11 Blological Chemistry Part (ii) Biosynthesis By T. J. SIMPSON Department of Chemistry University of Leicester University Road Leicester LEI 7RH 1 Introduction The previous Report' on biosynthetic studies of secondary metabolites and related work covered the period 1981-85. This Report will cover the period 1986-88. The area continues to be a tremendously active one and testing of biosynthetic mechan- isms and biogenetic ideas remains a great stimulus for synthetic organic chemistry and for the development and application of new methodologies. The application of genetic analysis and gene cloning methodologies have now come to occupy a dominant place in biosynthetic studies allowing such great advances in the study of the enzymology of the pathways of interest that the previous marked distinctions among primary and secondary metabolic and enzyme mechanistic studies are rapidly disappearing.Inevitably this Report must be highly selective with the main aim being to pick out some of the highlights and to convey an impression of how studies on the major biosynthetic pathways are developing. More comprehensive coverage of the areas to be discussed may be found in the regular specialized articles appearing in Natural Product Reports. 2 Fatty Acid and Polyketide Biosynthesis Fatty acid synthetase (FAS) is a multifunctional enzyme complex which shows various patterns of both structural and functional variation depending on its origin.2 In most prokaryotes and in chloroplasts ,the FAS comprises eight structurally independent and monofunctional enzymes.In vertebrates all these components are combined within a single octafunctional polypeptide chain. However in yeasts and lower fungi the FAS consists of two components. In Saccharomyces cerevisiae the two subunits are encoded by the FAS1 and FAS2 genes the former encoding acetyl- malonyl- and palmityl-transferases the dehydratase and the enoyl reductase active sites and the latter the P-ketoacyl synthetase P-ketoacyl reductase and acyl carrier protein domains. Both genes have been fully The fabB gene of T. J. Simpson Annu. Rep. Prog. Chem. Sect. B 1986 83,347. A. D. McCarthy and D. G. Hardie TIBS 1984,9 60. M. Schweizer L. M. Roberts H.-J. Holtke K. Takabayashi E. Hollerer B. Hoffmann G. Muller H. Kottig and E.Scheweizer Mol. Ges Genet. 1986 203 479. E. Schweizer G. Muller L. M. Roberts M. Schweizer J. Rosch P. Wiesner J. Beck D. Stratmann and I. Zanner Fat Sci. Technol. 1987 89 570. 32 I 322 T. J. Simpson Escherichiu coli encoding P-ketoacyi synthetase I has also been sequenced and the active site identified by tagging with [3H] cerulenin and sequencing of radiolabelled peptide fragments following proteoly~is.~ The E. coli and S. cerevisiae condensing enzymes have a common sequence of five amino acids (Ala-Cys-Ala-Thr-Ser) which are believed to contain the active-site cysteine. The stereochemical course of the hydration-dehydration reaction catalysed by the E. coli P-hydroxydecanoyl thioester dehydratase has been shown to be a syn process with protonation occurring on the si face at C2 of the dec-2-enoyl thioester (Scheme 1).6 The enzyme has been sequenced and the active site identified.7 'H C7H15 'H,O H COSR ' H $1:!3R c7H'5xxH H Scheme 1 There is growing evidence that acyl carrier proteins (ACPs) may play a crucial role in the overall control of both fatty acid and polyketide chain assembly.A three-dimensional structure for the E. coli ACP has been determined by n.m.r. methods.8 This structure shows the presence of a hydrophobic cleft which has been identified as a likely site of acyl chain binding. A heat-stable factor required for the synthesis of fatty acids in the erythromycin-producing organism Saccharopoly-spora erythruea has been purified and identified as a discrete ACP suggesting that the FAS of S.erythruea is a dissociable complex like that of E. coli.' The observation that antibiotic biosynthetic genes are not scattered but occur in closely linked clusters in Streptomycetes has led to great advances in understanding the genetics of polyketide biosynthesis in these organisms. Developments in gene cloning methodology have made it feasible to clone biosynthetic genes from one organism and to transfer the entire region or parts of it into other suitable hosts. This has enabled biosynthetic genes to be characterized and new 'hybrid' antibiotics to be created." The complete gene clusters for the biosynthesis of erythromycin in S. erythraea" and actinorhodin ( 1) an isochromanequinone antibiotic produced by Streptomyces coelicolor,'2 have been cloned as contiguous stretches of DNA.Co-synthetic studies in blocked mutants have identified at least six biosynthetic genes for actinorhodin 5 S. Kauppinen M. Siggaard-Andersen and P. von Wettstein-Knowles Carlsberg Res. Commun 1988 53 357. 6 J. M. Schwab J. B. Klassen and A. Habib J. Chem. SOC.,Chem. Commun. 1986 357. 7 J. E. Cronan W.-B. Li R. Coleman M. Narasimhan D. DeMendoza and J. M. Schwab J. Biol. Chem. 1988 263 4641. 8 T. A. Holak S. K. Kearsley Y. Kim and J. H. Prestegard Biochemistry 1988 27 6135. 9 R. S. Hale K. N. Jordan and P. F. Leadlay FEBS Left. 1987 224 133. 10 D. H. Sherman F. Malpardita M. J. Bibb H. M. Kieser S. E. Hallam J. A. Robinson S. Bergh M. Uhlen T. J. Simpson and D.A. Hopwood in 'Proceedings of the VlII International Congress of Biotechnology Paris' eds. G. Durand L. Bobichon and J. Florent SocietC Francais de Microbiologie 1988 p. 123. I1 P. Matsushima R. Stanzak R. N. Rao and R. H. Baltz Biotechnology 1986 4 229. 12 F. Malpardita and D. A. Hopwood Mol. Gen. Genet. 1986 205 60. Biological Chemistry- Part (ii) Biosynthesis (1) (actI ~c~III-VII).'~ Recent evidence indicates that the actI actIII and actVII genes code for the necessary condensation reduction and dehydration activities associated with the polyketide ~ynthase.'~ The act111 gene has been sequenced and does indeed show homology with other oxido-red~ctases.'~ The act1 and act111 genes have been used to probe restriction digests of genomic DNA from a large number of other Streptomycetes.16 Many of these showed the presence of DNA homologous to both act1 and act111.Interestingly the tetracenomycin (2) producer Streptomyces glaucescens contains DNA homologous to act I but not act111 con- sistent with the fact that tetracenomycin biosynthesis does not require a reduction step during polyketide chain assembly. These and many other aspects of genetic studies on polyketide assembly have been re~iewed.'~~'' Results over a number of years mainly from stable-isotope labelling studies," have indicated that necessary reductions and loss of oxygen from polyketide inter- mediates generally (but see below) occur concomitant with chain elongation as in fatty acid biosynthesis. Further support has come from the successful incorporation of partially elaborated chain elongation intermediates into tylosin (3) as indicated in Scheme 2.19 Success was achieved by feeding the intermediates as their N-acetyl cysteamine thioesters.When fed as their ethyl oxyesters incorporation of label occurred only after prior degradation. Similar results have been reported for eryth- romycin,20 nargenicin,21 and nonactin.22 Until recently there had been no reports l3 H. G. Floss S. P. Cole X.-G. He B. A. M. Rudd J. Duncan I. Fujii C. J. Chang and P. J. Keller in 'Regulation of Secondary Metabolite Formation' ed. H. Kleinkauf H. von Dohrern H. Dornauer and G. Nesemann VCH Meinheim 1986 pp. 283-304. 14 S. P. Cole B. A. M. Rudd D. A. Hopwood C. J. Chang and H. G.Floss J. Antibiot. 1987 40 340. 15 S. E. Hallam F. Malpardita and D. A. Hopwood Gene 1988 74 305. 16 F. Malpardita S. G. Hallam H. M. Kieser H. Motarnedi C. R. Hutchinson M. J. Butler D. A. Sugden M. Warren C. McKillop C. R. Bailey G. 0.Humphreys and D. A. Hopwood Nature (London) 1987 325 818. C. Chen in 'Antimicrobial Chemotherapy-Prospects for the Future' ed. S. Harnmon Wiley Chichester 1988. " T. J. Simpson Chem. SOC.Rev. 1987 16 123. 19 S. Yue J. S. Duncan Y. Yamamoto and C. R. Hutchinson J. Am. Chem. SOC., 1987 109 1253. 20 D. E. Cane and C. C. Yang J. Am. Chem. SOC.,1987 109 1255 *' D. E. Cane and W. R. Ott J. Am. Chem. SOC.,1988 110 4840. 22 J. A. Robinson and Z. Spavold J. Chem. SOC.,Chem. Commun. 1988 4. 324 T.J. Simpson 00 I Me / ;I Me Scheme 2 of the isolation of partially assembled polyketide intermediates from cultures of producing organisms. However compounds (4)-(6) which are clearly derived from the chain assembly process leading to the mycinamycins e.g. (7) have been isolated from mutants of Micromonospora gri~eorubida.~~ Me Me e H02C OH L MOH O D 0o 0H OH Ho2ve ke Me Me Me Me The biotin-mediated carboxylation of acetyl-CoA to malonyl-CoA is a key step in both fatty acid and polyketide biosynthesis. In contrast to previous proposals it has been shown by making use of the double isotope fractionation test that this is not a concerted process for the conversion of pyruvate into o~aloacetate.~~ The conversion of linoleic acid into the divinyl ether colneleic acid (8) and the a-and y-ketols (9) and (lo) and the analogous conversions of linolenic acid into 12-oxophytodienoic acid (1 1) by lipoxygenases from potato and flax seed have been studied with "0-and 2H-labelled s~bstrates.~~ The intermediacy of the correspond- ing 9-and 13-hydroperoxides and the allene epoxide (12) have been demonstrated.Two groups have reported the synthesis of chiral (R)-and (S)-[l-13C,2-2Hl]malonates.26927 The absolute configurations were confirmed by using them as 23 K. Kinoshita S. Takenaka and M. Hayashi J. Chem. Soc. Chem. Commun. 1988 943. 24 S. J. O'Keefe and J. R. Knowles J. Am. Chem. Soc. 1986 108 329; Biochemistry 1986 25 6077. 25 L. Crombie D. 0. Morgan and E. H. Smith J.Chem. Soc. Chem. Commun. 1987 502; L. Crombie and D. 0. Morgan ibid. 1987 503; 1988 556 558. 26 P. M. Jordan J. B. Spencer and D. L. Corina J. Chem. Soc. Chem. Commun.,1986 911. '' S. Huang J. M. Beale P. H. Keller and H. G. Floss J. Am. Chem. Soc. 1986 108 1100. Biological Chemistry- Part ( ii) Biosynthesis 325 0 Me CO2H OH (9) Me CO2H substrates for yeast FAS and by mass spectral analysis of the derived palmitic acid,26 and by n.m.r. analysis.27 The aflatoxin pathway continues to be of major interest and has been the subject of a number of studies. In contrast to previous results28 which indicated that hexanoyl-CoA acted as a chain primer for polyketide elongation incorporation of 13C-labelled malonate into averufin (13) provided evidence for an acetate starter eff e~t.~~ In order to obtain more information the stereochemistry of incorporation of acetate-derived hydrogen into averufin and the fatty acids of Aspergillus para- sitic~~~' and into cladosporin (14) and oleic acid in Cladosporium ~ladosporoides~~ &yyJ4 aI3 \ Me / HO Me HO 0 HH (13) (14) have been compared by incorporation of [2-13C,2-'H3]acetate and stereochemical analysis by 'H-decoupled 'H,13C heteronuclear shift correlation In clados-porin two acetate-derived 2H labels are incorporated at C13 but only one is retained at C11 and with opposite stereochemistry to that observed in the fatty acids of C.cladosporoides. In A. parasiticus however the single 2H labels retained at C2' and 28 For a review of this and related work see C.A. Townsend fire Appl. Chem. 1986 58 227. 29 I. M. Chandler and T. J. Simpson J. Chem. Soc. Chem. Comrnun. 1987 17. 30 C. A. Townsend S. W. Brobst S.E. Ramer and J. C. Vederas J. Am. Chem. Soc. 1988 110 318. 31 P. B. Reese B. J. Rawlings S. E. Ramer and J. C. Vederas J. Am. Chem. Soc. 1988 110 316. 32 P. B. Reese L. A. Trimble and J. C. Vederas Can. J. Chem. 1986 64,1427. 326 T. J. Simpson C4’ of averufin show the same absolute stereochemistry of incorporation as the fatty acids. This is taken to be consistent with the proposal of a hexanoyl-CoA primer derived from fatty acid metabolism. The next intermediate in the conversion of averufin into aflatoxin has been shown to be 1’-hydroxyversicolorone (15).This has been isolated from a new mutant strain of A. p~rasiticus~~ and its incorporation in labelled form into aflatoxin B1demon-~trated.~~ A later step in aflatoxin biosynthesis is the conversion of the anthraquinone versicolorin A into the xanthone sterigmatocystin. This necessitates the post-aromatic loss of a phenolic oxygen from versicolorin A. Precedent for this has now been established by the dem~nstration~~ that a cell-free system from Pyrenochaeta terrestris catalyses the NADPH-dependent reductive removal of oxygen from the 6-position of emodin (16) to give chrysosphanol (17). (16) R = OH (17) R = H A facile synthesis of isotopically labelled anthraquinones has been reported36 and used to prepare [methyl-2H3]chrysophanol whose specific incorporation into tajixanthone (18) in Aspergillus uariecolor was demonstrated by 2H n.m.r.In contrast it has been shown37 that the xanthone skeleton of mangostin (19) is derived from rn-hydroxybenzoate and three malonates via the benzophenone (20). An interesting proposal38 that alternariol (21) may be biosynthesized via ring cleavage and Me v MeYMe Me Me \o / Lo Me Me 33 C. A. Townsend K. A. Plavcan K. Pal S. W. Brobst M. S. Irish E. W. Ely and J. W. Bennett J. Org.. Chem. 1988 53 2472. 34 C. A. Townsend P. R. 0. Whittamore and S. W. Brobst J. Chem. SOC.,Chem. Commun. 1988 736. 35 J. A. Anderson B.-K. Lin H. J. Williams and A. I. Scott J. Am. Chem. SOC.,1988 110 5899. 36 S. A. Ahmed E. Bardshiri and T.J. Simpson J. Chem. SOC.,Chem. Commun. 1987 883. 37 G. J. Bennett and H.-H. Lee J. Chem. Soc. Chem. Commun. 1988 619. 38 E. E. Stinson W. B. Wise R. A. Moreau A. J. Jurewicz and P. E. Pfeffer Can.J. Chem. 1986,64 1590. Biological Chemistry- Part ( ii) Biosynthesis 327 OH rearrangement of lichexanthone has been shown39 not to be tenable by incorporation studies with [ l-'3C,'s02]acetate. In other work making use of oxygen labelling it has been shown by "0 n.m.r. that hydroxymellein (22) arises from direct benzylic hydroxylation of mellein follow- ing incorporation of "0-labelled acetate and oxygen and from direct conversion of 2H-labelled mellein.40 ''0 labelling studies have established the mechanism of formation of multicolosic acid (23) via cleavage of an aromatic prec~rsor.~' The HOzCFc07H OH 0 origins of the oxygen atoms in the phytotoxin betaenone B were e~tablished~~ by feeding [ l-'3C,'802]acetate and by treatment of cultures of Phoma betae with the cytochrome P-450 inhibitor ancymidol.As a result the intermediate (24) accumu-lated which suggests a biosynthetic pathway involving an intramolecular Diels- Alder cyclization at a late stage (Scheme 3). ' Me Me MeCazNa -+ H H OH Me Me Scheme 3 39 J. Dasenbrock and T. J. Simpson J. Chem. SOC.,Chem. Cornmun. 1087 1235. 40 C. Abell A. C. Sutkowski and J. Staunton J. Chem. SOC.,Chem. Cornmun. 1987 586. 41 J. S. E. Holker M. Kaneda S. E. Ramer and J. C. Vederas J. Chem. SOC.,Chem. Cornmun.,1987 1099. 42 H.Ockawa A. Ichichara and S. Sakamura J. Chem. SOC. Chem. Cornmun.,1988 600. 328 T. J. Simpson The origins of the oxygen atoms in the polyether antibiotics nara~in,~~ maduramy-cin,& and len~remycin~~ have been established. The most important of the polyethers is monesin whose biosynthesis is believed to proceed uia an epoxide-mediated cyclization cascade from a putative polyene precursor. A fully asymmetric synthesis of this has been described.46 The chemical feasibility of these cascades has been demonstrated by two group^^',^' who synthesized isomeric epoxides [e.g. (25)] and Me Me 0 Me 0 (25) demonstrated their rapid stereospecific cyclization to polyether structures upon saponification followed by acidification. The stereospecific synthesis of the diepoxide (26)has also been described.49 On treatment with pig-liver esterase this was converted in >70% yield into the cyclized product.However monitoring by n.m.r. showed that this was a stepwise process. A monocyclic lactone was formed within 1 h but further conversion into (27) occurred over a 24 h period (Scheme 4). This contrasts with the rapid cascade observed in aqueous solutions at low pH and has implications for the enzyme-catalysed process. -MeO2C OH 1 oQ-Q--coH Me OH Me Scheme 4 There has been much speculation concerning the biogenetic basis of the extensive structural and stereochemical homologies observed within the polyether and macrolide classes of antibiotics. These ideas have recently been extended to show that these homologies can transcend both classes.50 43 Z.Spavold J. A. Robinson and D. L. Turner Tetrahedron Lett. 1986 27 3299. 44 H. R. Tson S. Rajan T. T. Chang R. R. Fiala G. W. Stickton and M. W. Bullock J. Antibiot. 1987 40 94. 45 D. E. Cane and B. R. Hubbard J. Am. Chem. SOC.,1987 109 6533. 46 D. A. Evans and M. DiMare J. Am. Chem. Soc. 1986 108 2476. 47 W. C. Still and A. G. Romero J. Am. Chem. SOC.,1986 108 2105. 48 S. L. Schrieber T. Sammakia B. Huh and G. Schulte J. Am. Chem. SOC.,1986 108 2106. 49 S. J. Russell J. A. Robinson and D. J. Williams J. Chem. SOC.,Chem. Commun. 1987 351. SO D. O'Hagan Tetrahedron 1988 44 1691. Biological Chemistry- Part ( ii) Biosynthesis 329 Careful 13C labelling studies” have shown that the kinamycin antibiotics [e.g.(28)] must be derived via the benz[a]anthraquinone (29). One of the two carbons from the acetate-derived unit removed in this process is retained as the cyanamide carbon.52 &Me @ HO / \ Me \ -; ,;; __c /I I \ N bAc I \. OH 0 OH OH CN (29) (28) The origins of the carbon skeleton of the aurantinins (30) novel carbocyclic metabolites of Bacillus aurantinus have been determined by feeding 13C-labelled prec~rsors.’~ They contain five C-methyl groups derived from methionine two from the methyl of cleaved acetate units and a carboxyl derived from a cleaved acetate unit as indicated. The carbon skeleton of the indolizidine alkaloid cyclizidine (3 1) is derived entirely from acetate and propionate units with the cyclopropyl ring being derived from a single propionate unit.s4 Biosynthetic studies on marine polyketide-derived metabolites are rare.However [1-‘4C]propionate has been suc- cessfully incorporated into the dendiculatins (32) produced by the marine pulmonate Siphonaria dendic~lata.~~ Me Me w he -C0,Na I OH MeCH,CO,Na (32) methionine A P. J. Seaton and S. J. Could J. Am. Chem. SOC.,1987 109 5282. ’*P. J. Seaton and S. J. Could J. Am. Chem. SOC.,1988 110 5912. 53 A. Nakagawa Y. Konda A. Hatano Y. Hirigaya M. Onda and S. Omura J. Org. Chem. 1988,53,2660. 54 F. J. Leeper P. Padmanbhan G. W. Kirby and G. N. Sheldrake J. Chem. SOC.,Chem. Commun. 1987 505. 55 D. C. Manker M. J.Garson and D. J. Faulkner J. Chem. SOC.,Chem. Commun. 1988 1061. 330 T. J. Simpson 13 C-labelled acetates and methionine have also been incorporated into brevitoxin B,56,57and the complex results obtained were interpreted in terms of incorpora- tion of label via succinate 2-oxoglutarate propionate and 3-hydroxy-3-methyl- g~utarate.~’ 3 Terpenoids The role of leucine in mevalonate biosynthesis has been studied58 by incorporation of l3C-labe1led leucines into the sesquiterpenoid paniculide A in tissue cultures of Andrographis paniculata. The observed labelling pattern could be rationalized by the breakdown of leucine via HMG-CoA to acetyl-CoA and acetoacetate which are then incorporated into mevalonate via the anabolic HMG-CoA pathway. A number of papers have been published which provide further details on the molecular biology of HMG-CoA syntha~e~~ and HMG-CoA reductase.60 The purification of mevalonate Sdiphosphate decarboxylase and its inhibition by fluoromevalonate analogues have been described.61,62 Isopentenyl diphosphate (IPP) isomerase has been purified to homogeneity from fruits of Capsicum ann~um.~~ The preparation of phosphonylphosphinyl analogues of IPP and dimethylallyl diphosphate (DMAPP)64 and farnesyl diphosphate (FPP)65 and their interactions with prenyl- transferases and monoterpene cyclases has been described.66 The biosynthesis of cyclic monoterpenes has been the subject of an important review.67 The cyclization of geranyl diphosphate to the enantiomeric bornyl diphos- phates (33) has been studied68 using purified enzyme preparations from Salvia oficinalis and Tunacetum vulgare.Whereas the cyclase from T. vulgare accepts only (+)-(3S)-linalyl diphosphate to give only (-)-(33) the cyclase from S. oficinalis can cyclize both enantiomers of linalyl diphosphate. Further mechanistic on the pinene cyclases I and I1 isolated from S. oficinalis have also been published. 56 M. S. Lee D. J. Repeta K. Nakanishi and M. G. Zagorski J. Am. Chem. Soc. 1986 108 7855. 57 H.-N. Chou and Y. Shimizu J. Am. Chem. Soc. 1987 109 2184. 58 P. Anastasis I. Freer K. H. Overton D. Pickin D. S. Rycroft and S. B. Singh J. Chem. Soc. Perkin Trans. I 1987 2427. 59 G. Gil J. R. Smith J. L. Goldstein and M. S. Brown Proc. Natl. Acad.Sci. USA 1987 84 1863. 60 T. F. Osborne G. Gil M. S. Brown R. C. Kowal and J. L. Goldstein Roc. Natl. Acad. Sci. USA 1987 84 3614. 61 k.E. Chiew W. J. O’Sullivan and C. S. Lee Biochim. Biophys. Acta 1987 916 271. 62 J. E. Reardon and R. H. Abeles Biochemistry 1987 26 4717. 63 0. Dogbo and B. Camara Biochim. Biophys. Acta 1987 920 140. 64 R. W. McClard T. S. Jujita K. E. Stremler and C. D. Poulter J. Am. Chem. Soc. 1987 109 5542. 65 K. E. Stremler and C. D. Poulter J. Am. Chem. Soc. 1987 109 5542. 66 T. Gotoh T. Koyama and K. Ogura Chem. Lett. 1987 1627. 67 R. Croteau Chem. Rev. 1987 87 929. 68 R. Croteau D. M. Satterwhite D. E. Cane and C. C. Chang J. Biol. Chem. 1986 261 1348. 69 R. B. Croteau C. J. Wheeler D. E. Cane R. Ebert and H.-J.Ha Biochemistry 1987 26 5383. Biological Chemistry- Part (ii) Biosynthesis 33 1 Variations in natural abundance 2H levels have been used as a probe for biosyn- thetic mechanisms in monoterpene biosynthesis. 2H n.m.r. analysis of a-and p-pinenes suggested that isotopically sensitive partitioning within the pinene cyclase enzyme and a correlation between optical purity and site-specific deuterium-protium ratios has been observed for a-~inene.~' Natural (+)-limonene shows relative 2H depletion of 25% for the isopropenyl methyl hydrogens relative to the 7-methyl. This is attributed72 to the IPP-DMAPP isomerization. The isopropenyl methylene hydrogens are enhanced to a relative value of 2.61 (cJ the statistical value of 2 in the absence of any kinetic isotope effect).All of these data are consistent with a loss of a proton from C9 of the a-terpinyl cation during limonene biosynthesis (Scheme 5). I Me (3.0) Me *Me I I I + :-'Me Me. 'Me ( ) relative *H levels Scheme 5 The first step in iridoid biosynthesis is hydroxylation of geraniol to 8-hydroxy- geraniol(34). A detailed study of the regioselectivity and 2Hisotope effects observed in the hydroxylation of geraniol by a cytochrome P-450 mono-oxygenase from Catharanthus roseus has been reported.73 The further conversion of (34) into loganin has been studied74 in a cell-free-system from suspension cultures of Rauwo&a serpentina. Both monoaldehydes (35) and (36) have been isolated and both are converted into iridiodial.This is then cyclized to loganin via (37) rather than via iridiotrial. The cyclization of farnesyl idphosphate (FPP) to trichodiene (38) has been studied using an enzyme preparation from Trichotheciurn roseurn. It has been shown75 that (3R)-nerolidyl diphosphate is a true intermediate between FPP and trichodiene and that it cyclizes via an anti-boat conformation. The enzyme trichodiene cyclase has been purified to >95 YO homogeneity from Fusarium ~pororrichoides.~~ The same 70 R. A. Pascal M. W. Baum C. K. Wagner L. R. Rogers and D. Huang J. Am. Chem. Soc. 1986 108 7074. 71 G. J. Martin P. Janvier S. Akoka F. Mabon and J. Jurezak Tetrahedron Lett. 1986 27 2855. 72 M. F. Leopold W. W. Epstein and D. M. Grant J. Am. Chem. Soc. 1988 110 616.73 H. Fretz and W.-D. Woggon Helu. Chirn. Acta 1986 69 1959. 74 S. Uesato H. Ikeda T. Fujita H. Inouye and M. H. Zenk Tetrahedron Left. 1987 28 4431. 75 D. E. Cane and H.-J. Ha J. Am. Chem. Soc. 1986 108 3097; 1988 110 6865. 76 T. H. Hohn and F. van Middlesworth Arch. Biochem. Biophys. 1986 251 756. 332 T J. Simpson (34) R' = R2 = CH20H (35) R' = CHO R2 = CH,OH (36) R' = CH,OH R2 = CHO workers have shown77 that labels the oxygens indicated (a)in the conversion of trichodiene into T2-toxin (39). Treatment of cultures of Gibberella pulicaris with ancymidol suppressed biosynthesis of diacetoxyscirpenol (40) with concomitant accumulation of major quantities of trichodiene." Other significant work in the trichothecin area includes the preparation of u.v.-induced mutants of F.sporotrichoides which are blocked in T2-toxin production and accumulate diacetoxy- scirpen01,~~ and the incorporation of 2H- and 13C-labelled mevalonates into 3-acetyldeoxynivalenol (41) in Fusarium culmorum.80 A novel application of kinetic pulse labelling led to the identification of transient intermediates to (41) and sambucinol and to a new dead-end shunt metabolite.81 (39) R = OCOCH2Pri (40) R = H A number of papers dealing with the enzymatic cyclization of FPP to pentalene (42) have been published. Previous studies have shown that the first step in the cyclization of FPP to pentalene via humulene appears to be the intramolecular analogue of the prenyltransferase reaction.Incorporation of (R)-and (S)-[1-2H]FPP and 2H n.m.r. analysis show that inversion occurs at C1 of FPP during this cycliz- ation.'* (9R)- and (9S)-[9-3H,4,8-14C]FPP were prepared enzymatically from (1R)-and (1s)-[ 1 -3H]DMAPP and [4-14C]IPP. These were incubated83 with both crude and purified pentalene ~ynthetase,~~ and analysis of the resultant pentalene showed that during cyclization H9re of FPP becomes H8 of pentalene while H9 si undergoes 77 A. E. Desjardins R. D. Plattner and F. van Middlesworth Appl. Enoiron. Microbiol. 1986 51 493. " F. van Middlesworth A. E. Desjardins S. L. Taylor and R. D. Plattner J. Chem. SOC.,Chem. Commun. 1986 1156. '9 M. N. Beremand Appl. Environ. Microbiol. 1987 53 1855. 80 L. 0. Zamir K. A. Devor Y.Nadeau and F. Sauriol J. Biol. Chem. 1987 262 15 354. 81 L. 0. Zamir and K. A. Devor J. Biol. Chem. 1987 262 15 348. P. H. Harrison J. S. Oliver and D. E. Cane J. Am. Chem. SOC. 1988 110 5922. 83 D. E. Cane C. Abell R. Lattman C. T. Kane B. R. Hubbard and P. H. Harrison J. Am. Chern. Soc. 1988 110 4081. 84 D. E. Cane and C. Pargellis Arch. Biochem. Biophys. 1987 254 421. Biological Chemistry- Part ( ii) Biosynthesis 333 a net intramolecular transfer to Hla as indicated in Scheme 6. Since cyclization has already been shown to involve electrophilic attack on the si face of the C 10-C 11 double bond of FPP the formal SEr reaction takes place with net anti stereochemistry. CB-Enz cB-Enz I r= / I Scheme 6 ''C-labelled allicide (43) is incorporated into both alliacolide (44) and 12-hydroxyalliacolide (45) in cultures of Marasmius allia~eus.~~ Thus the 4-hydroxy group is introduced by direct hydroxylation rather than ria a 4,Sepoxide.A number of other papers of note describe the incorporation of I3C-labelled acetates into the fungal sesquiterpenoids quadrone (46),86 fomajorins (47),87 and 7,12-dihydroxyster- purene (48) .88 OH 0 R' (47) R = Me or CO,H RZ R1' 0& '0 Me 0 (43) R' = R2 = H (44) R' = OH R2 = H (45) R' = R2 = OH I 85 A. G. Avent J. R. Hanson and B. L. Yeoh .I.Chem. Res. (S) 1986 422. 86 D. E. Cane Y. G. Whittle and T. C. Liang Bioorg. Chem. 1986 14 417. 87 D. M. X. Donnelly J. O'Reilly J.Polonsky and M. H. Sheridan J. Chem. SOC.,Perkin Trans. 1 1987 1869. 88 C. Abell and A. P. Leech Tetrahedron Len. 1987 28 4887. 334 T. J. Simpson The stereochemistry of the formation of the exocyclic methylene function in ent- kaurene (49) has been examinedg9 using mevalonate containing a chirally label- led methyl group. Incorporation into (49) using a cell-free preparation from Marah macrocarpus indicated that the methyl to methylene elimination had occurred with the endo orientation indicated in Scheme 7. The mechanism usually proposed for the formation of the cation (51) via the pimarenyl cation (50) is in contravention of Baldwin’s rules. An alternative route via the cyclobutyl cation (52) has been examined.” However no incorporation of the appropriate cyclobutyl intermediates was observed.Model studies” have provided some support for the formally dis- allowed 5-endo-trig cyclization of (50) to (51). Enz-B 7 Scheme 7 A number of monoclonal antibodies which recognize specific features of the gibberellin molecule have been prepared9’ from antigenic gibberellin-protein com- plexes that are linked through C3 or C17. The same chemistry has been used to prepare gibberellins attached to photoaffinity labels and other probes for studying gibberellin biosynthesis and mode of action.93 The purification of a number of the enzymes on the gibberellin pathway have been described; these include a GA1-2P-hydroxyla~e~~ and Gb-~xidase.~’ 89 R. M. Coates S. C. Koch and S. Hedge J. Am. Chem.Sac. 1986 108 2762. 90 R. M. Coates and H.-Y. Yang J. Org. Chem. 1987 52 2065. 91 R. M. Coates and H.-Y. Yang 1. Chem. SOC.,Chem. Commun. 1987 232. 92 J. P. Knox M. H. Beale G. W. Butcher and J. MacMilIan PZunta 1987 167 9. 93 M. H. Beale R. Holley and J. MacMiIlan in ‘Plant Growth Substances 1985’ ed. M. Bopp Springer- Verlag Heidelberg 1986 p. 65. 94 V. A. Smith and J. MacMiIlan plant^ 1986 167 9. 95 S. J. Gilmour A. B. Bleecher and J. A. D. Zevaart Plant Physiol 1987 85 87. Biological Chemistry- Part ( ii) Biosynthesis 335 A novel antheridiogen has been isolated from Anemia rnexi~ana.~~ The structure (53) which was proposed on biogenetic grounds has been confirmed by total synthesis.97 The ;hj?-hydroxy group in aphidicolin (54) originates in water whereas the 17-hydroxy group is introduced from the atm~sphere.~~ -OH Incorporations of [5-'3C,5-2H2]mevalonate and [2-'3C,2H3]acetate into ursolic acid (55) and oleanolic acid (56) in tissue cultures of Rabdosia japonica confirm that the C12-Cl3 double bond of (55) is introduced by elimination of the 12-pro-R- hydrogen in a cis fashion and also confirm the occurrence of predicted 1,2-hydride shifts in the biosynthesis of (55) and (56).99 Me & Me Me 0,H Ho Me ,Me A number of papers on meroterpenoid biosynthesis have been published.Incor- porations of '3C,2H-labelled mevalonates into viridicatumtoxin (57)indicate that a 1,3-hydride shift from C15 to C19 occurs during the formation of the spirobicyclic ring system.'oo The biosynthesis of the austalides has been studied"' using I3C- and *H-labelled acetates and mevalonates and the synthesis and incorporation of 96 M.Furber L. N. Mander J. E. Nester N. Takahashi and H. Yamane Phytochemistry 1989 28 63. 97 M. Furber and L. N. Mander J. Am. Chem. SOC.,1988 110 4084. 98 M. J. Ackland J. F. Gordon J. R. Hanson B. L. Yeoh and A. H. Ratcliffe J. Chem. SOC.,Chem. Commun. 1987 1492. 99 S. Seo Y. Yoshimura A. Uomari K. Takeda H. Seto Y. Ebijuka and U. Sankawa J. Am. Chem. SOC. 1988 110 1740. 100 R. M. Horak V. J. Maharaj S. F. Marais F. R. van Heerden and R. Vleggaar J. Chem. Soc. Chem. Commun. 1988 1562. 101 A. E. de Jesus R. M. Horak P. S. Steyn and R. Vleggaar 1.Chem. SOC. Perkin Trans. 1 1987 2253.nMe 7'.J. Simpson OH CONH2 nu '3C,'80-labelled 3,5-dimethylorsellinate into austinlo2 and andilesin A'03 has been reported. 4 Shikimate and Related Metabolites The shikimate pathway continues to be of considerable interest not least because it offers an attractive target for selective antimicrobial and herbicidal agents. Progress has been greatly facilitated by the increased availability of the enzymes on the pathway by gene cloning to construct overproducing strains of E. coli. In this way the complete amino acid sequence of 3-dehydroquinate synthetase has been eluci- dated,Io4 and shikimate kinase which catalyses the conversion of shikimate into shikimate 3-phosphate has been cloned and purified.lo5 In E. coli the enzymes of the pathway are all monofunctional.However in S. cerevisiae one pentafunctional enzyme is found which is encoded by the arom gene. The nucleotide sequence of the arom gene has been determined and functional regions within the derived polypeptide have been compared with the corresponding enzymes from E. coIi.'06 This demonstrated that the arom polypeptide contains monofunctional domains and supports the theory that the S. cerevisiae evolved by gene fusion from monofunc- tional bacterial genes. 3-Dehydroquinate synthetase catalyses an apparently complex sequence of reac- tions which converts DAHP (58) into 3-dehydroquinate (59). Synthetic analogues of DAHP containing fluorine at C3 were converted by the enzyme into 6-fluoro derivatives of (59) and subsequently 3-dehydroshikimate and shikimate.lo7 The 2-deoxy analogue of (58) has been converted into the enol ether (60).The stereochemistry of the p-elimination of phosphate has been established'08 by stereo- specific labelling of (58) with 2H at C7. '% ' W H -02c&" -0zc OH o@ -02c OH 0 (58) (59) (60) F. E. Scott T. J. Simpson L. A. Trimble and J. C. Vederas J. Chem. Soc. Chem. Commun. 1986 214. 103 C. R. McIntyre F. E. Scott T. J. Simpson L. A. Trimble and J. C. Vederas J. Chem. Soc. Chem. Commun. 1986 501. I04 G. Miller and J. R. Coggins FEBS Lett. 1986 200 11. G. Miller A. Lewendon M. G. Hunter and J. R. Coggins Biochem. J. 1986 237 427. 106 K. Duncan R. M. Edwards and J. R. Coggins Biochem. J. 1987 246 375. 107 P. Le Marechal C.Froussios and R. Azerad Biochemie 1986 68 1211. 108 T. S. Widlanski S. L. Bender and J. R. Knowles J. Am. Chem. Soc. 1987 109 1873. Biological Chemistry- Part ( ii) Biosynthesis 337 The conversion of shikimate 3-phosphate (61) into EPSP (62) catalysed by EPSP synthase (the target for the herbicide glyphosphate) is thought to proceed via an addition-elimination mechanism involving nucleophilic attack of the 5-hydroxy group of (61) on C2 of PEP. The presence of an acid-labile intermediate was demon~trated'~~ and this has now been isolated by denaturing the enzyme under mild basic conditions followed by ion-exchange h.p.1.c. The intermediate has been shown by detailed n.m.r. analysis to correspond to the key tetrahedral intermediate (63).'lo @o"fiOH OH -~o..~o~co~-@o..~o~~o~ : OH o@ OH (61) (63) (62) 'H n.m.r.studies have shown that aqueous solutions of chorismic acid contain 10-20°/0 of the diaxial conformer (64) whereas only the diequatorial conformer (65) is observed in methanol."' The thermal Claisen rearrangement of chorismate to prephenate (66) is considerably slower in methanol than in aqueous solution. -02c< 0 0 c 'd H OH -0zc OH (64) (65) (66) The enzyme-catalysed process has been studied using secondary 3H isotope effects with chorismate labelled at C4. These allow mechanisms involving the 4-hydroxy group to be eliminated and an alternative to be proposed"* in which an enzyme- bound nucleophile stabilizes developing charge at C5.The thermal Claisen rear- rangement has also been studied using synthetic analogues of chorismate and by observing solvent and isotope effects. These indicate a transition state involving substantial C-0 bond cleavage but little formation of the C-C bond.'13 So-called C7N units which are believed to be derived from the shikimate pathway but not via shikimate have been implicated in the biosynthesis of many antibiotics. 4-Amino- and 3-hydroxyanthranilate have been shown .to be intermediates in the biosynthesis of ~treptonigrin"~ and LL-C10037a.''5 In contrast the C,N unit found I09 K. S. Anderson J. A. Sikorski and K. A. Johnson Biochemistry 1988 27 1604 7395. 110 K. S. Anderson J. A. Sikorski A. J. Benesi and K. A. Johson J. Am. Chem. SOC.,1988 110 6577.111 S. D. Copley and J. R. Knowles J. Am. Chem. Soc. 1987 109 5008. 112 W. J. Guilford S. D. Copley and J. R. Knowles J. Am. Chem. Soc. 1987 109 5013. 113 J. J. Gajewski J. Jurayi D. R. Kimbrough M. E. Gande B. Ganem and B. K. Carpenter J. Am. Chem. SOC.,1987 109 1170. 114 W. R. Erickson and S. J. Gould J. Am. Chem. SOC.,1987 109 620. 115 Y. G.Whittle and S. J. Gould J. Am. Chem. Soc. 1987 109 5043. 338 T. J. Simpson in asukamycin and manumycin appears to be derived via intermediates from the TCA cycle and triose phosphate p001."~ [l-13C]shikimate has been incorporated into the cyclohexanecarboxylic acid moiety but not the C7N unit of the ansatrienin antibiotics; no label from [6-2H]shikimate was retained however.' l7 Feeding experiments in young shoots of Acer nikoense have shown that phenyl- alanine and cinnamic acid are incorporated into the diarylheptanoid acerogenin A (67).The central carbon (0) is derived from C2 of acetate."' The synthase which catalyses the biosynthesis of the stilbene resveratrol (68) in Arachis hypogeae 'has been cloned and sequenced."' It shows a remarkable degree of homology with the consensus sequence of chalcone synthases from various sources. The synthesis of 6'-deoxychalcones has been shown to require the combined presence of chalcone synthase and an NADH-dependent reductase in cell-free extracts of Glycyrrhiza echinatat2' and Glycine max.121 The reductase from the latter has been purified to homogeneity. 6'-Deoxychalcones are precursors to both iso- flavanoids and rotenoids.The pattern of incorporation of 13C2-acetate into amor- phigenin (69) by seedlings of Amorpha fmticosa confirms'22 that deoxygenation must occur before cyclization of the polyketide-derived ring. An isoflavone synthase was recently isolated from G. max and a mechanism for the rearrangement proposed. However an alternative mechanism involving epoxidation of the aryl ring rather than the heterocyclic ring as originally proposed has been suggested.12* Both mechanisms involve spirodienone intermediates. I16 J. M. Beale R. E. Herrold H. G. Floss H. Nakagawa S. Omura R. Thiericke and A. Zeeck J. Am. Chem. SOC.,1988 110 4435. 117 R. Casati J. M. Beale and H. G. Floss J. Am. Chem. SOC.,1987 109 8102. 118 T.Inouye N. Kenmochi N. Furukawa and M. Fujita Phytochemistry 1987 26 1409. 119 G. Schroder J. W. S. Brown and J. Schroder Eur. J. Biochem. 1988 172 161. 120 S.-I. Ayake A. Udugawa and T. Furuya Arch. Biochem. Biophys. 1988 261 458. 121 R. Welle and H. Grisebach FEBS Lett. 1988 236 221. 122 L. Crombie 1. Holden N. van Bruggen and D. A. Whiting J. Chem. SOC.,Chem. Commun. 1986 1063. Biological Chemistry- Part (ii) Biosynthesis The formation of the chromene rings found in many rotenoids and other phenolics was thought to involve epoxidation of prenyl substituents and subsequent ring closure and dehydration. However although rot-2’-enoic acid (70) was well incorpor- ated into deguelin (71) by seedlings of Tephrosia vogelii the hydroxychroman (72) was not.123 Similar results were obtained with crude and purified enzyme prep- arations.These results make the intermediacy of epoxides unlikely. An alternative mechanism involves cyclization of the o-quinone methide (73). Incorporation of (70) labelled with 13C at C4’ produced deguelin labelled at both methyls but with a strong preference (73%) for C8’,indicating that the electrocyclization is stereoselec- tive but not stereo~pecific.’~~ The major route to amorphigenin (69) is also from rot-2’-enoic acid via stereospecific cyclization to rotenone in which the (E)-methyl of (70) becomes the methylene in rotenone. This is followed by a non-specific hydroxylation which results in a randomization of label between C7’ and C8’ in amorphigenin.However a secondary route appears to involve a positionally non- specific hydroxylation of the prenyl methyls in rot-2’-enoic acid followed by a chemospecific cyclization to (69).125 OH (72) Me” *Me -8‘ H -Mh OMe (71) OMe (69) 123 L. Crornbie J. Rossiter and D. A. Whiting J. Chem. SOC.,Chern. Cornrnun. 1986 352. 124 M. J. Begley L. Crornbie J. Rossiter M. Sanders and D. A. Whiting J. Chern. SOC.,Chem. Cornmun. 1986 353. lZ5 P. Bhandari L. Crornbie and D. A. Whiting J. Chern. SOC.,Chern. Comrnun. 1988 1085. 340 T. J. Simpson 5 Alkaloids and other Amino-acid-derived Metabolites The (S)-tropic acid moiety found in tropane alkaloids is formed from phenylalanine via an intramolecular 1,2-carboxyl migration. It has now been shown that during the concomitant 1,2-hydrogen shift it is the 3-pro-S-hydrogen of phenylalanine which migrates.'26 The generally accepted classical hypothesis for the biosynthesis of the tropane moiety of cocaine (74) needs to be modified as a result of experiments which showed that [2-I4C]-1 -methyl-A'-pyrrolinium chloride did not label C 1 of cocaine when fed to Erythoxylurn coca ~1ants.I~' However [l-'3C,'SN]-4-(methyl-amino)butanal did label the C5-N bond.The pathway shown in Scheme 8 was proposed.'28 Evidence for the intermediacy of (75) was obtained by a trapping experiment using 1-methylpyrrolidine-2-acetic acid. * -4: OEt ,COzMe COSCoA (74) (75) Scheme 8 The iminium ion (76) has been shown'29 to be an intermediate to the pyrrolizidine bases rosmarinecine (77) and retronecine (78).Both are known to be formed from two molecules of putrescine via hom~spermidine.'~~ The two pathways must diverge (76) 126 E. Leete Can. J. Chem. 1987 65 266. 127 E. Leete S. H. Kim and J. Rana Phytochemistry 1988 27 401. 128 E. Leete and S. H. Kim,J. Am. Chem. Soc. 1988 110 2976. 129 H. A. Kelly and D. J. Robins J. Chem. SOC.,Chem. Commun. 1988 329. 130 H. A. Kelly and D. J. Robins J. Chem. SOC.,Perkin Trans. l. 1987 177. Biological Chemistry- Part ( ii) Biosynthesis 341 after (76) because l~-hydroxymethyl-8a-pyrrolizidine is incorporated into ros- whereas the 1a-isomer is a better precursor of retr0ne~ine.l~~ rnarine~ine'~' Late stages in the biosynthesis of the indolizidine alkaloids salframine (79) and swain- sonine (80) have been studied'33 with the aid of specifically deuterated chiral precursors (81) and (82).The results indicate a pathway in which 1-oxoindolizidine acts as a branch point. (81) is converted into salframine via the corresponding 1,6-dihydroxy and 6-0x0 compounds. The route to swainsonine is via the 1,2-diol and the iminium ion (83). The biosynthesis of cyclizidine has been discussed above. Incorporations of variously deuterated cadaverines into various quinolizidine alkaloids have been reported,'34 providing useful information on the biochemical events occurring during the biosynthesis and interconversions of these alkaloids. .OH (82) (83) (80) The biosynthetic origins of all the carbons of ephedrine (84) and related alkaloids have been established finally.'35 Both labels from [2,3-'3C,]pyruvate were incorpor- ated as a unit into C2 and C3 of (85) in Ephedra gerardinia plants.Detailed '3C-labelling studies have shown'36 that the first benzylisoquinoline alkaloid from which the whole family are formed is (S)-norcoclaurine (85) rather than (S)-norlaudanosoline (86). Studies with S-adenosylmethionine containing chiral methyl groups showed that the 6-0-methylation of (87) to give reticuline (88) occurs with inversion of configuration at the meth~1.I~' During the further conversion of reticuline into jatrorhizine (89) the methoxy group at C2 of (89) has the same configuration as that at C6 of (88). The methyl group transfer occurs via the corresponding methylenedioxy bridge in berberine.Late intermediates in the biosynthesis of the ergot alkaloids chanoclavine and elymoclavine (90) have been established. While no incorporation of the previously proposed diol (91) could be observed,'38 the incorporation of *H label from (92) into (90) was detected by mass spectral analysis.'39 In addition an isotope trap 131 E. K. Kunec and D. J. Robins J. Chem. Soc. Chem. Commun. 1986 250. 132 E. Leete and J. Rana J. Nut. Prod. 1986 49 838. 133 C. M. Harris M. J. Schneider F. S. Ungemach J. E. Hill and T.M. Hams J. Am. Chem. SOC.,1988 110 940. 134 D. J. Robins and G. N. Sheldrake J. Chem Res. (S) 1987 159; 1988 230; T. Hemscheidt and I. D. Spenser Can. J. Chem. 1987 65 170.135 G. Grue-Sorensen and I. D. Spenser J. Am. Chem. SOC. 1988 110 3714. 136 S. Loeffler R. Stadler N. Nagakura and M. H. Zenk J. Chem. Soc. Chem. Commun. 1987 1160. 137 M. Kobayashi T. Frenzel J. P. Lee M. H. Zenk and H. G. Floss J. Am. Chem. SOC.,1987 109 6185. 138 A. P. Kozikowski M. Okita M. Kobayashi and H. G. Floss J. Org. Chem. 1988 53 863. 139 A. P. Kozikowski J.-P. Wu M. Shibuya and H. G. Floss J. Am. Chem. Soc.. 1988. 110 1970. 342 T.J. Simpson OH (85) R' = R2= R3 = R4 = H (86) R' = R3= R4 = H,R2= OH (87) R' = Me,R2= OH,R3= R4 = H (88) R' = R3= R4= Me,R2 = OH experiment with (92) anowed the presence of (93) in cultures of Claviceps to be detected. On the basis of these results a pathway was proposed for ring closure via the vinyloxirane (94).The isonitrile group in hapalindole A (95) a metabolite of 'ezH R%H \ H H (91)R = OH (931 (92) R = H 1 OH &$H H the marine cyanophyte Hapalosiphon fontinalis has been shown to be derived from glycine by incorporation of [2-'3C,'5N]glycine.'40 The intermediacy of the 5-formimino group in tetrahydrofolate was suggested which would be consistent with V.Bornemann G. M. L.Patterson and R. E. Moore J. Am. Chern. SOC.,1988 110 2339. Biological Chernistry- Part ( ii) Biosynthesis NC the observed incorporation of one methylene from glycine in the biosynthesis of the formamide moiety in tuberin.141 The isontrile carbon in the marine sponge metabolite diisocyanodociane (96) is labelled by ''C-cyanide.'42 L-[Methyl-13C]methionine labels the isonitrile group of the hazimycins in the bacterium Micrornonosporu ecbinosp~ra,'~~ but the carbon source of the isonitrile group in xanthocillin a metabolite of Penicilliurn notaturn remains obscure despite extensive labelling studies.'44 Biosynthetic work on the naturally occurring isocyanides is discussed in a recent comprehensive review of this varied group of rnetab01ites.l~~ Considerable progress continues to be made in understanding the biosynthesis of the p-lactam antibiotics and much of the important work on penicillins and cephalosporins has been discussed in a recent authoritative review.146 The key sequence in penicillin biosynthesis involves the double cyclization of the LLD-ACV tripeptide to give isopenicillin N catalysed by isopenicillin N synthase (IPNS).On the basis of present knowledge the catalytic cycle is proposed to proceed as shown in Scheme 9 the key feature being the intermediacy of highly reactive ferryl(~v) species such as Enz=Fe=O formed from an initial Fe -O2 complex at the active The mechanistic details of this sequence has been. probed by exhaustive studies14' using cloned IPNS and a huge variety of substrate analogues many of 141 K. M. Cable R. B. Herbert and J. Mann J. Chem. SOC.,Perkin Trans. 1 1987 1593. M. J. Carson J. Chem SOC.,Chem. Commun. 1988 35. 143 M. S. Puar H. Munayyer B. Hedge B. K. Lee and A. J. Wartz J. Antibiot. 1985 38 530. 144 K. M. Cable R. B. Herbert and J.Mann Tetrahedron Lett. 1987 28 3159. 145 M. S. Edenborough and R. B. Herbert Nut. Prod. Rep. 1988 5 229. 146 J. E. Baldwin and E. P. Abraham Nut. Prod. Rep. 1988 5 129. 147 For latest examples see J. E. Baldwin R. M. Adlington L. G. King M. F. Parisi,'W. G. Sobey J. D. Sutherland and H.-H. Ting J. Chem. SOC.,Chem. Commun. 1988 1635; J. E. Baldwin W. J. Norris R. T. Freeman M. Bradley R. M. Adlington S. Long-Fox and C. J. Schofield ibid. p. 1128; J. E. Baldwin R.M. Adlington M. Bradley W. J. Noms N. J. Turner and A. Yoshida ibid. p. 1125; J. E. Baldwin R. M. Adlington B. P. Domayne-Hayman G. Knight and H.-H. Ting ibid. 1987 1661. 344 T. J. Simpson ACV AA HOzC AA'\H / HOzC HOIC' Scheme 9 which have been converted into novel P-lactam structures.The IPNSs from multi- farious eukaryotic and prokaryotic sources have been cloned to provide a large body of comparative sequence data.14* They all appear to contain two highly conserved cysteine residues the importance of which have been tested by site- directed mutagenesis studies using Cephalosporium acremonium IPNS in which the cysteines are replaced by ~erines.'~~ The conversion of isopenicillin N into cephalo- sporins requires its isomerization to penicillin N followed by ring expansion and hydroxylation. These last two steps are catalysed by the single bifunctional enzyme deacetoxycephalosporin-C synthetase and hydr~xylase.'~' This has been cloned and shown to have significant sequence homology with IPNS.'" Studies with specifically 148 L.G. Carr P. L. Skatrud M. E. Scheetz S. W. Queener and T. D. Ingolia Gene 1986 48 257; B. K. Leskiw Y. Aharonowitz M. Mevarech S. Wolfe L. C. Vining D. W. S. Westlake and S. E. Jensen ibid. 1988 62 187; R. Ramon L. Carramoline C. Patino F. Sanchez and M. A. Penalva ibid. 1987 57 171. 149 S. M. Samson J. L. Chapman R. Belagaje S. W. Queener and R. D. Ingolia Roc. Nat. Acad. Sci. USA 1987 84 5705. 150 J. E. Baldwin E. P. Abraham R. M. Adlington J. D. Coates M. J. C. Crabbe N. P. Crouch J. W. Keeping G. C. Knight C. J. Schofield M. Thornally H.-H. Ting C. A. Vallejo and T. Wallis Biochem. J. 1987 245 831. 151 S. M. Samson J. E. Dotzlaf M. L. Slisz G. W. Becker R. M. van Frank L. E. Veal N.-K. Yeah J. R. Miller S. W.Queener and T. D. Ingolia Biotechnology 1985 5 1207. Biological Chemistry- Part (ii) Biosynthesis 345 deuterated samples of penicillin N suggest that the ring expansion occurs in two steps and that the required removal of a 2-methyl hydrogen precedes the loss of C3 h~dr0gen.l~~ The proposed mechanism is summarized in Scheme 10. OH Step 1 RHN EyFe=Enz I Enz=Fe=O + pencillin N 0 COZH OH RHN RHN r COzH C02H I Step 2 RHNx$ + Fe=Enz + H20 0 CO2H Scheme 10 Incorporations of chirally labelled samples of glycerol and of ornithine into clavulanic acid (97) have been reported."53 Both the ring and primary alcohol oxygens are enriched'54 by 1802. Two ornithine-containing metabolites clavaminic acid (98) and proclavaminic acid (99) have been isolated155 from a cell-free enzyme preparation from Streptomyces clavuligerus supplemented with a-ketoglutarate and Fe2+.Surprisingly (98) has the opposite stereochemistry at C3 and C5 relative to clavulanic The absolute configuration of proclavaminic acid has been estab- lished by a synthesis from a resolved sample of P-hydroxyornithine and its incorpor- ation into clavaminic acid by a partially purified synthase has been re~0rted.l~~ Both (98) and (99) have been incorporated into clavulanic acid by a broken-cell preparation from S. clavuligerus. N-Acetylglycylclavaminic acid (100) has been isolated from a clavulanic acid negative mutant of S. ~lavuligerus.'~~ [4-2H2,5-13 Clornithine is incorporated into (100) with retention of one 2H at C8 but no 2H I52 J.E. Baldwin R. M. Adlington R. T. Alpin N. P. Crouch C. J. Schofield and H.-H. Ting J. Chem. SOC.,Chem. Commun. 1987 1654. 153 C. A. Townsend and S.3. Mao J. Chem. SOC.,Chem. Commun. 1987 86; C. A. Townsend M.F. Ho and S.S. Mao ibid. 1986 638. I54 C. A. Townsend and W. J. Krol J. Chem. Soc. Chem. Commun. 1988 1235. 155 S. W. Elson K. H. Baggaley J. Gillett S. Holland N. H. Nicholson J. T. Sime and S. R. Woroniecki J. Chem. SOC.,Chem. Commun. 1987 1736. 15' K. H. Baggaley K. H. Nicholson and J. T. Sime J. Chem. Soc. Chem. Commun. 1988 567. 157 S. W. Elson K. H. Baggaley J. Gillett S. Holland N. H. Nicholson J. T. Sime and S. R. Woroniecki J. Chem. SOC., Chem. Commun. 1987 1739. 158 B. W. Bycroft A. Penrose J.Gillett and S. W. Elson J. Chem. SOC.,Chem. Commun. 1988 980. 346 T. J. Simpson incorporation into clavulanic acid is found. These results suggest that a P-keto intermediate (101) may be involved in the conversion of clavaminic acid into clavulanic acid. A similar type of intermediate has been invoked to explain the base-catalysed racemization of clavulanates. (99) (98) R = H (100) R = COCH,NHCOMe 1 H All the hydrogens in the hydroxyethyl side chain of thienamycin are derived from methionine and the stereochemical fate of the methyl group of methionine on incorporation into thienamycin (102) has been studied with chiral methyl-labelled rnethi~nine.'~~ OH [2-'3C,'5N]nocardicin G (103) has been shown to be incorporated intact into nocardicin A (104) whereas the 2'-epimer was degraded to (p-hydroxypheny1)gly- cine prior to incorporation.16' A partially purified cell-free system has been prepared from Nocardia uniformis which produces nocardicin A (104) from nocardicin E (105) and S-adenosylmethionine.'61Whole-cell experiments'62 have shown that the transfer of the 3-amino-3-carboxypropyl moiety from methionine to nocardicin A proceeds with inversion of configuration at C4.159 D. R. Houk K. Kobayashi J. M. Williamson and H. G. Floss 1. Am. Chem. SOC.,1986 108 5365. 160 C. A. Townsend and B. A. Wilson 1. Am. Chem. SOC.,1988 110 3320. 161 B. A. Wilson S. Bantia G. M. Salituro A. McE. Reeve and C. A. Townsend J. Am. Chem. SOC.,1988 110 8238. 162 C. A. Townsend A.McE. Reeve and G. M. Salituro J. Chem. SOC.,Chem. Cornmun. 1988 1579. Biological Chemistry- Part ( ii) Biosynthesis I CO2H (103) R = H X = H &NH2 (104) R = CH2CH2CH(NH2)C0,H X = NOH (105) R = H X = NOH Incorporation of 13C- and "N-labelled precursors demonstrate that the unusual amino acid 5-hydroxy-4-oxonorvaline(HON) is biosynthesized in Streptomyces akiyoshiensis uia condensation of an activated form of aspartate with acetyl- or malonyl-CoA to form a c6 intermediate that is converted into HON by loss of the acetate-derived carboxyl.163 A similar condensation between acetyl-CoA and glutamic semialdehyde has been proposed as the first step in the biosynthesis of carbapenam~.'~~ L-Nitrosuccinate has been identified as an intermediate in the biosynthesis of 3-nitropropanoic acid in Penicillium atro~eneturn.'~~ Both oxygens of the nitro group are derived from the atmosphere.'66 The results suggest the biosynthetic pathway shown in Scheme 11.Scheme 11 The modified uracil moiety in sparsomycin (106) has been sho~n'~' to be derived by extensive modification of tryptophan. The remaining carbons are derived from methionine cysteine and S-methylcysteine. HO \ 163 R. L. White A. C. DeMarco and I. C. C. Smith J. Am. Chem. Soc. 1988 110 8222. 164 B. W. Bycroft C. Maslen S. J. Box A. G. Brown and J. W. Tyler J. Chem. SOC.,Chem. Commun. 1987 1623. 165 R. L. Baxter A. B. Hanley and H. W.-S. Chan J. Chem. Soc. Chem. Commun. 1988 757. 166 R. L. Baxter and S.L. Greenwood J. Chem. Soc. Chem. Commun.,1986 175. 167 R. J. Parry and M. E. Eudy J. Am. Chem. Soc. 1988 110 2316. 348 T. J. Simpson 6 Porphyrins Porphobilinogen (PBG) chirally labelled at C11 has been incorporated into hydroxy- methylbilane by PBG deaminase in order to investigate whether hydroxymethyl- bilane is the true enzymatic intermediate or whether it is formed by trapping of an azafulvene by water. Degradation of the product to glycolic acid showed that the reaction proceeded with retention of configuration a result taken to be inconsistent with the release of a planar intermediate.168 There has been considerable progress in understanding the structure and mode of action of deaminase. F.p.1.c. has been used to purify deaminase from E.~oli'~~ and human erythrocyte^."^ Both genes have been cloned and sequenced and show a considerable degree of homology between the two enzymes.171 This has led to the overexpression of the E. coli enzyme by several groups of workers resulting in 100 to 200 fold overproduction.'72-174 Studies with overproduced enzyme have given the unexpected result that the enzyme group to which the first PBG molecule (to be incorporated into hydroxymethylbilane by deaminase) becomes bound is not an amino acid residue as previously thought but is in fact a novel covalently bound cofactor. 172,173,175 Even more surprising is the demonstration that this cofactor is in fact a dipyrromethane derived from two molecules of PBG! Confirmation of its derivation from PBG has been provided by incorporation of label from [5-14C]ALA,173 [5-13C]ALA,'75 and [1 1-I3C]ALA (ALA = 8-aminolaevulinic The 13C n.m.r.spectrum of the [5-13C]ALA-enriched deaminase shows a methylene carbon with a chemical shift in the range expected for an a-methylpyrrole suggesting that the cofactor is attached to one of the four cysteine residues in the enzyme. Previous result^"^ from 3H n.m.r. studies with 3H-labelled PBG which had suggested that the initial site of attachment of the initial PBG to be incorporated into hydroxy- methylbilane was a cysteine residue have been reinterpreted as being due to an initial attachment of the enzyme's substrate at a second cysteine residue. The gene for cosynthetase has been found to be adjacent to that for deaminase in E.coli and they appear to be translationally A favoured mechanism for the conversion of hydroxymethylbilane into uro'gen I11 involves the intermediacy of the spiro-pyrrolenine (107). Strong evidence for this has come from the synthesis of the analogous spiro-lactam (108) which appears to exist as two non-interconvert- ible conformers one of which is a powerful inhibitor of co~ynthetase.'~~ Studies with model pyrrolenines suggest that the further conversion of (107) into uro'gen 16' J.-R. Schauder S. Jendrezejewski C. Abell G. J. Hart and A. R. Battersby J. Chem. SOC.,Chem. Commun. 1987 436. 169 G. H. Hart C. Abell and A. R. Battersby Biochem. J. 1986 240 273. 170 R. W. M. De Rooij C. M. Hamer and J. H. P. Wilson Clin. Chim. Acta 1987 162 61.17' N.Raich P.H. Romeo A. Dubert D. Beaupain N. Cohen-Solal and M. Goosens Nucieic Acids Res. 1986 14 5955; S. D. Thomas and P. M. Jordan ibid. p. 6215. 172 G. H. Hart A. D. Miller F. J. Leeper and A. R. Battersby J. Chem. SOC. Chem. Commun 1987 1762. 173 P. M. Jordan and M. J. Warren FEBS Lett. 1987 225 879. 174 A. I. Scott T. 0. Baldwin M. Treat C. A. Roessner S. K. Grant N. J. Stolowich and H. J. Williams Roc. Nut. Acad. Sci. USA in press. 175 A. I. Scott N. J. Stolowich H. J. Williams M. D. Gonzales C. A. Roessner S. I. K. Grant and C. Pichon J. Am. Chem. SOC.,1988 110 5898. 1 76 J. N. S. Evans G. Burton P. E. Fagerness N. E. MacKenzie and A. I. Scott Biochemistry 1986,25,892. 177 A. Saserman A. Nepveau Y. Echelard J. Dymetryszyn M. Drolet and C.Goyer J. Bacteriol 1987 169 4257; P. M. Jordan B. I. A. Mgbeje S. D. Thomas and A. F. Alwan Biochem. J. 1988 249,613. 178 W. M. Stark G. J. Hart and A. R. Battersby J. Chem. SOC. Chem. Commun. 1986 465. Biological Chemistry- Part ( ii) Biosynthesis HO2C HO2C H02C C02H C02H (107)X-Y = -CH=N-0 II (108) X-Y = -C-NH-111 proceeds via a fragmentation-recombination mechanism rather than a series of [1,5]-sigmatropic rearrangements.’ 79 Full details have been published of earlier work which established the sequence of methylations in the biosynthesis of vitamin B12. This paper proposes a new nomenclature system of BI2biosynthetic In this system the reduced forms of Factors I 11 and I11 become ‘precorrins’ 1 2 and 3 in which the number indicates the total number of methyl groups introduced from S-admosylmethionine.Details of the system are best left to the cognoscenti. Further information on the sequence of events after precorrin-3 (109) has been obtained. Incorporations of precorrin-3 and the corresponding 12-methyl analogue (1 10) were compared under HOzC CO2H \ COzH (109) R = CHZCOZH (110) R = Me 179 A. R. Battersby M. G. Baker H. A. Broadbent C. J. R. Fookes and F. J. Leeper J. Chem. SOC.,Perkin Trans. 1 1987 2027; C. J. Hawker W. M. Stark and A. R. Battersby J. Chqm. SOC.,Chem. Commun. 1987 1313. 180 H. C. Uzar A. R. Battersby T. A. Carpenter and F. J. Leeper J. Chem. SOC.,Perkin Trans. 1 1987 1689. 350 T. J. Simpson strictly controlled conditions.lS1 Whereas (1 10) showed high incorporation (109) did not suggesting that precorrin-3 does not undergo decarboxylation.Previous results have shown that the fourth methylation occurs at C17 and if decarboxylation occurs before the fifth methylation at C12 as appears mechanistically reasonable then the intermediates (111) and (112) are likely to exist. COzH H02C (1 11) R = ,CH,COlH (112) R = Me Previous studies on the biomimetic methylation of pyrrocorphins have been extended to models with nitrile side chains analogous to the normal acetate and propionate side chains.'82 A review of the origins of the molecular structures of vitamin B, concludes that the corrin ring structure could readily have formed under prebiotic condition^.'^^ 7 Miscellaneous Metabolites Several papers have been published describing studies on the biosynthesis of the structurally varied cofactors found in methanogenic bacteria including methanop- terin,ls4 methanof~ran,'~' and pyrroloquinoline quinone.'86 Further evidence has appeared for stepwise enzymatic cyclopropane ring cleavage in the biosynthesis of ethylene."' The 2-amino-3-hydroxycyclopent-2-enone moiety found in reductio- mycin (113) and many other antibiotics has been shown to be derived uia an intramolecular cyclization of 5-aminolaevulinic acid.The remainder of the reduc- 181 F. Blanche S. Handa D. Thibaut C. L. Gibson F. J. Leeper and A. R. Battersby J. Chem. SOC.,Chem. Commun. 1988 1117. 182 C. Leumann T. Fruh M. Gobel and A.Eschenmoser Angew. Chem. Int. Ed. Engl. 1987 26 261. 183 A. Eschenmoser Angew. Chem. Int. Ed. Engl. 1988 27 5. 184 P. J. Keller H. G. Floss Q. Le Van B. Schwarzkopf and A. Bacher J. Chem. Am. Soc. 1986,108,344. W. Eisenreich B. Schwarzkopf Q. Le Van P. J. Keller and A. Bacher J. Chem. SOC.,Chem. Commun. 1988 1294. 186 D. R.Houck J. L. Hanners and C. J. Unkefer J. Am. Chem. SOC.,1988 110 6920. J. E. Baldwin R. M. Adlington G. A. Lajore C. Lowe P. D. Baird and K. Prout J. Chem. SOC.,Chem. Commun. 1988 775. Biological Chemistry- Part (ii) Biosynthesis tiomycin molecule is derived from 4-hydroxyben~oate.'~~ The elusive four-carbon precursor of the dimethylbenzenoid moiety of riboflavin has been shown to be 3,4-dihydroxy-butan-2-one 4-phosphate (1 14).Its formation from pentose phosphate by a highly purified enzyme from Candida guilliermondii has been observed directly by MF:H 08 J. M. Beale J. P. Lee A. Nakagawa S. Omura and H. G. Floss J. Am. Chem. SOC.,1986 108 331. R. Volk and A. Bacher J. Am. Chem. SOC.,1988 110 3651.