17 BIOSYNTHESIS By R. Ramage (The Robert Robinson Laboratories The University Oj'Liverpool) THE major concentration of effort in this field during the last year was con- cerned with the genesis and interrelationships of the indole alkaloids together with studies of the mechanisms involved in higher terpenoid biosynthesis. Alkaloids.-There has been a rapid decrease in effort devoted to the bio- synthesis of compounds derived from amino-acids with the exception of the isoprenoid alkaloids which will be discussed separately. In an excellent review' it has been shown that the biosynthesis of aromatic amino-acids is not regulated in the same way by all organisms. Two groups have studied'. the biosynthesis of capsaicin (1)and have showed that phenyl- alanine is the source of the benzylamine residue.The aliphatic moiety was found to be derived from valine. [~-'~C]Valine was incorporated into the terminal isopentylene part of the molecule in accord with the suggestion4 that valine is a precursor of isobutyryl coenzyme A which serves as a starter unit for the production of even-numbered iso-fatty acids. The unusual amino-acid p-aminophenylalanine (2) has been shown to be a precursor of chloramphenicol(4; R = NO,) produced by cultures of Strepto-myces species. [c1-l 4C a-' 'Nlp-Aminophenylalanine (2) was incorporated into the p-nitrophenylserinol moiety with only a small change in the 14C ''N ratio. Specific incorporation of ~~-threo-p-aminophenyl[c~rboxy-'~C]serine (3) and the amino analogue of chloramphenicol (4; R = NH') delineate the late stages of the biosynthesis of chloramphenicol(4; R = NO,) in considerable detail.' A study6V7 of gliotoxin (5) biosynthesis in Trichoderma viride made use of 13C 14C and "N as labelling isotopes.Incorporation of [l-I4C ''Nlphenyl- alanine indicated that phenylalanine only labelled N-5 in gliotoxin (5) but that transamination must also occur. [1-' 3C 3-'4C]Phenylalanine was however incorporated with unchanged isotope dilution showing that the carbon skeleton of phenylalanine remained intact. ["NlGlycine labelled both nitrogen atoms to different extents. F. Lingens Angew Chcm. 1968,350. ' D. J. Bennett and G. W. Kirby J. Chem. SOC. 1968,442. E. Leete and M. C. L. Louden J. Amer. Chem. SOC.,1968,90,6837. P. E. Kolattukudy Science 1968 159 498.' R. McGrath L. C. Vining F. Scala and D. W. S. Westlake Canad. J. Biochem. 1968,46,587. A. K. Bose K. G. Das P. T. Funke I. Kugajevsky 0.P. Shukla K. S. Khanchandani and R. J. Suhadolnik J. Amer. Chern. SOC.,1968,90 1038. ' A. K. Bose K. S. Khanchandani T. Tavares and P. T. Funke J. Amer. Chem. SOC. 1968,90 3593. 578 R. Ramage DL[~-’~C; 3,5-3H,]Tyrosine was fed to ‘Twink’ and ‘Deanna Durbin’ daffodils in an attempt’ to elucidate the relative stereochemistry of in vim protonation and hydroxylation at C-2 in the biosynthesis of norpluviine (6) and lycorine (7; R = OH) respectively. As expected the isolated alkaloids had lost half of the tritium fed but more significantly the biosynthetic lycorine (7; R = OH) had the same tritium content as norpluviine (6).This suggests that both protonation and hydroxylation at C-2 are stereospecific and further that the hydrogen removed on hydroxylation at C-2 is the same as that intro- duced in the formation of norpluviine (6). An intermediate in the biosynthesis of lycorine (7; R = OH) could conceivably be (8) which should yield the alkaloid by allylic rearrangement. This would also explain the result of Wild- man’ who found that caranine (7; R = 3H) stereospecifically labelled with tritium at C-2 was incorporated into lycorine (7; R = OH) with retention of tritium. Further work” on the biosynthesis of mesembrine (9) which has structural NMe (7) (8) (9) I. T. Bruce and G. W. Kirby Chem. Comm. 1968,207. W. C.Wildman and N. E. Heimer J. Amer. Chem. Soc. 1967,89,5265. lo P. W. Jeffs 1.U.P.A.C.Meeting London 1968. Biosynthesis 5 79 similarities to the Amaryllidaceae alkaloids showed that o-methylnorbelladine (10; R' = OMe R2= OH) was not an intermediate. However the isomer (10; R2= OH R2= OMe) was incorporated well into mesembrine (9) in Scelatium stricturn implicating the intermediacy of the spiro-dienone (11) which would be expected to fragment in the manner shown. (10) (11) Investigations"* l2 into the biosynthesis of pellotine (12) in the peyote cactus Lophophoru williamsii have shown the difficulties involved in a detailed examination of more primitive alkaloid structures. It was readily shown that dopamine was incorporated into the tetrahydroisoquinoline ring system however the timing of the hydroxylation of the aromatic ring and the subsequent methylation still remains uncertain.Feeding of [2-14C]acetate led to equal labelling of C-1 and C-9 but [l-'4C]acetate was incorporated better into C-1. [14C]Formic acid was a better precursor and again this produced almost equal labelling at C-1 and C-9. The work failed to identify the source of the 2-carbon unit but indicated the problems involved in the early stages of a1 k alo id bi osyn t he sis . [Il-14C]Tryptophan was administered13 to Nicotiana tabucum and the 6-hydroxykynurenic acid (13) isolated was shown to have the label in the carboxy-group. The indolenine peroxide (14) was postulated as a likely inter- mediate. The biosynthesis of psilocybin (15) was studiedI4 in Psilocybecubensio and the results indicate that hydroxylation of NN-dimethyltryptamine is the penultimate step.N-Acetyltryptamine (16) has been shown" to be the pre- cursor of harman (17) in PassiJloru edulis. l1 A. R. Battersby R. Binks and R. Huxtable Tetrahcdron Letters 1968,6111. l2 J. Lundstrom and S. Agurell Tetrahedron Letters 1968,4437. l3 M. Slaytor L. Copeland and P. K. Mac Nicol Phytochernistry 1968 1779. l4 S. Agurell and J. L. G. Nilsson Acta Chem. Scad. 1968,22 1210. is M. Slaytor and I. J. McFarlane Phytochemistry 1968 605. 580 R.Ramage Anthranilic acid and phenylalanine have been provenI6 to be the structural units of aborine (18) in Glycosrnis artoreu. Further r3H]anthranilic acid was found to be the precursor of ring A in the acridine alkaloid aborinine (19).A study of quinazoline alkaloid formation in Peganurn harnala revealed" that vasicine (20) was synthesised in uiuo from anthranilic acid and putrescine (21; R = H). ''N Studies showed that both the CI and 6 amino-functions of ornithine (21; R = C02H) were incorporated to the same extent suggesting the symmetrical intermediate (21 ;R = H). N-Methylisopelletierine (22 ; R = Me) formation in Sedurn sarrnentosurn has been found" to involve lysine. [6-I4C; 4,5-3H2]Lysine was utilised with the 3H :14C ratio unchanged and specific incorporation of 14C into C-6 of the alkaloid. This indicates a nonsymmetrical intermediate in the biosynthesis of N-methylisopelletierine (22; R = Me) unlike the degradation of ornithine to putrescine discussed earlier.[l-14C]Acetate was found' to be incorporated into the 2-position of the side-chain in (22; R = Me). Dimerisation of pelle-tierine (22; R = H) has been implicated" in the biosynthesis of lycopodine l6 D. Groger and S. Johne 2.Naturforsch. 1968,236 1072. 17 D. Liljegren Phytochemistry 1968 1299. l8 R. N. Gupta and I. D. Spenser Chem. Comm. 1968,85. l9 D. G. O'Donnovan and M. F. Keogh Tetrahedron Letters 1968,265. R. N. Gupta M. Castillo D. B. MacLean I. D. Spenser and J. T. Wrobel J. Amrr. Chem. Soc. 1968,90 1360. Biosynthesis 581 (23) in Lycopodium flabelliforme. Incorporation of [2-14C]- and [6-14C]-lysine into lycopodine (23) disposed of the idea that the lycopodium alkaloids are polyketide in origin.If the lysine were transformed into a symmetrical inter- mediate the lycopodine would be labelled as shown i.e. the carbonyl carbon would contain 25 % of the 14C activity. This was found to be the case which is interesting in view of the specific incorporation of [6-'4C]lysine into N-methylisopelletierine(22;R = Me) discussed previously. Isoprenoid Alkaloids.-The discovery2'-' that loganin (24) is a progenitor of the indole and ipecacuanha alkaloids has stimulated much effort in the later stages of the biosynthesis of these classes of alkaloids. Loganin was however never considered to be the actual unit which combined with trypta- mine or dopamine. Further oxidative processes namely cleavage of ring A to give the aldehyde (25) would have to be involved before such a union could be achieved in uiuo or in vitro.Proof of this came as a result of both structural .O Gluc. C.0 Me d02Me (27) a-@Me (28) \N OAc Me HO C02Me (29) 21 A. R. Battersby R. S. Kapil J. A. Martin and Mrs. L. Mo Chem. Comm. 1968 133. 22 P. Loew and D. Arigoni Chem. Comm. 1968,137. 23 A. R. Battersby and B. Gregory Chem. Comm. 1968 134. 582 R. Ramage elucidation of trace alkaloids and ultimately tracer methods. The E.T.H. group showed24 the glycoside foliamenthin ex. Menyanthes trfoliata to have the structure (26a) easily recognisable as a derivative of the elusive aldehyde (25). In keeping with this structure [4-14C] geraniol was incorporated into foliamenthin (26a) in Menyanthes trijoliata.Another important glycoside menthiafolin was identified25 as (26b) and [2-14C]geraniol feeding produced menthiafolin (26b) in which the 14C label was incorporated into the ester and lactol moieties in the rxtio 3:l. Menthiafolin (26b) thus gave a source of synthetic and biosynthetic secologanin (25) which was exploited26 in an elegant synthesis of ipecoside (27). That secologanin (25) was indeed the building block of the indole alkaloids was shown by feeding [O-r~thyl-~Hl- secologanin (25) to Vinca rosea. This afforded the following alkaloids with the corresponding incorporations of 14C ajmalicine (28) 0.55 % ; vindoline (29) 0.12 % ; catharanthine (30) 016 % ; and perivine (31) 0.13 %. The incorpora- tion2' of sweroside (32) into vindoline (29) in Vinca rosea and also28 reserpine and quinine probably proceeds by oxidation to secologanin (25).However since the mechanism of the transformation of loganin (24) to secologanin (25) is as yet unknown this result may have some deeper significance. The indole counterpart of ipecoside (27) would be expected to have structure (33) which has been assigned without stereochemical detail to a new alkaloid strictosidine isolated29 from Rhazia stricta and R. orientalis. [O-methyL3H]-Loganin (24) was incorporated3' efficiently into strictosidine (33) showing it to be an important intermediate in indole alkaloid biosynthesis. Condensation of [O-methyl-3H]secologanin (25) with tryptamine afforded31 the expected p-carbolines (33) epimeric at position 5.This mixture was shown to be efficiently 24 P. Loew Ch. V. Szczepanski C. J. Coscia and D. Arigoni Chem. Comm. 1968 1276. " A. R. Battersby A. R. Burnett G. D. Knowles and P. G. Parsons Chem. Comm. 1968 1277. " A. R. Battersby A. R. Burnett and P. G. Parsons Chem. Comm. 1968 1280. "l H. Inouye S. Ueda and Y. Takeda Tetrahedron Letters 1968,3453. 28 H. Inouye S. Ueda and Y.Takeda Tetrahedron Letters 1969,407. 29 G.N. Smith Chem. Comm. 1968,912. 'O R. T. Brown G. N. Smith and K. S. J. Stapleford Tetrahedron Letters 1968,4349. A. R. Battersby A. R. Burnett and P. G. Parsons Chem. Comm.,1968 1282. Biosy n thesis 583 incorporated into a ajmalicine (28) vindoline (29) catharanthine (30) and perivine (31).With doubly labelled p-carbolines (33)prepared from [O-rnethyl-3H]secologanin (25) and [3H]tryptamine it was shown that biosynthesis occurred without a change in the 3H:14C ratio.By dilution techniques it was shown that both p-carbolines and secologanin are present in Vinca rosea. C0,Me Another area of indole alkaloid biosynthesis which received much attention was the genesis of the three structural types corynanthe (34) aspidosperma (35) and iboga (36). The ready isomerisation of the Aspidosperma alkaloid (-)-tabersonine (37) to the Iboga alkaloid (+)-catharanthine (38) is thought32 to go via the intermediate (39) which could also be a transformation product of the alkaloid stemmadenine (40). This led to stemmadenine (40) being postulated 32*33 as a key intermediate for Aspidosperma and Iboga alkaloids.It is interesting to note that the alkaloids named secamines (41) isolated34 from Rhazia stricta may be regarded as dimers of (39) or closely related struc- ture. There are plausible pathways for the formation of stemmadenine (40) from the Corynanthe alkaloid geissoschizine (42). Acid-catalysed rearrange- ment3’ of (42) afforded catharanthine (30) together with pseudocatharanthine (39) (40) 32 A. A. Qureshi and A. I. Scott Chem. Comm. 1968,945. 33 J. P. Kutney C. Ehret. V. R. Nelson and D. C. Wigfield J. Amer. Chem. SOC. 1968,90 5929. 34 D. A. Evans G. F. Smith G. N. Smith and K. S. J. Stapleford Chem. Comm. 1968,859. 35 A. A. Qureshi and A. I. Scott Chem. Comm. 1968,947. 584 R. Ramage Me (42) OH (43). The latter can be considered to be derived from (44),an intermediate of the Strychnos alkaloid type.A novel approach36 to the elucidation of the late stages of indole alkaloid biosynthesis employed short-term germination (43) (44) of Vinca rosea seeds. The sequence of alkaloid formation was found to be Corynanthe Aspidosperma and Iboga. [O-methyL3H]Sternmadenine was incorporated well into catharanthine (30) and vindoline (29); similarly [O-methyL3H]tabersonine was found to be efficiently transformed into catharan- thine and vindoline. However cathatanthine was not incorporated into vindoline thus showing the irreversible nature of the change from Aspido- sperma to Iboga alkaloids. It was also found that specific incorporations of Corynanthe precursors in Vinca rosea were much higher in germinating seedlings than in the whole plant.By usine 6-month old Vinca rosea plants K~tney~~ also found that labelled tabersonine (37) was incorporated into catharanthine (30) and vindoline (29). The incorporation of DL-[~-'~C]-tryptophan into vincadine (45) and vincadifformine (46) was studied33 over different time intervals and the results obtained indicate that there is no direct biosynthetic relationship between the two alkaloids in spite of the smooth in vitro conversion of (45) into (46). 36 A. A. Qureshi and A. I. Scott Chem. Comrn. 1968,948. 37 J. P. Kutney W. J. Cretney J. R. Hadfield E. S. Hall V. R. Nelson and D. C. Wigfield J. her. Chem. Soc. 1968,90,3566. Biosynthesis 585 1soprenoids.-Monoterpenes.A of the biosynthesis of (+)-and ( -)-camphor (47) by using [2-'4C]mevalonate produced the interesting result that all the 14C was incorporated into the 6 position of camphor. Normal union of two C-5 units would have led to additional labelling at positions 8 and 9. This result agrees with earlier work3' on thujone (48) biosynthesis where [2-14C]me~alonate was incorporated specifically at the position shown. 0 9e (48) Obviously more work. must be done on the initial stages of terpene for- mation in the higher plants. Isopentenyl pyrophosphate isomerase has been isolated from liver.40 It was found to be activated by Mn2+ ions and the equilibrium between isopentenyl pyrophosphate and dimethylallyl pyro- phosphate favoured the latter.[2- ''C]Mevalonate feeding4' to Santolina chamaecyparissus gave a very low incorporation into artemesia ketone (49). The biosynthesis of nepetalactone (50) in Nepeta cataria L. was studied42 by using [2-'4C]mevalonate and it (49) 38 D. V. Banthorpe and D. Baxendale Chem. Comm. 1968,1553. 39 D. V. Banthorpe and K. W. Turnbull Chem. Comm. 1966,177. 40 P. W. Holloway and G. Popjak Bochom. J. 1968,106 835. 41 G. R. Waller G. M. Frost D. Burleson D. Brennon and L. H. Zalkow Phytochemistry 1968 213. 42 F. E. Regnier G. R. Waller. E. Z. Elsenbraun. and H. Anda Phytochmistry 1968 7 221. 586 R. Ramage was found that positions 3 8 6 and 9 had 36 17,29 and 18 % of the activity respectively. This indicates considerable randomisation at the isopentenyl stage.A cell-free system has been prepared43 from Mentha piperita which can convert pulegone (51) to menthone (52) and isomenthone (53) in the presence of NADPH,. [l-'4C]Geranyl pyrophosphate was tran~formed~~ @ GoMe& into cineole (54) in Rosrnarunus oficinalis 2nd degradation gave the expected equal labelling so shown. Qo Me Me Me Me Me Me (54) (51) (52) (53) Sesquiterpenes. In vitro acid-catalysed transformations4' of the epoxides (55) and (56) wlll undoubtedly have relevance to future biosynthetic studies in the endesmane and guaiane series. Rearrangement of (55) afforded (57) and (58) which is explicable in terms of a strict application of the Markownikoff rule; the epoxide (56) however gave the guaiane-type (59) predicted on steric grounds.This delicate balance between steric and electronic effects can be expected to play at least as important a role in sesquiterpene biosynthesis as in the biological transformations of 2,3-oxidosqualene leading to triterpenoids. The biosynthesis of ~antonin~~ (60) in Arternesia rnaritirna gave very low incorporations of all precursors including mevalonate and [l-3H]farnesol showing the great difficulty of transporting the precursor to the site of synthesis. Incorporation of the lactones (61) and its double-bond isomer indicate that the introduction of oxygen into ring A is a late step. Diterpenes. [l-3H]Geraniol pyrophosphate was incorporated specific- 48 into rosenonolactone (62) in Tricothecium roseurn.Administration of 43 J.Battail A. J. Burbott and W. D. Loomis Phytochemisrry 1968 1159. 44 B. Achilladelis and J. R. Hanson Phytochemistry 1968. 1317. 45 E. D. Brown and J. K. Sutherland Chem. Comm. 1968. 1060. 46 D. H. R. Barton G. P. Moss,and J. A. Whittle J. Chem. SOC.(C) 1968,1813. 47 B. Achilladellis and J. R. Hanson Phytochemistry 1968 589. 48 B. Achilladellis and J. R. Hanson Tetrahedron Letters 1968,4397. Biosynthesis 587 Me 4R-[4-3H 2-'4C]mevalonolactone afforded (62) with an unchanged 3H :14C ratio. Degradation placed the 3H labels as shown thus verifying the postu- lated4' hydrogen migration from C-9 to C-8 during biosynthesis and also eliminated the lactone formation via a A59 rosadiene intermediate. A studys0 of the biosynthesis of viridin (63) showed that the furan carbon indicated was not labelled by [2-14C]mevalonate and presumably arose by oxidation of a 3a-methyl group.[7-3H]Kaur-16-en-19-oic acid (64; R = H) was founds1 to be an effective precursor of steviol (64; R = OH) in Stevia rebandiana suggesting bridgehead hydroxylation as the final biosynthetic step. In a further investigation5 into the biosynthesis of tetracyclic diterpenes the Sussex group fed 4R-[4-'HH 2-'4C]mevalonate to Gibberella fujikuroi and isolated labelled gibberellic acid (65) plus 4,18-dihydroxykaurenolide(66). The 3H labels in the biosynthetic diterpenes were shown to be at the positions indicated and the P-CH,OH grouping in (66) found to be derived from [2-14C]mevalo- nate. Since 4R-[13H]mevalonate should give 3H at C-2 of gibberellic acid (65) in the P-configuration it follows that hydroxylation at this centre must have occurred with inversion.Retention of mevalonate-derived hydrogen at C-4b and C-lOa in (65) excludes the formation of either 4a 4b- or 4a 10a-double- bonds during loss of the angular methyl group and lactone formation. Also retention of the 3H at C-9 in (66) rules out any intermediacy of pimara-8,9- diene (67) during the biosynthesis of tetracyclic diterpenes of the kaurane type. In order to determine whether pimara-7,8-diene was in fact involved [Z3H, 49 A. J. Birch R. W. Richards H. Smith A. Harris and N. B. Whalley Tetrahedron 1968 7,241. M. M.Blight J. J. W. Coppen and J. F. Grove Chem. Comm. 1968,1117. J. R. Hanson and A. F.White Phytochemistry 1968,595. '2 J. R. Hanson A. Hough and A. F. White Chem. Comm. 1968,467. 588 R. Ramage Me 10a HO 10 OH Me 98 C0,H (65) (66) (67) 2-'4C]mevalonate was fed5 to Gibberellafujikuroi which should label position 7 of kaurene (68) nonstereospecifically with 3H.No loss of 3H from this position was observed which eliminated pimara-7,g-diene as a precursor. Pimara- 8,14-diene was specifically incorporated into the tetracyclic diterpenes (65) and (66) although with low efficiency. In studiess4. 55 designed to determine the oxidative sequence leading from a tetracyclic gibberane skeleton to gib- berellic acid (65) structure (69) and gibberellin A14 (70) were found to be Me Me 8Me Me (68) 'OZR CHO effective precursors of gibberellic acid (65) and gibberellin A13 (71).The latter was found not to be an intermediate in the biosynthesis of gibberellic acid (65). These workers also synthesised (69) by base treatment of (72). Bearing in mind the oxygenation pattern of (66) which co-occurs with gibberellic acid (65) this transformation must be closely analogous to the biosynthetic pathway. 53 J. R.Hanson and A. F. White Chem. Comm. 1968 1689. 54 B.E. Cross R. H. B. Galt and K. Norton Tetrahedron 1968,24,231. 55 B.E. Cross K. Norton and J. C. Stewart J. Chem. SOC.(C),1968 1054. Biosynthesis 589 Steroids and Triterpenes. Investigations into the reductive dimerisation of two CI5 units in the biosynthesis of squalene have shown that thiamine pyrophosphate is implicateds6 in the formation of squalene.Stereoisomers having structure (73) assigned to an isolated intermediate in squalene bio- synthesiss7 were synthesised” and shown to be different from the natural material. H R (74) 0& RR (77) In order to elucidate details of substrate specificity of 2,3-epoxysqualene cyclase unnatural epoxides (74) were treated with 100,OOO g. supernatant preparation of rat liver microsomes.s9~60 Only the trans-oxide (74; R1= Me ” G. E. Risinger and H. D. Durst Tetrahedron Letters 1968,3133. ” H. C. Rilling J. Bid. Chem. 1966,241 3233. ’* E. J. Corey P. R. Ortiz de Montellano Tetrahedron Letters 1968 5113. ” R. B. Clayton E. E. van Tamelon and R. G. Nadean J. Amer. Chem. Soc. 1968,90 820. 6o E.J. Corey K. Lin and M. Jantelar J. Amer. Chem. Soc. 1968,90. 2724. 590 R.Ramage R2 = H) yielded the 4-desmethyllanosterol analogue (75 ;R' = Me R2 = H). [23-14C]29,30-Bisnor-2,3-epoxysqualene(76; R' = H R2= R3= Me) was treated with a particle-free solution of 2,3-epoxysqualene-amyrincyclase from pea seedlings6' and gave 29,30-bisnoramyrin (77; R = H). Although the ion (78 ;R = Me) has been proposed62 as an intermediate in p-amyrin (77 ;R = Me) biogenesis it would appear unlikely that the corresponding primary carbonium ion (78 ;R = H)would be involved in the bisnor series. In an attempt to separate the cyclisation and rearrangement processes involved in sterol biosynthesis the bisnor-2,3-epoxysqualene(76; R' = Me R2 = R3= H) was treated with the cyclase enzyme system.63 One ofthe important driving forces causing rearrange- ment of the primary cyclic intermediate (79; R' = R2 = Me) to lanosterol (75; R' = R2 = Me) is the relief of repulsive interactions in ring B (twist boat) R HO (79) (80) due to the methyl group at C-8.In the bisnor intermediate (79; R' = R2= H) no such interactions exist and not surprisingly the product of cyclisation was found to be (80) formed by deprotonation of (79; R' = R2 = H). Thus the important function of the enzyme is the cyclisation stage after which the relative stabilities of carbonium ions control the product formation. Another interesting example which showed the importance of the C-8 methyl group in (79) causing rearrangement to the lanostane skeleton was given by treatment of 15-nor-2,3-epoxysqualene (76; R2= H R' = R3= Me) with 2,3-epoxysqualene- lanosterol ~yclase.~~ The product (75; R' = H R2 = Me) was shown not to E.J. Corey and S. K. Gross J. Amer. Chem. SOC. 1968,90,5045. A. Eschenmoser L.Ruzicka 0.Jeger and D. Arigoni Helu. Chim. Acta 1955,38 1890. 63 E. J. Corey P. R. Ortiz de Montellano and H. Yamamoto J. Amer. Chem. SOC.,1968,90,6254. 64 E.E.van Tamlen R. P. Hanzlik K. B. Sharpless,R. B. Clayton W. J. Richter and A. L. Burlin-game J. Amer. Chem. SOC.,1968,90,3284. Biosynthesis 591 have incoporated 3H from the medium hence cyclisation and subsequent rearrangement of the 15-nor series is identical to that occurring in 2,3-epoxy- squalene. HO' In a study6' of the biosynthesis of the antibiotic fusidic acid (81),in Fusidiurn coccineum it was shown that 2,3-epoxysqualene was incorporated intact confirming the earlier work66 with [2-14C]mevalonate.The formation of helvolic acid (82) in Cephalosporiurn caerulens has been shown6' to involve 3~-hydroxy-4-~-hydroxymethylfusida-l7(20)[ 16,20-cis]24-diene (83). In the transformation of (83) to helvolic acid (82) the 4P-hydroxymethyl group is eliminated and thus gives a clue to the possible mechanism and sequence of demethylation of 4 :4-dimethyl sterols. An investigation6' designed to give more information on this problem revealed that 4P-methyl-4a-hydroxymethyl-cholestanol (84 ; R' = CH20H R2 = Me) and 4a-hydroxymethylcholestanol (84; R' = CH20H R2 = H) were converted efficiently into cholestanol (84; R' = R2 = H) by rat liver homogenates.However in contrast to the helvolic acid biosynthesis it was found that. 4a-methyl-4P-hydroxymethyl-cholestanol(84; R' = Me R2 = CH20H) was not transformed into cholestanol (84; R' = R2 = H). These results indicating initial removal of the 4a-methyl group should be compared with earlier on the biosynthesis of choleste- rol from [2-14C]mevalonate which suggested that the 4P-methyl group is the first to be eliminated as in the case of helvolic acid (82). The idea that cycloartenol (85) is the first product of cyclisation and re- arrangement of 2,3-epoxysqualene is supported by the incorporation of 2,3-epoxysqualene into cycloartenol (85) in a cell-free system from newly '' W.0.Godtfredsen H. Lorck E. E. van Tamelen J. D. Willett and C. B. Clayton J. Amer. Chem. SOC.,1968,90,208. 66 D. Arigoni Conference on the Biogenesis of Natural Products Academia Nzlzionale dei Lincei Rome 1964. '' I. Okuda Y. Sato T. Hattori and H. Igarashi Tetrahedron Letters 1968,4769. 68 K. B. Sharpless T. E. Snyder T. A. Spencer K. K. Maheshwari G. Guhn and R. B. Clayton J. Amer. Chem. SOC.,1968,90,6874. 69 J. L. Gaylor and C. V. Delwiche Steroids 1964 4 207. 592 R. Ramage HO developed bean leaves.” The biosynthesis of (85) was studied71 by using 4R[2-14C 4-3Hl]mevalonate and it was found that the 3H:14C ratio in cycloartenol (85) was the same as in the squalene isolated. This indicates a hydrogen migration from C-9 to C-8 with concomitant cyclopropane-forma- tion ; this eliminates lanosterol (86) as an intermediate in the biosynthesis of cycloartenol(85).2,3-Epoxysqualene was shown to be a precursor of lano- sterol (86)in yeast and thus the low turnover of lanostadiene (8,24) to lanosterol must be due to hydroxylation of an unnatural precursor.72 The remarkable triterpene tetrahymanol (87) which is a metabolite of the protozoan Tetrahymena pyricforrnis has been shown7 3*74 to be produced by acid-catalysed cyclisation of squalene terminated by nucleophilic attack at C-21 rather than from 2,3-epoxysqualene. Since deoxytetrahymanol has C2 symmetry both biosynthetic routes were possible. Elimination of the 14~-methyl group of lanosterol (86) in the biosynthesis of sterols is accompanied by stereospecific removal of the 15a-hydr0gen.~ 5-77 4,4-Dimethyl-5a-cholesta-8,14-dien-3~-ol(88) produced by such a transforma- tion has been shown7* to be an intermediate between lanosterol (86) and 70 H.H. Rees L. J. Goad and T. W. Goodwin Tetrahedron Letters 1968 723. 71 H. H. Rees L. J. Goad and T. W. Goodwin Bochem. J. 1968,107,417. 72 D. H. R. Barton A. F. Gosden G. Mellows and D. A. Widdowson Chem. Comm. 1968 1067. 73 E. Caspi J. B. Greig and J. M. Zander Bochem. J. 1968,109,931. l4 E. Caspi J. M. Zander J. B. Greig F. B. Mallory R. L. Couner and J. R. Landrey J. Amer. Chem. SOC. 1968.90.3563. 75 L. Canonica H. Fiecchi M. G. Kienle H. Scala G. Galli E. G. Paoletti and R. Paoletti J. Amer. Chem. SOC., 1968,90,3597. 76 G.F. Gibbons L. J. Goad and T. W. Goodwin Chem. Comm. 1968 1458. 77 E. Caspi J. B. Greig P. J. Ramm,and K. R. Varna Tetrahedron Lettcrs 1968,3829. 78 L. Canonica A. Fiecchi M. Gallikiente A. Scale G. Galli E. Grossi E. G. Paoletti and R. Paoletti J. Amer. Chem. SOC. 1968,90 6532. Biosyn thesis 593 (87) (88) HO cholesterol (89) using rat liver homogenates in the presence of oxygen. Under anaerobic conditions the conversion of (88) is stopped at C-29-A8-sterols (90; R = Me). The biological conversion of 5a-cholest-8-en-3P-01 to 5a-cholest-7-en-3P-01 during biosynthesis of cholesterol (89) has been shown to involve elimination of the 7fLhydrogen. No migration of hydrogen from C-7 to C-9 was observed ;799 8o the hydrogen attached to C-9 in (91) being acquired from the medium.The remaining steps between (91) and cholesterol (89) involve introduction of a A5-double bond to give (92) followed by reduction of the A'-double-bond. cis-Elimination of the 5a-and 6a-hydrogens have been sh~wn,~'-~~ to be involved in the formation of the As-double-bond. In agree- ment with this the formation of the phytosterol periferasterol (93) in Ochro-monas malhamensis has been shown,84 by feeding 3R-[2-14C (5R)-R-3Hl]-mevalonate to involve elimination of the 6a-hydrogen and also the 23-pro-R- hydrogen. A study of the biological reduction of (92) to cholesterol revealed that the 8P-hydrogen in cholesterol was obtained from the medium and the 7a-hydrogen from NADH. The removal of the C-19 methyl group in the biological conversion of androsterone (94; R = Me) to oestrone (95) has been showng5 to involve the 19-hydroxymethylandrosterone(94; R = CH,*OH) and the 19-formylandrosterone (94; R = CHO).Polyketide-derived Compounds.-Bohlmann has studied the biosynthetic relationships of many natural acetylenes. The biosynthesis of phenylhep- 79 L. Canonica A. Fiecchi M. Gallikienle A. Scala G. Galli E. Grossi E. G. Paoletti and R. Paoletti Steroids 1968 287. M. Akhtar and A. D. Rahimtala Chem. Comm. 1968,259. M. Akhtar and S. Marsh Biochem. J. 1967,102,462. S. Dewhurst and M. Akhtar Biochem. J. 1967,105 1187. 83 A. M. Paliokes and G. J. Schroepfer J. Biol. Chem. 1968,243,453. 84 A. R. H. Smith L. J. Goad and T. W. Goodwin Chem. Comm. 1968 1259. M. Akhtar and S.J. M. Skinner Biochem. J. 1968,109 318. 594 R.Ramage HO (94) (95) tatriyne (96) was investigated86 by feeding [2,3-3H2 14-14C]tetradec-5-en-8,10,12-triyn-l-ol (97) to Coreopsis lanceolata L. A possible intermediate (98) would result in the loss of all 3H at C-3 but this is not observed. The origin of the thiomethyl grouping in (99) and (100) was studied87 by using C3H 35S]-methionine which was administered to Anthemis tinctoria L. Isolated esters (99) and (100) had a different 3H:35Sratio from the methionine fed hence the sulphur and methyl group were donated independently to matricaris ester (101) which was also shown to be a precursor of (99) and (100) in Anthemis tinctoria. In order to investigategg the formation of the highly unsaturated hydrocarbons (102) (103) and (104) in Centaurea dilute L.the alcohol (105) was fed and found to be incorporated into the hydrocarbons. This gives some idea of the sequence of triple-bond formation and it was suggested that the terminal olefinic linkage is formed by breakdown of the P-hydroxy-acid intermediate (105). A startg9 has been made in the isolation of the important 86 F. Bohlmann H. Bonnet and K. Jente Chem. Ber. 1968,101,855. '' F. Bohlmann and T. Burkhardt Chem. Ber. 1968,101 861. F. Bohlmann M. Wolschokowsky J. Laser C. Zdero and K. D. Bach Chem. Ber. 1968,101 2056. Biosynthesis 595 enzymes concerned with the biosynthesis of naturally occurring acetylenes with the conversion of oleic acid (106)into (107)with a cell-free system.Linoleic acid (109) had earlierg0 been shown to be an intermediate in the biosynthesis of crepenynic acid (108).It was also proven that neither the diol nor epoxide was an intermediate in the conversion of (109) into (108). Ph[-],Me (96) Me[Cx] 3-CH *CHLCH. [CH,] .CH2 OH (97) (98) MeC=C.CHACH.C==CH-CH=kH*CO,Me $Me (99) 09 F. Bohlmann and H. Schulz Tetrahedron Letters 1968,4795. 90 F. Bohlmann and H. Schulz Tetrahedron Letters 1968 1801. 596 R.Ramage Me [kC] ,CH= w (107) Me[CH,] C=C* CH CH=CH. [CH2] CO,H (108) MeCCH,] .H&H CH,CH=CH * [CH,] * CO H (109) There has been a reviewg1 of the prostaglandins and much work on the biosynthesis of this important class of biologically active substances.The mechanism of the conversion of eicosa-8,11,14-trienoic acid (110) into prosta- glandin El (111) and prostaglandin F1a (112) was studiedg2 with [13D-3H 3-14C]- and [13L-3H 3-14C]-eicosca-8,1 1,14-trienoic acid. The hydrogen lost from C-13 was shown to have the L-configuration and (112) was shown not to have involved (111) as an intermediate. It is suggested that the loss of hydrogen from C-13 leads to the intermediate (113) which can form the cyclic peroxide (114) by a concerted process. Reductive fission of the peroxide would yield prostaglandin F a (112) and oxidative cleavage would afford prosta- glanding El (111). Additional evidence for the cyclic process was providedg3 by the biosynthesis of 12-hydroxyheptadeca-8,lO-dienoicacid (115) and malondialdehyde from eicosa-8,11,14-trienoic acid The fragmentation may be formally represented as a cyclic process.Biosynthesis of 8-isoprostaglandin El 94 (1 11 epimer at C-8) involves isomerisation of prostaglandin El (111). Earlier workgs on the biosynthesis of glauconic acid (116) had shown the pathway involved dimerisation of (117). Further studies96 strongly suggest the 91 S. Bergstrom and B. Sammelsson Science 1968 109. 92 M. Hamberg and B. Sammelsson. J. Biol. Chem. 1967,242,5336. 93 M. Hamberg and B. Sammelsson J. Biol. Chem. 1967,242 5344. 94 E. G. Daniels W. C. Krueger F. P. Kupiecki J. E. Pike and W. P. Schneider J. Amet. Chem. Soc. 1968,90 5894. " C. E. Moppett and J. K. Sutherland Chem. Comm. 1966,772. 96 J. L. Bloomer L. E. Moppett and J.K. Sutherland J. Chrm. SOC. (C),1968 588. Biosynthesis 597 COOH C,-dicarboxylic acid oxalacetic acid as the source of C-8 C-6 and C-7 in (1 17) the remainder being polyketide in origin. The aliphatic lichen acid proto- lichesterinic acid (118) has been shownQ7 to be derived from a C-16 polyketide unit together with a C-3 fragment from another source. An investigationg8 of the synthesis of triacetic acid lactone (119;R = H) by pigeon liver fatty acid synthetase showed that (119;R = H) was formed in the absence of TPNH but that palmitic acid was obtained in the presence of TPNH. Although methyltriacetic lactone (119; R = Me) is not incorporated into stripitatic acid (120) in P. Stipitatum it has been foundQ9 that [14C]-formate is a source of C-1 units for both structures at the positions indicated.These results indicate that the hydroxy-pyrones are free forms of enzyme-held polyketides which are true tropolone precursors and that addition of the C-1 unit occurs at the polyketide level. By using [2-' 3C] -,[1-' 3C]-acetate and [ 3C]formate to study the biosynthesis of sepedonin (121) it was found that formate labelled C-8 specifically in agreement with the previous example of stipitatic acid (120). loo Studies of the biosynthesis of terreic acid (122) indicatelo' that it is formed from 6-methylsalicyclic acid in Aspergillus terreus the epoxide oxygen being 97 J. L. Boomer W. R. Eder W. F. Hoffmann Chem. Comm. 1968,354. 98 J. E. Nixon G. R.Puty and J. W. Porter J. w'ol. Chem. 1968 5471.99 G. S. Man and S.W. Tananbaum J. Amer. Chem. SOC.,1968,90,5302. loo A. G. McInnes D. G. Smith L. C. Vining and J. L. C. Wright Chem. Comm. 1968 1669. G. Read and L. C. Vining Chem. Comm. 1968,935. 598 R. Ramage Me (121) derived from the atmosphere. The very remarkable natural product giorosein (123) which prefers not to assume an aromatic structure has been shownlo2 to be produced by Gliocladiurn roseurn viu (124) and the quinone (125). "'0"' 0 HOOMe 6:: 0 Me0 "//,Me HO\ HO ' 'Me 0 CHO 0 (123) (124) (125) N ~ OH zOH :OH OH H 2 Early worklo3 on the biosynthesis of tetracycline had shown that 6-methyl- pretetramid (126) was a precursor of 7-chlorotetracycline. Recent worklo4- lo6 in this field employed mutants in order to identify the intermediates.In this way the anthraquinone (127) was isolated however on treatment with a tetracycline producing organism it was shown to be a shunt-product from the main pathway. Another such metabolite was (128) in which the 6-methyl group was introduced at an early stage before cyclisation of the tetracene system was complete. It is interesting to note that the oxygen at C-8 is already missing even at this relatively early stage of the biosynthesis. The biosynthesi~'~~-'~~ of lo2 M. W. Steward and N. M. Packter Bochcm. J. 1968,109 1. lo3 J. R. D. McCormick S. Johnson and N. 0.Sjolander J. Amer. Chem. SOC.,1963,85,1692. lo4 J. R. D. McCormick and E. R. Jensen J. her. Chem. SOC.,1968,90,7126. lo5 J. R. D. McCormick E.R. Jensen N. H. Arnold M. S. Carey H. H. Joachim S. Johnson P. A. Miller and N. 0.Sjolander J. Amer. Chem. SOC.,1968,90 7127. Io6 J. R. D. McCormick E. R. Jensen S. Johnson and N. 0.Sjolander,J. Amer. Chem. SOC.,1968 90 2201. lo' M. Biollaz G. Buchi and G. Milne J. Amer. Chem. SOC.,1968,90 5017. Io8 M. Biollaz G. Buchi and G. Milne J. Amer. Chem. SOC.,1968,90,5019. lo9 J. A. Dunkerslott D. P. H. Hsieh and R. I. Matelas J. Amcr. Chem. SOC. 1968 90,5020. Biosynthesis 599 (127) 0 45 (129) aflatoxin B (129) has been studied with [1-14C]- and [2-14C]-acetate feedings to Aspergillusflauus. One fascinating result which emerged was that C-11 and C-14 are both derived from [2-'4C]acetate. A very interesting biogenetic route was proposed'08 involving a tetracene (130; R = H or OH) similar to that involved in tetracycline biosynthesis.The closely related metabolite of Asper-gillus versicolor sterigmatacystin (131; R = H) has also been shown''O to involve head-to-head coupling of two acetate units. A mutant of Aspergillus uersicolor produced by irradiation yielded' ' 5-methoxysterigmatacystin (131;R = OMe).The co-occurrence of versicolorin A (132) and sterigmatacystin in certain strains of the organism suggest that the xanthone may be derived from a related anthraquinone which in turn could be produced via a tetracycline intermediate as suggested by Buchifo8 for the aflatoxin B1 biosynthesis. -0 -70 R 0 OH OH OH (130) (U1) OH 'lo J. S. E. Holker and L. J. Mulheirn Chem.Comm. 1968 1576. J. S. E.Holker and S. A. Kagal Chem. Comm. 1968 1574. 600 R. Ramage The role of aren-oxide-oxepin systems in the metabolism of aromatic systems has assumed greater importance with the synthesis of (4-2H] 3,4-epoxytoluene (133) and its subsequent rearrangement to [3-2H] -p-cre~ol."~ From the meta- bolism of naphthalene' l by rat-liver microsomes 2,3-epoxynaphthalene (135) was isolated by dilution techniques and was found to be hydrated enzymatically to trans-1,2-dihydro-l,2-dihydroxynaphthalene (136). Acetylaranotin (137) is a naturally occurring substance recently isolated which contains the dihydro- oxepin system.'I4 Baldwin has discussed the cleavage of aromatic rings in terms of enzymic generation of a species equivalent in its powers to singlet oxygen.' It was shown that 1,4-peroxides formed from aromatic rings may be trans- formed by acid to cleavage products of the type found in biological systems.'12 D. M. Jerina J. W. Daly and B. Witkop J. Amer. Chem. SOC. 1968,90 6523. D. M. Jerina J. W. Daly B. Witkop P. Zaltzman-Nirenberg and S. Udenfriend J. Amer. Chem. SOC.,1968,90,6525. li4 R. Nagarajan N. Neuss and M. M. Marsh J. Amer. Chem. SOC. 1968,90,6519. '15 J. E. Baldwin H. H. Busson and H. Krauss Chem. Comm. 1968 984.