) Tarn. J.P. (1988) Proc. Natl Acad. Set USA 85. 5409-5413. 115)DriJfhout. J.W. and Bloemhoff. W. (1991) Int. J. Peptide Portein Res. 37. 2732. 116) Spetzler. J.C. and Tarn. J.P. (1995) Int. J. Peptide Protein Res. 45. 78-85. 117) Huang, W.. NardeUi. B. and Tarn, J.P. (1994) Mol Immunol 3 1 . 1191-1199. 118)Defoort. J.-P.. Nardelli. B.. Huang. W.. Ho. D.D. and Tarn. J.P. (1992) Proc. Natl Acad. Set USA 89. 3879-3883. 119) NardeUi. B.. Defoort. J.-P.. Huang. W. and Tarn. J.P. (1992) AIDS Res. Hum. Retroviruses S, 1405-1407. 120) Tarn. J.P. and Lu, Y.A. (1989) Proc. Natl Acad. Set USA 86. 9084-9088. 121)Tarn. J.P., Clavijo, P., Lu. Y.A.. Nussenzweig, V.. Nussenzweig. R. and Zavala, F. (1990) J. Exp. Med. 171. 299-306. 122)Troalen. F.. Razafindratsita. A.. Puisieux. A.. Voeltzel. T.. Bohuon. C , Bellet. D. and Bidart. J.M. (1990) Mol Immunol 27. 363-368. 123) Marguerite. M.. Bossus, M.. Mazingue. C . Wolowczuk, I.. Gras-Masse. H.. Tartar. A.. Capron. A. and Auriault. C. (1992) Mol Immunol 29. 793-800. 124)Briand. J.-P.. Barin. C . Van Regenmortel. M.H.V. and MuUer. S. (1992) J. JmnumoL MetK 156. 255-265. 125)Esposito. G.. Fogolari. F.. Viglino. P.. Cattarinussi, S.. De Magistris. M.T.. Chiappinelli. L. and Pessi, A. (1993) Eur. J. Biochem 217. 171-187. 126)Modrow. S.. H5flacher. R , Mertz, H. and Wolff. J. (1989) J. Immunol Meth. 118. 1-7.

964

127)Chong. P.. Sydor. M.. Wu. E. and Klein, M. (1992) MoL Immunol 29, 443446. 128) Fischer. P.M.. Comis. A. and Howden. M.E. (1989) J. Immunol Meth. 118. 119-123. 129)Kanda. P.. Kennedy, R.C. and Aparrow. J.T. (1991) Int. J. Peptide Protein Res. 3 8 . 385-391. 130)Goddard. P.. McMurray. J.S.. Sheppard. R.C. and Emson. P. (1988) J. Chem. Soc. Chem. Comnum. 1025-1027. 131) Hegel. M.. Pichova. D., Minarlk. P. and Sheppard. R.C. (1990) in: Peptides 1990 (Giralt, E. and Andreu, D.. eds.) pp. 837-838. ESCOM. Leiden. 132)Rapp, W.. Zhang, L.. Beck-Sickinger. A.G.. Deres. K., WiesmuUer. K.H.. Jung, G. and Bayer. E. (1990) in: Peptides 1990 (Giralt. E. and Andreu. D.. eds.) pp. 849-850. ESCOM. Leiden. 133)Butz. S.. Rawer. S.. Rapp. W. and Birsner. U. (1994) Peptide Res. 7. 20-23. 134) Gregoriadis. G. (1990) Immunol Today 11. 89-97. 135) Alving. C.R. (1989) J. Immunol Meth. 140. 1-13. 136)Ostro. M.J. (1987) Liposomes from Biophysics to Therapeutics, M. Dekker. New York. 137) Goodman-Snitkoff. G.. Eisele, L.E., Heimer, E.P.. Felix, A.M., Anderson, T.T., Fuerst. T.R. and Mannino. R.J. (1990) Vaccine 8, 257-262. 138) Goodman-Snitkoff, G., Good, M.F., Berzofsl^. J.A. and Mannino, R.J. (1991) J. Immunol 147. 410-415. 139)Watari. E., Dietzschold, B.. Szokan. G. and Heber-Katz, E. (1987) J. Exp. Med. 165. 459-470. 140) Dal Monte, P.R. and Szoka, F.C. (1989) J. Immunol 142, 1437-1443. 141) Dal Monte, P.R. and Szoka, F.C. (1989) Vaccine 7. 401-408. 142)Frisch. B.. Muller, S., Briand, J.P., Regenmortel, M.H.V. and Schuber, F. (1991) Eur. J. Immunol 21, 185-193. 143)Friede, M., Muller, S., Briand, J.P., Regenmortel, M.H.V. and Schuber, F. (1993) MoL Imnumol 30, 539-547. 145) Yasuda, T., Kanegasaki, S., Tsumita. T., Tadakuma. T., Ikewaki, N., Homma, J.Y., Inage, M., Shiba, T. and Kusumoto, S. (1984) Eur. J. Biochem 140, 245248. 146) Alving. C.R. and Richardson, E.G. (1984) Rev. Infect. Dis. 6. 493-496. 147) Lederer. E. (1980) J. Med. Chem. 23, 819-825. 148)Baschang, G. (1989) Tetrahedron 45, 6331-6360. 149) Georgiev, V.S. (1990) TIPS 11, 373-378. 150)Moroder, L., Musiol, H.-J. and Siglmuller, G. (1990) Synthesis 10, 889-892. 151)Gemeiner, M„ Leidinger, E., Miller, I. and Moroder, L. (1992) Biol Chem Hoppe-Seyler 373, 1085-1094. 152)Ellouz, F., Adam, R., Ciorbaru, R. and Lederer, E. (1974) Biochem. Biuj Res. Commun. 59, 1317-1325. 153)Matsumoto, K.. Otani, T.. Une, T.. Osada. Y., Ogawa, H. and Azuma. I. (1983) Jn/ect. J/Ti/n. 39, 1029-1040. 154) Romano, R., Musiol, H.-J.. Weyher, E., Dufresne, M. and Moroder. L. (1992) Biopolymers 32, 1545-1558.

965 155) Romano, R., Dufresne. M., Prost. M.-C. Ball. J.-P., Bayerl. T.M. and Moroder. L. (1993) Biochim Biophys. Acta 1145, 235-242. 156)Machy, P. and Leserman, L.D. (1983) Biochinh Biophys. Acta 730, 313-320. 157) Heath, T.D., Lopez, N.G. and Papahadjopoulos, D. (1985) Biochim. Biophys, Acta 820, 74-84. 158) Rude, E., Meyer-Dellus. M. and Gundelach, M.-L. (1971) Eur, J. Immunol 1, 113-123. 159) Stark. J.M., Locke, J. and Healtley. R.V. (1980) Immunology. 39, 345-352. 160)Hopp, T.P. (1984) Mat Immunol 1, 13-16. 161)Rouaix, F., Grass-Masse, H., Mazingue, C Diesis, E.. Ridel. P.R.. Estaquler. J., Capron, A., Tartar, A. and Auriault. C. (1994) Vaccine 12, 1209-1206. 162) Martinon, F., Grass-Mass, H., Boutlllon, C., Chlrat, F., Deprez, B., Gulllet, J.-G., Gomard, E., Tartar, A., Levy, J.-P. (1992) J. Immunol 149, 3416-3422. 163)Deres, K., Schlld, H., Wiesmuller, K.-H., Jung. G. and Rammensee H.G. (1989) Nature 342, 561-564. 164)Schild, H., Norda, M., Deres, K, Falk, K., RStzschke, O., WiesmuUer, K.-H., Jung, G. and Rammensee, H.-G. (1991) J, Exp, Med, 174, 1665-1668. 165)Seifert, R., Schultz, G., Richter-Freund, M., Metzger, J., Wiesmuller, K.-H., Jung, G., Bessler, W.G. and Hauschildt, S. (1990) Biochem. J. 267, 795-802. 166) Metzger, J., Jung, G., Bessler, W.G., Hoffmann. P., Streaker, M., Ueberknecht. A. and Schmidt, U. (1991) J. Med, Chem. 34, 1969-1974. 167) Wiesmuller, K.-H., Besslerr, W. and Jung, G. (1983) Hoppe-Seyler's Z. Physiol Chem, 364, 593-606. 168) Bessler, W.G., Suhr, B., Buhrlng. H.-J., Muller, D.P., Wiesmuller, K.-H., Becker, G. and Jung, G. (1985) Immunbiol 170, 239-244. 169) Jung, G., Wiesmuller, K.-H.. Becker, G., Buhring, H.-J. and Bessler, W.G. (1985) Angew, Chem. 97, 883-885. 170)Reichert. C.M., Carelli, C , JoUvet, M., Audibert, F.. Lefrancier, P. and Chedid, L. (1980) MoL Immunol 17, 357-363. 171) Carelli, C , Audibert, F., GaiUard, J. and Chedid, L. (1982) Proc, Natl Acad. Set USA 79, 5392-5395. 172) Carelli, C . Ralamboranto, L.. Audibert. F., Gaillard, J.. Briquelet. N., Dray, F., Fafeur, V.. Haour, F. and Chedid. L. (1985) Int, J. Imnumpharmacol 7, 215-224. 173)Hosmalin. A., Carelli, C , Gaillaerd, J., Lefrancier, P., Drobeco, H., Leclerc, C , Amar, D., Audibert, F. and Chedid, L. (1987) Clin, Immun. Imnvunpath, 45, 447-460. 174)Moroder, L., Gobbo, M., Becker, G. and Wunsch, E. (1989) BioL Chem, Hoppe-Seyler 370. 365-375. 175)Moroder, L.. Dufresne, M., Gohring, W., Wunsch, E., Leidinger, E. and Gemeiner, M. (1989) Biol Chem. Hoppe-Seyler 370. 1209-1214. 176)Schutze, M.P., Leclerc, C , JoUvet, M., Audibert, F. and Chedid, L. (1985) J. Immunol 135. 2319-2322. 177)Dryberg. T. and Oldstone, M.B.A. (1986) J. Exp, Med, 164. 1344-1349. 178)Schutze, M.P.. Deraud. E., Przewlocki, G. and Leclerc, C. (1989) J. Immunol 142, 2635-2640. 179) Amon, R.. Teicher, E. and Scheraga. H.A. (1974) J. Mol Biol 90, 403-407.

966 180)Niman. H.L.. Houghten. R.A.. Walker. L.E.. Relsfeld, RA.. Wilson. IJL. Hogle. J.M. and Lemer. R.A. (1983) Proc. Natl Acad. Set USA 80. 4949-4953. 181) Jacob. C O . . Amon. R and Sela. M. (1985) AfoL Immunol 22. 1333-1339. 182)Borras-Cuesta. F.. Fedon. Y. and Petlt-Camurdan. A. (1988) Eur. J. Immunol 18. 199-202. 183) Broekhuljsen. M.P.. Van Rijin. J.M.M:. Blom. A.J.M.. Pouwels. P.H.. EngerValk. B.E.. Brown. F. and Francis. M.J. (1987) J. Geru Virol 6 8 . 3137-3143. 184)Leclerc. C . Przewlocki. G.. Schutze. M.P. and Chedid. L. (1987) Eur. J. Immunol 17. 269-273. 185)BoiTas-Cuesta. F.. Petit-Camurdan. A. and Fedon. Y. (1987) Eur. J. Immunol 17. 1213-1215. 186)Francis.M.J.. Hastings. G.Z.. Syred. A.D.. McGinn. B., Brown. F. and Rowlands. D.J. (1987) Nature 330. 168-170. 187) Cox. J.H.. Ivanyi. J.. Young. D.B.. Lamb. J . R . Syred. A.D. and Francis, M.J. (1988) Eur. J. Immunol 18. 2015-2019. 188) Partidos. C . Stanley. C. and Steward. M. (1992) Eur. J. Immunol 22. 26752680. 189) Partidos. C . Stanley. C. and Steward. M. (1992) Mol Immunol 29. 651-658o 190) Good. M.F.. Maloy. W.L.. Lunde. M.N.. Morglalt. H.. Comette. J.L.. Smith. G.L.. Moss. B.. Miller. L.M. and Berzofsky. J.A. (1987) Science 235. 10591062. 191)Rouaix. F.. Gras-Masse. H.. Mazingue. C . Ridel. P.-R. Diesis. E.. Marguerite, M.. Estaquier. J.. Capron, A.. Tartar. A. and Auriault. C. (1994) Immunopharmacol 28, 137-143. 192) Good. M.F.. Pombo. D.. Lunde. M.N.. Maloy. W.L.. Halenbeck. R . Koths. K.. Miller. L.H. and Berzofsky. J.A. (1988) J. Immunol 141. 972-977. 193) Partidos. CD.. Obeid, O.E. and Steward. M.W. (1992) Immunology 77. 262266. 194)Richman. S.J. and Reese. RT. (1988) Proc. Natl Acad. Set USA 85. 16621666. 195) Gras-Masse. H.. Jolivet. M.. Drobecq. H.. Aubert. J.P.. Beachey. E.H.. Audibert. F.. Chedid. L. and Tartar. A. (1988) Mol Immunol 25. 673-678. 196)Dorow. D.S.. Shi, P.-T.. Carbone. F.R, Minasian, E., Todd, P.E.E. and Leach, S.J. (1985) Mol Immunol 22. 1255-1265. 197) Dyson, H.J., Cross, K.J., Houghten. R.A.. Wilson. I.A.. Wright. P.E. and Lemer, R.A. (1986) Nature 318. 480-483. 198) Dyson, H.J.. Ranee. M.. Houghten, RA., Wright, P.E. and Lemer. RA. (1988) J. Mol Biol 201. 201-217. 199)Kaumaya, P.T.P., Bemdt, K.D.. Heidom. D.B.. Trwhella. J.. Kezdy. F.J. and Goldberg, E. (1990) Biochemistry 29. 13-23. 200)Kaumaya. P.T.P.. Kobs-Conrad. S.. Digeorge. N.M. and Stevens. V.C. (1994) in: Peptides - Design, Synthesis and Biological Activity (Basawa. A. and Anantharamaiah. G.M.. eds.) pp. 133-164. Birkh§user. Boston. 201) Kaumaya. P.T.P.. VanBuskirk, A., Goldberg, E. and Pierce, S.K. (1992) J. Biol Chem. 267, 6338-6346. 202) Kobs-Conrad, S.. Lee, H., DiGeorge, A.M. and Kaumaya, P.T.P. (1993) J. BioL Chem. 268. 25285-25295. 203)Tuchscherer, G. and Mutter. M. (1995) J. Peptide Set 1, 3-10.

967 204) Mutter. M. and Villeumler (1989) Angew. Chem. Int. Ed. Engl 28. 535-554. 205)Tuchscherer. G.. Servls. C . Corradln. G.. Bliim. U.. Rivier. J . and Mutter. M. (1992) Protein Set 1. 1377-1386. 206)Moroder. L.. Bovermann. G. and Wunsch, E. (1988) in: Peptide Chemistry 1987 (Shiba. T. and Sakaklbara. S.. eds.) pp. 759-764. Protein Research Foundation. Osaka. 207) Bovermann. G.. Moroder. L. and Wunsch. E. (1989) in: Peptides 1988 (Jung. G. and Bayer. E.. eds.) pp. 748-750. Walter de Gruyter, Berlin. New. York. 208) Marquart. M. and Deisenhofer. J. (1982) Immunol Today 3, 160-166. 209) Burton. D.R (1985) Mol Imnumol 22. 161-206. 210)Feinstein. A.. Richardson. N.E. and Taussig. M.J. (1986) Imnumol Today 7. 169-174. 211) Deisenhofer. J. (1981) Biochemistry 20, 2361-2370. 212) Ito. W. and Arata Y. (1985) Biochemistry 24. 6467-6474. 213) Endo, S. and Arata. Y. (1985) Biochemistry 24. 1561-1568. 214) Kim. H.-H. Matsunaga. C.. Yoshlmo, A.. Kato. K, and Arata. Y. (1994) J. Mol Biol 236. 300-309. 215) Marquart. M.. Deisenhofer. J.. Huber. R. and Palm. W. (1980) J. Mol Biol 141. 369-391. 216) Wunsch. E.. Moroder. L.. G6hring-Romani. S.. Musiol. H.-J.. G6hrlng. W. and Bovermann. G. (1988) Int. J. Peptide Protein Res. 32. 368-383. 217) Zhang. R. and Snyder. G.H. (1989) J. Biol Chem. 264. 18472-18478. 218) Moroder. L.. Gemeiner. M.. GShring. W., Jaeger. E.. Thamm. P. and Wunsch, E. (1981) Biopolymers 20, 17-37. 219)Mukaijama. T. and Takahashi. K. (1968) Tetrahedron Lett. 5907-5909. 220) Helmkamp. G.K.. Cassey. H.N.. Olsen, B.A. and Pettitt. D.J. (1965) J. Org. Chem. 30. 933-935. 221) Moroder. L.. Hubener. G.. Gohring-Romani. S.. G6hring, W.. Musiol, H.-J. and Wunsch. E. (1990) Tetrahedron 46, 3305-3314. 222)Heaton. G.S.. I^don, H.N. and Schofield, J.A. (1956) J. Chem. Soc (C) 31573168. 223) Large. D.G.. Rydon. H.N. and Schofield. J A . (1961) J. Chem. Soc. (C) 17491752. 224)Yarvis, D., I^don. H.N. and Schofield, J.A. (1961) J. Chem. Soc. (C) 17521765. 225) Hardy, P.M.. Ridge, B..R^don. H.N. and dos S.P. Serrao. F.O. (1971) J. Chem. Soc. (C) 1722-1731. 226)Siedler. F.. Rudolph-Bohner, S., Doi. M.. Musiol, H.-J. and Moroder. L. (1993) Biochemistry 32, 7488-7495. 227)Siedler. F., Quarzago, D., Rudolph-Bohner, S. and Moroder, L. (1994) Biopolymers 34, 1563-1572. 228) Chou, P.Y. and Fasman, G.D. (1977) J. Mol Biol 115, 135-175. 229) Scheraga, H.A. (1980) in: Protein Folding (Jaenicke, E., ed.) pp. 261-286. Elsevier. Amsterdam. 230)Wetlaufer. D.B. (1981) Adv. Protein Chem. 34, 61-92. 231) Kim, P.S. and Baldwin, R.L. (1982) AnniL Rev. Biochem. 51, 459-489. 232) Montelione, G.T. and Scheraga, H.A. (1989) Ace. Chem. Res. 22, 70-76.

968

233) Oas. T.G. and Kim. P.S. (1988) Nature 336. 42-48. 234) Montelione. G.T.. Arnold. E.. Meinwald. Y.C.. Stlmson. E.R.. Denton. J.B.. Huang. S.-G., Clardy. J. and Scheraga. H.A. (1984) J. Am. Chent Soc. 106, 7946-7958. 235) Bergmann. L.W. and Kuehl. W.M. (1979) J. Biol Chem. 254. 5690-5694. 236)Kessler. H.. Mronga. S.. MuUer, G.. Moroder. L. and Huber. R (1991) Biopolymers 31. 1189-1204. 237) Sreerama. N. and Woody. R.W. (1994) Biochemistry 33. 10022-10025. 238)Relchlln. M. (1975) Adv. Immunol 20. 71-123. 239) Moroder. L. and Lutz. J. (1995) in: Chemistry of Natural Products (Atta-urRahman, ed.). Elsevier. Amsterdam, in press. 240) Sieber. P.. Kamber. B.. Riniker. B. and Rittel. W. (1980) Helu. ChinL Acta 63. 2358-2363. 241)Wunsch. E.. Moroder. L.. G6hrlng-Romanl, S.. Musiol. H.-J.. Gohring. W. and Scharf. R. (1991) Int, J. Peptide Protein Res, 37. 61-71. 242)Hubener. G.. Gohring, W.. Musiol. H.-J. and Moroder, L. (1992) Peptide Res. 5. 287-292. 243)Dockray. G.J.. Gregory. R.A.. Hood. L.E. and Hunkapillar. M.W. (1979) Bioorg. Chem. 8. 465-470. 244)Woody. RW. (1985) in: The Peptides (Hruby. V.. ed.) Vol. 7. pp. 15-114. Academic Press. Orlando. Fl. 245) Moroder. L.. Musiol. H.-J.. K6cher, K.. Bali. J.P., Schneider. C.H.. Guba, W.. Muller, G., Mierke. D.F. and Kessler. H. (1993) Eur. J. Biochem. 212. 325333. 246) Jackson, M. and Mantsch. H.H. (1992) Biochim. Biophys. Acta 1118, 139143. 247) Sonnichsen, F.D., van Eyk. J.E., Hodges, R.S. and Sykes, B.D. (1992) Biochemistry 31, 8790-8798. 248)Kondo, M.. Kodama, H.. Costa. T. and Shimohigashl. Y. (1986) Int. J. Peptide PorteinRes. 27. 153-159. 249) Kodama, H., Shimohigashi. Y., Costa, T. and Kondo, M. (1988) Int. J. Peptide Protein Res. 32, 41-46. 250) Stolz, M.B. and Fauchere, J.-L. (1988) Helv. Chint Acta 7 1 , 1421-1428. 251) Sakaguchi. K., Kodama, H., Matsumoto, H., Yoshida, M., Takano, Y., Kamiya, H., Waki, M. and Shimohigashi, Y. (1989) Peptide Res. 2, 345-351. 252) Stepinski, Y., Zajaczkowski, I., Kazem-Bek, D., Temeriusz, A., Lipkowski. A. and Tam, S.W, (1991) Int. J. Peptide Protein Res. 38, 588-592. 253) Cheronis, J . C , Whalley, E.T., Nguyen, K.T., Eubanks, S.R., Allen, L.G.. Duggan, M.J., Loy, S.D., Bonham, K.A. and Blodgett, J.K. (1992) J. Med. Chem. 35, 1563-1572. 254) Moroder, L. and Wunsch, E. (1987) in: Gastrin and Cholecystokintn Chemistry, Physiology and Pharmacology (Bali, J.-P. and Martinez, J., eds.) pp. 21-32, Elsevier, Amsterdam. 255) Birdsall, H.H. and Rossen, R.D. (1984) J. Immunol 133, 1257-1264. 256) Manheimer, A. and Bona, C. (1985) Eur. J. Immunol 15, 718-722. 257) Susal, C , Guo, Z., Temess, P., Padanyi, A., Petranyi, G. and Opelz, G. (1990) Transplant. Proc. 22, 1893-1894.

969 258)Temess. P.. Marx, U.. Sandilands. G.. Roelcke. D.. Welschof. M. and Opelz, G. (1993) Clin. Exp. Immunol 933. 253-258. 259)Temess. P.. Kohl, I.. Hubener, G.. Battlstuta. R . Moroder. L.. Welschof, M.. Dufter. C . Finger. M.. Hain. C . Jung. M. and Opelz. G. (1995) J. Immunol, in press. 260) Lund. J.. Winter. G., Jones, P.T.. Pound. J.D.. Tanaka. T.. Walker. M.R. Artymluk. P.J.. Arata. Y., Burton. D.R. JeflFeris. R and Woof. J.M. (1991) J. Immunol 147. 2657-2662. 261)Roosneck. E. and Lanzavecchla. A. (1991) J. Exp. Med 173. 487-489. 262) Seed. B. (1994) Chemistry and Biology (1994) 1. 125-129. 263) Wank. SA..Pisegna. J . R and de Weerth. A. (1992) Proc. Natl Acad. Set USA 89. 8691-8695. 264)Reth. M. (1992) Arum. Rev. Immunol 10. 97-101. 265) Gambler, J.C.. Pleiman. CM. and Clark, M.R (1994) AnntL Rev. Immunol 12. 457-486. 266) Ravetch. J.V. and Kinet. J.P. (1991) Annu. Rev. Immunol 9. 457-492. 267) Watts. C. and Davidson. H.W. (1988) EMBOJ. 7. 1937-1945. 268) DeFranco, A.L. (1987) Anna. Rev. Cell Biol 3, 143-178.



971

C-NMR Spectroscopy of Coumarins and their Derivatives : A Comprehensive Review

B. Mikhova and Helmut Duddeck

1.

INTRODUCTION Coumarins constitute an important class in the realm of natural products with significant biolo-

gical activity (1). Although many books and articles have appeared since 1970 containing ^^C NMR data of various classes of natural products, only a very few of them deal with coumarins derivatives, and these mainly cover the literature of the 1970s only (2,3). Since that time, however, the data of many more coumarins have been published and NMR spectroscopy has seen a revolution. Thus, we believe that it is time to update the earlier reviews. Since the eariy 1970s, ^^C NMR spectroscopy has developed into one of the most valuable tools for structure elucidation of organic compounds and natural products because the ^^c NMR spectrum is a fmgerprint of a given compound. Moreover, ^^c data of a derivative not yet reported can often be extrapolated from the chemical shifts of compounds with related structural features. Nevertheless, it is still mandatory to perform a safe ^^C signal assignment of unknown molecules in order to avoid misinterpretations which may lead to erroneous conclusions. Therefore, we present a brief overview of NMR methods (section 2.1) which can be divided into two parts: a) The classical procedures have already been summarized by us before (3); nevertheless we include some of them here for the sake of completeness, b) New one- and two-dimensional NMR experiments have been designed during the 1980s which make some of the classical methods obsolete. The data in this review have been compiled in a data base using MDL ISIS-Base. Literature has been covered until spring 1995.

972 2.

METHODS OF i^C SIGNAL ASSIGNMENTS In general, ^^C NMR spectra are recorded under proton broad-band decoupling in order to

avoid the severe signal overlap which can easily occur because of the large one-bond carbonhydrogen coupling constants ^J(C,H) = 120-250 Hz. This procedure results in a breakdown of all signal splittings due to such couphngs. Owing to the low natural abundance of the ^^C isotope (ca. 1.1%), ^-^C NMR signals appear as narrow singlets if no further NMR-active nuclei with high natural abundance (e.g. ^^F or îp) are present in the molecule. This spectral simplification, however, produces a serious drawback in signal assignments since valuable coupling information is destroyed. Thus, a variety of assignment methods have been developed some of which are introduced in this chapter. There are five main areas: experimental NMR techniques, coupling constants, solvent effects, presence of auxiliaries and derivatization.

2.1

Experimental NMR Techniques

In the 1970s these techniques consisted mainly of ^H decoupling methods, the most prominent ones of which were the ^H broad-band (BB) decoupling and the single-frequency off-resonance decoupling (SFORD) techniques. In SFORD, the decoupler frequency is positioned outside the ^H resonance range (off-resonance). Thus, all carbon-hydrogen couplings are reduced to such an extent that only the largest coupling constants, namely ^J(C,H), give rise to small residual splittings, from which the number of hydrogens adjacent to the respective carbon atoms can be read directly; singlets correspond to quaternary carbons, doublets to methine, triplets to methylene and quartets to methyl groups. Nevertheless, in these spectra, signal overlap and second-order effects sometimes prohibit unambiguous interpretations in unfavourable cases. After 1980, NMR spectroscopy saw a revolution due to the introduction of new one- and twodimensional experiments, new superconducting magnets providing magnetic fields up to 17.6 T (^Hresonance frequency: 750 MHz) and an enormous progress in computer technology. New techniques, such as the recording of J-coupled spin echoes (Attached Proton Test), and INEPT and DEPT have been developed to circumvent these difficulties (4). By executing INEPT or DEPT involving polarization transfer from ^H to ^^C, the information about the number of adjacent hydrogens is also no longer reflected in residual signal splittings as in SFORD but in signal phases and intensities; CH and CH3 signals appear as positive and those of CH2 as negative singlets (Fig. Ic). Alternatively, it is possible to suppress all signals except those of CH (Fig. lb). Therefore, a comparison of these two DEPT spectra with the ^H BB-decoupled spectrum (Fig. la) allows an unambiguous assignment of all four sorts of CHn fragments (n = 0-3).

973

9-0-^0

b •'-'^

i II*' i>A hi»M' wi*i

Studies in Natural Product Chemistry, Volume 18: Stereoselective Synthesis, Part K (Studies in Natural Products Chemistry)

Studies in Natural Product Chemistry, Volume 27: Bioactive Natural Products, Part H (Studies in Natural Products Chemistry)

Studies in Natural Products Chemistry, Volume 29: Bioactive Natural Products (Part J) (Studies in Natural Product Chemistry Series)

Studies in Natural Products Chemistry: Bioactive Natural Products Part L

Studies in Natural Products Chemistry, Volume 35

Studies in Natural Product Chemistry, Volume 15: Structure and Chemistry, Part C (Studies in Natural Products Chemistry)

Studies in Natural Products Chemistry, Volume 31: Indices Part A

Studies in Natural Products Chemistry, Bioactive Natural Products

Studies in Natural Product Chemistry, Volume 25: Bioactive Natural Products, Part F

Studies in Natural Product Chemistry, Volume 22: Bioactive Natural Products, Part C

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

Studies in Natural Products Chemistry, Volume 33: Bioactive Natural Products (Part M)

Studies in Natural Product Chemistry, Volume 20: Structure and Chemistry, Part F, (Including Cumulative Index Volumes 1-20) (Studies in Natural Products Chemistry)

Cp2TiCl in Natural Product Synthesis

Natural Product Chemistry at glance

Progress in the Chemistry of Organic Natural Products, Volume 93

Progress in the Chemistry of Organic Natural Products, Volume 93

Progress in the Chemistry of Organic Natural Products, Volume 89

Progress in the Chemistry of Organic Natural Products, Volume 92

Progress in the Chemistry of Organic Natural Products, Volume 94

Natural Product Chemistry at a Glance

Saponins (Chemistry and Pharmacology of Natural Products)

Medicinal chemistry of bioactive natural products

Medicinal Chemistry of Bioactive Natural Products

Comprehensive Natural Products II Vol.9 Modern Methods in Natural Products Chemistry

Selected Topics in the Chemistry of Natural Products

Bioactive Natural Products (Part I)

Biodiversity and Natural Product Diversity, Volume 21 (Tetrahedron Organic Chemistry)

Monosaccharides Their Chemistry and Their Roles in Natural Products

Monosaccharides: Their Chemistry and Their Roles in Natural Products

Progress in the Chemistry of Organic Natural Products v.93

Studies in Natural Product Chemistry, Volume 18: Stereoselective Synthesis, Part K (Studies in Natural Products Chemistry)

Studies in Natural Product Chemistry, Volume 27: Bioactive Natural Products, Part H (Studies in Natural Products Chemistry)

Studies in Natural Products Chemistry, Volume 29: Bioactive Natural Products (Part J) (Studies in Natural Product Chemistry Series)

Studies in Natural Products Chemistry: Bioactive Natural Products Part L

Studies in Natural Products Chemistry, Volume 35

Studies in Natural Product Chemistry, Volume 15: Structure and Chemistry, Part C (Studies in Natural Products Chemistry)

Studies in Natural Products Chemistry, Volume 31: Indices Part A

Studies in Natural Products Chemistry, Bioactive Natural Products

Studies in Natural Product Chemistry, Volume 25: Bioactive Natural Products, Part F

Studies in Natural Product Chemistry, Volume 22: Bioactive Natural Products, Part C

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

Studies in Natural Products Chemistry, Volume 33: Bioactive Natural Products (Part M)

Studies in Natural Product Chemistry, Volume 20: Structure and Chemistry, Part F, (Including Cumulative Index Volumes 1-20) (Studies in Natural Products Chemistry)

Cp2TiCl in Natural Product Synthesis

Natural Product Chemistry at glance

Progress in the Chemistry of Organic Natural Products, Volume 93

Progress in the Chemistry of Organic Natural Products, Volume 93

Progress in the Chemistry of Organic Natural Products, Volume 89

Progress in the Chemistry of Organic Natural Products, Volume 92

Progress in the Chemistry of Organic Natural Products, Volume 94

Natural Product Chemistry at a Glance

Saponins (Chemistry and Pharmacology of Natural Products)

Medicinal chemistry of bioactive natural products

Medicinal Chemistry of Bioactive Natural Products

Comprehensive Natural Products II Vol.9 Modern Methods in Natural Products Chemistry

Selected Topics in the Chemistry of Natural Products

Bioactive Natural Products (Part I)

Biodiversity and Natural Product Diversity, Volume 21 (Tetrahedron Organic Chemistry)

Monosaccharides Their Chemistry and Their Roles in Natural Products

Monosaccharides: Their Chemistry and Their Roles in Natural Products

Progress in the Chemistry of Organic Natural Products v.93

Recommend Documents