Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies on the biosynthesis, degradation and synthesis of olivacine-ellipticine type indole alkaloids Grierson, David Scott 1975

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1975_A1 G75_1.pdf [ 11.12MB ]
Metadata
JSON: 831-1.0061014.json
JSON-LD: 831-1.0061014-ld.json
RDF/XML (Pretty): 831-1.0061014-rdf.xml
RDF/JSON: 831-1.0061014-rdf.json
Turtle: 831-1.0061014-turtle.txt
N-Triples: 831-1.0061014-rdf-ntriples.txt
Original Record: 831-1.0061014-source.json
Full Text
831-1.0061014-fulltext.txt
Citation
831-1.0061014.ris

Full Text

STUDIES ON THE BIOSYNTHESIS, DEGRADATION AND SYNTHESIS OF OLIVACINE-ELLIPTICINE TYPE INDOLE ALKALOIDS BY DAVID SCOTT GRIERSON B.Sc. The University of Bri t i s h Columbia (1970) A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of CHEMISTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June, 1975 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of Brit ish Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of CAfl»MIS4/• The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1WS Date ABSTRACT Part I of this thesis describes the isolation of representatives of a class of indole alkaloids, lacking the 3-S-ethyla^ino side chain, from two plant sources Aspidospema australe, and Aspidosperna vargasii. A preliminary investigation of the biosynthesis of several of these compounds was conducted in Aspidosperca vargasii. From crude extracts of Aspicosperma australe the pyridocarbazole alkaloids olivacine (16) and guatambuine (25) were isolated. From Aspidosperra vargasii uleine (18), apparicine (19), desmethyluleine (85 ) and the pyridocarbazoles 9-methoxyolivacine (82) and guatanbuine (25) were isolated. Aromatic tritium labelled tryptophan (27) and stemoadenine (13) were shown to be incorporated into 9-methoxyolivacine (82) and tryptophan (27) was also incorporated into guatanbuine (25) in A^p^djpsjjejr^a vargasii. Neither precursor was incorporated into uleine (18). In part II a degradation scheme was developed for the isolation of the C-l methyl, C-2 methyl(N-methyl) and C-3 methylene groups of the "D" ring of the olivacine (16) and e l l i p t i c i n e (17) systems. Both e l l i p t i c i n e (17) and olivacine (16) were converted to their N-methyl tetrahydro derivatives guatam-buine (25) and N-inethyltetrahydroellipticine (26) via formation of the methiodide sa l t s of 16 and 17 followed by reduction with sodium borohydride. Compounds 25 and 26 were converted to their corresponding methiodides 86 and 95 and reacted under Hofmann reaction conditions. Olefins 88 and 97 were obtained from gua-tambuine methiodide (86) and ole f i n 102 was obtained from 95. Olefins 88 and 102 were reacted with ozone and the formaldehyde produced was isolated as the bisdimedone derivative. i i i The C-2 vinyl compound 97 was elaborated into the C-3 v i n y l compound 112 by hydrogenation of 97 to 103, formation of the nethiodide 111 and reaction of 111 with sodium hydride in dinethy1formamideo The nethiodides 86 and 95 were also ring opened to 89 and 107 by reaction with l i t h i u n aluminum hydride. These cocpounds were in turn converted to their methiodides 90 and 108 and reacted with potassium t-butoxide i n t-butanol. The trimethylaaine produced during the reactions was isolated as the tetramethyl-amaonium iodide s a l t . The efficiency of the N-nethyl group isolation was determined by degrading (N-^C methyl)-guatambuine methiodide (86) and N-nethyl-tetrahydroellipticine nethiodide (95) via the lithium aluminum hydride ring-opening sequence. Guatambuine (25) was also ring-opened to a C-3 v i n y l derivative 125 by reaction with acetic anhydride and sodium acetate. Part III was concerned with the synthesis of olivacine (16). Two approaches were developed; i n sequence A the reaction of tryptophyl brocide (207) with methylacetoacetate (205) gave 3-carboaethoxy-5-(3-indolyl)-2- pentanone (204). Cyclization of 204 led to an equal mixture of l-oethyl-2-carbomethoxycarbazole (134) and l-methyl-2-carboaethoxy-l,2,3,4-tetrahydrocarbazole (209) formed by disproportionation of the i n i t i a l l y forried 208. Dehydrogenation of the mix-ture of 134 and 209 over"Pd/C gave 134. The carbazole ester 134 was also obtained d i r e c t l y from 204 by cyclization i n the presence of chloranil as the hydrogen acceptor. Compound 134 was reduced to the alcohol 157 with lithium aluminum hydride and the alcohol 157 was oxidized to the aldehyde 152 with Jones reagent. The aldehyde 152 was converted to olivacine (16) and guataxabuine i v (25) by a known procedure. In sequence j i . , when 9-benzyltetrahydrocarbazole (217) was reacted under Vilsceier-Haack conditions l-methyl-3-forj2yl-9-benzylcarbazole (219) was forced. Compound 219 was elaborated to the aminoacetal 224 by two routes; condensation with asiinoacetaldehyde diethylacetal (171) led to the inine acetal 221 which was alkylated with methylmagnesium chloride to give 224. Alternatively 219' was alkylated to give the cx-hydroxyethyl carbazole 222 which was converted to i t s corresponding acetate 223. The acetate group was displaced by aainoace-taldehyde diethylacetal (171) to give 224. The cyclization of 22'4 to 6-benzo-olivacine (225) followed by debenzylation to olivacine (16) was not attempted, however the conditions necessary for the cyclization have been worked out for the synthesis of the closely related molecule, e l l i p t i c i n e (17). V TABLE OF CONTENTS Page TITLE PAGE. . . . . . . . . . . . i ABSTRACT • • • • • o « o « * o * « » < » » * « » » * ° » < » * ' » * » TABLE OF CONTENTS v LIST OF FIGURES ' . v i LIST OF TABLES ix ACKNOWLEDGEMENTS . • x PART I INTRODUCTION . . . . . . . . . . . . . . . 1 DISCUSSION 37 EXPERIMENTAL 54 PART II INTRODUCTION 62 DISCUSSION 68 EXPERIMENTAL 137 PART III INTRODUCTION 166 DISCUSSION . 189 EXPERIMENTAL 235 BIBLIOGRAPHY 256 v i LIST OF FIGURES Figure Page 1 Compounds of Major Pharnacological I n t e r e s t . 2 2 Some Representative Indole A l k a l o i d s 5 3 Indole A l k a l o i d s Occurring i n Genus Aspidospema 6 4 The Barger-Hahn-Robinson-Woodward Hypothesis 9 5 The Kenkert Prephenate Postulate 10 6 The Acetate Postulate 11 7 The Thomas-Kenkert Monoterpene Postulate 12 8 Rearrangement of >tonoterpene Unit i n t o the Three A l k a l o i d a l Skeletons 15 9 The Early Stages of I n d o l e ~ A l k a l o i d Biosynthesis"' as Proven by Experiment , .....17 10 Biogenesis o f the Cory-nan the Family.... 20 11 The Postulated O r i g i n s of the Strychnos family 22 12 The Kenkert Postulate f o r the Biosynthesis o f the Iboga and Aspidosperma f a m i l i e s 24 13 The A c r y l i c E s t e r Postulate f o r the Biosynthesis o f the Aspidospenr.a and Iboga Families 28 14 Tne Kenkert Postulate f o r the Biosynthesis o f O l i v a c i n e (16) E l l i p t i c i n e (17) and Uleine (18) 32 15 The D j e r a s s i Postulate f o r the Biosynthesis o f Apparicine (19) 33 16 Kim and Erickson's i n v i t r o Study of Uleine Biosynthesis 33 v i i Figure Page 17 Postulated Synthesis of Apparicine (19) via Fragmentation of the Tryptaraine Bridge 33 18 The Potier-Janot Postulate for the Biosynthesis of the Non Tryptamine Bridged Alkaloids 35 19 The N.M.R. spectrum of Guatambuine (25) 42 20 The Labelling Pattern in the Biosynthesis of the Olivacine (16) and E l l i p t i c i n e (17) Series According to the Potier-Janot Postulate..... 64 21 The Structure Elucidation of Olivacine (16) by Ondetti and Deulofeu (1961).. 65 22 The Conversion of Olivacine (16) and E l l i p t i c i n e (17) to their N-Methyltetrahydro forms. 69 23 The N.M.R. Spectrum of the Crude Reaction Mixture 24 The N.M.R. Spectrum of the Crude Reaction Mixture KOt-Bu/t-BuOH) 82 25 The N.M.R. Spectrum of l-Methyl-2-vinyl-3-(a-dimethylaminoethyl)carbazole (97) 84 26 The N.M.R. Spectrum of l-Methyl-2-(B-dimethylamino-ethyl)-3-vinylcarbazole (88) 88 27 A Comparison of the UV. Spectra for the Olefins 88 and 97 with the Normal Carbazole Spectrum of 98 91 28 Mass Spectrum of Carbazole Compound 103 112 29 Mass Spectrum of Carbazole Compound 88 112 30 Mass Spectrum of Carbazole Compound 109 112 31 Plausible Mass Spectral Fragmentation Pattern for Compound 103 114 v i i i Figure Page 32 Plausible Mass Spectral Fragmentation Pattern for Compound 88 115 33 Plausible Mass Spectral Fragmentation Pattern for Compound 109 115 34 Approaches to the Isolation of the C-l Methyl Group Involving a C-l* Oxygen Functionality 118 35 Acetate Substitution Approach Studied on Model Compounds 122 36 N.M.R. Spectrum of 1,4-Dimethy1-2-(3-(N,N-methyacetylamino) ethyl)-3-acetoxyEethylcarbazole (121) 123 37 N.M.R. Spectrum of l-Methyl-2-(3- (N,N-methyl-acetylamino)ethyl-3-vinylcarbazole (125) 129 38 The UV. Spectrum of the Iminium Cation (142), Enamine (126), and Reaction Product of 142 Treated with Dinethylsulphate in Base 134 Methylation of the Anhydro Base 136 40 Synthesis of Olivacine (16) and Guatambuine (25) by Schrcutz and Wittwer (I960) 168 41 Scheme A Synthesis of Olivacine (16) According to Wenkert and Dave 171 41 Scheme B 172 42 Synthesis of Olivacine (16) by Mosher et a l . (1966)....174 43 The Kametani Benzocyclobutene Analog Approach (1975) 176 44 Woodward (1959) Synthesis of E l l i p t i c i n e (17) 177 45 Synthesis of E l l i p t i c i n e (17) by Cranwell and Saxton (1962) 178 46 Synthesis of E l l i p t i c i n e (17) by Govindachari et a l . (1963) Scheme A - C 181 j F i fiure Page 47 Kilrains tor and Sa l i s b u r y (1972) Synthesis o f E l l i p t i c i n e (17) ' 1 8 5 4S The Synthesis o f E l l i p t i c i n e (17) by Le G o f f i c , Goyettc, and Ahond (1973) 1 8 7 49 An Ir.proved Synthesis of O l i v a c i n e (16) and Guatanbuir.e (25), Sequence A 192 50 Oxalyl Chloride Route to Tryptophol (206) 193 51 N.M.R. Spectrum of 5-Carboir.ethoxy-5-(3-indoly 1)-2-pentanone (204) ' 197 52 P l a u s i b l e Mass S p e c t r a l Fragmentation Pattern f o r the A l k y l a t i o n Product 204 ' 199 53 N.M.R. Spectrum of the C y c l i z a t i o n Reaction Mixture, 202 54 P l a u s i b l e Mass S p e c t r a l Fragmentation Modes for Components 134 and 209 o f the C y c l i z a t i o n Reaction ?-lixture 204 55. The Synthesis of O l i v a c i n e (16) f r o a Tetrahydro-carbazole (216) 219 56. Proposed Mechanism f o r the V i l s s e i e r - R a a c k F o r o y l a t i o n of 9-Benzyltetrahydrocarbazole (216) 224 57. N.M.R. Spectrum, of l-Methyl-3-(2,p-diethoxyethyl iiainomethyl) carbazole (221) 229 58. UV. Spectrum of Compound 221 i n Water and i n D i l u t e Hydrochloric Acid 230 X LIST OF TABLES TABLE PAGE I Column Chromatography Results on Extracts from A. vargasii. 46 I I Results of Incorporation of Tryptophan (27) and Ste:madenine (13) into A. vargasii 51 I I I Characteristic N.M.R. Chemical Shifts for the Degradation Products of Guatanbuine (25) 86 T T T H O Q . , l ^ r A . . - . " J-l-^^n r- C \ 1 r ^ V 1 <- f l O i l x i ACKNOWLEDGEMENTS I would l i k e to express my gratitude to Dr. James P. Kutney for his guidance, optimism and the opportunity to learn through the course of th i s research. I would also l i k e to thank Dr. George Fu l l e r and Mr. Harald Hanssen for t he i r co l laborat ion with me in th i s research, and also the other members of the group, past and present, fo r helpful discussions and suggestions. Thanks are due to Dr. P h i l i p J . Sal isbury for his expert ise and wi l l ingness to help me in the propogating and handling of the p lants. Special thanks are due to my fami ly, fr iends and typ i s t s fo r t he i r help and preserverance during the preparation of th i s manuscript. Receipt of a National Research Council Postgraduate Scholarship i s g ra te fu l l y acknowledged. INTRODUCTION PART 1 The plant kingdom plays a paramount role in both the existence and maintenance of a l l animal l i f e on this planet. The existence of the present life-sustaining oxygen-rich atmosphere is considered in large part to be a consequence of the photosynthetic production of oxygen by primitive plant 1,2,3 forms during the early stages of the development of the earth. This same process, harnessing the power of the sun has sustained l i f e by enabling plants to produce consumable energy-containing compounds v i t a l to the function of the animal organism. A considerable number of these consumable plants have been found over the ages to be beneficial to man in a medicinal or maintenance manner, aiding in his fight against the diseases which threaten the longevity of his l i f e . The use of such herbal remedies to cure ailments is generally considered to precede even the origins of agriculture. Accounts of their reported use are found in the inscriptions and writings of the ancient Egyptian, Babylonian and Chinese c i v i l i z a t i o n s . Rivalled, perhaps only by the Ayurvedic medicine in India, Chinese herbal medicine is the oldest continuous surviving tradition in their culture. The SHANG-HAN LUN ("Treatise on Fevers") written by Chang Chung-Ching (ca A.D. 195) is a classic of Chinese c l i n i c a l medicine right down 4 to the present day. In modern times the s c i e n t i f i c world has come to view folklore and t r i b a l medicines with considerable interest. Ignoring the shrouds of mystic and - 2 -(3) (4) Figure 1. Compounds of Major Pharmacological Interest. considering the possible curative effects of these herbal extracts, i t has been realized that many possess pharmacological activity. Such investigations have often led to the isolation of the biologically active component. This work has been spurred on by such medically important discoveries as: Reserpine (1) for the treatment of mental disorders, isolated from the Indian snake root plant Rauwolfia serpentina; morphine (2) an analgesic from the opium poppy Papaver somniferum; strychnine (3) the cardiac principle and stimulant from Strychnos nux vomica; and quinine (4) an antimalarial from Cinchona bark (Figure 1) . These land mark discoveries have provided the incentive for the presently active and intense worldwide investigation of the plant kingdom for compounds of possible medicinal value. The subsequent efforts towards the synthesis of these - 3 -molecules has been the foundation of the modern drug industry. Although a tremendous number of synthetically prepared drugs are in use today, i t is estimated that approximately 50% of the medicines prescribed are s t i l l 5,6 derived from natural plant sources. It was recognized early in the investigations that some indole alkaloids possess beneficial medicinal properties. They have been found to occur widely among flowering plants particularly of the Apocynaceae, Loganiaceae, 7 and Rubiaceae plant families. These three families stand close together in 8 the phylogenetic charts of the taxonomists. The family Apocynaceae is particularly rich in alkaloids, especially the genera Rauwolfia, Vinca (Catharanthus), Alstonia and Aspidosperma. The great majority of complex indole alkaloids can be visualized to con-s i s t of a tryptamine unit and a Cg-C^n unit which is monoterpene derived. 9 Approximately 1000 indole alkaloids are known to date. These compounds in general offering minor variations in functionality, oxidation state and stereochemistry lend themselves to a formal division into four main groups according to their skeletal features. These four groups, the corynanthe (5), strychnos (5), aspidosperma (6) and iboga (7) are represented by three distinct skeletal arrangements of the carbon atoms in the CQ-C^Q unit as i l l u trated below. (5) (6) ( 7 ) i - 4 -Examples of these groups are the ring closed forms: geissoschizine (8), preakuammicine (9), vindoline (10) and catharanthine (11) and the corresponding ring opened forms: picraphylline (12), stemmadenine (13), vincadine (14) and 18-carbomethoxycleavamine (15) (Figure 2). As a result of an intensive search of the Apocynaceae family for both biological and taxonomical purposes, i t has been found that many plants of the genera Aspidosperma and Ochrosia contain complex alkaloids possessing novel 7,9,10 structures. These compounds are exemplified by the structures of olivacine (16), e l l i p t i c i n e (17), uleine (18), apparicine (19) and vallesamine (20). They are novel in that they do not possess the 3-3-ethylamino side chain (tryptamine bridge) indigenous to the majority of naturally occurring indole alkaloids. It can be seen that uleine (18), apparicine (19), and vallesamine (20), have only a single carbon atom separating the 3-position of (Nb) the indole nucleus from the basic nitrogen, whereas, olivacine (16), and e l l i p -11* ticine (17), being substituted 6H-(4,3-b)-pyrido carbazoles exhibit a three carbon bridge (Figure 3 ) . In terms of formality, these compounds can be considered to possess the corynanthe-strychnos skeleton 5, however by virtue of their almost exclusive presence in Aspidosperma plants, they are considered to be aspidosperma alka-loids. They are generally found to co-occur in many of these plant systems and are not, therefore, simply isolated examples of biological peculiarities. They are also found to co-occur with a small number of normal tryptamine bridged alkaloidal systems such as N-acetylaspidospermidine (21), aspidocarpine (22), N-acetyl-ll-hydroxyaspidospermatidine (23), and ll-methoxy-4,19-dihydrocondylo-10 carpine (24) (Figure 3 ) . V * Two numbering systems have been used to number the pyridocarbazole skeleton. The system used to label this system and the Uleine (18) and apparicine (19) skeletons in this thesis was that accepted by Chem. Abstracts (Ref. 11). - 5 -Figure 2. Some Representative Indole Alkaloids. - 6 -CHOOC CH^DH (20) COOCH (24) (23) Figure 3. Indole Alkaloids Occurring in Genus Aspidosperma. - 7 -In the latter two cases i t is tempting to envisage the build-up of the 114 uleine skeleton by extension of the two carbon tryptamine bridge. Both olivacine (16) and e l l i p t i c i n e (17) have received considerable interest in recent years as a result of their antitumor activity, however our interest in these compounds relates to their biosynthetic build-up in the plant system. At the i n i t i a t i o n of the present work, the occurrence of these compounds posed the important questions: What is the mechanistic path-way of their in vivo synthesis and is there some connection between their occurrence and that of the normal tryptamine bridged alkaloids found in the plant. In other words, are these alkaloids closely related in terms of the presently accepted indole alkaloid biosynthetic scheme? The answers to these questions, i t was hoped, would be found through a study of the biosyntheses of these compounds in Aspidosperma plants. A considerable number of reviews have been published concerning different aspects of indole alkaloid biosynthesis. A detailed discussion of the present state of the biosynthesis and occurrence of this class of compounds is contained 12 13 in the recent reviews by Schmid and Cordell . As a consequence, therefore, only a brief overview of the biosynthesis of indole alkaloids w i l l be discussed in this presentation with emphasis on the present knowledge concerning the biosynthesis of olivacine (16), e l l i p t i c i n e (17), uleine (18), apparicine (19), and the N-methyltetrahydro derivatives of the pyrido carbazoles, guatambuine (21) and N-methyltetrahydroellipticine (22). (22) - 8 -There is general agreement that the tryptamine segment of the indole alkaloids is derived from tryptophan (27), in accord with the postulate of 14 Pictet in 1906. With the development of tracer techniques in the last 15-18 decade, this hypothesis has received ample j u s t i f i c a t i o n . Labelled tryptophan (27) has been found to be successfully incorporated into a large variety of indole alkaloids. Tryptamine, (28) however, the other logical (27) (28) 18-20 precursor has been incorporated with mixed success. The exact nature of this result is d i f f i c u l t to interpret, i t could suggest that the decar-boxylation i s delayed in some cases, or that tryptamine (28) is not trans-ported to the site of biosynthesis as is the amino acid. The biogenetic origin of the remaining Cg-C^Q unit, in contrast to the source of the 8-aminoethylindole moeity, was the subject of much controversy. The earliest speculations as to the origin of this non-tryptophan unit were 20 based upon the structural similarities of indole alkaloids. Prior to the f i r s t tracer experiments concerned with the origin of this 21-2 unit, there were three main postulates in existence. The original Barger-Hahn 23 24-6 postulate had been elaborated by Robinson and Woodward to the point where i t could accommodate the formation of alkaloids of the strychnos and corynanthe, as well as the original yohimbe, alkaloid skeletons as shown in figure 4. The combined hypothesis involved the condensation of 3,4-dihydroxy-phenylacetaldehyde (29) derived from tyrosine with tryptamine (28) either at the 3-position of the indole system to y i e l d strychnine (3), or at the 2-position - 9 -(31) Figure 4 . Hie Barger - Hahn - Robinson - Woodward Hypothesis. - 10 -(34) Figure 5. The Wenkert Prephenate Postulate. - 11 -Figure 7. The Thomas-Wenkert Monoterpene Postulate. - 12 -to yield 30 which contains an aromatic ring E. Fission of the aromatic ring E between the two hydroxyl groups and subsequent combination with appropriate units leads to yohimbine ( 3 1 ) . A number of deficiencies were immediately evident in these specula-27 tions and in 1959 Wenkert and Bringi proposed an elegant alternative. They i n i t i a l l y proposed that a hydrated prephenic acid was the intermediate, 28 but later Wenkert modified this hypothesis so that prephenic acid (32) i t s e l f was the direct progenitor of the indole alkaloids. The latter rearranges according to the scheme shown in figure 5 to afford a crucial intermediate, the seco-prephenate-formaldehyde (SPF) unit ( 3 3 ) which can be elaborated into yohimbine (27) and corynantheine ( 3 4 ) . This ingenious alternative scheme uses the a l i c y c l i c precursors of phenylalanine directly to account for the oxidation state of the E ring. A key step i n this hypothesis, the 1,2-migration of the pyruvate side chain of prephenic acid with retention of configuration, explains the absolute configuration at C-15 29 in corynanthe-strychnos alkaloids. The carbomethoxy group i s an integral part of the hydroaromatic progenitor instead of being attached whenever necessary. 30 31-33 S c h l i t t l e r and Taylor in 1960 and Leete in 1961 postulated that the non-tryptophan portion of the indole alkaloids was derived via the ace-tate pathway. The suggestion u t i l i z e d a six carbon chain derived from three acetate units which condensed with malonic acid and a one carbon unit (biologi cally equivalent to formaldehyde) yielding the desired C^Q unit (Figure 6 ) . At this time the structure elucidation of a number of cyclopentane mono-34-37 terpenic glucosides (iridoids) had been achieved, exemplified by verbena-l i n ( 3 5 ) , genipin ( 3 6 ) , aucubin ( 3 7 ) , and asperuloside ( 3 8 ) . - 13 -(39) Realizing that these glucosides, and in particular the seco-cyclopentane unit of swertiamarin (39) had the corynanthe-strychnoslike skeleton 5 as well as 29 the same stereochemistry in the appropriate position, as C-15 in the alkaloids , and that the carbomethoxy function (or derivative thereof) also appeared at the 28 corresponding position, led Wenkert and simultaneously and independently 38 Thomas to propose a monoterpenoid based hypothesis for the origin of the C9"10 u n i t (Pig" 1" 6 7) • The key intermediates in their pathway are the cyclopentanoid unit 40 having the corynanthe skeleton 5 and i t s cyclic form 41 analgous to swertiamarin (39). The unit 41 i s derived intact without the need for an additional carbon from formaldehyde or glycine as required by the earlier hypotheses. The cyclopentanoid monoterpenes genipin (36) and aucubin (37) have a related biosynthetic pathway and have been shown to co-occur with the alkaloids 38 in the genus Strychnos . - 14 -The i n i t i a l biosynthetic experiments using radioactive precursors 39,40 disproved al1 the hypotheses concerning the genesis of the Cg-Cjg unit. However, further experiments by Battersby and co-workers with (2-14rj)-mevalonate afforded low incorporations of activity into the ubiquitous 40,43 non-tryptophan unit. These results were rapidly confirmed by two other 41,42 research groups working with the plant species Vinca rosea and Vinca major. The Thomas-Wenkert monoterpene postulate consequently became widely accepted. More experiments by these same workers u t i l i z i n g s p e c i f i c a l l y labelled mevalonates coupled with degradative data demonstrated the intact incorporation of this unit into the representative alkaloids catharanthine (11), vindoline (10), 43-6 ajmalicine (44), serpentine (42) and perivine (43). Subsequently i t was 44,46-52 44,50 shown that geraniol pyrophosphate (45) and nerol (46) could also (42) (43) serve as precursors of these alkaloids (Figure 9). In addition, deuterium labelled mevalonolactone fed to V. rosea produced alkaloids whose mass spectral 47-8 fragmentation patterns substantiated the radioactive label findings. From these experiments a pattern emerged for the rearrangement of the mono-terpene unit into the three skeletal units 5, 6, and 7 for each of the alkaloidal - 15 -(11) (10) Figure 8. Rearrangement of Monoterpene Unit into the Three Alkaloid Skeletons families, which was consistent with the position of the radiolabel i n the various isolated alkaloids, (Figure 8). It appeared as though the corynanthe skeleton was formed f i r s t during the biosynthesis and i t subsequently rearranged by the manner shown to the aspidosperma (6) and iboga (7) skeletons. It was not known however, how this process occurred in the plant or whether there were terpenoid glycosides, as yet not isolated, having skeleta of types 6 and 7 which independently reacted with tryptamine (28). With the monoterpene hypothesis clearly substantiated, an elegant series of experiments by Battersby and co-workers determined the exact nature of the sought after cyclopentanoid intermediate. The iridoid compounds genipin (36) 28 and verbenalin (35) and others proposed by Wenkert were found not to be incor-50,53 porated into any of the alkaloids studied. However a very similar monoter-pene loganin (49) exhibiting essentially the correct oxidation level and stereo-- 16 -chemistry anticipated for the non-tryptophan precursor was found to be specifically incorporated into the alkaloids catharantl.ine (11), vindoline 53-5 (10), ajmalicine (44), serpentine (42) and perivine (43) in V. rosea , 55 as well as into the alkaloids in Rauwolfia serpentina , and Cephaelis 64 ipecacuanha. Furthermore, through isotopic dilution studies i t was shown 53-4,57 that loganin (49) co-occurred with the indole alkaloids in V.rosea 57 and Strychnos nux vomica. It has since been shown that loganin (49) occurs 13 in many alkaloid containing plants. The isoprenoid origin of loganin (49) was subsequently demonstrated by feeding specifically labelled forms of mevalonate, geraniol, and nerol to 53-4,56-8 54-5,59-61 57-8,62 62-3 V.rosea , NL_ t r i f o l i a t a , nux vomica , and carliensis. Battersby demonstrated also the incorporation in rosea of 10-hydroxy-geranial and 10-hydroxynerol (47) into both loganin (49) and the aforementioned 65-6 alkaloids (Figure 9). Randomization of the label at the positions marked 2,6 indicated that oxidation at both of these carbons i s a necessary part of the sequence to deoxyloganin (48). Deoxyloganin (48) was found to be a constituent of rosea and nux vomica and was shown to be a precursor of loganin (49) and the indole constituents of the 67-8 former plant system. The cleaved monoterpene derivative of loganin (49), secologanin (50) was isolated from rosea and found once more to be specifically incorporated into the alkaloids in the plant. Loganin (49) was shown also, to be a precursor of this 69 compound. 28 An interesting observation i s that the central intermediate of Wenkert's prephenic acid hypothesis, the SPF unit (33) is almost identical in structure, stereochemistry and oxidation level to the c r i t i c a l monoterpene intermediate - 17 -Figure 9. The Early Stages of Indole Alkaloid Biosynthesis. - 18 -secologanin (50), although i t was unsupported by experimental evidence. With the origin of the Cg-CjQ unit firmly supported by experiment, efforts were directed towards determining how secologanin (50) was u t i l i z e d by the plant system in the biosynthesis of indole alkaloids. Simultaneously and independently, vincoside (51) and isovincoside (52) 70 71 were isolated from V. rosea , and strictosidine (52) from Rhazya s t r i c t a . 72-3 X-ray analysis has proven the relative configurations about C-3 to be as shown. In addition i t has been shown that vincoside (51) is s p e c i f i c a l l y 70,74-5 incorporated by V. rosea plants into a l l three types of indole alkaloids, 70 and that i t is i t s e l f derived from tryptophan (27) and loganin (49) (Figure 9). 74 Isovincoside (52) was not incorporated into any of the alkaloids studied (SI) (52) which was surprising since i t rather than vincoside (51) exhibits the configur ation at C-3 found in the corynanthe alkaloids. To date i t has been demonstrated that the hydrogen at C-3 is epimerized on incorporation into the alkaloids and that through incorporation of (5-%) loganin (corresponding to C-^ H) into the alkaloids that retention of tritium 76 occurs in the requisite epimerization . Much remains to be done however to - 19 -determine the fate of the vincoside molecules during biosynthesis. These results in rosea at least strongly suggest that a crucial inter-mediate, vincoside (51), is formed by convergent pathways involving separate biosystheses of tryptophan (27) and secologanin (50), and that this interme-diate then undergoes the appropriate rearrangements to the various families of complex indole alkaloids. Vincoside (51), then, serves as a convenient dividing line, i t s biosynthesis may be thought of as the early stages of biosynthesis, while the subsequent rearrangements make up the latter stages of the biosynthesis. The bioconversion of vincoside (51) into the corynanthe alkaloids entails an unexceptional enzymic hydrolysis of the glucosidic residue followed by reductive condensation of the nascent aldehyde (53) as shown in figure 10. Corynantheine aldehyde (54), corynantheine (34) and geissochizine (8) are the immediate products of the cyclization while ajmalicine (44), an abundant cory-nanthe compound in V. rosea is subsequently reached by cyclization of 8. The former three compounds have been detected to be actively involved in the bio-Jr synthesis through sequential feeding and deuterium labelling experiments in 77-79 V. rosea , thus lending support to the proposed transformations. The strychnos alkaloids d i f f e r from the corynanthe formally only in the position of attachment of the same C^ Q unit 5 to tryptamine. Two pathways have 80 81 been proposed by Wenkert and Scott respectively depicting the rearrange-ments involved in transforming the corynanthe into the strychnos alkaloids as il l u s t r a t e d in figure 11 for the major strychnos compounds akuammicine (59) and stemmadenine (13). geissochizine (8) containing a reactive 3-aldehydo ester - 20 -Figure 10. The Biogenesis of the Corynanthe Family. - 21 -function is the central intermediate in both pathways. Credence but not precedence is given to the Wenkert scheme (A) by the presence of the formylstrictamine 55 along with pleiocarpamine (56) in 82,13 Rhazya s t r i c t a . (55) (56) By the same token Scott's postulate (B) is supported by the isolation of the 6-hydroxyindolenine 57, geissochizine oxindole (58) and preakuammicine 79 (9) from sequential feedings in V. rosea and the recently reported forma-83-4 tion of oxindole alkaloids in Mitragyna parv i f o l i a . Also, the incorporation 79 of geissochizine oxindole (58) into akuammicine (59) lends credence to the postulate. Much remains to be done however, to determine whether one or both path-ways are important in the plant, and whether or not the pathway taken depends upon the individual plant. Throughout the investigations of indole alkaloid biosynthesis a major concern has been to determine the manner in which the aspidosperma and iboga families are derived through rearrangement of the monoterpene portion of the corynanthe skeleton 5 as depicted in figure 8. Figure 11. The Postulated Origins of the Strychnos Family. - 23 -Through the large body of incorporation experiments dealing with the early stages of biosynthesis, vincoside (51) has been shown to be a precursor of both the aspidosperma and iboga alkaloids. Subsequent incorporations of 85,79 77-79 labelled forms of geissochizine (8) corynantheine aldehyde (54) and 77 particularly stemmadenine (13) into vindoline (11) and catharanthine (10) has demonstrated that the corynanthe-strychnos alkaloids are also precursors 77,86 of the two families. Sequential feeding experiments by Scott in V. rosea seedings further confirms that the strychnos alkaloids are the immediate pre-cursors of the aspidosperma and iboga and that the postulated sequence: corynanthe strychnos -»• aspidospermaiboga i s correct. 87 The isolation and incorporation of stemmadenine (13) in rosea was of particular significance because i t was recognized that i t occurs furthest along the strychnos pathway towards the aspidosperma and iboga bases. 28 Wenkert i n i t i a l l y introduced the concept that a seco C3-C7 intermediate derived from the strychnos skeleton could act as a pivotal precursor to both the aspidosperma and iboga alkaloids. The Wenkert postulate portrayed i n figure 12 suggests that via intramolecular Michael and Mannich condensations the seco C3-C7 piperidiene 62 obtained from a 1,5-dicarbonyl stemmadenine-like compound 61 (analogous to the iminium intermediate 60) is converted into the nine membered ring systems 63 and 64. Subsequent transannular cyclization of these components leads to the six membered ring bases 65 and 66 from which the aspidosperma and iboga alkaloids are derived. 88,13 A similar intermediate 67 was proposed by Levy to account for the biosynthesis of the quebrachamine (68) as well as the aspidosperma and iboga bases. Figure 12. The Wenkert Postulate for the Biosynthesis of the Iboga and Aspidosperma Families. - 25 -CH|XOC CHOH (67) Fundamental to Wenkert's postulate was the transannular cyclization step for conversion of the tetracyclic nine membered ring systems into the pentacyclic ring systems of the aspidosperma and iboga alkaloids. In an effort to test this transformation, i n i t i a l experiments determined i t to be 89-93 a facile i n vitro process as exemplified by the conversion of quebrach-amine (68) and 16-carbomethoxycleavamine (15) (via their N(b) iminium salts) to their pentacyclic analogs aspidospermidine (69) and catharanthine (11). H H (68) (69) However, a l l subsequent labelling experiments to test the paral l e l in vivo conversion to either the pentacyclic aspidosperma or iboga alkaloids 94-5 failed, no incorporation was found into any of the alkaloids studied. - 26 -These results strongly suggested that the latt e r portion of Wenkert*s postulate was incorrect and that i f a pivotal intermediate such as the seco C3-C7 piperidiene is involved i n the biosynthesis of the aspidosperma and iboga bases, then a rational whereby i t could account for the natural tetra-cyclic and pentacyclic structures would have to be determined. Support was given the intermediacy of a pivotal seco C3-C7 precursor by the observation that tabersonine (70) was e f f i c i e n t l y incorporated into both the aspidosperma alkaloid vindoline (10) and the iboga alkaloid catharanthine 94 77 (11) i n V. rosea plants and seedlings. This interconversion hot only involves a considerable rearrangement of the aspidosperma skeleton of taber-soninine (70) including a reversal of the transannular cyclization, but i t further demonstrates the sequence corynanthe-strychnos -*• aspidosperma -*• iboga. C O O C H 3 (70) It i s pertinent to point out that the reverse process, the conversion of 77 catharanthine (11) to tabersonine (70)» was found not to occur suggesting that equilibration with a pivotal intermediate li e s only to the side of the aspido-sperma representative tabersonine (70). Observations by Scott based upon the in_ v i t r o interconversions of selective - 27 -alkaloids in hot acetic acid led to the proposal that the acrylic ester 72 available by requisite rearrangement of stemmadenine (13) may be involved as the sought after pivotal intermediate in the biosynthesis of the aspido-96-7 sperma and iboga bases. Through the transformations outlined in figure 13 the tetracyclic and pentacyclic ring systems can be generated without having to invoke the transannular cyclization process. Evidence for the occurrence of a seco intermediate such as the acrylic ester 72 was obtained by the isolation of the dimeric indole compounds, the 98-9 secamines, from Rhazya species. Of particular significance was the pre-100-1 sence of the monomeric secodines 73, 74 and 75. (73) (74) (75) Due to the inherent i n s t a b i l i t y of the dihydropyridinium system in the acrylic ester 72 since named dehydrosecodine, the synthesis of labelled forms of this compound for the testing of the Scott proposal has not to date been possible. In view of this, the more stable reduced form of the acrylic ester 102-3 72, secodine (76) and the hydroxy ester 77 both available by synthesis were considered as useful intermediates in the biosynthetic investigations. - 28 -COOCH 3 COOCH 3 (77) (76) Figure 13. The Acrylic Ester Postulate for the B'iosyntheses of the Aspidosperma and Iboga Families. - 2 9 -This view received support when Battersby proposed, on the basis of isotopic dilution experiments that 16,17-dihydrosecodin-17-ol (77) occurs naturally 104 in rosea and R. orientalis (figure 13, bottom). 105-8 Administration of 16,17-dihydrosecodin-17-ol-(Ar-^H) (77) into V. rosea , 106,109 106-8,110 V. minor, and Aspidosperma pyricollum resulted in each case in deterioration of the plants with no detectable incorporation into the 103,105-6,109 alkaloids studied. Subsequent experiments with (Ar-^H)-Secodine (76) have been more successful, however the results must remain tentative as only very low but constant incorporations of this precursor have been obtained into Vindoline (10) and catharanthine ( 11) in rosea, and the alkaloids of minor. Further studies using various forms of doubly labelled secodine (76) coupled with degradation of the alkaloids isolated lent stronger support to the conviction that the secodine skeleton is incorporated intact into the aspido-105-112 sperma and iboga alkaloids. It i s evident from the secodine results that the inherent d i f f i c u l t i e s of low incorporation and the necessity of feeding a precursor of the wrong oxida-tion state w i l l have to be overcome before the acrylic ester, dehydrosecodine (72) can be firmly accepted as the pivotal precursor of the aspidosperma and iboga alkaloids. The abnormally bridged indole alkaloids uleine (18), apparicine (19), olivacine (16) and e l l i p t i c i n e (17) (as well as the reduced froms of the latter) are presently believed to be derived from late stage intermediates i n the biosynthetic pathway. Very l i t t l e is presently known regarding their bio-synthesis and as a consequence a number of different postulates have been put - 30 -forward to rationalize their presence in the Aspidosperma and Ochrosia plant systems. The anomolous structures of these compounds i n i t i a l l y led investigators to believe however, that by contrast to the normal 3-ethylamine bridged com-pounds they were not tryptophan derived. In conjunction with the proposal of the SPF unit 33 as the progenitor of the monoterpene portion of the alka-28 loids, Wenkert proposed that the biosynthetic pathway to uleine (18), olivacine (16) and e l l i p t i c i n e (17) involved reaction of the SPF unit 33 with glycosylideneanthranilic acid (78) a precursor of tryptophan (27). Condensa-tion of the requisite carbonyl groups of the SPF unit with a nitrogen source (such as methylamine) and reaction at the C-3 of indole with loss of the glycosyl function leads to the alkaloids (Figure 14). 113 With the structure elucidation of apparicine (19) in 1965, Djerassi proposed that Wenkert's intermediate 79 could serve as a precursor in i t s biosynthesis. Isomerization to the exocyclic iminium species 80 followed by cyclization with the indole nucleus as shown in figure 15 would y i e l d apparicine (19). 28 Wenkert's postulate invokes a separate biosynthetic pathway to these alkaloids not involving tryptophan (27) so as to extrude the two carbon chain, and as such i t is i n discord with the opinion that these indole alkaloids are tryptophan (27) derived via the pathway elucidated for the form alkaloidal families. In order to test the basic assumption behind this proposal, tryptophan -(Ar-3H) (27) was administered to A.pyricollum, a plant system reported to - 31 -9 contain apparicine (19) and uleine (19) . Significant incorporation was 110 obtained into apparicine (19) which indicates i t to be tryptophan derived, 113 contrary to Djerassi's proposal. Insufficient quantities of uleine were isolated however to permit a determination of any incorporation into i t . Despite subsequent attempts 119,120 to radiolabel this molecule in australe, no biosynthetic data i s presently available for i t or for the pyridocarbazoles olivacine (16), e l l i p t i c i n e (17), N-methyltetrehydroellipticine (26) and guatambuine (25). With tryptophan (27) as the precursor of the indole portion of apparicine (19), i t i s necessary to determine the mechanism whereby one or both of the carbons of the B-ethylamine side chain are extruded. Some insight was gained when (Ar-^H) and either C-2 or C-3-14c doubly labelled tryptophan (27) was fed to A. pyricollum. It was found that C-3 of tryptophan (27) is incorporated into apparicine (19) with retention of the 3H/14C ratio whereas 108 over 97% of the label at C-2 was shown to be lost. Extrapolating the knowledge that tryptophan is the likely precursor of 114 apparicine (19) to include uleine (18), Kim and Erickson studied a 115 mechanism proposed by Joule ejt al whereby the two carbon bridge of uleine (18) is extruded. Their experimental model for uleine (18) biosynthesis was based upon the co-occurrence and close similarity between uleine (18) and alkaloids of the condylocarpine type such as 24. They attempted without success to accomplish the in vitro expulsion of the two carbon unit from the oxidation produce of a condylocarpine system 81 (figure 16). This mechanism i s probably of l i t t l e biosynthetic significance however as i t i s specific to uleine (18) only and cannot readily be adapted to account for i Figure 14. The Wenkert Postulate for the Biosyntheses of Olivacine (16), E l l i p t i c i n e (17) and Uleine (18). - 33 -OHC COOCH. OHC COOCH 3 Figure 15. The Djerassi Postulate for the Biosynthesis of Apparicine (19) Figure 16. Kim and Erickson's in vitro Study of Uleine Biosynthesis. (13) -~ ,H / A N (19) Figure 17. Postulated Synthesis of Apparicine (19) via Fragmentation of the Tryptamine Bridge. - 34 -the lack of the 3-ethylamine bridge in the closely related apparicine (19) structure. Stemmadenine (13) known to co-occur with apparicine (19) and uleine (18) 116 in A. pyrico1lum was found to be e f f i c i e n t l y incorporated into apparicine 108 (19). This important result places apparicine (19) and very likely uleine (18) and the pyridocarbazole alkaloids of the olivacine (16) and e l l i p t i c i n e (17) at a late stage in the biosynthetic scheme. 108 The subsequent low incorporation of vallesamine (20) an alkaloid closely reminiscent of Stemmadenine (13) but having only a methylene bridge suggests but does not prove that the extrusion of the carbon from the tryptophan bridge and the necessary modification to the exocyclic methylene at C-3 of stemmadenine may occur as a concerted process. 106-8,110 Incorporation of various forms of secodine (76) in /u_ pyricollum coupled with degradation data places the rearrangement processes which lead to the abnormal skeleton of apparicine (19) at even a later stage i n the biosynthetic scheme. It i s not known at the present time which i f either of these two precursors secodine (76) or stemmadenine (13) is the immediate precursor of the apparicine system for i t is known that an equilibrium exists between stemmadenine (13), dehydrosecodine (72) and the aspidosperma alkaloids. A more enlightened rationale for the biosynthesis of these non-tryptamine 117 bridged alkaloids was deduced from a study by a French group of the in vitro fragmentation of the tryptamine bridge under modified Polonovski reaction con-ditions (i.e. reaction of the N-oxide with trifluoroacetic acid). It was suggested that a similar fragmentation mechanism could be involved in the extrusion of the one carbon unit during the biosynthesis of apparicine (19) (figure 17). - 35 -This proposal encompases a large portion of the known incorporation data, including the involvement of stemmadenine (13) as a crucial precursor and the extrusion of the C-2 and not C-3 of tryptophan in the formation of 108 apparicine (19). A minor discrepancy exists however, for i t has been shown that the hydroxymethylene and not the carbomethoxy functionality at C-16 is lost during the rearrangement process. 118 Subsequently this mechanistic proposal has been extended to account for biosynthesis of uleine (18) and the oxidized and reduced members of the olivacine (16) and e l l i p t i c i n e (17) series (figure 18). As such, i t represents the abnormal alkaloids to be derived from the convergent indole biosynthetic 28 113 pathway and is thus a very viable alternative to the Wenkert and Djerassi postulates. Also, the generation of uleine (18), apparicine (19) and the pyridocarbazole alkaloids 16, 17 25 and 26 from common intermediates suggesting the involvement of similar enzymic processes in the biosynthetic pathway is in accord with the observed frequent co-occurrence of these compounds in Aspidosperma and Ochrosia plants. As mentioned however, no biosynthetic data exists at the moment to prove 118 that the unified hypothesis as proposed by the French group for olivacine (16), e l l i p t i c i n e (17), uleine (18), guatambuine (25), or N-methyltetrahydro-28 e l l i p t i c i n e (26) i s correct (figure 18), or whether the Wenkert postulate can be substantiated (figure 14). 18. The Potier-Janot Postulate for the Biosynthesis of the Non-Tryptamine Bridged Alkaloids. - 37 -DISCUSSION - PROLOGUE The work presented i n this thesis represents part of a long-range program in our laboratory for the investigation of the biosynthesis of representative alkaloids of the uleine (18), apparicine (19), and pyrido-carbazole systems i n a variety of Aspidosperma plants. Previous work on this project has concerned i t s e l f primarily with the biosynthesis of 106, 108, 110, 119 apparicine (19) and uleine (18) in A. pyricollum. Some preliminary feeding experiments have also been conducted on the biosynthesis 119, 120 of uleine (18), olivacine (16) and guatambuine (25) in A. australe. The emphasis in the present study however has been on developing the background work necessary for a detailed study of the biosynthesis of the pyridocarbazole systems represented by olivacine (16), guatambuine (25), e l l i p t i c i n e (17) and N-methyltetrahydroellipticine (26). Four basic requirements were considered i n i t i a l l y to be essential to this detailed biosynthetic study: 1. a suitable plant source that contains the compounds of interest and is actively biosynthesizing them. 2. a synthetic scheme for the specific labelling of proposed precursors, . 3. an eff i c i e n t degradation scheme for the isolation of specific radio-labelled atoms. 4. an adequate source of unlabelled compounds for purposes of comparison, for dilution studies, and for chemical transformation. - 38 -Different aspects of these requirements were developed for biosynthetic work i n two plant systems A. australe, and A. vargasii as described in the three parts of this thesis. Part 1 i s concerned with plant extractions so as to obtain the required alkaloids from a crude A. australe extract, and to determine the f e a s i b i l i t y of plant feeding experiments in A. vargasii. Part 11 describes the development of a degradation scheme for both the olivacine (16) and e l l i p t i c i n e (17) series, and Part 111 describes the synthesis of olivacine (16) and guatambuine (25) for the continuation of future biosynthetic work. - 39 -DISCUSSION - PART 1 Section A, Plant Extractions:  Aspidosperma Australe (Mull. Argov.) 121-3 A. australe has been reported by Ondetti and Deulofeu to contain as their principle alkaloidal components olivacine (16) and both enantiomers of guatambuine (25). The a v a i l a b i l i t y of approximately one hundred grams of 124 a crude methanolic extract from a large scale extraction of A. australe therefore prompted the development of an isolation procedure whereby con-venient but limited supplies of these two alkaloids could be obtained for degradation work. A general alkaloid extraction procedure described by G i l b e r t 1 0 was used. This method involved treatment of the methanolic extract with 15% acetic acid and extraction with petroleum ether to remove neutral and acidic plant material. By subsequent basification and extraction with chloroform the basic alkaloidal fraction was isolated. Thin layer chromatography (TLC) on this extract showed the presence of two major components, a visible bright yellow spot which fluoresced under UV and sprayed brown with eerie sulphate which corresponded to olivacine (16) and a dark blue spot under UV which sprayed turquoise with eerie sulphate corres-ponding to guatambuine (25). For their isolation, advantage was taken of the observation on TLC that both compounds possessed a similar retention time on alumina whereas they exhibited very different Rf values on s i l i c a gel. On s i l i c a chromatoplates - 40 -guatambuine (25) occurred at very low Rf whereas olivacine (16) moved with the solvent front, this behavior reflecting the considerable differences in 125 their known bas i c i t i e s . This characteristic difference made i t possible to f i r s t separate both components from the small amounts of other materials present in the extract by column chromatography on alumina and subsequently to separate them from each other by chromatography on s i l i c a gel. Considerable d i f f i c u l t i e s were encountered, however, in the application of the alkaloidal extract onto the column. Olivacine (16) and guatambuine (25) are soluble only in polarsolvents, for example methanol, acetonitrile and pyridine. It was found, however, that a chloroform solution containing minimum amounts of methanol and pyridine would solubilize the extract without interfering with the separations. Upon recrystallization from methanol, i t was determined that olivacine (16) and ( - ) - guatambuine (25) were present in the crude extract to the extent of 1.8 and 1.6% respectively. Olivacine (16), obtained crystalline as yellow needles was characterized by i t s N.M.R. and mass spectra, as well as the superimposability of i t s UV. and 126 IR. spectra with the reported spectra. The N.M.R. spectrum exhibited two singlets at 62.80 and 63.15 for the C-5 and C-l methyl hydrogens respectively. The mass spectrum exhibited a parent peak,at m/e = 246 and very l i t t l e fragmen-tation due to the aromatic nature of the molecule. ( - ) - guatambuine (25), obtained as cream coloured cubes was also charac-126 terized by i t s spectral data. The mass spectrum exhibited a parent peak at - 41 -m/e = 264 and c h a r a c t e r i s t i c fragments at m/e = 249, 233, 221-218, 204. The N.M.R. spectrum i s shown i n f i g u r e 19 because i t possesses several important features which w i l l be alluded to repeatedly i n subsequent discussions. Of note i s the presence of a doublet at 1.52 (J = 7 Hz) f o r the C - l methyl group and the presence of a quartet at 3.90 (J = 7 Hz) f o r the methine hydrogen at the same p o s i t i o n . The chemical s h i f t o f the methine hydrogen i s very c h a r a c t e r i s t i c as i t i s influenced by the f a c t that i t i s both b e n z y l i c and adjacent to the basic nitrogen atom. Also present i n the N.M.R. spectrum i s a s i n g l e t at 2.40 (N-CH^) and a s i n g l e t at 2.54 (C-5,CH 3). 0 ' 1 • 1 i 1 1 • ' i ' 1 ' ' i i i i M 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 r 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 i i i i i i i i i i i i i i j. (25) H .•If*" 1 1 1 1 1 1 1 1 1 1 1 1 1 I i i i i -I r i i i I i i i i 1 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 r i i . i I i i i i I I i i j [ i i 10 Figure 19. The N.M.R. Spectrum of Guatambuine (25) - 43 -Aspidosperma vargasii (A.DC.) A. vargasii plants when mature are trees standing up to twenty meters 127 in height with relatively slender branches and a close thin bark. They are indigenous to rocky arid slopes and transition forest in Venezuala, and adjacent Columbia and Guiana. The bark of A. vargasii has been studied 128 previously by Burnell and Delia Casa (1967) and has been shown to contain as i t s principle constituents three pyridocarbazole alkaloids, guatambuine (25), N-methyltetrahydroellipticine (26) and 9-methoxyolivacine (82). (82) A. vargasii represents, therefore, an ideal plant in which to further extend our biosynthetic program for the two components 9-methoxyolivacine (82) and N-methyltetrahydroellipticine (26) are not present in the previously studied plants A. pyricollum and A. australe. Prior to any radioactive feeding experiments the s u i t a b i l i t y of A. vargasii as a plant source for biosynthetic study had to be determined. This involved f i r s t l y the determination of the presence of these constituents in the live plants i n quantities that could be isolated and secondly the development of a feasible isolation scheme whereby the desired constituents could be obtained in pure crystalline form for radioactive counting. - 44 -A typical masceration-extraction procedure was u t i l i z e d to obtain a crude methanolic extract of the plant. U t i l i z i n g the observation by 128 Burnell and Delia Casa that guatambuine (25) was extractable from aqueous acidic solution with chloroform an extraction procedure was devised to partition the alkaloids present in the crude methanolic extract into acidic-chloroform extractable and basic-chloroform extractable fractions. Suspending the crude methanolic extract in 10% hydrochloric acid and extracting f i r s t with petroleum ether to remove non-polar material and then with chloroform (acidic-chloroform extract) afforded a mixture of three major alkaloidal components. One of these components provided an identical retention time and colour development with either guatambuine (25) or N-methyltetrahydroellipticine (26), TLC characteris-tics of these two alkaloids being very similar. Subsequent basification of the aqueous phase and extraction with chloroform (basic-chloroform extract) yielded two further major components and one minor component the least polar of which corresponded to an olivacine-like compound by comparison of i t s Rf value and fluorescent response to UV with authentic olivacine (16). By means of thin layer chromatography i t appeared as though the p a r t i -tioning was highly efficient as very l i t t l e of either of the basic-chloroform extract components could be detected in the acid-chloroform extract, and similarly only small amounts of the guatambuine-like (25) component could be detected in the basic-chloroform extract. Of particular significance in the above extraction was the detection of three previously unreported components in relatively large amounts. - 45 -The separation of the various components in each extract was accom-plished by column chromatography on alumina. Table 1 provides a summary of the results obtained during this separation. In most instances the column chromatography had to be repeated on the various combined fractions in order to obtain sufficiently pure material for recrystallization. This was particularly true for the separation of Uleine (18), guatambuine (25) and the minor "uleine l i k e " component. In each case the chromatography fractions were monitored by TLC, UV. and, where necessary, Fourier transform N.M.R. spectroscopy. The UV. spectrum was a particularly important diagnostic tool in revealing the presence of the minor "uleine l i k e " component on chromatography of the combined guatambuine (25) fractions. This component was present in very small quantities and could be detected only on TLC as a light blue colouration in the t a i l of the turquoise-green spot for guatambuine (25) . On the basis of the UV. spectrum, this component in the extract mixture was shown to possess the chromophoric system characteristic of Uleine (18) and apparicine (19). No subsequent data was obtained to further determine i t s structure. The Fourier transform N.M.R. spectrum of the guatambuine (25) fraction demonstrated i t to be homogeneous and not containing any N-methyltetrahydro-e l l i p t i c i n e (26) which possesses almost identical properties. Although this 128 latter component was anticipated to be present in the plant extract, no trace of i t was ever detected in any of the several plant extraction experiments conducted - 46 -Table 1 - Column Chromatography Results on Extracts from A. vargasii A) Acid Chloroform Extract: x 10z Frac. No. Solvent Wt.(mg.) % of Total Plant Structure 1 - 8 CHC1, 105 3.88 apparicine (19) 10 - 20 21 - 44 CHCl3:EtOAc 33 20:60% CHCl3:EtOAc 47 Et0Ac:Me0H 8% 1.22 1.74 uleine (18) guatambuine (25) From rechromatography of fracs. 21 - 44 "Uleine like component" B) Base Chloroform Extract: 1 - 1 5 CHC13 to 100 CHCl3:EtOAc 20% 3.70 9-Methoxyolivacine (82) 17 - 20 CHCl3:EtOAc 40% dihydroolivacine (141) 21 - 35 CHCl3:EtOAc 40% to EtOAc: MeOH 5% 46 1.73 Desmethyluleine (84) - 47 -The absence of N-methyltetrahydroellipticine was disappointing since i t would have been desirable to evaluate the biosynthesis of the whole series of these alkaloids in the same plant species. On the other hand, i t s presence would have complicated the isolation procedure greatly, for even t r i a l separations of a mixture of pure guatambuine (25) and N-methyltetrahydro-e l l i p t i c i n e (26) proved to be very d i f f i c u l t on a small scale. The identities of apparicine (19), uleine (18) and desmethyluleine (84) were determined by a comparison of mass spectral, N.M.R. and UV. data with 129,113 that already reported. A characteristic feature in the N.M.R. spectra of a l l three compounds i s a pair of singlets in the o l e f i n i c region <54-5 for the two protons of the exocyclic methylene group. The presence of a doublet at 6 4.11 i n the spectrum of the uleine compounds and assigned to the C-1H (see table 1) which i s both benzylic and adjacent to the basic nitrogen, also pro-vided a fac i l e differentiation from apparicine (19) since the latter possesses geminal hydrogens at the C-l position and exhibits an AB quartet in the same region. Of particular note also was the absence of the N-methyl signal in the spectrum of desmethyluleine (84). The structure of guatambuine (25) was based upon the superimposability of both N.M.R. and IR spectra with those of an authentic sample. The structure of 9-methoxyolivacine (82) was determined from i t s N.M.R. spectrum which exhibited two singlets at 6 2.80 and 3.16 for the aromatic methyl protons and a singlet at 6 4.00 for the methoxy group. It is noteworthy that the position of these absorptions is close in value to those obtained for olivacine (16), and markedly different to the position of the singlet peaks in - 48 -e l l i p t i c i n e (17). The UV. spectrum was almost identical to that reported for 9-methoxyolivacine (82) and distinctly different than for olivacine (16). The mass spectrum possessed a parent peak at m/e = 276 consistent with the required molecular formula and like olivacine (16) i t possessed an almost negligible fragmentation pattern. The minor component in the basic chloroform extract (fractions 17 - 20, Table 1) of which only several milligrams were isolated was tentatively determined to be 3,4-dihydroolivacine (83) on the basis of the TLC and UV comparisons with 3,4-dihydroolivacine (83) available by synthesis (See sequence A, page 190). It i s clearly evident from the isolation results that A. vargasii is a very suitable plant source for biosynthetic investigations. Not only does i t contain representatives of uleine (18), apparicine (19), and the olivacine series of compounds, but each of the five major components are present in 30-100 mg. quantities and can readily be purified by either crystallization or sublimation techniques. It was observed however that the quantities of uleine (18) obtained depended markedly upon the size (i.e. age) of the plant used in the experi-ment. Only large and thick woody stemmed plants yielded relatively large quantities of uleine (18) (230 mg.). Smaller plants yielded considerably less of this valuable alkaloid (5-10 mg.). It is of interest to note that the A. vargasii results further demonstrate the frequent co-occurrence of the different types of non-tryptamine alkaloids with each other in Aspidosperma plants. - 49 -Section B, Incorporation Experiments i n A. v a r g a s i i (A.DC.) Only a preliminary i n v e s t i g a t i o n of the incorporation of r a d i o a c t i v i t y into the a l k a l o i d s of A. v a r g a s i i has been conducted at t h i s time. The objective behind the experiments was to determine whether ac t i v e biosynthesis was occurring i n the plant and not n e c e s s a r i l y to determine any information regarding the rearrangements that take place on conversion of precursor mole-cules i n t o the a l k a l o i d s . With regards to the present knowledge of indole a l k a l o i d biosynthesis as o u t l i n e d i n the Introduction, the precursors that were l o g i c a l l y chosen f o r feeding experiments were aromatic t r i t i u m l a b e l l e d tryptophan (27) and stemmadenine (13). The (Ar-3H)-tryptophan (27) experiment was of primary importance f o r i t would demonstrate that o l i v a c i n e (16), guatambuine (25), and u l e i n e (18) are derived from the elucidated indole a l k a l o i d pathway, thus i n v a l i d a t i n g 28 113 the Wenkert and D j e r a s s i proposals. The incorporation of (Ar-3H)-Stemmadenine (13) was also of major importance because by the Potier-Janot 118 postulate i t represents the immediate precursor of the whole s e r i e s o f non-tryptamine bridged a l k a l o i d s . Both precursors were l a b e l l e d with t r i t i u m i n the aromatic r i n g f o r reasons that the t r i t i u m exchange synthesis of the precursors i s r e l a t i v e l y simple and there i s no s i g n i f i c a n t loss of l a b e l by transformations occurring during biosynthesis. Also u n t i l synthetic stemmadenine (13) becomes a v a i l a b l e 119 120 130 i n our laboratory, ' ' Stemmadenine (13) l a b e l l e d i n the aromatic r i n g i s the only r a d i o a c t i v e form of t h i s precursor a v a i l a b l e . - 50 -The aromatically l a b e l l e d precursors were prepared by an exchange reaction i n t r i t i a t e d t r i f l u o r o a c e t i c a c i d , c r y s t a l l i z e d to constant a c t i v i t y and incorporated i n 1-3 mg. q u a n t i t i e s to approximately f i v e year o l d A. v a r g a s i i p l a n t s . The precursors were administered by the cotton wick technique and the biosynthesis was allowed to proceed f o r a one-week period. The plants were then mascerated and the a l k a l o i d a l components i s o l a t e d i n an i d e n t i c a l manner to that described i n section A f o r A. v a r g a s i i . The compounds studied were r e c r y s t a l l i z e d to constant a c t i v i t y meaning that the d i f f e r e n c e between successive countings was less than 5%. The e f f e c t s o f 134 colour quenching on the e f f i c i e n c y of s c i n t i l l a t i o n counting o f methoxyolivacine were standardized by preparing counting samples of i d e n t i c a l concentration a f t e r each r e c r y s t a l l i z a t i o n . A summary of the incorporation r e s u l t s i s presented i n Table I I. The r e s u l t s show that there has been a d e f i n i t e i n c o r p o r a t i o n of both tryptophan (27) and stemmadenine (13) i n t o methoxyolivacine (82) and that tryptophan has also been incorporated i n t o guatambuine (25). Both precursors however, were not incorporated i n t o u l e i n e (18). The r a d i o a c t i v e incorporation of these two compounds in t o the other constituents of A* v a r g a s i i was not determined during these i n i t i a l experiments. The absence of any incorporation into u l e i n e (18) was f r u s t r a t i n g because s i m i l a r r e s u l t s have been obtained f o r t h i s molecule during incorporation experiments i n other plant systems. Nothing i s presently known concerning the nature of the enzymic processes that produce these compounds within the plant or what conditions are necessary to stimulate Table II. Results of Incorporation of Tryptophan (27) and Stemmadenine (13) into A. vargasii. n , Activity Compound ^ ^  ' Isolated (DPM) Specific Activity isolated (DPM/mmole) Weight isolated (mg.) % incorporation Experiment I (Ar-3H)-Tryptophan (27). (2.50 x 10 1 1 DPM/mmole) 9-Methoxyolivacine c 0 0 1 A11 (82) 5 , 8 2 X 1 0 3.17 x 105 91 0.0049 Guatambuine _ Q O i r . l l (25) 5 , 8 2 x 1 0 i 1.71 x 10 5 40 0.0018 U 1r ^ " e 5.82 x 10 1 1 0.00 6' — Experiment II (Ar-3H)-Stemmadenine (13). (3.21 x 10 1 0 DPM/mmole) 9-Methoxyolivacine , o r i -,rt10 (82) 6 ' 2 9 X 1 0 5.74 x 104 85 0.0077 U}?j?e 6.29 x 10 1 0 0.00 10 — - 52 -these processes. It i s only p o s s i b l e therefore to put forward speculations as to the nature of the d i f f i c u l t i e s associated with u l e i n e (18) biosynthesis. It could be that at the time of year at which the feeding experiments were conducted that u l e i n e (18) was not being a c t i v e l y biosynthesized, or that some environmental or external c o n d i t i o n was not being met which would stimulate i t s production. It could also be that the time scale of the experiments was too long or of i n s u f f i c i e n t length f o r i t s biosynthesis to be detected. The problem of membrane transport can also be suggested to account i n part f o r the low incorporation of tryptophan (27) and stemmadenine (13) in t o the a l k a l o i d s which did possess a c t i v i t y . I t i s undoubtedly probable that the precursors e s p e c i a l l y tryptophan (27) were also u t i l i z e d f o r purposes other than a l k a l o i d production which would also contribute to the low a c t i v i t i e s obtained. Due to the nature of the feeding technique however, i t was h i g h l y probable that the proposed precursors were absorbed i n t o the plant before s i g n i f i c a n t external decomposition could take place (absorption time 2-5 h r . ) . The only s i g n i f i c a n c e that can be a t t r i b u t e d to these preliminary incorporation r e s u l t s i s that both stemmadenine (13) and tryptophan (27) have been u t i l i z e d by the plant system to produce methoxyolivacine (82) and guatambuine (25). This does not imply i n the absence o f enzymic studies that these two compounds are n e c e s s a r i l y the precursors e x i s t i n g n a t u r a l l y within the plant. It also cannot be said that because stemmadenine (13) was incorporated into methoxyolivacine (82) that the - 53 -transformations that occurred proceeded according to the Potier-Janot postulate. The bi o s y n t h e t i c r e s u l t s presented here are however the f i r s t to be obtained f o r the pyridocarbazole a l k a l o i d system and do give some i n d i c a t i o n as to intermediacy o f tryptophan (27) and stemmadenine (13) i n t h e i r biosynthesis. Further b i o s y n t h e t i c i n v e s t i g a t i o n s are c u r r e n t l y being conducted i n our laboratory to elucidate the pathway leading to t h i s c l a s s of indole a l k a l o i d s . - 54 -EXPERIMENTAL - PART I Melting points were determined on a Kofler block and are uncorrected. A l l ultraviolet (UV.) spectra were recorded in methanol using a Cary 15 recording spectrophotometer. The infrared (IR.) spectra were recorded with a Perkin-Elmer Model 457 spectrophotometer either in chloroform solution (cavity cells 0.5 mm.) or as a nujol mull between sodium chloride plates as indicated. A l l measurements were made in cm _l and calibration of the spectra was achieved using the 1604 air* absorption band of polystyrene. Nuclear magnetic resonance spectra (N.M.R.) were obtained in deuterochloro-form solutions (unless otherwise indicated) at 100 MHz on a Varian HA-100 or a Varian XL-100 nuclear magnetic resonance spectrometer. A l l N.M.R. spectra obtained via the Fourier Transform technique (F.T.) w i l l be so noted and were obtained with the Varian XL-100 instrument. Chemical shifts were given in (ppm) with reference to tetramethylsilane as the internal standard. The multiplicity, integrated areas, and proton assignments are given in paren-theses. Mass spectra were determined on an AEI-MS-902 or an Atlas CH-4B mass spectrometer, with high resolution mass spectra determined with the former. Woelm neutral alumina and EM Reagents GF254 s i l i c a gel were used for thin and preparative layer chromatography. In part I, unless otherwise specified, ethyl acetate was used as the solvent for development of the chromatoplates. Woelm neutral alumina (Act. I l l ) and Merck s i l i c a gel (Act. 2-3) were used for column chromatography. Radioactivity was measured with a Nuclear-Chicago Mark II liquid scin-t i l l a t i o n counter in counts per minute (cpm). The radioactivity of the sample - 55 -in disintegrations per minute (dpm) was subsequently determined by the 133a,b external standard technique using a b u i l t - i n Barium-133 source of gamma radiation. S c i n t i l l a t i o n cocktails used were either a prepared solution of toluene and a Nuclear-Chicago PPO-POPOP concentrate or a premixed solution of Nuclear-Chicago PCS cocktail. A s c i n t i l l a t i o n counting sample consisted of a solution of the sample dissolved in 1 ml. of methanol and 14 ml. of the appropriate cocktail (total volume 15 ml.). For each sample the background of the v i a l was predetermined and subsequently subtracted from the measured dpm's for the sample i n subsequent calculations. Each sample was counted for a time period long enough for the total counts for the sample, less the total counts for the background to exceed ten thousand counts. The A. vargasii plants used in this study were grown in the Horticulture Department greenhouse, the University of British Columbia. Extraction of A. australe (Mull. Argov.) 124 The Crude methanolic extract (20 gm.) was suspended in 15% acetic acid (350 ml.) and stirred for 1.5 hr. The aqueous suspension was then extracted with petroleum ether (3 x 200 ml.) and insoluble particles that remained in the aqueous phase were removed by suction f i l t r a t i o n . The aqueous phase was basi-fied with 10% sodium hydroxide solution to pH 10 and extracted with chloroform (3 x 200 ml.). The chloroform layer was dried and concentrated to give a so l i d alkaloid extract (1.26 gm.). The very polar components were removed by rapid - 56 -f i l t r a t i o n through an alumina column (Act I I I ) . The olivacine-guatambuine f r a c t i o n obtained from the f i l t r a t i o n was dis s o l v e d i n chloroform containing a small amount of methanol and py r i d i n e (1 ml.) and then applied to a s i l i c a -gel column (100 gm., Act 2-3). I n i t i a l , e l u t i o n with chloroform, followed by ethyl:acetate and ethy l acetate-methanol (10%) gave o l i v a c i n e (16) (360 mg., 1.8%). Subsequent e l u t i o n with methanol y i e l d e d guatambuine (25) (320 mgs., 1.6%). The o l i v a c i n e (16) was r e c r y s t a l l i z e d from methanol to give yellow needles, m.p. 318-325° ( l i t . , m.p. 3 1 8 - 3 2 4 ° ) . 1 2 6 UV.; Amax (log e): 374 (3.56), 325 (3.76), 291 (4.84), 284 (4.88), 275 (4.72), 265 (4.57), 235 (4.33), 222 (4.40). N.M.R. (FT.): 3.15 ( s i n g l e t , 3H, C - l CH 3), 2.80 ( s i n g l e t , 3H, C-5 CH 3). Mass spectrum: M+, m/e = 246. Guatambuine (25) was obtained as cream coloured cubes by r e c r y s t a l l i z a t i o n from methanol, m.p. 242-250° ( l i t . , 2 4 8 - 2 5 0 ° ) . 1 2 6 UV.; Amax (log e): 341 (3.48), 327 (3.63), 297 (4.26), 288(sh) (4.08), 262 (4.36), 250 (4.50), 240 (4.64). N.M.R. (FT.) (Figure 19): 3.90 (quartet, IH, J = .7 Hz, C - l H), 2.54 (s i n g l e t , 3H, C-5CH 3), 2.40 ( s i n g l e t , 3H, N-CHj), 1.52 (doublet, 3H, J = 7 Hz, C-l CH 3). Mass spectrum: M+, m/e = 264; main peaks: 249, 233, 221-218, 204. Extraction of A. Va r g a s i i (A.DC.) Two woody A. v a r g a s i i plants (270 gm.) (approximately 5 years o l d , and containing an extensive root system) were ground up i n a Wiley m i l l (not predried). The c o l l e c t e d pulp was suspended i n methanol (3 x 300 ml.) and - 57 -furt h e r mascerated i n a Waring blendor (x3). A f t e r each masceration the s o l i d plant residues were separated by suction f i l t r a t i o n and washed l i b e r a l with methanol. The combined methanol f i l t r a t e s were concentrated to dryness. The crude methanol extract was suspended i n 10% hydrochloric acid (500 ml.) and s t i r r e d at room temperature f o r 0.5 hr. It was then extracted successively with petroleum ether 65-110° (3 x 250 ml.) and chloroform (3 x 250 ml.). The petroleum ether extract was discarded while the chloroform extract (Acid chloroform extract) was dr i e d over sodium sulphate, and concentrated to give a s o l i d mixture (500 mg.) The aqueous phase was then b a s i f i e d to pH 11 with concentrated ammonium hydroxide and extracted with chloroform (3 x 250 ml.). The chloroform s o l u t i o n (Base chloroform extract) was d r i e d over sodium sulphate and concentrated to give a yellow s o l i d mixture (300 mg.). The a c i d chloroform extract was column chromatographed on alumina (30 gm.). The extract was d i s s o l v e d i n hot chloroform and applied immediately to the column. The column was eluted (5 ml. f r a c t i o n s ) by gradient e l u t i o n s t a r t i n g with chloroform followed by chloroform:ethyl acetate mixtures (10% increment increases of ethyl acetate) and f i n a l l y ethyl acetate-methanol (9:1) (See Table 1). Fractions 1-8 eluted with chloroform contained apparicine (19) which a f t e r r e c r y s t a l l i z a t i o n from acetone y i e l d e d c o l o u r l e s s needles (105 mg., 3.88 x 10" 2 % ) , m.p. 192-194° ( l i t . , m.p. 192-194°), [ a ] 2 6 = -165.2° (C. 1.0 i n CHC1 3), ( l i t . , [ a ] 2 7 = ± 177° i(C. 2.16 i n C H C 1 3 ) . 1 1 3 UV.; Xmax - 58 -(log e): 303 (4.50); Xmin (log e): 265 (3.70). N.M.R.: 7.90 (broad s i n g l e t , 1H, N-H), 5.42, 5.28 (two s i n g l e t s , 1H each, =CH 2), 5.30 (multiplet, 1H, =CH-CH3, p a r t l y obscurred by exocy c l i c methylene s i n g l e t ) , 4.28, 4.58 (AB quartet, 2H, J = 18 Hz, C-l CH 2), 1.47 ( s p l i t doublet, 3H, J = 8 Hz, J = 1.5 Hz, =CH-CH,j), mass spectrum: M+, m/e = 264; main peaks: 249, 235, 222, 208. Found: C, 81.51; H, 7.63; N, 10.45%. Calc. f o r C 1 0 H _ n N „ : C, 81.78; H, 7.63; N, 10.60%. Fractions 10-20 eluted with chloroform:ethyl acetate (20-60%) contained uleine (18) which a f t e r r e c r y s t a l l i z a t i o n from methanol y i e l d e d c o l o u r l e s s cubes (33 mg., 1.22 x 10 % ) . UV.; Amax (log e): 301 (4.50). N.M.R. : 8.22 (broad s i n g l e t , 1H, N-H), 7.6-7.1 (multiplet, 4H, aromatic), 5.28, 5.00 (two s i n g l e t s , 1H each, =CH 2), 4.11 (doublet, 1H, J = 2 Hz, C - l H), 2.30 (s i n g l e t , 3H, N-CH^). Mass spectrum: M+, m/e = 266; main peaks: 251, 237, 223, 222, 209, 208, 207, 194, 180. Fractions 21-44 eluted with chloroform:ethyl acetate 1:1 up to ethyl acetate:methanol 8% contained guatambuine (25) which a f t e r r e c r y s t a l l i z a t i o n -2 from methanol y i e l d e d cream coloured cubes (47 mg., 1.74 x 10 % ) , m.p. 250-252°, ( l i t . , 2 4 8 - 2 5 0 ° ) . 1 2 8 . UV.; Xmax (log E ) : 340 (3.50), 325 (3.63), 297 (4.25), 287(sh) (4.08), 260 (4.36), 248 (4.50), 239 (4.64). N.M.R. (FT.): 3.86 (quartet, 1H, J = 6 Hz, C - l H), 2.46 ( s i n g l e t , 3H, C-5 CH 3), 2.34 (s i n g l e t , 3H, N-CH 3), 1.45 (doublet, 3H, 3=1 Hz, C-l CH 3). Mass spectrum: M+, m/e = 264; main peaks: 249, 233, 221-218, 204. Found: C, 55.85; H, 5.62; N, 6.73%. Calc. f o r C i g H 2 3 N 2 I : C, 56.17; H, 5.71; N, 6.89%. By preparative layer chromatography (alumina, 1.0 mm., EtOAc) o f a - 59 -small portion of the Fra c t i o n 21-44 material the blue c o l o u r a t i o n at the t a i l of the guatambuine (25) spot (turquoise, eerie sulphate) was i s o l a t e d . UV.; Amax: 303; Amin: 265. The base chloroform extract was chromatographed on alumina (20 gm.) i n an i d e n t i c a l fashion. Fractions 1-15 eluted with chloroform contained 9-methoxyolivacine (82) which a f t e r r e c r y s t a l l i z a t i o n from methanol y i e l d e d yellow needles (100 mg., 3.70 x 1 0 _ 2 % ) , m.p. 290-292° ( l i t . , m.p. 291-293°). 1 2 UV.; Amax (log e): 330 (3.80), 295 (4.70), 269 (4.59), 239 (4.37), 222 (4.32). N.M.R. (FT.): 4.00 ( s i n g l e t , 3H, 0CH 3), 3.16 ( s i n g l e t , 3H, C-l CH 3), 2.80 ( s i n g l e t , 3H, C-5 CH 3). Mass spectrum: M+, m/e = 276; main peaks: 261, 245. Found: C, 74.50; H, 6.35; N, 9.02. Calc. f o r C l gH 1 6N 20(CH 30H): C, 74.02; H, 6.49; N, 9.09. Fractions 17-20 eluted with chloroform:ethyl acetate (20-40%) contained a yellowish s o l i d which was furthe r p u r i f i e d by preparative l a y e r chromatography on alumina (1.0 mm., EtOAc:MeOH 20%). A small undetermined amount of a yellow powder corresponding to 3,4-dihydroolivacine (141) by TLC (alumina, EtOAc) was obtained. UV.; Amax: 367, 310, 294, 280, 275, 235. Fractions 21-35 eluted with chloroform:ethyl acetate (40%) to ethyl acetate:methanol (5%) contained desmethyluleine (84) as a yellow f i l m (46 mg., 1.73 x 10" 2%). UV.; Amax: 303; Amin: 265. N.M.R. (FT.): 5.30, 5.00 (two s i n g l e t s , IH each, =CH 2), 4.34 (multiplet, IH, C - l H). Mass spectrum: M+, m/e = 252; main peaks: 235, 233, 209-8, 194-5, 180. - 60 -(Ar-3H)-Stemmadenine (13) T r i t i a t e d t r i f l u o r o a c e t i c a c i d was prepared by d i s t i l l a t i o n of t r i f l u o r o a c e t i c anhydride (0.80 ml., 5.55 x 10 mole) in t o t r i t i a t e d water -3 (100 y l . , 5.55 x 10 mole, 100 mCi) using a vacuum t r a n s f e r system. The t r i t i a t e d t r i f l u o r o a c e t i c a c i d was then combined with stemmadenine (13) (25 mg., 7.0 x 10 mole) by means of the vacuum t r a n s f e r system and the r e s u l t i n g a c i d s o l u t i o n was maintained under a dry nitrogen atmosphere at room temperature f o r 48 hr. The t r i t i a t e d t r i f l u o r o a c e t i c a c i d was then d i s t i l l e d o f f and the residue was taken up i n d i l u t e ammonium hydroxide (75 ml.) and extracted with methylene c h l o r i d e (3 x 30 ml.). The methylene c h l o r i d e f r a c t i o n s were combined, d r i e d over sodium sulphate, and concen-t r a t e d to a colou r l e s s foam (17 mg., 68%). A f t e r s i x r e c r y s t a l l i z a t i o n s 7 from methanol the measured r a d i o a c t i v i t y was constant (8.98 x 10 DPM/mg., 3.21 x 1 0 1 0 DPM/mmole.) (PP0-P0P0P c o c k t a i l ) . Plant Incorporation Experiment: The Administration of (Ar-3H)-Tryptophan (27) to A. v a r g a s i i (Ar-3H)-Tryptophan (27) (2.33 mg., 2.50 x 1 0 1 1 DPM/mmole) was dis s o l v e d i n a s o l u t i o n of methanol:water:acetic a c i d (1 ml.:4 ml.: 1 drop) and administered to an A. v a r g a s i i plant (^130 gm.) by the cotton wick technique. This method required the threading of a cotton s t r i n g through the stem of the growing plant at a point above the ground, but below the branching point. The intertwined ends of the wick were placed i n a 18 x 75 mm. t e s t - 61 -tube containing the precursor s o l u t i o n located at the base of the pl a n t . For the remaining period of the incubation a f t e r the precursor had been absorbed (2-5 hr.) the t e s t tube was kept f u l l by repeated additions o f d i s t i l l e d water. A f t e r 7 days under intermittent fluorescent lamp i l l u m i n a t i o n the plants were extracted f o r t h e i r a l k a l o i d content by the procedure p r e v i o u s l y described. Methoxyolivacine (82), guatambuine (25) and u l e i n e (18) were i s o l a t e d by column chromatography on alumina and r e c r y s t a l l i z e d to constant a c t i v i t y (PPO-POPOP c o c k t a i l ) . The q u a n t i t i e s i s o l a t e d , s p e c i f i c a c t i v i t i e s and percentage incorporations f o r these compounds are presented i n Table I I , page 51. The Administration of (Ar-3H)-Stemmadenine (13) to A. v a r g a s i i (Ar-3H)-Stemmadenine (13) (1.96 mg., 3.21 x 1 0 1 0 DPM/mmole) was dis s o l v e d i n methanol:water (1:4) (5 ml.) and administered to an A. v a r g a s i i plant (^130 gm.) by the cotton wick technique. A f t e r 7 days under inte r m i t t e n t fluorescent lamp i l l u m i n a t i o n the plants were extracted as •7 p r e v i o u s l y described: Acid chloroform extract (312 mg. 5.13 x 10 DPM); 7 Base chloroform extract (247 mg. 1.07 x 10 DPM). Methoxyolivacine (82), and Uleine (18) were i s o l a t e d by column chromatography on alumina and r e c r y s t a l l i z e d to constant a c t i v i t y (PPO-POPOP c o c k t a i l ) . The q u a n t i t i e s i s o l a t e d , s p e c i f i c a c t i v i t i e s and percentage incorporations f o r these compounds are presented i n Table I I , page 51. - 62 -INTRODUCTION - PART II The elucidation of the transformations that take place in the biosynthetic pathway leading to the non-tryptamine bridged indole alkaloids by the radioactive labelling technique requires both the synthesis of specifically labelled mole-cules that are postulated to be their precursors and the development of a degra-dation scheme for determining the position of the incorporated radiolabels into the alkaloidal products. To p a r a l l e l , therefore, the biosynthetic feeding experiments that have 119 been i n i t i a t e d in our laboratory on olivacine (16) and guatambuine (25) in A. australe, a degradation scheme has been developed for these molecules. Further, 131 as a consequence of the ava i l a b i l i t y of large amounts of e l l i p t i c i n e (17) and i n i t i a l l y of only limited quantities of olivacine (16) and guatambuine (17) from natural sources, reactions studied in the e l l i p t i c i n e (17) series provided both a model system and an extension of the degradation scheme to include this series. I n i t i a l experiments in A. australe involved feeding labelled forms of secodine (76), however, i t i s currently thought that secodine (76) occurs to far along the pathway to be a precursor of the non-tryptamine bridged alkaloids. Stemmadenine (13) is considered to be the direct progenetor of these compounds and i t i s also considered that the equilibrium between dehydrosecodine (72) and stemmadenine (13) i s responsible for the observed low incorporations of seco-dine (76) into olivacine (16) and guatambuine (25). As a result, stemmadenine 119,120,130 (13) i s currently in the stages of being synthesized in our laboratory with the objective in mind of being able to unambiguously introduce labelled - 63 -atoms into the positions indicated in figure 20. 118 In accord with the Potier-Janot postulate, the labelled atoms in stemmadenine (13) would end up in the positions indicated in olivacine (16), guatambuine (25), e l l i p t i c i n e (17) and N-methyltetrahydroellipticine (26). The arrow connecting tryptophan (27) and stemmadenine (13) indicates where a label in the C-2 position of tryptophan (27) would occur in stemmadenine (13) and subsequently in the alkaloids. As can be seen, with the exception of the labelled carbomethoxy grouping which may be incorporated into the C-5 position 108 in the "C" ring, the labelled atoms a l l occur in the "D" ring. The degradation scheme developed, therefore, was devised to permit the isolation of the C-l methyl, C-2 methyl (N-methyl), and the C-3 carbon atom in the olivacine (16) series and similarly the C-2 methyl and the C-3 carbon atom in the e l l i p t i c i n e (17) series. This scheme necessarily had to take into account that generally only small amounts of alkaloids (20-50 mg.) with low specific activities were available from plant incorporation experiments. The demands of having to work on a small scale without the option of diluting with unlabelled materials placed constraints upon the developed sequence. The reactions involved had to be simple i.e. not involving any complex work-ups, etc., very e f f i c i e n t and reproducible in terms of yields, and also, the overall sequence should involve a minimum number of reactions. A further and important stipulation was that the end products had to be easily purified and in crystalline form either in their natural state, or as a suitable derivative for radioactive counting. Figure 20. The Labelling Pattern for the Biosynthesis of the Olivacine (16) and E l l i p t i c i n e (17) Series According to the Potier-Janot Postulate. - 65 -(92) (91) Figure 21. The Structure Elucidation of Olivacine (16) by Ondetti and Deulofeu (1961). - 66 -Very l i t t l e chemistry of the olivacine (16)-guatambuine (25) system has 121-3 been studied to date, however, the work by Ondetti and Deulofeu (1961) was particularly applicable to the desired degradation of the olivacine (16) and e l l i p t i c i n e (17) "D" rings (Figure 21). These workers elucidated the structure of olivacine (16) by i t s conversion via i t s methiodide (85) to guatambuine (25) and a subsequent comparison of the Hofmann degradation pro-ducts obtained from guatambuine methiodide (86) with those obtained from uleine (18) via i t s methiodide (87). It was found that guatambuine methiodide (86) undergoes ring opening under Hofmann conditions (NaOH/EtOH) to give two major products. The main component A (32%) was not characterized, but i t was determined by lack of hydrogen uptake to be a non-olefinic material, and the microanalysis of i t s methiodide correlated with the composition C2QH25N2I. The second component (22%) after hydrogenation to 89 and conversion to i t s methiodide 90 was proposed to have the o l e f i n i c structure 88 by comparison of IR., UV., melting point, and mixed melting point with methiodide 90 shown by 132 Schmutz et al_. to be obtained from uleine (18) . Support for their assignments were further strengthened by a comparison of the IR., UV., and melting points of the second Hofmann product 91. 135 Buchi et_ a l . working concurrently in this area cleaved the styryl side chain of 91 under osmium tetroxide-periodate conditions to give the carbazole aldehyde 92. The work of these authors was however incomplete in many respects. The correlation of the degradation products of the two alkaloids led to the correct assignment of the structure of olivacine (16) and guatambuine (25), however i t i s - 67 -d i f f i c u l t to indicate with certainty that the comparisons arc valid based upon the IR., UV. and rudimentary physical data. No data were obtained for the alternate llofmann product 97 that could arise from guatambuine (25), a compound whose spectral data, and quite possibly whose melting point would bear a close resemblance to compounds 88 and 89. Only through the applica-tion of N.M.R. and mass spectroscopic techniques (which were not available to these authors) could the structure of ole f i n 88 be unambiguously determined. However, this sequence potentially permits the isolation of a l l three desired centers in olivacine (16) and guatambuine (25) and consequently i t provided the basis on which the degradation of these two alkaloids as well as the e l l i p t i c i n e s (17) and 26 was developed. - 68 -DISCUSSION - PART II 1_. Interconversions i n the Olivacine (16)-Guatambuine (25) and E l l i p t i c i n e (17)-N-Methyltetrahydroellipticine (26) Series.  The conversion of olivacine (16) and e l l i p t i c i n e (17) to their correspond-ing N-Methyltetrahydro counterparts guatambuine (25) and N-methyltetrahydro-e l l i p t i c i n e (26) was necessary for the totally aromatic structures were not in a suitable form to undergo reactions that would effect scission of the pyrido "D" ring. 121-3 Ohdetti and Deulofeu have converted olivacine (16) to guatambuine (25) by formation of the methiodide of olivacine 85 followed by catalytic hydrogena-tion under lew pressure (Figure 22). In a similar fashion Goodwin, Smith, and 136 Horning converted e l l i p t i c i n e (17) to N-Methyltetrahydroellipticine (26) by formation of i t s methiodide 93 and subsequent reduction of the pyridinium 137 salt with sodium borohydride. Buchi et a l . developed the f i r s t portion of an alternate method of transforming the "D" ring of e l l i p t i c i n e (17). They found that e l l i p t i c i n e (17) underwent facile hydrogenation at atmospheric pressure to give the tetrahydroderivative 94. By subsequent simple methylation i t would be possible to obtain the methiodide of N-methyltetrahydroellipticine (95) . The transformation of e l l i p t i c i n e (17) to i t s N-methyltetrahydro counter-part presented no d i f f i c u l t i e s . Both routes were investigated but i t was found - 69 -(94) (95) Figure 22. The Conversion of Olivacine (16) and E l l i p t i c i n e to their N-Methyltetrahydro Forms. 136 that the sequence of Goodwin et^ al_, was the simplest to follow. The formation of e l l i p t i c i n e methiodide 93 never went totally to completion despite the use of a great excess of methyl iodide. This necessitated the removal of the small amount of residual starting material before reduction because i t s presence made subsequent purification by either recrystallization or column chromatography d i f f i c u l t . This was most easily accomplished by a rapid column chromatography on alumina. The methiodide was e f f i c i e n t l y eluted from the column with methanol:ammonium hydroxide (1%) to give bright orange crystals after concentration (77%). - 70 -Reduction of the pyridinium system with sodium borohydride gave N-methyl-tetrahydroellipticine (26) as cream coloured crystals. It was observed that the crude reaction mixture began to yellow on standing, even in a vacuum dessicator. This decomposition was checked by recrystallization from methanol to give cream coloured needles (91%). The spectral data (UV., N.M.R.) for 26 agreed closely with that previously 126, 128 reported. The N.M.R. spectrum possessed two singlets at 62.56 and 2.70 for the aromatically substituted methyl groups at C-5 and C - l l respectively. Two other singlets were also observed, one for the N-methyl group (62.38), and the other for the isolated methylene group at C-l (63.74). A t r i p l e t occurred at 62.98 (J = 5Hz) for the C-4 methylene protons while the corresponding t r i p l e t for the C-3 methylene protons (62.76) was partly obscured by the s i n g l e t (62.70) for the C - l l methyl group. The conversion of 26 to i t s methiodide 95, in preparation for Hofmann degradation, by reaction with methyl iodide i n methanol proceeded in quantita-tive y i e l d to give colourless needles. The catalytic hydrogenation of e l l i p t i c i n e (17) with platinum oxide de-137 veloped by Buchi et a l . was slightly modified by substituting acetic acid for ethanol as the solvent. It was found that with acetic acid a cleaner product was obtained. The hydrogenation was complete after several hours at atmospheric pressure to give a highly fluorescent pale blue solution which after work-up yielded a colourless crystalline product 94. As reported, the tetrahydro compound 94 was considerably sensitive to air, turning yellow on standing. For this reason a small portion of i t was characterized as it s - 71 -N-acetyl derivative 96 and the remainder was converted directly to N-methyl-tetrahydroellipticine methiodide 95 by reaction with methyl iodide. (96) The ef f i c i e n t conversion of olivacine (16) to guatambuine (25) unlike the transformation of e l l i p t i c i n e (17) was unexpectedly found to be very d i f f i c u l t . The major hurdle in the synthesis proved to be the methylation of olivacine (16) to i t s methiodide 85. Reaction of olivacine (16) with methyl iodide i n ethanol under a wide range of conditions from room temperature to a sealed tube at 150° resulted consistently i n only a 30% conversion to i t s methiodide. A persistent problem with the reaction also was the presence of considerable quantities of uniden-t i f i e d dark coloured material which hampered work-up. The salt 85 was obtained in low yi e l d as a yellow microcrystalline s o l i d . For purposes of degradation the yi e l d of the methiodide was increased to 80% by repeated recycling of reaction mother liquors which contained unreacted olivacine (16). The problem of formation of methiodide 85 was eventually overcome by 138 using methylfluorosulphonate as the alkylating agent and acetonitrile as - 72 -the solvent medium. The reaction proceeded practically instantaneously with methofluorosulphate formation. The salt was not isolated (due to the presence of flurosulphuric acid as a high boiling contaminant) but reduced directly to give guatambuine (25) as described below. In a similar fashion to methiodide formation i t was found that catalytic hydrogenation of olivacine methiodide (85) with platinum oxide at low pressure (5 atm., 80°) proceeded in poor or more correctly inconsistent y i e l d . On the other hand reduction of the pyridinium ring of 85 (methiodide or fluorosulphate) proceeded readily and i n excellent y i e l d by reaction with sodium borohydride. In this manner guatambuine (25) was obtained as a pure cream coloured crystalline s o l i d (93%). The spectral data for this compound was identical to that previously obtained (Part I). The subsequent conversion of guatambuine (25) to i t s methiodide 86 pro-ceeded in quantitative y i e l d by reaction with methyl iodide. A preliminary attempt was made to by-pass olivacine methiodide (85) forma-tion by direct hydrogenation of olivacine (16) i n an analogous fashion to that developed for e l l i p t i c i n e (17). Even after prolonged hydrogenation under pressure (2.5 atm.) considerable starting material remained and the product that was formed, as previously observed, proved to be unstable in air thereby adding complications to i t s isolation and purification. This approach was consequently abandoned. - 73 -2_. Hofmann Elimination: From a consideration of the mechanism of the Hofmann elimination reaction two structurally isomeric olefinic products 88 and 97 can be anticipated to be formed during the base induced (Hofmann conditions) ring opening of the methiodide of guatambuine (25). (§6) (88) (97) 139 In accord with the Hofmann rule i t could be anticipated that i n an unsymmetrical cyclic amine like guatambuine (25) there would be a preferential formation of one o l e f i n i c product over the other. In the vast majority of cases the; hydrogen removed during elimination derives from the least substi-tuted carbon atom, this leading in turn to the least substituted o l e f i n . In the guatambuine system elimination in either direction leads to an unsubstituted olefinic product, however i t might be expected on the above consideration that elimination in the direction of the C-l methyl group leading to the C-3 olefin 88 would be the preferred course. However, an overriding influence to this general trend results from the - 74 -presence of a relatively acidic benzylic hydrogen on the C-4 carbon atom. In situations where a benzylic hydrogen can become involved in the 6-elimina-tion process, elimination in its direction tends to predominate. The predominant elimination product anticipated for guatambuine (25) would therefore be the C-2 olefin 97. For purposes of degrading guatambuine (25) the production of both ol e f i n i c products by Hofmann elimination would be ideal for then a simple oxidative cleavage of both double bonds would complete the isolation of the two desired carbon atoms. In the situation where only one of the two possible olefins i s isolated oxidative cleavage followed by a second Hofmann reaction again completes the sequence. 121-3 From the results of Ondetti and Deulofeu i t appeared feasible that the C-3 olefin 88 proposed by them to be the only unsaturated product formed by reaction with sodium hydroxide in refluxing ethanol could be isolated i n quantities large enough to degrade. On repeating their reaction on a small scale (50 mg.) two products were again observed, however i t was determined from the measured integration of the N.M.R. signals that the product ratio had altered from that reported (Figure 23), only 5-10% of the product mixture corres-ponded to o l e f i n i c material. The major non-olefinic component isolated in 60% y i e l d exhibited spectral data consistent with structure 98. The N.M.R. spectrum had a quartet at 64.85 (J = 6Hz) and a corresponding doublet at 61.53 (J = 6Hz) for the methine hydrogen and the methyl group which are substituted at the C - l 1 carbon atom (see 97 for numbering system) and are Figure 23. The N.M.R. Spectrum of the Crude Reaction Mixture (NaOH/EtOH). - 76 -adjacent to the oxygen atom. The increased deshielding effect of the oxygen compared to nitrogen resulted in the substantial downfield s h i f t of the methine quartet with respect to the corresponding quartet in the spectrum for guatambuine (25) (Figure 19., Table III). Two singlets occurring at 62.49 and 2.40 were attributed to the C-l methyl group and the dimethylamino group respectively. A multiplet centered at 63.04 was attributed to the C-3' methylene protons whereas the multiplet for the C-4' methylene protons was obscured by the singlets for the above mentioned methyl groups. The quartet (63.38, J = 7Hz) and t r i p l e t (61.18, J = 7Hz) absorptions were assigned to the methylene and methyl group respectively of the substituted ethoxy group. (98) The mass spectrum possessed a parent peak at m/e = 324 and only two major fragmentations at m/e = 295, 278 for the a-cleavage of the ether linkage in both possible directions. The high resolution mass spectrum parent peak at m/e = 324.2207 was within acceptable limits of the value 324.2200 calculated for C 2 1H 2 gN 20. For purposes of comparison with the unassigned component A isolated by 121-3 Ondetti and Deulofeu, compound 98 also possesses a normal carbazole UV. - 77 -spectrum. It is highly probable that the same product has been isolated in both instances. This assumption is strongly supported by the observation that their microanalytical data is within experimental error for both the molecular composition C 2 0 H 2 5 N 2 I proposed by them for component A and the calculated values for the composition C22H31N2OI f ° r the methiodide derivative of compound 98. It i s apparent that 98 i s formed by substitution of the quaternary nitro-gen of guatambuine methiodide by ethoxide ion generated in the reaction medium. Substitution reactions are always a competing process with elimination and as w i l l become more evident in later discussion, the benzylic center (C-l) onto 140 which the nitrogen is attached in guatambuine (25) as well as in uleine (18) 113 and apparicine (19) is considerably activated towards substitution processes. This is to a major extent a consequence of stabilization of the benzylic carbonium ion that i s formed during reaction by the lone pair of electrons on the indole nitrogen oriented para to i t . The intermediacy of such a stabilized benzylic carbonium ion implies that the substitution reaction occurs by a SNj^  type mechanism. This further implies that on formation of substitution products' the optical activity of the asym-metric benzylic center in guatambuine (25) would be lost. Such has been found 78 to be the case when the ( - ) -methiodide 86 was t r e a t e d under the s t r o n g l y b a s i c r e a c t i o n condi t ions employed by Ondet t i and Deulofeu-. I t has been demonstrated a l s o i n c l o s e l y r e l a t e d systems e x e m p l i f i e d by the a r o m a t i c a l l y oxygenated t e t r a h y d r o i s o q u i n o l i n e a l k a l o i d s such as emetine (99) (99) that the c y c l i c six-membered r i n g s t ruc ture does not undergo s u b s t i t u t i o n r e a c t i o n unless s t rong b a s i c condi t ions normally a s s o c i a t e d wx-vh- o l v ^ - i ' c o t i . i w o 139,141-3 are employed. The probable r a t i o n a l e f o r t h i s requirement i s that s u f f i c i e n t d r i v i n g force must be provided to overcome the tendency o f the ni trogenous l e a v i n g group to r e v e r t back t o the quaternary amine by i n t e r n a l r e t u r n . Considerable e f f o r t was made t o i s o l a t e the minor o l e f i n i c product from the crude mixture , however even by r e p e t i t i v e column chromatography of o l e f i n e n r i c h e d f r a c t i o n s i t was not p o s s i b l e to adequately e l i m i n a t e the presence o f the s u b s t i t u t i o n product 98 as contaminant. • I t was e v e n t u a l l y p o s s i b l e though to determinejwith a good degree o f confidence the i d e n t i t y o f the o l e f i n p r o -duced d u r i n g the r e a c t i o n . This assignment was i n f e r r e d from the N . M . R . spec-trum (Figure 23) by a comparison with the chemical s h i f t s f o r the - 79 -components of the a- and 3-dimethylaminoethyl, and vinyl side chains found in the o l e f i n i c and saturated forms of the ring-opened products of guatambuine (25) (Table III). As w i l l be alluded to again in subsequent discussion the two major dis-tinguishing features between the two possible elimination products are: the relative positions of the singlet for the protons of the dimethylamino group as i t occurs in the C-2 and C-3 side chains and i i ) the chemical shifts for the cis and trans - coupled olefinic hydrogens. It can be seen from Table III that the singlet for the dimethylamino group in the C-3 side chain of compounds 97, 103, and 98 occurs at higher f i e l d than does the singlet for the same group in the C-2 side chain of com-pounds 88 and 89. The absence therefore of a peak at <v 62.25, for the C-3 side chain, integrating for 10% of the dimethylamino group singlet at 62.40 would suggest that the elimination product possesses a 3-dimethylaminoethyl side chain at C-2 and that i t has structure 88. This assignment was confirmed as a result of the occurrence of the trans coupled hydrogen of the double bond to lowerfield (65.68, J p ^ 1 = 17Hz) t n a n the cis coupled hydrogen (65.29, J i i 2 » = 11Hz). This justaposition of the o l e f i n i c signals i s characteristic of the C-3 vinyl substituent. The N.M.R. data substantiated the production of the ring-opened olefin 88 121-3 under Hofmann conditions as predicted by Ondetti and Deulofeu. It was somewhat surprising that i t was the only unsaturated product produced by the reaction since the majority of structurally similar molecules such as the tetra-144 145 hydroisoquinolines salsolidine (100) and lophocerine (101) ring open so - 80 -as to give the alternate Hofmann product which i s predicted by the Hofmann rule to be the predominant product. (101) It would appear evident from the observed formation of the substitution product 98 with the olefin 88 as the minor side product that the benzylic carbonium ion mechanism plays a central role in the reactivity of guatambuine methiodide (86) under sodium hydroxide in ethanol conditions. The formation of the small amount of the elimination product 88 during the reaction can be envisaged to result from a competing neutralization of the carbonium ion by loss of a proton from the adjacent methyl center (Ej elimination). The lack of substantial quantities of the elimination product 88 in the reaction mixture necessitated the study of other reaction conditions to increase - 81 -its overall y i e l d or that of the alternate elimination product 97. The substitution product formed in 60% yield wasn't, but could have been used for subsequent isolations of the N-methyl group (see Section 3.) because the ether functionality would be essentially inert to the basic reaction conditions used and i t s presence therefore would not lead to any complications. To try and enhance the formation of ol e f i n i c product during the Hofmann reaction, potassium t-butoxide in t-butanol was selected as the base. The reason for this choice being that potassium t-butoxide i s a stronger base and weaker nucleophile than sodium ethoxide and i t would be less prone therefore to produce undesirable substitution products during reaction. The N.M.R. spectrum for the crude reaction mixture did indicate that the product ratios had altered significantly. A small percentage of substitution product was produced as deduced from the presence of a singlet at 6 1.28 for the hydrogens of the t-butyl group, however the predominant portion of the reaction mixture consisted of a mixture of two o l e f i n i c products (Figure 24). From the measured integration of the two singlets for the dimethylamino groups the ole f i n i c product ratio was determined to be 60:40 (compound 97:88). During i n i t i a l experiments (30 mg. scale) isolation of the components of the reaction mixture was accomplished by preparative layer chromatography. Only one product corresponding to the major spot on TLC could be isolated in yields (30%) which enabled spectral data to be obtained. From the spectral data the identity of the o l e f i n i c compound was consis-tent with the structure 97 which is the expected product on the basis of the 121-3 Hofmann rule, but which i s opposite to that reported by Ondetti and Deulofeu . Figure 24. The N.M.R. Spectrum of the Crude Reaction Mixture (KOt-Bu/t-BuQH). - 83 -N H N / \ (97) Identifying features in the N.M.R. were the doublet (61.39, J = 6Hz), quartet (63.75, J = 6Hz), and singlet (62.25 for the methyl group, methine hydrogen and dimethylamino group respectively for the side chain at C-3 of the carbazole system (Figure 25). A singlet was also present at 62.50 for the protons of the C-l methyl group. The signals for the three hydrogens of the vinyl side chain at C-2 portrayed a typical ABX system. A pair of doublet of doublets occurred in the o l e f i n i c region, one at 65.20 ( J 3 1 ^ 4 1 = 17Hz, ^ ' 1 4 1 = 2Hz) for the trans C-4' hydrogen, the other at 5.65 ^ 3 1 ^ 4 1 = 11Hz, J 4 i > 4 i = 2Hz) for the cis C-4' hydrogen. A series of four lines was also observed downfield at 66.98 ( J 3 1 41 = 17Hz, ^ 1 ^ 4 1 = 11Hz) for the C-3' hydrogen of the double bond. The difference in the magnitude of the coupling constant enabled the cis and trans hydrogens to be distinguished and i t i s of interest that the chemical sh i f t positions of these resonances in olefin 97 are reversed from that observed for styrene. This characteristic feature may be a consequence of steric or electronic interactions of the N-dimethylamino group with the vinyl system which results i n a reversal of the influence that the aromatic ring has on the two terminal hydrogen atoms. - 84 -- 85 -A parent peak was observed at m/e = 278 in the mass spectrum of 97, and major fragmentation peaks came at m/e = 263, 234, 233, 219, 218, 204. The fragmentation reactions that occur for the series of C-2 and C-3 unsaturated and saturated ring-opened products are very characteristic and for this reason a discussion of the mass spectrum w i l l be reserved t i l l later (Section 5). The parent peak at m/e = 278.1780 i n the high resolution spectrum was within acceptable limits of the calculated value m/e = 278.1782 for the composition C ^ r ^ ^ . The interaction of the vinyl system with the carbazole back bone caused only a slight modification of the normal carbazole UV. spectrum. There was a bathochromic s h i f t of 10mm to A m a x247 for the predominant absorption. The spectrum obtained for olefin 97 bore no resemblance to 121-3 the highly modified carbazole UV. spectrum reported for olefin 88. Prior to the eventual isolation of the second elimination product from the reaction by column chromatography i t s identity was also deduced by i n -ference from the N.M.R. spectrum of the crude reaction (Figure 24). As previously discussed, i t has been shown from the data presented in Table III that the two different nitrogen containing side chains for the ring-opened products could be distinguished by the position of the chemical s h i f t for the respective dimethylamino group. Two singlets for the dimethylamino group of both side chains are present in the spectrum of the crude reaction mixture. The singlet at 62.32 corres-ponded to the C-3 side chain of the major o l e f i n i c product 97 isolated from Table III: Characteristic N.M.R. Chemical Shifts for the Degradation Products of Guatambuine (25) Structure: Chemical Shift: l a lb l c Id 6C-1 CH3 6N(CH 3) 2 6C-1'H SC-l'O^ 6 = CH 2(trans) 6 = CH 2(cis) (25) 2.54 f 2.40 3.90 1.52 (97) 2.50 2.25 3.75 1.39 5.20 5.65 (103) 2.49 2.49 < (88) 2.50 2.29 3.62 2.40 4.85 2.38 1.42 1.53 5.68 5.29 00 ON N ( (89) 2.64 le 2.54 (89) from LAH on (25) 2.52 2.46 (89) from (18) degradation 2.48 2.38 la . (q,J = 6Hz), b. (d,J = 6Hz), c. ( J ^ - 17 Hz,J B C = -2.0Hz) d. ( J x c = 11 Hz) e. (FTl N.M.R.. verv low concentration. f in guatambuine This carbon i s C-_5_ - 87 -the reaction mixture. It was very probable therefore from the correlations that the remaining singlet peak (62.40) corresponded to the C-2 side chain present in olefin 88. A pair of doublet of doublets in the o l e f i n i c region (65.70 (J = 17Hz, 2Hz); 65.32 (J = 11Hz, 2Hz)) could be distinguished from these attributed to olefin 97. From the measured coupling constants i t was determined that the chemical shifts of the cis and trans-oriented terminal hydrogens coincided with those for styrene but were reversed with respect to the resonances for the elimination product 97. This would be expected for compound 88 for no s t e r i c or electronic interaction would be anticipated between the C-3 vinyl group and the dimethylamino group. It was not possible to conclusively prove the structure of the unisolated olefin from the N.M.R. spectrum of the crude mixture, however the data are consistent with the postulated structure 88. During the i n i t i a l experiments i t was thought that only the predominant Hofmann product 97 could be isolated i n adequate y i e l d and that i t would not be possible to isolate the C-3 o l e f i n i c product 88 directly by the Hofmann degradation approach. It became necessary in order to determine any activity at the C-l methyl of guatambuine (25) in subsequent incorporation experiments to find an alternate method which would generate the C-3 vinyl side chain in a ring-opened compound. To this, end the C-2 olefin 97 became of particular importance for not only could i t be oxidatively degraded so as to isolate the C-3 carbon atom of guatambuine (25), but i t s potential second Hofmann elimination product would contain the desired C-3 (carbazole numbering) vinyl side chain. The - 88 -- 89 -development of this latter chemistry and the study of a number of alternate methods aimed at producing a ring-opened compound with the C-3 vinyl side chain comprises a considerable portion of the work on the degradation sequence described in subsequent sections. It has been found however from further work that by purifying the crude reaction mixture by column chromatography on alumina that i t became possible to isolate the C-3 olefin 88 in 18% y i e l d (olefin 97 was isolated in 36% y i e l d ) . (88) The N.M.R. spectrum (Figure 2 6 . ) as anticipated possessed a singlet at 6 2 . 3 8 which was in close proximity ( A 6 0 . 1 2 ) to the singlet for the C-l methyl group ( 6 2 . 5 0 ) . A multiplet centered at 6 3 . 0 8 integrated for the benzylic methylene protons of the 3-dimethylamino side chain while the multiplet for the remaining two protons was buried beneath the two singlet absorptions for the respective methyl groups. Two pairs of multiplets (doublets of doublets) occurred i n the o l e f i n i c region for the terminal C - 2 1 hydrogen atoms. The shif t of the trans coupled C - 2 ' hydrogen 6 5 . 6 8 ( J i » , 2 » = 1 7 H z > J 2 ' , 2 " = 2 . 0 Hz) to lower f i e l d with respect to the one which i s coupled in a cis fashion 6 5 . 2 9 ( J j , 2 i = 11 H z > 2 ' = H z ^ W a S s n o w n t o ^ e characteristic for the C-3 o l e f i n i c side chain as suspected from the N.M.R. spectrum of the crude -90-reaction mixture. The mass spectrum of 88 possessed a parent peak at m/e = 278 with major fragmentations found at m/e = 263, 249, 234, 233, 220, 219-8, 204 (see section 5 f o r a d e t a i l e d d i s c u s s i o n ) . The high r e s o l u t i o n spectrum possessed a parent peak at 278.1780 which was consistent with the composition ^1^22^2 ^ o r c o m P o u n c i 88. The UV. spectrum o f 88 bore no resemblance to e i t h e r the normal carbazole spectrum or the spectrum obtained f o r the C-2 o l e f i n 97 (Figure 27). The gross d i s s i m i l a r i t i e s i n the UV. spectra of 88 and 97 and the close coincidence of the values obtained i n the spectrum of 88 with the 121-3 reported values by Ondetti and Deulofeu proved c o n c l u s i v e l y that they had proposed the c o r r e c t structure f o r the e l i m i n a t i o n product i s o l a t e d i n t h e i r study. It would have been d i f f i c u l t to have predicted beforehand e i t h e r the shape of the absorption patterns f o r the two e l i m i n a t i o n products 88 and 97 or the f a c t that the d i f f e r e n c e i n C-2, C-3 s u b s t i t u t i o n patterns would r e s u l t i n completely d i f f e r e n t spectra f o r the two compounds. Now that these data are a v a i l a b l e i t i s p o s s i b l e to consider the i n t e r a c t i o n s of the v i n y l groups i n the two p o s i t i o n s with respect to the carbazole r i n g system and postulate that the d i f f e r e n c e s may be due to the d i f f e r e n t resonance contributions i n the excited states of 88 and 97 (as shown). 1 I 1 1 1 1 220 250 300 350 400 Figure 27. A Comparison of the UV. Spectra for the Olefins 88 and 97 with the Normal Carbazole Spectrum of 98. - 92 -N < (88) < (97) Further confirmation for the production of olefin 88 was provided by reducing i t to i t s saturated derivative 89, and comparing i t s spectral data with 89 obtained both by uleine degradation and lithium aluminum hydride ring opening of guatambuine methiodide (86) (Section 3). The product distribution (;2:1 by weight) for the ring opening of guatambuine methiodide (86) using potassium t-butoxide in t-butanol reflects the enhancement of the E 2 elimination mechanism over the mechanism observed with the previous conditions. The isolation of the C-3 olefin 88 directly from the Hofmann reaction was an important result because as discussed in later sections i t i s ques-tionable whether or not i t would be feasible to isolate the C-l methyl group of guatambuine (25) by the alternate routes developed. However, even the < (89) - 93 -yi e l d of 88 obtained under these conditions (18%) becomes of marginal prac-t i c a l i t y when the reaction is repeated on a projected plant isolation scale (30-40 mg.). To see i f the product distribution could be shifted or the overall y i e l d of the two elimination products could be increased, hot aqueous hydroxide was studied as the basic medium. It was found however that the product ratio did not alter significantly (N.M. R, integration) and that the overall reaction y i e l d was lower (60%) . The use of sodium hydride as the base in hot dimethylformamide was considered in the hope that by using a very powerful base the relative acidi-ties of the B-hydrogens i n guatambuine might become the sole determining fac-tor i n the reaction and as a consequence the product ratio might be shifted completely towards the C-2 olefin 97. It was found quite unexpectedly after reaction, however, that exactly the opposite had occurred. The C-3 olefin 88 obtained in 70% y i e l d was the only reaction product isolated. The spectral data was identical to that obtained for 88 above using potassium t-butoxide as the base. In terms of the mechanistic rationale i t was not entirely clear why olefin 88 was the exclusive reaction product, however, the practical aspect of this result was that the C-3 olefin 88 became available in adequate quan-t i t i e s such that subsequent oxidative cleavage (see Section 3) enabled the C-l methyl groups of guatambuine to be unambiguously isolated. The application of the Hofmann elimination reaction to the degradation of N-methyltetrahydroellipticine methiodide (95) was straight forward for there i s only one direction in which B-elimination can occur. - 94 -< (95) (102) To reduce the occurrence of the competing substitution reaction the reaction was conducted using potassium t-butoxide in refluxing t-butanol. A single product was formed during the reaction. (TLC) in 96% y i e l d and the spectral data was consistent with the o l e f i n i c product 102. The N.M. R, possessed two singlets at 62.50 and 2.90 for the protons of the aromatically substituted methyl groups at C-l and C-4 respectively. A third singlet occurred at 62.24 for the dimethylamino group. As in the guatambuine (25) series the proximity of this signal reflected the presence of the dimethylamino group on the C-3 position. Another singlet (63.64) was attributed to the methylene protons both benzylic (C-l) and adjacent to the nitrogen atom. Two pairs of multiplets (doublets of doublets) came at 65.20 (J2',3' = 1 7 H z » J3«,3» = 2 H z ) a n d 6 5 - 6 4 (J2',3' = 1 1 H z> J3',3' = 2 H z ) and once again the upfield position of the trans coupled hydrogen with respect to the cis hydrogen was characteristic of the C-2 vinyl side chain. The mass spectrum possessed a parent peak at m/e = 278 and major fragmen-tations at 263, 234, 233, 219, 218, 204. The high resolution spectrum possessed a parent peak at 278.1780 which was within acceptable limits of the value 278.1782 calculated for the composition C j g r ^ ^ . - 95 -The UV . spectrum bore a distinct resemblance to that obtained for carbazole, this behavior again being characteristic of the presence of the C-2 vinyl side chain in the elimination product. The oxidative degradation of the Hofmann product i s discussed in section 3. - 96 -3. N-Methyl, C-3 Methylene and C-l Methyl Group Isolations. Due to the scale of the biosynthetic experiments, i t was accepted that for the guatambuine system (25) a l l three carbon centers could not be isolated from the same experiment. With the Hofmann elimination reaction developed to the point where the ring-opened olefins 88 and 97 from guatambuine (25) and 102 from N-methyltetrahydroellipticine (26) were available in adequate yields, the two carbon centers (C-l and C-3) could then be isolated individually by ozonolytic cleavage. Determination of the N-methyl group would then be a separate experi-ment involving a Hofmann reaction from the methine methiodides of any of the above olefins or from their saturated derivatives 89, 103 or 104. (89) (103) (104) To gain the most information from the biosynthetic experiments u t i l i z i n g the Hofmann reaction chemistry, the isolation experiments were divided into two catagories. In the f i r s t , the C-l methyl group was unique to guatambuine (25) and i t s isolation v i a the ozonolytic cleavage of C-3 olefin 88 was considered as a separate experiment. In the second, the N-methyl and C-3 methylene groups were common to both 25 and N-methyltetrahydroellipticine (26) and efforts were therefore directed at developing a combined experiment whereby both centers could be isolated sequentially. - 97 -Before attempting the ozonolysis of the valuable C-3 olef i n 88, the reaction conditions were worked out for compound 102 derived from N-methyltetrahydroellipt 146 cine (26). The procedure u t i l i z e d by Battersby and Harper was adopted, being modified only in that methylene chloride was used as the solvent. The formalde-hyde produced (C-3 carbon) on reductive work-up was isolated as the bisdimedone 2/0 derivative by steam d i s t i l l a t i o n of the reaction mixture into a saturated dimedone solution. (102) (105) + CHO 2 No attempt was made to isolate the carbazole aldehyde residue 105 and i t was thus not known whether ozone also attacked the benzylic dimethylamino group. The formaldehyde dimedone was isolated as long crystalline needles in 87% yi e l d . The ease with which the C-3 carbon could be isolated by ozonolysis indicated that the technique would be generally applicable within the series of C-2 and C-3 vinyl carbazole derivatives. C-3 vinyl system of Olefin 88 was similarly readily cleaved on ozonolysis. The formaldehyde dimedone (C-l methyl) was isolated as long needles in 60% yi e l d . Again, no attempt was made to isolate the aldehyde 106 also proposed to be formed during the reaction. This then completed the degradation of guatambuine so as to isolate the C-l methyl group. - 98 -(88) (106) For the isolation of the remaining two centers in 25 and 26, i t was con-sidered that the most eff i c i e n t order i n which the degradation could be conducted would be by sequential Hofmann reaction which would eliminate the N-methyl group and produce the C-2 vinyl system that would subsequently be cleaved oxidatively. ^ + (CH^N — \ C H Q + CHO By conducting the reaction in the reverse sequence i.e. oxidative cleavage, f i r s t i t was thought that problems would arise, for i t was known that in general, 145-6 the techniques used for isolating the one carbon unit as the dimedone derivative lead to poor recovery of the remainder of the molecule. It was necessary to u t i l i z e the B-dimethylaminoethyl side chain for this degradation because as w i l l become clearer i n Section 4, considerable d i f f i c u l t i e s arise during methiodide formation of compound 103 containing the benzylic dimethyl-amino group. To produce a suitable derivative i n both the guatambuine (25) and N-Methyl-tetrahydroellipticine (26) advantage was taken of the reactivity of the C-l ben-140 zylic center towards nucleophiles. It has been shown that in both uleine and 113 apparicine that reaction of the quaternary salts with lithium aluminum hydride - 99 -effected displacement of the nitrogen center by hydride. In a s i m i l a r manner i t was found that the methiodides of 25 and 26 were attacked by l i t h i u m aluminum hydride to give i n good y i e l d s compound 89 and 107. Compound 89 was prepared also for the purpose of comparing i t s physical and 132 spectral properties with those for the degradation product of uleine and with the hydrogenated Hofmann product of O l e f i n 88 prepared both i n the present work 121-3 and previously (Table I I I ) . The N.M.R. spectrum of compound 89 possessed a m u l t i p l e t 53.00, and a t r i p l e t (61.36, J = 7 Hz) f o r the newly formed ethyl side chain. Present also were the s i n g l e t s at 62.52 and 2.46 for the C-l methyl and dimethylamino groups respectively. As mentioned previously the chemical s h i f t f o r the l a t t e r group was c h a r a c t e r i s t i c for the presence of the B-dimethylaminoethyl side chain at the C-2 p o s i t i o n . The mass spectrum possessed a parent peak at m/e = 280 and major fragments at m/e = 266, 236, 222, 207, 206, 204 (Section 5). The high res o l u t i o n mass spectrum parent peak at m/e = 280.1975 was consistent with that calculated (m/e = 280.1938) fo r the composition C 1 Q H 2 4 N 2 . For further comparison and proof of structure 89 the methiodide of uleine 87 120,132 was ring-opened under Hofmann conditions to give the compound 89. The spectral data f o r 89 prepared by the two d i f f e r e n t routes were almost i d e n t i c a l . An added feature to conversion of 87 •*• 89 was that the subsequent i s o l a t i o n of the N-methyl and C-3 methylene groups from compound 89 constituted a degradation of - 100 -both uleine (18) and guatambuine (25). The N.M.R. spectrum of compound 107 exhibited three singlets at 62.75, 2.42, and 2.36 for the C-4, C-l and C-3 methyl groups respectively. A fourth singlet was present at 62.32 for the dimethylamino group. The mass spectrum possessed a parent peak at m/e = 280 with major fragments at m/e = 266, 249, 236, 222, 207-6, 204, 191 (Section 5). The parent peak in the high resolution mass spectrum m/e = 280.1903 was again consistent with the calculated value m/e = 280.1938 for C-^rL^N^ Both compounds 89 and 107 were converted to their methiodides 90 and 108 in quantitative yields in preparation for the Hofmann reaction. The procedure used for isolating the N-methyl group w i l l be described below, at the present time attention w i l l be focussed on the formation of the olefins 91 and 109. (90) (91) + N(CH 3) 4 (110) 1 V H 1 +N(CH3)4 (108) (109) H (111) The N-methyltetrahydroellipticine derived methiodide 108 was reacted with potassium t-butoxide in t-butanol in the same manner as described i n Section 2, The desired olefin 109 was obtained as a brown o i l in 85% yie l d . The N.M.R. spectrum for the product showed a doublet of doublets at 65.68 (Jj, 4 , = 12 Hz, J^, 4 , = 2 Hz) for the cis C-4H' and a complimentary doublet - 101 -of doublets at 65.24 ( J3 ' ,4•= 1 8 H z » J4',4« = 2 H z ) f o r t h e trans C-4*H. The singlets for the three methyl groups occurred at 62.84 (C-4), 2.52 (C-l), and 2.40 (C-3). The mass spectrum possessed a parent peak at m/e = 235 with main peaks at m/e = 220, and 205 (Section 5). The high resolution spectrum possessed a parent peak at m/e = 235.1325 which was within acceptable limits for that calculated (m/e = 235,1360) for C 1 7H 1 7N. The reaction of the methiodide 90 derived from the guatambuine system with potassium t-butoxide in t-butanol did not proceed as well as anticipated, a dark brown product containing considerable quantities of decomposition material was obtained. The ol e f i n 91 was obtained as a brown o i l i n 57% yi e l d after purification by preparative layer chromatography on alumina. A cleaner product was obtained by reaction of the methiodide 90 with sodium hydride in dimethylformamide. However in this instance as well, the product yield was low (40%). The N.M.R. spectrum for this product (of poor quality) showed the pair of doublets of doublets, 65.65 ( J 3 , 4 , = 12 Hz, J 4 , 4 , = 2 Hz) for the cis_ C-4' H and 65.29 (J31 41 = 18 Hz, J 4 1 J 4 I = 2 Hz) for the trans C-4' H. The singlet for the C-l methyl group was readily discerned at 62.54 but the signals for ethyl groups occurred as complex multiplets due to the presence of impurities. The mass spectrum for 91 possessed a parent peak at m/e = 235 and major fragments at m/e = 220 and 205. The high resolution spectrum possessed a parent peak at m/e = 235.1384 within acceptable limits of the value 235.1360 calculated for CjyHjyN. The N-methyl group of guatambuine (25) and N-methyltetrahydroellipticine (26) was isolated during the Hofmann reactions by converting the trimethyla-mine that was formed to i t s tetramethylammonium iodide derivative. This was - 102 -done by passing the gases produced during the reaction through a solution of methyl iodide. To determine the efficiency of the trapping technique, a mock/biosynthetic 1 experiment was conducted where (N-C^ methyl) guatambuine (25) and N-Methyltetra-hydroellipticine (26) were converted to their ring-opened methiodides 90 and 108, and subjected to the Hofmann reaction. A total of 51 % and 72% of the a c t i v i t y present in 25 and 26 was recovered as the tetramethylammonium iodide s a l t . These results demonstrated that the N-methyl group could be readily isolated and that the isolation technique would be suitable to future biosynthetic experiments. The subsequent ozonolyses of olefins 91 and 109 so as to isolate the C-3 methylene group of 25 and 26 was unfortunately not investigated at this time. In order for the ozonolysis reaction to be feasible i n the guatambuine series how-ever, i t would be f i r s t necessary to further develop the Hofmann reaction of 108 so as to obtain a higher yield of the o l e f i n 109. Considering the success of the ozonolysis reaction in the cleavage of the C-3 vinyl side chain of olefins 88 and 102, i t would be anticipated that no major d i f f i c u l t i e s would be encoun-tered during ozonolysis of 91 and 109. An overall view of the degradation scheme for the olivacine (16) system involving the Hofmann reaction followed by ozonolytic cleavage reactions i s ? presented in below. Considerable work was also done to develop alternate routes whereby the N-methyl, C-l methyl and C-3 methylene groups could be isolated. This work i s presented in subsequent sections. 103 -- 104 -4. The Synthesis of l-Methyl-2-ethyl-3 vinylcarbazole (114) from the C-2  Olefin 97. To reiterate for a moment, the synthetic schemes described in the following sections were aimed at devising a feasible alternative to the direct Hofmann reaction of guatambuine methiodide (86) for creating a C-3 vinyl carbazole derivative for the subsequent degradative isolation of the C-l methyl carbon of guatambuine (25). This work was prompted by the i n i t i a l b e l i e f that the C-3 olefin 88 would be unavailable from the potassium t-butoxide-t-butanol reactions. Two different approaches were developed, an alternate ring opening procedure (Section 6) and the second Hofmann reaction approach starting from the relatively abundant C-2 olefin 97. This latter approach to be discussed in this section is il l u s t r a t e d below. (112) (111) In preparation for the second Hofmann reaction the C-2 vinyl side chain of 97 was reduced to compound 103 in near quantitative y i e l d by - 105 -catalytic hydrogenation over platinum oxide. This compound like i t s saturated Hofmann counterpart 89 was also prepared as an aid for interpreting the N.M.R. of the crude Hofmann elimination reaction mix-tures (Table III, section 2.). The N.M. R. spectrum for the reduced compound 103 showed a t r i p l e t (61.20, J = 6 Hz) and a quartet (62.94, J = 6 Hz) for the methyl and methylene hydrogens respectively of the newly formed C-2 ethyl side chain. Present also was a doublet (61.42, J = 6 Hz) and a quartet (63.62, J = 6 Hz) for the C-l' methyl group and the methine hydrogen substituted both benzylic and adjacent to nitrogen in the C-3 side chain. Two singlets occurred at 61.48 and 1.28 for the C-l methyl and dimethylamino group, the position of the later absorption being characteristic for the C-3 nitrogen bearing side chain. The mass spectrum possessed a parent peak at m/e = 280 and major fragmentations occurred at m/e = 265, 236, 235, 220, 207-5, 204 (see dis-cussion section 5). Compound 103 exhibited a normal carbazole UV . spectrum as expected. L i t t l e information was obtained from the I P. spectrum other than i t resembled very closely the spectrum obtained for the saturated component 89. The subsequent formation of the methiodide 111 in pure form on a small scale proved to be extremely d i f f i c u l t . Reaction of 103 with methyl iodide as a neat solution or in a variety of solvents such as ether, chloroform, and methanol resulted in the co-formation of several non-polar products. This was a consequence of the i n s t a b i l i t y of the newly formed quaternary nitrogen - 106 -with respect to nucleophilic displacement, which is inherent to i t s 147 gramine like structure. Purification of the methiodide 111 by column chromatography was not attempted as i t was thought that the molecule would decompose. Re c r y s t a l l i -zation from polar solvents like methanol was unsuccessful since TLC showed that with increasing time in methanol the concentration of non-polar side product increased. The methiodide was sufficiently purified for subsequent reaction by simply washing the concentrated reaction product with chloroform which removed the soluble non-polar materials and l e f t behind the insoluble methiodide 111. In the case where the methiodide 111 was formed by reaction with methyl iodide in methanol i t was shown by N.M. R. that the major non-polar com-ponent corresponded as anticipated to the 0-methyl ether 113, formed by 143, 147 displacement. (113) Unfortunately the i n a b i l i t y to obtain 111 in pure form made i t impossible 121-3 to compare i t s physical properties with those previously reported. - 107 -The choice of reaction conditions for the subsequent elimination reaction to 112 had to take into account the labile nature of the quaternary nitrogen group, substituted at the benzylic C - l ' center, towards displacement reaction. It has been shown by Norcross and 143 Openshaw in a study of the cyclic and acyclic methiodides of model compounds related to the emitene system (99) that in the acyclic case exemplified by 114, the benzylically substituted quaternary amino group, which has an electron donating group oriented para to i t , possesses an exceptional reactivity towards nucleophilic substitution. This reactivity was attributed to the a b i l i t y of a stabilized carbonium ion to be formed in the transition state of the reaction (SNj mechanism). - 108 -The close parallel between these systems and the methiodide 111 derived from guatambuine (25) was s e l f evident. It was apparent from 148 this and other studies that the basic aqueous or alcoholic condi-tions normally employed for the Hofmann degradation would f a i l to pro-duce the desired o l e f i n i c product 112 . To circumvent this same problem in the emetine (99) series, Openshaw 142-3 et a l . found that pyrolysis of the acyclic methiodides in an aprotic solvent like diethyl ketone resulted in olefin formation. It was also observed, however, that in accord with the carbonium ion mechanism, the y i e l d of the reaction depended markedly on the mesomeric influence of the alkyl substituent ( R) on the side chain. As the size of the R group decreased, so did the yield and i t could be anticipated from this trend that the side chain where R = H indiginous to 111 may result in a poor yie l d of o l e f i n 112. With this possibility i n mind, i t was decided to study the f e a s i b i l i t y of the pyrolytic elimination reaction on a model compound (117) derived from salsolidine. methiodide (115) instead of using valuable quantities of H I . (115) (118) - 109 -It was found that prolonged heating of 117 in diethyl ketone resulted only in the isolation of small amounts of unidentified materials. None of the desired olefin 118 was detected, and this approach was consequently abandoned. The C-3 olefin 112 was eventually obtained in 70% yield from 111 by reaction with sodium hydride as the base in dimethylformamide. This reaction had previously proven successful i n ring opening guatambuine methiodide (86) to the analogous ol e f i n 88. To avoid confusion at this point i t must be mentioned that the successful completion of the sequence to the C-3 vinylcarbazole 112 (the compound that was supposed to take the place of 88 in the degradation sequence) came after the conditions were found to obtain 88 directly. The spectral data for the ole f i n i c product was consistent with i t s structure 112 . The presence of the vinyl grouping at C-3 was readily (112) determined by i t s characteristic UV . spectrum and the position of the jolefinic absorptions (5.69 (Jj i 2 i = I 7 H z » ^2' 2' = 2 , ( 5 5.28(Jp 2»= 11 H J 2 , 2, = 2 Hz) ) in the N.M.R. spectrum. - 110 -The mass spectrum possessed a parent peak at m/e = 235 and major fragmentations at m/e = 220 and 205. The parent peak in the high resolu-tion spectrum 235.1380 was consistent with the composition CjyH^yN. The considerable d i f f i c u l t y i n preparing the methiodide 113 makes the sequence to compound 112 unreliable and low yielding in terms of the quantities of olefin obtained. It would be doubtful i f sufficient quantities of formaldehyde (as i t s dimedone derivative) would be obtained from the oxidation of 112 to substantiate any incorporation experiments by this route. It was for this reason that the work presented in section 6 was init i a t e d . - I l l -5. Mass Spectral Correlations of Ring "D" Opened Derivatives of Guatambuine (25) and E l l i p t i c i n e (17).  A discussion of the mass spectra of the o l e f i n i c (88, 97, 102) and reduced (89, 103, 107) ring-opened products of guatambuine (25) and e l l i p t i c i n e (17) as well as their second Hofmann products (91, 112, 109) has been reserved for a separate section because close similarities existed in the fragmentation patterns of these derivatives. It appears as though the occurrence of several characteristic fragments in the spectra i s internally consistent with their t r i - and tetra- substituted carbazole structures. Characteristic features i n the spectra of the ring opened products are a substantial parent peak either at m/e = 278 (unsaturated) or m/e = 280 (saturated) and a fragmentation pattern consisting of successive losses of 14 and 15 mass units which corresponds to the formal loss of successive methylene and methyl groups (Figure 28 $ 29). These characteristic fragments occur at m/e = ^ 2 8 0 ) , 265, 236, 235, 222, 220, 207, 205 and m/e =(M+:278), 263, 249, 234, 233, 220, 219, 218, 205, 204. Weaker fragments are also observed in the region m/e = 191-193, 180, and 167 for further losses of 13-15 mass units. The m/e = 167 fragment presumably represents the carbazole ion. Except in the spectra of the o l e f i n i c compounds the peak at m/e = 249 corresponding to loss of 29 mass units occurred only to a minor extent, and with the exception of the compounds 97 and 103 with the a-dimethylaminoethyl side chain the ^-15 peak was also not intense. In the latter instance, the presence of a substantial M*-15 i s probably a consequence of a f a c i l e cleavage - i ' 2 -JJl L L 4J 200 210 220 230 240 2S0 260 270 210 F i g u r e 28. Mass Spectrum of Carbazole Compound (]03) l » ° . » ° » ° 230 240 250 260 270 in F i g u r e 29. Mass Spectrum of Carbazole Compound (88). •Ill -11111 . .HI. 200 210 220 230 240 Figure 30. Mass Spectrum of Carbazole Compound ('09), - 113 -of the C - l 1 methyl group so as to form an even electron system (Figure 28). This same cleavage is an important fragment in the mass spectrum of guatam-buine (25) and is only of minor occurrence in the spectrum of N-methyltetra-hydroellipticine (26). The f i r s t major fragment after the parent occurs at m/e = 266, 265 (264, 263) for loss of 44 and 45 mass units which corresponds to the cleavage of the dimethylamino group plus one hydrogen (equivalent to three methyl groups). Plausible fragmentation pathways for compounds 97 and 88 in the guatam-buine series which are consistent with the presence of the observed fragments and the compositions for these fragments determined from the high resolution spectrum are presented in figures 31 § 32. Pathways involving similar pro-cesses can be formulated for their isomeric ring-opened counterparts 103 and 89. It should be emphasized however that in the absence of any selective labelling and other experiments to validate the structures of the fragment ions the proposed structures and fragmentation patterns are purely hypothetical. Several comments can be made concerning the postulated breakdown of compound 103 (M+, m/e = 280). Only a single metastable at m/e = 206 corres-ponding to the fragmentation 235 •*• 220 was evident in i t s spectrum. Considerably more information was obtained from the spectrum of i t s unsaturated analog 97 (M*, m/e = 278) where metastables were present for the analogous fragmentations: M1", m/e = 234 for m/e = 263 248 205 234 -»- 219 191 218 204 184 263 + 220 - 114 -(C 1 6H 1 4N) 220 Figure 31. Plausible Mass Spectral Fragmentation Pattern for Compound 103. - 115 -263 (C 1 5H UN) 205 Figure 32. Plausable Mass Spectral Fragmentation Pattern for Compound 88. (C 1 6H 1 2N) 218 Figure 33. Plausable Mass Spectral Fragmentation Pattern for Compound 109. - 116 -The fragmentation m/e = 220 •> 205 involved cleavage of the C-l methyl from the aromatic ring. Such processes are well known, and i t i s currently believed that the species obtained is not a phenyl type radical but a ring 149 opened species containing an acetylenic linkage. Cleavage of the vinyl side chain to liberate an acetylenic system was not postulated since i t s involvement in the fragmentation processes would lead to a more complex spectrum than was observed. Precedent for the formulation of the structures of a number of the fragments was available from a detailed study of the mass spectrum of 140 uleine (18) and i t s derivatives. Uleine fragments in an expected manner to the ring-opened trisubstituted carbazole derivative 119 which i s very similar to compound 89. The mass spectrum of this synthesized compound exhibits many of the fragments observed for the guatambuine (25) derived components. (119) The fragmentation processes postulated for the elimination product 88 are straightforward, being governed by the a- and 8- cleavage of the C-2 side chain. On the other hand the fragmentation processes occurring for the - 117 -N-methyltetrahydroellipticine derived compounds 102 and 107 are more d i f f i -cult to postulate structures for because the fragmentation process involves successive losses of methyl groups from the carbazole ring. The second Hofmann products (91 , 112 and 109 ) possess structures which potentially correspond to the postulated fragments at m/e = 235 for the f i r s t Hofmann products. It i s not too surprising, therefore, that their spectra also possesses dominant fragments at ^ -15 (m/e = 220) and M^ -30 (m/e = 205) (Figure 30). This does not necessitate however that the fragmentation pro-cesses are the same in both cases. In the spectrum of 109 metastables are present at m/e = 206 for the frag-mentation m/e = 235 220 and m/e = 191 for the process m/e = 220 205. The second metastable i s very weak and the later fragmentation process may not be of major importance (Figure 33) . The general features of the mass spectra of these compounds are also found in the spectra of many of the other C-2 and C-3 substituted 1-methylcarbazole derivatives discussed in this thesis. - 118 -6. Alternate Rj ng " D' Opening Re actions of Guatambuine (25). This section is concerned with the work that was directed toward ring-opening guatambuine (25) in a manner which would spe c i f i c a l l y cleave the C-l carbon-nitrogen bond and enable the C-l methyl group to be isolated. Two approaches to the problem were considered : i) the direct opening of the "D" ring by pyrolytic elimination to give the C-3 vinyl system, and i i ) the development of substitution and oxidation reactions which would introduce an oxygen bearing functionality into the C-l' position (carbazole numbering). This functionality could subsequently be elaborated into a derivative suitable for cleavage by one of three routes depicted in figure 34 . Of these different routes the dehydration-ozonolysis was the most suitable 118 because i t would be anticipated from the Potier-Janot postulate that radioactive stemmadenine (13) would label the C-l methyl group with tritium only. The Baeyer-V i l l i ger-acetate hydrolysis approach would permit the isolation of the total Figure 34. Approaches to the Isolation of the C-l Methyl Group Involving a C-l' Oxygen Functionality. - 119 -specific activity, however, considerable d i f f i c u l t i e s would be associated with isolating the acetate salt. The Haloform reaction would be the least useful since a l l three of the radioactive hydrogens would be lost. A. Pyrolytic Elimination ; The pyrolysis of guatambuine methiodide (86) as with the base induced Hofmann reaction can lead to formation of either/or both olefins 88 and 97. It was hoped, however, that pyrolytic cleavage would lead to a preference for the production of the C-3 olefin 88. It was found that pyrolysis of guatam-buine methiodide (86) in diethyl ketone at 100° did not induce cleavage of the quaternized tetrahydropyridine ring system. This result parrallels the obser-vation by Norcross and Openshaw 143 that the 2-phenylpiperidinium ring system of 120 also failed to open under these conditions. This they attributed to the s t a b i l i t y of the six membered ring since the anologeous ring-opened com-pound!^ cleaved readily on pyrolysis in diethyl ketone. (120) Pyrolysis of the methiodide 86 in the absence of solvent at 200° also f a i l e d to induce ol e f i n formation. It was determined from the N.M.R. spectrum that instead of ring-opening under these more drastic conditions, the nitrogen dequaternized with reformation of guatambuine (25). This was not a total l y unexpected result since some dequaternization generally occurs during Hofmann - 120 -139 elimination, and dequaternization of methoacetates by pyrolysis is a 150 known synthetically applicable procedure. The pyrolysis of the quaternary ammonium hydroxide (classical Hofmann reaction) was not tried although i t was probably that elimination would have occurred under these conditions. This was due to technical d i f f i c u l t i e s that were encountered during the pyrolysis of the methiodide 86. Very l i t t l e product was formed and i t was always contaminated by considerable quantities of decomposition products. Total selectivity towards formation of the C-3 olefin could have been 151 achieved by pyrolysis of the amine oxide of guatambuine (25). This reaction would have been limited to elimination in the direction of the C-l methyl group because unlike the 8-hydrogens in the six membered ring, the hydrogens of the methyl group are able to adopt the cis-coplanar orientation in the transition state, a condition which i s necessary for elimination. The reaction has the major drawback however, in that elimination can occur from the cis (a,e) configuration only, this limits the maximum yie l d to 50%. For this reason and due to the d i f f i c u l t i e s previously encountered in the pyrolysis technique the amine oxide reaction was not attempted and the pyrolysis approach to the ring-opening of guatambuine was abandoned. Acetate Substitution Ring-Opening : The susceptibility of the C-l benzylic center of guatambuine towards attack by a nucleophile provided an opportunity to introduce an oxygen functionality into this position. From a consideration of the possible degradation routes - 121 -presented in figure 34 the introduction of a hydroxyl group to give the alcohol derivative became the goal. This could not be achieved directly because i t was known that reaction of guatambuine methiodide (86) with aqueous sodium hydroxide resulted in elimination rather than substitution, and solvolysis with water failed to effect reaction of the six membered ring. It has been shown, however, that reaction of the closely related gramine system under acetylation conditions 147 resulted in a facile displacement of the tertiary nitrogen by acetate. Application of this reaction to the cleavage of guatambuine (25) followed by subsequent hydrolysis of the acetate presented an entirely feasible route to the formation of the desired a-hydroxyethyl side chain. Due to the limited resources of guatambuine, this sequence and the subsequent dehydration were developed beforehand using N-methyltetrahydro-e l l i p t i c i n e (26) and l-methyl-3-(a-acetoxyethyl)-9-benzylcarbazole £23 ) (available from the synthesis presented in sequence B, part III) as model systems f i g u r e 35). The acetylation ring-opening conditions were worked out using compound 26 and the conditions necessary for the hydrolysis were developed using both compounds 26 and 223. The dehydration reaction was developed using the carbazole alcohol 222 . 140 Joule and Djerassi observed that when uleine was treated with acetic anhydride in pyridine, i t was pyridine and not acetate that displaced the i n i -t i a l l y formed N +(b)-acyl ion. This somewhat surprising result was probably unique to the uleine system because treatment of N-methyltetrahydroellipticine - 122 -Ac 20 NaOAc / 7 t-BuOH/aq.NaOM (26) (121) / O (122) OAc OH (223) (222) (123) Figure 35. Acetate Substitution Approach Studied on Model Compounds, under the same conditions led to the formation of the expected N-acetyl acetate 121. Singlets i n the N.M.R. spectrum for the acetyl methyl groups were readily discernable i n the region of 62.1 as was the characteristic downfield position of the C-l' methylene singlet. The crude product mixture was dark in colour which was typical for acetylation reaction in refluxing pyridine. A cleaner product mixture which spontaneously crystallized in benzene was obtained by reaction with acetic anhydride and sodium acetate (76% yield ) . The spectral data for the acetate formed under these conditions was identical to that obtained by reaction in pyridine. The N.M. R. spectrum of 121 was considerably more complex than anticipated (Figure 36). A mixture of two separate conformers of 121 (a,b) (about 2:1 by Figure 36. N.M.R. Spectrum of 1,4- Dimethyl-2-($-(N,N-methylacetylamino)ethyJ)-3-acetoxymethylcarbazole (121). - 124 -integration) could be distinguished which most probably arose from a com-152-4 bination of hindered internal rotation about the amide linkage and steric crowding of the C-3 side chain by the bulky N-acetyl group. These two conformers influenced the environments of the C-l methyl (62.60, 2.52) and N-methyl groups (63.05, 3.00) as well as the N-acetyl methyl (62.08, 1.98) i (121a) (121b) and the C-l* methylene group (65.50, 5.46). This resulted i n twin sets of peaks being present for each of these groups. Singlet resonances occurred at 62.88 and 2.03 for the C-4 methyl and the acetate methyl groups, respec-tively. These groups were not influenced by the existence of the molecule i n two distinct conformation. The IR. spectrum possessed absorptions at 1730 cm"* and 1635 enf* for the carbonyls of the acetate and N-acetyl groups, respectively. A normal carbazole spectrum was obtained in the UV. of compound 121 . The mass spectrum possessed a parent peak at m/e = 366 with major fragmentations at m/e = 306 (M* - HOAc), 280 (M+ - C4H8NO), 263, 233, 222-1. The parent peak at 366.1956 was within acceptable limits of the calculated value 366.1943 for the composi-tion C22H26N203)« - 125 -The hydrolysis of the acetate group of 121 to the corresponding alcohol 122 proved to be more d i f f i c u l t than expected. The N-acetyl group was com-pletely inert to the basic media whereas the acetate group was readily dis-placed by alkoxide when either aqueous methanol or ethanol was used as sol-vent. A two phase system of t-butanol and aqueous sodium hydroxide was found to effect hydrolysis without any competing nucleophilic substitution by t-butoxide ion. The reaction proceeded slowly at room temperature. After 24 hr. unreacted starting material was detected. When the temperature was raised to 100°, the acetate group of 121 was completely hydrolyzed after 1 hr. Only a single peak at 1635 cm - 1 for the carbonyl of the N-acetyl group was present in the IR. spectrum which was consistent with the hydrolysis to the alcohol 122 . The UV . spectrum remained essentially unchanged as expected. The mass spectrum possessed a weak parent peak at m/e = 324 and a major frag-mentation at m/e = 308 for loss of 16 mass units. The parent peak at m/e = 324.1814 in the high resolution mass spectrum was consistent with the composi-tion C20H24N2O2. The N.M.R. spectrum for 122 was too complex to properly analyze, again as a result of the presence of conformers. The complexity of the spectrum may also have been increased by the a b i l i t y of the hydroxyl group to hydrogen bond to the amide carbonyl oxygen. A pair of multiplets were discernable at 66.20 and 5.00 for the methylene hydrogens at C-l*. The remaining singlets for the various methyl groups were s p l i t in a complex manner which made their assign-ment ambiguous. - 126 -The hydrolysis of the 3-(pi -acetoxyethyl) carbazole derivative 223 using the same two phase basic media also proceeded to completion after 1 hr. at 100° (isolated yield 79%). A l l the spectral and analytical data for the pro-duct alcohol 222 was identical to that obtained for the same compound 222 syn-thesized in sequence B, part III (page 225)* Pertinent at this point was the loss of the acetate carbonyl peak at 1720 cm-1 in the IR spectrum. Basic media was required for the subsequent dehydration of the alcohol 222 to the olefin 123;. because acidic media (for example 20% r^SO^) would result in concomitant hydrolysis of the amide group. The dehydration was effected by taking advantage of the known i n s t a b i l i t y of benzylic tosylates with respect to elimination. By refluxing the alcohol 222 in pyridine containing p-toluene-,155 sulphonyl chloride the olefin 123 was obtained in 60% y i e l d . The N.M.R. spectrum possessed a doublet of doublets at 65.78 (Jj, 2 i = 17 Hz, <J2,,2' = **z) ^ o r ^ e t r a n s C-2* Hydrogen and a second doublet of dou-blets at 65.21 ( J i i ^ ' = 1 1 H z » J2',2' = 1 > 5 H z ) f o r t h e cis_ C-2» hydrogen. The corresponding doublet of doublets for the C - l 1 hydrogen was partly obscured by the aromatic signals. The chemical shifts for the cis and trans C-2* hydrogens coincided closely with those found for olefin 88 which also possesses the C-3 vinyl side chain. Singlets occurred at 65.72 and 2.60 respectively for the methylene and C-l methyl group. The UV . spectrum also bore a close resemblance to that obtained for ol e f i n 88 as expected. The mass spectrum possessed a parent peak at m/e = 297 and - 127 -almost no fragment peaks. The parent peak at 297.1472 in the high resolution spectrum was within acceptable limits of the calculated value Having worked out the conditions necessary for formation of the ring-opened acetate, i t s hydrolysis, and dehydration of the alcohol for the model compounds, attention was directed towards applying these reactions to guatam-buine (25). The acetylation reaction was i n i t i a l l y conducted using acetic anhydride in pyridine. Under these conditions the reaction proceeded as anti-cipated to give the N-acetyl acetate 124. The presence of two conformers of 124 was again observed i n the N.M.R. The N-acetyl and C-l methyl groups and the C - l ' methine hydrogen were influenced the most by the two separate conformations of the molecule. Shoulders were observed on the singlet peak for the N-methyl group (63.00) and on the doublet peak for the C - l 1 methyl group (61.64, J = 6 Hz). Also, a multiplet instead of the anticipated quartet was obtained for the C - l ' methine hydrogen (66.34). The resonances for the C-l methyl (62.48, 2.40) and N-acetyl methyl groups (62.08, 1.98) were present as a distinctly separated pair of singlets. The singlet for for the composition C22H19N. (124) - 128 -the acetate methyl group (62.08) was superimposed upon the singlet for N-acetyl methyl group of the major conformer of 124. A single peak was thus obtained for both groups. The mass spectrum possessed a weak parent peak at m/e = 366 and a major fragment at m/e = 306 for loss of the elements of acetic acid. Two peaks were present at 1710 and 1620 cm"* i n the IR. for the carbonyls of the acetate and N-acetyl groups respectively. A carbazole UV. spectrum was obtained for 124 as expected. When sodium acetate i n acetic anhydride was used for the acetylation ring-opening, i t was observed that considerable quantities of o l e f i n i c material was present in the reaction mixture. By conducting the reaction at reflux tempera-ture for 10 hr. the olefin 125 was obtained as the major product in 64% yield. The spectral data for 125 was consistent with i t s assigned structure. The occurrence of two distinct conformers of 125 resulting from hindered rotation about the N-methylamide linkage was again observed in the N.M.R. spectrum (Figure 37). Twin sets of peaks were observed for the N-methyl (63.00, 2.92), C-l methyl (62.58, 2.52) and N-acetyl methyl groups (62.10, 1.96), and the C-2' cis oriented vinyl proton (65.37, 5.34). A doublet of doublets was present at 65.72 for the trans C-2' hydrogen. This proton was far enough removed from the (125) - 129 -- 130 -amide group to be uninfluenced by i t . The mass spectrum possessed a parent peak at m/e = 306 and major frag-ments at 234, 220 and 204. The parent peak in the high resolution spectrum at m/e = 306.1736 was within acceptable limits of the calculated value m/e = 306.1732 for C 2 o H22 N2°* A s i n g l e carbonyl absorption at itCO cm - 1 was present in the IR. spectrum for the N-acetyl group. The UV. spectrum for compound 125 was similar to that for the C-3 olefin 88 in that the absorption maxima consisted of two broad humps i n the region of 280 mm. and 240 mm. It was apparent from these results that reaction with acetic anhydride i n sodium acetate induced the pyrolytic elimination of the i n i t i a l l y formed acetate to occur. Identical reactivity has been observed for a-arylalkyl quaternary 143 ammonium salts on pyrolysis. This property was of considerable value since i t meant that guatambuine (25) could be transformed i n one step and in good yield to the desired C-3 vinyl derivative. In terms of the development of the degradation sequence for the isolation of the C-l methyl group either the acetate ring-opening reaction or the Hofmann reaction using sodium hydride in dimethylformamide could be used. The only advan-tage that the Hofmann reaction offered was that the yie l d was somewhat higher and the reaction time was considerably shorter. C. Oxidation Reactions This part of the discussion i s concerned with a pa r t i a l review of the possible application of several different types of oxidation-hydrolysis reactions towards the isolation of the C-l methyl group of guatambuine (25). Very l i t t l e work was done i n this area however, due to the successes achieved with the other approaches. - 131 -One area that was not studied at a l l was the p o s s i b i l i t y of the d i r e c t oxidative cleavage of the C-l methyl from the parent aromatic, compound olivacine (16). The pyridine nitrogen imparts a r e a c t i v i t y to the a-substituted C-l methyl which i s not possessed by the C-5 methyl group. I t should be possible therefore through the choice of the proper o x i d i z i n g conditions to s e l e c t i v e l y oxidize the C-l methyl group p r e f e r e n t i a l l y to the C-5 methyl group. IV I I 156-8 Both Ce and Ag o x i d i z i n g reagents have been shown to s e l e c t i v e l y oxidize a single a l k y l group of a p o l y a l k y l benzene system. I t has been further shown that i t was possible to conduct a stepwise oxidation of that methyl group from the alcohol to the benzoic acid derivative. The control over the oxida-t i o n process exhibited by these compounds makes t h e i r application to the oxida-t i o n of the C-l methyl of olivacine (16) e n t i r e l y possible. Permanganate ion would also be a s e l e c t i v e o x i d i z i n g agent since i t i s known that 2-methylpyridine i s r e a d i l y oxidized to pyridine-2-carboxylic a c i d . 159 A number of methods including the use of permanganate ion, manganese 160 I I 161-2 dioxide, and Ag s a l t s are available f o r the oxidation of amines to the corresponding carbonyl compound. They are generally applicable'only to primary or secondary amines however since the intermediate imines which are formed are h y d r o l y t i c a l l y unstable with respect to the aldehydes or ketones. Dimethyl s u l -163 phoxide has been shown to be a r e l a t i v e l y good ox i d i z i n g agent for t e r t i a r y amine hydrochlorides and quaternary ammonium s a l t s . However, the oxidation of 1-phenylethylamine s a l t s which were closely related to guatambuine (25) f a i l e d due to the i n s t a b i l i t y o f the ketonic products formed i n the reaction media. These types of reactions were therefore not attempted on the guatambuine system. The majority of reactions f o r o x i d i z i n g amines to the carbonyl compounds - 132 -involve formation of the imine or imminium cation as the starting material 164 or as the reaction intermediate. Mercuric acetate has been u t i l i z e d successfully for the oxidation of tertiary amines to their carbonyl counter-parts. The reaction involves prior formation of the iminium cation which readily hydrolyzes in aqueous base. The application of this reaction to cyc l i c systems such as guatambuine (25) however does not result i n ketone formation, basification results i n the formation of the ot,B-enamine. The subject of ena-mine formation w i l l be returned to shortly. CH,I +/ OH" CH=N— 3 - CH^rM^— CH=0 165 Ozonolysis has been used to cleave the carbon-nitrogen double bond of Schiffs bases and nitrones to the corresponding aldehydes or ketones. Oxaziranes and amides are generally formed also during the reaction and i n comparable yields to the carbonyl component. This would be a major drawback to i t s application to small scale degradation work. A method which has been developed s p e c i f i c a l l y for the oxidation of ter-tiary amines involves treatment of the amine or the iminium salt with warm 166 buffered permanganate followed by rapid product removal. It would be necessary - 133 -however for the degradation of 142 to modify the i s o l a t i o n procedure f o r the reaction to accommodate working on a small scale with n o n - v o l i t i l e carbonyl compounds. Some preliminary experiments that were t r i e d involved an attempt to hydrolyze the iminium s a l t of guatambuine 142 i n a c i d i c and basic media. The iminium s a l t 142 was available as a stable c r y s t a l l i n e s o l i d from the synthesis presented i n sequence A, part I I I . Ketone formation could not be detected under either conditions. In aqueous base however, an intense yellow coloured solution was produced and an amorphous yellow p r e c i p i t a t e developed. The formation of t h i s intense colouration has been ascribed i n related systems to the formation 136,167-8 of the a,8-enamine (anhydro base). ' In comparison to the iminium cation 142 only minor s h i f t s i n the UV. spectrum were observed to accompany the formation of the enamine 126 (Figure 38). (142) (126) Attempts to i s o l a t e the enamine 126 (anhydro base) by extraction i n t o ether f a i l e d . This was not a t o t a l l y unexpected r e s u l t since they are known to be unstable with respect to the quaternary i o n . A ring-opening of the B-carboline r i n g system has been developed by Gupta 167-8 and Spenser during t h e i r investigation of the methylation of the anhydro bases of Harman derivatives (Figure 39). Figure 38.. The UV. Spectrum of the iminium Cation (142) Enamine (126), and Reaction Product of 142 Treated with Dimethylsulphate in Strong Base. - 135 -They found that the enamine of Harmaline 128 could be isolated after base treatment of the iminium salt 127. When the enamine was subsequently treated with methyl iodide, the quaternary ammonium iodide produced was displaced by hydroxide to give the ring opened ketone 129. A Hofmann elimination of the amine function of the C-3 side chain led to the C-3 vinyl product 130. This same process has also been shown to occur for the l-methyl-3,4 dihydroisoquino-169 line systems by treatment with dimethylsulphate in concentrated base. It i s entirely feasible that this sequence of reactions could be carried out on guatambuine (25). The enamine 126 could be made available by mercuric 164 acetate oxidation of guatambuine (25). The reaction with dimethyl sulphate in 20% aqueous sodium hydroxide has been conducted on a small scale. Within 5 min. after heating the reaction mixture at 100° the yellow colouration for the enamine 126 disappeared. At the present time, however, only the UV data was available (Figure 38). The absorption curve for the reaction mixture was quite reminiscent of both the imine (141) and the C-3 o l e f i n 88 (Figure 27) which gives some indication at least that the starting material 142 alters in the reaction media and that perhaps the keto-olefin (131) was produced. If the chemistry of this ring-opening reaction could be developed to the point where good yields were obtained, then i t could serve as a feasible alternative to the Hofmann pr acetate substitution approaches previously dis-cussed. By a proper choice of reactions, the selective isolation of the C-l methyl, N-methyl and C-3 carbons of guatambuine could be achieved, i . e . boro-hydrive reduction of the ketone to the alcohol, then ozonolysis of the double 170-171 bond followed by a Haloform reaction on the secondary alcohol. -136-(129) OH-(130) Figure 39. Ring Opening of the 3-Carboline Ring System by Methylation of the Anhydro Base. - 137 -EXPERIMENTAL - PART II For a description of the general experimental information, see Experimental part 1. A l l T.L.C. plates were developed in chloroform ( C H C I 3 ) or ethyl acetate unless otherwise indicated, and the alumina for column chroma-tography was deactivated to Activity III by the addition of water (6%). N-Methyl-1,2,3,4-tetrahydroellipticine (26) A solution of e l l i p t i c i n e (17) (500 mg., 2.03 x 10"^ mole) dissolved in methanol (400 ml.) containing an excess of methyl iodide (5 ml.) was stir r e d at room temperature for 24 hr. The reaction mixture was then concentrated to dryness, taken up i n methanol (200 ml.) and preadsorbed onto alumina (10 gm.), and applied to the top of an alumina column (40 gm.). Traces of unreacted starting material were eluted with ethyl acetate:methanol 5% and the desired methiodide product 93 was subsequently eluted with methanol: ammonium hydroxide 1%. The methiodide 93 was obtained as a bright orange crystalline s o l i d (630 mg., 77%). -3 The methiodide 93 (500 mg. 1.24 x 10 mole) was dissolved in aqueous - 138 -ethanol (250 ml.) and reacted with an excess of sodium borohydride at room temperature for 15 hr. The reaction mixture was then concentrated, taken up in water (150 ml.) and extracted with chloroform ( 3 x 50 ml.). The combined chloroform fractions were dried over sodium sulphate and concentrated to give 26 as a pure colourless crystalline s o l i d (300 mg., 91%) which turned yellow on prolonged standing. The product 26 was stabilized by recrystallization from methanol to give cream coloured needles, m.p. 217-220° ( l i t . , m.p. 215-126 220°). UV.; m a v ( l o g e ) : 340 (3.49), 326 (3.62), 296 (4.25), 286 (sh)(4.02), 260 (4.49), 248 (4.60), 238 (4.74). N.M.R. (F.T.): 8.20 (doublet, J = 7Hz, C-10H), 7.84 (broad singlet, N-H), 3.76 (singlet, C-l CH 2), 3.00 ( t r i p l e t , J • 5Hz, C-4 CH 2), 2.76 ( t r i p l e t , J = SHz, (partly obscured by a singlet at 3.70), C-3 CH 2), 3.70 (singlet, C - l l CH 3), 2.56 (singlet, C-5 CH 3), 2.38 (singlet, N-CH3). Mass spectrum: M+; m/e = 264; main peaks: 263 (base peak), 249, 233, 221, 204-5. Found: C, 81.67; H, 7.64; N, 10.30. Calc. for C 1 8H 2pN 2 : C, 81.78; H, 7.63; N, 10.60. N-Methyl-l,2,3,4-tetrahydroellipticine methiodide (95). A solution of 26 (265 mg. 1.00 x 10~3 mole) i n methanol (25 ml.) con-taining an excess of methyl iodide was allowed to stand at 0° for 24 hr. The colourless crystals of 95 that precipitated were collected by suction f i l t r a -tion, washed with methanol and dried under vacuum (396 mg. 98%), m.p.307 - 308° Found: C, 55.91; H, 5.87; N, 6.50. Calc. for C19 H23 N2 I : C ' 5 6 - 2 9 J H» 5 - 6 9 J N» 6 « 9 1 ' - 139 -1,2,3,4-Tetrahydroellipticine (94). E l l i p t i c i n e (17) (100 mg., 4.06 x 10~* mole) was dissolved i n glacial acetic acid (20 ml.) and the solution was hydrogenated at room temperature and atmospheric pressure over Adams catalyst (Pt0 2, 85 mg.) for 12 hr. The mixture was f i l t e r e d to remove the catalyst which was carefully washed with additional acetic acid (10 ml.). The combined f i l t r a t e s were concentrated, taken up i n water (75 ml.), basified with 10% sodium hydroxide and the resulting white suspension was extracted with chloroform (3 x 30 ml.). The combined chloroform fractions were dried over sodium sulphate, and concen-trated to give 94 as colourless crystals (95 mg.,93 % ) , which rapidly turned yellow on standing. UV.; X m a x: 340, 324, 293, 284, 261, 250, 241 (log (e) values were not obtained, however, the relative extinction coefficients for the absorption maxima were v i r t u a l l y identical to those observed for N-methyl-tetrahydroellipticine (17)). Compound 94 was converted to i t s N-acetyl derivative for further characterization. The preparation of the methiodide 95 of compound 94 was carried out in an identical manner to the reaction of N-methyltetrahydroellipticine (17). The melting point and microanalytical data agreed sa t i s f a c t o r i l y with the values previously obtained. 137 N - A c e t y l - l ^ S ^ - t e t r a h y d r o e l l i p t i c i n e (96). Compound 94 (35 mg. 1.40 x 10~* mole) was dissolved in a mixture of acetic anhydride (3 ml.) and pyridine (3 ml.), and stirred at 70° for 3 hr. - 140 -The reaction mixture was then concentrated, taken up in water (75 ml.), basified with 10% sodium hydroxide solution, and extracted with chloroform (3 x 30 ml.). The combined chloroform fractions were washed with water, dried over sodium sulphate, and concentrated to give crude 96 (40 mg., 97%) Several recrystallizations from methanol afforded 96 as colourless plates, m.p. 262-265° ( l i t . m.p. 272.5-273°). Found: C, 78.01; H, 7.02; N, 9.30. Calc. for C i gH 2 0N 2O: C, 78.05; H, 6.89; N, 5.47. Olivacine methofluorosulphate (85) and i t s reduction to Guatambuine (25). Olivacine (16) (20mg., 8.13 x 10"5 mole) dissolved i n acetonitrile (25 ml.) was treated with methylfluorosulphate (30ul, 3 equiv.) (freshly d i s t i l l e d ) at room temperature. A yellow precipitate formed within minutes and after 15 min. the reaction mixture was concentrated to give the fluorosulphate salt 85 as a yellow paste. The crude product 85 was suspended in aqueous ethanol (100 ml.) and reacted with an excess of sodium borohydride at room temperature, for 15 hr. The yellow colouration for the olivacine salt disappeared almost instantaneously on addition of the reducing agent. The reaction mixture was then concentrated to a paste, taken up i n water (75 ml.), and extracted with chloroform (3 x 30 ml.). The combined chloroform fractions were dried over sodium sulphate, and concentrated to give guatambuine (25) as a cream coloured s o l i d (20 mg., 93%). Recrystallization from methanol gave colour-less needles, m.p. 248° ( l i t . , m.p. 248-250°). The UV. and N.M.R. and low resolution mass spectra for the reaction product was consistent with that previously obtained for guatambuine (25). High resolution mass spectrum: Calc. for C. H N : 264.1626. Found: 264.1663. lo Z\j I - 141 Guatambuine methiodide (86). The methiodide 86 was prepared in quantitative y i e l d by reaction of 25 with an excess of methyl iodide at 0° for 24 hr. Colourless needles were obtained which were recrystallized from methanol, m.p. 299° ( l i t . , 299-121-3 301°). N.M.R. (DMS0-d6): 8.10 (doublet, IH, J = 8Hz, C-10 H), 2.90 (singlet, IH, C - l l H), 4.96 (quartet, 1 H, J = 6Hz, C-l H), 3.80 (broad multiplet, 2 H, C-4CH2), 3.30 (multiplet, obscured by singlet absorption at 3.18, C-3 CH 2), 3.18, 3.12 (two singlets, 3 H, N(CH 3)2), 1.74 (doublet, 3H, J = 6Hz, C-l CH 3). Found: C, 56.34; H, 5.69; N, 6.63. Calc. for C19 H23 N2 I : C» 5 6 - 2 9 J H» 5 - 6 8 J N» 6-80. Olivacine methiodide (85) and i t s Reduction to Guatambuine (25). Olivacine (16) (15 mg., 6.09 x 10"^ mole) dissolved i n ethanol (25 ml.) was treated with an excess of methyl iodide and refluxed for 0.5 hr. The crystalling precipitate was collected and the mother liquors were concentrated, taken up in a minimum amount of methanol and treated again with methyliodide. The second crop of crystalline material was collected and the mother liquors were again concentrated and rereacted with methyl iodide. By recycling the mother liquors i n this fashion four* times, an overall 80% y i e l d of impure olivacine methiodide (85) (18.4 mg.) was obtained. The crude product 85 (14 mg., 3.49 x 10"^ mole) was reduced with sodium borohydride in an identical manner to the methofluorosulphate salt 85 described i n the previous experiment. The spectral data was consistent - 142 -with that previously obtained for guatambuine (25). Hofmann Degradation of (-)-Guatambuine Methiodide (86). i ) Preparation of l-Methyl-2-(6-dimethylarainoethyl)-3-(a-ethoxyethyl) 121-3 carbazole (98) and C-3 olefin 8 8 .  (-)-guatambuine methiodide ( 8 6 ) (40.2 mg., 9.92 x 10"5 mole) was suspended in a 10% sodium hydroxide in 9 5 % ethanol solution (60 ml.) and refluxed for 2 . 5 hr. The reaction mixture was then concentrated to a paste, diluted with water ( 7 5 ml.) and extracted with ether (3 x 30 ml.). The combined ether fractions were dried over sodium sulphate, and concentrated to give a light brown o i l containing a mixture of com-pounds 98 and 8 8 (40 mg.,>100%) N.M.R.: (See Figure 23). Compound 98 was isolated as a transparent o i l by successive pre-parative layer chromatography on alumina (1 mm., EtOAc) (24 mg., 7 5 % ) and s i l i c a gel (19 mg., 60%). An analytical sample was obtained by dis-25 25 t i l l a t i o n of 98 at 160-180° at 0.07-0.01 mm. (a D) = 0° (before: (a n) = -112°). UV.; X m a x: 340, 326, 297, 287 (sh), 261, 250, 240. N.M.R. (Figure 23): 4.85 (quartet, 1 H, J = 6 H z , C-l'H), 3.42 (quartet, 2 H, J = 7 H z , O C H 2 C H 3 ) , 3.05 (multiplet, 2 H, C-3» CH 2), 2 . 4 9 (singlet, 3 H, C-l CH 3), 2.40 (singlet, 6 H, N(CH 3) 2), 1.53 (doublet, 3 H, J = 6 H z , C-l' CH 3), 1.20 (t r i p l e t , 3 H, J = 7 H z , O-CH2-CH3). Mass spectrum: M*, m/e = 324; main peaks: 295, 278. High resolution mass spectrum: Calc. for C 2 1 H 2 8 N 2 0 : 3 2 4 - 2 2 0 0 - F o u n d : 324.2207. Found: C, 56.85; H, 6.60; N, - 143 -5.75. Calc. for C 2 2H 3 1N 2IO: C, 56.80; H, 6.67; N, 6.0. ) Preparation of l-Methyl-2-vinyl-3-("-dimethylaminoethyl)carbazole (97),  and l-Methyl-2-(o-dimethylaminoethyl)-3-vinylcarbazole (88). A) Guatambuine methiodide (86) (20 mg., 4.93 x 10"^ mole) was suspended in t-butanol (12 ml.) containing potassium t-butoxide (30 mg.) and heated at reflux for 1.5 hr. The solution was then concentrated to a paste, diluted with water (75 ml.) and extracted with ether ( 3 x 30 ml.). The combined ether fractions were dried over sodium sulphate, and concentrated to give a light brown o i l containing a mixture of compound 97 and 88 (10 mg., 72%). N.M.R. (Figure 24): 5.70 (distorted doublet of doublets, 1 H, J r 2' = 1 7 H z » J2' 2' ~ 2 H z * C " 2 ' H t r a n s ( 8 8) overlapped with C-41 H cis (97) ) 5.68 (doublet of doublets, 1 H, J 3 1 ^ , = 11 Hz, ^ 1 ^ 4 1 = 2 Hz, C-41 H cis (97) ), 5.23 (doublet of doublets, 1 H, J 3 1 < 4 i = 17 Hz, J 4 1 > 4 t = 2 Hz, C-41 trans (97) ), 5.32 (doublet of doublets, 1 H, J j , 2 , = 11 Hz, J 2 , 2, = 2 H z , C-21 H cis_ (88) ), 3.82 (quartet, J = 6 H z , C - l ' H), 2.52 (singlet, C - I C H 3 ) , 2.40 (singlet, N(CH 3) 2), 2.32 (singlet, N(CH 3) 2), 1.44 (doublet, J = 6 Hz, C - l 1 CH 3), 1.28 (singlet, 0(CH 3) 3). The major compound 97 was isolated by preparative layer chromotography on alumina (1 mm., ethyl acetate) (4 mg., 30%). UV. (figure 27); A : HI CLA. 340, 326, 297, 248. N.M.R. (Figure 25): 8.08 (multiplet, 214, C-4, C-5 H), 6.98 (doublet of doublets, 1 H, J 3 , > 4 , = 11 Hz, J 3 ^ 4 , = 17 Hz, C-3'H), 5.65 (doublet of doublets, 1 H, J 3 , 4 , = 11 Hz, J 4 , 4 , = 2 Hz, C-4' H c i s ) , - 144 -5.20 (doublet of doublets, 1 H, J 3 i # 4 i = 17 Hz, ^ . ^ 4 1 = 2 Hz, C-4' H trans), 3.75 (quartet, 1 H, J = 6Hz, C-l' H), 2.50 (singlet, 3 II, C-l CH3), 2.25 (singlet, 6 H, N(CH 3) 2), 1.39 (doublet, 3 H, J = 6 Hz, C-1»CH 3). 0 Mass spectrum: M*, m/e = 278 (base peak); main peaks: 263, 249, 235-3, 220-217, 205, 204. High resolution mass spectrum: Calc. for C 1 9H 2 2N 2: 278.1782. Found: 278.1780. Found: C, 56.92; H, 5.71; N, 6.63. Calc. for C2 0H 2 5N 2I: C, 57.27; N, 5.96; N, 6.68. B) In a second experiment, guatambuine methiodide (86) (80 mg., 1.97 x 10~4 mole) was suspended i n t-butanol (30 ml.) containing potassium t-butoxide (80 mg.) and heated at reflux for 2.5 hr. On work-up a mixture of olefins 88 and 97 was obtained (40 mg., 72%). The mixture was separated by column chromatography on alumina (5 gm.). By elution with benzene: chloroform 1:1 ole f i n 97 was obtained (20 mg., 36%). The spectral data for 97 was identical to that obtained in experiment A. By increasing the solvent polarity to benzene:chloroform 70%, olefin 88 was eluted to give a light brown o i l (10 mg., 18%). UV. (Figure 27); Xmov.: 325 (broad hump), 295 (sh), 280, 267 (sh), 241. N.M.R. (Figure 26): 8.00 (multiplet, 3 H, C-4, C-5 H, N-H), approximately 7.2 (doublet of doublets, 1 H, C-l'H (obscured by CHC1 peak) ), 5.68 (doublet of doublets, 1 H, J l ' 2' = 1 7 H z » J2' 2' = 2 H z » c " 2 ' n t r a n s ) » 5 - 2 9 (doublet of doublets, 1 H, J i ^ 2 i = 11 Hz, J 2 t j 2 i = 2 Hz, C-2'H c i s ) , 3.10 (multiplet, 2 H, C-3'CH2), 2.50 (singlet, 3 H, C - I C H 3 ) , 2.38 (singlet, 6 H, N(CH 3) 2). Mass spectrum: M+, m/e = 278 (base peak); main peaks; 263, 249, 234, 233, - 145 -220, 219-19, 204. High resolution mass spectrum: Calc. for CigH22 N2 : 278.1782. Found: 278.1780. Found: C, 56.86; H, 6.27; N, 6.22. Calc. for C H N I: C,57.15; H, 5.99; N, 6.66, For the methiodide of 88. m.p. 280-282°. UV.;AmQV (loge) : 330(3.25), 296(sh)(3.98), 278(4.23), 240(4.41). i) Preparation of Olefins 88 and 97. Guatambuine methiodide (86) (15 mg., 3.70 x 10"5 mole) was sus-pended i n a 40% aqueous sodium hydroxide solution (20 ml.) and heated at reflux temperature for 3 hr. The cooled reaction mixture was diluted with water (50 ml.) and extracted with chloroform (3 x 30 ml.). The .combined chloroform extracts were washed with water, dried over sodium sulphate, and concentrated to give a brown o i l (6mg., 60%). The spectral data for the crude reaction mixture was identical to that using potassium t-butoxide as the base in t-butanol (reaction i i ) . ) Preparation of l-Methyl-2-(8^dimethylaminoethyl)-3-vinylcarbazole (88). -4 Guatambuine methiodide (86) (76 mg., 2.42 x 10 mole) and sodium hydride (50 mg.) i n dry dimethylformamide was heated at 100° for 2 min. after which time the excess hydride was destroyed by the careful addition of water. The reaction mixture was then diluted with water (75 ml.) and the resultant white suspension was extracted with ether (4 x 30 ml.). The combined ether layers were washed with water (3 x 30 ml.), dried over sodium sulphate and concentrated to give olefin 88 as a colourless o i l which slowly s o l i d i f i e s to an amorphous white solid under vacuum (37 mg., 7 0 % ) . UV. (Figure 27); X m a v: 325 (broad hump), 295 (sh), 280, 267 (sh), - 146 -(Figure 26) 241. N.M.R.: 8.00 (multiplet, 3 H, C-4, C-5, N-H), 5.68 (doublet of doublets, 1 H, J J f 2 , = 17 Hz, J 2 , 2 , = 2 Hz, C-2'H trans), 5.29 (doublet of doublets, 1 H, Ji» t2' = 1 1 H z » J2',2' = 2 H z » C ~ 2 ' H c i s ) , 3.10 (multiplet 2 H, C-3'CH2), 2.50 (singlet, 3 H, C - I C H 3 ) , 2.38 (singlet, 6 H, N(CH 3) 2). Hofmann Degradation of N-Methyltetrahydroellipticine methodide (95). Preparation of l,4-Dimethyl-2-vinyl-3-dimethylaminomethylcarbazole (102). N-Methyltetrahydroellipticine methiodide (95) (50 mg., 1.23 x 10~4 mole) was suspended in t-butanol (25 ml.) containing potassium t-butoxide (40 mg.) and heated at reflux for 4 hr. The solution was concentrated to a paste, diluted with water (75 ml.) and extracted with ether (3 x 30 ml.). The com-bined ether fractions were dried over sodium sulphate, and concentrated to give 102 as a tan coloured o i l (33 mg., 96%). UV.: Ara : 335, 322, 293, 250. N.M.R.: 8.25 (broad doublet, 1 H, J = 6 Hz, C-10 H), 8.02 (broad hump, 1 H, N-H), 7.46-6.96 (multiplet, 5 H, 4 aromatic H, C-3'H), 5.64 (doublet of doublets, 1 H, J j , 4 , = 11 Hz, J 4 , 4 , = 2 Hz, C-4'H, c i s ) , 5.20 (doublet of doublets, 1 H, J j , 4,,."= 17 Hz, J 4 , 4 , = 2 Hz, C-4'H, trans), 3.64 (singlet, 2 H, C-l'CH ), 2.90 (singlet, 3 H,C-4 CH_), 2.50 (singlet, 3 H, C-l CH), £ O «3 2.24 (singlet, 6 H, N(CH 3) 2). Mass spectrum: M*, m/e = 278; main peaks: 263, 249, 248, 234, 233 (base peak), 219, 218, 204. High resolution spectrum: Calc. for C 1 QH„N,: 278.1782. Found: 278.1780. - 147 -l-Methyl-2-(3-dimethylaminoethyl)-3-ethylcarbazole (89) from (88). Olefin 88 (2 mg., 7.19 x 10"6 mole) dissolved in methanol (3 ml.) was hydrogenated at room temperature and atmospheric pressure over Adams catalyst (Pt0 2, 5 mg.) for 0.5 hr. The mixture was fi l t e r e d to remove the catalyst which was carefully washed with additional methanol (10 ml.). The combined f i l t r a t e s were concentrated under vacuum to give 89 as a colourless o i l (2 mg.). The spectral data for the reaction product 89 was consistent with that for the same product 89 obtained through reaction of guatambuine methiodide (86) with lithium aluminum hydride and through Hofmann degradation of Uleine (18) (to be subsequently described). N.M.R. (FT): 2.64 (singlet, 3H, C-l CH 3), 2.54 (singlet, 6 H, N(CH 3) 2). l-Methyl-2-(B-dimethylaminoethyl)-3-ethylcarbazole (89) from (86). Guatambuine methiodide (86) (50 mg., 1.23 x 10~* mole) and lithium aluminum hydride (50 mg.) suspended in dry tetrahydrofuran (25 ml.) was heated at reflux temperature for 4 hr. The excess hydride reagent was destroyed by the successive addition of water (1 ml.), 10% aqueous sodium hydroxide (2 ml.), and water (1 ml.). The resultant precipitate was removed by suction f i l t r a t i o n , and washed with methanol. The combined f i l t r a t e s were concentrated, taken up in ether (75 ml.) and washed with water (3 x 30 ml.). The ether layer was then dried over sodium sulphate, and concentrated to give 89 as a transparent o i l (30 mg., 88%). UV: Xmax: 3 4 0 » 3 2 6 » 296, 2 8 5 (sh), 260, 247, 239. N.M.R.: 8.03 (doublet, 1 H, J = 7 Hz, C-5 H), 7.78 (singlet, 1 H, C-4H), 3.2-2.6 (multiplet, 6 H, Ar-CH2-CH3, CH2-CH2-N(CH3)2), 2.52 (singlet, 3 H, C-l CH 3), 2.46 (singlet, 6 H, - 148 -N(CH 3) 2), 1.36 ( t r i p l e t , 3 H, J = 7 Hz, Ar-CH2-CH3). Mass spectrum: M+, m/e = 280 (base peak); main peaks: 266, 236, 222, 207, 206, 204. High resolution mass spectrum: Calc. for cIQ H24 N2 : 280.1938. Found: 280.1975 Compound 89 obtained from guatambuine methiodide (86) was further characterized as i t s methiodide salt 113 which was formed in quantitative yi e l d by reaction with an excess of methyl iodide in methanol at 0°C for 121-3 24 hr. m.p. 291-293° ( l i t . , m.p. 287-8 ). Found: C, 56.72; H, 6.40; N, 6.42. Calc. for C 2 f JH 2 7N 2I: C, 56.88; H, 6.44; N, 6.63. UV.; A m a x (loge) 340 (3.32), 325 (3.40), 297 (4.12), 285 (3.88), 260 (4.09), 239 (4.52). - 149 -l-Methyl-2-(B-dimethylaminoethyl)-3-ethylcarbazole (89) from the Hofmann  Degradation of Uleine Methiodide (87). 132 Uleine methiodide (87) was prepared according to Schmutz et_ a l . ixo and a modification of their procedure was used to degrade this compound. Uleine methiodide (87) (60 mg., 1.47 x 10~4 mole) was suspended i n a solution of 5% sodium hydroxide i n 95% ethanol (25 ml.) and refluxed for 2 hr. The reaction mixture was then concentrated to a paste, diluted with water (75 ml.) and extracted with ether (3 x 30 ml.). The combined ether fractions were dried over sodium sulphate, and concentrated to give a brown o i l . Polar contaminants were removed by column chromatography on alumina (5 gm.). Elution with chloroform yielded 89 as a light brown o i l (31 mg. 75%) UV; X : 337, 324, 296, 285 (sh), 260, 245, 237. N.M.R.: 7.74 (singlet, max 1H, C-4H), 3.2-2.7 (multiplet, 6 H, Ar-CH2CH3, -CH2-CH2-N(CH3)2)» 2.48 (singlet 3H, C-l CH 3), 2.36 (singlet, 6 H, N(CH 3) 2), 1.30 (t r i p l e t , 3H, J = 7 Hz, Ar-CH2CH3). Mass spectrum: M+, m/e = 280 (base peak); main peaks: 266, 236, 222, 207, 206, 204. High resolution mass spectrum: calc. for C19 H24 N2 : 2 8 0 « 1 9 3 8 - Found: 280.1915. Compound 89 obtained from uleine methiodide (87) was further charac-terized as i t s methiodide s a l t 113, which was formed in quantitative y i e l d by reaction with an excess of methyl iodide i n methanol at 0°C for 24 hr., 121-3 m.p. 285° ( l i t . m.p. 287-8°). Found: C, 56.89; N, 6.47; N, 6.70. Calc. for C 2 ( )H 2 7N 2I: C, 56.88; H, 6.44; N, 6.63. - 150 -l,3,4-Trimethyl-2-(8-dimethylaminoethyl)carbazole CI07) . -4 N-Methyltetrahydroellipticine methiodide (95) (50 mg., 1.23 x 10 mole) and lithium aluminum hydride (50 mg.) suspended in dry tetrahydrofuran (25 ml.) was heated at reflux temperature for 4 hr. The excess hydride reagent was destroyed by the successive addition of water (1 ml.), 10% aqueous sodium hydroxide (3 ml.), and water (1 ml.). The resultant precipitate was removed by suction f i l t r a t i o n , and washed with methanol. The combined f i l t r a t e s were concentrated, taken up in ether (75 ml.) and washed with water (3 x 30 ml.). The ether layer was then dried over sodium sulphate, and concentrated to give 107 as an opaque s o l i d (30 mg., 88%). Ifecrystallization from chloroform yielded 107 as colourless crystals, m.p. 302-304°. UV.; \ i a x '• 340, 326, 297, 284 (sh), 262, 241. N.M.R : 8.18 (doublet, 1 H, J = 7Hz, C-5 H), 7.90 (broad hump, 1 H, N-H), 3.04 (multiplet, 4 H, CH2-CH2-N(CH 3) 2), 2.75 (singlet, 3 H, C-4 CH 3), 2.42 (singlet, 3 H, C-l CH 3), 2.36 (singlet, 3 H, C-3 CH 3), 2.32 (singlet, 6 H, N(CH 3) 2). Mass spectrum : M*, m/e = 280 (base peak); main peaks : 266, 249, 236, 222, 207, 206, 204, 191. High resolution mass spectrum : Calc. for C i gH 2 4N 2 : 280.1938. Found: 280.1903 Compound 107 obtained from N-methyltetrahydroellipticine methiodide (95) was further characterized as i t s methiodide salt 108, which was formed i n quantitative y i e l d by reaction with an excess of methyl iodide i n methanol at 0° for 24 hr. Found : C, 56.93; H, 6.36; N, 6.36. Calc. for C20 H27 N2 I : C ' 5 6 - 8 8 ; H» 6 - 4 4 i N» 6 * 6 3 * -151 -l,3 )4-Trimethyl-2-vinylcarbazole (109). Methiodide 108 (25 mg., 5.93 x 10~5 mole) was suspended in t-butanol (20 ml.) containing potassium t-butoxide (30 mg.) and heated at reflux for 2.5 hr. The reaction mixture was concentrated to a paste, diluted with water (75 ml.) and extracted with ether (3 x 30 ml.). The combined ether fractions, were dried over sodium sulphate, and concentrated to give a dark coloured o i l . Polar contaminants were removed from the crude product mixture by column chromatography on alumina. By elution with chloroform the desired olefin 109. was obtained as a tan coloured o i l (12 mg., 85%). UV.; * m a x : 335, 322, 293, 250. N.M.R. : 8.26 (doublet, 1 H, J = 7 Hz, C-5H), 6.94 (doublet of doublets, 1 H, J 3 i , 4 i = 18 Hz, J 3 i , 4 i = 12 Hz, C-3'H), 5.68 (doublet of doublets 1 H, J 3 , 4 , = 12 Hz, J 4 I J 4 I = 2 Hz, C-4«H, c i s ) , 5.24 (doublet of doublets, 1 H? J 3 1 41 = 18 Hz, J 4 I > 4 I = 2 Hz, G-4'H, trans), 2.84 (singlet, 3 H, C-4 CH 3), 2.52 (singlet, 3 H, C-l CH3), 2.40 (singlet, 3 H, C-3 CH3) . Mass spectrum: M+, m/e = 235; main peaks : 220, 205. High resolution spectrum : Calc. for C 1 7H 1 7N : 235.1360. Found : 235.1325. :- - - - - - '.' 1. l-Methyl-2-vinyl-5-ethylcarbazole (91). A. Methiodide 90 (25 mg., 5.93 x 10"5 mole) was suspended i n t-butanol (IS ml.) containing potassium t-butoxide (40 mg.) and heated at reflux for 2.5 hr. The reaction mixture was then concentrated to a paste, diluted with water (75 mi.) and extracted with ether (3 x 30 ml.). The combined ether fractions were dried over sodium sulphate, and concentrated to give a dark coloured o i l (17 mg.). Column chromatography on alumina (3 gm.) (elution with chloroform) - 152 -removed a portion of the dark coloured contaminants (14 mg.). The crude pro-duct was further purified by preparative layer chromatography on alumina (1 mm., Bz:CHCl3 1:1). Compound 91 was obtained as a dark brown o i l (about 8 mg., 57%). UV.: Indicated the presence of contaminants. N.M.R.: 5.65 (doublet of doublets, 1 H, ^ 1 ^ 4 1 = 12 Hz, J 4 1 41 = 2 Hz, C-4'H c i s ) , 5.29 (doublet of doublets, 1 H, J 3 1 41 = 18 Hz, ^ ^ 4 1 = 2 Hz, C-4'H trans), 2.54 (singlet, 3 H, C-l CH 3). Mass spectrum: M*, m/e = 235; main peaks: 220, 205. High resolution mass spectrum: Calc. for C 1 7H 1 7N: 235.1360. Found: 235.1384. B_. Methiodide 90 (40 mg., 9.50 x 10~5 mole) was dissolved in dime thy lformamide (5 ml.) containing sodium hydride (10 mg.) and heated at 100° for 2 min. The reaction mixture was then diluted carefully with water (70 ml.) and extracted with ether (3 x 30 ml.). The combined ether fractions were washed with water (3 x 30 ml.), dried over sodium hydride, and concentrated to give 91 as an opaque o i l (9 mg., 40%). The spectral data for compound 91 was described i n experiment A . (N-*4C methyl) Guatambuine methiodide (86). Guatambuine (5 mg., 1.89 x 10"^ mole) dissolved in methanol (2 ml.) was reacted with 1 4C-methyl iodide (1 ml., /.ox 10?DPM/ml.) for 15 hr. at room temperature. The solvent was removed and the crude methiodide 86 was diluted with unlabelled 86 (50 mg.) and recrystallized from methanol (36 mg.), m.p. 299° ( /.tfx lO^ DPM/m mole). - 153 -(N-*4C methyl) N-methyltetrahydroellipticine methiodide (95). N-methyltetrahydroellipticine (5 mg., 1.89 x 10" 5 mole) dissolved i n methanol (2 ml.) was reacted with 1 4C-methyliodide (1 ml., 1 ,Q x 10?DPM/ml.) for 15 hr. at room temperature. The solvent was removed and the crude methiodide 95 was diluted with unlabelled 95 (50 mg.) and recrystallized from methanol (47 mg.), ( 3 .8? x 10 DPM/m mole). Lithium Aluminum Hydride Ring-Opening of (86), Hofmann Degradation of (90),  and Isolation of the N-Methyl Group of (25). Methiodide 86 (36 mg., . 4«.x 10*DPM) was suspended i n tetrahydrofuran (25 ml.) and reacted with lithium aluminum hydride (50 mg.) at reflux tempera-ture for 4 hr. The product 89 was isolated and converted to i t s methiodide 90 (19 mg., 51%) as previously described. The methiodide 90 (19 mg., 4.51 x 10 - 5 mole) was reacted with potassium t-butoxide (30 mg.) in t-butanol (20 ml.) at reflux temperature for 2.5 hr. A slow stream of nitrogen gas was passed through the solution and the effluent gas containing the trimethylamine was passed through methyl iodide solution. The methyl iodide solution was concentrated and the (N-*4C methyl) tetramethyl-5 ammonium iodide was recrystallized from methanol (2.5 mg.,2 . fe1 x 10 DPM) PCS cocktail). Isolation Efficiency: 5? %. - 154 -Lithium Aluminum Hydride Ring-Opening of (95), Hofmann Degradation of (108),  and Isolation of the N-Methyl Group of (26). Methiodide 95 (47 mg., «9 , / a 10 DPM) was suspended in tetrahydrofuran (25 ml.) and reacted with lithium aluminum hydride (50 mg.) at reflux tempera-ture for 4 hr. The product 107 was isolated and converted to i t s methiodide 108 (25 mg., 51%) as previously described. The methiodide 108 (25 mg., 5.93 x 10"5 mole) was reacted with potassium t-butoxide (30 mg.) in t-butanol (20 ml.) at reflux temperature for 2.5 hr. A slow stream of nitrogen gas was passed through the solution and the effluent gas containing the trimethylamine was passed through methyl iodide solution This methyl iodide solution was concentrated and the (N-*4C methyl)tetramethyl-ammonium iodide was recrystallized from methanol ( 3 mg. Z.io x 10 DPM) (PCS cocktail). Isolation Efficiency: 17. %. Ozonolysis of (102), Isolation of the C-3 Methylene Group of (26). 146 The procedure followed was essentially that by Battersby and Harper. Ozonized oxygen (200 bubbles per minute) was passed through a solution of olefin 102 (44 mg., 1.58 x 10"4 mole) in methylene chloride (10 ml.) at -78° and then through acidified potassium iodide solution. The gas was passed for two times the interval required to produce the f i r s t colour of iodine in the potassium iodide solution (3-4 min.). After the methylene chloride had been evaporated the ozonide was decomposed by being heated under reflux for 0.5 hr. with water (20 ml.), zinc dust (200 mg.), and s i l v e r nitrate (10 mg.). Half the water was - 155 -then d i s t i l l e d at atmospheric pressure into a solution of dimedone (300 mg.) in water (20 ml.), and ethanol (8 ml.), water (10 ml.) was added to the dis-t i l l i n g flask and the d i s t i l l a t i o n to half volume was repeated into the same dimedone solution. Microcrystalline needles of the dimedone derivative began to separate out during d i s t i l l a t i o n . After 15 hr. at 0° the crystals were collected by centrifugation (27 mg., 87%) and recrystallized from 50% aqueous ethanol to affort the pure derivative, \ ;: Ozonolysis of (88), Isolation of the C-l Methyl Group of (25). Olefin 88 (50 mg., 1.18 x 10~4 mole) was ozonolyzed i n an identical manner to compound 102. Microcrystalline needles of the dimedone derivative began to separate out during d i s t i l l a t i o n . After 15 hr. at 0° the crystals were collected by centrifugation (20 mg., 60%) and recrystallized from 50% aqueous ethanol to afford the pure derivative., \ - 156 -l-Methyl-2-ethyl-3-(g-dimethylaminoethyl)carbazole (103). Olefin 97 (18 mg., 6.47 x 10"*"' mole) dissolved in methanol (10 ml.) was hydrogenated at room temperature and atmospheric pressure over Adams catalyst (Pt0 2, 10 mg.) for 0.5 hr. The mixture was f i l t e r e d to remove the catalyst which was carefully washed with additional methanol (10 ml.). The combined f i l t r a t e s were concentrated under vacuum to give 103 as a colourless film (18 mg., 99%). UV.; \ m a x : 337, 324, 297, 286 (sh), 261, 248, 239. N.M.R. : 3.62 (quartet, 1 H, J = 6 Hz, C-1'H), 2.94 (quartet, 2 H, J = 7 Hz, C-3»CH 2), 2.50 (singlet, 3 H, C-l CH3), 2.38 (singlet, 6 H, N(CH 3) 2), 1.34 (doublet, 3 H, J = 6 Hz, C-1'CH3), 1.20 ( t r i p l e t , 3 H, J = 7 Hz, C-4'CH3). Mass spectrum M*", m/e = 280; main peaks : 265, 236, 235 (base peak), 220, 207-204. High resolution mass spectrum : Calc. for C i gH 2 4N 2 : 280.1938. Found: 280.1923. Attempts were made to further characterize compound 103 as i t s methiodide 113, (see following experiment). l-Methyl-2-ethyl-3-(a-trimethylammonioethyl)carbazole iodide (111). The hydrogenated compound 103 (18 mg., 6.4 x 10"5 mole) dissolved in chloroform methanol 1:1 was reacted with an excess of methyl iodide at 0° for 12 hr. The reaction mixture was concentrated to give a granular o i l (27 mg.) consisting of three products. The desired methiodide 111 was obtained in p a r t i a l l y pure form by washing the crude product mixture with chloroform (2 x 10 ml.). The partually purified methiodide i l l was obtained as an opaque soli d (15 mg., 55%). No physical data was obtained for this compound. - 157 -The chloroform washings were combined and concentrated to a yellow o i l . The major non-polar component 113 was isolated (12 mg.) by elution of an alumina column (5 gm.) with chloroform. N.M.R. (60 MHz) : 3.72 (quartet, 1 H, J = 6 Hz, C-l'H), 3.26 (singlet, 3 H, 0CH3), 2.50 (singlet, 3 H, C-l CH 3), 1.52 (doublet, 3 H, J = 6 Hz, C-1'CH_). l-Methyl-2-ethyl-3-vinylcarbazole (112) . Methiodide 111 (15 mg., 3.60 x 10"5 mole) and sodium hydride (15 mg.) in dry dimethylformamide was heated at 100° for 2 min. after which time the excess hydride was destroyed by the careful addition of water. The reaction mixture was then diluted with water (75 ml.) and the resultant white suspension was extracted with ether (4 x 30 ml.). The combined ether layers were washed with water (3 x 30 ml.), dried over sodium sulphate, and concentrated to give olefin 112 as a brown o i l (7 mg.). The crude product was purified by chroma-tography on alumina. By elution with benzene 112 was obtained as a light •brown o i l ( 4 mg.). UV.; X ^ : 3 2 5 < 295(sh), 280, 267{sh), 241. N.M.R. : 5. 6 9 ( J 1 , j 2 l = 17 Hz, J 2» > 2» = 2Hz., C-2' H trans>, 5.28( J-j, 2,= 11 Hz, J2',2' = 2 H z » C " 2 ' H c i s ) ' 2 - 4 6 ( s i n g l e t , C-l CH 3). Mass spectrum: M+, m/e = 235; main peaks : 220, 204. High resolution mass spectrum: Calc. for C H 1 7N : 235.1360. Found : 235.1380. - 158 -Salsolidine methiodide (115). 144 A mixture of salsolidine (100) (3.0 gm., 1.71 x 10"2 mole), methyl iodide (28 gm.), and aqueous sodium carbonate (6 gm. in 40 ml.) was warmed under gentle reflux overnight. The reaction mixture was then concentrated to dryness and taken up in water (50 ml.) whereupon tan coloured crystals immediately precipitated (2.3 gm.). The aqueous mixture was basified with 5% sodium hydroxide and extracted with chloroform (3 x 30 ml.), The com-: bined chloroform fractions were dried over sodium sulphate, and concentrated to an amber coloured foam. By recrystallization from methanol-ether beige crystals of 115 were obtained (2.20 gm., 35%), m.p. 205° ( l i t . , m.p. 229-235*4,) Found: C, 46.13; H, 6.10; N, 3.59. Calc. for C 1 4H 2 20 2NI : C, 46.41; H, 6.07; N, 3.87, l-(a-dimethylaminoethyl)-2-vinyl-4,5-dimethoxybenzene (116), , Salsolidine methiodide (115) (50 mg., 1.38 x 10"4 mole) was suspended in a 10% sodium hydroxide in 95% ethanol solution (30 ml.) and refluxed for 1.5 hr. The reaction mixture was then concentrated to a paste, diluted with water (75 ml.) and extracted with ether (3 x 30 ml.). The combined ether fractions were dried oyer sodium sulphate, and concentrated to give a light brown o i l (20 mg,, 61%). UV.; : 310 (sh), 290, 262, N.M.R,: 7.23 (doublet of doublets, 1 H, J J I ^ I = 16 Hz, J 3 , < 4 , = 10 Hz, C-3'H)f 7,02, 6.94 (2 singlets, 1 H each, C-3, C-6 H), 5.50 (doublet of doublets, 1 H, J31 41 = 16 Hz, J 4 i j 4 i = 2 Hz, C-4'H), 5.16 (doublet of doublets, 1 H, J j , 4 , «= 10 Hz, J 4 , 4 , - 2 Hz, C-4'M), 3.96 (singlet, 6 H, 0CH3), 3.50 (quartet, - 159 -1 H, J = 7 Hz, C-l'H), 2.33 (singlet, 6 H, N(CH 3) 2), 1-35 (doublet, 3 H, J = 6 Hz, C-l* CH 3), Reduction of (116) to 1-(a-dimethylaminoethyl) -2-ethyl-4,5-dimethoxybenzene  and formation of methiodide 117 . Olefin 116 (64 mg., 2.72 x 10"4 mole) dissolved in methanol (10 ml.) was hydrogenated at room temperature and atmospheric pressure over Adams catalyst (Pt02, 10 mg.) for 0.5 hr. The mixture was f i l t e r e d to remove the catalyst which was carefully washed with additional methanol (10 ml.). The combined f i l t r a t e s were concentrated under vacuum to give 117 as a colourless o i l (64 mg. 98%). UV; \ v : 280. N.M.R.: 7.10 (singlet, III, C-6H),6.64 (singlet, 1 H , C-3 H), 3.90 (split singlet, 6 H, 0CH3), 3.43 (quartet, 1 H, J = 7 Hz, C-l'H), 2.65 (quartet, 2 H, J = 7 Hz, Ar-CH2CH3), 2.33 (singlet, 6 H, N(CH 3) 2), 1.40 (doublet, 3 H, J = 6 Hz, C-1'CH3 (overlapped by t r i p l e t at 1.13) ), 1.13 ( t r i -plet, 3 H , J = 6 Hz, ArO^CHj (overlapped by doublet at 1.40) ). The hydrogenated compound (64 ml.) was converted to i t s corresponding methiodide 117 by reaction with an excess of methyl iodide i n methanol. Recry-s t a l l i z a t i o n from methanol, m.p. 151-153°. Found : C, 47.97; H, 6.20; N, 3.29. Calc. for C 1 5H 2 6N0 2I : C, 47.59; H, 6.88; N, 3.70 (analysis unsatisfactory). Attempted formation of 1-V inyl-2-ethyl-4,5-dimethoxybenzene (H8). The methiodide (30 mg., 8.24 x 10" ^ mole) was refluxed in diethyl ketone (10 ml.) for 2 hr. The solvent was then removed under vacuum to give - 160 -an amber o i l . The crude product was d i s t i l l e d at 120-180° at 0.05 mm. A yellow film was obtained which was dissolved in ether (25 ml.) and washed with 10% sodium thiosulphate solution to give a colourless film of product (yield undetermined). From separate experiments the N.M.R. spectrum for the crude and purified reaction product showed that the starting material had totally decomposed. Thermal Pyrolysis of Guatambuine Methiodide (86). Guatambuine methiodide (about 10 mg.) was heated under vacuum for 12 hr. at 0.05 mm. and 200°C. A small yield of yellowish coloured crystals sublimed during this period, m.p. 200-210°, The N.M.R. spectrum was identical to that obtained for guatambuine (25) . - 161 -l-Methyl-3-(ot-hydroxyethyl)-9-benzylcarbazole ( 2 2 2 ) . Carbazole acetate 223 (100 mg., 2,80 x 10~4 mole) was partitioned between a two phase medium consisting of t-butanol (5 ml.) and 20% aqueous sodium hydroxide (5 ml.), and heated at 100° with vigorous s t i r r i n g for 1 hr. The reaction mixture was then concentrated to remove the t-butanol, diluted with water (75 ml.) and extracted with ether (3 x 30 ml.). The combined ether layers were washed with water, dried over sodium sulphate and concentrated to give a colourless foam. The crude product was purified by column chromatography on alumina (10 gm.). Trace amounts of unreacted carbazole acetate and other non polar contaminants were removed by elution with benzene. By subsequent elution with chloroform and concen-tration, the desired alcohol 222 was obtained as a colourless foam (70 mg., 79%). UV.; \ „ : 342, 327, 293, 287, 283, 263, 237, N.M.R. : 8.08 (doublet 1 H, J = 2 Hz, C-5H), 7.98 (multiplet, 1 H, C-4H), 7.2, 6.95 (2 multiplets, 9 H, aromatic), 5.66 (singlet, 2 H, N-CH2-C6H5), 5.00 (quartet, 1 H, J = 6 Hz ArCH(0H)CH3), 2.54 (singlet, 3 H, C-l CH 3), 1,97 (singlet, 1 H, OH), 1.54 (doublet, 3 H, ArCH(OH)CH3). Mass spectrum ; M , m/e = 315 (base peak); main peaks : 300, 297. High resolution mass spectrum : calc. for C 2 2 H 2 1 N O : 315.1623. Found: 315.1632. l-Methyl-3-vinyl-9-benzylcarbazole ft23 )• Carbazole alcohol 222 (75 mg., 2,38 x 10"4 mole) and toluenesulphonyl chloride (58 mg.,.1.3 equiv.) in pyridine (8 ml.) was heated at reflux temperature for 6 hr. The pyridine was then removed under vacuum and the - 162 -residue was dissolved in chloroform (75 ml.) and washed successively with dilute hydrochloric acid, 5% sodium hydroxide and water. The chloroform layer was dried over sodium sulphate, and concentrated to give a dark coloured o i l (54 mg.). The crude product was purified by f i l t r a t i o n through a column of alumina (3 gm.). By elution with chloroform the ol e f i n 123 was obtained as an amber coloured o i l (42 mg., 60%). UV.; ^  : 350, 336, 279, 270 (sh), 241. N.M.R.: 8.14 (multiplet, 1 H, C-5H), 8.05 (multiplet ( s p l i t singlet), 1 H, C-4 H), 6.90 (doublet of doublets, C-l'H, obscured by aromatic signals), 5.78 (doublet of doublets, 1 H, 2 , = 17 Hz, J 2 , 2 , = 1.5 Hz, C-2'H trans (partly obscured by 5.72 singlet), 5.72 (singlet, 2 H, Ar-CH2-N), 5.21 (doublet of doublets, 1 H, J V t 2 1 = 11 Hz, J 2 , > 2 , = 1.5 Hz, C-2'H c i s ) , 2.60 (singlet, 3 H, C-l CH 3). - 163 -1,4-Dimethyl-2-(B-(N,N-methylacetylamino)ethyl)-3-acetoxymethylcarbazole (12l) N-Methyltetrahydroellipticine (26) (50 mg. 1.89 x 10~4 mole) and anhydrous sodium acetate (20 mg.) in acetic anhydride (3 ml.) was heated at reflux tempera-ture for 8 hr. The cooled reaction mixture was then diluted with 5% aqueous sodium hydroxide solution (75 ml.) and stirred for 15 min. after which time the resultant suspension was extracted with chloroform (3 x 30 ml.). The combined chloroform fractions were dried over sodium sulphate, and concentrated to a tan coloured glass which readily crystallized from benzene to give colourless cry-stals of i'2-i (53 mg., 76%). UV.; A m a x : 335 (3.50), 321 (3.56), 293 (4.13), 285 (3.95), 263 (4.49), 250 (4.70), 243 (4.74). IR. ( C H C I 3 ) : 1730, 1635 cm"1. N.M.R.: 5.50, 5.46 (2 x singlet, 2 H, ArCH2OAc), 3.50 (multiplet, 2 H, C-2'CH2), 3.05, 3.00 (2 x singlet, 3 H, NCH3), 2.88 (singlet, 3 H, C - 4 C H 3 ) , 2.60, 2.52 (2 x singlet, 3 H, C-l CH 3), 2.08, 1.98 (singlet, 3 H, C - l 1 N C 0 C H 3 ) , 2.03 (singlet, 3 H, N C O C H 3 ) . Mass spectrum: M+, m/e = 366; main peaks: 306, 280, 263, 233, 222-1. High reso-lution mass spectrum: Calc. for C 2 2H 2 6N 20 3: 366.1943. Found: 366.1956. Found: C, 72.20; H, 7.11; N, 7.20. Calc. for C 2 2H 2 6N 20 3: C, 72.11; H, 7.18; N, 7.64. 1,4-Dimethyl-2- (8 -(N,N-methylacetylamino)ethyl)-3-hydroxymethylcarbazole (122) Carbazole acetate 121 (50 mg., 1.36 x 10"4 mole) was partitioned between a two phase medium consisting of t-butanol (5 ml.) and 20% aqueous sodium hydroxide (5 ml.), and heated at 100°C with vigorous s t i r r i n g for 1 hr. The reaction mix-ture was then concentrated to remove the t-butanol, diluted with water (75 ml.), and extracted with ether (3 x 30 ml.). The combined ether fractions were washed with water, dried over sodium sulphate and concentrated to give 122 as a colour-less solid. The crude product was purified by preparative layer chromatography - 164 -on alumina (1 mm., EtOAc:MeOH 20%). UV.; J j ^ : 335, 321, 293, 285, 263, 250, 243. IR.: 1635 cm-1, N.M.R.: 6.20, 5.00 (multiplets, 2 H, -CH2OH). Mass spectrum: M*, m/e = 324; main peak: 308. High resolution mass spectrum: Calc. for C 2 0 H 2 4 N 2 O 2 : 324.1837 . Found: 324.1814. l-Methyl-2-(B -(N,N-methylacetylamino)ethyl)-3-(ct -acetoxyethyl) carbazole(124) • Guatambuine (60 mg., 2.27 x 10~4 mole) was dissolved in pyridine (4 ml.) containing acetic anhydride (2 ml.) and Heated at reflux temperature for 2 hr. The solvent was then removed under high vacuum to give a dark red o i l . The crude product was p a r t i a l l y purified by column chromatography on alumina (3 gm.). By elution with chloroform 124 was obtained as an amber coloured o i l (58 mg., 70%). UV.; X^y.: 335, 323, 295, 285, 260, 247, 239. IR.: 1710, 1620. N.M.R.: 6.34 (multiplet, 1 H, C-l'H), 3.50 (multiplet, 2 H, C-3' CH 2), 3.00 (singlet, 3 H N-CH3), 2.48, 2.40 (2 x singlet, 3 H, C-l CH 3), 2.08 (Singlet, 3 H, O C O C H 3 ) , 2.08 1.98 (2 x singlet, 3 H, N C O C H 3 ) . Mass spectrum: M+, m/e = 366, main peaks: 306 234, 233, 221, 220 (base peak), 205,204. 1-Methy 1-2-(B-(N,N-methylacetyland.no)ethyl) -3-vinylcarbazole (125 ) . Guatambuine (25) (60 mg., 2.27 x 10~ 4 mole) and anhydrous sodium acetate i n acetic anhydride was heated at reflux temperature for 10 hr. The cooled reaction mixture was then diluted with 5% aqueous sodium hydroxide solution (75 ml.) and stirred for 15 min. after which time the resultant suspension was extracted with chloroform (3 x 30 ml.). The combined chloroform fractions were dried over sodium sulphate, and concentrated to a light brown o i l . The crude reaction mix-ture was purified by column chromatography on alumina (5 gm.). By elution with - 165 -chloroform the olefin 125 was obtained as a faintly coloured o i l (45 mg., 64%), UV; Xj^xt 325 (broad hump), 295 (sh), 285-275 (broad peak), 262, 240, IR,: 1630 cm"1. N.M.R.: 5.72 (doublet of doublets, 1 H, Jii ) 2» = 1 7 H z » J2',2' 2 Hz, C-2'H trans.), 5.37, 5.34 (2 x doublet of doublets, 1 H, J i i ^ i = 1 1 H z J 2 t >2« = 2 Hz, C-2' H c i s ) , 2.50 (multiplet, 2 H, C-3' CH 2), 2.20 (multiplet, 2 H, C-4' CH 2), 3.00, 2.92 (2 x singlet, 3 H, NCH3), 2.58, 2.52 (2 x singlet, 3 H, C-l CH 3), 2.10, 1.96 (doublet(J = 2 Hz) and singlet, 3 H, NC0CH3), Mass, spectrum: M+, m/e = 306; main peaks: 234, 220 (base peak), 204, High resolu-tion mass spectrum: Calc, for C 2 ( )H 2 2N 20: 306,1732, Found: 306,1736 Formation of the Enamine (126), and Reaction with Dimethylsulphate. In a small scale experiment the enamine 126 was found to be precipitated as an amorphous yellow s o l i d after the addition of 20% aq, sodium hydroxide solution (10 ml.) to a water solution (10 ml.) of the iminium cation (about 3 mg.). UV. (20% aq. NaOH) (Figure 38); A,^: 365, 335, 310, 294, 280, 270 (sh), 243, 237, Addition of dimethylsulphate (1 ml.) to the basic solution of the enamine and heating at 100 for 10 min, resulted in the total disappearance of the yellow colouration, UV, (Figure 38) l ^ ^ : 278, 245 (sh), 235, - 166 -INTRODUCTION: (PART III) During the course of experiments on the biosynthesis and degradation of the pyridocarbazole alkaloids olivacine (16) and guatambuine (25), i t became apparent that for the continuation of the work that larger quantities of these compounds would be required than were available from plant extracts 124 of Aspidosperma australe. Attention was, therefore, directed toward the development of an e f f i c i e n t synthesis of these two alkaloids adaptable to large scale preparation (1 - 10 gm.). Both the olivacine (16) and e l l i p t i c i n e (17) systems have been previously synthesized as a consequence of two separate interests. In the early sixties 172-6 synthetic corroboration was necessary following their isolation from various Aspidosperma and Ochrosia plants, and the elucidation of their novel tetracyclic structure. The subsequent determination that olivacine (16), e l l i p t i c i n e (17) and their analogs exhibited anti-tumor a c t i v i t y , led again to 177-180 the development of syntheses of these alkaloids. The approach adopted in the present synthesis of olivacine (16) and guatambuine (25) was designed in view of what was known about the chemistry of these molecules. It i s , therefore, necessary to b r i e f l y review the relevant details of the established syntheses. 172 For purposes of structure elucidation Schmutz and Wittwer (1960) were the f i r s t to publish a synthesis of olivacine (16). They attacked the problem of construction of the complex tetracyclic ring system by dividing the synthesis into two distinct stages. The f i r s t stage was the synthesis of - 167 -an appropriately substituted t r i c y c l i c carbazole intermediate containing the "A, B, and C" rings of olivacine (16) . The second stage was the 11 construction of the fourth or (4,3-b) fused pyrido "D" ring. The f i r s t stage carbazole intermediate required a methyl group in the 1-position and a versatile group in either the 2- or 3- position which could subsequently be elaborated to give the remaining "D" ring. Bearing in mind the proposed design of the "D" ring synthetic efforts were aimed at producing as the intermediate, l-methylcarbazole-2-carboxylic acid chloride (135). This compound was prepared in six steps from 2-methyl-3-aminobenzonitile (132) (Figure 40). By either of two well known condensation reactions, the modified Bischler or the Borsche synthesis, the starting material 132 was converted to 7-cyano-8-methyl-l,2,3,4-tetrahydrocarbazole (133). Hydrolysis of the cyano group and subsequent esterification, followed by dehydrogenation of the saturated ring gave the carbazole ester 134. By reaction with thionyl chloride the carbazole ester 134 was transformed into the desired acid chloride 135. Wolff rearrangement of the diazoketone 136 formed by reaction of the acid chloride 135 with diazomethane produced the corresponding homologous amide 137 . This amide 137 was transformed into the N-acetyl derivative 140 by dehydration to the n i t r i l e 138 with phosphorus oxychloride followed by reduction with Raney nickel to the amine 139 and subsequent acetylation with acetic anhydride in pyridine. Bischler cyclization of the amide 140 to the imine, 3,4-dihydro olivacine (141), completed the synthesis of the tetracyclic olivacine skeleton. The imine 141 was easily converted to guatambuine (25) by formation of the methiodide salt 142 and reduction over platinum oxide. Alternatively the imine 1 4 1 was - 168 -(25) (16) gure 40. Synthesis of Olivacine (16) and Guatambuine (25) by Schmutz and Wittwer (1960). - 169 -readily dehydrogenated with palladium on charcoal to give olivacine (16). The cyclization of 140 was aided by the electronic influence of the indole nitrogen para to the C-3 position in the "C" ring. A stable resonance contributing structure can be drawn depicting increased electron density centered at the C-3 position which enhances the attack on the carbonyl carbon of the amide. (140) The synthesis has the advantage that i t can for the most part be adapted to large scale preparation of either guatambuine (25) or olivacine (16), The construction of the "D" ring however suffered from several drawbacks, i t was both indirect and necessitated the use of large quantities of diazomethane, The major drawback to the overall sequence was however, i t s excessive length, seventeen steps in total are involved when i t i s considered that i t was necessary to f i r s t synthesize the starting material 132 from o-toluidine (143),* - 170 -A second synthesis of olivacine (16) was published by Wenkert and Dave 173 (1962) shortly after the work by the Swiss group. As with their predecessors they aimed at the synthesis of an intermediate carbazole 152 which possessed an aldehyde functionality at the C-2 position required for the elaboration of the "D" ring. A considerably different approach for the synthesis of this intermediate was chosen however, u t i l i z i n g 1-ketotetrahydrocarbazole (144) as the starting material, this compound being readily available by Fisher indole synthesis. The ketone group of this molecule served the dual role of directing the course of functionalization onto the C-2 position and subsequently the preferential methylation onto the C-l position. The two schemes that were developed are presented in figure 41. Considering Scheme A, sodium ethoxide induced acylation of 1-ketotetra-hydrocarbazole (144) with ethyl oxalate gave the ethoxalyl derivative 145 (R=H), which on acetylation yielded the diester 145 (R=0Ac). Zinc and acetic acid reduction of either the diester 145 (R=0Ac) or the a,3-unsaturated keto-ester 146 (obtained by partial hydrogenation of the diester) produced the saturated keto-ester 147. Reaction of the keto-ester 147 with methyllithium gave two products 148 and 149, which on palladium on charcoal dehydrogenation led to the corresponding carbazoles 150 and 151. Ozonolysis of the o l e f i n i c side chain of compound 150 gave the carbazole aldehyde 152, the v i t a l intermediate for the total synthesis. In the study of the haloform reaction on the acetone side chain of carbazole 151 i t was hoped that the presence of the C-methyl group would provide sufficient s t e r i c hindrance to prevent reaction at the benzylic carbon over reaction at the methyl carbon thus giving rise to the - 171 -(152) (153) Figure 41. Scheme A, Synthesis of Olivacine (16) According to Wenkert and Dave (1962). - 172 -(144) NH2OH (159) HCOOEt O R (154) R = H = i-pr o (158) PCI* pyr Base 79% (141) CH 3Li 72% (152) (155) CHO CHO Pd/C 78% (16) Figure 41. Scheme B, Synthesis of Olivacine According to Wenkert § Dave (1962) acetic acid derivative. However in analogy with phenylacetone the corresponding carbazole carbocylic acid 153 was obtained. In view of the fact that carbazoles with only one carbon side chains were available from the above sequence, i t was decided to study the same reaction with a one-carbon acylating agent, Scheme B_. Condensation of 1-ketotetrahydrocarbazole (144) with ethyl formate gave the formyl derivative 154 (=H) . Reaction with i-propyl iodide produced the i-propyl ether 154 (R=i-pr), which on treatment with methyllithium yielded the 1,2-dihydrocarbazole aldehyde 155 . Upon dehydrogenation of this aldehyde 155, i t was surprising to observe that the major isolated product was 1,2-dimethylcarba-zole (156). Equally surprising was the extreme ease with which the dehydrogen-ation could be carried out, i t was found that by merely heating a benzene solution - 173 of the 1,2-dihydrocarbazole aldehyde 155 in the presence of palladium on charcoal for one-half hour gave the desired carbazole aldehyde 152 . However, even in this extremely mild case, small amounts of the 1,2-dimethylcarbazole (156) was produced. It is probable that the f a c i l i t y of the reaction was due to a rapid disproportionation of the 1,2-dihydrocarbazole aldehyde 155 to the. carbazole alcohol 157 prior to i t s reduction to the 1,2-dimethyl compound 156 or reoxidation to the carbazole aldehyde 152 (isolated in small quantities). 182 Such disproportionation reactions have been observed previously although as yet not much is recorded regarding the energetics of the process. CHO C H O (155) (156) The "D" ring was constructed by base catalyzed condensation of the inter-mediate aldehyde 152 with acetone to give the chalcone 158. Hydrogenation of the latter and reaction with hydroxylamine gave the oxime of the dihydrochalcone 159. Beckmann rearrangement of the oxime 28 occurred with the concomitant cyclization to give the imine, 3,4-dihydroolivacine (141). Dehydrogenation of the imine 141 yielded olivacine (16). The synthesis developed by Wenkert's group was efficient in that olivacine (16) could be obtained i n only nine steps from starting material. However, i t was adapted to only small scale preparations. A further synthesis of olivacine and i t s analogs has been developed by 177 Mosher et al (1966) as part of a program on cancer chemotherapy. This synthesis represents a refinement of the latter stages of the scheme developed 172 by Schmutz and Wittwer (1960) . Instead of u t i l i z i n g the Wolff rearrangement - 174 -COOCH. (132) LAH 92% (134) ( P y r ) 2 C r 0 3 (157) IHH2 A c 2 ° (160) (139) Pd/C POC13-toluene (16) (141) Figure 42. Synthesis of Olivacine (16) by Mosher ejt al_. (1966). 176 they used the nitromethane condensation, developed by Govindachari et^ a l in the e l l i p t i c i n e (17) series as a more direct means of constructing the "D" ring, In essence the synthesis represents the better halves of two other syntheses united together to give the f i r s t synthesis of olivacine (16) adaptable to large scale preparation (Figure 42). The intermediate carbazole aldehyde 152 was obtained by reduction of the ester 134 followed by Sarett oxidation of the alcohol 157 . Condensation - 175 -of the aldehyde 152 with nitre-methane led to the nitrostyrene 160 in high yield, which on lithium aluminum hydride reduction produced the amine 139 , again i n high yi e l d . Conversion of the corresponding amide 140 to the imine 172 141 was again carried out according to the scheme of Schmutz and Wittwer, Olivacine (16) was immediately available from this imine Recently some exploratory work on the synthesis of the pyridocarbazole 184-5 skeleton has been published by T. Kametani et al_ (1974) . For several years they have been u t i l i z i n g benzocyclobutene derivatives as precursors for the 183 synthesis of isoquinoline alkaloids , and are at present attempting to expand the scope of their benzocyclobutene approach to include the synthesis of indole alkaloids of the olivacine (16) and e l l i p t i c i n e (17) type (Figure 43). It has 184 been found that on intermolecular cycloaddition of indole (161) with the benzocyclobutene analog 4,5-dibromomethyl-3- hydroxy-2-methylpyridine hydrobroraii (162) in dimethylformamide for four hours, followed by acetylation gave the expected dihydropyridocarbazole derivative 163 in 4% y i e l d and i t s structural isomer 164 in 15% yield. Subsequent dehydrogenation produced the olivacine type compound 165 and i t s corresponding isomer 166 . The various syntheses that have been described represent the current state of the synthesis of the olivacine (16) skeleton. The design of the present synthesis of this system was not however derived solely from the chemistry of olivacine (16) but to a considerable extent from a study of the synthesis of the closely related pyridocarbazole, e l l i p t i c i n e (17). It i s relevant, therefore, to discuss the chemistry of the e l l i p t i c i n e (17) series since the difference in the position of one methyl group has resulted i n a number of quite different approaches to the synthesis of this system. - 176 -(165) Figure 43 . The Kametani Benzocyclobutene Analog Approach (1975) - 177 -(17) Figure 44 . Woodward (1959) Synthesis of E l l i p t i c i n e (17), The f i r s t synthesis of e l l i p t i c i n e (17) was achieved i n a somewhat 174 unusual and elegant manner by Woodward and co-workers (1959) as part of structure elucidation work on Aspidosperma alkaloids. They were able to incorporate both the desired methyl functionality in the C-5 position of the "C" ring and the complete pyrido - "D" ring in one step through the condensa-tion of indole (161) with 3-acetyl pyridine to give 1,1-bis-(3_indolyl)1-1 (3-pyridyl)ethane (167) as ill u s t r a t e d in figure 44. The pyridine ring of this dimeric product was reduced with zinc in acetic anhydride to the - 178 -(173) Figure 45 . The Synthesis of E l l i p t i c i n e (17) by Cranwell and Saxton (1962) 1,4-diacetylpyridine derivative 168 the 4-acetyl group of which was ideally set-up for condensation with the electron rich 2-position of the indole molecule in a manner identical to the i n i t i a l condensation with 3-acetyl pyridine. The condensation was effected by pyrolysis under vacuum at 200°C. Unfortunately however, even though the process was intramolecular, e l l i p t i c i n e (17) was isolated in only a 2% yield. This approach was, therefore, incompatible with any scheme for large scale preparation of the alkaloid. A different and more elementary approach was devised several years later 175 by Cranwell and Saxton (1962) . Following the theme developed for the synthesis of olivacine (16) they considered a two stage synthesis of e l l i p -ticine (17) to avoid the necessity of a lengthy Fisher indole synthesis. By - 179 -reaction of indole (161) with hexane-2,5-dione in ethanol-HCl 1,4-dimethyl-carbazole (169) was obtained in 36% y i e l d . Subsequent formylation under con-troll e d conditions gave the desired l,4-dimethyl-3- formyl carbazole (170) as the predominant product. The direction of formylation was influenced by the presence of the methyl groups in the "C" ring (Figure 45). The construction of the "D" ring from the aldehyde 170 required the direct condensation of nitrogen onto the carbonyl functionality with subsequent provision for a ring closure. Such requirements form the basis of a well known Pomeranz-Fritsch reaction between an aldehyde and an amino acetal. Reaction of the aldehyde 170 with aminoacetaldehyde diethylacetal (171) yielded in 85% yi e l d the requisite imine acetal 172. It was found however, that reaction of this imine acetal 172 under any of the normally employed acidic conditions, i . e . sulphuric acid, polyphosphoric acid, boron trifluo r i d e etherate, or arsenic pentoxide failed to effect cyclization. Cyclization under the above conditions was also tried without success on the amino acetal 173 and on the corresponding acid derived from condensation of the aldehyde 170 with glycine. Cyclization was eventually achieved i n low yi e l d on the amino acetal 173 by refluxing in dry ethanol - HCl. The proposed product obtained 3,4-dihydroellipticine (191) was not purified but immediately dehydrogenated with palladium on charcoal to give e l l i p t i c i n e (17) in very low y i e l d . The failure of the cyclization process to proceed in a facile manner can be rationalized to be a consequence of the poor nucleophilic character - 180 -of the C-2 position of the carbazole skeleton. It may be seen that resonance structures invoking participation of the indole nitrogen in the build up of election density at C-2 are unfavourable due to the required disruption of the aromatic resonance of the "A" ring. On the basis of i t s design this synthesis was very practical, however the d i f f i c u l t i e s encountered in the cyclization step totally deplete i t s value, Almost simultaneously with the publication of the above work, Govindachari, 176 Rajappa and Sudarsanam (1963) reported a new synthesis of e l l i p t i c i n e (17). They u t i l i z e d the classical Fisher indole approach for the synthesis of appro^ priate carbazole intermediates. D i f f i c u l t i e s were encountered during two key stages of their sequence requiring in each instance complete redesign of the i approach. The three sequences^are~presented in figure 46, schemes A to C_. In scheme A, 2,5-dimethyl-4-nitrobenzoic acid (174) was converted to the tetrahydrocarbazole 175 which on dehydrogenation yielded the desired 1,4^ dimethyl carbazole-3-ester (176). The subsequent objective was to convert the ester 176 into the corresponding aldehyde 178 by lithium aluminum hydride reduction and oxidation of the alcohol 177. However, reaction with lithium aluminum hydride resulted unexpectedly in total reduction of the ester 176 to 1,3,4-trimethylcarbazole (179). Similar total reduction has been observed on - 181 -- 182 -- 183 -186 reaction of 3-acetyl and 3-formylindole with lithium aluminum hydride. This result can be rationalized by proposing that in the basic solution i t is the conjugate base that i s the reactive species, this intermediate being hydrogenolyzed. O " (176) Attempting to modify the synthesis, in order to alleviate the necessity of having to reduce a C-3 ester, the Scheme B_ was undertaken. In this approach 5-cyano-2-nitro-p-xylene (180 ) was converted via the appropriate arylhydrazine to 6-cyano-l,2,3,4-tetrahydro-5,8-dimethylcarbazole (181). Dehydrogenation of this compound and subsequent Raney nickel reduction of the n i t r i l e 182 gave the amine 183. i t was thought that condensation of the amine 183 with glyoxyl diethoxyacetal followed by Pomeranz-Fritsch cyclization of 184 would then y i e l d e l l i p t i c i n e (17) directly. However the same d i f f i -culties were encountered with this conversion as were experienced by Cranwell 175 and Saxton and the approach was abandoned. Scheme C_ involves the conversion of 3,6-dimethyl-2-nitrobenzaldehyde (185) to an intermediate, 6 -amino-2-cyano-p-xylene (186) very reminiscent of 172 the starting material u t i l i z e d by Schmutz 5 Wittwer (1960) . (Figure 40) Fisher indolization to the tetrahydrocarbazole 187 , hydrolysis of the cyano group, and subsequent dehydrogenation afforded l,4-dimethyl-2-carbomethoxy-carbazole (188). It was found that this C-2 ester could easily be reduced to - 184 -the corresponding alcohol 189 without the complication experienced in Scheme A, this again probably being a result of low electron density at C-2. Sarett oxidation of the alcohol 189 yielded the aldehyde intermediate 190.. The u i-nitrostyryl approach to the synthesis of the "D" ring developed by these workers has been discussed previously in conjunction with Mosher's 177 synthesis of olivacine (16) . The dehydrogenation of the imine, 3,4-dihydroellipticine Q.91 ) obtained in four steps from the aldehyde provided e l l i p t i c i n e (17). The f i r s t stage of the work portrayed in Scheme C parallels closely 172 that developed by Schmutz § Wittwer (1960), and consequently i t suffers from the same drawback, i t s excessive length. The nitromethane condensation reaction 177 however as Mosher realized was a very direct and highly e f f i c i e n t method for construction of the "D" ring. 178 Subsequently Dalton and co-workers (1967) were able to overcome the problem of the cyclization of the imine acetal 172 to produce e l l i p t i c i n e (17). 175 The synthetic scheme of Cranwell and Saxton's was adopted (Figure 45 ) with improvement of the i n i t i a l condensation step between indole (161) and hexane-2, 5-dione. By substituting p-toluenesulphonic acid-ethanol for this condensation i t was found that the yie l d could be increased to 51%, compared with the 36% previously obtained. It was found that the crucial cyclization of the imine acetal 172 could then be effected in 33-40% yields using o-phosphoric acid. 175 Contrary to the results of Cranwell and Saxton who claimed that ethanol - HC1 cyclization of the amino acetal 173 produced 3,4-dihydroellipticine (191) (not characterized) i t was found that when the amino acetal 173 was heated - 185 -in o-phosphoric acid that e l l i p t i c i n e (17) was isolated directly. It i s known that 1,2-dihydroisoquinolines produced by similar cyclizations undergo 187 disproportionation and oxidation to isoquinolines, so that the previous authors may have produced e l l i p t i c i n e (17) directly and their dehydrogenation step was probably unnecessary. This synthesis proved to be of considerable value for large scale pre-paration of e l l i p t i c i n e (17) due to i t s brevity although the yields were uniformly low. Recently there has been a revival of interest in the chemistry and biological activity of e l l i p t i c i n e (17). Two new syntheses have been reported 1 *7 A and both adopt the approach, f i r s t devised by Woodward (1959) , involving the condensation of an appropriate unit containing a preformed pyrido ring onto the indole nucleus. The general view taken with this strategy i s that ring closure of the MC" ring would require fewer conversions than are involved in the build up of the pyridine ring. 179 Kilminster § Sainsbury (1972) reinvestigated Woodward's condensation of 3-acetylpyridine onto indole (161). They preferred however to study the condensation into respective indolin-2 and -3-ones 192 and 193 . It was found that although i n i t i a l condensation products would form with ease the subsequent ring closure reaction could not be induced, presumably due to the inert character of the carbonyl groups. From these observations they were, however, successful i n synthesizing e l l i p t i c i n e (17) through condensation of a substituted 3-acetylpyridine with l-acetylindol-3-yl acetate (196) (Figure 47 ). The diacetyldihydropyridine 194 was prepared, oxidized to the acetyl - 186 -H H (192) (193) (17) (198) Figure 47 . Kilminster and Sainsbury (1972) Synthesis of E l l i p t i c i n e , pyridine 195, and condensed with the indole 196 affording a mixture of both geometrical isomers 197. Reduction of either isomer with sodium borohydride followed by acidification gave the compound 198 . This material when heated in aqueous hydrogen bromide, followed by neutra-lization and absorptions onto s i l i c a gel afforded e l l i p t i c i n e (17) in 40% y i e l d . This synthesis offers both a short and e f f i c i e n t route to e l l i p t i c i n e (17), 180 In the second synthesis, Le Goffic, Gouyette, and Ahond (1973) , proposed to unite the preformed "D" ring through a piperidone enamine conden-sation with an appropriately substituted gramine molecule. - 187 -(17) (200) Figure 48. The Synthesis of E l l i p t i c i n e (17) by Le Goffic, Goyette, and Ahond (1973). Consideration was i n i t i a l l y given to the condensation between the pyrrolidine enamine of l-benzyl-4-piperidone (199) and 2-ethyl gramine (200) (Figure 48). This would ideally set the stage for a two-step synthesis of e l l i p t i c i n e (17). However i t was found that the attempted condensation reaction gave only dimeric pxoducts and none of the desired material. To circumvent this problem the ethyl group was incorporated onto the ring "D" instead. This was accomplished by reaction of the condensation product 201 with sodium acetylide, epimeric tertiary alcohols 202 being isolated. The alcohols were reacted in the presence of formic anhydride to - 188 -give the desired tetracyclic carbazole product 203 possessing the proper orientation of the two methyl groups. Debenzylation § dehydrogenation over palladium on charcoal produced e l l i p t i c i n e (17) in 24% overall yield from indole (161). The excellent work done by the French group represents the culmination of the work to date in the f i e l d of the synthesis of the pyridocarbazoles olivacine (16) and e l l i p t i c i n e (17). - 189 -DISCUSSION - PART III Two distinct synthetic themes presided in the many syntheses of olivacine (16) and e l l i p t i c i n e (17) presented in the preceding introduction, the two-stage approach where a suitable t r i c y c l i c carbazole intermediate was f i r s t formed and subsequently elaborated into the tetracyclic pyridocar-bazole system, and the approach involving the condensation of a preformed "D" ring onto the indole nucleus, followed by ring closure to give the " C " ring and the complete tetracyclic structure. The former, two-stage approach was adopted in the design of the present syntheses of olivacine (16) and guatambuine (25). In this design consideration was taken of the fact that 177 the w-nitrostyryl approach u t i l i z e d by Mosher provided an e f f i c i e n t means of constructing the pyrido "D" ring of olivacine (16), whereas The Pomeranz-Fritsch approach adopted i n the e l l i p t i c i n e (17) series provided a considerably shorter but lower yielding method of pyrido ring construction. An eff i c i e n t synthesis of these two pyridocarbazole alkaloids could be obtained, therefore, i f a short 5 high yielding synthesis of either of the two formylcarbazole compounds l-methyl-2-formylcarbazole 0-52) or 1-methy 1-3-formylcarbazole (219) could be devised. These considerations formed the basis behind the developed syntheses of olivacine (16) and guatambuine (25) presented i n Sequences A § B. - 190 -Sequence A: 175 By analogy with Cranwell and Saxton's synthesis of e l l i p t i c i n e (17) the condensation reaction between a methyl ketone functionality and indole was adopted for the synthesis of the carbazole-2-aldehyde 152. Unlike e l l i p t i c i n e (17) however, the absence of a symmetrical methyl group substi-tution pattern in olivacine (16) necessitated the prior construction of an appropriately functionalized compound 204 where a subsequent methyl ketone condensation with C-2 of the indole moeity would comprise the second step in the construction of the carbazole skeleton. R = CHO = COOCH, (204) (205) An inspection of the structure of compound 204 immediately indicated that i t could be synthesized by the alkylation reaction of an appropriate 1,3-dicar-bonyl with a 3-substituted indole derivative like tryptophyl bromide (207). Methyl acetoacetate (205, R = COOCHj) was chosen as the 1,3-dicarbonyl component instead of the corresponding aldehydo compound (205, R = CHO) for reasons that the carbonyl of the carbomethoxy group i s less reactive towards nucleophilic attack than the carbonyl of the ketone, thus eliminating the competition between these two centers during the subsequent condensation reaction. Also, the trans-formation of the carbomethoxygroup into the desired aldehyde functionality could be effected via a published procedure. - 191 -The synthesis of compound (204, R = COOCH^) from commerically available tryptophol (206) and i t s cyclization with subsequent dehydrogenation to the carbazole ester 134 or i t s cyclization in the presence of a hydrogen acceptor reagent to give the carbazole ester 134 directly is presented in figure 49. The reduction of the ester 134 to the carbazole alcohol 157 and i t s subsequent oxidation to the desired l-methyl-2-formylcarbazole (152 ) is also depicted. Elaboration of this aldehyde 152 into both olivacine (16) and guatambuine (25) via the t*>-nitrostyryl approach proceeded in an anologous fashion to that 177 reported by Mosher e_t al_ (1966) The present synthesis of the carbazole ester 134 in three high yielding steps from the readily available starting material tryptophol (206 ) represents a considerable improvement in efficiency over the alternate nine-step Fisher indolization approach previously employed. The overall synthesis consisting of ten steps to olivacine (16) is thus more ef f i c i e n t than those previously 172-3,177 . „ , , reported and i s easily adaptable to the large-scale preparation of both 188 olivacine (16) and guatambuine (25). - 192 -Figure 49 . An Improved Synthesis of Olivacine (16) and Guatambuine (25), Sequence A. - 193 -Tryptophol (206) Tryptophol (206) was obtained both commercially and via a simple synthesis. Before entering into a discussion of the synthesis of olivacine (16) from this starting material, i t is perhaps advantageous to b r i e f l y discuss i t s synthesis 189,190-1 Several methods are reported, but for the purposes of the present work the reaction of indole (161) with oxalyl chloride (210) was considered to be the most adaptable to large-scale preparation (Figure 50 ). C P H (161) (206) cy^o 93% (210) LAH 60% (211) Me OH 92% (212) Figure 50. Oxalyl Chloride Route to Tryptophol (206). This synthesis follows quite closely in experimental detail with the 190 parallel synthesis of tryptamines by the same method . The glyoxyl chloride 211 was obtained as easily isolated bright yellow crystals in 93% y i e l d . Recrystallization proved unnecessary since the microanalysis was in agreement with the anticipated structure. The product proved to be moderately stable, not reacting perceptibly with water, and discolouring on standing in a i r only - 194 -after a period of several days. Reaction of the glyoxyl chloride 211 with methanol gave the corresponding ester 212 in 92% yie l d . Again, the microanalytical data supported the structural assignment. Reduction of the ester 212 i n the presence of an excess of lithium aluminum hydride gave tryptophol C206 ) in 60% y i e l d . Column chromatography on alumina was sufficient to remove base line contaminants present i n the crude product mixture. Tryptophol £06 ) was identified by i t s characteristic N.M.R. spectrum consisting of a pair of well separated tr i p l e t s centered at 62.96 (J = 6Hz) and 3.86 (J = 6Hz) for the C-l (CH20H) and C-2 (CH2) protons respectively. A sharp singlet was present at 61.53 for the hydroxy 1 proton..:. Tryptophol (206) could readily be converted to the bromide 207 by reaction 192 with phosphorus tribromide. Purification by recrystallization proved to be more d i f f i c u l t than expected, and the product was subsequently purified by column chromatography on alumina. The mass spectrum exhibited the expected doublet for the parent ion at m/e = 223 and 225 due to the equal abundance of the two bromine isotopes. The N.M.R. spectrum possessed a single complex multiplet centered at 43.6 for the four methylene hydrogens. The t r i p l e t at 6 3.0 for the methylene protons adjacent to the oxygen atom in tryptophol (206 ) has shifted upfield in the spectrum of the bromide. 3-Carbomethoxy-5-(3-indolyl) -2-pentanone (204 ) Several different reaction conditions were studied for the alkylation reaction i n order to obtain a reproducible result. A 2:1 ratio of the sodium - 195 -salt of methyl acetoacetate (205) to tryptophol bromide (207) was normally employed to compensate for any neutralization of the anion that might occur as a result of removal of the indolic hydrogen atom. Conducting the condensation in refluxing tetrahydrofuran proved to be generally unsuccessful. Varying periods of reflux from three to fifteen hours were tried i n an effort to complete the reaction, however, the N.M.R. and mass spectral data indicated that normally only a mixture of starting materials was obtained. 193 Zaugg e_t al_ (1960) indicated dimethylformamide to be a superior solvent for the alkylation of the anion of ethyl acetoacetate with a series of primary and secondary aliphatic halides. Usual yields of about 70% were obtained when the reaction was conducted at 100°- for three to four hours. In a preliminary experiment with tryptophol bromide (207) using dimethylformamide the reaction was permitted to remain at 100° for fifteen hours. The N.M.R. spectrum of the isolated product showed several expected absorptions; a singlet due to the ketone methyl at 52.20 and a group of distorted t r i p l e t s at 6 2.3 and 2.8 due to the methylene hydrogen atoms. However, concomitant with these observations was the absence of a singlet for the methyl hydrogens of the carbomethoxy group, indicating that decarboxylation had accompanied the cyclization process. The mass spectrum supported this interpretation, having a parent peak at m/fe = 201 consistent with an alkylation product lacking the ester functionality. This phenomenon was thought to be a result of excessive reaction time and that by shortening the reaction period the undesirable side reaction might be eliminated. - 196 -Thus, i t was found that conducting the reaction in dimethylformamide at 100° for three to four hours gave one predominant product as depicted from T.L.C. By column chromatography on alumina the desired product was isolated as a pure yellow o i l . The yields obtained were found to be very dependent on the scale of the chromatography, the y i e l d varying from 75% to 50% on proceeding from a small scale (1-2 gm.) to large scale (10-20 gm.) purification. No attempts were made to further optimize the aklylation reaction in , terms of time or temperature of reaction, although i t was known that the reactivity of the corresponding tryptophyl tosylate (213) towards nucleophilic displacement is considerably greater than that of simple primary al i p h a t i c 194 tosylates. This rate enhancement has been demonstrated to be the result of a unique a b i l i t y of the 3-position of the indole ring to participate in the displacement of the leaving group through formation of the spiro-indolenine 214. c n f % ~ a / — - O c n H V . H H (213) (214) Subsequent solvolysis or reaction of the cyclopropyl intermediate with a nucleophile being a facile step, thereby providing the substituted tryptophol derivative. - 197 -- 198 -Under reaction conditions where an appropriate nucleophile i s either 111 very weak or absent, the spiro indolenine 214 has been isolated. The reaction conditions employed for the present alkylation were such, however, that i t s presence would not be detected as a reaction product. Conclusive identification of the alkylation product 204 was made on the basis of it s spectral data. The N.M.R. spectrum (Figure 51) was very characteristic. Two singlets for methyl protons were readily assigned, one at 6 2.12 for the ketone methyl, the other at 63.66 for the carbomethoxy methyl group. Both signals were shifted s l i g h t l y upfield from those obtained for the corresponding groups (6 2.20 and 3.76) in methyl acetoacetate. Three distinct sets of distorted triplets were assignable, one for each of the two methylene groups 6 2.28, (C-4) and 62.76 (J = 7Hz,C-5), and the third for the methine hydrogen at 6 3.48 (J = 7Hz,C-3). Of particular note was the presence of a spike at 6 6.94, integrating for the one hydrogen atom at the C-2 position of the indole ring. This signal indicated that under the reaction condition the alky-lation product had remained in tact without subsequent cyclization. Further confirmation that cyclization had not occurred was derived from the U.V. spectrum which showed a typical indole absorption. The I.R. spectrum con-tained two carbonyl absorptions at 1750 and 1725 cm _l for the ester and ketone carbonyls respectively. In the low resolution mass spectrum the parent peak at m/e = 259 was relatively intense (25%) . Important fragmentations were seen at m/e = 228, 186, 184, 144, 143, and 130. A possible mode of fragmentation i s presented in Figure 52. The high resolution spectrum was within accepted limits and corroborates the parent peak at m/fe = 259.1212 (calc. 259.1207) for C 1 5 H 1 7 O 3 N . - 199 -It further corroborates the formulation of the fragment peaks giving the expected compositions for each fragment. CXif H CC9H8N) 130.0656 ^OCH (C 1 5H 1 70 3N) 259.1207 C C14 H14°2 N ) 228.0918 H (C 1 0H 1 0N) 144.0812 (C 1 0H 9N) 143.0734 0 ^ H ( C U H 1 0 N ) 1S6.0812 (C 1 2H 1 2ON) 186.0918 gure 52. Plausable Mass Spectral Fragmentation Pattern for the Alkylation Product 204. - 200 -Cyclization Reaction 178 The conditions employed by Dalton et_ al_ in the similar condensation of a methyl ketone center with the C-2 position of indole presented an obvious f i r s t choice for the cyclization of the alkylation product 204. It was consequently found that a reaction did occur with complete consumption of the starting material when the alkylation product 204 was treated with toluene sulfonic acid in refluxing ethanol. T.L.C, however, indicated the presence of an equal mixture of two products. These results were unexpected in that the spectral data for the products were inconsistent with the a n t i c i -pated product, l-methyl-2-carbomethoxy-3,4-dihydrocarbazole (208) • Alternate conditions were sought i n an effort to control the course of the cyclization so as to produce the desired compound 208 as the exclusive product. Bischler-Napieralski cyclization conditions, phosphorous oxychloride in toluene were thus studied. Reaction conditions ranging from refluxing i n toluene for one hour to s t i r r i n g at room temperature for fifteen minutes were a l l found to effect quantitative conversion of the alkylation product 204 . T.L.C. and N.M.R. indicated that the product composition obtained was identical with that observed using Dalton's conditions. 175 Methanol-HCl, conditions u t i l i z e d by Cranwell and Saxton resulted once again in a quantitative conversion to a mixture of the same two products. T.L.C. demonstrated that the reaction had occurred instantaneously on mixing the reactants. It was evident from these results that the cyclization under a variety of acidic conditions leads to the formation of a mixture of two products. - 201 -A detailed examination of the spectral data revealed the nature of the processes that occurred. The N.M.R. spectrum, (Figure 53) portrayed the presence of two carbomethoxy methyl group singlets at 5 3.84 and 3.68 as well as a pair of doublets at 6 1.00 and 1.15 (J = 7Hz) indicative of two different methyl groups attached to a saturated carbon atom bearing a methine hydrogen. These two methyl doublets integrated for a total of three hydrogen atoms thereby suggesting that they are due to the presence of epimers. The methylene hydrogens were spread out over the region 6 1.5 to 3.5 and the integration of these signals was d i f f i c u l t to ascertain with any degree of accuracy. The presence of a singlet at 6 2.67, integrating for three hydrogens was indicative of a methyl attached to an unsaturated center. The multiplet in the aromatic region centered at 6 7.2 integrated for ten hydrogens, two more than twice the number of hydrogens expected in the desired substance. This fact, coupled with the presence of two signals for carbomethoxy methyl and methyl groups suggested that the product mixture consisted of two components each containing the indole moeity, a methyl group and a carbomethoxy group. The absence of a peak at 5 6.94 for the indole C-2 hydrogen was important i n that i t meant that cyclization onto the C-2 position had occurred as expected. Based on a comparison of the relative intensities of the carbomethoxy signals i t was deduced that the reaction mixture consisted of a 1:1 mixture of the two products. The identity of the two components could not be deduced however, from the N.M.R. spectrum alone, but with the added support of the mass spectral data, their structures were determined. 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 ( 1 i | t i i i f i i i f i i i i' i i i I'M i i' i i i i' sooo J J O O iJgo s o J 0 l i e I l | I .1 l ,i I I l ' I I | I I i i I i I i i i i | i i 'i i I i 'i i i M i i i I i i Figure 53. N.M.R. Spectrum of the Cyclization Reaction Mixture. - 203 -The region of the mass spectrum corresponding to the expected molecular ion was characterized by the presence of two intense peaks at m/e = 239 (100%) and m/e = 243 (50%), and an almost negligible peak at m/e = 241, the expected molecular ion for the proposed 3,4-dihydrocarbazole 208. These two peaks were considered to be the parent peaks for each of the two isolated components. That the m/e = 239 peak did not arise v i a fragmentation of the m/e = 243 peak was strongly supported by the presence of fragmentations for loss of CH 3 , O C H 3 , and COOCH3 from each of these parent peaks, figure 54 il l u s t r a t e s the characteristic fragmentations observed in the low resolution spectrum. Of great importance was the presence of fragmentation peaks at m/e = 169, 168, 157, 143, and 130 which could only arise through disruptions of a saturated carbazole C" ring. Such an appropriately methyl and carbomethoxy substituted tetrahydro carbazole would give rise to the observed m/e = 243 peak. The m/e =239 peak can thus be considered to arise from loss of four hydrogens in the former to give the f u l l y aromatized carbazole compound. To determine whether or not the species observed in the mass spectrum arose from a d i s p r o p o r t i o n a t e of the 3,4-dihydrocarbazole 208 during cyclizatioi or was simply an event i n the spectrometer, experiments were conducted where low ionizing voltages of 53eV and low temperatures for v o l a t i l i z a t i o n were employed. In a l l cases the spectra were identical to that obtained under normal operating conditions (i.e. 70eV) T — i — r 5 0 . 0 S 8 mmif , 100.0 150.0 CO 8 COOCH-(134) M^m/e = 239 COOCH. (209) M+,m/e - 243 2 0 0 . 0 M/E T 2 5 0 . 0 r — i — r — i — | — i — i — r — i — r — i — i — i — r 3 0 0 . 0 3 5 0 . 0 1 4 0 0 . 0 gure 54. Mass Spectral Fragmentation Patterns for Cimponents 134 and 209 of the Cyclization feaction Mixture. - 205 -On the basis of the N.M.R. and mass spectral analysis i t was proposed that the cyclization reaction produced the f u l l y unsaturated compound, l-methyl-2-carbomethoxy carbazole (134) and i t s f u l l y saturated counterpart l-methyl-2-carbomethoxy-l,2,3,4-tetrahydrocarbazole (209) . To accommodate the existence of the saturated methyl group in two separate environments, two configurations of the methyl with respect to the carbomethoxy groups were inferred. (134) (209a) (209b) Conclusive proof of the above assignments was not possible however without separation of the two components and provision of appropriate characterization data. Considering that the carbomethoxy functionality is essentially neutral, i t was not surprising to find that the T.L.C. characteristics for the cycliza-tion product mixture were identical to those obtained for an equal mixture of carbazole and tetrahydrocarbazole. I N both cases two closely overlapping spots were obtained giving identical colour reactions when sprayed with eerie sulphate. Separation of the two reaction components proved to be very d i f f i c u l t . On column chromatography an enrichment of the impure tetrahydro compound . was obtained, whereas, pure carbazole component 134 was obtained in good yield (about one half of the material applied). This latter material crystallized spontaneously from highly concentrated chloroform solution to give colourless - 206 -crystals, M.P. = 138-140°. Very poor weight recovery of the tetrahydrocar-^ bazole 209 from the column was achieved. This situation also prevailed when this material was subsequently further purified by preparative layer chroma^ tography. After several purifications small amounts of carbazole 134 were s t i l l detected. Complete analytical and spectral data were obtained for the crystalline carbazole 134 and these were in compliance with i t s structure. The N,M,R, spectrum exhibited the anticipated singlets at 63.84 and 2.70 for the hydrogens of the two methyl groups, while signals for methylene protons observed in the above mixture were now totally absent. The U.V. spectrum was surprising i n that i t was identical in shape to that for the crude reaction mixture. The measured extinction coefficients compared favourably with those for the corresponding ethyl ester prepared by 172 Schmutz and Wittwer (I960) , however a discrepancy arose as to the position of the maximum. The maximum for the methyl ester (measured in methanol) was reported at 303 nm and that for the ethyl ester (measured in ethanol) i s observed at 308 nm. The fragmentation pattern i n the low resolution mas spectrum corresponded to one half of the observed fragments in the composite spectrum for the crude reaction mixture. The parent peak in the high resolution spectrum m/e = 239.0946 was found to be consistent with the correct molecular composition C 1 5 H 1 3 O 2 N , as were the compositions for the fragment ions corresponding to the loss of CH3, OC H 3 and COO C H 3 . - 207 -The spectral data determined for the tetrahydrocarbazole 209 was poor in that inadequate quantities of the compound could be obtained in pure form. Crystallization could not be induced, consequently i t was studied as a viscous o i l . The U.V. spectrum exhibited a typical indole absorption having maxima at 290, 282, and 275 nm, this being expected for a tetrahydrocarbazole. The extinction coefficients for the indole chromophore are low, compared to a carbazole chromophore, and they may represent only a small contribution to the composite spectrum which may possibly explain the similarity of the spectra of the carbazole 134 and the crude mixture. Fourier transform N.M.R. portrayed the presence of the carbomethoxy functionality and the saturated methyl as a singlet at 6 3.68 and a pair of high f i e l d doublets respectively (the latter partly obscured by impurities) . Particularly significant were the small set of multiplets centered at 6 3.38 and 3.96 attributed to the methine protons in their two respective configurations (209 a and b). The absorption for the methine hydrogen geminal with the carbomethoxy group was masked by the singlet peak at 6 3.68 for this group. The parent peak in the mass spectrum at m/e = 243 (85%) was clearly evident, as were the fragments corresponding to loss of C H 3 , OCH3 and COOCH3 respectively. Those fragments corresponding to the disruption of the saturated "C" ring were also evident. Thus the cyclization of the alkylation product^04 did not give the expected 3,4-dihydrocarbazole 208 , but proceeded further to yi e l d the carbazole 134 and tetrahydrocarbazoles (709 a and b). Their formation can be mechanistically - 208 -r envisaged to result from a total disproportionation of the i n i t i a l l y formed 3,4-dihydrocarbazole 208 under the reaction conditions employed (figure 49 ) f (204) -»• |(208)| -+ (134) + (209). Such a disproportionation reflects the inherent i n s t a b i l i t y of the dihydrosystem with respect to i t s anomatic counterpart, Other examples of this occurrence exist, particularly in cases where potential dihydropyridine systems are involved such as the 3,4-dihydroiso-r 195 168 quinolines and 3,4-dihydro-8-carbolines . It i s of interest to note that the 3,4-dihydrocarbazole aldehyde 155 synthesized by Wenkert and Dave 173 (1962) did not undergo disproportionation on i t s formation. Whether this reflects the difference in the method of preparation of this compound or i t s (155) s t a b i l i t y i s not known. The aldehyde 155 did disproportionate (see page 172) however even on mild dehydrogenation treatment with palladium on charcoal. Methanol - HC1 (2%) conditions were employed for a l l large scale c y c l i r zations due to the simplicity of the reaction and the quantitative yields of the two products. Column chromatography was not used to separate the two components during preparative work because subsequent dehydrogenation gave pure carbazole 134. The overall yield for the subsequent dehydrogenation could be increased by prior selective recrystallization of the carbazole - 209 -component 134 from a highly concentrated chloroform solution of the mixture. Dehydrogenation was, therefore, conducted only on the mother liquors from the cuptallization of 13^. l-Methyl-2-carbomethoxycarbazole (134). The use of palladium on charcoal for the dehydrogenation of tetrahydro-carbazole and other dihydro compounds has considerable precedence particularly for the systems dealt with in the present work (see introduction part III). Palladium on charcoal was, therefore, the obvious choice for the conver-sion of the cyclization reaction mixture into the carbazole 134 . Under typical reaction conditions, however, only par t i a l dehydrogenation was observed. Changing to platinum oxide as the catalyst enabled complete dehydrogenation in a l l experiments conducted. However, the large quantities necessary made i t s use prohibitive on a preparative scale. Attention was thus turned towards the use of quinone dehydrogenating agents. Quinones, especially chloranil and dicyanodichlorobenzoquinone (DDQ) have been u t i l i z e d with great success in the dehydrogenation of a wide range 196-7 of carbazole compounds. With chloranil a homogeneous carbazole ester 134 reaction product mixture was obtained. Traces of coloured impurities were removed from the crude product either by recrystallization from chloroform or by column chromatography. The product was eluted with benzene as a yellow o i l which subsequently crystallized to yi e l d colourless crystals. The yields were quite variable ranging normally between 70-80% although occasionally dropping as low as 60%. To prevent manipulative d i f f i c u l t i e s during chromatography due to the non-polar nature of both the carbazole compound and the coloured contaminants 1 - 210 -purification was normally postponed un t i l after reduction to the corresponding alcohol 157 . The reduction was found not to be affected by the presence of these impurities and the alcohol 157 being more polar was more easily separated from them. The spectral and analytical data obtained for the dehydrogenation product were identical to that obtained for the carbazole component 134 of the cycliza-tion mixture, thus confirming the above conversion, Cyclization - Dehydrogenation Advantage was taken of the disproportionation reaction to achieve the direct conversion of the alkylation product 204 into the carbazole 134 in one step by conducting the cyclization reaction i n the presence of an excess of chloranil as the hydrogen acceptor. The rationale behind this reaction was that the 3,4-dihydrocarbazole208;would act as a poor hydrogen acceptor com* pared with chloranil. On this basis the alkylation product 204 dissolved in benzene containing excess chloranil was treated with methanol - HC1 (2%) . The isolated product, formed i n 84% yield, was indeed the desired carbazole 134. Thus the necessity for a distinct dehydrogenation step was eliminated, 1-Methy1-2-hydroxymethylcarbazole (157) . The conversion of the ester group of the carbazole 134 to the corresponding alcohol 157 was readily accomplished by making use of lithium aluminum hydride 172 reduction conditions developed by Schmutz and Wittwer . Reaction of pure carbazole ester 134 gave the desired carbazole alcohol 157 as a colourless cry-stalline s o l i d in 95% yie l d . The corresponding reduction of impure carbazole ester 134 obtained via 211 chloranil dehydrogenation was conducted in an identical manner. Chromatogra-phic purification of the product provided an overall 60% yield of the alcohol 157.* The spectral data were consistent with the structure 157. The mass spectrum + showed a parent peak at m/e = 211 as well as an M -17 for loss of hydroxyl. The absence of the carbonyl absorption at 1710 cm"* in the I.R. spectrum was noted. The U.V. spectrum exhibited a typical carbazole absorption as expected due to loss of conjugation with the carbomethoxy functionality. The singlet at 6 3.84 in the N.M.R. for the ester methyl group was absent, having been replaced by a singlet at 64.88 ( C H 2 O H ) and a singlet at 61.50 (OH). l-Methyl-2-formylcarbazole (152) The conversion of the benzylic alcohol 157 to the aldehyde 152 proved to be 177 a low yielding step. Mosher et_ al_ used bispyridine-chromium trioxide ( C i ^ - p y ^ ) in pyridine, i.e. Sarett conditions to oxidize this alcohol 157, reporting a y i e l d of about 70% for a large scale reaction (16 gm). Govindachari et al report a higher y i e l d of 84% for a small scale (1 gm) Sarett oxidation of 1 7r» the closely related alcohol 189 in the e l l i p t i c i n e (17) synthesis. It was found during the present work, however, that consistently lower yields ( 60%) were obtained for the Sarett oxidation of the alcohol 157. In an attempt to improve upon this y i e l d a considerable number of oxidation conditions were studied during the present synthesis. A tabulation of the various oxidation •footnote: For ease of manipulation the entire sequence could be conducted with-out purification at each step. In terms of overall y i e l d starting from trypto-phyl bromide, a comparison of both procedures showed there to be no advantage in not purifying each product (see experimental). 212 methods including the Sarett, the conditions under which the reactions were conducted and yields or comments on the reaction outcome i s presented i n Table III. From this study i t was determined that an ef f i c i e n t oxidation procedure depended upon two c r i t e r i a , i) The solubility of the alcohol in the reaction media and i i ) An efficient uncomplicated work-up. 198 Collins conditions were an obvious alternative to the Sarett reaction with yields of 90% being reported for the related conversion of benzyl alcohol to benzaldehyde. Similarly high yields are obtained from Silver^**) oxide*"'*' or C e ^ ^ oxidation*" of the benzylic alcohols. Hov/ever in a l l three cases poor solubility of the alcohol 157 in the reaction solvent hindered oxi-dation and under the conditions quoted for the reaction no detectible formation of aldehyde 152 was observed. Considerable d i f f i c u l t i e s during work-up resulting from mutual s o l u b i l i -ties of the extracting solvents and reaction reagents were encountered for the pyridine-based oxidation methods and this was one of the major drawbacks to the Sarett oxidation. The influence of the work-up on product yield was particularly drastic in the case of the HOAc/Cr03-pyr2 oxidation 2 0 0 where a drop from 85% to 35% y i e l d accompanied the scaling of the reaction to larger quantities. The pyridine-S03/DMS0-Et3N oxidation method developed by Doering and co-2 0 1 workers for work in the steroid f i e l d gave a mixture of three products on work-up . Separation of the aldehyde component from this mixture by either 213 Table IV. Results for Oxidation of Alcohol (157) Method Conditions Yield or Comment Sarett Cr03«pyr2 in pyr 24 hrs., RT. 60-65%d HOAc/Cr03.pyr in HOAc -5 to +16° 15 min. 85%, 35% d Collins Cr0 3*pyr 2 in CH2Cl2 RT., 0.5 hr. no product detected 3 Ag I X0 1 M H 3P0 4, 25l 3-4 hr. no reaction a,b C e ™ H0Ac/H20 90° 12 hr. no reaction 3 Pyr'SOj/DMSO Et3N RT., 24 hr. inseparable mixture 1 Pb(0Ac) 4/pyr RT., 24 hr 50-60%d Jones Cr0 3, H 2S0 4(aq.) i n Acetone 1 min., RT, 70-75% a. TLC detection (A1 20 3, CHC1 ), b. UV detection, c. small scale $ 300 mg., d. large scale > 2.0 gm. - 214 -alumina or s i l i c a chromatography was unsuccessful as was attempted recry-st a l l i z a t i o n from chloroform. The problems associated with the presence of pyridine during extraction 202 were eliminated by using the lead tetraacetate in pyridine conditions . At the completion of reaction the solvent was f i r s t removed in vacuo before extrac-tion of the product. The aldehyde 152 was obtained as the major product i . e . (50-60% yield) after column chromatography. Several minor by-products were detected in this reaction. The of a non-polar component coinciding on T.L.C. with authentic l-methylcarbazole-2-acetate (215), this material being an expected by-product. (215) Jones oxidation^ 0" 5 proved to be a viable alternative to the Sarett reaction yielding the aldehyde 152 as the sole product. This reaction often leads directly to the carboxylic acid , but in this instance the reaction stopped at the aldehyde stage. The reaction did not appear to be dependent on either the time span or the amount of oxidant present. Attempted further oxidation of the aldehyde 152 resulted i n the re-isolation of starting material. Although a single reaction product was produced the yields were not exceptionally high (about 70%), this again reflecting the inefficiency of product extraction during chromium (VI) oxidations. Despite the only moderate, i f any, improvement in yields, the Jones conditions were considerably more - 2 15 -e f f i c i e n t than the Sarett i n terms of reaction time and ease of work-up. Proof of the aldehyde structure 1 5 2 was derived from a comparison of the melting point and quantitative U.V. spectrum with that reported by 1 7 7 Mosher et^ a_l, as well as from the other spectral and analytical data. The N.M.R. spectrum exhibited two singlets, one at 62.87 due to the aromatic methyl, and the other at 610.39 for the aldehydic hydrogen. The I.R. spectrum indicated the expected aldehyde carbonyl band (1665 cm-"*) and in the mass spectrum parent peak was at m/e = 209. High resolution mass measurement provided the value m/e = 209.083] which established the desired molecular formula (C^H^NO requires:209.0842). "D" Ring Synthesis The construction of the "D" ring of olivacine through formation of the imine, 3,4-dihydro olivacine ( 1 4 1 ) from the aldehyde intermediate 1 52 was 177 accomplished by following the procedure published by Mosher e_t al_. No unexpected complications were encountered during the interconversions and a l l analytical and spectral data were in accord with the proposed structures and compared within acceptable limits with literature values where quoted. Condensation of the aldehyde 1 52 with nitromethane in the presence of ammonium acetate produced the nitrostyrene 1 6 0 as the sole reaction product i n 94% yield. Recrystallization to crimson needles was possible from large volumes of ethanol. The nitrostyrene 1 60 was reduced to the amine 139 by lithium aluminum hydride reduction in tetrahydrofuran at room temperature. The product isolated as a white crystalline solid was obtained in 95% y i e l d . The amine 1 3 9 was converted without purification to the N-Acetylamine by reaction - 216 -with acetic anhydride in pyridine. The product was a white crystalline solid obtained in 95% yield. The Bischler-Napieralski cyclization of the latter to 3,4-dihydroolivacine 141 was accomplished by refluxing the amide 140 with phosphorous oxychloride in toluene. The ultraviolet spectrum of 141 run i n water and dilute hydrochloric acid showed the expected batho-125 chromic shift accompanying the transformation C = N •*• C = NH*. Reaction of the imine 141 with methyl iodide i n methanol produced the methiodide salt 142 i n quantitative yield. The shifts observed in the UV. spectrum of this iminium cation i n base reflected i t s rearrangement to i t s corresponding enamine (anhydro base) (Section 6., part II). Reduction 142 of the methiodide either by sodium borohydride in aqueous ethanol or by catalytic hydrogenation over platinum oxide gave in high y i e l d the tetracyclic alkaloid (±)-guatambuine (25). Proof of the structure of guatambuine (25) was derived from the superimposibility of the UV. and IR. spectra with the 126 reported spectra and by the close comparison of the N.M.R. and mass spec-trum with that obtained for guatambuine (25) isolated from australe (Figure 20). Dehydrogenation of the imine 141 by reaction over palladium on charcoal i n refluxing decalin produced the tetracyclic alkaloid olivacine (16) in 70% yield. Total characterization of this product was also derived from a com-parison of the N.M.R. and mass spectral data with that obtained for olivacine isolated from A. australe. The N.M.R. spectrum displayed a pair of singlets at 62.82 and 3.16 for the C-l and C-5 methyl groups respectively. The mass spectrum possessed a parent peak at m/e = 246 corroborated by the high resolu-tion spectrum parent peak at m/e = 246.1165 (Ci7H 1 4N 2 requires m/e = 246.1156). - 217 -Sequence B A short route to the synthesis of the carbazole-3-aldehyde 219 was provided from the reported observation 2^ that the Vilsmeier-Haack formy-lation of 9-benzyl-I,2,3,4-tetrahydrocarbazole (217) instead of leading to 205 the expected 7-formyl-9-benzyl-l,2,3,4-tetrahydrocarbazole (218) gives l-methyl-3- formyl-9-benzylcarbazole (219). The y i e l d of about 50% for this conversion was acceptable since this provided a one step synthesis of this desired intermediate aldehyde from a readily available starting material (217). OHC-: II I N Bz (217) N Bz (218) CHO (219) In the application of the Pomeranz-Fritsch reaction for the synthesis of the pyrido ring the necessity of incorporating a methyl group at C-l of the olivacine system (16) created complications. It is known that the con-206 densation of amino acetals with ketones i s far less e f f i c i e n t than the - 218 -corresponding reaction with aldehydes. Hence the most direct route to olivacine (16) by reaction of aminoacetaldehyde diethylacetal (171) with the 3-acetylcarbazole 220 had l i t t l e chance of being successful. Two alternatives to this condensation were available. On the one hand the l-methyl-3-formylcarbazole intermediate (219) was f i r s t condensed with aminoacetaldehyde diethylacetal (171) and the resulting iminomethyl acetal 221 was then methylated with methylmagnesium chloride to give the aminoethyl acetal 224. Alternatively, the 3-formylcarbazole compound 219 was reacted with methylmagnesium chloride to give the a-carbazolyl ethanol system 222 which was converted to the corresponding acetate 223. Substitution of the acetate func-tionality by the aminogroup of aminoacetaldehyde diethylacetal (171) again produced the desired aminoethyl acetal 224 (Figure 55) . Cyclization of this aminoethyl acetal 224 to 6-benzoolivacine (225) f o l -lowed by debenzylation to olivacine (16) was not attempted. However, the con-ditions necessary for the cyclization have been worked out for the synthesis of the closely related molecule, e l l i p t i c i n e (17). (220) - 219 -R I P=Bz (225) =H (16) Figure 55. Synthesis of Olivacine (16) from Tetrahydrocarbazole (216). This synthesis as i t presently stands is generally low yielding in each of the individual steps, however, i t is potentially a short and simple route to olivacine (16) starting from readily available starting material. It could easily be adapted to large scale preparation and towards the prepara-tion of various forms of radioactive olivacine (16) (particularly the C-l CH^ group) which may be required in future biosynthetic experiments. - 220 -9-Benzyltetrahydrocarbazole (217) Tetrahydrocarbazole (216) was readily available in large quantities as a starting material through the condensation of phenylhydrazine with cyclo-207 hexanone in the presence of acid. The indole nitrogen (Na) of tetrahydro-carbazole (216) was converted to i t s benzyl derivative 217 by reaction of ben-zyl bromide and sodium hydride. A colourless o i l obtained in 77% yield possessed the expected spectral characteristics. Thus the mass spectrum showed the expected parent peak at m/e = 261 while the N.M.R. spectrum exhibited the presence of the methylene of the benzyl group (6 5.10, singlet) and the methylenes of the carbocyclic ring system (pair of multiplets centered at 62.60 and 1.86). The absence of the N-H stretch signal in the region of 3500 cm - 1 in the IR spectrum was also indicative of reaction at the indole nitrogen. The 9-benzyl derivative 217 was formed for three reasons. F i r s t l y , the sub-sequent Vilsmeier-Haack formylation to be carried out has been shown to proceed to the desired l-methyl-3-formylcarbazole compound 219 only when the indole nitrogen i s protected by an alkyl group i.e. methyl, benzyl, etc. Secondly, having the indole nitrogen protected eliminates any undesirable side reactions that may occur during methylation of the formyl or iminomethyl group under Grignard conditions. Thirdly, the benzyl group i s practically the only alkyl protecting group which can be effectively removed in the last step to yield olivacine (16). - 221 -l-Methyl-3-formyl-9-benzylcarbazole (219) . Reaction of 9-benzyltetrahydrocarbazole (217) with 1.3 equivalents of phosphorus oxychloride in dimethylformamide (Vilsmeier-Haack conditions) proceeded as reported 2^ 4 to give the desired l-methyl-3-formyl-9benzylcarbazole (219) . The reaction mixture was purified by column chromatography on alumina. Whereupon the aldehyde 219 was recovered in 48% yield as a pure yellow o i l which was subsequently crystallized to give pale yellow rosettes. The N.M.R. spectrum of the product 219 possessed a singlet at 6 1 0 . 0 0 for the aldehyde proton as well as singlets at 6 5.72 and 2.60 for the methylene of the benzyl group and the C-l methyl groups respectively. The correct justa-position of the methyl (C-l) and formyl (C-3) groups to each other was clear from the presence of two distorted singlets (meta coupling not resolved) at 68.46 and 7.66 for the C-4 and C-2 hydrogens respectively. The absence of any multiplet absorptions for the methylene protons characteristic of a tetra-hydrocarbazole system confirmed the aromatic nature of the molecule. The UV. spectrum was altered considerably from the normal carbazole spectrum as a consequence of the extension of the carbazole chromophore by the formyl group. The IR spectrum possessed a carbonyl absorption at 1670 cm"* for the aldehyde carbonyl in conjugation with the aromatic ring. The mass spectrum displayed a very dominant parent peak at m/e = 299 ( 1 0 0 % ) and very l i t t l e fragmentation. The position of the parent peak was corroborated by the high resolution spectrum which had a peak at m/e = 299.1309 which was within accepted limits of the calculated mass at m/e = 299.1305 for C21H17NO. - 222 -A slightly less polar component eluted during column chromatography was the only detectable side product of the reaction. Attempts to separate i t completely from the carbazole aldehyde 219 failed, however i t s N.M.R. spectrum gave every indication that i t corresponded to l-formyl-9- benzyl-tetrahydrocarbazole (226). Thus present in the N.M.R. spectrum was a singlet at 5.13 for the benzyl methylene and two multiplets at 6 2.60 and 1.80 for the methylenes of the tetrahydro ring. The aldehyde proton absorption occurred at 6 9.83 slightly up f i e l d from the aldehyde absorption in 219. C H O (226) The presence of the formyl tetrahydro compound 226 was of considerable significance for although l i t t l e i s presently known about the mechanism of 208 the alkylation-formylation-dehydration process, i t has been shown that Vilsmeier-Haack formylation of 6-chloro-9-methyltetrahydrocarbazole (227) under mild conditions produces as the reaction product l-formyl-6-chloro-9-methyltetrahydrocarbazole (228) and that subsequent further reaction of this compound 228 at elevated temperature resulted in conversion to the l-methyl-3-formyl carbazole 229. - 223 -Cl Cl CHO Cl-N A (226) (227) (228) It i s apparent from this observation that in the present reaction of 9-benzyltetrahydrocarbazole (217) that the C-l methyl of the product aldehyde 219 is also derived via an i n i t i a l C-l formylation of 217. On mechanistic grounds i t was anticipated that l-methyl-6-formyl-9-benzylcarbazole (230) would also be produced during the formylation reaction. However a detailed examination of the aromatic region of the N.M.R. spectra of respective column fractions gave no indication of i t s presence. In light of these results a plausible mechanism can be put forward which suggests that both alkylation and formylation takes place before aromatization of the tetrahydro ring (Figure 56). I f aromatization occurred f i r s t , then a mixture of the 3- and 6-formyl carbazole compounds would have been observed. (230) - 224 -Figure 56. Proposed Mechanism for the Vilsmeier Haack Formylation of 9-Benzyltetrahydrocarbazole (216). - 225 -l-Methyl-3-(g-hydroxyethyl) carbazole (222) To explore the route whereby the methyl group in the potential C-l position of olivacine (16) i s introduced before reaction with aminoacetalde-hyde diethylacetal (171) the 3-formyl group of 219 was methylated with methylmagnesium chloride in tetrahydrofuran. The reaction proceeded to com-pletion (TLC) within minutes at room temperature to yield the alcohol 222 as the predominant product. Purification by column chromatography provided a 64% yield of the desired product. The methylation reaction was also accomplished using methyllithium instead of the Grignard reagent, however i t offered no advantages in terms of reaction time, yields, or ease of handling so i t was not employed routinely. Minor impurities were removed from the crude reaction mixture by rapid chromatography on alumina. The alcohol 222 was obtained as a colourless foam which subsequently proved to be extremely d i f f i c u l t to recrystallize. It was eventually necessary to convert the alcohol to i t s corresponding acetate i n order to obtain a satisfactory microanalysis. The N.M.R. spectrum of this product exhibited a doublet at 6 1.55 (J * 6 H z ) for the benzylic methyl group and a corresponding quartet at 6 5.03 (J = 6 H z ) for the methine proton which is both benzylic and adjacent to the oxygen of the hydroxyl group. Three singlets were also evident at 6 1.83 (OHJ), 2.60 ( C - I C H 3 ) , and 5.70 (N-CH2-C5H5) respectively. The mass spectrum possessed a parent peak at m/e = 315 (100%) and only two major fragments at m/e = 300 for (M -CH3) and m/e = 297 for (M -H20). The high resolution mass spectrum parent - 226 -peak at 315.1632 was within acceptable limits of the calculated mass of 315.1623 for C22H2iNO. The absence of a carbonyl absorption at 1670 cm"^  in the IR. spectrum was consistent with the proposed structure as was the UV. spectrum which exhibited a typical carbazole absorption, l-Methyl-3-(a-acetoxyethyl) carbazole (223) The acetate derivative 223 was prepared by reaction of the alcohol 222 with acetic anhydride in pyridine. It was obtained as a colourless crystalline s o l i d in high yi e l d after careful recrystallization from methanol. The acetate group proved to be somewhat labile however, as an overexposure to hot methanol resulted in a substantial conversion of the acetate to the methyl ether by solvent displacement of the acetoxy group. It was also observed (N.M.R.) that small amounts of ole f i n i c products accompanied formation and subsequent heating of the carbazole acetate (section 6, part II), The spectral data was very characteristic for the acetate derivative 223, The N.M.R. spectrum showed a characteristic downfield shift in the position of the benzylic methine quarter to 66.00 (J = 7Hz) due to the added deshielding influence of the carbonyl group. The corresponding benzylic methyl group removed from the influence of the acetate function absorbed at 61.56 (J = 7Hz), Singlets were observed at 6 1.98 and 2.52 for the protons of the acetate methyl and the methyl group at C-l respectively. A singlet was also observed at 6 5.68 for the methylene of the benzyl group substituted at the indole nitrogen. The change in the election density of the "C" ring in the conversion of the - 227 -conjugated aldehyde 219 to the non conjugated acetate 223 reflected i t s e l f in the position of the absorptions for the C-2 and C-4 hydrogens of the "C" ring. Both of these signals shifted upfield, the C-4 proton signal to 6 7.98 while that of the C-2 hydrogen had shifted to the region occupied by the large aromatic multiplet (centered at 6 7.10). The mass spectrum possessed a strong parent peak at m/e = 357 (100%) and a major fragment at m/e = 298 for loss of the acetate group. Ibis frag-mentation was accompanied by the presence of a metastable at m/e = 249 corresponding to m/e = 357 —• m/e = 298. The high resolution spectrum parent peak at 357.1676 was within accepted limits of the value 357.1729 calculated for C 2 4 H 2 3 N O 2 . The UV spectrum portrayed a typical carbazole absorption as anticipated and the IR spectrum has a carbonyl absorption for the acetate group at 1720 cm"1. l-Methyl-3-(a-(N-2* >2'-diethoxyethylamino)ethyl)carbazole (224) from acetate 223 In an attempt to synthesize the aminoethyl acetal 224 from the carbazole acetate 223 i t was determined that high temperature and prolonged heating times were necessary in order to induce displacement of the acetate function with the amino group of aminoacetaldehyde diethylacetal (171). Even after fifteen hours at 160° considerable starting material was present in the reaction mixture which by this time had turned very dark. By column chromatography on alumina a small quantity of material was - 228 -isolated whose spectral properties corresponded with the anticipated amino-ethyl acetal 224 (see conversion (221) —> (224) for a description of spectral data). It was concluded in view of the drastic reaction conditions that were necessary to provide a small percent conversion to product that the acetate functionality was not a sufficiently good leaving group for this purpose. However, instead of exploring the properties of the corresponding bromide as a leaving group, i t was decided to study the methylation of the iminomethyl acetal 221 as a means of obtaining the same aminoethyl acetal 224. l-Methyl-3-(3,g-diethoxyethyliminomethyl)carbazole (221) The aldehyde condensation reaction had already been developed in the 175 synthesis of e l l i p t i c i n e (17) and i t s application to the 3-formyl carbazole intermediate 219 lacking one methyl group at the C-4 position presented no d i f f i c u l t i e s . Reaction of the aldehyde 219 with an equimolar quantity of aminoacetaldehyde diethylacetal (171) with the azeotropic removal of the water formed gave the iminomethyl acetal in 85% yie l d . The product was purified by recrystallization from a petroleum ether-benzene combination. An attempt to purify the product by column chromatography on alumina resulted in hydrolysis on the column of the imine bond to give back the starting aldehyde. The N.M.R. spectrum (Figure 57) of the crystalline product 221 showed a t r i p l e t at 64.81 (J = 5Hz) for the hydrogen of the acetal carbon (CH(OEt) 2), a second t r i p l e t at 6 1.14 (J = 7Hz) for the two methyls of the diethylacetal and a complex multiplet centered at 6 3.70 for the methylenes of the acetal and for - 229 -o neutral 220 250 300 350 nm 400 Figure 58. UV. Spectrum of Compound 221 in Water and in Dilute Hydrochloric Acid. - 231 -the methylene adjacent to the imine nitrogen. Two singlets were also present at 65.72 and 2.58 for the benzyl methylene and the C-l methyl group respectively. A distorted singlet at 68.32 was attributed to the proton on the imine carbon while the singlets at 68.40 and 7.60 having moved downfield with respect to the acetate 223 were assigned to the C-4 and C-2 hydrogens respectively. The mass spectrum exhibited a parent peak at m/e = 414 and important frag-mentations at m/e = 369, 339, 311, 298, 284. Metastable peaks were observed for the major fragmentation processes, m/e = 414 -»• m/e = 369 and m/e = 311 m/e = 284. The high resolution mass spectrum corroborated the parent peak at 414.2250 and the molecular compositions for the postulated fragments. The presence of the conjugated imine system was detected by an absorption at 1690 cm-1 in the IR. spectrum. The ultraviolet spectrum of 221 run i n water and dilute hydrochloric acid showed the expected bathochromic shift accompanying the transformation C = N -> C = NH+. l-Methyl-3-(a-(N-B,B-diethoxyethylamino)ethyl)carbazole (224) • Methylation of the imine 221 as with the formyl compound 219 could be effected through reaction with methylmagnesium chloride i n tetrahydrofuran. The less reactive nature of the imine system however, necessitated reaction at reflux tem-perature for twenty-four hours. Even after prolonged heating some starting material remained. Three major products were observed on TLC with the amino acetal 224 being the most polar. The three components were e f f i c i e n t l y separated from each other by column chromatography on alumina. - 232 -The least polar component (bright orange spot on TLC) eluted with benzene remains uncharacterized. The N.M.R. spectrum possessed a quartet (63.93, J = 6 Hz) and doublet (61.36, J = 6 Hz) for the C-l* methine and C - l 1 methyl group which indicated that the molecule was a 3-(a-substituted ethyl)carbazole derivative. The nature of the substituted group could not be discerned from the N.M.R. or the IR spectrum. The mass spectrum possessed a parent peak at m/e = 368 with a major fragment at m/e = 299 for the loss of the unknown substituent. It was not known which functionality was substituted onto the C - l ' position of the unknown carbazole component with a mass of 69 and no characteristic absorp-tions in the N.M.R. The second component isolated by elution with benzene corresponded to the aldehyde 219. The isolation of this compound resulted from the presence of unreacted starting material in the crude reaction jnixture which hydrolyzed during the column chromatography. The desired amino-acetal 224 was isolated in 20% yield by elution with benzene:chloroform mixtures and chloroform. The N.M.R. spectrum for 224 possessed a multiplet at 68.14 for the C=5H and a singlet at 67.98 for the C-4 H. The singlet for the C-2 H had moved upfield into the region of the aromatic multiplet compared to the position of the C-2 H in the imino acetal 221. A quartet was present at 63.95 (J = 6 Hz) for the methine hydrogen on the alkylated C-l' carbon. The doublet for the corresponding C-l ' methyl group occurred at 61.47 (J » 6 Hz). The doublet (62.69, J = 6 Hz) for the methylene group of the diethoxy-ethylamino group in 224 had shifted upfield from i t s position under the multiplet for the methylenes of the 0-ethyl groups in compound 221. The signals for the - 233 -two methyls of the O-ethyl groups were very complex compared to the t r i p l e t for these groups in the imino acetal 221. The mass spectrum possessed a parent peak at m/e = 430 with major frag-mentation peaks at m/e = 415, 369, 298, 288. The high resolution mass spectrum possessed a parent peak at m/e = 430.2633 which was within acceptable limits of the value m/e = 430.2620 calculated for C 2 8 H 3 4 N 2 O 2 . No attempt was made to improve the reaction y i e l d either by changing the solvent, or the Grignard reagent or by conducting the reaction for a longer period of time. The cyclization of the amino acetal 224 to 6-benzoolivacine (225) was not attempted at this time. It was known however that o-phosphoric acid effected the analogous cyclization in the synthesis of e l l i p t i c i n e (30% y i e l d ) . It would be anticipated therefore that the cyclization reaction would also work for the olivacine system (16). There is the p o s s i b i l i t y however, which does not exist for e l l i p t i c i n e , that cyclization can occur in either of two directions to give compounds 225 or 231. Whether or not this w i l l be a problem remains to be determined. The subsequent debenzylation of 225 to olivacine (16) would be a t r i v i a l step. (16) (231) - 234 -This synthesis has the potential of being a short and simple route to olivacine (16) which could readily be adapted to a large scale preparation of this material for future biosynthetic work. By minor modifications i t would be possible to obtain quantities of guatambuine (25) also. - 235 -EXPERIMENTAL - PART III For a description of the general experimental information, see Experimental part 1. A l l T.L.C. plates were developed in chloroform (CHCI3) unless otherwise indicated, and the alumina for column chromatography was deactivated to Activity III by the addition of water (6%). Sequence A Indole -3-glyoxylic acid chloride (21l) To a stirred solution of Indole (25 gm., 2.14 x 10"1 mole) in anhydrous ether (400 ml) at 0° was added an ether solution of oxalyl chloride (210) (31.2 ml, 3.64 x 101 mole) over a 0.5 hr period. A bright yellow precipitate was produced almost immediately, however the reaction was allowed to continue for 0.5 hr after addition was completed. The product was collected by vacuum f i l t r a t i o n , and washed liberally with dry ether to give the acid chloride as yellow crystals (41 gm. 93%), m.p 123-126* Found: C,58.13; H,3.10,N,6.38%. Calc. for C 1 0H 60 2NC1: C,58.40; H,2.92; N,6.81% Methyl indole-3-glyoxylate (212) To the acid chloride 211 (39gm.,1.88 x 10"1 mole) dissolved in dry tetra-hydrofuran (500 ml) and stirred at room temperature was added a large excess of dry methanol. A beigh precipitate formed almost immediately, however the reaction - 236 -was allowed to s t i r for 0.5 hr. The precipitate was collected by suction f i l t r a t i o n (29 gm.) and washed liberally with tetrahydrofuran. The mother liquors were concentrated and the bright pink solid obtained (6 gm.) was washed also with tetrahydrofuran. The s o l i d products were combined and recrystallized from methanol-tetrahydrofuran to give the ester 212 as beige coloured crystals (35 gm., 92% yield), m.p. 227-8°. Found: C,64.76; p H,4.44; N,6.62%. Calc. for C nH 9N03: C.65.02; H,4.46; N.6.89%. 190 Indole-3-ethanol (Tryptophol) (206) The glyoxylate ester 212 (20.7 gm., 1.0 x 10"1 moles) was dissolved in dry tetrahydrofuran (600 ml) under gentle reflux and under a nitrogen atmos-phere (mechanical s t i r r e r ) . Lithium aluminum hydride (7.6 gm., 2.0 x 10" moles) was added to the refluxing solution in small portions either as a powder or as a slurry in tetrahydrofuran. After addition was completed, a further quantity of tetrahydrofuran (200 ml) was added, f i n a l volume (800 ml). The reaction mixture was refluxed gently for 15 hr. after which time the reaction vessel was cooled i n ice and the excess lithium aluminum hydride destroyed by the sequential addition of water (8 ml), 15% sodium hydroxide (8 ml), and water (24 ml). The resultant granular precipitate was removed by suction f i l t r a t i o n and washed twice with tetrahydrofuran (2 x 200 ml). The combined tetrahydrofuran solutions were concentrated to 50 ml and taken up in water (600 ml). The aqueous mixture was extracted repeatedly with ether (200 ml portions). The ether fractions were combined, washed with water, dried over - 237 -sodium sulphate, and concentrated to dryness (benzene azeotrope) to give a colourless crystalline product (9.6 gm., 60%). The crude product was purified by column chromatography on alumina (100 gm.). The product was eluted with chloroform to give colourless crystals, m.p. = 0 ( l i t ° ) . N.M.R. (60 MHz): 3.86 (tri p l e t , J = 6Hz,2H,C-2, a i 2 ) , 2.96 (tri p l e t , J = 6Hz2H,C-l, CH2), 1.56 (Singlet, 1H, 'OH) 192 Indole-3-(2-ethylbromide) (Tryptophyl Bromide) (207) A solution of phosphorous tribromide (2ml,2.07 x 10"2 mole) in ether (25 ml.) was added dropwise over 0.5 hr. to a solution of tryptophol (206) (10 gm. 6.22 x 10"2 mole) in ether (200 ml.) at 0 ° . The reaction mixture was stirred at 0°C for 15 hr. after which time the supernatant was decanted, the red syrupy residue was washed with several portions of ether (100 ml), and the combined ether extracts were washed with saturated sodium bicarbonate and water. The ether layer was dried over sodium sulphate and concentrated (benzene azeotrope) to give a colourless crude crystalline product (12.5 gm., 91%). The crude product was purified by either recrystallization from benzene:hexane or column chromatography on alumina (200 gm.), eluting with benzene. The bromide was obtained as colourless crystals, m.p. = 0 ( l i t . °) . N.M.R. (60 MHz): 3.40 (multiplet,4H,C-l and 2CH2). Mass spectrum: M,+m/e = 223,225. 3-Carbomethoxy-5-(3-indolyl)-2-pentanone (204) Methyl acetoacetate (205) (12.66 ml, 1.18 x 10"1 mole) was added in four portions to a suspension of sodium hydride (2.88 gm., 1.20 x 10"1 mole) - 238 -(prepared by washing a 57% o i l dispersion (5.05 gm.) with benzene) in dimethylformamide (500 ml). The mixture was stirred at room temperature until gas evolution had ceased and a clear solution was obtained (about 1 hr.). Tryptophyl bromide (207) (12.3 gm., 5.5 x 10 - 2 moles) was then added in one portion and the reaction was stirred at 100°C for 3.5 hours under a nitrogen atmosphere. A white precipitate was formed during the course of the reaction. The cooled reaction mixture was diluted with water (700 ml) and extracted repeatedly with ether. The combined ether extracts were washed with water, dried over sodium sulfate and concentrated to give an amber coloured viscous o i l (11.5 gm. 80%). The crude product was either cyclized directly or purified by column chromatography on Alumina (500 gm.). Elution of the column with benzene removed a l l contaminants and subsequent elution with benzene:chloroform 1:1 and f i n a l l y with chloroform yi e l d microanalyti-cally pure 204 as a yellow o i l (7.90 gm., 50% from tryptophyl bromide (207) ), b.p. = 135+10 at 0.05 m.m. From column chromatography on a small scale (1-2 gm) typical yields were 70-75%. IR. ( C H C 1 3 ) : 3500, 1750, 1725. UV. X m a x ( l o g e): 290 (4.30), 280(4.37), 272(4,34). N.M.R. (see Figure 51): 8.04 (broad singlet, IH, N-H), 7.6-7.0 (2 multiplets, 4H, aromatic), 6.94 (Spike, IH, C-2H), 3.66 (Singlet, 3H, C O O C H 3 ) , 3.48 ( t r i p l e t , J = 7 H z , IH, C-3 CH), 2.76 (triplet, J = 7 H z , 2H, C-5 CH 2), 2.28 (multiplet, 2H, C-4 CH2)» 2.12 (Singlet, 3H, C O C H 3 ) . Mass spectrum: M+, m/e = 259; main peaks: 228, 186, 184, 144,143 (base peak), 130. High resolution mass spectrum: calc. for C 1 5 H 1 7 O 3 N : 259.1207. Found: 259.1212. Found: C, 69.36; H,6.52; N,5.20%. Calc for - 239 -C 1 5H 1 70 3N: C, 69.48; H,6.61; N, 5.40% Cyclization Reaction to (134) and (209) A To a solution of the alkylation product 204 (7.90 gm., 3.05 x 1 0 " 2 mole) in methanol ( 1 0 0 ml.) at room temperature was added an equal volume of methanol - HCL ( 5 % ) , and the reaction mixture was allowed to s t i r for one minute (f i n a l H C 1 concentration 2 . 5 % ) . On mixing, the reaction was exothermic and accompanied by a change to a darker colour. The solvent was then removed to give a dark foam (7.15 gm., 98%) consisting of an equal mixture of 1-methyl-2-carbomethoxy-carbazole (134) and l-methyl-2-carbomethoxy-l,2,3,4-tetrahydrocarbazole (209)' IR. ( C H C 1 3 ) : 3480, 1740-1680. UV; A m a x : 340, 303, 295 (shoulder), 248. N.M.R. (see Figure 53): 8.44 (broad singlet, 1 H , N-H), 8.0-7.0 ( 2 multiplets, 1 1 H , aromatic protons), 3.84 (singlet, 3 H , C O O C H 3 ) , 3.68 (singlet, 3 H , C O O C H 3 ) , 2.67 (Singlet, 3 H , CH3), 1 . 0 0 , 1.15 ( 2 doublets, J = 7 H z , 3 H , C H 3 ) . Mass spectrum: M-, m/e = 243, 239 (base peak); main peaks: m/e = 228, 224, 2 1 2 , 208, 184, 180, 169, 168, 157, 143, 130. High resolution mass spectrum: calc. for C 1 5 H 1 7 O 2 N : 243.1258, and for C 1 5H 1 30 2N: 239.0946. Found: 243.1248 and 239.0933. B_ The identical procedure was followed for the cyclization of impure alkylation material. In this manner the alkylation product ( 1 1 . 5 gm., 4.4 x 1 0 " 2 mole) was reacted with methanol-HCl ( 2 . 5 % ) to give an equal mixture of compounds 134 and 209 ( 1 0 . 1 2 gm., 88%). The spectral data were identical to those quoted in part A. Separation of Compounds 134 and 209 Recrystallization of the crude mixture of compounds 134 and 209 (7.15 gm.) - 240 -from concentrated chloroform solution afforded pure compound 1 3 4 as colourless crystals. Column chromatography of a small portion of the mixture (500 mg.) on alumina (20 gm.,) using benzene gave impure compound 209 in the f i r s t fraction and pure compound 134 in subsequent fractions (about 200 mgs., 40%). The latter component was recrystallized from chloroform to give colourless crystals of l-methyl-2-carbomethoxycarbazole (134) with m.p. 138-140°, ( l i t . , m.p. 126-127° for ethyl ester). IR. (nujol): 3425, 1710. UV. A m a x (log £ ) : 340(3.62), 303(4.38), 246(4.69). N.M.R. (F.T.): 3.84 (Singlet, C O O C H 3 ) , 2.70 (Singlet, C H 3 ) , Mass spectrum: M+, m/e = 239 (base peak); main peaks: m/e = 224, 208, 180, High resolution mass spectrum: calc. for C 1 5 H 1 3 O 2 N : 239.0946. Found: 239.0948, Found: C,75.15; H,5.38; N,5.77%. Calc. for C 1 5H 1 30 2N: C,75.30; H,5.48; N,5.85%. By preparative layer chromatography using alumina plates ( 1 m.m.) (Benzene; chloroform 1:1) a small quantity of l-methyl-2-carbomethoxy-l,2,3,4-tetrahydro* carbazole (209) was isolated as an o i l in about 90% purity. UV.; A m a x : 290, 282, 275, N.M.R. (F.T.): 3.96 multiplet, C-l CH), 3.68 (Singlet, C O O C H 3 ) , 3.38 (multiplet, C-l CH). Mass spectrum: M*, m/e = 243; main peaks: 228, 212, 184 (base peak), 169, 168, 157, 144, 143, 130. 1-Methy1-2-carbomethoxycarbazole (134) The cyclization mixture of compounds 134 and 209 (7.15 gm., 2.98 x 10~2 mole) was dissolved in m-xylene (300 ml.) and chloranil (7.15 gm., 2.90 x 10"2 mole) was added. The resultant dark solution was refluxed for 24 hr. after which time i t was diluted with water (300 ml) and washed repeatedly with 10% sodium hydroxide solution to remove dark coloured contaminants. The - 241 -organic layer was then washed with water (3 x 200 ml.), dried over sodium sulphate and concentrated to give a dark crystalline produce (6.5 gm.), Recrystallization from chloroform gave grey coloured crystals of ester (2.6 gm.). Column chromatography of the mother liquors on alumina (200 gm.) using benzene yielded the ester as a yellow o i l which crystallized to colourless crystals on standing (2.7 gm.), m.p. 138-140°. An overall y i e l d of (5.3 gm., 81%) was obtained. IR. (nujol): 3425, 1710. U V . ; A m a x (log§: 340 (3.62), 303 (4.38), 246 (4.69). N.M.R. (F.T.): 3.84 (Singlet, COOCH3), 2.70 (Singlet, CH3). Mass spectrum: M+, m/e = 239 (base peak); main peaks: m/e = 224, 208, 180. Found: C, 75.53: H, 5.65; N, 5.48%, Calc. for C 1 5H 1 30 2N: C, 75.30; H, 5.48; N, 5.85%. Cyclization - Dehydrogenation of 204 to 1-Methyl-2-carbomethoxycarbazole  (134). A solution of methanolic-HCl (2.0%) (10 ml.) was added dropwise over several minutes to a stirred solution of the alkylation product 204 -3 (1.0 gm., 3.86 x 10 mole) and chloranil (1.5 equiv.) in benzene (15 ml.). The reaction mixture was stirred at room temperature for an additional 5 min. after which time i t was concentrated to dryness redissolved in chloroform (50 ml.) and washed successively with 10% aqueous sodium hydroxide (5 x 30 ml.) and water (3 x 30 ml.). The chloroform layer was dried over sodium sulphate and concentrated to give light brown - 242 -crystals. Recrystallization from concentrated chloroform solution yielded 204 as colourless crystals (774 mg., 84%), m.p. 138-140°. The spectral data obtained for 134 obtained from 204 was identical to that previously described. l-Methyl-2-formylcarbazole (152) 203 Jones reagent (5M) was titrated into a solution of the carbazole -3 ' alcohol 157 (1.35 gm., 6.4 x 10 mole) dissolved in acetone at room temperature unt i l a presistently red coloured reaction mixture was obtained (about 1.5 equiv.). The reaction was stirred for an additional five minutes after which time methanol was added u n t i l a green colouration was obtained. The reaction mixture was then suction f i l t e r e d and the residue washed with copious amounts of methanol. The f i l t r a t e was concentrated, taken up in water and exhaustively extracted with ether. The ether extracts were dried over sodium sulphate, concentrated and applied to an alumina column (10 gm.). The product was eluted with chloroform to yield yellow crystalline needles (1.00 gm., 70%). Recrystallization from chloroform afforded an analytical sample, m.p. 165-166° ( l i t . , m.p. 164-164.5).8 IR. (nujol): 1665. UV.; Amax (log e): 365 (3.52), 317 (4.42), 255 (4.63). N.M.R. (FT.): 10.39 (singlet, CHO), 2.87 (singlet, CH^). Mass spectrum: M+, m/e = 209 (base peak); main peaks: 180, 152. High resolution spectrum: Calc. for C 1 4 H i ; l N 0 : 209.0840. Found: 209.0831. Found: C, 80.08; H, 5.30; N, 6.39%. Calc. for C^H^NO: C, 80.36; H, 5.30; N, 6.69%. - 243 -1-Methyl-2-hydroxymethy1carbazole (157) -2 A The Carbazole ester 134 (6.6 gm., 2.77 x 10 mole) dissolved in anhydrous ether (100 ml.) was added dropwise over a period of 0.5 hr. to a _2 slurry of lithium aluminum hydride (2 gm., 5 x 10 mole) in anhydrous ether (250 ml.) at room temperature. Large aggregate lumps of lithium aluminum hydride formed during the addition. The reaction was stirred for an additional 0.5 hr. after which time excess hydride reagent was destroyed by the addition of water (2 ml.), 15% sodium hydroxide (2 ml.) and f i n a l l y water (6 ml.). The resultant granular precipitate was removed by suction f i l t r a t i o n and washed l i b e r a l l y with ether. The combined ether fraction was washed with water (3 x 100 ml.), dried over sodium sulphate, and concentrated to give a colourless crystalline solid (5.26 gm., 90%). Recrystallization from chloroform gave an analytical sample of the alcohol 15 7 with m.p. 184 - 186° ( l i t . , m.p. 187 - 188°). IR. (nujol): 3420, 3400-3325. UV.; Amax (log e): 335 (3.63), 324 (3.68), 294 (4.36), 285 (4.16), 257 (4.33), 247 (4.56), 238 (4.73). N.M.R. (FT.): 4.88 (singlet, Ar-CH20H), 2.58 (singlet, CH 3), 1.50 (singlet, OH). Mass spectrum: M+, m/e = 211 (base peak); main peaks: 194, 180, 167. High resolution mass spectrum: Calc. for C^H^NO; 211.0996. Found: 211.0982. Found: C, 79.63; H, 6.49; N, 6.37%. Calc. for C^H^NO: C, 79.59; H, 6.20; N, 6.63%. -2 Impure carbazole ester 134 (9.3 gm., 3.89 x 10 mole) was reduced to the alcohol 157 in an identical manner. The dark crystalline material (7.2 gm.) - 244 -that was obtained was p u r i f i e d by column chromatography on alumina (350 gm.). El u t i o n with benzene:chloroform 1:1 followed by chloroform:methanol 5% yielded the alcohol 157 as colou r l e s s c r y s t a l s (4.9 gm., 60%), 209 l-Methyl-2- (co-nitrovinyl) carbazole (160) _2 The aldehyde 152 (2.50 gm., 1.19 x 10 mole) was taken up i n nitromethane (20 ml.), ammonium acetate (0.50 gm.) was added and the mixture was heated at 100°C f o r 0.5 hr. During the course of heating the s t a r t i n g material f i r s t d issolved then an orange p r e c i p i t a t e of product p r e c i p i t a t e d out. The cooled r e a c t i o n mixture was suction f i l t e r e d and the product obtained was washed with col d ethanol and dr i e d to give orange needles (2.62 gm.). Concentration of the mother l i q u o r s , d i s s o l u t i o n i n water and repeated e x t r a c t i o n with chloroform y i e l d e d f u r t h e r q u a n t i t i e s of orange c r y s t a l l i n e product (0.230 gm., o v e r a l l y i e l d 95%). R e c r y s t a l l ^ i z a t i o n from ethanol (250 ml/gm.) gave long crimson needles, m.p. 242-244° ( l i t . , m.p. 244-247°). IR. (n u j o l ) : 3340, 1360. UV.; Amax (log e): 375 (4.39), 292 (3.87), 248 (4.50), 237 (4.46). N.M.R. (FT.): 8.51 (doublet, J = 14 Hz, -CH=CH-N02), 7.59 (doublet, J = 14 Hz, -CH=CH-N02), 2.64 ( s i n g l e t , CH^). Mass spectrum: M+, m/e = 252; main peaks: 220, 205 (base peak), 180. High r e s o l u t i o n spectrum: Calc, f o r C j ^ H ^ N ^ ^ 252.0938. Found: 252.0905. Found: C, 71.70; H, 4.97; N, 10.82%, Calc. f o r C 1 C H 1 0 N 0 0 0 : C, 71.42; H, 4.79; N, 11.10%. - 245 -l-Methyl-2-(g-aminoethyl)carbazole (139) This compound was prepared according to the procedure developed by 177 Moster et ad. -2 A solution of the nitrostyrene 160 (2.85 gm., 1.13 x 10 mole) in tetrahydrofuran (50 ml.) was added dropwise over 0.5 hr. to a stirred suspension of lithium aluminum hydride (2 gm.) in tetrahydrofuran (250 ml.) at room temperature. Stirring was continued for an additional 20 min., and the hydride was then decomposed by the successive addition of water (2 ml.), 10% aqueous sodium hydroxide (6 ml.), water (2 ml.). The resultant granular precipitate was f i l t e r e d and washed with tetrahydrofuran (200 ml.). The combined f i l t r a t e s were concentrated to a paste, taken up in ether (75 ml.) and washed with water (3 x 30 ml.). The ether layer was dried over sodium sulphate and concentrated to give 139 as a s l i g h t l y yellowed crystalline solid (2.52 gm., 95%). UV., X^: 335,320,293,284(sh) 257,245,236. N.M.R. (for HC1 s a l t ) : 4.76 (singlet, 2H, NHz), 2.40 (singlet, 3H, C-l CH,). Mass spectrum: M+, m/e = 224. - 246 -l-Methyl-2-(B-(N-acetylamino)ethyl)carbazole (140) This compound was prepared following the procedure developed by 172 Schmutz and Wittwer. _2 The amine 139 (2.52 gm., 1.12 x 10 mole) was dissolved in a mixture of pyridine (30 ml.) and acetic anhydride (3 ml.) and stirred at 60° for 0.5 hr. The solvent was removed under vacuum and the residue was dissolved in ether (75 ml.) and washed with dilute hydrochloric acid (3 x 30 ml.) and water (3 x 30 ml.). The ether layer was then dried over sodium sulphate and concentrated to give 140 as a colourless crystalline solid (2.93 gm., 95%). UV.; X m a x: 335, 320, 294, 283 (sh), 257, 245, 238. IR. (CHCI3): 1660 cm"1. N.M.R: 5.56 (broad hump, 1 H, NHCOCH3); 3.46 (triplet, 2 H, J = 6 Hz, ArCH2CH N), 3.00 (multiplet, 2 H, ArO^O^N), 2.46 (singlet, 3 H, C-l CH 3), 1.93 (singlet, 3 H, COCH3). mass spectrum: M+, m/e = 266; main peaks: 223, 220, 207 (base peak), 193, 180, 167. 3,4-Dihydro-l,5-dimethyl-6H-pyrido-(4,3-b)carbazole (141) . Phosphorous oxychloride (9 ml.) was added to a solution of the amide 140 (1.45gm., 5.45 x 10"3 mole) in hot toluene (250 ml.). The reaction mixture was refluxed for 1 hr. after which time the solvent was removed to give a green cry-stalline solid which was extracted with hot dilute hydrochloric acid (3 x 250 ml.) Insoluble residues were removed by suction f i l t r a t i o n . The combined acid frac-tions were basified to pH 11 with concentrated ammonium hydroxide and extracted with chloroform (4 x 300 ml.). The combined chloroform extracts were dried over - 247 -sodium sulphate and concentrated to give brown crystalline material. The crude product was recrystallized from chloroform or chloroform:methanol mix-tures to give 3,4-dihydroolivacine (141) as colourless crystals (0.85 gm., 63%), m.p. 310° ( l i t . , m.p. 300-315°). UV. (H20 $ aq HC1); X m a x log (e): 370 (3.93), 340 (3.71), 324 (3.84), 310 (4.15), 290 (sh) (4.35), 280 (4.56), 273 (4.54), 245 (4.36), 234 (4.44). UV. (aq. NaOH); X m a x: 283, 272, 245, 236. N.M.R.: 3.76 (triplet, 2 H, J = 6Hz, C-4 CH 2), 2.90 (tri p l e t , 2 H, C-3 CH 2), 2.60 (distorted singlet, 3 H, C-l CH 3), 2.50 (singlet, 3 H, C-5 CH 3). Mass spectrum: M*, m/e = 248 (base peak); main peaks: 233, 220-217, 204, 191. Found: C, 82.14; H, 6.55; N, 10.99. Calc. f o r C 1 7 H 1 6 N 2 : C, 82.22; H, 6.49; N, 11.28. The imine 141 was converted to i t s methiodide 142 in quantitative yield by reaction with methyl iodide i n methanol:chloroform solution at room tempera-ture for 4-5 hr. The pale orange methiodide 142 was recrystallized from methanol, m.p. > 300° ( l i t . , m.p. 320°). UV. (H20) (Figure 38); X ^ t l o g e): 365 (4.33) 308 (4.32), 300 (4.17), 279 (4.61), 265 (sh) (4.37), 225 (4.48). UV. (aq. NaOH); Xm : 365, 335, 310, 294, 280, 270 (sh), 243, 237. N.M.R.: 8.86 (singlet, 1 H, C- l l H), 3.98 (multiplet, 2 H, C-4 CH 2), 3.68 (singlet, 3 H, N-CH3), 3.20 (multiplet 2 H, C-3 CH 2), 2.90 (singlet, 3 H, C-l CH3), 2.48 (singlet, 3 H, C-5 CH 3). Found: C, 55.00; H, 4.90; N, 7.18. Calc. for C 1 8H 1 RN~I: C, 55.40; H, 4.87; N, 7.06. - 248 -l,5-dimethyl-l,2,3,4-tetrahydro-6H-pyrido-4(,3-b)carbazole (25)  (Guatambuine) A. The methiodide 142 (100 mg., 2.64 x 10~4 mole) was dissolved i n methanol (20 ml.) and the solution was hydrogenated at room temperature and atmospheric pressure over platinum oxide (Pt02» 20 mg.) for 2 hr. The mix-ture was f i l t e r e d to remove the catalyst which was carefully washed with additional methanol (10 ml.). The combined f i l t r a t e s were concentrated to give a pale yellow solid. The product was obtained as a colourless crystalline solid by f i l t r a t i o n through a small alumina column (CHCl3:MeOH 5%). A l l analy-t i c a l and spectral data i s presented below. B. The methiodide 142 (250 mg., 6.17 x 10"4 mole) was dissolved in aqueous ethanol (100 ml.) and reacted with an excess of sodium borohydride at room temperature for 2 hr. The reaction mixture was then concentrated, taken up i n water (75 ml.) and extracted with chloroform (3 x 30 ml.). The combined chloroform fractions were dried over sodium sulphate, and concentrated to give colourless crystals (160 mg., 98%). The product was recrystallized from 128 methanol to give 25 as cream coloured cubes m.p. 248-250° ( l i t . , 248-250°). UV.; (log e): 341 (3.48), 327 (3.63), 297 (4.26), 288 (sh) (4.08), 262 (4.36), 250 (4.50), 240 (4.64). N.M.R. (FT.): 3.90 (quartet, 1 H, J = 7 Hz, C-l H), 2.40 (singlet, 3 H, C-5 CH 3), (singlet, 3 H, N-CH3), 1.52 (doublet, 3 H, J = 7 Hz, C-l CH 3). Mass spectrum: M+, m/e = 264; main peaks: 249, 233, 221-218, 204. High resolution mass spectrum: Calc. for C j g ^ o ^ : 264.1626. Found: 264.1626. Found: 264.1656. Found: C, 81.55; H, 7173; N, 10.76. Calc. for C 1 8H 2 0N 2: C, 81.78; H, 7.63; N, 10.60. - 249 -l,5-dimethyl-6H-pyrido-(4,3-b)carbazole (16) (Olivacine) Dihydroolivacine (83) (30 mg., 1.20 x 10~4 mole) was suspended in deca-l i n (4 ml.) containing 5% palladium on charcoal (50 mg.) and refluxed for 1.5 hr. The reaction mixture was cooled to 0° and the solvent was removed by suction f i l t r a t i o n . The residue containing the olivacine (16) was taken up in hot chloroform (200 ml.) and re f i l t e r e d . The washing procedure was repeated (3 x 200 ml.) and the combined chloroform f i l t r a t e s were concentrated to give an orange coloured so l i d . The crude product mixture was recrystallized from methanol to give olivacine (16) as bright yellow needles (20 mg., 70%), m.p. 126 318-324° ( l i t . , m.p. 318-326°). UV.; Vax( l o2 e ) : 3 7 4 ( 3- 5 6)» 3 2 5 ( 3- 76), 291 (4.84), 275 (4.72), 265 (4.57),235 (4.33), 222 (4.40). N.M.R. (FT.): 3.16 (singlet, 3 H, C-l CH 3), 2.82 (singlet, 3 H, C-l CH 3). Mass spectrum: M+, m/e = 246. High resolution mass spectrum: Calc. for C-^H^^: 246.1156. Found: 246.1165. Found: C, 82.46; H, 5.94; N, 11.20. Calc. for C J J H J ^ ^ : C, 82.90; H, 5.73; N, 11.37. - 250 -Sequence B. 1,2,3,4-tetrahydrocarbazole (216) This compound was prepared according to the procedure of Rogers and 207 Corson. Cyclohexanone (98 gm., 1.0 mole) and phenylhydrazine (108 gm., 1.0 mole) were reacted in acetic acid (360 ml.) at reflux temperature for 1 hour. The reaction mixture was then cooled to about 5° and the crystalline mass which precipitated was collected by suction f i l t r a t i o n . The f i l t e r cake was washed with water (100 ml.), then with 75% ethanol (100 ml.) and dried under vacuum. The crude product was crystallized from methanol (700 ml.) after treatment with activated charcoal to y i e l d colourless crystals of tetrahydro-carbazole (216) (135 gm., 79%), m.p. 113-115° ( l i t . , m.p. 115-116°). 9-Benzyl-l,2,3,4-tetrahydrocarbazole (217) Tetrahydrocarbazole (216) (10 gm., 5.84 x 10~2 mole) and benzyl bromide (7.10 ml., 5.84 x 10"2 mole) were dissolved in dry dimethylformamide (100 ml.) and stirred at 0°C. Sodium hydride (1.44 gm., 6.00 x 10 - 2 mole) was added in four portions over one hour. The reaction mixture was stirred for an additional hour after which time i t was carefully diluted with water (400 ml.) and extracted with ether (3 x 200 ml.). The ether fractions were combined and washed with water, dried over sodium sulphate and concentrated to give a colourless mobile o i l . Small amounts of starting material were removed by column chromatography on alumina (400 gm., Act III). Addition of the crude reaction mixture to the column i n benzene and elution with petroleum ether yielded the desired 9-benzylated tetrahydrocarbazole 217 as a transparent o i l (11.6 gm. 77%). N.M.R.: 5.10 (Singlet, 2H, NCH^CftHQ, 2.60 (multiplet, 4H, C-2,3CH?), 1.86 - 251 -(multiplet, 4H, C-1,4CH ). mass spectrum: M+, m/e = 261. 1-Methy1-3-formy1-9-benzylcarbazole (219) A solution of 9-benzyltetrahydrocarbazole (217) (10 gm., 3.83 x 10"2 mole) in dry dimethylformamide (50 ml.) with phosphorous oxychloride (4.55 ml., 1.3 equiv.) was stirred at 100°C for eight hours. Hydrolysis of the reaction mixture was then effected by addition of 30% potassium acetate solution (20 ml.) with heating for an additional 20 minute period. The cooled mixture was then diluted with water (200 ml.) and extracted with ether (5 x 100 ml.). The combined ether fractions were washed with water, dried over sodium sulphate and concentrated to give a viscous dark o i l . The crude product was purified i n i t i a l l y by column chromatography on alumina (250 gm. Act III). The crude mixture was applied to the column in benzene and the desired carbazole aldehyde 219 was eluted with petroleum ether:benzene 1:1 to give an orange glass (5.47 gm. 48%). By subsequent recrystallization form petroleum ether-benzene colourless needles were obtained, m.p. 102-106°. m ' » max ( l o2 ) : 3 4 0 ( s h) (4.22), 327 (4.30), 289 (4.60), 273 (4.72), 244 (4.60), 235 (4.66). IR. ( C H C I 3 ) : 1670. N.M.R. (100 Mhz): 10.00 (Singlet, IH, CHO), 8.46 (distorted singlet, IH, C-4H), 8-10 (multiplet, IH, C-5H), 7.66 (distorted singlet, IH, C-2H), 7.4-7.l(multiplet, 6H, aromatic) 6.9 (multiplet, 2H, aromatic), 5.72 (Singlet, 2H, benzyl CH2)» 2 - 6 0 (Singlet, 3H, C - I C H 3 ) . Mass spectrum: M+> m/e = 299 (base peak). High resolution mass spectrum: calc. for C 2 1 H 1 7 N O : 299.1305. Found: 299.1309. Found: C,8405; H,5.62; N,4.57. Calc. for C21H17NO: C,84.25; H,5.72; N,4.68. - 252 -l-Methyl-3-(a-hydroxyethyD-9-benzylcarbazole (222) To the carbazole aldehyde 219 (112 gm., 3.74 x 10~2 mole) in dry tetra-hydrofuran (100 ml.) was added in one portion an excess of methylmagnesium * chloride - THF solution (20 ml., 3.16M). The reaction was stirred at room temperature for 0.5 hr. after which time the excess Grignard reagent was destroyed by the careful addition of dilute hydrochloric acid. The mixture was concentrated until a l l the tetrahydrofuran was removed, diluted with water (250 ml.) and extracted with ether (3 x 100 ml.). The ether fractions were combined, washed with water, dried over sodium sulphate and then concen-trated to give a pale yellow o i l (11 gm.). The crude reaction mixture was purified by column chromatography on alumina (200 gm., Act III). By elution with chloroform and ethyl acetate the carbazole alcohol 222 was obtained as a colourless foam (7.5 gm., 64%). UV.; X m a x: 342, 327, 293, 287, 283, 263, 237. N.M.R.: 8.08 (doublet, 1H, J = 2Hz, C-5H), 7.98 (multiplet, 1H, C-4H), 7.2, 6.95 (2 multiplets, 9H, aromatic), 5.66 (singlet, 2H, N-O^-C^), 5.00 (quartet, 1H, J = 6Hz, ArpH_(0H) CH^, 2.54 (Singlet, 3H, C-1CH3), 1.97 (Singlet, 1H, OH), 1.54 (doublet, 5H,ArCH(OH)CH3) . Mass spectrum: M+, m/e = 315 (base peak); main peaks: 300, 297. High resolution mass spectrum: calc. for C 2 2H 2iNO: 315.1623. Found: 315.1632. l-Methyl-3-("-acetoxyethyl)-9-benzylcarbazole (223) The Carbazole alcohol 222 (7 gm., 2.22 x 10"2 mole) was taken up in pyridine (75 ml.) and acetic anhydride (10 ml.) and gently heated at 60°C for one hour. The solvent was then removed and the resulting residue taken up in ether (350 * Fisher Reagent. - 253 -ml.) and washed with 5 % sodium hydroxide solution and with 5% hydrochloric acid respectively. The ether layer was then dried over sodium sulphate and concentrated to give a tan coloured gum. The crude reaction product was fil t e r e d through a short alumina column and recrystallized from methanol, m.p. 137-140 . UV; X m Q V (loge): 341 ( 3 . 7 5 ) , 327 (3.71), 293 (4.30), 283 (4.04), 263 ( 4 . 5 5 ) , 238 ( 4 . 7 9 ) , 230 ( 4 . 7 7 ) . IR.: 1720. N.M.R.: 8.05 (multiplet, 1H, C-5H), 7.98 (distorted singlet, 1H, C-4H), 7 . 3 - 6 . 8 (two multiplets, 9 H , aromatic), 6.00 (quartet, 1H, J = 7 H z , Ar-CH(CH 3)0A c), 5 . 6 8 (Singlet, 2H, benzyl O^), 2.52 (Singlet, 3H, C-ICH3) 1.98 (Singlet, 3H, - C O C H 3 ) , 1.56 (doublet, 3H, J = 7 H z , A r - C H ( C H 3 ) 0 A c ) . Mass spectrum: M+, m/e = 3 5 7 (base peak); main peak: 298. High resolution mass spectrum: Calc. for C 2 4 H 2 3 N 0 2 : 357.1727. Found: 357.1676. Found: C,80.40; N, 6.38; N, 3 . 9 3 . Calc. for C 2 4 H 2 3 N 0 2 : C, 80.64; H, 6 . 4 9 ; N, 3.92. l-Methyl-3-(ct- (N- B ,.g -diethoxyethylamino) ethyl) -9-benzylcarbazole (2Z4) from (323) .  The carbazole acetate 223 ( 5 0 mg. 1.40 x 1 0 " 4 mole) was dissolved in aminoacetaldehyde diethylacetal (171) ( 1 m l . ) and heated at reflux temperature (160°) for 15 hr. after which time the excess amine 171 was removed under high vacuum. The dark orange o i l which was obtained was column chromatographed on alumina ( 3 gm.). A small quantity of the desired aminoethyl acetal 224 was obtained by elution with benzene. N.M.R.: 8.14 (multiplet, 1H, C-5H), 7.98 (Singlet, 1H, C-4H), 7.2 (complex multiplet, aromatic), 5.72 (singlet, 2H, N-CH2 C 6H 5), 4.64 (t r i p l e t , 1H, J = 6 H z , C H ( 0 E + ) 2 ) , 3.95 (quartet, 1H, J = 6 H z , - 254 -Ar-CH(CH3)-N), 3.58 (complex multiplet, 4 H , ffl(OCH2CH3)2), 2.69 (doublet, 2H, J = 6 H z , -N-CH2-), 2.60 (Singlet, 3H, C - 1 C H 3 ) , 2.0 (broad singlet, 1H, N(b)H), 1.47 (doublet, 3H, J = 6 H z , Ar-CH(CH3)-N), 1.20 (complex multiplet, l-Methyl-3-(g,g-diethoxyethyliminomethy1)-9-benzylcarbazole (221) l-Methyl-3-formylcarbazole (219) ( 2 0 gm. 6.68 x 10"3 mole) and amino-acetaldehyde diethylacetal (171) (1 ml., 8.26 x 10 3 mole) were heated at 100° for two hours, then benzene (15 ml.) was added and d i s t i l l e d o f f to remove water formed in the reaction. The residue crystallized from petroleum ether (65-110°) - benzene to give a colourless crystalline powder (2.34 gm., 85%), m.p. 92-94 . UV. (MeOH) Amax ( l o g s ) : 3 4 5 (3.87), 330(sh) (4.07), 310 (4.23), 285 (4.62), 275 (4.61), 247 (4.5 9 ) , 237 (4.60). (MeOH-HCl) A r a a x: 380, 313, 305 280, 270, 233. IR.: 1690. N.M.R.: 8.40 (Singlet, 1H, C-4H), 8.32 (distorted singlet, 1H, CH = N-), 8.10 (multiplet, 1H, C-5H), 7.60 (Singlet, 1H, C-2H), 7.4 7.1 (multiplet, 6 H , aromatic), 7.0 (multiplet, 2H, aromatic), 5.72 (Singlet, 2 H , benzyl H), 4.81 ( t r i p l e t , 1H, J = 5 H z , C H ( 0 E t ) 2 ) , 3.70 (complex multiplet, 6H, C H ( 0 C H 2 C H 3 ) 2 ) , = N-CH2-), 2.58 (Singlet, 3H, C - I C H 3 ) , 1.14 ( t r i p l e t , 6 H , J = 7 H z CH(OCH2CH3) ). Mass spectrum: M+, m/e = 414 (base peak); main peaks: 369, 3 3 9 , 311, 298, 284. High resolution mass spectrum: Calc. for C 2 7 H 3 0 N 2 O 2 ; 414.2307. Found: 414.2250. Found: C.77.95, H, 7 . 2 0 ; N, 6.56. Calc. for C 2 7 H 3 0 N 2 O 2 : C, 78.23; H, 7.29; N, 6.76. - 255 l-Methyl-3-(g-(N- 3, g-diethoxyethylanu.no)ethyl)-9-benzylcarbazole (224) The iminomethyl acetal 221 (1.6 gm., 3.86 x 10~3 mole) in dry tetrahy-drofuran (50 ml.) was treated with an excess of methylmagnesium chloride r (4.8 ml., 3.16M) at reflux temperature for 24 hr. The excess Grignard reagent was then destroyed by the addition of water and the resultant precis pitate was removed by suction f i l t r a t i o n and washed lib e r a l l y with ether, The f i l t r a t e was concentrated, taken up in water (250 ml.) and extracted with ether (3 x 100 ml.). The ether layer was washed with water, dried over sodium sulphate and concentrated to give an orange foam (1.52 gm.). The crude product mixture was column chromatographed on alumina (100 gm., Act III) and the desired aminoethyl acetal was eluted with benzene:chloroform (1:1) and chloroform to give a faintly coloured o i l (330 mg. 20%). UV; ^ m S L X - 341, 327, 293, 284, 263, 239, 231. N.M.R.: 8.14 (multiplet, IH, C-5H), 7.98 (Singlet, IH, C-4H), 7.2 (complex multiplet, aromatic ( C H C I 3 contamination)), 5.72 (Singlet, 2H, N-CH2 C 6H 5), 4.64 (t r i p l e t , IH, J = 6Hz, CH(0Et) 2), 3.95 (quartet, IH, J = 6Hz, Ar-CH(CH.3)-N), 3.58 (complex multiplet, 4H, OUOCH^CH^^), 2.69 (doublet, 2H, J = 6Hz, -N-CH2-), 2.60 (Singlet, 3H, C - I C H 3 ) , 2.0 (broad singlet, IH, N(b)H), 1.47 (doublet, 3H, J = 6Hz, Ar-CH(CH3)-N), 1.20 (complex multiplet, + 6H, CH(OCH2CH3)). Mass spectrum: M , m/e = 430; main peaks: 415, 369, 298 (base peak), 288. High resolution mass spectrum: Calc. for C 2 8 N 3 4 N 2 O 2 : 430.2620. Found: 430.2633. - 256 -Bibliography ( For Part I ) 1. M. Calvin, Chemical Evolution, Oxford University Press (1969) 2. L. E. Orgel, The Origins of Life - Molecules and Natural Selection, J. Wiley § Sons (1973). 3. C. Ponamperuma, The Origins of Life, Thames and Hudson, London (1972). 4. R. C. Crozier, Traditional Medicine in Modern China, Harvard University Press (1968). 5. L. Aikman, Nature's Gifts to Medicine, Journal of National Geographic Society, 420 , 1974. 6. M. B. Krieg, Green Medicine, Rand McNally § Company (1964), 7. V. Snieckus, The Alkaloids, XI, 1, R. F. Manske, ed., Academic Press, New York (1968). 8. R. Hegnauer, Planta Med., 1 (1958), 9. M. Hesse, Indolalkaloide in Tabellenen, Springer, Berlin, 1_, (1964), 2, (1968). 10. B. Gilbert, A. P. Duarte, Y. Nakagawa, J. A. Joule, S. E. Flores, J. Aguayo Brissolese, J. Campello, E. P. Carrazzoni, R. J . Owellen, E. C. Blossey, K. S. Brown Jr., C. Djerassi, Tetrahedron, 21, 1141 (1965) . 11. Chemical Abstracts, Eighth Collective Index, Index Guide, 66-75 (1971), 12. I. Kompis, M. Hesse, H. Schmid, Lloydia, 34(3), 269 (1971). 13. G. A. Cordell, Lloydia, 37 (2), 219 (1974). 14. A. Pictet, Arch. Pharm., 224, 389 (1906), 15. E. Leete, Tetrahedron, 14, 35 (1961). 16. E. Leete, JN_ Am. Chem. Soc, 82, 6338 (1960), 17. E. Leete, A. Ahmad, I. Kompis, Jj_ Am. Chem. Soc., 87, 4168 (1965). 18. J. P. Kutney, W. J. Cretney, J. R. Hadfield, E. S. Hall, V. R. Nelson, D. C. Wigfield, J ^ Am. Chem. Soc, 90, 3566 (1968). 19. A. I. Scott, Acc. of Chem. Res., 3, 151 (1970) and references therein, - 257 -20. A. R. Battersby, Specialist Periodical Reports, The Chemical Society, The Alkaloids, 1, 31, (1970) and references therein. 21. G. Barger, C. Scholz^Helv. Chim. Acta., 16, 1343 (1933). 22. G. Hahn, H. Werner, Liebigs Ann. Chem., 520, 123 (1935). 23. R. Robinson, The Structural Relation of Natural Products, Claredon Press Oxford (1955). 24. R. B. Woodward, Angew. Chem., 68, 13 (1956). 25. R. B. Woodward, Nature, 162, 155 (1948). 26. R. B. Turner, R. B. Woodward, The Alkaloids, VIII, R. F. Manske, ed., Academic Press, New York (1953). 27. E. Wenkert, N. V. Bringi, J . Am. Chem. Soc, 81, 1474 (1959). 28. E. Wenkert, J. Am. Chem. Soc, 84, 98 (1962). 29. J. LeMen, W. I. Taylor, Experientia,21, 508 (1965). 30. E. Sc h l i t t l e r , W. I. Taylor, Experientia, 16, 224 (1960). 31. E. Leete, S. Ghosal, P. N. Edwards, J. Am. Chem. S o c , 84, 1068 (1962). 32. P. N. Edwards, E. Leete, Chem and Ind., London, 1666 (1961). 33. E. Leete, S. Ghosal, Tetrahedron Letters, 1179 (1962). 34. G. Buchi, R. E. Manning, Tetrahedron Letters, 5 (1960). 35. C. Djerassi et a l . , J . Org. Chem., 26, 1192 (1961). 36. W. Haegele, F. Kaplan, H. Schmid, Tetrahedron Letters, 110 (1961). 37. J. Grimshaw, Chem. and Ind., 403 (1961). 38. R. Thomas, Tetrahedron Letters, 544 (1961). 39. A. R. Battersby, R. Binks, W. Laurie, G. V. Parry, B. R. Webster, Proc. Chem. Soc, 369 (1963). 40. A. R. Battersby, R. Binks, W.-Laurie, G. V. Parry, B. R. Webster, J. Chem. Soc, 7459 (1965). 41. F. McCapra, T. Money, A. I. Scott, I. G. Wright, Chem. Comm., 537 (1965) 42. H. Goeggel, D. Arigoni, Chem. Comm., 538 (1965). 258 -43. A. R. Battersby, R. T. Brown, R. S. Kapil, A. 0. Plunkett, J. B. Taylor, Chem. Comm., 46 (1965). 44. A. R. Battersby, R. T. Brown, R. S. Kapil, J. A. Knight, J . A. Martin A. 0. Plunkett, Chem. Comm., 888 (1966). 45. A. I. Scott, Karl Folkers Lecture Series, University of Wisconsin, December 1964. 46. T. Money, I. G. Wight, F. McCapra, A. I. Scott, Pros. Nat. Acad. Sci . U.S., 53, 901 (1965). 47. E. S. Hall, F. McCapra, T. Money, K. Fukumoto, J. R. Hanson, B.S. Mootoo, G. T. Ph i l l i p s , A. I. Scott, Chem. Comm., 348 (1966). 48. T. Money, I. G. Wright, F. McCapra, E. S. Hall, A. I. Scott, J. Am. Chem. Soc, 90, 4144 (1968). 49. A. R. Battersby, R. T. Brown, J. A. Knight, J. A. Martin, A. 0. Plunkett, Chem. Comm., 346 (1966). 50. A. R. Battersby, Pure Appl. Chem., JL4, 117, (1967). 51. P. Loew, H. Goeggel, D. Arigoni, Chem. Comm., 342 (1966). 52. E. Leete, S. Ueda, Tetrahedron Letters, 4915 (1966). 53. A. R. Battersby, P. T. Brown, R. S. Kapil, A. 0. Plunkett, Chem. Comm., 890 (1966). 54. P. Loew, D. Arigoni, Chem. Comm., 137 (1968). 55. A. R. Battersby, R. S. Kapil, J . A. Martin, Mrs. Lucy Mo, Chem. Comm., 133 (1968). 56. A. R. Battersby, J . C. Byrne, R. S. Kapil, J. A. Martin, T. G. Payne, D. Arigoni, P. Loew, Chem. Comm., 951 (1968). 57. R. Guarnaccia, L. Botta, C. J. Coscia, J. Am. Chem. Soc, 92, 6098 (1970). 58. R. Guarnaccia, L. Botta, C. J. Coscia, J. Am. Chem. Soc, 96, 7079 (1974). 59. A. R. Battersby, R. S. Kapil, R. Southgate, Chem. Comm., 131, (1968). 60. A. R. Battersby, E. S. Hall, R. Southgate, J. Chem. Soc. (c), 721 (1969). 61. S. Brechbuhler-Bader, C. J . Coscia, P. Loew, C. von Szezepanski, D. Arigoni, Chem. Comm., 136 (1968). - 259 -62. R. Guarnaccia, L. Botta, C. J. Coscia, Am. Chem. Soc., 91, 204 (1969). 63. C. J. Coscia, R. Guarnaccia, Chem. Comm., 138 (1968). 64. A. R. Battersby, B. Gregory, Chem. Comm., 134 (1968). 65. S. Escher, P. Loew, D. Arigoni, Chem. Comm., 823 (1970). 66. A. R. Battersby, S. H. Brown, T. G. Payne, Chem. Comm., 827 (1970). 67. A. R. Battersby, A. R. Burnett, P. G. Parsons, Chem. Comm., 826 (1970). 68. H. Inouye, S. Ueda, Y. Aoki, Y. Takeda, Tetrahedron Letters, 2351 (1969). 69. A. R. Battersby, A. R. Burnett, P. G. 1187 (1969). Parsons, J. Chem. S o c , (c), 70. A. R. Battersby, A. R. Burnett, P. G. (1969). Parsons, J. Chem. S o c , (c), 1193 71. G. N. Smith, Chem. Comm., 912 (1968). 72. W. P. Blackstock, R. T. Brown, G. K. Lee, Chem . Comm., 910 (1971). 73. K. T. D. DeSilva, G. N. Smith, K. E. H. Warren , Chem. Comm., 905 (1971). 74. A. R. Battersby, A. R. Burnett, E. S. Hall, P. G. Parsons, Chem. Comm., 1582 (1968). 75. R. T. Brown, G. N. Smith, S. J. Stapleford, Tetrahedron Letters, 4349 (1969) 76. A. R. Battersby, K. H. Gibson, Chem. Comm., 902 (1971). 77. A. A. Qureshi, A. I. Scott, Chem. Comm., 948 (1968). 78. A. I. Scott, P. C. Cherry, A. A. Qureshi, Am. Chem. Soc , 91, 4932 (1969). 79. A. I. Scott, Accts. Chem. Res., 3, 151 (1970). 80. E. Wenkert, B. Wickberg, J. Am. Chem. Soc, 87, 1580 (1965). 81. A. I. Scott, A. A. Qureshi, J. Am. Chem. S o c , 91, 5874 (1969). 82. G. A. Cordell, G. F. Smith, Manuscript in preparation for J. Chem . Soc 83. E. J. Shellard, P. J. Houghton, Planta Med., 21, 16 (1972). 84. E. J. Shellard, K. Sarpong, P. J. Houghton, J. Pharm. Pharmacol., 2445 (1971). 23, 85. A. R. Battersby, E. S. Hall, Chem. Comm., 793 (1969). - 260 -86. A. I. Scott, P. B. Reighardt, M. B. Slaytor, J. G. Sweeny, Bioorg. Chem., 1_, 157 (1971). 87. A. Sandoval, F. Walls, J. N. Shoolery, J. M. Wilson, H. Budzikiewicz, C. Djerassi, Tetrahedron Letters, 409 (1962). 88. J. Levy, Doctoral Thesis, University of Paris (1962). 89. J . P. Kutney, E. Piers, R. T. Brown, Am. Chem. Soc, 92, 1700 (1970). 90. J . P. Kutney, W. J. Cretney, J. R. Hadfield, E. S. Hall, V. R. Nelson, J. Am. Chem. Soc., 92, 1704 (1970). 91. J . P. Kutney, R. T. Brown, E. Piers, J . R. Hadfield, Am. Chem. Soc., 92,.1708 (1970). 92. J. P. Kutney, W. J. Cretney, P. Le Quesne, B. McKaque, E. Piers, J . Am. Chem. Soc, 92, 1712 (1970). 93. J . P. Kutney, N. Abduraham, C. Gletsos, P. Le Quesne, E. Piers, I. Vlattas, Am. Chem. Soc, 92, 1727 (1970). 94. J. P. Kutney, W. J. Cretney, J . R. Hadfield, E. S. Hall, V. R. Nelson, D. C. Wigfield, Am. Chem. Soc., 90, 3566 (1968). 95. J . P. Kutney, C. Ehret, V. R. Nelson, W. C. Wigfield, J_^ _ Am. Chem. Soc, 90, 5929 (1968). 96. A. A. Qureshi, A. I. Scott, Chem. Comm., 945 (1968). 97. A. A. Qureshi, A. I. Scott, Chem. Comm., 949 (1968). 98. D. A. Evans, G. F. Smith, G. N. Smith, K. S. J. Stapleford, Chem. Comm., 859 (1968). 99. D. A. Evans, J. A. Joule, G. F. Smith, Phytochemistry, 7, 1429 (1968). 100. G. A. Cordell, G. F. Smith, G. N. Smith, Chem. Comm., 193 (1970). 101. R. T. Brown, G. F. Smith, K. S. Stapleford, D. A. Taylor, Chem. Comm., 190 (1970). 102. A. R. Battersby, A. K. Bhatnager, Chem. Comm., 193 (1970). 103. R. S. Sood, Studies Related to the Biosynthesis of Indole Alkaloids, PhD. Thesis, The University of British Columbia (1970). 104. A. R. Battersby, A. K. Bhatnager, Chem. Comm., 189 (1970). 105. J. P. Kutney, J . F. Beck, N. J. Eggers, H. W. Hanssen, R. S. Sood, N. D. Westcott, J. Am. Chem. Soc, 93, 7322 (1971). - 261 -106. J . F. Beck, Studies on the Biosynthesis of Indole Alkaloids, PhD. Thesis, The University of British Columbia (1971). 107. J. P. Kutney, C. Ehret, G. Poulton, R. S. Sood, N. D. Westcott, Bioorg. Chem., 1, 194 (1971). 108. J. P. Kutney, J. Heterocyclic Chem., 9, Supplementary Issue, S-l (1972). 109. J . P. Kutney, J. F. Beck, V. R. Nelson, R. S. Sood, J. Am. Chem. S o c , 93, 255 (1971). 110. J . P. Kutney, V. R. Nelson, D. C. Wigfield, J . Am. Chem. Soc, 91, 4298 (1969) , 111. N. J. Eggers, Studies on the Biosynthesis of Indole Alkaloids, PhD. Thesis, The University of British Columbia (1973). 112. J . P. Kutney, N. J. Eggers, Heterocycles, 1, 11 (1973), 113. J . A. Joule, H. Monteiro, L. J. Durham, B. Gilbert, C. Djerassi, J ^ Chem. Soc, 4773 (1965). 114. I. K. Kim, K. L. Erickson, Tetrahedron, 2_7, 3979 (1971). 115. J. A. Joule, M. Okashi, B. Gilbert, C. Djerassi, Tetrahedron, 21, 1717 (1965). 116. R. R. Arndt, S. H. Brown, N. C. Ling, P. Roller, C. Djerassi, J. M. Ferreira, B. Gilbert, E. C. Miranda, S. E. Flores, A. P. Duarte, E. D. Carrazzoni, Phytochemistry, 6, 1652 (1967). 117. A. Ahond, A. Cave, C. Kan-Fan, Y. Langlois, P. Potier, Chem. Comm., 157 (1970) , 118. P. Potier, M. M. Janot, C. R. Acad. Sc. Paris, 276, 1927 (1973). 119. G. B. Fuller, Studies on the Synthesis and Biosynthesis of Indole Alkaloids, PhD. Thesis, The University of British Columbia (1974). - 262 -120. This work i s presently being conducted in our laboratory by H. W. Hanssen whose collaboration i s gratefully acknowledged. 121. M. A. Ondetti, V. Deulofeu, Tetrahedron Letters, (7), 1 (1959). 122. M. A. Ondetti, V. Deulofeu, Tetrahedron Letters, (1), 18 (1960). 123. M. A. Ondetti, V. Deulofeu', Tetrahedron, 15_, 160 (1961). 124. Kindly provided by Professor B. Gilbert, Research Center for National Products, Faculty of Pharmacy, Rio de Janeiro, Br a z i l . 125. G. B. Marini-Bettolo, J. Schmutz, Helv. Chim. Acta., 42_, 2146 (1959). 126. Physical Data of Indole and Dihydroindole Alkaloids, 1_§ 2, E l i L i l l y and Company (1960). 127. R. E. Woodson Jr., Missouri Botanical Garden Annals, 38, 119 (1951). 128. R. H. Burnell, D. Delia Casa, Can. J. Chem., 45, 89 (1967). 129. J . A. Joule, M. Osashi, B. Gilbert, C. Djerassi, Tetrahedron, 21, 1717 (1965). 130. This work i s presently being conducted i n our laboratory by N. G. Lewis, whose collaboration i s gratefully acknowledged. 131. Kindly provided by Dr. D. Henry, Stanford Research Institute, Menlo Park, California. 132. J . Schmutz, F. Hunziker, Helv. Chim. Acta., 41, 288 (1958). 133a. Mark I Liquid S c i n t i l l a t i o n Systems Instruction Manual, Section 1, Nuclear Chicago Corporation, Des Plaines, I l l i n o i s (1966). b. R. W. Hendler, Ana. Biochem.. ]_, 110 (1964). 134. 135. G. Buchi, E. W. Wamhoff, J. Am. Chem. Soc, 81, 4433 (1959). 136. S. Goodwin, A. F. Smith, B. C. Homing, J . Am. Chem. Soc, 81,1903 (1959). - 263 -137. G. Buchi, D. W. Mayo, F. A. Hochstein, Tetrahedron, 15, 167 (1961). 138. M. G. Ahmed, R. W. Alder, G. H. James. M. L. Sinnolt, M. C. Whiting, Chem. Comm., 1533 (1968). 139. A. C. Cope, E. R. Trumbull, Org. Reactions. 140. J . A. Joule, C. Djerassi, J . Chem. Soc, 2777 (1964). 141. M. Pailer, L. Bilek, Monatsh., 79, 135 (1948). 142. A. R. Battersby, H. T. Openshaw, J. Chem. Soc, S59 (1949). 143. G. Norcross, H. T. Openshaw, J. Chem. Soc, 1174 (1949). 144. A. R. Battersby, T. P. Edwards, J. Chem. Soc, 1214 (1960). 145. D. G. 0'Donovan, H. Horan, J. Chem. Soc. (C) 2791 (1968), 146. A. R. Battersby, B. J. T. Harper, J . Chem. Soc, 3526 (1962) • 147. T. A. Geissman, A. Armen, J. Am. Chem. Soc, 74, 3916 (1952) 148. E. D. Hughes, C. K. Ingold et a l . , J . Chem. Soc, 899 (1940). 149. H. Budzikiewicz, C. Djerassi, D. H. Williams, Mass Spectrometry of Organic Compounds, Holden-Day (1967). 150. N. D. V. Wilson, J . A. Joule, Tetrahedron, 24, 5493 (1968), 151. A. C. Cope, N. A. LeBel, J. Am. Chem. Soc, 82, 4656 (1960), 152. W. A. Szarek, S. Wolfe, J. K. N. Jones, Tetrahedron Letters, (38), 2743 (1964). 153. H. S. Gutowsky, C. H. Holm, J . Chem. Phys., 25, 1288 (1956). 154. M. T. Rogers, J . C. Woodbrey, J. Phys. Chem., 66, 540 (1962) • 155. 0. Wintersteiner, M. Moore, J. Am. Chem. Soc, 65, 1503, 1507 (1943). 156. L. Syper, Tetrahedron Letters, (42), 4198 (1967). 157. L. Syper, Tetrahedron Letters, (37), 4493 (1966). - 264 -158. T. L. Ho, Synthesis, 347 (1973). 159. M. Wei, R. Stewart, J . Am. Chem. Soc, 88, 1974 (1966). 160. S. W. Pelletier, W. A. Jacobs, J. Am. Chem. Soc., 76_, 4496 (1954), and references therein. 161. J . B. Lee, C. Parkin, M. J . Shaw, N. A. Hampson, K. I. MacDonald, Tetrahedron, 29_, 751 (1973). 162. J. B. Lee, T. G. Clarke, Tetrahedron Letters, (5), 415 (1967), 163. V. J. Traynelis, R. H. Ode, J . Org. Chem., 35, 2207 (1970). 164. N. J . Leonard, F. P. Hauck, J. Am. Chem. Soc, 79_, 5279 (1957). 165. A. H. Riebel, R. E. Erickson, C. J . Abshire, P. S. Bailey, J . Am. Chem. Soc, 82, 1801 (1960). 166. S. S. Rawalay, H. Shechter, J . Org. Chem., _32, 3129 (1967), 167. R. N. Gupta, I. D. Spenser, Can. J . Chem., 40, 2041 (1962). 168. R. N. Gupta, I. D. Spenser, Can. J. Chem., 40, 2049 (1962), 169. W. J . Gensler, E. M. Healy, I. Anshuus, A. L. Bluhm, J. Am. Chem. Soc, 78, 1713 (1956). 170. R. C. Fuson, B. A. Bull, Chem. Rev., 15, 275 (1934). 171. M. H. Hashmi, Pakistan J. Science, 10, 159 (1958). 172. von J. Schmutz, H. Wittwer, Helv. Chim. Acta., 43, 793 (1960), 173. E. Wenkert, K. G. Dave, J . Am. Chem. Soc, 84, 94 (1962). 174. R. B. Woodward, G. A. Jacobucci, F. A. Hochstein, J. Am. Chem. Soc, 81_, 4434 (1959). 175. P. A. Cranwell, J. E. Saxton, J. Chem. Soc, 3482 (1962). 176. T. K. Govindachari, S. Rajappa, V. Sudarsanam, Ind. J. Chem., 1_, 47 (1963). 177. C. W. Mosher, 0. P. Crews, E. M. Acton, L. Goodman, J. Med. Chem., 9_, 237 (1966). - 265 -178. L. K. Dalton, S. Demerac, B. C. Elmes, J . W. Leder, J . M. Swan, T. T e i t e i , Aust. J . Chem., 20, 2715 (1965). 179. K. N. Kilminster, M. Sainsbury, J. Chem. Soc. Perkin 1, 2264 (1972). 180. F. LeGoffic, A. Gouyette, A. Ahond, Tetrahedron, 29, 3357 (1973). 181. E. Noelting, Ber. deutsch. Chem. Ges., 37, 1015 (1904). 182. E. Wenkert, B. F. Barnett, J . Am. Chem. Soc, 82, 1671 (1960). 183. T. Kametani, K. Ogasawara, T. Takahashi, Tetrahedron, 29, 73 (1973). 184. T. Kametani, Y. Ichikawa, T. Suzuki, K. Fukumoto, Tetrahedron, 30, 3713 (1974). 185. T. Kametani, T. Suzuki, K. Takahashi, K. Fukwmoto, Tetrahedron, 30, 2207 (1974). 186. E. Leete, L. Marion, Can. J. Chem., 31, 775 (1953). 187. J . M. Bobbit, J . Kiely, K. L. Khanna, R. Ebermann, J . Org. Chem. 30, 2247 (1965). 188. J. P. Kutney, D. S. Grierson, Heterocycles, 3, 171 (1975). 189. M. E. Speeter, W. C. Anthony, J. Am. Chem. Soc, 76 , 6209 (1954). 190. R. B. Woodward, F. E. Bader, H. Bickel, A. J . Frey, K. W. Kierstead, Tetrahedron, 2, 1 (1958). 191. M. J u l i a , H. Ingolen, J. Lenzl, Bull . Soc. Chim. France, 2291 (1966) 192. E. Wenkert, R. A. Massey-Westropp, R. G. Lewis, J . Am. Chem. Soc, 84 3732 (1962). 193. H. E. Zaugg, D. A. Dunnigan, R. J . Michaelis, L. R. Sevitt, T. S. Wang, A. H. Soramers, R. W. DeNet, J . Org. Chem., 26, 644 (1960). 194. W. D. Closson, S. A. Roman, G. T. Kwistkowski, D. A. Corwin; Tetrahedron Letters, (21), 2271, (1966). 195. C. I. Brodrick, W. F. Short; J . Chem. Soc, 2587 (1949). - 266 -196. L. M. Jackman; Hydrogenation-Dehydrogenation Reactions, Adv. in Org. Chem., 2_, 329 Interscience (1962). 197. B. M. Barclay, N. Campbell; J . Chem. Soc, 530 (1945). 198. R. Ratcliffe, R. Rodehorst; J. Org. Chem., 35_, 4000 (1970). 199. W. S. Trahanovsky, L. B. Young, G. L. Brown; J. Org. Chem., 32, 3865 (1967). 200. K. E. Stensio; Acta. Chimica.Scand, 25, 1125 (1971). 201. J. R. Parikh, W. von E. Doering; J. Am. Chem. Soc, 89, 5507 (1967). 202. R. E. Partch; Tetrahedron Letters,(41), 3071 (1964). 203. Fieser § Fieser, Reagents in Org. Synthesis, 1_, 142 J . Wiley § Sons (1967). 204. P. Bruck, Chem. Comm., 1690 (1970). 205. N. F. Kucherova, V. P. Eudakov, N. K. Kochetkov, Zhur. Obschchei Khim., 27, 1049 (1957). 206. G. F. Smith, Heterocyclic Chemistry, Van Nostrand Reinhold (1972). 207. C. U. Corson, C. U. Rogers, Org. Syn. Col l . Vol., 4, 884 (1972). 208. Private communication with Dr. Y. Murakami, our laboratory, whose contribution i s gratefully acknowledged. 209. B. A. Whittle, E. N. P. Young, J. Med. Chem., 6, 378 (1963). 210. E. C. Horning, M. G. Horning, J. Org. Chem., 11, 95 (1946). GRADUATE STUDIES CONT'D Seminar in Special Topic (Natural Products) J.P. Kutney Seminar i n Chemistry D.G. Clark PUBLICATIONS I.H. Rogers, D. Grierson Can. Dep. Fish. For., Bi-Mo. Res. Notes, 25, (4), 33 (1969). Extractives from the bark of Grand F i r [Abies grandis (Dougl.) Lindl] I.H. Rogers, D. Grierson Wood & Fiber, 4, (1), 33 (1972) Extractives from Grand F i r [Abies grandis (Dougl.) Lindl] Bark J.P. Kutney, D;S. Grierson, G.D. Knowles and N.D, Westcott Tetrahedron, 29, 13 (1973) Studies on Constituents of Abies grandis. The Structure and Absolute Stereochemistry of Cyclograndisolide and Epicyclograndi-solide, Novel Cyclopropane Triterpene Lactones J.P. Kutney, D.S. Grierson Heterocycles, 3 (2), 171 (1975) An Improved Synthesis of the Olivacine type Indole Alkaloids 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061014/manifest

Comment

Related Items