15 Alkaloids By H. F. HODSON The Wellcome Research Laboratories Beckenham Kent BR3 3BS 1 Introduction The excellent comprehensive coverage provided by Volume 1 of the Specialist Periodical Reports on Alkaloids has been maintained in Volume 2,2 which covers the period July 1970 to June 1971 and deals with the whole alkaloid field with the exception of the steroidal bases of the Solanurn and Veratrurn groups. The first volume of the Specialist Periodical Reports on Biosynthesis3 has also been published the chapter on alkaloid biosynthesis includes a tabular survey of all incorporations reported during 1971. Two papers4,’ describe in tabular form the screening for alkaloid content of plant extracts which had previously been tested for antitumour activity and found to be inactive.The 2000 extracts examined to date represent 1688 species of diverse geographical origin ;490 gave positive alkaloid tests and 288 of these had not previously been reported to contain alkaloids. As more species are examined and as separation techniques and structural elucidations become more refined there have been an increasing number of reports in which alkaloids or alkaloid types previously thought to be unique to one particular genus or family have been isolated from a completely unrelated plant species ; this trend has continued during the year under review. Repeating the pattern noted last year most newly isolated alkaloids are either examples of known types or have structures which can readily be accommodated within biogenetic schemes which are well established or currently accepted.A considerable amount of synthetic work has appeared but unlike that discussed last year6 most of the notable synthetic achievements involve ingenious applica- tions of known reactions rather than new ones. It is striking that so many of the syntheses reported this year have been motivated (at least in part) by the actual or potential therapeutic value of the target compound ;thus there have been four ‘The Alkaloids’ ed. J. E. Saxton (Specialist Periodical Reports) The Chemical Society London 197 1 vol. 1. ’ ‘The Alkaloids’ ed. J. E. Saxton (Specialist Periodical Reports) The Chemical Society London 1972 vol. 2. ‘Biosynthesis’ ed. T. A. Geissman (Specialist Periodical Reports) The Chemica Society, London 1972 vol.1. S. J. Smolenski H. Silinis and N. R. Farnsworth Lloydia 1972 35 1. H. H. S. Fong M. Trojankova J. Trojanek and N. R. Farnsworth Lloydia 972 35 117. ‘ H. F. Hodson Ann. Reporrs (B) 1971 68 493. 487 488 H. F. Hodson syntheses of camptothecin two of cephalotaxine and two of ellipticine all alkaloids with antitumour activity. 2 Pyridine and Pyrrolizidine Alkaloids In a biomimetic synthesis,' (f)-nicotine has been prepared (Scheme 1) in a yield of 7 % from an aqueous solution of glutaraldehyde ammonia and 1-methyl-A'- pyrrolinium acetate in aqueous solution at pH 10 in the presence of air. Reagents i NH,; ii ;iii [O] Me Scheme I Swazine (l),a new alkaloid from Senecio swaziensis is unique in having an epoxide ring in the diacid moiety.Acidic hydrolysis of (1) gave the base retro- necine together with the novel spirodilactone (2),whose structure was deduced' by X-ray crystallography of the p-bromobenzoate. The mode of attachment of the diacid is the reverse of that at first suggested and was establishedg by an X-ray study of the methiodide of (1). It is now well established that the dramatic hepatotoxicity of the toxic pyrrolizi- dine alkaloids depends on their metabolic conversion into pyrrole esters of part structure (3) these being potent alkylating agents. A full account of the recent extensive chemical and biochemical work in this area has been published." ' E. Leete J.C.S. Chem. Comm. 1972 1091. * C. G. Gordon-Gray R. B. Wells N. Hallak M.B. Hursthouse S. Neidle and T. P. Toube Tetrahedron Letters 1972 707. ' M. Laing and P. Sommerville Tetrahedron Letters 1972. 5183. A. R. Mattocks in 'Phytochemical Ecology' Academic Press London and New York 1972 p. 179. Alkaloids 489 3 Quinoline Alkaloids A new and convenient general procedure' 'for the preparation of furanoquinoline alkaloids of the dictamnine type is illustrated by the synthesis (Scheme 2) of dictamnine (4; R' = R2 = H) itself. R'@ / OMe i,ii \ ~ Rlw R2 R2 (6)R' = R2 = Me0 liii Reagents i BuLi then BrCH,CH =CMe,; ii HCl; iii 0 or Os0,-10,-; iv PPA Scheme2 In two plant species labelled dictamnine [cf (4)]prepared by the method of Scheme 2 was efficiently incorporated (1.7-2.1 % and 1.2-3.5 %) into the dimethoxy-alkaloid skimmianine (4; R1= R2 = OMe).' ' Since platydesmine (5 ;R = H) is known to be a highly efficient precursor of dictamnine this suggests the pathway platydesmine (5; R = H)+ dictamnine (4; R' = R2 = H)-+ skimmianine (4;R' = R2 = OMe) with aromatic hydroxylation and methyla- tion as the last steps.However a newly isolated alkaloid from Dictamnus albus has been proved by synthesis,I2 to have structure (6),suggesting itself as a likely biogenetic precursor via (5; R = OMe) of skimmianine which has long been J. F. Collins W. J. Donnelly M. F. Grundon D. M. Harrison and C. G. Spyropoulos J.C.S. Chem. Comm. 1972 1029. I' R. Storer and D. W. Young Tetrahedron Letters 1972 2199. 490 H. F. Hodson known as a constituent of this plant ;this new base has been named preskimmia- nine.If (6) is indeed a precursor of skimmianine then different pathways may be followed in different plants or possibly even in the same plant. However pre- skimmianine is not proved to be a precursor of skimmianine and it should be noted that in several recent cases detailed investigations have shown that the ‘obvious’ chemical pathway is not the one followed in the plant. 4 Isoquinoline Alkaloids A reviewI3 provides an account of developments in this field over the past five years. Imeluteine (7; R = OMe) and rufescine (7; R = H) from Abuta spp. (Meni- spermaceae) are alkaloids of a new structural type. l4 The azafluoranthene skeleton of (7) must surely be derived from 1-phenylisoquinoline precursors ; alkaloids of this latter type are rare and have only recently been recognized although not in the Menispermaceae.The structures (7) were suggested by spectroscopic data and confirmed by unambiguous syntheses which incorporated a Pschorr cyclization of the appropriate 1-phenyldihydroisoquinolinediazonium salts (8). OMe OMe 6 OMe (7) Two new alkaloids are minor but intriguing variants of aporphine bases. The first of these variabiline (lo),” a unique example of an amino-substituted benzylisoquinoline-derived alkaloid was isolated from Ocotea uariabilis where it occurs together with the aporphine (+)-apoglaziovine (11) and its precursor (+)-glaziovine (9). Variabiline was synthesized by heating (9) with a mixture of dibenzylamine and dibenzylamine hydrochloride; it appears not to be an artefact ; dibenzylamine could not be detected in the crude extract and does not react with glaziovine under the conditions of the extraction.The other aporphine modifica- tion from Thalictrum pofygumum is the quaternary alkaloid thalphenine (12),16 with a biogenetically interesting methylenedioxy bridge which figures pro- minently in the n.m.r. spectrum; the structure and absolute configuration were established by X-ray analysis of the iodide. The methine base derived from l3 T. R. Govindachari and N. Viswanathan J. Sci.Ind. Res. India 1972 31 244. l4 M. P. Cava K. T. Buck and A. I. da Rocha J. Amer. Chem. Soc. 1972.94 5931. l5 M. P. Cava M. Behforouz and M. J. Mitchell Tetrahedron Letters 1972 4647.’‘ M.Shamma J. L. Moniot S. Y. Yao and J. A. Stanko J.C.S. Chem. Comm. 1972,408. Alkaloids 491 Me0 ,-’ Me0 ,-’ (10) RR\ H°FMe = (PhCH,),N( k) HFMe 0 (9) (11) R = OH I (12) thalphenine was also present in T. polygamum and was recently isolated” from T. rugosum. The benzylisoquinoline alkaloid reticuline has been converted into ( & )-cepharamine (15) one of the simpler alkaloids with the hasubanan skeleton in a sequence’ * which commenced with the ring-opening [by (F3€C0)20] of 2’-bromoreticuline to give a stilbene which was reduced to (13; X = Br). Photo- lytic dehydrobromination’ (path a) produced the dienone (14) and hydrolytic removal of the N-acyl group in (14) was followed by internal Michael addition to give the hasubanan skeleton ;the product an isomer of (+)-cepharamine was converted into (1 5) by acid-catalysed transetherification.In an earlier approach2’ ” N. M. Mollov L. N. Thuan and P. P. Panov Compt. rend. Acad. bulg. Sci.,1971 24 1047. T. Kametani H. Nemoto T. Kobari K. Shishido and K. Fukumoto Chem. andInd. 1971 538. l9 T. Kametani and K. Fukumoto Accounts Chem. Res. 1972. 5 212. 2o T. Kametani T. Kobari and K. Fukumoto J.C.S. Chem. Comm. 1972 288. 492 H. F. Hodson the dihydrostilbene (13; X = H) from reticuline was subjected to phenolic oxidation ;ppcoupling (path b)gave an intermediate with the ‘wrong’ oxygena- tion pattern. Most synthetic routes to the spirobenzylisoquinoline system have employed a Pictet-Spengler condensation.In a new approach a synthesis of (i-)-ochro-birine (17)2’ makes use of the Bobbitt modification of the Pomeranz-Fritsch reaction to convert (16) into the corresponding dihydroisoquinoline. EtO OEt Several benzylisoquinoline-aporphine ‘dimer’ alkaloids are now known and presumably arise from bisbenzylisoquinoline precursors by way of benzyliso- quinoline-proaporphine bases. Pakistanamine ( 18)22 from Berberis baluchi-stanica now provides the first example of a benzylisoquinoline-proaporphine alkaloid ; it underwent acid-catalysed dienone-phenol rearrangement to give a benzylisoquinoline-aporphine base differing only in its degree of methylation from a new alkaloid pakistanine isolated from the same source. ’0-cz Another interesting bisbenzylisoquinoline-derived structure is provided by stepinone (19)from Stephaniajaponica.Structure (19) was deduced23 by standard methods which included cleavage with sodium in liquid ammonia of NO-dimethyltetrahydrostepinone to give (S)-(-)-armepavine [A portion cf (19)] N. E. Cundasawmy and D. B. MacLean Canad. J. Chem. 1972,50 3028. 22 M. Shamma J. L. Moniot S. Y. Yao. G. A. Miana and M. Ikram J. Amer. Chem. SOC. 1972,94 1382. 23 T. Ibuka T. Konoshima and Y. Inubushi Tetrahedron Letters 1972 4001. Alkaloids 493 and a new phenolic base. The 0-ethyl derivative of this new base was identical with the phenylbenzazepine (20) synthesized by an unambiguous route,24 thus establishing the B portion [cf. (19)] of stepinone itself. Deuterium exchange prior to cleavage followed by location (n.m.r.) of deuterium in the cleavage product^,^' helped to define the C-terminal positions of the head-to-head ether linkage.This method was also used by other workers in the structure determina- tion of a new bisbenzylisoquinoline alkaloid neumarine.26 The past few years have seen the isolation from CephaEotaxus harringtonia of a unique series of bases all with the same skeleton exemplified by the principal alkaloid cephalotaxine (21). Interest in these alkaloids has been heightened by the antitumour activity of the esters e.g.harringtonine (22),27 and two syntheses of ( &)-cephalotaxine have now been reported. One synthesis” proceeded via the tricyclic enamine (23) the elements of the fourth ring being introduced by acylation to give (24),which was converted into the wdicarbonyl compound (25).Cyclization of (25)with magnesium methoxide gave ( +)-demethylcephalotaxinone (26) which has recently been isolated29 from C. harringtonia; methylation and reduction of (26)gave (+)-cephalotaxine (21). The enamine (23) was prepared independently in connection with another 24 Y. Inubushi T. Harayama and K. Takeshima Chem. and Pharm. Bull. (Japan) 1972 20 689. 25 Y. Inubushi T. Kikuchi T. Ibuka and I. Saji Tetrahedron Letters 1972,423. 26 I. R. C. Bick H. M. Leow and N. W. Preston J.C.S. Chem. Comm. 1972,980. 27 K. L. Mikolajczak R. G. Powell and C. R. Smith Tetrahedron 1972 28 1995 and references there cited. 28 J. Auerbach and S. M. Weinreb J. Amer.Chem. Soc. 1972,94 7172. 29 Ref. 28 footnote I I. 494 H. F. Hodson (9 0 RO R (21) R = H OMe (23) R = H (24) R = COCH(0Ac)Me(25) R = CO-COMe OHI OHI I(22) R = Me2CCH2CH2.CCH2C02Me co- projected ~ynthesis.~' The second synthesis3 I involved the assembly of the spiro-compound (27). Generation of the benzyne carbanion (2 equivalents of potassium triphenylmethide) was followed by internal nucleophilic addition to give (+)-cephalotaxinone (28) in yields of 13-16 %; the synthesis was completed by reduction to (-t)-(21). 0 0 OMe OMe (27) X = C1 Br or I (28) The Cephalotaxus bases are tantalizingly similar to the Erythrina alkaloids and a plausible biosynthetic origin from an Erythrina precursor has been sug-ge~ted.~~ of C.harringtonia has shown however that Further in~estigation~~ in this species the typical Cephabtaxus alkaloids are accompanied by small amounts of five Homoerythrina alkaloids closely related to schelhammericine 30 L. J. Dolby S. J. Nelson and D. Senkovich J. Urg. Chem. 1972,37 3691. 31 M. F. Semmelhack B. P. Chong and L. D. Jones J. Amer. Chem. SOC.,1972,94,8629. 32 V. Snieckus in Ref. 2. 33 R. G. Powell Phytochemistry 1972 11 1467. Alkaloids 495 (29);also a hitherto unexamined species C. wils~niana,~~ has furnished cephalo- taxine together with two Homoerythrina alkaloids of the 6,7-epoxy type [cf.(29)] previously encountered only in the completely unrelated plant Phelline comma. These observations suggest3 an equally plausible biogenetic origin from a Homoerythrina precursor such as (30),via a tetracyclic intermediate such as (31).5 Amaryllidaceae Alkaloids A synthesis35 ofanhydrolycorine(33) from (32)in 67% yield provides yet another” example of the application in alkaloid synthesis of the formation of a biphenyl linkage by photochemical dehydrobromination. (33) The lactam alkaloid narciclasine of widespread occurrence among plants of the Amaryllidaceae has been the subject of a number of recent biosynthetic studies which suggest that it is derived from a precursor of the crinine type [cf. (35)] by elimination of the two-carbon bridge. An X-ray of the tetra-acetate has now firmly established the structure (34)for narciclasine which differs from the previously revised37 structure only in the configuration at C-2; further degradative and n.m.r.studies38 are also in accord with the formulation (34). Feeding experiment^^^ demonstrate that vittatine (39 of known stereo- chemistry is incorporated (0.8%) into narciclasine thus supporting the depicted absolute stereochemistry (34). 34 R. G. Powell K. L. Mikolajczak D. Weisleder and C. R. Smith Phytochemistry 1972 11 3317. 35 H. Hara 0.Hoshino and B. Umezawa Tetrahedron Letters 1972 5031. 36 A. Immirzi and C. Fuganti J.C.S. Chem. Comm. 1972,240. ’’ H. F. Hodson Ann. Reports (B),1970 67,476. 38 A. Mondon and K. Krohn Tetrahedron Letters 1972 2085. 39 C. Fuganti and M. Mozza J.C.S. Chem. Comm. 1972 239. 496 H. F. Hohon -OH OH 0 (35) (34) 6 Indole Alkaloids Paraensine (36)from Euxylophora paraensis (Rutaceae) is an addition to the small group of quinazolinocarboline alkaloids and the first one to possess an isoprenoid moiety.40 It is interesting to note that isoprenoid units abound in other alkaloids (furanoquinolines acridones) from the Rutaceae.Me Me (36) Glycosides and Covynanfk-Sfvychnos Alkaloids.-Vincoside (37 ; R = H) derived from tryptamine and secologanin is now unequivocally established as the key early intermediate in the biosynthesis of the vast array of indole mono- terpenoid alkaloids.41 Last year a 5-carboxyisovincoside (38 ; R = CO,H) (37) 38-H (38) 3a-H (39) 3B-H (40) 3a-H 40 B. Danieli P. Manitto F. Ronchetti G. Russo and G. Ferrari Experientia 1972 28 249.41 See for example A. R. Battersby in ref. 1 p. 31. Alkaloids 497 derived from tryptophan and secologanin was isolated for the first time. This has now been followed by the isolation42 from Adina rubescens in a work-up which involved acetylation and methylation of the two epimeric lactams (39) and (40),derivatives of 5-carboxyvincoside (37 ;R = C0,H) and 5-carboxyiso- vincoside (38 ;R = C0,H) respectively. Structures (39) and (40)were confirmed by synthesis from L-tryptophan and secologanin and the depicted stereo-chemistry was firmly e~tablished.~~ 5-Carboxyvincoside (37 ; R = C02H) could be an end-product off the main biosynthetic pathway or it could serve as well as vincoside as a precursor of some monoterpenoid indole alkaloids.More interestingly it might function as the precursor of a whole range of more highly evolved carboxy-substituted alkaloids. This possibility stimulated a deliberate search for such alkaloids which has been rewarded by the discovery43 in Adina rubescens of adirubine shown to have the tryptophan- based Corynanthi-type structure (41). Et The biogenetically interesting structure (42) earlier (without stereochemistry) for roxburghine D has been confirmed by a biosynthetically modelled via the acid (43; R = OH) and its tryptamide (43;R = p-Ind.-CH,CH,-NH); detailed n.m.r. support the stereochemistry depicted in (42). Me MeO,CH CORD (42) (43) 42 W. P. Blackstock R.T. Brown C. L. Chapple and S. B. Fraser J.C.S. Chem. Comm.1972 1006. 43 R. T. Brown C. L. Chapple and G. K. Lee J.C.S. Chem. Comm. 1972 1007. 44 Ref. 37 p. 479. 4s H. Riesner and E. Winterfeldt J.C.S. Chem. Comm. 1972 786. 46 C. Cistaro L. Merlini R. Mondelli and G. Nasini J.C.S. Chem. Comm. 1972 785. 498 H. F. Hodson In suaveoline (45;R = H)from Rauwo@a suaveofens the monoterpenoid moiety is incorporated into a pyridine ring." The structure was confirmed by synthesis4' of Nb-methylsuaveoline (45;R = Me) via the compound (44) obtained from ajmaline. Many (but not all) of the simple monoterpenoid pyridine bases are artefacts arising from seco-iridoid precursors by the action of ammonia used in the work-up. Suaveoline was obtained in an ammonia-free extraction process and is not therefore an artefact.H (44) (45) A new variation of the Nb-C-21secosarpagine [cf. (44)]carbon skeleton seen in suaveoline is provided by talcarpine (47)from Pleiocarpa talbotii :48 talpinine (46)from the same species was methylated with ring-opening to give the C-20 epimer of talcarpine :mild base treatment of talcarpine (47)gave this C-20epimer exclusively. H HI H (46) (47) Plants of the genus Alstonia have furnished a number of bisindole alkaloids all with macroline (48)as one half being linked via the a-methyleneketone function to either the aromatic or alicyclic portion of the other 'half' molecule. Now three of these alkaloids have been synthe~ized~~ in biomimetic reactions which begin with the acid-catalysed Michael addition between (48)and an appropriate nucleophilic centre in the other 'half '.Thus quebrachidine [Aportion of (50)]and (48)gave an adduct which was characterized as a labile hemiacetal with part structure (49)and which could be cyclized (BF,,Et,O) to give the alkaloid alstonisidine which is now therefore formulated as (50);490 this differs from the structure advanced last year5' only in the reversal of the A-B linkages. 47 S. P. Majumdar P. Potier and J. Poisson Tetrahedron Letters 1972 1563. 48 J. Narango M. Pinar M. Hesse and H. Schmid Helv. Chim. .4cta 1972 55 752. 49 (a) D. E. Burke J. M. Cook and P. W. Le Quesne J.C.S. Chern. Comm. 1972 697; (6) D. E. Burke and P. W. Le Quesne ibid. p. 678; (c) D. E. Burke C. A. De Markey P. W. Le Quesne and J.M. Cook ihid. p. 1346. 50 Ref. 6 p. 507. Alkaloids 499 H CH,OH -'N' I (50) Similar syntheses of two other bisindole alkaloids ~illalstonine~~~ and macral~tonine,~~~ involved Michael additions of (48)to an indole fi-position and an activated aromatic position respectively. An elegant new synthesis of (f)-yohimbine5' centred on the formation (Scheme 3) of the decahydroisoquinoline (52) with the full D,E-ring functionality and stereochemistry of the alkaloid itself. Application of a new enamine annelation reaction to give (51) was followed by successive stereospecific (trans ring junction) and stereoselective (mainly axial alcohol) reduction steps finally the N-methyl group was removed by von Braun cleavage followed by reductive decyanation to give (52),which led to yohimbine by standard operations.Me Me H -(i--HB + Me02C$ -P 0 Me0,C Me0,C" 0 OH (51 1 (52) Reagents i PhH -MeOH reflux ii Li-NH ,-Et,O-Bu'OH ; iii Pt-H ; iv BrCN and chromatog.; v Zn-82 AcOH Scheme 3 Several natural oxindole alkaloids have now been converted into their indole counterparts a transformation not hitherto realized although the reverse process G. Stork and P. N. Guthikonda J. Amer. Chem. SOC.,1972,94 5109. 500 H. F. Hodson is well documented. The sequence is illustrateds2 for pteropodine (53) ;O-alkyla-tion to the imido-ester (mixture of C-7 epimers) was followed by reduction (NaBH in acetic acid) to the indole (54). Oxidative cyclization of this 2,3- seco-alkaloid (54) with mercuric acetate in the usual manner gave a mixture of tetrahydroalstonine (55) and akuammigine (56).An alternative cyclizationS3 employed a modified Polonovski reaction; (54) was converted into the N-oxide (epimeric mixture) which on treatment with trifluoroacetic anhydride furnished akuammigine (56) unaccompanied by its 3-epimer. A similar ferrous-ion- catalysed cyclization of model N-oxides has also been reported.54 Me02Cbo (53) (54) ,Me (55) 3a-H (56)38-H The ester (57),prepared in connection with the recent extensive synthetic efforts in the quinine series has been utilized in a synthesis (Scheme 4) of (57) 0 Reagents i THF 90 "C; ii SiO sepn. of epimers; iii NaNH,; iv MeMgI-fi Scheme 4 52 N. Aimi E.Yamanaka J. Endo S. Sakai and J. Haginiwa Tetrahedron Letters 1972 1081. 53 H.-P. Husson L. Chevolot Y. Langlois C. Thal and P. Potier J.C.S. Chem. Comm. 1972,930. 54 C. A. Scherer C. A. Dorschel J. M. Cook and P. W. Le Quesne J. Org. Chem. 1972 37 1083. Alkaloids 50 1 ( & )-dihydrocinchonamine,' incorporating a Madelung synthesis in the indole- forming stage. Aspidosperma Types.-Rhazinilam is an interesting neutral compound first isolated from Rhazya stricta and also from Melodinus australis and Aspidosperrna quebracho-blanco; it gradually accumulates in basic fractions of R. stricta presumably arising from an alkaloid precursor. It is converted by acid into a mixture of closely related bases with the chromophore of a 3,4-dialkylpyrrolo- quinoline (58),and a consideration of the probable nature of this change together with spectroscopic studies provided convincing evidence in favour of structure (59) for rhazinilam;56 this was confirmed in an independent X-ray A plausible derivation from a hypothetical Aspidosperma precursor such as (60)is ~uggested.~~ I.l-i Iboga Types.-The absolute stereochemistry of catharanthine (61) was un-equivocally established some ten years ago by anomalous dispersion studies and although the opposite configuration had earlier been suggested for ibogaine this assignment has not generally been accepted ; the chemical correlations within this group have not included optical data. It has now been shown5* that the c.d. spectra of a number of Zboga alkaloids including coronaridine (62) voacangine ibogamine and tabernanthine are similar but with Cotton effects of reverse sign to those of catharanthine (61) and the absolute configuration of these alkaloids must therefore be enantiomeric with that of catharanthine.Skeletal Interconversions-In 1968 Scott and Qureshi reported that in refluxing acetic acid (+)-tabersonine (66) rearranged to a mixture of (&)-catharanthine (70) (12%) and pseudocatharanthine (68) (28%) and that (+)-stemmadenine 55 G. Grethe H. L. Lee and M. R. Uskokovic Synthetic Comm. 1972 2 55. " K. T. De Silva A. H. Ratcliffe G. F. Smith and G. N. Smith Tetrahedron Letters 1972,913. 57 D. J. Abraham and R. D. Rosenstein Tetrahedron Letters 1972 909. '* K. Blaka Z. Koblicova and J.Trojanek Tetrahedron Letters 1972 2763. 502 H. F. Hodson (63) under these conditions gave 12% of (66) 9% of (70) and 16% of (68).59 These rearrangements which were postulated to proceed via the achiral dihydro- pyridine acrylic ester (71) appeared to establish an in vitro link between alkaloids of the Covynanthd-Strychnos [e.g. (63)] Aspidospema [e.g. (66)] and Zboga [e.g. (70)]skeletal types and provided a welcome analogy for the biosynthetic skeletal transformations of these alkaloids. (63) R = H (66) (64) R = Ac 19,20-dihydro (67) 14,15-dihydro (65) R = AC C!O,Me C0,Me // (68) (69) 15,20-dihydro Other workers though were unable to effect these rearrangements under the conditions described and full details of their efforts have now been published.60,61 A further chapter in the story is provided by a series of papers62 from Scott and Wei who show that the skeletal rearrangements can indeed be detected in vitro and can plausibly take place via the intermediate (71) and double-bond isomers thereof.However the conditions under which they have now been realized and the yields obtained bear little resemblance to those described in the 1968 paper and it is therefore not surprising63 that this earlier work could not be reproduced s9 J. A. Joule in Ann. Reports (B) 1969 p. 483 and in ref. 1 p. 193. R. T. Brown J. S. Hill G. F. Smith and K. S. J. Stapleford Tetrahedron 1971,27,5217. 61 M. Muquet N. Kunesch and J. Poisson Tetrahedron 1972 28 1363.A. I. Scott and C. C. Wei J. Amer. Chem. SOC.,1972,94 8263,8264 8266. 63 Cf.A. I. Scott J. Amer. Chem. SOC.,1972.94 8262. Alkaloids 503 by others; it is also now clear that aerial oxidation and oxidation-reduction disproportionation processes are intimately involved. Thus for example dihydrostemmadenine 0-acetate (64) adsorbed on to silica gel and heated to 150 "C in air for 45 minutes gave (+)-pseudocatharanthine (68) (1 %) and dihydropseudocatharanthine (69) (0.5%) ; a higher yield (2.5 %) of (68) was obtained through the intermediacy of dihydropreakuammicine acetate (721 prepared from (64) by a regiospecific catalytic oxygenation. Exemplifying the transformation to an Aspidosperrna skeleton thermolysis on silica gel of the hydrochloride of (65) gave ( +)-vincadifformine (67) (0.2 %).Alternatively stemmadenine acetate (65) was converted regiospecifically by platinum-catalysed oxidation into (73) which was reduced to dihydroprecondylocarpine acetate (74);thermolysis of (74) on silica gel gave 0.2 % each of (+)-vincadifformine (67) and (f)-tabersonine (64). Me(02C' 'CH20Ac MeO,C' 'CH OAc (73) (74) 19,20-dihydro 7 Quinoline Alkaloids Biogenetically Derived from Indoles Previously obtained only from Carnptothecaacurninata (0.005 % yield) campto- thecin (79) has now been isolated in yields approaching 0.1% from Mappia foetida (Nyssaceae) where it is accompanied by the new base 9-methoxycampto- thecin [cJ (79)].64 Underlining the keen interest in this alkaloid the two syntheses of (*)-camptothecin reported last year have been followed by no fewer than four all six together providing a superb illustration of the diversity possible in approaches to a complex target molecule.One of the new syntheseP is a culmination of the Bu'02C *C02Bu' Bu'0,C *C02BuL (75) (76) 64 T. R. Govindachari and N. Viswanathan Indian J. Chem. 1972,10,453;Phytochemistry 1972,11 3529. E. Winterfeldt T. Korth D. Pike and M. Boch Angew. Chem. Internat. Edn. 1972 11 289; M. Boch T. Korth J. M. Nelke D. Pike H. Radunz and E. Winterfeldt Chem. Ber. 1972 105 2126. 504 H. F. Hodson 0 (77)R' = R2 = H (78)R' = Et R2 = H (79)R' = a-Et R2 = P-OH biogenetically motivated approach described previously66 in which suitably substituted tetrahydrocarbolines are autoxidized to bases with the camptothecin chromophore.Thus the key indole intermediate (75) was autoxidized to the corresponding quinolone (76; R = OH) and thence via (76 R = Cl) to (76; R = H). Selective reduction (di-isobutylaluminium hydride) of the methoxy- carbonyl function of (76; R = H) gave an alcohol which was readily converted into the lactone (77); ethylation to (78) was followed by autoxidation under carefully controlled conditions [DMF-Et,N-Cu(OH),-O,] to give (*)-campto- thecin (79). 7-Chlorocamptothecin [cf. (79)] was also prepared from (76; R = Cl). Another synthesis6' also proceeded by way of deoxyde-ethylcamptothecin (77) constructed in this case (Scheme 5) by attaching the ring E elements to the pyridone (80) to give (81) followed by reduction of the aldehyde function lac- tonization (with deformylation) and finally dehydrogenation.(77) OMe (80) (81) Reagents i Vilsmeir; ii NaH-CH,(CO,Bu'),; iii NaBH,; iv HCl; v DDQ Scheme 5 From the laboratory in which camptothecin was first isolated comes a synthesis6* in which the key step is the Michael condensation of (82) and (83) to give (84) the cyanohydrin of which contains all the necessary structural elements of the alkaloid. 66 Ref. 6 p. 510; J. Warneke and E. Winterfeldt Chem. Ber. 1972 105 2120. '' T. Sugasawa T. Toyoda and K. Sasakura Tetrahedron Letters 1972 5109. 68 M. C. Wani H. F. Campbell G.A. Brine J. A. Kepler M. E. Wall and S. G. Levine J. Amer. Chem. Soc. 1972,94,3631; M.E. Wall H. F. Campbell M. C. Wani and S. G. Levine ibid. p. 3632. Alkaloids 505 C0,Me C0,Me (84) 0t (83) Pyridine-2,5-dicarboxylicacid providing the D-ring of camptothecin in the fourth ~ynthesis,~’ was converted into the diol(85) and thence utilizing a Claisen rearrangement into (86).The ketone derived from (86) was then used in a Fried- lander quinoline synthesis to provide (87) which was transformed via (fj-deoxycamptothecin (78) into (fj-camptothecin (79) in 11 % overall yield from the pyridine diacid. EtACO,Me (87) In a new general method” for the introduction of alkyl or alkenyl groups into heterocycles a Wittig reagent displaces halide ion from a suitable halogeno- heterocycle [e.g.(SS)] to give a new ylide (e.g.(89)l which can be hydrolysed to an alkyl heterocycle or can be subjected to the usual reaction of Wittig reagents with carbonyl compounds to give an olefin.The latter modification has been incor- porated into a new synthesis’l of (+)-quinine via (88) (89j and (90); further C. Tong and H. Rapoport J. Amer. Chem. SOC.,1972,94 8615; J. Plattner R. D. Gless and H. Rapoport ibid. p. 8613. 70 E. C. Taylor and S. F. Martin J. Amer. Chem. Soc. 1972 94 2874. ” E. C. Taylor and S. F. Martin J. Amer. Chem. Soc. 1972 94 6218. 506 H. F. Hodson (88) 2CH 2= PPh H 1 VNA~ CH=PPh IH CHO -6J (89) elaboration of (90) followed the methods employed by the Roche group in their 1970 synthesis.72 8 Lycopodium Alkaloids A new alkaloid gymnamine (91) very closely related to lycodine (92),has been isolated from the leaves of the flowering plant GyrnnernasyIvestre (Asclepiadaceae) and is the first base of this type to be encountered outside the Lycopodiaceae.Structure (91) was established by the conversion of gymnamine into lycodine (92),as ill~strated.'~ A synthe~is'~ of (+)-luciduline (Scheme 6) depended on the stereoselective formation of the cis-decalin (94). Noteworthy features of the sequence are the diene synthesis in the first stage and the stereoselective oxy-Cope rearrangement of (93). l2 Ref. 37 p. 470. 73 G. S. Rao J. E. Sinsheimer and H.M. McIlhenny Chem. and Ind. 1972 537. 74 W. L. Scott and D. A. Evans J. Amor. Cliem. Soc. 1972 94 4779. Alkaloids 507 C1 I + H,C=C.CN -+ A OMe (93) NHMe Me t " 0 H 0DMe (95I MgBr Reagents i A ; ii 250°C; iii (CH,OH),-H,O+; iv HCHO Scheme 6 9 Terpenoid Bases Staphisine first isolated from Delphinium stuphisagria over thirty years ago has now been shown by an X-ray study75 of its monomethiodide to have the novel bis-diterpenoid structure (96).Both halves are clearly derived from atisine- type units although the B unit has suffered an unprecedented rearrangement. Me /N\ Me -75 S. W. Pelletier A. H. Kapadi I. H. Wright S. W. Page and M. G. Newton J. Amer. Chem. SOC.,1972,94 1754. 508 H. F. Hodson The twenty or so alkaloids hitherto isolated from Daphniphyllum macrophyllum comprise a fascinating set of complex bases with a common carbon skeleton classified into three types on the basis of their N-heterocyclic m~ieties.’~ X-Ray studies of two new alkaloids from this species daphnilactone A (97)” and daphnilactone B (98),’*show them to be examples of a fourth and fifth type.0 H MeNgH H’ 0 Hitherto only the carbon skeleton of (+)-dendrobine [(99) as natural (5)-enanti~mer’~] of this orchid had been prepared but now two total syntheses80*81 alkaloid have been reported; nothing less than a full account could do justice to the skill and ingenuity of these notable achievements. 7h 0.E. Edwards in Ref. 1 p. 375; M. Toda Y. Hirata and S. Yamamura Tetrahedron 1972 28 1477; H. Irikawa S. Yamamura and Y. Hirata ibid. p. 3727. 7’ K. Sasaki and Y.Hirata Tetrahedron Letters 1972 1275; J.C.S. Perkin II 1972 1411. 78 H. Niwa M. Toda and Y. Hirata Tetrahedron Letters 1972 2697; K. Sasaki and Y. Hirata ibid. p. 1891. 79 D. Behr and K. Leander Acra Chem. Scand. 1972,26,3196. Y. Inubushi T. Kikuchi T. Ibuka T. Tanaka I. Saji and K. Tokane J.C.S. Chem. Comm. 1972 1252. 81 K. Yamada M. Suzuki Y. Hayakawa K. Aoki H. Nakamura H. Naguse and Y. Hirata J. Amer. Chem. SOC. 1972 94 8278.