The Biosynthesis of Plant Alkaloids and Nitrogenous Microbial Metabolites Richard B. Herbert School of Chemistry University of Leeds Leeds LS2 9JT UK Reviewing the literature published in 1994 (Continuing the coverage of literature in Natural Product Reports 1995 Vol. 12 p. 445) 1 Pyrrolidine and Piperidine Alkaloids 1.1 Tropane Alkaloids 1.2 Cyclizidine and Swainsonine 1.3 Pyrrolizidine and Quinolizidine Alkaloids 2 Isoquinoline Alkaloids 2.1 Morp hinan A1 kaloids 2.2 Protoberberine and Benzophenanthridine Alkaloids 2.3 Colchicine and Ipecac Alkaloids 3 Metabolites derived from Tryptophan 3.1 Terpenoid Indole Alkaloids 3.2 Eudistomins Pyrrolnitrin Xenorhabdus Metabolites and Indolic Phytoalexins 4 Other Metabolites of the Shikimate Pathway 4.1 Pyoverdines and Gliovirin 4.2 Ephedra Alkaloids and TaxoP (Paclitaxel) 4.3 Pentabromopseudilin Ascomycin and Naphthomycin 4.4 DIMBOA 4.5 Phenazines and Actinomycins 5 P-Lactams 5.1 Clavulanic Acid and Valclavam 5.2 Penicillins 6 Miscellaneous Metabolites 6.1 And rimid Lac t acys tin and Eus ta tins 6.2 3-Nitropropionic Acid and Valanimycin 6.3 Tetraponerine-8 and Lysolipins 6.4 Nucleoside Antibiotics 6.5 Coronatine 7 References Year by year some groups of secondary metabolites remain like hardy perennials the subject of consistent and interesting research whilst some groups burst fresh upon the biosynthetic scene and flower briefly; both types can reveal exciting new aspects of secondary metabolism.Increasingly there is a shift to more biologically based investigations. One way or another it is a delight to read and report on the ever evolving biosynthetic studies of this group of nitrogen-containing metabolites. The time-honoured pattern of previous reports is continued in this year's account with appropriate specific reference to earlier reports in the series and to other reviews for background material.l-s Access to the literature was obtained in substantial measure through the excellent IS1 Data Service at Bath. The stereochemistry of enzymic reactions which take place in association with a-amino acids has been very nicely re~iewed.~ The biomolecular and genetic aspects of nicotine and iso- quinoline terpenoid indole and tropane alkaloids -for which there is now extensive such information -has been surveyed.'O 1 Pyrrolidine and Piperidine Alkaloids 1.1 Tropane Alkaloids The biosynthesis of the tropic acid (2) moiety in tropane alkaloids such as hyoscyamine (3) begins with phenylalanine (1).In the course of biosynthesis the skeleton of (1) rearranges such that the carboxyl group undergoes an intramolecular 1,2- (3) R = 'HYoH 0 ?H (4) R = pph 0 (5) R = 0 shift with retention of configuration at C-3 and that the %pro- S hydrogen of (1) migrates in the reverse direction.s Recently it has been shown that phenyllactic acid (6) is a key late intermediate; it is a more immediate precursor than phenyl- pyruvic acid (ref.5 p. 446). This latter conclusion has been confirmed by the results of an experiment with (RS)-[2-13C,-2-2H]-3-phenyllactic acid [as (6)].11 Incubation of this material with transformed root cultures of Datura stramonium gave labelled littorine (4) and hyoscyamine (3). Detailed examination (NMR MS) of the latter showed that the 13C-2H bond of the precursor was incorporated intact into the hydroxymethyl group of the tropate moiety in (3). Significant loss of deuterium was also observed for both (3) and (4)which may be attributed to interconversion of (6) and phenylpyruvic acid. Furthermore [1,2-13C2]phenyllactic acid [as (6)] was incorporated in D.stramonium into the tropic acid moieties of hyoscyamine (3) and scopolamine (7) in a manner such that the two labels become contiguous (labelling of C-1' and C-2'),12 i.e.intra-molecular 1,2-shift of the carboxyl group as observed for phenylalanine [see above cf. (I) (2) and (3)]. In transformed Datura root cultures tropic acid was found to inhibit alkaloid production whilst phenyllactic acid was less inhibitory; the former had little influence on the specific incorporation of phenyllactic acid which was an efficient alkaloid precur~or.~~ This indicates strongly that free tropic acid is not an alkaloid precursor and this is supported by additional definitive results. NMe (RS)-Phenyl[1,3-*3C2]lactoyl[rnethyl-2H3] tropine [as(4)] on incubation with root cultures gave quintuply labelled hyos- cyamine (3) and apoatropine (9 i.e.intact incorporation ; 45 some of the label in the alkaloids formed and of the littorine recovered arose from hydrolysis and recombination ;neither of the hydrolysis products phenyllactic acid and tropine when added to the cultures led to dilution of the quintuply labelled (3) and (5) which was produced; the rearrangement of the labelled phenyllactoyl moiety was observed in the expected way.14 Clearly littorine (4) can serve as a direct precursor in vivo for hyoscyamine (3) and is thus the substrate on which rearrangement can occur. (An additional path through the CoA ester of phenyllactate cannot yet be excluded. The early stages of tropane alkaloid biosynthesis can involve putrescine (8) as an intermediate (cf.ref. 3 p. 575) which becomes N-methylated to give N-methylputrescine as the next biosynthetic intermediate. Further work on the enzyme responsible putrescine N-methyltransferase has been reported. The enzyme was purified 700-fold and characterized. The latter included testing the activity of analogues of putrescine there was a requirement for two amino groups in a trans conformation separated by four carbon atoms; the C diamine cadaverine was a very poor substrate.l5 (For the ensuing stages of tropane biosynthesis see ref. 5 p. 445 and earlier reports.) Two tropinone reductases have been identified in vivo; one reversibly converts tropinone (9) into tropine (10) and the other yields Y-tropine [the C-3 epimer of (lo)] irreversibly (ref.4 p. 55). Both of these compounds have been found to accumulate in transformed root cultures of Atropa belladonna. Results of feeding experiments indicate that tropine and Y-tropine do not isomerize and only the former is incorporated into hyoscyamine (3).16 The two reductases have been purified from the transformed root cultures and characterized.l63 Basic simi- larities were seen with tropinone reductases from other sources mainly regarding kinetic properties and analogue acceptance ; there were also clear differences." NMe NMe 5l 0 (9) The gene for hyoscyamine 6P-hydroxylase [see (3)] from Hyoscyamus niger has been expressed in hairy routes of A. belladonna (ref. 5 p. 446). This gene fused to a reporter gene has been expressed in A.belladonna H. niger and Nicotiana tabacum.18 Expression of the hydroxylation gene was examined and it was concluded from the results that it is controlled by some genetic regulation specific to scopolamine-producing plants [6/?-hydroxyhyoscyamine is a precursor for scopolamine (711. Tropane alkaloid patterns in plants and hairy roots of Hyoscyamus albus have been examined in detail.Ig The incorporation of 14C-labelled phenylalanine ornithine and arginine into tropane alkaloids in suspension-cultured cells and aseptic roots of intact A. belladonna has been studied." The following have also been examined scopolamine production in root cultures of Duboisia myoporoides obtained by repeated selectionz1 and by a new two-stage culture method;22 the effects of oxygen on the production of nicotine and tropane alkaloids in cultured roots of D.my~poroides.~~ Anatoxin-a (1 1) is produced by a number of cyanobacterial species including Anabaena jlos-aquae. It has been found that like other pyrrolidine alkaloids,8 it is derived (in part) from arginine ornithine putrescine (13) and A'-pyrroline (12) (incorporation of 'T-labelled materials and identification of appropriate enzyme activities in A.fl~s-aquae).~~ NATURAL PRODUCT REPORTS 1996 N-H 1.2 Cyclizidine and Swainsonine Cyclizidine (14) is an unusual metabolite containing a cyclo- propane ring produced by a Streptomyces species. It is biosynthesized from three propionate units (this includes the provenance of the cyclopropane ring) and four acetates [see ( 14)].8-25 New evidence26 demonstrates (a) that the oxygen at C-2 is carried over from a molecule of acetate (presumably the other two oxygen atoms originate in aerial oxygen); (b) C-16 as well as C-9 and C-17 originate in C-3 of propionate; (c) in providing C-15 of (14) C-2 of propionate is utilized with retention of only one proton that is the 2-pro-S and in addition inversion of configuration occurs (determined with deuterium labelling).A mechanism consistent with these data is shown in Scheme 1; the initial steps account for observation (c). COSCoA COSCoA COSCoA )-H* )V-HR ->;H~* Me Hs* X S Scheme 1 Swainsona galegifolia produces the indolizine alkaloid swain- sonine.It has been found that the addition of copper sulfate reduction of medium pH and feeding with swainsonine precursors enhanced swainsonine production in transformed root cultures of S. galegifolia; stimulation of the release of swainsonine into the culture medium was also 1.3 Pyrrolizidine and Quinolizidine Alkaloids A notable contribution has been made to our understanding of the biosynthesis of pyrrolizidine alkaloids through the identi- fication in Senecio vulgaris of homospermidine synthase (HSS) which catalyses the combination of two molecules of putrescine (1 5) to give homospermidine (1 7) an intermediate in pyrrolizidine alkaloid biosynthesis. The reaction is NAD' dependent (ref. 5 p. 447). Welcome further progress has now been reported.28 Both plant and bacterial HSS have been examined.First it was shown that label from both (S)-and (R) -[1-*H]putrescine could be completely retained in the homospermidine produced that is to say hydrogen is transferred to and from C-1 of putrescine in what is essentially an intramolecular process (by way of NADSNADH). Some conflict with previous results using whole plants was noted. Second in the presence of putrescine (1 5) spermidine (1 6) could substitute for one of the molecules of (1 5) required to make (1 7). It is suggested that in NATURAL PRODUCT REPORTS 1996R. B. HERBERT the biosynthesis of pyrrolizine alkaloids in vivo half of (1 7) has its provenance in (15) the other in (16). The way it is deduced that the enzyme works is shown in Scheme 2.Protein N I I I I H2Nd NH3 + or H2N-NH2 Transaldimination I 1 I 1 NH2 I+ I I NH2 Scheme 2 Details were recently reported of a study on the biosynthesis of Lythraceae alkaloidsz9 (ref. 5 p. 447) and work has again been published in this area. Incorporation of alkaloid pre- cursors as [3H]lysine/[14C]phenylalanine and [3H]lysine/[14C]- p-coumaric acid has been used to establish relationships between the three alkaloid types found in Heimia salicifolia plants.30 Whereas within an alkaloid type both the cis- and trans-fused variations originate from a common metabolic pool the three types of alkaloid are formed from different precursor pools. Where they overlap the earlier resultszg are consistent with the findings reported here.30 Quinolizidine alkaloids are constituted from multiples of cadaverine (18) units two units for instance being used to construct lupinine (19).The fate of the cadaverine protons on /OH & 4 3 incorporation into a range of quinolizidine alkaloids has been examined in detail. Results for the incorporation of [l-2H]- and 4-zH,]Cadaverine [as (1 S)] gave lupinine (22) labelled as shown; (-)-sparteine was also isolated in this experiment and it had a labelling pattern similar to that found previously for the enantiomeric compound. The chirally labelled compounds (20) and (21) gave lupinine with the labelling patterns (23) and (24) respectively. It is notable that the carbon which becomes C-1 of (19) is associated with stereospecific retention of a 2-pro-R proton from cadaverine and loss of a 2-pro-S proton.Retention of deuterium at C-1 and C-3 means that imine/enamine equilibration does not occur during biosynthesis as it does during the biosynthesis of other quinolizidine alkaloids such as sparteine with consequent deuterium loss. R' d:: R2 D~; D (20)R' = D;R2 = H (21) R1=H; R2=D Work on actyltransferases associated with the biosynthesis of quinolizidine alkaloids (ref. 5 p. 447) has been extended with the isolation purification (to homogeneity) and charac- terization of a novel 0-trigoyltransferase from Lupinus termis seedlings.33 Two isoforms of the enzyme were identified both of which were specific in catalysing transfer of tiglic acid from tigloyl CoA to (-)-13a-hydroxymultiflorine (25) thus yielding (26) and to (+)-13a-hydro~ylupanine.~~ Both of these compounds are tetracyclic with an axial hydroxy group at C- 13 and are of 7S,9S configuration.No enzyme activity was detected towards the 7R,9R alkaloid (-)-baptfoline or towards the bicyclic quinolizidine alkaloids (+)-epilupinine and (-)-lupinhe. (-)-3P-Hydroxy- 13a-tigloyloxylupanine has been isolated as a new alkaloid from seedlings of Cytisus scoparius and tigloyltransferase activity associated with the formation of the alkaloid was detected in cell-free extracts of this plant.34 Clarification has been given to the interspecies distribution of acyltransferases in relation to alkaloid pattern in the various plants.33 2 lsoquinoline Alkaloids 2.1 Morphinan Alkaloids In a remarkable body of work nigh on all the enzymes involved in the biosynthesis of morphine have been described (ref.5 p. 446; and earlier reports). One of the remaining enzymes has been isolated this past year (from poppy cell cultures) and has been purified to homogeneity and characterized. It catalyses the closure of the oxide bridge in the conversion of salutaridinol (27) into thebaine (29)3s and there is a surprise associated with this. Enzyme activity was found to be associated with intact [3-zH]cadaverine into lupinine have been reported previo~sly.~~ mitochondria. Since intact mitochondria are known to generate In new work the incorporation of [2-2H]cadaverine [as( 18)] ATP this suggested that phosphate might serve as a leaving into lupinine in Lupinus luteus has been [2,2,4 group in the ring closure.But when an enzyme preparation NATURAL PRODUCT REPORTS 1996 was incubated with ATP/Mg2+ or any other nucleoside triphosphate conversion was not enhanced. On the other hand use of a cofactor mix which included coenzyme A resulted in a 100°/~ enhancement. Quite surprisingly it turns out that the enzyme (acetyl coenzyme A salutaridinol-7-0-acetyltrans-ferase) is acetyl CoA dependent and that salutaridinol 7-0-acetate (28) is formed which then undergoes spontaneous ring closure in vivo at slightly alkaline pH to give (29) with acetic acid as the leaving group [an enzyme catalysing this step could not be found; the C-7 epimer of (27) was inert with the acetylating enzyme].It may be noted that one other step in morphine biosynthesis namely the conversion of neopinone into codeinone (ref. 5 p. 447) is also spontaneous. Me0 HO& Mz$,NMe - NMe Me0 ,' Me0 Hd H Acd H 1 Me0 ' Me0 S\ NMe 2.2 Protoberberine and Benzophenanthridine Alkaloids The biosynthesis of sanguinarine (3 1) involves many steps the last of which takes place with the oxidation of dihydro- sanguinarine (30).36 Five further enzyme-catalysed steps beginning with (30) lead to macarpine (36) (Scheme 3). The first L6 (30) Lo (32) R=H 1 E(33) R = Me I (34) R=H C(35)R=Me J OMe Scheme 3 two steps involve hydroxylation to (32) then 0-methylation to (33).The enzymes for these steps have been characterized; they are highly substrate specifi~.~'Two novel enzymes dihydrochelirubine- 12-hydroxylase and SAM 12-hydroxy-dihydrochelirubine- 12-@methyltransferase have been found in cell-free extracts of yeast-elicited ThaZictrum bulgaricum (and also Eschscholtzia calif~rnica).~~ They are responsible for the next two steps (33) +(34) +(35) in macarpine biosynthesis. The hydroxylase is a microsomal-associated cytochrome P-450-dependent mono-oxygenase specific for C-12 of (33). The 0-methyltransferase was purified; it appears to be highly specific for (34) as substrate. The last step (35) +(36) is catalysed by non-specific dihydrobenzophenanthridine oxi-da~e.~~ In the biosynthesis of macarpine (36) highly substrate- specific microsome-bound cytochrome P-450 enzymes play a major role in catalysing a number of steps.All of these enzymes are activated (three to more than ten-fold) by yeast elicitation. On the other hand the cytosolic methyltransferases are not activated by the elicitation process. It is that there is interplay between cytosolic and membrane-associated enzymes which may have regulatory roles. The bio-synthesis of macarpine has been authoritatively reviewed.40 Evidence has been obtained that an external source of calcium ions is required for benzophenanthridine alkaloid accumulation which has been induced by a fungal elicitor in cell-suspension cultures of Sanguinaria canadensi~.~~ was It suggested therefore that calcium and possibly calmodulin and/or protein kinase C may participate in a signal transduction system which leads to benzophenanthridine alkaloid pro-duction.Early in the course of the biosynthesis of benzyl- isoquinoline alkaloids e.g. berberine (39) norcoclaurine (37) undergoes 0-methylation to give (38) (ref. 2 p. 512; see Scheme 8). The effect of cytochinins on the activities of early biosynthetic enzymes in ThaZictrum minus cell cultures has been examined and it has been found that of these enzymes norcoclaurine-6-0-methyltransferasewas markedly activated by cytochinins and especially by 6-ben~ylaminopurine.~~ The results suggest that the induction of berberine production in these cell cultures by cytochinins is primarily attributable to the increase in this 0-methyltransferase activity.Other work has been reported on the stimulation of berberine production in T. minus cell "H <% HO 0Y ''H 0 OMe ' HO '*' OMe (37) R = H (39) Berberine (38) R = Me S-Adenosyl-L-methionine:tetrahydroberberine-cis-N-methyl-transferase catalyses the specific cis-N-methylation of e.g. canadine (40) to give the N-methyl derivative (41). This 0 OMe OMe uOMe uOMe enzyme has been purified to homogeneity from S. canadensis cell cultures and has been ~haracterized.~~ Of several substrates tested the enzyme was only fully active with (40) that is to say it is highly substrate specific; surprisingly stylopine (42) was not a substrate (cf. ref. 40). Previously an N-methyltransferase NATURAL PRODUCT REPORTS 1996-R.B. HERBERT catalysing the same reaction had been partially purified from E. ~alifornica.~~ Stylopine(42)and canadine (40)were found to be equally effective and the best substrates for this enzyme. The relationship of the enzymes from the two sources remains to be explored further. It has been that the results with the E. californica enzyme arise from the use of partially purified protein. Stylopine (42) is biosynthesized in two steps from scoulerine (43) each step involving the formation of a methylenedioxy group by oxidation of an O-methoxyphenol. The two enzymes from E. californica,which are involved have been characterized they are P-450 enzymes (ref. 3 p. 578; ref. 2 p. 511) and thus ring closure probably involves (formally) the radical (45) rather than the cation (46).A P-450enzyme has been detected in microsomal preparations from different Ranunculaceae and Berberidaceae cell cultures and partly characterized in prep- arations from a Thalictrum tuberosum cell line.46 This enzyme catalyses the conversion of (S)-tetrahydrocolumbamine (44) into (Qcanadine (40) and is hghly specific for this substrate. This conversion represents the penultimate step in the bio- synthesis of berberine (39) and this enzyme is the last to be characterized in the biosynthetic pathway. The substrate specificity of the enzyme confirms the deduced role for canadine (40) as an intermediate in berberine biosynthesis. The cDNA from E. californica that encodes berberine-bridge enzyme (see e.g.ref 1 p. 188) has been heterologously expressed in a cell culture of the fall army worm Spodoptera fr~giperda.~’ The expression resulted in overproduction of the plant enzyme in a catalytically active form. Jasmonic acid (47) is deduced to be an integral part of a general signal transduction system which regulates inducible defence genes in plants (ref. 5 p. 449). Induction of berberine- bridge enzyme was part of this study. It has likewise been part of a study in which coronatine (48) was found to mimic octadecanoid (jasmonic acid) signalling in higher plants ;(48) does not elicit the accumulation of endogenous jasmonic acid. Most interestingly modelling reveals that coronatine is a structural analogue of the cyclopentane octadecanoid pre- cursors [as (49) cis isomer] of jasmonic 0 (49) Clear evidence has been obtained that tetrahydroproto- berberine alkaloids with the unusual 14-R configuration [as (50)] are derived by reduction of alkaloids which are at the oxidation level of berberine (39) (ref.3 p. 578). This has now been confirmed in precursor feeding experiments in Corydalis cava for corydaline (51) tetrahydrocorysamine (52) and thalictricavine (53).49 A partially enriched protein fraction from C. cava was found to catalyse reduction and methylation of berberine (39) to give (53) (other protoberberines showed varying substrate acceptability). The conversion is dependent on SAM and NADPH (B-type reductase). 7,8-Dihydro- berberine (54) was an excellent enzyme substrate.Thus the sequence to e.g. thalictricavine (53) is very reasonably (39) + (54) -+ (53). (51) R1 = R2= R3 = R4 = Me (52) R1R2= R3R4= CH2 (53) R1R2= CH2; R3 = R4 = Me The X-ray crystal structure of a salt of (-)-corycavinium (56) has been carried This alkaloid and its desoxy derivative (55) were shown to be efficiently transformed into corynoline (57) and its C-14 epimer in tissue cultures of C. cava;(55) was also transformed into (56). (For earlier related work see ref. 5 p. 450; ref. 2 p. 514; ref. 1 p. 189.) (55) R = H (SS)R=OH The aberrant bioconversion of unnatural 2'-aminoreticuline (58) into unnatural alkaloids such as 12-aminoberberine [as (39)] has been reported; incorporation as measured by radioactivity was less than 0.7 %.51 MeomNMe HO OMe 2.3 Colchicine and the Ipecac Alkaloids The bioconversion of colchicine and thiocolchicine into their 3-demethyl derivatives has been studied in plant cell and bacterial cultures.52 The effect on cephaeline and emetine production of adding known precursors of ipecac alkaloids to tissue cultures of Cephaelis ipecacuanha has been 3 Metabolites derived from Tryptophan 3.1 Terpenoid Indole Alkaloids The cDNA encoding strictosidine synthase from RauwolJia serpentina has been heterologously expressed in a cell culture of the fall army worm.47 It has been shown that the indole-diterpenoid paxilline (59) is an efficient precursor for the penitrems of Penicillium janczewskii e.g.penitrem A (60) (ref. 1 p. 192). Recent results confirm this finding; (59) was a similarly efficient precursor for the janthitrems of Penicillium janthinellum e.g. janthitrem B (61).54 Excellent incorporations of 1OP-hydroxypaxilline (62) into both series of metabolites were also observed; the 10a- epimer of (62) was poorly incorporated which indicates that (62) is normally involved in biosynthesis with enzymic attention being paid to the stereochemistry at C-10. 10P-0-Acetyl- paxilline was incorporated at a modest level presumably after in vivo hydrolysis to (62). 7a-Hydroxy- 13-desoxypaxilline [as (59)] and 10P-hydroxy- 13-desoxypaxilline [as (62)] have been identified as novel metabolites of P. pa~illi.~~ 7a-Hydroxy-paxilline [as (59)] was isolated from P.paxilli and was shown to derive from paxilline. Acremonium lolii was found to produce metabolites similar to P. paxilli. Possible biosynthetic relationships between the different indole-diterpenoids have been 3.2 Eudistomins Pyrrolnitrin Xenorhabdus Metabolites and Indolic Phytoalexins Results of feeding experiments with radioactive precursors demonstrate that eudistomin I (63) is biosynthesized in the Floridian tunicate Eudistoma olivaceum from L-tryptophan via tryptamine and from L-proline ; L-ornithine and L-arginine were not used significantly for bio~ynthesis.~~ Aromatic nitro groups are found rarely in secondary metabolites examples being chloramphenicol,8 obafluorin (ref. 4 p. 60) and pyrrolnitrin (66).s These nitro groups arise by oxidation of amino functions and a valuable preliminary contribution to our understanding of what is going on has been NATURAL PRODUCT REPORTS 1996 H (59) H H OH made.A chloroperoxidase has been isolated from cultures of the pyrrolnitrin producer Pseudumonas pyrrocinia which chlori- nates (64) to give (65).57 The pure enzyme (cloned and overexpressed) has now been found to oxidize (65) into pyrrolinitrin (66) in the presence of hydrogen peroxide.58 Further developments are awaited with interest. (For recent work on the formation of 3-nitropropionate see Section 6.2). Tryptophan (68) is incorporated with retention of C-3 and loss of C-1 into the indole metabolites (67) of the entomo- pathogenic bacterium Xenorhabdus nematophil~s.~~ 0 (67)R' = H or COMe R2= H or Me NATURAL PRODUCT REPORTS 1996-R.B. HERBERT SMe H L-Cysteine Other indolic phytoalexins ii (68)L-Tryptophan H (72) H (73) SMe Scheme 4 The biosynthesis of the cruciferous phytoalexins brassinin (71) cyclobrassinin (73) and spirobrassinin (72) has been examined in detail in UV-irradiated turnip slices. Preliminary results (ref. 3 p. 580) are now incorporated in a full paper.6o The combined data allow a clear description of the biosynthetic pathway to be that shown in Scheme 4. A key putative intermediate is (70) and there is new persuasive evidence that it is a normal turnip metabolite. Thus a turnip homogenate when incubated with sodium methanethiolate gave (71) and also (74); neither metabolite was produced in the absence of methanethiolate.Also the moderately reactive benzyl isot- hiocyanate gave (75) on incubation with the homogenate. L-[methyl-3H,35S]Methionine afforded doubly labelled (71) and (72) without change of isotope ratio i.e. the methylthio group in these metabolites arises intact from methionine an unusual observation. Products obtained from the metabolism of (76) support a role for the epoxide (77) (or equivalent) in the steps beyond (71). Finally a close relationshp in the biosynthesis of glucosinolates [as (69)] and that of the phytoalexins has been deduced. The completed pathway whch involves a key rearrangement of the tryptophan (68) skeleton (see * atoms) is shown as Scheme 4.H PhT%N IfSMe S (74)n= 2 (75)n= 1 ~-@-~~C]Tryptophan and ~-['~CH,]rnethionine have been found to be incorporated into the phytoalexins cyclobrassinone and 1-methoxyspirobrassin in UV-irradiated slices of tubers of kohlrabi6' This is of course consistent with the foregoing. On the other hand in Arabidopsis thaliana it has been found62 that the phytoalexin camalexin (78) is formed directly from 9 H H2N-NmoH N 'OH H oHN \ AH NH2 (79) anthranilic acid rather than its later metabolite tryptophan. It was concluded that the pathway to (78) diverts from an intermediate in primary metabolism that lies between an-thranilic acid and indole. H H (77) (78) 4 Other Metabolites of the Shikimate Pathway The biosynthesis of glucosinolates and cyanogenic glycosides is the subject of detailed discussion in a companion 4.1 Pyoverdins and Gliovirin The biosynthesis of the pyoverdins (ref.4 p. 61) has been reviewed.64 Pseudobactin (79) is a member of this group of metabolites. The aromatic moiety of (79) has been shown to originate in Pseudomonas fiuorescens from DL-tyrosine [as (80)].65 Dopa is not a precursor so further aromatic hy- droxylation occurs after joining of the tyrosine residue to at least part of the rest of the molecule. Diketopiperazine metabolitesg constitute a diverse and interesting group of metabolites. One such metabolite is gliotoxin (82) which has been extensively investigated (ref. 3 p. 580); it originates in part in phenylalanine.Gliotoxin (82) and gliovirin (83) are produced by different strains of the fungus Gliocladium virens. It has now been shown66 that (83) is formed from two molecules of L-phenylalanine (8 1) (labelling of C-1 and C-3 of (83) by the [l-13C]-labelled amino acid). Rm:H (80)R= OH (81) R = H OMe NATURAL PRODUCT REPORTS 1996 4.2 Ephedra Alkaloids and TaxoP (Paclitaxel) A full paper has appeared6’ on the biosynthesis of Ephedra alkaloids whch proceeds via a novel route from phenylalanine by way of benzoic acid (as its CoA ester?) (ref. 5 p. 456). The origin of the N-benzoylphenylisoserine side chain in TaxoP (84) has been the subject of recent scrutiny (ref. 5 p. 455). It has now been demonstrated by making elegant use of multiply deuteriated precursors that the sequence of side- chain attachment in Taxol biosynthesis is baccatin-I11 (86) (intact incorporation of 10-acetyl-2H, 13-2H,-labelled material) HN4’ +(87) (intact incorporation of material deuteriated thrice in the acetyl group and five-fold in the phenyl ring) -P Taxol (84).6s N-Benzoylated phenylisoserine when used as a pre-cursor was hydrolysed prior to incorporation.It was established H H that cephalomannine (85) is formed by N-acylation of (87) (90) (Scheme 5). 0 PhKNH 0 (84) R= MeoYYMe PhU Taxow AH 0 0 (85) R = I OH (86) R =H *p* Me *‘OH (91) t NH2 0 0 PhKSCoA OH Hd (87) I Scheme 5 4.3 Pentabromopseudilin Ascomycin and Naphthomycin Pentabromopseudilin (89) is a potent marine antibiotic which has been isolated together with violacein (90) from Chromo-Hca; bacteria and Alteromonas luteoviolaceus.[For the bio-synthesis of (90) see ref. 5 p. 454.1 Exploratory feeding Hd . OH experiments with differently labelled glucose samples in cultures OH of A. luteoviolaceus showed that the benzene ring of (89) has its provenance in carbohydrate metab~lism.~~ The labelling pat- terns observed for (89) and (90) were unexpected but they could reasonably be attributed to a lack of triosephosphate isomerase in the organism. The results did show however that the shikimate pathway was involved [presumably due to per- meability problems shikimic acid itself was incorporated into neither (89) nor (90)].The labelling patterns for (89) indicated that a symmetrical intermediate was implicated in its bio- synthesis. This was identified as p-hydroxybenzoic acid (88) and it was firmly identified as a biosynthetic precursor in feeding J experiments with 2H-and lT-labelled (88). Curiously the pyrrole ring was not labelled by labelled samples of acetate tryptophan benzoic acid glycerol or glucose.69 -(91) Following work on related metabolites the biosynthesis of the dioxygenated cyclohexane ring in ascomycin (91) has been studied.’* [2-13C]Shikimic acid [as (92)] specifically labelled OH C-34 of (91). Further experiments were carried out with Scheme 6 NATURAL PRODUCT REPORTS 1996R. B. HERBERT deuteriated shikimic acid samples from the results of which the sequence and stereochemistry of the reactions could be deduced; the two acids (93) and (94) were also found to act as precursors and can be located as biosynthetic intermediates.It is interesting to note that there are significant stereochemical differences between this case (Scheme 6) and that of cyclohexanecarboxylic acid which also originates from shikimic acid (ref. 5 p. 456). Results of an extensive study with 13C-labelled precursors demonstrate that naphthomycin A (95) is assembled in Streptornyces collinus via a polyketide pathway from 3-amino- 5-hydroxybenzoic acid (96) as a starter unit plus seven propionate and six acetate chain extension units as A m-C,N unit [as (96)] is common to a number of antibiotics8 and this unit may or may not derive via the shikimate pathway (ref.4 p. 59; ref. 3 p. 584; ref. 2 p. 522; ref. 1 p. 195). Results of a feeding experiment with [13C3]glycerol establish that the rn-C,N unit in (95) is elaborated via the shikimate pathway (see ref. 4 p. 59 for details of the likely pathway). Finally it was shown that [7-13C]-3-amino-5-hydroxybenzoic acid [as (96)] was efficiently and specifically incorporated into (95) (labelling of C-27).'l H0$ 4.4 DIMBOA DIMBOA (99) is a major defence compound in maize. It has long been known to derive from anthranilic acid (97) and ribose. Recently it was shown that tetradeuterio-anthranilic acid (98) was incorporated into (99) with retention of three deuterium atoms consistent with entry of a hydroxyl group at C-1 by a mechanism of hydroxylative decarb~xylation.~~ The effect of growing cultures of maize in D,O has been explored.73 Glucosyltransferase and N-hydroxylase activity associated with DIMBOA biosynthesis has been identified in maize and ~haracterized.,~2H- 1,4-Benzoxzin-3(4H)-one (101) has been valine and good evidence, has been obtained that the necessary inversion of configuration in the course of biosynthesis occurs in the enzyme-bound form of the amino acid.This is of likely relevance to the biosynthesis of other peptidic antibiotics which contain D-amino acids e.g. the penicillins (Section 5.2) and virginiamycins. 5 B-Lactams 5.1 Clavulanic Acid and Valclavam Beautiful work has been reported on the biosynthesis of clavulanic acid (1 06) in Streptomyces clavuligerus most recently the stages leading up to the intermediate claviminic acid (102) (ref.5 p. 458). Two things need to happen in the conversion of (102) into (106); that is to say change of an amino group to a hydroxyl [the 9-oxygen in (106) derives from molecular oxygen indicating the transformation is oxidative] and a curious inversion of stereochemistry at C-3 and C-5. These points are reasonably accommodated by proposing that the aldehyde (103)/(105) is an intermediate between (102) and (106); the change in stereochemistry could occur simply via (104). This aldehyde has now been identified as present in S. clavuligerus cultures and it has the (3R,5R) stereochemistry (105).78 Further an enzyme clavulanic acid dehydrogenase has been isolated from S.clavuligerus which converts (105) into clavulanic acid (106) in the presence of NADPH. The enzyme has been purified and the N-terminal sequence has been determined (it is PSALQGKVALITGASSGIGE -amended seq~ence'~). Evi- dence was obtained that the benzyl ester of (105) undergoes slow spontaneous racemization in solution presumably via the ester of (104). All this evidence points strongly at the aldehyde (105) being an intermediate in clavulanic acid biosynthesis. The aldehyde (105) undergoes ready decarboxylation thus providing a reasoned entry to non-carboxylated clavam~.,~ H H m 0-C02H c02x firmly identified as an intermediate in DIMBOA bio~ynthesis.~~ It was an efficient precursor for (99) and could be isolated in a radioactivity trapping experiment.Further it was converted into (100) in an NADPH- and oxygen-dependent reaction with maize microsomes. R A H (97) R = H (99) R=OMe (101) (98) R = 0 (100) R=H 4.5 Phenazines and Actinomycins The genetic cloning of a phenazine biosynthetic locus from Pseudomonas aureofaciens and analysis of its expression has been reported.s6 Multifunctional actinomycin synthase I1 assembles the intermediate containing the first three residues of the acti- nomycin molecule namely 4-methyl-3-hydroxyanthranilic acid L-threonine and D-valine. The enzyme activates L- but not D- Claviminic acid synthase (CAS) is responsible for the two- step conversion of proclaviminic acid (109) into claviminic acid (102) which occurs via (1 10).CAS also catalyses the earlier step whereby (107) is converted into (108). Following upon earlier work (ref. 5 p. 458; ref. 80) two isozymes of CAS have been purified from S. clavuligerus.81One of the isozymes has been cloned and expressed in Escherichia coli. The recombinant enzyme was able to catalyse the above three steps thus confirming a trifunctional role for the enzyme. U 0JPNKNH2 NH C02H (107) R = H (108) R=OH NATURAL PRODUCT REPORTS 1996 During the purification of ACV synthetase (penicillin biosynthesis) from S. clavuligerus a small protein corre-sponding to an amidinohydrolase was identified which is essential for clavulanic acid production.82 The corresponding gene was located near the penicillin cephamycin and clav- ulanic acid biosynthetic genes.It seems that this protein may be the same as PAH which catalyses the conversion of (108) into (109) (ref. 5 p. 458). Valclavam (1 11) is produced by Streptomyces antibioticus ssp. antibioticus Tii718. Primary precursors83 for (1 11) are L- valine the methyl group of L-methionine a C pool metabolite and significantly arginine rather than ornithine (as for clavulanic acid ref. 5 p. 458). Work by others shows that proclaviminic acid (109) and valine are specific precursor^.^^ A link between the biosynthesis of clavulanic acid and that of (1 11) is apparent. This is strengthened by the identification in S. antibioticus of clavulinate biosynthesis enzymes namely CAS and PAH.85 5.2 Penicillins A review on the reactions of non-haem iron with dioxygen in biology and chemistry includes a discussion of the mechanism of action of isopenicillin N synthase.86 Another review which is on 2-oxoglutarate dependent dioxygenase and related enzymes includes a discussion of penicillin and cephalosporin biosynthesi~.~~ A stimulating essay has been publisheds8 on genetic engineering in the synthesis of natural products.It includes penicillins and cephalosporins. Biomimetic conversion of ACV tripeptide analogues into p-lactams has been rep~rted.~’ ACV synthetase couples together L-a-aminoadipate L-cysteine and L-valine to give L-&(a-aminoadipo1y)-L-cysteinyl-D-valine (ACV). In the search for analogues acceptable as substrates for the synthetase it has been founds0 that (S)-carboxymethylcysteine is an effective substitute for a-aminoadipate and both allylglycine and vinylglycine could substitute for cysteine indicating that the thiol group of the latter is not essential for peptide formation.L-Allo-isoleucine but not L-isoleucine could effectively substitute for valine. In common with other non-ribosomal peptide synthetases ACV synthetase appears to have a relatively broad substrate specificity. Amino acid sequence alignment of the Cephalosporium acremonium isopenicillin N synthase (IPNS) to similar non- haem Fe2+ containing enzymes from 28 different sources reveals a homologous region of high sequence conservation with an invariant histidine residue at position 272 in IPNS.’l Site-directed mutagenesis for IPNS confirms the importance of this amino acid residue it is essential for catalytic activity.For related spectroscopic work see ref. 3 p. 585. Isopenicillin N is the progenitor of other penicillins with different side chains the conversion of the former into the latter involving replacement of the L-a-aminoadipoyl chain in isopenicillin N and this is catalysed by an acyltransferase. The effect of site-directed mutagenesis on proenzyme cleavage and the catalytic activity of the acyl coenzyme A isopenicillin N acyltransferase from Penicillium chrysogenum has been stud- ied.s2 A separation of proenzyme cleavage and catalytic activity was noted. A broad-range disulfide reductase has been isolated from P.chrysogenum which when coupled with IPNS results in the conversion of the disulfide form of ACV into isopenicillin N.93 It is suggested that the reductase might have a normal role in penicillin biosynthesis. Expression of penicillin biosynthetic genes in P. chrysogenum has been found to be regulated by nitrogen repression glucose repression and growth stage a-Aminoadipic acid is a key building block in the biosynthesis of penicillins. It is known to be synthesized in P. chrysogenum from a-ketoglutaric acid and acetyl CoA. It has now been found that catabolism of lysine which may occur by two different routes also affords a-aminoadipic acid in this organism.s5 6 Miscellaneous Metabolites The levels of caffeine and threobromine and the metabolism of [8-14C]adenine in developing leaves of C0Jg-a arabica have been examined.s6 Soluble glyucosyltransferase(s) have been identified in Solunum melongena leaves which catalyse the glucosylation and galactosylation of solasodine and diosgenin and related 6.1 Andrimid Lactacystin and Eustatins Results of experiments with precursors labelled with stable isotopes clearly define the origins of the unusual acylsuccinimide moiety of andrimid (112) in cultures of a marine isolate of Pseudomonas $uorescens ; excellent incorporations were ob- tained.ss [1-l3C]Valine labelled C-1’ of (1 12) [13C2]acetate labelled C-2 and C-3 and curiously C-6 (also C-1”’ through C-8”’,but not C-4 and C-5) whilst glycine provides C-4 C-5 and the nitrogen atom as an intact unit.A pathway consistent with the results is shown in Scheme 7.98 Scheme 7 NATURAL PRODUCT REPORTS 1996R. B. HERBERT Interestingly [1,2-13C,]glycine in addition to giving doubly labelled (1 12) also gave some material whch bore a single label at C-5 (corresponds to C-2 of glycine). This is consistent with operation of the tartronic semialdehyde and glyoxylate path- ways recently observed in terrestrial P. fluorescen~.~~ Streptomyces sp. commonly elaborate metabolites with unique structures. Such a metabolite is lactacystin (1 13). Study of its biosynthesis with precursors labelled with stable isotopes shows that C-4 C-5 C-9 C-10 C-11 and C-12 have their genesis in L-leucine the cysteine unit derives from L-cysteine and C-6 C-7 C-8 and C-13 are provided by isobutyrate ([l-13C]isobutyricacid was found to label C-8 C-1 C-4 and C-14; the latter three labelling sites are accounted for by unexceptional isobutyrate metabolism).loO It is suggested that isobutyrate condenses with leucine via methylmalonic semialdehyde.eH HOpC. ,N. ' $OH lsobutyrate A 12 A Leucine L-Leucine and L-ornithine are precursors for eurystatins A (1 14) and B (1 15) in cultures of Streptomyces eurythermus."' Valine and isoleucine could be substituted for leucine thus affording new eurystatins. H (114) R= Me (115) R= Et 6.2 3-Nitropropionic Acid and Valanimycin Penicilliurn utrovenetum biosynthesizes 3-nitropropionic acid (119) from L-aspartic acid (116) (ref.2 p. 206; ref. 4 p. 64). The incorporation of the diethyl ester (117) into (119) after anticipated hydrolysis in the culture indicates (118) is a biosynthetic intermediate. It has now been shown that hydrolysis of racemic (1 17) by pig liver esterase (chosen for mild hydrolysis) affords (1 19).lo2 Since no decarboxylase activity could be associated with the esterase decarboxylation of (1 18) formed in the hydrolysis must be spontaneous. This raises the question as to whether decarboxylation of (118) is normally spontaneous in P. atrovenetum or is enzyme catalysed. Attempts to resolve this question failed. (For other work on the formation of nitro groups see Section 3.2.) fC02R (117) R= Et (118) R= H The provenance of valanimycin (124) in Streptomyces viri- difaciens is in isobutylamine (1 20) via isobutylhydroxylamine (121) and serine (122) (ref.4 p. 64). The role of the hydroxylamine (1 2 1) in biosynthesis has been strongly sup- ported by the identificationlo3 ofenzyme activity in s.viridifuciens which catalyses the conversion of (120) into (121) in the presence of FAD plus NADH. Of a range of amines tested (120) was the best substrate. Specific incorporation of [4-13C]( 123) and incorporation of [15N,]( 123) without fracture of the N-N bond establishes (123) formed reasonably from the reaction of (121) and (122) as an intermediate in valanimycin biosynthesis.lo3 (121) R =OH (123) 6.3 Tetraponerine-8 and Lysolipins The tetraponerines are a group of tricyclic defence alkaloids produced by Tetraponera ants.Feeding experiments (using about a thousand ants per batch) with 14C-labelled compounds indicate that the origins of tetraponerine-8 (125) are as follows.1o4 The pyrrolidine ring arises from glutamate via L-ornithine and putrescine whilst the remaining 12-carbon chain derives from a combination of six acetate units. HH The lysolipins e.g. lysolipin X (126) are members of a small family of xanthone metabolites. ResultsLo5 of a study of the biosynthesis on the lisolipins revealed two particularly in- teresting features. First the oxygen incorporation pattern [see (126)] was unexpected and will repay further investigation. Second the skeleton is constructed entirely from malonate. It serves as starter unit (C-21 C-23 C-24) for what turns out to be a single polyketide chain.The use of malonate here is in the opposite sense to that in cycloheximide and the tetracyclines; that is to say the activated carbon of malonyl CoA is attached to the ring nitrogen of (126) whereas the C0,-derived carbon atom is the one which initiates the polykedite chain. Further results are awaited with interest. *o-o* 6Me Me 6.4 Nucleoside Antibiotics A-Factor is a small y-lactone which induces streptomycin production in Streptomyces griseus. Based on recent work a model has been proposedlo6 for an A-factor regulation cascade which ultimately switches on the streptomycin biosynthetic genes.lo6 Quite unusually the uracil moiety in sparsomycin (131) origmates in tryptophan (1 27).A plausible route to (1 3 1) would be by way of the kynurenine pathway of tryptophan degra- dation whch involves scission of ring B ; in the case of (13 1) scission of ring A would follow. However two intermediates on this pathway namely N-methylanthranilic acid (ref. 4 p. 58) and N'-formylkynurenine (1 28),lo7 were not intact precursors for (1 3 1) in Streptomyces sparsogenes. This argues against the involvement of the kynurenine pathway in sparsomycin biosynthesis and an alternative is favoured where ring A of (127) is cleaved before ring B.lo7 The pyrimidine (1 29) is a specific precursor for sparsomycin (1 3 1) (ref. 4 p. 58) and its potential normal intermediacy in the biosynthesis of (1 3 1) is considerably strengthened by isolating an enzyme from S.sparsogenes that catalyses the conversion of (129) into (1 30). The enzyme requires NAD' and it is suggested that its mechanism of action is similar to that of IMP dehydrogenase which catalyses the conversion of inosine 5'- monophosphate (IMP) into xanthosine 5'-rnonopho~phate.'~~ 0 0 It has been shown recently that neplanicin A (132) is an intermediate in the biosynthetic pathway to aristeromycin (133) in Streptomyces citricozor (ref. 5 p. 460). In strong support of this enzyme activity has been identified in and partially purified from S. citricolor that catalyses the NADPH-dependent reduction of (132) to (133).lo8 It was shown further that reduction proceeds with anti geometry and involves the 4- pro-R hydrogen atom of NADPH (Scheme 8).In the search for intermediates between glucose and neplanicin A/aristeromycin it has been found that the tetraol (134) is not an intact precursor for aristeromycin (133); the incorporation was very 10w .lo9 *HR Hs NATURAL PRODUCT REPORTS 1996 sine D-glucose L-a-arginine and the methyl group of meth- ionine. Carefully chosen enzyme inhibitors have been used to distort biosynthesis in S. griseochromogenes."O Three known previously minor metabolites [including (1 36) and (1 37)] and two new ones accumulated in usefully substantial quantities. This aided elucidation of the biosynthetic intermediates en route to (135) (137) is the first nucleoside intermediatelll and (136) is the last intermediate;l12 the mechanism and stereo- chemistry of C-3' deoxygenation has been e~tab1ished.l~~ (135) R=Me (136)R = H 6.5 Coronatine Coronatine (138) which has a quite unique structure is a phytotoxin produced by many pathovars of Pseudomonas syringae.The coronafacic acid (139) moiety is a polyketide derived from three acetate units one butyrate unit and one pyruvate unit; the two oxygen atoms are acetate oxygens. The coronamic acid (141) portion is formed from L-isoleucine by way of L-alloisoleucine (140); the nitrogen atom of the latter is retained into (141). Cyclopropane formation occurs with loss of the C-2 proton and one proton from C-6; a C-6 hydroxy derivative is not involved. A mechanism for ring closure involving iron 0x0 species may reasonably be drawn.Finally coronamic acid (141) is a highly efficient precursor for coronatine (1 38). Some of these results which had appeared in preliminary communications (ref. 3 p. 588; ref. 114) are now available in a full paper.l15 The intermediacy of coronamic acid (141) in the biosynthesis of (138) has been confirmed by the work of others.116 Normally present in cultures in minute amounts (141) accumulated in a mutant blocked in cor- onatine biosynthesis. It was also labelled by administered ~-[U-~~C]isoleucine and coronamic acid was an efficient coronatine precursor in a blocked mutant. It is clear that cyclopropane formation precedes linkage to the coronofacic acid moiety of (138). "QH H i (138)R= HNq H02C '-(139)R=OH CONH2 OH 6 R H2N H Me bAd Hob..m I ... a . Hd 'OH HO OH HO' OH (132) (133) (1 34) Ad = adenine Scheme 8 Blasticidin S (135) is an antifungal antibiotic elaborated by Streptomyces griseochromogenes. It is constructed from cyto- Work on the cloning and expression of genes responsible for coronamic acid biosynthesis has been reported.'17 Interestingly the production of (141) can occur in P. syringae which lacks the gene cluster for the synthesis of (138) i.e. production of (139) and (141) can occur independently. NATURAL PRODUCT REPORTS 1996R. B. HERBERT 7 References 1 R. B. Herbert Nat. Prod. Rep. 1991 8 185. 2 R. B. Herbert Nat. Prod. Rep. 1992 9 507. 3 R. B. Herbert Nat. Prod. Rep. 1993 10 575.4 R. B. Herbert Nat. Prod. Rep. 1995 12 55. R. B. Herbert Nat. Prod. Rep. 1995 12 445 6 R. B. Herbert in Rodds Chemistry of Carbon Compoundrs ed. S. Coffey Elsevier Amsterdam 1980 2nd edn. Vol. IV Part L p. 291. 7 R. B. Herbert in Rodd’s Chemistry of Carbon Compounds ed. M. F. Ansell Elsevier Amsterdam 1988 2nd edn. Vol. IV Part L supplement p. 155. 8 R. B. Herbert The Biosynthesis of Secondary Metabolites 2nd edn. Chapman and Hall London 1989. 9 D. W. Young Topics in Stereochemistry 1994 21 381. T. Hashimoto and Y. Yamada Annu. Rev. Plant Physiol. Plant Mot. Biol. 1994 45 257. 11 N. C. J. E. Chesters D. O’Hagan and R. J. Robins J. Chem. SOC. Perkin Trans. I 1994 1159. 12 A. Ansarin and J. G. Woolley Phytochemistry 1994 35 935.13 R. J. Robins J. G. Woolley M. Ansarin J. Eagles and B. J. Goodfellow Planta 1994 194 86. 14 R. J. Robins P. Bachmann and J. G. Woolley J. Chem. SOC. Perkin Trans. I 1994 615. N. J. Walton A. C. J. Peerless R. J. Robins M. J. C. Rhodes H. D. Boswell and D. J. Robins Planta 1994 193 9. 16 B. Drager and A. Schaal Phytochemistry 1994 35 1441. 17 A. Portsteffen B. Drager and N. Nahrstedt Phytochemistry 1994 37 391. 18 T. Kanegae H. Kajiya Y. Amano T. Hasimoto and Y. Yamada Plant Physiol. 1994 105 481. 19 K. Doerk-Schmitz L. Witte and A. W. Alfermann Phyto-chemistry 1994 35 107. S. Nyman J. Experimental Botany 1994 45 979. 21 Y. Yukimune Y. Hara and Y. Yamada Biosci. Biotech. Biochem. 1994 58 1443. 22 Y. Yukimune H. Tabata Y.Hara and Y. Yamada Biosci. Biotech. Biochem. 1994 58 1820. 23 Y. Yukimune H. Yamagata Y. Hara and Y. Yamada Biosci. Biotech. Biochem. 1994 58 1824. 24 J. R. Gallon P. Kittakoop and E. G. Brown Phytochemistry 1994 35 1195. F. J. Leeper P. Padmanabhan G. W. Kirby and G. N. Sheldrake J. Chem. SOC.,Chem. Commun. 1987 505. 26 F. J. Leeper S. E. Shaw and P. Satish Can. J. Chem. 1994 72 131. 27 T. M. Ermayanti J. A. McComb and P. A. O’Brien Phyto-chemistry 1994 36 313. 28 F. Bottcher D. Ober and T. Hartmann Can. J. Chem. 1994,72 80. 29 S. H. Hedges R. B. Herbert and P. C. Wormald J. Nat. Prod. 1993 56 1259. A. Rother and J. M. Edwards Phytochemistry 1994 36 911. 31 R. B. Herbert Nat. Prod. Rep. 1988 5 525; 1990 7 107. 32 D.J. Robins and G. N. Sheldrake J.Chem.SOC.,Chem. Commun. 1994 1331. 33 H. Suzuki I. Murakoshi and K. Saito J. Biol. Chem. 1994 269 15853. 34 K. Saito H. Suzuki Y. Yamashita and I. Murakoshi Phyto-chemistry 1994 36 309. R. Lenz and M. H. Zenk Tetrahedron Lett. 1994 35 3897. 36 T. M. Kutchan and M. H. Zenk J. Plant Res. 1993 3 165. 37 W. De-Eknamukul T. Tanahashi and M. H. Zenk Phyto-chemistry 1992 31 2713. 38 L. Kammerer W. De-Eknamkul and M. H. Zenk Phyto-chemistry 1994 36 1409. 39 H.-M. Schumacher and M. H. Zenk Plant Cell Rep. 1988,7, 43; H. Arakawa W. G. Clark and C. J. Coscia Arch. Biochem. Biophys. 1992 299 1. M. H. Zenk Pure and Appl. Chem. 1994,66 2023. 41 G. B. Mahady and C. W. W. Beecher Phytochemistry 1994 37 415. 42 M.Ihara S. Tanaka and M. Tabata Phytochemistry 1994 36 327. 43 D. D. Smolko and S. W. Peretti Plant Cell Rep. 1994 14 131. 44 B. R. O’Keefe and C. W. W. Beecher Plant Physiol. 1994 105 395. M. Rueffer G. Zumstein and M. H. Zenk Phytochemistry 1990 29 3727. 46 M. Rueffer and M. H. Zenk Phytochemistry 1994 36 1219. 47 T. M. Kutchan A. Bock and H. Dittrich Phytochemistry 1994 35 353. 48 E. Weiler T. M. Kutchan T. Gorba W. Brodschelm U. Niesel and F. Bublitz FEBS Lett. 1994 345 9. 49 M. Rueffer W. Bauer and M. H. Zenk Can. J. Chem. 1994,72 170. 50 M. Kamigauchi Y. Noda K. Iwasa Z. Nishijo T. Ishida Y. In and W. Wiegrebe Helv. Chim. Acta 1994 77 243. 51 D. S. Bhakuni and S. Jain J. Chem. SOC. Perkin Trans. I 1994 323. 52 A.Pouler E. Bombardelli C. Ponzone and M. H. Zenk J. Ferment. Biogeng. 1995 79 33. 53 C. Veeresham C. Kokate and V. Venkateshwarlu Phytochemistry 1994 35 947. 54 J. Penn and P. G. Mantle Phytochemistry 1994 35 921. 55 P. G. Mantle and C. M. Weedon Phytochemistry 1994,36 1209. 56 G. Q. Shen and B. J. Baker Tetrahedron Lett. 1994 35 1141. 57 W. Wiesner K.-H. van Pte and F. Lingens J. Biol. Chem. 1988 263 13725. 58 S. Kirner and K.-H. van Pie Angew. Chem. Int. Ed. Engl. 1994 33 352. 59 L. Sundar and F. N. Chang J. Gen. Microbiol. 1993 139 3139. 60 K. Monde M. Takasugi and T. Ohnishi J.Am. Chem. SOC. 1994 116 6650. 61 D. Gross A. Porzel and J. Schmidt 2. Naturforsch. Teil C Biosci. 1994 49 281. 62 J. Tsuji M. Zook S. C. Somerville R.L. Last and R. Ham- merschmidt Physiol. Mol. Plant Pathol. 1993 43 221. 63 P. M. Dewick Nat. Prod. Rep. 1995 12 579. 64 H. Budzikiewicz Pure and Appl. Chem. 1994 66 2207. 65 B. Nowak-Thompson and S. J. Gould Tetrahedron 1994 50 9865. 66 R. D. Stipanovic C. R. Howell and P. A. Hedin J. Antibiot. 1994 47 942. 67 G. Grue-Ssrensen and I. D. Spenser J. Am. Chem. SOC. 1994 116 6195. 68 P. E. Fleming A. R. Knaggs X.-G. He U. Mocek and H. G. Floss J. Am. Chem. SOC. 1994 116 4137. 69 U. Hanefeld H. G. Floss and H. Laatsch J. Org. Chem. 1994,59 3604. 70 K. L. Wallace K. A. Reynolds K. Koch H. A. I. McArthur M. S. Brown R. G. Wax and B. S. Moore J. Am. Chem. SOC. 1994 116 11600. 71 J. P. Lee S.-W. Tsao C. Chang X.-G. He and H.G. Floss Can J. Chem. 1994 72 182. 72 P. Kumar and W. S. Chilton Tetrahedron Lett. 1994 35 3247. 73 S. Peng and W. S. Chilton Phytochemistry 1994 37 167. 74 V. Leighton H. M. Niemeyer and L. M. V. Jonsson Phyto- chemistry 1994 36 887. 75 P. Kumar D. E. Moreland and W. S. Chilton Phytochemistry 1994 36 893. 76 D. G. Georgakopoulis M. Hendson N. J. Panopoulos and M. N. Schroth Appl. Environ. Biol. 1994 60,293 1. 77 A. Stindl and U. Keller Biochemistry 1994 33 9358. 78 N. H. Nicholson K. H. Baggaley R. Cassels M. Davison S. W. Elson M. Fulston J. W. Tyler and S. R. Woroniecki J. Chem. SOC. Chem. Commun. 1994 1281. 79 S. W. Elson private communication. 80 E. N. Marsh M. D.-T. Chang and C. A. Townsend Biochemistry 1993 31 12648.81 E. J. Lawlor S. W. Elson S. Holland R. Cassels J. E. Hodgson M. D. Lloyd J. E. Baldwin and C. J. Schofield Tetrahedron 1994 50 8737. 82 K. A. Aidoo A. Wong D. C. Alexander R. A. R. Rittammer and S. E. Jensen Gene 1994 147 41. 83 J. E. Baldwin K.-C. Goh and C. J. Schofield Tetrahedron Lett. 1994 35 2779. 84 J. W. Janc L. A. Egan and C. A. Townsend Bioorg. Med. Chem. Lett. 1993 3 2313. 85 J. E. Baldwin Y. Fujishima K.-C. Goh and C. J. Schofield Tetrahedron Lett. 1994 35 2783; J. E. Baldwin K.-C. Goh and C. J. Schofield Tetrahedron Lett. 1994 35 2779. 86 A. L. Feig and S. J. Lippard Chem. Rev. 1994 94 759. 87 E. de Carolis and V. de Luca Phytochemistry 1994. 36 1093. 88 A. I. Scott J. Nat. Prod. 1994 57 557. 89 Y. Kita N. Shibata N.Kawano T. Tohjo C. Fujimori and H. Ohishi J.Am. Chem. Soc. 1994 116 5116. 90 J. E. Baldwin C.-Y. Shiau M. F. Byford and C. J. Schofield Biochem. J. 1994 301 367. 91 S. Tow-Suan and D. S. H. Tan FEMS Microbiology Lett. 1994 120 241. 92 M. B. Tobin S. J. C. Cole S. Kovacevic J. R. Miller J. E. Baldwin and J. D. Sutherland FEMS Microbiol. Lett. 1994 121 39. 93 G. Cohen A. Argaman R. Schreiber M. Mislovati and Y. Aharonowitz J. Bacteriol. 1994 176 973. 94 B. Feng E. Friedlin and G. A. Marzluf Appl. Environ. Microbiol. 1994,60,4432. 95 C. Esmahan E. Alvarez E. Montenegro and J. F. Martin Appl. Environ. Microbiol. 1994 60 1705. 96 N. Fujimori and H. Ashihara Phytochemistry 1994 36 1359. 97 C. Paczowski and Z. A. Wojciechowski Phytochemistry 1994,35 1429.98 J. Needham M. T. Kelly M. Ishige and R. J. Andersen J. Org. Chem. 1994 59 2058. 99 R. B. Herbert and A. R. Knaggs J. Chem. SOC. Perkin Trans. I 1992 109. 100 A. Nakagawa S. Takahashi K. Uchida K. Matsuzaki S. Omura A. Nakamura N. Kurihara T. Nakamatsu Y. Miyaki K. Take and M. Kainosho Tetrahedron Lett. 1994 35 5009. 101 K. Suzuki S. Toda T. Furumai Y. Fukagawa and T. Oki J. Antibiot. 194 47 982. 102 R. L. Baxter S. L. Smith J. R. Martin and A. B. Hanley J. Chem. SOC. Perkin Trans. 1 1994 2297. 103 R. J. Parry and W. Li J. Chem. SOC.,Chem. Commun. 1994,995. NATURAL PRODUCT REPORTS 1996 104 B. Renson P. Merlin D. Daloze J. C. Braekman Y. Roisin and J. M. Pasteels Can.J Chem. 1994 72,105. 105 H.Bockholt G. Udvarnoki J. Rohr U. Mocek J. M. Beale and H. G. Floss J. Org. Chem. 1994 59 2064. 106 S. Horinouchi and T. Beppu Antonie van Leeuwenhoek 1993,64 177. 107 R. J. Parry J. C. Hoyt and Y. Li Tetrahedron Lett. 1994 35 7497. 108 R. J. Parry and Y. Jiang Tetrahedron Lett. 1994 35 9665. 109 J. Hanrahan H. Navratilova K. A. Shoberu D. H. G. Crout D. W.Hutchinson,S. M.RobertsandE. M. H.Wellington,J. Chem. SOC. Perkin Trans. I 1994 3533. 110 S. J. Gould J. Guo A. Geitmann and K. Dejesus Can. J. Chem. 1994 72 6. 111 J. Guo and S. J. Gould J. Am. Chem. SOC. 1991 113 5898. 1 12 J. Guo and S. J. Gould Bioorg.Med. Chem. Lett. 1991 1 497. 113 S. J. Gould and J. Guo J. Am. Chem. SOC. 1992 114 10176. 114 R. B. Herbert Nut. Prod. Rep. 1988 5 537. 115 R. J. Parry S. V. Mhaskar M.-T. Lin A. E. Walker and R. Mafoti Can. J. Chem. 1994 72,86. 116 R. E. Mitchell S. A. Young and C. L. Bender Phytochemistry 1994 35 343. 11 7 M. Ullrich A. C. Guenzi R. E. Mitchell and C. L. Bender Appl. Environ. Microbiol. 1994 60 2890.