The Biosynthesis of Shikimate Metabolites P. M. Dewick Department of Pharmacy University of Nottingham Nottingham NG7 2RD Reviewing the literature published during 1986 (Continuing the coverage of literature in Natural Product Reports 1986 Vol. 3 p. 565) 1 The Shikimate Pathway 1.1 DAHP Synthase 1.2 3-Dehydroquinate Synthase 1.3 3-Dehydroquinase 1.4 Shikimate Dehydrogenase 1.5 Quinate Dehydrogenase 1.6 Shikimate Kinase 1.7 EPSP Synthase 1.8 Chorismate Synthase 1.9 Chorismate Mutase 1.10 Phenylalanine and Tyrosine 1.1 1 Anthranilic Acid p-Aminobenzoic Acid and Related Compounds 2 Tryptophan and Related Compounds 2.1 Tryptophan Synthase 2.2 Indole-3-acetic Acid and Related Metabolites of Tryptophan 3 Phenols and Phenolic Acids 4 Phenylpropanoids 4.1 Phenylalanine Ammonia-lyase 4.2 Cinnamic Acids and Esters 4.3 Phenylpropane and Phenylethane Derivatives 4.4 Coumarins 4.5 Lignins 4.6 Lignans 5 Flavonoids 5.1 General Aspects 5.2 Chalcone Synthase 5.3 Flavanones Dihydroflavonols Flavonols and F1avo n e s 5.4 Catechins and Proanthocyanidins 5.5 Anthocyanidins 5.6 Methylation and Glycosylation of Flavonoids 5.7 Sulphation of Flavonoids 5.8 Isoflavonoids 6 Quinones 7 Miscellaneous Shikimate Metabolites 7.1 Cyclohexyl Fatty Acids 7.2 Sporopollenin 8 References This report reviews the literature that was published during 1986 on the biosynthesis of non-nitrogenous compounds that are derived wholly or partly from shikimate and continues the coverage in Volume 3 of Natural Product Rep0rts.l A book entitled ‘The Shikimic Acid Pathway’ has been published,2 containing the invited papers that were presented at The Phytochemical Society of North America Symposium held in June 1985.This book does not aim to provide a totally comprehensive coverage but it contains excellent reviews describing recent advances in many areas relating to shikimate- derived materials. Most of these articles will be cited in the following sections under the appropriate subject material. An early chapter presents a valuable overview of many aspect^.^ 1 The Shikimate Pathway 1.1 DAHP Synthase The first reaction in the shikimate pathway (Scheme 1) involves the condensation of phosphoenolpyruvate (PEP) (1) with erythrose 4-phosphate (2) and is catalysed by the enzyme 3- deoxy-~-arabino-heptulosonate-7-phosphate synthase (DAHP synthase) [phospho-2-dehydro- 3-deoxyheptonate aldolase ; E.C.4.1.2.151. In any particular organism one or more forms of this enzyme may exist and these forms are characterized by different sensitivities towards feedback inhibition especially by the aromatic amino acids. Members of the group of prokaryotes which includes Escherichia coli may possess one or more of the co; I PEP (1) + 4 NAD+ 00 ““tLOH H 0“ OH DAHP (3) OH Erythrose 4-phosphate (2) co; I iii 0OOH 0 I I OH OH 3 -De>ydroshiki mate (5) 3-Dehydroquinate (4) co, I 6 HO’ OH H OR*OOH I I OH OH Shi kimate (6) Quinate (7) Enzymes i DAHP synthase; ii 3-dehydroquinate synthase iii 3-dehydroquinase ; iv shikimate dehydrogenase ; v quinate deh ydrogenase Scheme 1 73 four distinct isozymes DAHP synthase-0 (insensitive to feedback inhibition) DAHP synthase-Phe DAHP synthase- Tyr or DAHP synthase-Trp the latter isozymes being sensitive to feedback inhibition by the appropriate aromatic amino acid.4 It is thought that DAHP synthase-0 and DAHP synthase- Tyr were present in a common ancestor of these organisms and that the newly evolved DAHP synthase-Trp once possessed sensitivity to feedback inhibition by chorismate as well.Sensitivity to chorismate is retained in several pseudomonads. DAHP synthase-Phe probably evolved recently and is presently found only in E. coli and in its close relatives within this group of prokaryotes. The enzyme activity from Bacillus polymyxa has been found to be subject to feedback inhibition by all three aromatic amino acids,5 and that from a Pseudomonas species is subject to inhibition by tyrosine tryptophan anthranilate and phenylpyruvate.6 Similarly isozymes of DAHP synthase have been shown to exist in plants. In cultured cells of Nicotiana silvestris the two isozymes DAHP synthase-Mn and DAHP synthase-Co have been identified,' the abbreviations indicating the requirement of the isozyme for a divalent cation (either Mn2+ or Co2+ respectively).The different properties of the two isozymes allowed assays to be developed for the detection of either isozyme in plant extracts and a variety of monocots and dicots were analysed. In general the pair of isozymes seems to be present in higher plants with higher activity being observed for DAHP synthase-Co in all cases that have yet been studied. Surprisingly only DAHP synthase-Mn had been recognized in many earlier studies. The DAHP synthase from tubers of potato (Solanum tuberosum) closely resembles that which had previously been isolated from carrot (Daucus carota); it shows a hysteretic lag is stimulated by Mn2+ and L-tryptophan and can be resolved into two forms8 1.2 3-Dehydroquinate Synthase The complete sequence of amino-acid residues of the enzyme 3-dehydroquinate synthase [E.C.4.6.1.31 which catalyses the conversion of DAHP (3) into 3-dehydroquinate (4) has been el~cidated.~ The enzyme catalyses a complex sequence of reactions requiring an oxidation a p-elimination a reduction and an intramolecular aldol condensation and the study of its NATURAL PRODUCT REPORTS 1988 properties has been hampered by the low natural levels of this enzyme. The sequencing of the enzyme was made possible because larger amounts were accessible through gene cloning and the construction of overproducing strains of Escherichia coli. Synthetic analogues of DAHP containing fluorine at position 3 were transformed by the 3-dehydroquinate synthase from E.coli and their further conversion into 6-fluoro-derivatives of 3-dehydroshikimate and shikimate was also investigated.lo 1.3 3-Dehydroquinase 3-Dehydroquinase [3-dehydroquinate dehydratase; E.C. 4.2.1.101 catalyses the dehydration of 3-dehydroquinate (4) to 3-dehydroshikimate (9,and has now been isolated in relatively large amounts by gene cloning and by constructing a strain of E. coli that overproduces this enzyme." This has allowed the complete amino-acid sequence to be determined. The enzyme appears to be dimeric and has properties similar to those of the 3-dehydroquinase component of the pentafunctional arom enzyme complex of Neurospora crassa. l2 The monofunctional 3-dehydroquinase has not previously been purified.1.4 Shikimate Dehydrogenase Shikimate dehydrogenase [E.C. 1.1.1.251 is an NADP+-linked dehydrogenase that catalyses the reversible reduction of 3- dehydroshikimate (5) to shikimate (6). The enzyme from the conifer Pinus sylvestris has been isolated and partially purified ; it has been shown to be separable into three isozyme~.'~ In conifers in contrast to other plants all three isozymes are active not only with NADP' but also with NAD'. It has been suggested that the NAD+-dependent shikimate dehydrogenase catalyses the initial reaction of an alternative pathway in which shikimate is converted into hydroxybenzoic acids. 1.5 Quinate Dehydrogenase Quinate (7) is frequently encountered in Nature especially in plants usually in combination with other materials (e.g.with caffeic acid in chlorogenic acid; see Section 4.2). The reversible reduction of 3-dehydroquinate (4) to quinate (7) is catalysed by co co2-co; I I I OH OH OH Shikimate (6) Shikimate 3-phosphate (8) EPSP (9) I OH 6H Prephenate (11) Chorismate (10) Enzymes i shikimate kinase ; ii EPSP synthase; iii chorismate synthase; iv chorismate mutase Scheme 2 NATURAL PRODUCT REPORTS 1988 -P. M. DEWICK quinate dehydrogenase [E.C. 1.l. 1.241. The enzyme activity has been observed in needles of larch (Larix sibirica) and of pine (Pinus sylvestris) and both NAD+-dependent and NADP+- dependent activities could be demonstrated. l4 The action of larch quinate dehydrogenase which had been separated from shikimate dehydrogenase on quinic acid led to the accumulation of protocatechuic and gallic acids though the intermediates 3-dehydroquinate and 3-dehydroshikimate were detectable in labelling studies.The cofactor requirement varied according to conditions; it has been suggested that in conifers quinate can be metabolized either via shikimate or by alternative pathways and the functional role of quinate dehydrogenase depends on the nature of the cofactor. 1.6 Shikimate Kinase Phosphorylation of shikimate (6) to shikimate 3-phosphate (8) is brought about by shikimate kinase [E.C. 2.7.1.711 in the presence of ATP (Scheme 2). Gene cloning has allowed an overproducing strain of Escherichia coli to be constructed. This has for the first time yielded a purified monofunctional enzyme shikimate kinase II.15 The enzyme is monomeric and examination of its amino-acid sequence has shown a region that is homologous with other kinases and ATP-requiring enzymes.Shikimate kinase I1 has very different Michaelis constants from the isozyme shikimate kinase I suggesting that shikimate kinase I1 is the isozyme that normally functions in the biosynthesis of aromatic compounds. Metal ions preferably Mg2+ are necessary for its activity.16 1.7 EPSP Synthase The condensation of shikimate 3-phosphate (8) with phospho- enolpyruvate to produce 5-enolpyruvylshikimate 3-phosphate (EPSP) (9) is catalysed by EPSP synthase [3-phosphoshiki- mate 1-carboxyvinyltransferase; E.C. 2.5.1.191. This enzyme is inhibited strongly by the herbicide glyphosate.Adapta- tion of plants to this herbicide is accompanied by an over- production of EPSP synthase to ensure that the shikimate pathway functions normally. Further plants that are capable of achieving this have been identified. Cultures of a glyphosate- tolerant cell line of Petunia hybrida had 15-20-fold increased levels of the enzyme but the enzyme itself was still herbicide- sensitive." Similar results were found in cultured cells of a tomato hybrid Lycopersicon esculentum x L.peruvianum." The hybrid could tolerate 100 times the concentration of glyphosate that was needed to affect wild-type cells and when it was treated with the herbicide it produced 8-13 times as much EPSP synthase activity. Again the enzyme was glyphosate- sensitive.Although increased levels of EPSP synthase may be obtained by using plants that are adapted to the presence of glyphosate its isolation and subsequent purification is still rather tedious. Since the enzyme is predominantly found in the chloroplasts preliminary manipulation of the plant tissue to give a chloroplast fraction facilitates the purification of the enzyme.19 The inhibition of biosynthesis of aromatic amino acids by glyphosate is covered in detail in a recent review.2o 1.8 Chorismate Synthase The elimination of phosphoric acid from EPSP by the enzyme chorismate synthase [E.C. 4.6.1.43 yields chorismate (10). This enzyme has been characterized from microbial sources but had not previously been identified in a plant.An anaerobic chorismate synthase from pea (Pisum sativum) has now been detected in tissue extracts and in chloroplast preparations.21 The reaction that is catalysed is formally a trans-1,4-elimination but a 1,6elimination might be expected to proceed more favourably if the substituents were cis. An alternative mechanism that has been proposed is via an initial suprafacial 3,3-rearrangement [to form the allylic isomer iso-EPSP (12)] c 0; * co, I 1 OH bH EPSP (9) Chorisrnate (10) OH is0 -EPSP (12) Enzyme i chorismate synthase Scheme 3 followed by a trans-1,2-elimination (Scheme 3). Synthetic iso- EPSP (12) has been evaluated as a substrate for the chorismate synthase of Neurospora crassa. It proved to be an inhibitor of the enzyme and not an alternative substrate.22 1.9 Chorismate Mutase Chorismate mutase [E.C.5.4.99.51 catalyses the Claisen-like rearrangement of chorismate (10) to prephenate (1 l) trans- ferring the phosphoenolpyruvate-derivedside-chain so that this becomes directly bonded to the carbocycle and generates the basic skeleton of the phenylpropanoids. A monofunctional enzyme (CM-F) and a further isozyme (CM-P) the latter being a component of the bifunctional P-protein [chorismate mutase- prephenate dehydratase ; E.C. 5.4.99.5 and E.C. 4.2.1.51 are known to exist in prokaryotes. In plants two mono-functional isozymes (CM- 1 and CM-2) have been identified. Antibodies to these two isozymes from Sorghum bicolorx S. sudanensis have been raised,23 and the lack of any cross- reaction between the isozymes and their antisera indicates that there is no immunological similarity between the two isozymes from this source.CM- 1 is activated by tryptophan but inhibited by tyrosine and by phenylalanine whereas CM-2 is not regulated by any of these amino acids. During further studies on a wider range of plants24 it was demonstrated that the regulated form CM-1 was always present and that most plants also contain the unregulated form CM-2. Nevertheless the regulatory properties of CM- 1 were found to differ significantly between plants. The antisera that were raised for the Sorghum isozymes showed good cross-reactivity with the isozymes from maize (Zea mays) and some reactivity with those from barley (Hordeum vulgare) but all other species that were examined were antigenically distinct from Sorghum bicolor.1.10 Phenylalanine and Tyrosine The biosynthesis of the aromatic amino acids L-phenylalanine (14) and L-tyrosine (17) from chorismate may occur by several NATURAL PRODUCT REPORTS 1988 PL P + dNH2 Prephenate (11) Phenylpyruvate (13) L -Phenylalanine (14) (enz yme-bound1 - t. c0; I I iV PLP I OH OH Chorismate (10) Prephenate (111 L -Arogenate (15) (PLP = pyridoxal 5'-phosphate) viii kNH2 PLP OH OH 4 -Hydroxyphenylpyruvate (16) L -Tyrosine (17) Enzymes i chorismate mutase (monofunctional); ii chorismate mutase-prephenate dehydratase (bifunctional); iii phenylpyruvate aminotransferase ;iv prephenate aminotransferase ;v arogenatedehydratase ;vi arogenate dehydrogenase ;vii prephenate dehydrogenase ;viii 4-hydroxyphenylpyruvate aminotransferase Scheme 4 pathways (Scheme 4).The pathway that is used is dependent proceeds via arogenate.27 High arogenate dehydrogenase on the organism and sometimes several routes operate in activity could be detected but there was no evidence of a particular species. A study of several alkaloid-producing prephenate dehydrogenase activity thus indicating that trans-strains of the ergot-producing fungi Claviceps purpurea and amination of prephenate precedes aromatization. The aro-C. fusiformis indicated that all of the strains that were genate dehydrogenase was strongly inhibited by tyrosine but investigated utilize both arogenate (15) and phenylpyruvate unaffected by phenylalanine prephenate or tryptophan.The (13) as intermediates in the biosynthesis of phen~lalanine.~~properties suggest that this enzyme plays an important role in Thus the enzymes prephenate dehydratase [E.C. 4.2.1Sl] prephenate aminotransferase and arogenate dehydratase were detected. Tyrosine was preferentially or exclusively synthesized via the arogenate pathway as demonstrated by the presence of arogenate dehydrogenase in all strains and of prephenate dehydrogenase [E.C. 1.3.1.121 or prephenate dehydrogenase (NADP') [E.C. 1.3.1.131 in some. The enzymes of the arogenate pathway were insensitive to any feedback inhibition by the two aromatic amino acids. Cultured cells of Nicotiana silvestris contain arogenate dehydratase but not prephenate dehydratase thus establishing that phenylalanine in this species is derived from arogenate but not from phenylpyruvate.26 Arogenate dehydratase was also found in chloroplasts of spinach (Spinacia oleracea) and the enzymes from both plants were shown to be specific for L-arogenate as substrate being inhibited by L-phenylalanine but activated by L-tyrosine.The biosynthesis of tyrosine in sorghum (Sorghum bicolor x S. sudanensis) also the regulation of biosynthesis of tyrosine in sorghum and that arogenate could also function as a precursor of phenylalanine in this plant. The production of arogenate from prephenate by prephenate aminotransferase was also demonstrated in Sor-ghum bicolor x S. sudanensis.28The enzyme utilized glutamate as the amino donor though aspartate could serve as a less efficient substrate.The reaction was freely reversible and in the forward direction was unaffected by tyrosine phenylalanine or tryptophan. The sensitivity of prephenate dehydratase to feedback inhibi-tion by phenylalanine that had been demonstrated in a wild-type strain of Bacillus polymyxa was found to have disappeared in the enzyme from a mutant that is resistant to analogues of ~henylalanine.~ The deregulation and increased activity of the enzyme led to overproduction of phenylalanine in the culture. The presence or absence of enzymes that are involved in the conversion of chorismate into phenylalanine and tyrosine in NATURAL PRODUCT REPORTS 1988 -P.M.DEWICK Shiki mate NIH shift C horismate X OH Prep hen ate co2-H?f I t)H bH L -Arogenate (15) Spiro-arogenate (181 Scheme 5 several pseudomonads has been used to indicate how the various species may have In this respect changes in regulatory properties also become important markers. The biosynthesis of the aromatic amino acids and the role of arogenate are discussed in recent reviews.3o* 31 A spiro-lactam variant of arogenate called spiro-arogenate (18) had been identified in cultures of a mutant Neurospora crassa though the significance of this metabolite remains to be established. The origins of this compound have been studied further in experiments in which labelled shikimate was supplied to the mutant.32 Label quickly appeared in various related metabolites in the sequence chorismate prephenate and finally arogenate and then labelled spiro-arogenate appeared but only after a delay of several hours.This establishes a likely biosynthetic sequence (Scheme 5) but suggests that the postulated ‘arogenate spirase ’ enzyme may have a relatively low affinity for arogenate. A further route to tyrosine is by direct aromatic hydroxylation of phenylalanine. The sequence is employed particularly by mammalian systems. Phenylalanine hydroxylase [phenylalanine 4-mono-oxygenase;E.C. 1.14.16.13 requires molecular oxygen as the oxygen source and a tetrahydropterin as cofactor. The relative activities of this enzyme in the presence of tetra-hydrobiopterin (as the natural cofactor) or of 6-methyltetra- hydropterin were markedly different depending on whether rat liver or rat kidney was the source of the The enzymes of kidney and liver appear to be in different states of activation (probably a dimer for the highly active kidney hydroxylase) and may be regulated in different ways.Phenyl- alanine hydroxylase from Chromobacterium violaceum con-tains a stoicheiometric amount of The hydroxylation step is accompanied by an NIH shift of the 4-hydrogen to position 3 or 5 and this phenomenon has now been detected in Man.35 Oral doses of ~-[ring-~H,]phenylalanine were admin- istered and the concentrations of labelled L-phenylalanine and L-tyrosine in the plasma were detected via mass spectro- metry. The distribution of 2H label in tyrosine was calculated to be consistent only with the NIH shift mechanism followed by random loss of one of the substituents from position 3 or 5 (Scheme 6).The formation of tyrosine and the related o-tyrosine and m-tyrosine from L-phenylalanine in rats has been shown 6 @H X OH OH Scheme 6 co; NANPI + PH H’z 0 (19) [20) Enzyme 4-aminobenzoate hydroxylase Scheme 7 OH NH* (21) to depend on hydroxylases but phenylalanine hydroxylase catalyses only the 4-hydroxylation and not the 2-or the 3-hydro~ylation.~~ The isolation purification and properties of phenylalanine hydroxylase3’ and the mechanism of action of the enzyme38 have been reviewed. 1.11 Anthranilic Acid p-Aminobenzoic Acid and Related Compounds Anthranilate (o-aminobenzoate) is an intermediate in the biosynthetic pathway to L-tryptophan and is derived from chorismate in a reaction that is catalysed by anthranilate synthase [E.C.4.1.3.271. Cell lines of tobacco (Nicotiana tab- acum) contain two forms of the enzyme one being resistant to feedback inhibition by tryptophan and the other being almost completely inhibited by low levels of the amino acid.39 p-Aminobenzoate (19) derives from chorismate by a similar sequence to that which yields anthranilate (see ref. 1). Further metabolism gives the 4-hydroxyaniline moiety of N-(L-glutam- 5-yl)-4-hydroxyaniline (2 l) which occurs in high concentration in the fruiting body of the edible mushroom (Agaricus bisporus). A mono-oxygenase 4-aminobenzoate hydroxylase which catalyses the conversion of (19) into 4-hydroxyaniline (20) has been identified.40 The decarboxylative hydroxylation (Scheme 7) requires either NADH or NADPH and molecular oxygen is needed as the source of the oxygen; the enzyme is FAD-dependent.Other substituted benzoates with free amino-groups and carboxyl groups in an ortho or para relationship (e.g. anthranilate or 4-aminosalicylate) would serve as sub- strates but in these cases the hydroxylation was accompanied by the formation of hydrogen peroxide. The reaction is analogous to that catalysed by salicylate hydroxylase [salicyl- ate 1-mono-oxygenase ;E.C. 1.14.13. I] but 4-aminobenzoate hydroxylase does not act on salicylate. CH20@ Ho-%hO H “2 02 QE) -Ql3 H H (22) L -Tryptophan (23) QoH (241 (cc and 0 refer to subunits of tryptophan synthase) Scheme 8 E H Internal aldimine L -Serine form of enzyme -bound PLP E + 0 @A-N +/ H NHZ H NATURAL PRODUCT REPORTS 1988 2 Tryptophan and Related Compounds 2.1 Tryptophan Synthase Tryptophan synthase [E.C.4.2.1.201 is a multi-enzyme complex (the a2p complex) comprising two a subunits and a p2dimeric subunit. This enzyme catalyses the final reaction in the biosynthesis of tryptophan [i.e. the synthesis of L-tryptophan (23) from L-serine and indole-3-glycerol phosphate (22)] and both the aldolytic cleavage of indole-3-glycerol phosphate to indole (24) and the formation of tryptophan from L-serine and indole (Scheme 8).The second and third reactions are also catalysed by the a and p2 subunits respectively of tryptophan synthase. The replacement of the P-hydroxyl group of L-serine with indole which is catalysed by the p subunit requires pyridoxal 5’-phosphate and the distinctive ultraviolet-visible spectral properties of this compound provide a sensitive probe for the detection and identification of intermediates during the reaction. In continuation of earlier work (see ref. l) but using more refined techniques the reaction of L-serine with indole that is catalysed by the a,P2 complex from Escherichia coli has been studied via rapid-scanning stopped-flow ultraviolet-visible ~pectroscopy.~~ These studies have confirmed the reaction sequence that was proposed previously (Scheme 9) and have indicated both that the a-aminoacrylate Schiff s-base intermediate (25) is highly reactive towards indole and that the steps in the sequence before this one limit the rate of the reaction.Binding of indole alters the spectrum of the a-aminoacrylate prior to the formation of a covalent bond. Monoclonal antibodies have been used during investigations of the conformational effects of the binding of ligands on the p2 Extended conformational rearrangements of the protein are brought about by fixation of the coenzyme pyridoxal 5’-phosphate and the substrate L-serine. The associ- ation of the a and p subunits is also accompanied by 0- H enzyme-bound oc-aminoacrylate Schiff’s base with PLP (25) lndole I L -Tryptophan H Scheme 9 NATURAL PRODUCT REPORTS 1988 -P.M. DEWICK H ti L-Tryptophan (23) Indole (24) Scheme 10 H H H H H Schiff's base of (35)-5-f I uoro-2,3 -dihydro-L-Trp (26) Scheme 11 H CH CO HQrzJH C HZCH CO HI (2 8) IAA (29) important conformational changes on the p2 subunit. This brings about an increase in tryptophan synthase activity and abolishes the serine deaminase activity of the p2 subunit. Although tryptophan synthase catalyses a number of pyridoxal-5'-phosphate-dependent/?-elimination reactions and /?-replacement reactions that are also catalysed by trypto- phanase [E.C. 4.1.99.11 a principal and puzzling difference between the two enzymes has been the apparent inability of tryptophan synthase to catalyse the p-elimination of indole from L-tryptophan.It has now been shown for the first time that the /?, subunit and the a2P2complex of tryptophan synthase from E. coli and from Salmonella typhimurium do catalyse a slow /?-elimination reaction (Scheme 10) which produces indole pyruvate and ammonia.43 The rate of reaction was significantly higher in the presence of the a subunit and could be increased by enzymically removing the products of the reaction (by supplying substrates and thus exploiting the other catalytic activities of tryptophan synthase). That the reaction was not due to contaminating tryptophanase was demonstrated by using specific inhibitors of the enzymes. (3R)- 2,3-Dihydro-~-tryptophan, which is a specific inhibitor of tryptophanase had no effect but (3S)-2,3-dihydro-~-trypto-phan which is a specific inhibitor of tryptophan synthase completely inhibited the reaction.The cleavage reaction was also inhibited by D-tryptophan which is the product of a slow racemization reaction. It was concluded that tryptophan synthase is similar to tryptophanase in its reaction mechanisms and specificity but several of the reactions are catalysed at very different rates. To explore aspects of the reactions that are catalysed by the a2,8 complex of tryptophan synthase from Escherichia coli several 5-fluorinated substrates and reaction-intermediate ana- logues were adaed to the enzyme and their fates were followed by 19Fn.m.r.meas~rements.~~ Tryptophan synthase was shown to bind the (3S)-diastereoisomers of both 5-fluoro-2,3-dihydro- D-tryptophan and 5-fluoro-2,3-dihydro-~-tryptophan both specifically and tightly but it bound 5-fluoro-~-tryptophan more tightly than 5-fluoro-~-tryptophan. Unexpectedly a slow isomerization of 5-fluorotryptophan of tryptophan and of (3$)-5-fluoro-2,3-dihydrotryptophanwas detected. The reac- tion was extremely slow being lo3 to lo5times slower than the /?-replacement and /?-elimination reactions that are catalysed by tryptophan synthase and probably has no biochemical sig- nificance in vivo. Whether tryptophan synthase itself serves a catalytic role in the isomerization or the enzyme simply binds the substrate and pyridoxal 5'-phosphate so that chemical isomerization of the Schiff s base (26) may occur is not known (Scheme 11).A novel tryptophan analogue 3-(indazol- 1 -yl)-L-alanine (27) was produced when the tryptophan synthase a2/?, complex from E. coli was supplied with the indole analogue indazole (28).45This is the first example of a /?-replacement reaction that is catalysed by tryptophan synthase but which occurs at a position other than position 3 of indole analogues. Several reviews relating to the biosynthesis of tryptophan have been published including descriptions of genetic aspects,46 ~egulation,~' and conformational studies on tryptophan synthase.48 2.2 Indole-3-acetic Acid and Related Metabolites of Tryptophan Indole-3-acetic acid (IAA) (29) is an essential plant growth hormone (auxin).It is derived in Nature by oxidative NATURAL PRODUCT REPORTS 1988 (30) R = GIC (31) R= NHCHC0,H I CH,CO H H (34) R = H (35) R =OH Indole-3 -acet aldox ime / Indole-3- acet amide -H (32) R = AC (331 R =H bNH";"'02~ C H CO H H (36) R =H (37) R = GIC Tryptophan I I ndole -3 -pyr u v ic acid Tryptamine I Indole-3- acetaldehyde Indole -3 -e t ha n o I (Tryptophol) Indole -3-acetic acid Indole-3-methanol Indole -3-car balde h yde Indole-3- car box y lic acid Scheme 12 metabolism from tryptophan. Several pathways may operate via the intermediates indole-3-pyruvic acid tryptamine indole- 3-acetamide or indole-3-acetaldoxime (Scheme 12).Indole-3- acetic acid may be subjected to further metabolism and can yield a wide variety of indole and oxindole derivatives. Labelling studies using chloroplast fractions from pea (Pisum sativum) have shown that indole-3-methanol is a major catabolite in this In leaf segments of wheat (Triticum aestivum) 14C-labelled IAA gave glucosides of indole-3-methanol and indole- 3-carboxylic acid as did labelled indole-3-methanol when it was admini~tered.~~ Conjugates of IAA such as indole-3-acetyl-P-D-glucose (30) and indole-3-acetylaspartic acid (3 l) may also be formed.51 In addition a non-decarboxylative pathway that leads to a number of oxindole derivatives exists in T. aestivum. The concentration of IAA increases after wheat has been infected with the rust fungus Puccinia graminis and this may be due to a contribution from the fungus itself.Studies of non-transformed and Agrobacterium-tumefaciens-transformed cells of tobacco (Nicotiana tabacum) have shown that labelled tryptophan is converted into indole-3-acet-aldoxime and indole-3-ethanol in both cell types but indole-3- acetamide occurs exclusively in the transformed cells.52 81 NATURAL PRODUCT REPORTS 1988 -P. M. DEWICK dOH HO" I I AH OH OH Quinate (7) 3 -Dehydroquinat e (4 1 3-Dehydroshi kimate (5) Shikimate (6) 6 HO \ OH OH OH Protocat echuate (38) Gallate (39) Enzymes i quinate dehydrogenase ;ii 3-dehydroquinate dehydratase ;iii shikimate dehydrogenase ;iv 3-dehydroshikimate hydro-lyase Scheme 13 OH HO OH 11 0 0 -Glucogallin (401 Indole-3-acetaldoxime yielded IAA and indole-3-ethanol in both tissues with indole-3-acetamide again occurring only in transformed cells.Labelled IAA was rapidly conjugated with aspartic acid to give indole-3-acetylaspartic acid though with increased efficiency in the transformed cells. The overall results suggested that in the transformed cells auxin genes from the Agrobacterium are integrated into the host altering the host's capacity for both biosynthesis and conjugation of IAA and thus producing the characteristic tissue of a crown gall tumour. In Pseudomonas syringae pv. savastanoi which is the causative agent of olive knot disease and oleander knot disease IAA is derived from tryptophan via indole-3-acetamide.Water- soluble lysine conjugates e.g. (32) and the previously reported compound (33) have been detected in cultures of strains of the pathovar from oleander though not from those that affect olive.53 A main catabolite of IAA in seedlings of maize (Zea mays) is the oxindole derivative oxindole-3-acetic acid (34).54 In roots of broad bean (Viciafaba) the formation of conjugates of 3- hydroxyoxindole-3-aceticacid (dioxindole-3-acetic acid) (35) appears to be a major route of metabolism of IAA.55 Over 70 % of the activity from labelled IAA was recovered in the two aspartic acid conjugates (36) and (37) after a period of 24 hours. The metabolic sequence to these compounds was investigated by feeding experiments in which labelled precursors were Indole-3-acetylaspartic acid (3 l) but not di-oxindole-3-acetic acid (35) was found to be a precursor of dioxindole-3-acetylasparticacid (36) and double-labelling experiments proved that the conjugate was incorporated intact.The hydroxy-compound (36) was glucosylated to (37) in cotyledons of V.faba though the transformation was very slow in its roots. The glucoside appears to be formed in the cotyledons and is then transported to the roots. The biosynthesis and regulation of the formation of IAA has been re~iewed.~' 3 Phenols and Phenolic Acids Some simple phenolic acids are derived directly from 3-dehydroshikimate (which is an intermediate on the shikimate pathway) and retain the full carbon skeleton of this acid. Studies of biosynthesis in young seedlings of Pinus sylvestris have shown that protocatechuic acid (3,4-dihydroxybenzoic acid) vanillic acid (4-hydroxy-3-methoxybenzoicacid) and gallic acid (3,4,5-trihydroxybenzoicacid) are all formed from either [14C]quinate or [14C]shikimate precursor^.^^ The presence of key enzymes of the shikimate pathway e.g.quinate dehydrogenase shikimate dehydrogenase 3-dehydroquinate dehydratase and 3-dehydroshikimate hydro-lyase was also demonstrated. Protocatechuic acid (38) and gallic acid (39) were produced when quinic acid was treated with a quinate dehydrogenase preparation from larch (Larix sibirica). l4 The shikimate dehydrogenase activity had been removed from this preparation so these acids result from the action of other enzymes on 3-dehydroquinate and 3-dehydroshikimate.Trace amounts of these intermediates could be demonstrated to be formed by labelling studies (Scheme 13). An enzyme that has been isolated from leaves of oak (Quercus robur) catalyses an exchange reaction between P-glucogallin (1 -0-galloyl-P-D-glucose) (40) and free glucose. 59 This acyltransferase achieves no overall chemical change but if ['4C]glucose was supplied to the preparation it was incorporated into the P-glucogallin. Gallic acid itself was not exchanged and other potential sugar donors (such as @-glucose 1-phosphate a-D-glucose 1-phosphate D-glucose &phosphate gentiobiose or sucrose) were not accepted. However the enzyme would react with benzoylglucose p-coumaroylglucose and sinapoyl- glucose.The role if any of this acyltransferase in the biosynthesis and metabolism of gallotannins is unknown but 82 its use provides a very convenient method to prepare labelled /3-glucogallin and related 1-0-acyl esters. The biosynthesis of hydroxybenzoic acids and the role of gallic acid in secondary metabolism form the subject of a detailed review.6o 4 Phenylpropanoids 4.1 Phenylalanine Ammonia-lyase The enzyme phenylalanine ammonia-lyase (PAL) [E.C. 4.3.1.51 catalyses the elimination of ammonia from L-phenylalanine producing trans-cinnamic acid and effectively controls the flow of material from the shikimate pathway into many important phenylpropanoid-based secondary metabolites. Recent work has shown that purified enzyme preparations from cell suspension cultures of French bean (Phuseolus vulgaris) that had been exposed to a polysaccharide elicitor from the cell walls of the phytopathogenic fungus Col-letotrichum lindemuthianum are inherently unstable.61 The native subunit (M 77000) breaks down to yield partial degradation products (M 70000 and 53000) both in vitro and in vivo.This degradation was noted with four different forms of the enzyme that could be isolated from P.vulgaris. The four forms are resolvable and are characterized by different pl values. The elicitor-induced PAL activity in P. vulgaris cells OH I PhC H,CH CO,H PhCH,CH-P=O I It ONH NH2 OH L -AOPP (41) APEP (42) HO Ho*o&oH \ OH + OH 0 NATURAL PRODUCT REPORTS 1988 declines rapidly if the cells are subsequently treated with the reaction product trans-cinnamic acid.62 Other demonstrated inhibitors of PAL from germinating seeds of lettuce (Lactuca sativa) include D-phenylalanine 4-fluorophenylalanine p-phenyl-lactic acid and ~-tryptophan.~~ The synthetic L-phenylalanine analogue ~-a-amino-oxy-,8-phenylpropionic acid (L-AOPP) (41) also causes significant inhibition of PAL activity at very low concentrations.Other evidence from experiments in which cell cultures of Cryptomeriu juponica and Perilla frutescens var. crispa were used suggests that this compound may inhibit the formation of PAL although it is usually considered to be a competitive inhibitor of the enzyme.64 Phenylalanine ammonia-lyase from buckwheat (Fagopyvum esculenturn) has been used to test other synthetic analogues for potential inhibitory action.65 The phosphonic analogue (1 -amino-2-phenylethyl)phosphonic acid (APEP) (42) was found to competitively inhibit buckwheat PAL the (R)enantiomer being more effective than the (5') form.The corresponding phosphonous analogues were less inhibitory. If APEP was applied to hypocotyls of buckwheat seedlings it inhibited the synthesis of anthocyanins and caused an increase in the concentration of phenylalanine. Seedlings of kohlrabi developed normally in the presence of APEP though their anthocyanin content was greatly reduced. The isolation and subsequent purification of the enzyme from developing fruit of oranges (Citrus sinensis) has been described.66 Specific inhibitors of PAL are described in a recent review.2o 4.2 Cinnamic Acids and Esters Hydroxycinnamic acids are in most cases obtained by aromatic hydroxylation of the cinnamic acid that is formed by the action of PAL gradually building up the oxygenation pattern and 0qnC02H -dOH OH AH (431 Quinic acid (44) OH Chlorogenic acid (45) Scheme 14 H H o s o d OOMe HO OH 0 (47) O h O\ MOH e OMe IL6) OMe Me,:-oH 0 Sinapine (48) Choline (49) NATURAL PRODUCT REPORTS 1988 -P.M. DEWICK giving the commonly encountered 4- 3,4- and 3,4,5-oxy- genated products. There is ample evidence to demonstrate that 3-hydroxylation of 4-hydroxy-substituted phenylpropanoids occurs but surprisingly the process has not been well studied at the enzymic level.In a recent preparations from tubers of potato (Solanum tuberosum) have been shown to catalyse the 3-hydroxylation of 4-hydroxyphenylpropanoid carboxylic acids including p-coumaric acid and tyrosine ; NADH (or NADPH) and FAD (or FMN) were necessary cofactors. Among a range of 4-hydroxylated C,C and C,C compounds that were tested only 4-hydroxyphenylacetic acid and p-cresol were hydroxylated. The hydroxylase showed some features of phenolase hydroxylation and is probably concerned in the biosynthesis of chlorogenic acid in potato. Further modification of cinnamic acids often requires an activated form of the acid to be produced initially and glucose esters are frequently encountered in this role as well as esters of coenzyme A.The biosynthesis of chlorogenic acid (5-0-caffeoylquinic acid) (45) involves a transesterification reaction between 1-0-caffeoyl-@-glucose (43) and quinic acid (44) (Scheme 14). The enzyme that catalyses this reaction in roots of sweet potato (Ipomoea batatas) has been purified and its substrate specificity investigated.,’ Although the enzyme demonstrated strict specificity towards compounds that are related to quinic acid as acceptors (shikimic acid was also transformed but less effectively than quinic acid) it had broad substrate specificity towards hydroxycinnamoyl-D-glucoses as donors. 1-0-Cinnamoyl-D-glucose I -0-p-coumaroyl-~-glucose and I -0-caffeoyl-D-glucose were all readily trans-formed into the corresponding products.The hydroxycinnamoylglucose substrate may however function as both the acyl donor and the acyl acceptor and such a process is observed in the biosynthesis of 1,2-di-O-sinapoyl-P-D-glucose (46) in seedlings of radish (Raphanus sativus var. sativu~).~~ The enzyme which is a hydroxycinnamoyl-transferase catalyses the formation of (46) from two molecules of 1-0-sinapoyl-P-D-glucose (47) showing strict specificity of transfer to the 2-hydroxyl of the acceptor. However it has broad substrate specificity towards glucose esters of phenyl- propane acids transforming esters of sinapic ferulic and p- coumaric acids yet shows no activity towards glucose esters of C,C acids e.g. 1 -0-benzoylglucose and 1-0-galloylglucose.Sinapine synthase enzymes have been isolated from seeds of radish (Raphanus sativus var. sativus) and of mustard (Sinapis aIba).’O Sinapine (48) is the choline ester of sinapic acid and the hydroxycinnamoyltransferase sinapine synthase [sinapoyl- CO H + HO-CH / C(0)SCoA ‘CO H (50)R =H Tartronic acid (53) (51) R =OH (52)R =OMe + HSCoA (54)R =H (55)R =OH (56)R =OMe Scheme 15 glucose-choline sinapoyltransferase ; E.C. 2.3.1.911 uses 1-0-sinapoyl-6-D-glucose (47) as the acyl donor and choline (49) as the acceptor. The purified enzymes from these two sources had similar properties and though sinapoylglucose was the favoured substrate ferulic and p-coumaric esters of glucose were also transformed. Acyl donors that were not acceptable included the 6- and 3-0-sinapoylglucoses 1-0-benzoylglucose and 1-0-galloylglucose.Conjugates of p-coumaric acid caffeic acid and ferulic acid with tartronic acid (53) are present in young plants of mung bean (Phaseoh radiatus L. [syn. P. aureus Roxb. and Vigna radiata (L.) R. Wilczek]). In contrast to the above examples glucose esters are not involved in the formation of the tartronates (54F(56) but instead the coenzyme A esters (50)-(52) feature as the activated cinnamic acids (Scheme 15).71 Thus no products were formed when an enzyme preparation was incubated with 1-0-(p-coumaroy1)glucose but p-coumaroyl and caffeoyl coenzyme A esters were utilized. Assessment of the substrate specificity for the reaction was facilitated by exploiting its freely reversible nature.No reaction was observed with esters of tartaric malic or quinic acid. Seedlings of Stachys albens and other Stachys species synthesize a number of caffeic acid esters including chlorogenic acid (49 verbascoside (57) and stachyoside (58). Feeding experiments using leaves of S. albens have demonstrated that phenylalanine and cinnamic acid are incorporated into the caffeoyl moiety of all three compounds with the 3,4-dihydroxyphenylethyl moiety of (57) and (58) arising from tyrosine or better from t~ramine.~’ Incorporation data indicate that stachyoside probably arises by further glucosylation of verbascoside and the results are consistent with earlier studies on the biosynthesis of verbascoside in lilac (Syringa vulgaris) (Scheme 16).How tyramine is transformed into the dihydroxyphenylethyl portion of (57) and (58) has yet to be established. However the tyrosine decarboxylase [E.C. 4.1.1.251 in cell cultures of Syringa vuIgaris appears to be strongly inhibited by the phenylalanine analogue L-AOPP as is PAL (see Section 4.1).73 It is suggested that there may be some metabolic co-ordination between the two convergent pathways that lead to verbascoside and related compounds. 4.3 Phenylpropane and Phenylethane Derivatives Genetic analysis of hybrids of Perilla frutescens has been undertaken to outline the biosynthetic sequences leading to a number of allylbenzenes that are produced by this plant.74 A Phenylalan ine Ty rosine Cinnamic acid Tyramine 0 OH R OH Verbascoside (57) R = Rha Stachyoside (58) R = Rha-Glc Enzymes i phenylalanine ammonia-lyase ; ii tyrosine decarboxylase Scheme 16 NATURAL PRODUCT REPORTS I988 u Me0 / Dillapiole (611 Methyleugenol (59) OMe MeoooMe Elemicin (62) Scheme 17 pathway that was proposed many years ago (Scheme 17) appears to be operative since biosyntheses of dillapiole (61) and elemicin (62) are controlled by two independent genes expressing for reactions a and b respectively.Only myristicin (60) is produced in the absence of both of these dominant genes. The role of methyleugenol (59) is unfortunately still hypothetical because it has not yet been detected in Perilla OH OH frutescens. Homogentisic acid (63) is a tyrosine-derived phenylethane (16) Homogentisic acid (63) metabolite that is formed by the action of 4-hydroxy-phenylpyruvate dioxygenase [E.C.1.13.1 1.271 on 4-hydroxy- Enzyme i 4-hydroxyphenylpyruvate dioxygenase phenylpyruvate (16) (Scheme 18). Whilst many features of this Scheme 18 complex conversion are being resolved (see ref. l) a new piece of information has recently been presented. 75 A spectroscopic investigation of the non-haem iron-containing dioxygenase from a species of Pseudornonas has shown the presence of tyrosine that is co-ordinated to iron in the active site. The PhycoZH enzyme thus belongs to the class of iron-tyrosinate proteins NH2 but the precise function of the tyrosine has yet to be established. During biosynthesis of the antibiotic virginiamycin S by Streptornyces virginiae L-phenylalanine is incorporated by way of L-phenylglycine (64).In theory the transformation of phenylalanine into (64) could occur via an intermolecular P h w CO H transfer of nitrogen e.g. as shown in Scheme 19 or via some Ph-Co2H intramolecular pro~ess.'~ That an intermolecular pathway is 0 involved has been confirmed by a feeding experiment with DL-[3-13C l5N]pheny1alanine. Carbon- 13 n.m.r. analysis of the phenylglycine portion showed no labelled nitrogen remaining [ -COzI and it has been suggested that the rearrangement (Scheme 19 path b) resembles that which occurs during the formation of p-tyrosine from tyrosine. 0 The chromenes demethoxyencecalin (65) O-demethyl-PhE C 02 H encecalin (66) and encecalin (67) appear to be related by a logical biosynthetic sequence that involves hydroxylation and then methylation based on observations of the levels of these compounds in Ageratina adenophora during the plant's t ransami not\ion development and also from feeding experiments with unlabelled material^.^^ That the compounds may be derived from shikimate is suggested because of the massive reduction in the total concentration of chromenes if the plant is treated with glyphosate (an inhibitor of EPSP synthase; see Section 1.7).(641 However based on earlier studies with related compounds the acetophenone moiety appears to be more likely to be acetate- Scheme 19 derived than shikimate-derived. NATURAL PRODUCT REPORTS 1988 -P.M.DEWICK 4.4 Coumarins A series of feeding experiments in which shoots of Daphne mezereum were used have been conducted to investigate the biosynthesis of daphnetin (7,8-dihydroxycoumarin) (69).78 Umbelliferone (68) was incorporated more efficiently than p-coumaric acid and caffeic acid was poorly utilized. The findings support the idea of umbelliferone being a rather general precursor of other plant coumarins that bear further oxygen R (65) R = H Umbelliferone(68) R = H (66) R = OH Daphnetin (69) R = OH (67) R = OMe Meorno HO \ Scopolet in (70) Ayapin (71 1 GlcO \ Meorno Scopolin (72) Scheme 20 substituents on the aromatic ring rather than that the additional substitution occurs at the cinnamic acid stage.Daphne mezereum also contains skimmin which is the 7-0-glucoside of daphnetin and it has been suggested that glucosylation occurs at the coumarin stage. Coumarins accumulate in stem sections of sunflower (Helianthus annuus) after they have been infected with fungi and although scopoletin (70) and scopolin (72) may be detected in uninoculated tissue their concentrations increase markedly in infected stems. Ayapin (71) is found only in infected plant^.'^ If labelled scopoletin was supplied to stem sections of Helminthosporium-carbonum-infected sunflower it was incorporated into the glucoside scopolin (72) and par- ticularly efficiently into the methylenedioxy-derivative ayapin (Scheme 20). Both ayapin and scopoletin were rapidly degraded by the sunflower pathogen Alternaria helianthi.(E)-2-Hydroxy-4-methoxycinnamic acid (73) has been isolated from Artemisia dracunculus and has been proposed as the precursor of herniarin (74) in this plant.*O Indeed this cinnamic acid is unstable in ultraviolet light and rapidly cyclizes to herniarin. Compound (73) could not be detected in Lavandula oficinafis when studies on the biosynthesis of herniarin were first conducted some years ago. Many natural phenolic derivatives contain a furano sub- stituent that is known to be derived by loss of a three-carbon unit from an isopropyldihydrofurano-system which itself arises by cyclization of a prenyl substituent with an ortho-hydroxyl group (see also Section 5.8). Thus (+)-marmesin (75) has been shown (from incorporation studies) to be a precursor of the furanocoumarin psoralen (76) (Scheme 21).Microsomal fractions from cell suspension cultures of parsley (Petroselinum crispum) which had been challenged with an elicitor from either Alternaria carthami or Phytophthora megasperma fsp. gfycinea have now been shown to catalyse the conversion of (+)-marmesin into psoralen.81 The system requires NADPH and molecular oxygen as cofactors the reaction being catalysed by an elicitor-induced cytochrome-P-450-dependent enzyme psoralen synthase. Although 14C-labelled (& )-marmesin was used in the incubation experiments dilution experiments with either (+)-or (-) -marmesin demonstrated that (+)-marmesin was in fact the substrate; (-)-marmesin was not converted.The observations support the hypothesis that a 3’-hydroxylation step may be involved perhaps via the mechanism in Scheme 21 if this hydroxyl group is cis to the hydroxyisopropyl group. The cis configuration is found in most naturally occurring examples of such compounds. A 3’-hydroxylated intermediate may not accumulate if hydroxylation is the rate-limiting step ;indeed no r Me0m0 MeO\ COHO Z H \ intermediates between marmesin and psoralen were detectable. However an unidentified product that was neither a precursor (731 Herniarin (74) nor a product of psoralen was isolated from incubations with Psoralen (76) Scheme 21 86 Alternaria-induced microsomes but not from Phytophthora- induced preparations. Methylation of furanocoumarins is a further transformation that has been encountered in elicitor-stimulated parsley cells.Cultured cells that had been treated with elicitor from Phytophthora megasperma f.sp. glycinea contained two 0-methyltransferases which catalyse the S-adenosylmethionine- dependent methylation of xanthotoxol(77) to xanthotoxin (78) and of bergaptol (79) to bergapten (80) respectively.82 The 5 R Xanthotoxol (77)R =OH Ber gapto‘ (79) = OH Xanthotoxin (78)R =OMe Bergapten(80) R =OMe R U3Xo OMe 5-Hydroxyxanthotoxin (811R =OH Isopimpinellin (82) R =OMe NATURAL PRODUCT REPORTS 1988 latter enzyme was also shown to catalyse the methylation of 5-hydroxyxanthotoxin (8 1) to isopimpinellin (82). Simple coumarins were either not converted or very poor substrates and the utilization of 5,8-dihydroxypsoralen and 8-hydroxy- bergapten by either enzyme was only minimal.The activities of both enzymes showed transient increases when the elicitor was applied and these occurred a few hours later than the increases in PAL and 4-coumarate-CoA ligase activities. A review on the biosynthesis of plant coumarins has been published.83 4.5 Lignins Infection of leaves of wheat (Triticum aestivum) with a biotic elicitor that can be isolated from the rust fungus Puccinia graminis f.sp. tritici results in lignification preceded by an increase in PAL This increase in enzyme activity is then accompanied by increases in the activity of other enzymes of the general phenylpropanoid pathway and of the specific pathway of lignin biosynthesis including 4-coumarate-CoA ligase cinnamyl-alcohol dehydrogenase and peroxidase.Lignification is assumed to occur exclusively by oxidative polymerization of the (E) monomers p-coumaryl alcohol (83) coniferyl alcohol (84) and sinapyl alcohol (85) (monolignols). The isolation of the (2)-monolignols cis-coniferyl alcohol (86) and cis-sinapyl alcohol (87) from bark of beech (Fagus grandif~lia)~~ suggests that lignification might also involve the (Z) monomers. Polymeric lignins that were derived by incubating H202 and peroxidase with either (9-or (2)-coniferyl alcohol appeared to be identical. Thus both (E)-and (2)-monolignols may be used for lignification or alternatively the lignifying enzymes are indeed highly specific and (9-monolignols accumulate in beech bark because they are not suitable substrates.4.6 Lignans It has been shown that a range of tumour-inhibitory aryltetralin lignans in Podophyllum hexandrum can be subdivided bio- synthetically into two groups. One group contains a 3,4,5-trimethoxy-substituted pendent aromatic ring and its members are derived from desoxypodophyllotoxin (88). Members of the other group contain a 4-hydroxy-3,5-dimethoxy-substituted pendent ring and are derived from 4’-O-demethyldesoxy- KHzoH R \ OMe podophyllotoxin (89). The major lignans podophyllotoxin (90) OH OH and 4’-O-demethylpodophyllotoxin(91) are produced by hydroxylation of (88) or (89) at C-4 (Scheme 22). Further (83) R’ = R2 = H (86) R =H feeding experiments in which intact plants of P.peltaturn were (84) R’ = OMe R2 = H (87) R = OMe used,86 have confirmed earlier suggestions that the peltatins (85) R’ = R2 = OMe P-peltatin (92) and a-peltatin (93) are derived from (88) and ?H ____) OR OR OR O-Peltatin (92) R =Me Desoxypodophyllotoxin (88)R = Me Podophyllotoxin(90) R = Me a-Peltatin (93) R = H (89) R = H L’-O-Demethylpodophyllotoxin(91) R = H Scheme 22 NATURAL PRODUCT REPORTS 1988 -P. M. DEWICK OH OH Matarresinol (94) (95) t;’ Me0Met&o \ OMe +OMe 0 Yatein (96) (97) A nh ydr opodor hi zol (98) t;‘ H OMe OMe (88) Podorhizol (99) Podop hyllot oxin Scheme 23 (89) respectively by aromatic hydroxylation at C-5. The incorporation of desoxypodophyllotoxin (88) into podophyllo- toxin (90) and P-peltatin (92) was also demonstrated to occur in P.hexandrum. Intermediates in the biosynthesis of podophyllotoxin prior to desoxypodophyllotoxin have been investigated by feeding ex- periments using P.hexandrum plants that involved a range of dibenzylbutyrolactone lignans that are structurally related to the Podophyllum lignans and known to co-occur with aryl- tetralin lactone lignans in various planks7 Yatein (96) proved to be a satisfactory precursor; the corresponding cis-isomer (100) was also incorporated to a lower extent probably via yatein. Podorhizol (99) epipodorhizol (10I) and anhydro- podorhizol (98) were not incorporated into podophyllotoxin despite the demonstrated presence of both podorhizol and anhydropodorhizol in P.hexandrum plants. A biosynthetic sequence (Scheme 23) from yatein to podophyllotoxin via a key quinone methide intermediate (97) which can cyclize to NATURAL PRODUCT REPORTS 1988 H OMe OMe (100) OH 0 (2s)-Nor ingenin (102) R = H Dihydrokaempferol (104)R = H (1061 (2Sl-EriodictyoL (103)R =OH Dihydroquercetin (105) R = OH KaempferoI (107) R = H Quercetin (108) R = OH desoxypodophyllotoxin (88) has been proposed. This inter- mediate is probably also the percursor of both podorhizol (99) and anhydropodorhizol (98) by addition of water or by loss of a proton respectively. It is known that the 4-0-methyl series and the 4’-0-demethyl series of Podophyllurn lignans are formed separately from some common precursor.This is logically suggested to be matairesinol (94) which could be modified to either yatein (96) or 4’-0-demethylyatein (95) prior to cyclization giving the two groups of aryltetralin lactones. That the pattern of trimethoxy-substitution of podophyllotoxin is derived by a sequence that does not involve the hydroxy- dimethoxy pattern (as in 4’-0-demethylpodophyllotoxin) is demonstrated by an analysis of the labelling in these methyl substituents for samples of (90) and of (91) that had been obtained from feeding [S-rnethyl-14C]methionine. Quite differ- ent labelling patterns were observed and it was concluded from these that the branch-point compound will be either 4’- hydroxy- 3'-methox y-or 4’ 5’-dihydroxy-3'-methox y-substituted in the aromatic ring that ultimately becomes the pendent aryl group.The biosynthesis of lignans is briefly covered in a review of the chemistry of lignans.88 5 Flavonoids 5.1 General Aspects An overview of flavonoid biosynthesis has been published.8g 0 Apigeni n (109) R = H Lut eolin (110)R = OH 5.2 Chalcone Synthase Chalcone synthase [naringenin-chalcone synthase; E.C. 2.3.1.741 appears to be the rate-limiting enzyme for biosynthesis of flavonoids in the primary leaves of oat (Avena ~ativa).~~ However the purified enzyme is not inhibited by C-glucosyl- flavones which are the end-products of the biosynthesis of flavonoids in oat and to date there is no indication that chalcone synthase is regulated by feedback or by similar enzyme-modulation mechanisms in this plant.Nevertheless there is a clear correlation between chalcone synthase activity and the rate of accumulation of flavonoids during the development of a leaf. The binding of one of the substrates of the chalcone synthase reaction i.e. 4-coumaroyl-CoA was competitively and strongly inhibited by the flavone apigenin though this compound has not been implicated as an intermediate in the biosynthetic pathway to the major oat flavonoids. The chalcone synthase from buckwheat (Fagopyrurn esculenturn) has also been isolated and purified.g1 The enzyme utilizes malonyl-CoA and 4-coumaroyl-CoA as substrates though caffeoyl-CoA or feruloyl-CoA could substitute for the latter material albeit less effectively.Antibodies to the enzyme have been developed and characterized for specificity. In mixing experiments using extracts from two cell lines of carrot (Daucus carota) tissue cultures that either contained or were devoid of chalcone synthase activity a strong inhibition of chalcone synthase could be dem~nstrated.~~ The inhibition was NATURAL PRODUCT REPORTS 1988 -P. M. DEWICK (+I-Dihydroquercetin (105) (111) (+)-Catechin (11 2) Scheme 24 OMe 0 (-)-Epicatechin (113) (114) Malvidin (115) traced to the presence of a heat-labile 3’-nucleotidase [E.C. 3.1.3.61 in the cells that lacked chalcone synthase activity. The the flavone apigenin (109) giving eriodictyol (103) quercetin (108) and luteolin (1 10) respe~tively.~~ Its cofactor requirements enzyme hydrolyses the phosphate group at C-3’ of adenosine for 0 and NADPH together with other properties suggest in the CoA thioester substrates.Although 4-coumaroyl-3’- that the enzyme is a cytochrome-P-450-type mono-oxygenase. dephospho-CoA was still able to act as a starter group for the reaction of chalcone synthase the dephosphorylated malonyl- CoA could not be used as a chain extender and indeed acted 5.4 Catechins and Proanthocyanidins as an efficient inhibitor of chalcone synthase. Coenzyme A itself Flavan-3,4-diols (leucoanthocyanidins) and flavan-3-01s (cat- lost its inhibitory action after it had been 3’-dephosphorylated. echins) arise by successive reduction steps from dihydro- Such processes may play an important role in regulation of the flavonols.The double reduction step has now been demon- biosynthesis of flavonoids. strated with an enzyme preparation from maturing grains of barley (Hordeurn v~lgare).~~ A soluble NADPH-dependent 5.3 Flavanones Dihydroflavonols Flavonols and Flavones Dihydroflavonols are important intermediates in the bio-synthetic sequences to other flavonoid derivatives and are formed by 3-hydroxylation of flavanone precursors. A (2s)- reductase converted (+)-2,3-dihydroquercetin (105) into the (2R,3&4S)-flavan-3,4-diol (+)-2,3-trans-3,4-cis-leucocyanidin (1 11) but was strongly inhibited by the product of the reaction. A second less-active NADPH-dependent reductase catalysed the reduction of (111) to (+)-catechin (112) (Scheme 24). flavanone 3-hydroxylase [3-dioxygenase] from flowers of Proanthocyanidins (or condensed tannins) contain catechin Petunia hybrida has been purified and shown to have similar elements that possess 2,3-trans [e.g.(+)-catechin (1 12)] and properties to enzymes from other The enzyme 2,3-cis stereochemistry [e.g.(-)-epicatechin (1 13)]. There is is a typical 2-oxoglutarate-dependentdioxygenase requiring much speculation but little agreement on how the 2,3-cis- 2-oxoglutarate7 oxygen Fe2+ and ascorbate as cofactors. (2s)- compounds are derived in Nature. The recent isolation of the Naringenin (102) is converted by the enzyme into (2R,3R)-2,3- first natural 2,3-cis-dihydroflavonol (I 14) from Acacia melano- dihydrokaempferol (104) but (2R)-naringenin is not an xylong7 reinforces the suggestion that this stereochemical acceptable substrate.Similarly (2S)-eriodictyol (1 03) was transformed into (2R,3R)-2,3-dihydroquercetin(lOS) though the (2S)-3’,4’,5’-trioxygenated flavanone 2,3-dihydromyricetin (106) was not metabolized. In addition if 5,7-dihydroxy-flavanone (pinocembrin) or if either of the related flavanones naringin or prunin (which are naringenin 7-0-glycosides) was added to the standard enzyme incubation it had no effect on the conversion of naringenin into 2,3-dihydrokaempferol. Thus the enzyme appears to have high stereospecificity and a rather narrow substrate specificity. Independent studies with the same plant have demonstrated the same cofactor requirements and substrate specifi~ity.~~ In addition there was substantial correlation between the activity of this enzyme in a series of genetic lines of P.hybrida and the flavonol content in their buds feature may be introduced at the dihydroflavonol stage.Perhaps 3-hydroxylation of flavanones may occur giving either configuration at this centre. However since all flavanone 3-hydroxylase enzymes that have so far been investigated yield 2,3-trans stereochemistry in the products an alternative sequence (with the involvement of a C-3 epimerase and conversion of 2,3-trans-compounds into 2,3-cis-compounds) must also be considered. Although leaves of Ginkgo biloba and Ribes sanguineum contain major amounts of catechins and dimers in which there is 2,3-cis stereochemistry cell cultures that had been derived from the leaves tended to synthesize 2,3- trans-isomers instead.98 Several other variations in patterns of proanthocyanidins catechins and their flavonoid precursors between leaf and cultures were noted.. and flowers. Hydroxylation at C-3’ of the aromatic ring occurs at the flavonoid level but is not restricted to any particular oxidation 5.5 Anthocyanidins state of the flavonoid skeleton. Thus a microsomal broad- spectrum flavonoid 3’-mono-oxygenase from seedlings of maize (Zea mays) has been observed to 3’-hydroxylate the Anthocyanidins represent a further group of flavonoids that are derived from dihydroflavonols via flavan-3,4-diols. The violet colour of flowers of Hedysarum carnosum is attributable flavanone naringenin (1 02) the flavonol kaempferol (1 07) and to the presence of the anthocyanidin malvidin (1 1S) whereas NATURAL PRODUCT REPORTS 1988 OMe 0 0 (116) (1171 Gossypetin(ll8) R = H Corniculatusin(ll9) R = Me 0 = H Hesperetin (122) Isovitexin (123) 8-Hydroxykaernpfer'ol (120) R Sexangularetin (121) R = Me R2 HO Pelargonidin (124) R' = R2 = H Cyanidin (125) R'=OH R2=H Delphinidin (126) R' = R2= OH white or violet-spotted mutants are deficient in this com-pound containing colourless flavonols instead.These derive by dehydrogenation of the dihydroflavonol intermediates. The accumulation of flavonoid derivatives in mutants of H. carnosum has been shown to be controlled by genes relating to one or other of the two processes that are involved in the biosynthesis of anthocyanidins i.e. the conversion of dihydroflavonols into flavan-3,4-diols and then of flavan- 3,4-diols into an tho cyan id in^.^^ 5.6 Methylation and Glycosylation of Flavonoids A flavonol 0-methyltransferase from flowers of Chrysosplenium americanum has been demonstrated to methylate the 2'-or the 5'-hydroxyl group of the flavonol glucosides (1 16) and (1 17) respectively.loO Compounds (I 16) and (1 17) and their sub- sequently formed methyl ethers represent four of the major partially methylated flavonol glucosides that have been found in this plant.The development of a yellow colour in flowers of Lotus corniculatus can be correlated with the formation of 8-substituted flavonols particularly gossypetin (8-hydroxy- quercetin) (118) and corniculatusin (8-methoxyquercetin) (1 19) together with smaller amounts of sexangularetin (8-meth- oxykaempferol) (121).lol The accumulation of these pigments involves an 0-methyltransferase activity that uses 8-hydroxy- Isorhamnetin (127) quercetin and 8-hydroxykaempferol (1 20) respectively as its substrates. The albedo of grapefruit (Citrus paradisi) is characterized by the presence of the very bitter-tasting flavanone glycoside naringin (naringenin 7-0-neohesperidoside). Grapefruit cells in suspension culture do not accumulate flavanone glycosides but are able to specifically 0-glucosylate naringenin (1 02) and hesperetin (122) giving the 7-0-glucosides prunin and hes- peretin 7-0-glucoside respectively."' These products have not previously been reported in grapefruit but may well serve as precursors for the biosynthesis of flavanone rhamnoglucosides in the intact tissue.Further glycosylation of the flavone C-glucoside isovitexin (6-C-glucosylapigenin) (123) in the various parts of white campion (Silene pratensis) has been studied in relation to the enzymes controlling the processes and to the genetic factors that control the production of enzyrnes.lo3 Genes controlling glucosylation of the 7-hydroxyl group were expressed in all parts of the plant but those controlling 7-0-xylosylation were expressed only in the petals. In vegetative parts 7-0-xylosylation appeared to be replaced by 7-O-galactosylation but the two reactions do not originate from the same gene. Two different enzymes that catalyse the biosynthesis of isovitexin 7- 0-galactoside were identified.Enzymes catalysing 7-0-glu- cosylation also show variation in substrate specificity and NATURAL PRODUCT REPORTS 1988 -P. M. DEWICK (25)-Naringenin (102) R OH (2s)-Liquiritigenin (128) R H Genistein (131) R =OH (130) (129) Oaidzein (132) R =H Scheme 25 genetic aspects have been investigated.lo4 One 7-O-glu-cosyltransferase transfers glucose to isovitexin but not to isovitexin 2”-O-rhamnoside ;the other glucosylates isovitexin 2”-O-rhamnoside but not isovitexin. Glucosylation of anthocyanidin derivatives at position 3 or 5 has been demonstrated by using enzyme extracts from flowers of stock (Matthiola incana). The enzyme that catalyses the transfer of glucose from uridinediphosphoglucose to the 3- hydroxyl group of pelargonidin (124) was also able to catalyse a similar reaction with the flavonol quercetin (108).’05 Acyanic or pale mutants that lack anthocyanin pigments were blocked in the pathway after dihydroflavonol intermediates but also had drastically reduced glucosyltransferase activity.Cell cul- tures of carrot (Daucus carota) contain an enzyme with similar properties.1o6 Although the cultures contain cyanidin (125) as the only flavonoid aglycon the enzyme glucosylated the anthocyanidins perlargonidin (1 24) cyanidin (1 25) and del- phinidin (1 26) and the flavonols kaempferol(lO7) and quercetin (108) at position 3. The flavanones naringenin and eriodictyol and the dihydroflavonol dihydrokampferol were not glu-cosylated. For this enzyme evidence was presented for the requirement for the presence of some heat-stable organic cofactor that is separable by ultrafiltration.A Matthiola- incana-derived UDPglucose :anthocyanin 05-glucosyltransfer- ase was able to glucosylate perlargonidin and cyanidin 3-O- glycosides and their acylated derivatives. lo’ The best substrates were 3-xylosylglucosides that were acylated with p-coumarate followed by 3-xylosylglucosides and then the 3-glucosides that were acylated with p-coumarate. The 3-glucosides themselves were very poor substrates. Levels of enzyme activity correlated exactly with the formation of 5-glucosylated anthocyanins during the development of a bud. Similarly a xylosyltransferase that catalyses the transfer of xylose from UDPxylose to the glucose of cyanidin 3-glucoside has been isolated from flowers of M.incana.’08 The 3-glucosides of pelargonidin and del- phinidin together with cyanidin 3-(p-coumaroyl)-glucoside and cyanidin 3-(caffeoy1)-glucoside also acted as substrates for the enzyme. Again the accumulation of 3-glucoside derivatives that occurred during the development of a flower could be correlated with this enzyme activity. 5.7 Sulphation of Flavonoids The composite Flaveria bidentis contains a range of flavonol 3-O-glucosides together with a number of flavonol sulphate esters based on quercetin (108) and isorhamnetin (127).lo9 The sulphate esters include 3-sulphates 3,7-disulphates 3,3’,7- trisulphates and 3,3’,4’,7-tetrasulphates.Labelled cinnamic acid was predominantly incorporated into the flavonol glu- cosides suggesting that the sulphation step is likely to be a late stage in the biosynthesis of the sulphate esters.Sulphate that was labelled with 35Swas well incorporated into these esters. 5.8 Isoflavonoids During the biosynthesis of isoflavonoids from flavonoid C,C,C precursors the shikimate-derived aromatic ring migrates to the adjacent carbon of the C unit. An isoflavone synthase activity that is capable of effecting this crucial step was recently detected in a microsomal preparation from cell suspension cultures of soybean (Glycine max) that had been challenged with an elicitor from Phytophthora megasperma f.sp. glycinea. It was shown to convert the flavanone substrates (2q-naringenin (102) or (2S)-liquiritigenin (1 28) into the isoflavones genistein (13 1) or daidzein (132) respectively.The enzyme is a mono-oxygenase requiring NADPH and molecular oxygen as cofactors and a hypothetical pathway via epoxi-dation of the enol form of the flavanone was proposed (Scheme 25). This mechanism has been criticizedl10 and a more favourable one in which the migrating aryl ring is epoxidized rather than the heterocyclic ring has been suggested. A spiro-dienone intermediate analogous to (129) is included however NATURAL PRODUCT REPORTS 1988 (2Sl-Nari ngenin or (2s)-Liquir it igeni n 0 0 0 Genist ein HO or f-Doid zei n Scheme 26 (Scheme 26). Further study of the isoflavone synthase pre- paration has resulted in the isolation of an intermediate believed to be 2,4’,5,7-tetrahydroxyisoflavanone(133) in the transformation of (2q-naringenin (102) into genistein (131).ll1 HO The conversion of (102) into (133) requires NADPH oxygen OH and cytochrome P-450 but the formation of genistein from this intermediate which is formally a dehydration requires none of (133) (134) these cofactors.The 2-hydroxyisoflavanone (1 33) has been suggested to be derived from the carbo-cation (130). A 2’- hydroxylating system could also be demonstrated to be present in the microsomal preparation via the isolation of small amounts of 2’-hydroxygenistein (134). The incorporation of [13C,]acetate into the rotenoid amor- CH,-CO,H phigenin (135) by seedlings of Amorpha fruticosa has been ++ OwAr demonstrated to be analogous to its incorporation into other isoflavonoids which have resorcinol oxygenation patterns in 00 the acetate-derived ring.110 Although only low levels of enrichment were attained the labelling pattern (Scheme 27) was established by using the INADEQUATE pulse sequence to obtain a 13C n.m.r.spectrum. This labelling pattern confirms that the ‘missing’ oxygen function is removed before the polyketide portion cyclizes. The rotenoid skeleton is derived from 2’-methoxyisoflavones ; J the majority of natural examples contain an additional prenyl substituent that is added after this basic skeleton has been formed. Typically the prenyl substituent cyclizes to form an isopropyldihydrofuran ring (as in amorphigenin) or a di-methylchromene system [as in deguelin (1 38)] though these structures are by no means restricted to rotenoids but occur in many classes of natural phenolic compounds.A widely held view is that these rings arise via epoxidation and ring-closure OMe followed by dehydration (Scheme 28 path a) but an alternative sequence (via an ortho-quinone methide and cyclization) is also OMe chemically feasible for the formation of chromenes (Scheme 28 path b).Labelled (6aS 12aS)-rot-2-enonic acid (136) has been Amor p higenin (1351 demonstrated to be well incorporated into deguelin (138) in seedlings of Tephrosia vogelii,’12 but the (6aS 5’RS 12aS)-Scheme 27 alcohol (137) was not significantly utilized as a precursor. A NATURAL PRODUCT REPORTS 1988 -P.M. DEWICK 93 4' 5' v (137) HO OMe Rot-Z'-enonic acid (136) \ b\ OMe (139) Deguelin (138) Scheme 28 (+)-Maackiain (140) (1411 (+) -Pisatin (142) Scheme 29 crude enzyme preparation from the seedlings and a purified deguelin cyclase enzyme from seeds of T. vogelii efficiently converted (136) into deguelin requiring 0 as cofactor. No stable intermediates could be detected and (1 37) was similarly not converted by the enzyme. These results appear to rule out the epoxide sequence but leave the ortho-quinone methide (139) as a plausible intermediate. An interesting observation from the studies was the relatively high efficiency of seedlings of T. vogelii for carrying out the later stages of biosynthesis of deguelin contrasting with the rather poor incorporation of phenylalanine.This observation indicates that the biosynthesis of rotenoids de novo is very slow. After (6aS 12aS)-rot-2'-enonic acid (136) that was labelled with 13C at C-4' was incubated with the enzyme preparation from seeds of T. vogelii 13C-labelled deguelin was isolated and analysed by 13C n.m.r. ~pectro~copy."~ The label was located in both methyl groups of the dimethylchromene though prefer- entially (73%) at C-8' [i.e. in the (pro-R) position]. Thus the enzymic cyclization with respect to the prochiral methyl groups is stereoselective but not stereospecific. No stereo-chemical bias occurred if the same change was achieved by using a chemical conversion so these effects are enzyme-mediated.Whatever the precise nature of the electrocyclization of (139) to deguelin an anticlockwise rotation is apparently favoured relative to a clockwise one (Scheme 28). Experiments in which precursors were fed to the pea (Pisum sativum) had established that the later stages in the biosynthesis of the pterocarpan phytoalexin (+)-pisatin (142) were 6a-hydroxylation of (+)-maackiain (140) followed by methylation of the phenolic group (Scheme 29). This sequence has been con- firmed by studies in which it was found that extracts from pea seedlings were able to methylate (+)-6a-hydroxymaackiain (141) to (+)-pisatin using S-adenosylmethionine as the methyl donor.l14 Although some enzyme activity was present in healthy seedling tissue infection with a microbe or treatment with CuCl (each of which can elicit the synthesis of pisatin) resulted in a much greater activity.The extract showed no methyltransferase activity towards either (-)-maackiain or (-)-6a-hydroxymaackiain though slow methylation of (+)-maackiain (140) was detectable. This suggests that the ability of CuC1,-treated pea tissue to transform (-)-maackiain and (-)-6a-hydroxymaackian into (-)-pisatin as reported in other studies is dependent on a new enzymic pathway that is induced by these precursors and is not present in seedlings that have been treated with CuCl alone. Enzymes that are involved in the biosynthesis of glyceollins [which are phytoalexins of soybean (Glycine max)] were shown to be induced rapidly after the plants had been infected with Phytophthora megasperma f.sp.glycinea only if an incompatible race of fungus was used that did not lead to successful colonization of the ~1ant.l'~ The activities of PAL chalcone synthase isoflavone synthase and dihydroxypterocarpan 6a-hydroxylase all increased markedly after the plant had been infected with an incompatible race of the fungus but they were not particularly different after it had been infected with a compatible race to which the host plant was susceptible. C.H ,OR (143)R = C(O)CH,CO,H (144) R = H c'pp (0 OH k H+ TPP -I Isochorismate (147) R (146) 4 J. 4 q 0 OSB (148) (R = CH,CH,CO,H) (TPP = thiamin diphosphate) Scheme 30 A specific isoflavone 7-0-glucoside-6"-malonatemalonyl-esterase activity has been detected and purified from roots of chickpea (Cicer arietinurn).ll6 This enzyme hydrolysed sub- strates such as biochanin A 7-0-glucoside 6"-malonate (143) to the isoflavone 7-0-glucoside (144) but it had extremely low activity with a range of synthetic esterase substrates and was also insensitive to typical inhibitors of esterases.The mal- onyltransferase thus appeared to differ greatly from other known esterases and it may have a biological function in C. arietinum of releasing isoflavone materials from malonate hemiesters of isoflavone 7-0-glucosides that have accumulated in this plant so that they may be available for the biosynthesis of pterocarpan phytoalexins. The biosynthesis elicitation and biological activities of isoflavonoid phytoalexins have recently been reviewed."' NATURAL PRODUCT REPORTS 1988 6 Quinones Many natural naphthoquinones and anthraquinones including menaquinone (vitamin Kz) are biosynthesized via the shiki- mate-derived intermediate o-succinylbenzoic acid (OSB) (148).This compound arises from isochorismate (147) 2- oxoglutarate (145) and thiamin diphosphate (TPP) pre-sumably via the succinic semialdehyde-TPP anion (146) that is derived by decarboxylation of 2-oxoglutarate (Scheme 30). Evidence has been presented1lS that this decarboxylation is not a function of the oxoglutarate dehydrogenase complex [E.C. 1.2.4.21but is carried out by a separate activity. Thus cell-free extracts from Escherichia coli without added TPP lose o-succinylbenzoate synthase activity but retain all of the activities of the oxoglutarate dehydrogenase complex.Secondly o-succinylbenzoate synthase activity is inhibited by the addition of tetrahydrothiamin diphosphate but the activities of the oxoglutarate dehydrogenase complex are only slowly affected. In confirmation it has now proved to be possible to separate the two enzyme activities by chromatography. The naphthoquinone shikonin (1 52) is an example of another group of natural shikimate-derived quinones which originate via 4-hydroxybenzoic acid (149). The postulated biosynthetic pathway (Scheme 31) to shikonin and related materials is supported by the isolation of intermediates from shikonin- producing cell cultures of Lithospermum erythrorhi~on."~ Thus 3-geranyl-4-hydroxybenzoicacid (150) and geranylhydro-quinone (151) have been isolated for the first time.In non- shikonin-producing cultures the only intermediate that was detected was (150) suggesting that the important decarb- oxylation/ hydroxylation step is repressed in these cultures. Cultures that are capable of synthesizing shikonin but which are being grown in a medium that is not conducive to the production of this quinone were found to accumulate the glucoside of 4-hydroxybenzoic acid (1 53).lZo The concentration of this metabolite decreased rapidly when the cells were transferred to a production medium which stimulated the synthesis of shikonin. This suggested that the precursor 4- hydroxybenzoic acid was being stored in the form of the glucoside when the cells were not synthesizing shikonin and could be released when it was required for metabolism.Two reviews on the biosynthesis of quinones have been published.lZ1.lZ2 7 Miscellaneous Shikimate Metabolites 7.1 Cyclohexyl Fatty Acids A major component of the cellular fatty acids of Curtobacterium pusillum is 1 1-cyclohexylundecanoic acid (1 55). Trace amounts of 13-cyclohexyltridecanoicacid (1 56) together with other non- cyclohexyl fatty acids are also found. These acids are formed by fatty-acid synthase activity and involve extension of the chain of the appropriate starter units by malonate units. A variety of CoA esters of starter acids were accepted when acyl-carrier protein (ACP) and NADPH were added to a crude enzyme preparati~n,"~ though surprisingly acetyl-CoA itself was not satisfactory.Since acetyl-ACP could function as a starter the specificity of the system results from the initial acyl-CoA :[acyl-carrier-protein] S-acyltransferase activity rather than from subsequent chain-elongation reactions. The cyclohexane-con- taining fatty acids arise from cyclohexanecarbonyl-CoA (154) as the starter unit which in turn is formed from shikimic acid (Scheme 32) as demonstrated by feeding experiments in which labelled shikimic acid was used. This necessitates the reduction of the functional groups of shikimic acid and stereochemical aspects of the reduction of the double-bond have been investigated in further experiments.lZ4Labelled 11 -cyclohexyl- undecanoic acid (157) (from the feeding of ~-[6,6-'H,]-glucose to C. pusillum) was partially degraded (by using a combination of microbiological and chemical means) to give (9-cyclohexylphenylcarbinol,which was analysed by 'H n.m.r. spectroscopy. Most of the deuterium enrichment was located at NATURAL PRODUCT REPORTS 1988 -P. M. DEWICK CO,H CO,H CO,H I 1 I geranyl d 1 phosphate OGlc OH (150) maIonyl -CoA C (0)-SCOA ++-.) ICH,I,&O,H [CH 1,,CO2H Ho-P-co2H HO -(154) (155) (156) Scheme 32 I HO' 0 OH (a) glycolysis; (b) pentose phosphate pathway Scheme 33 the axial proton on C-2 of the cyclohexyl moiety with smaller amounts at both 6-H, and 6-H, (157). Labelling at other positions was insignificant.This labelling pattern established that the reduction of the double-bond of shikimic acid had been stereospecific and involved syn addition of hydrogen at the re faces of C-1 and C-2 (Scheme 33). 7.2 Sporopollenin Sporopollenin is the general name for a group of C, polymers that are thought to be derived by oxidation of carotenoids and which are found as components in the walls of pollen and spores. Phenolic acids e.g. rn-hydroxybenzoic acid p-hydroxy- benzoic acid and protocatechuic acid result from the alkaline fusion of such materials. In recent it was shown that inhibitors of carotenoid formation had minimal effects on the biosynthesis of sporopollenin in the anthers of Cucurbita pepo. In the anthers of species of Tulipa tracer studies showed that phenylalanine was well incorporated whilst mevalonic acid was a poor precursor.Glucose malonic acid and p-coumaric acid were also utilized. These findings question the role of carotenoids as precursors but suggest that metabolism of phenolics may be an essential part of the biosynthetic sequence. @+OH 0 OHqOH S hikonin (152) (151) Scheme 31 8 References 1 P. M. Dewick Nut. Prod. Rep. 1986 3 565. 2 ‘The Shikimic Acid Pathway’ ed. E. E. Conn (Recent Advances in Phytochemistry Vol. 20) Plenum New York and London 1986. 3 H. G. Floss in ref. 2 p. 13. 4 S. Ahmad B. Rightmire and R. A. Jensen J. Bacteriol. 1986 165 146. 5 K. Shetty D. L. Crawford and A. L. Pometto Appl. Environ.Microbiol. 1986 52 637. 6 I. N. Olekhnovich N. P. Maksimova and Yu. K. Fomichev Mol. Genet. Mikrobiol. Virusol. 1986 No. 12 p. 34 (Chem. Abstr. 1987 106 64 168). 7 R. J. Ganson T. A. D’Amato and R. A. Jensen Plant Physiol. 1986 82 203. 8 J. E. B. P. Pinto J. A. Suzich and K. M. Herrmann Plant Physiol. 1986 82 1040. 9 G. Millar and J. R. Coggins FEBS Lett. 1986 200 11. 10 P. Le Marechal C. Froussios and R. Azerad Biochemie 1986 68 1211. 11 K. Duncan S. Chaudhuri M. S. Campbell and J. R. Coggins Biochem. J. 1986 238 475. 12 S. Chaudhuri J. M. Lambert L. A. McColl and J. R. Coggins Biochem. J. 1986 239 699. 13 V. I. Osipov and I. V. Shein Biokhimiya (Moscow) 1986 51 9 (Chem. Abstr. 1986 104 125 520). 14 V. I. Osipov and I. V. Shein Biokhimiya (Moscow) 1986,51 230 (Chem.Abstr. 1986 105 3603). 15 G. Millar A. Lewendon M. G. Hunter and J. R. Coggins Bio-chem. J. 1986 237 427. 16 R. C. DeFeyter and J. Pittard J. Bacteriol. 1986 165 331. 17 H. C. Steinrucken A. Schultz N. Amrhein C. A. Porter and R. T. Fraley Arch. Biochem. Biophys. 1986 244 169. 18 C. M. Smith D. Pratt and G. A. Thompson Plant Cell Rep. 1986 5 298. 19 D. M. Mousedale and J. R. Coggins J. Chromatogr. 1986 367 217. 20 N. Amrhein in ref. 2 p. 83. 21 D. M. Mousedale and J. R. Coggins FEBS Lett. 1986 205 328. 22 P. A. Bartlett U. Maitra and P. M. Chouinard J. Am. Chem. Soc. 1986 108 8068. 23 B. K. Singh and E. E. Conn Arch. Biochem. Biophys. 1986 246 617. 24 B. K. Singh S. C. Lonergan and E. E.Conn Plant Physiol. 1986 81 717. 25 H.-P. Schmauder and D. Groger Planta Med. 1986 395. 26 E. Jung L. 0.Zamir and R. A. Jensen Proc. Natl. Acad. Sci. USA 1986 83 7231. 27 J. A. Connelly and E. E. Conn Z. Naturforsch. Sect. C 1986,41 69. 28 D. L. Siehl J. A. Connelly and E. E. Conn Z. Naturforsch. Sect. C 1986 41 79. 29 G. S. Byng A. Berry and R. A. Jensen Arch. Microbiol. 1986 144,222. 30 R. A. Jensen in ref. 2 p. 57. 31 R. A. Jensen Physiol. Plant. 1986 66 164. 32 L. 0.Zamir and K. A. Devor Can. J. Microbial. 1986 32 215. 33 D. N. Rao and S. Kaufman J. Biol. Chem. 1986 261 8866. 34 S. 0.Pember J. J. Villafranca and S. J. Benkovic Biochemistry 1986 25 6611. 35 W. D. Lehmann and H. C. Heinrich Arch. Biochem. Biophys. 1986 250 180.36 S. Ishimitsu S. Fujimoto and A. Ohara Chem. Pharm. Bull. 1986 34 768. 37 V. V. Chestkov Yu. V. Kugukova L. I. Kovalev and S. S. Shish-kin Vopr. Med Khim. 1986 32 123 (Chem. Abstr. 1986 104 182 138). 38 S. J. Benkovic L. M. Bloom G. Bollag T. A. Dix B. J. Gaffney and S. Pember Ann. N.Y. Acad. Sci. 1986 471 226. 39 J. E. Brotherton R. M. Hauptmann and J. M. Widholm Planta 1986 168,214. 40 H. Tsuji T. Ogawa N. Bando and K. Sasaoka J. Biol. Chem. 1986 261 13 203. 41 W. F. Drewe and M. F. Dunn Biochemistry 1986 25 2494. 42 L. Djavadi-Ohaniance B. Friguet and M. E. Goldberg Bio-chemistry 1986 25 2502. 43 S. A. Ahmed B. Martin and E. W. Miles Biochemistry 1986,25 4233. NATURAL PRODUCT REPORTS 1988 44 E. W. Miles R. S. Phillips H. J. C.Yeh and L. A. Cohen Bio-chemistry 1986 25 4240. 45 H. Tanaka K. Tanizawa T. Arai K. Saito T. Arai and K. Soda FEBS Lett. 1986 196 357. 46 R. Hutter P. Niederberger and J. A. DeMoss Annu. Rev. Micro- biol. 1986 40,55. 47 I. P. Crawford Bacteria 1986 10 25 1. 48 N. Tweedy and C. R. Matthews Comments Mol. Cell. Biophys. 1986 3 295. 49 B. H. Brown A. Crozier and G. Sandberg Plant Cell Environ. 1986 9 527. 50 G. Wiese and H. J. Grambow Phytochemistry 1986 25 2451. 51 G. Wiese and H. J. Grambow 2. Naturforsch. Sect. C 1986 41 1023. 52 T. Rausch G. Kahl and W. Hilgenberg Physiol. Plant. 1986,68 458. 53 A. Evidente G. Surico N. S. Iacobellis and G. Randazzo Phyto-chemistry 1986 25 125. 54 H. M. Nonhebel J. Exp. Bot. 1986 37 1691. 55 S.Tsurumi and S. Wada Plant Cell Physiol. 1986 27 559. 56 S. Tsurumi and S. Wada Plant Cell Physiol. 1986 27 1513. 57 T. Kosuge and M. Sanger in ref. 2 p. 147. 58 V. I. Osipov and L. P. Aleksandrova Izv. Sib. Otd. Akad. Nauk SSSR Ser. Biol. Nauk 1986 No. 1 p. 83 (Chem. Abstr. 1986 105 187 673). 59 G. G. Gross S. W. Schmidt and K. Denzel J. Plant Physiol. 1986 126 173. 60 E. Haslam in ref. 2 p. 163. 61 G. P. Bolwell J. Sap C. L. Cramer C. J. Lamb W. Schuch and R. A. Dixon Biochim. Biophys. Acta 1986 881 210. 62 G. P. Bolwell C. L. Cramer C. J. Lamb W. Schuch and R. A. Dixon Planta 1986 169 97. 63 G. J. Kudakasseril and S. C. Minocha Plant Cell Physiol. 1986 27 1499. 64 N. Ishikura S. Teramoto Y. Takeshima and S. Mitsui Plant Cell Physiol.1986 27 677. 65 B. Laber H.-H. Kiltz and N. Amrhein Z. Naturforsch. Sect. C 1986 41 49. 66 I. A. Dubery and J. C. Schabort Biochem. hi. 1986 13 579. 67 J. M. Boniwell and V. S. Butt 2.Naturforsch. Sect. C 1986 41 56. 68 R. J. A. Villegas and M. Kojima J. Biol. Chem. 1986 261 8729. 69 B. Dahlbender and D. Strack Phytochemistry 1986 25 1043. 70 W. Grawe and D. Strack 2.Naturforsch. Sect. C,1986 41 28. 71 D. Strack R. Ruhoff and W. Grawe Phytochemistry 1986 25 833. 72 C. Andary and R. K. Ibrahim 2.Naturforsch. Sect. C 1986,41 18. 73 C. C. S. Chapple M. A. Walker and B. E. Ellis Planta 1986 167 101. 74 Y. Koezuka G. Honda and M. Tabata Phytochemistry 1986 25 2085. 75 F. C. Bradley S. Lindstedt J. D. Lipscomb L. Que A. L. Rowe and M.Grundgren J. Biol. Chem. 1986 261 11 693. 76 J. W. Reed and D. G. 1. Kingston J. Nut. Prod. 1986 49 626. 77 P. Proksch J. Palmer and T. Hartmann Planta 1986 169 130. 78 S. A. Brown Z. Naturforsch. Sect. C 1986 41 247. 79 B. Tal and D. J. Robeson Plant Physiol. 1986 82 167. 80 0.Hofer G. Szabo and H. Greger Monatsh. Chem. 1986 117 1219. 81 H. Wendorff and U. Matern Eur. J. Biochem. 1986 161 391. 82 K. D. Hauffe K. Hahlbrock and D. Scheel Z. Naturforsch. Sect. C 1986 41 228. 83 S. A. Brown in ref. 2 p. 287. 84 B. Moerschbacher B. Heck K. H. Kogel 0.Obst and H. J. Reisener 2. Naturforsch. Sect. C 1986 41 839. 85 E. Morelli R. N. Rej N. G. Lewis G. Just and G. H. N. Towers Phytochemistry 1986 25 1701. 86 W. M. Kamil and P. M. Dewick Phytochemistry 1986 25 2089.87 W. M. Kamil and P. M. Dewick Phytochemistry 1986 25 2093. 88 A. Pelter in ref. 2 p. 201. 89 W. Heller in ‘Plant Flavonoids in Biology and Medicine Biochemical Pharmacological and Structure-Activity Relation- ships’ ed. E. Middleton and J. B. Harborne (Progress in Clinical and Biological Research Vol. 213) Alan R. Liss New York 1986 p. 25. NATURAL PRODUCT REPORTS 1988 -P. M. DEWICK 90 W. Knogge E. Schmelzer and G. Weissenbock Arch. Biochem. Biophys. 1986 250 364. 91 G. Hrazdina E. Lifson and N. F. Weeden Arch. Biochem. Bio- phys. 1986 247 414. 92 W. Hinderer and H. U. Seitz Arch. Biochem. Biophys. 1986,246 217. 93 L. Britsch and H. Grisebach Eur. J. Biochem. 1986 156 569. 94 G. Forkmann P.de Vlaming R. Spribille H. Wiering and A. W. Schram Z. Naturforsch. Sect. C 1986 41 179. 95 R. L. Larson and J. B. Bussard Plant Physiol. 1986 80 483. 96 K. Nyegaard Kristiansen Carlsberg Res. Commun. 1986 51 51 (Chem. Abstr. 1986 105 129 745). 97 L. Y. Foo J. Chem. Soc. Chem. Commun. 1986 675. 98 H. A. Stafford K. S. Kreitlow and H. H. Lester Plant Physiol. 1986 82 1132. 99 A. Chriki D. Combes and M. Marrakchi C. R. Acad. Sci. Ser. 3 1986 302 585. 100 H. E. Khouri N. Ishikura and R. K. Ibrahim Phytochemistry 1986 25 2475. 101 M. Jay and R. K. Ibrahim Biochem. Physiol. P$anz. 1986 181 199. 102 E. Lewinsohn E. Berman Y. Mazur and J. Gressel Phyto-chemistry 1986 25 2531. 103 J. M. Steyns and J. van Brederode Z. Nuturforsch. Sect. C 1986 41 9.104 J. M. Steyns and J. van Brederode Biochem. Genet. 1986 24 349. 105 M. Teusch G. Forkmann and W. Seyffert Z. Nuturforsch. Sect. C 1986 41 699. 106 M. Petersen and H. U. Seitz J. Plant Physiol. 1986 125 383. 107 M. Teusch G. Forkmann and W. Seyffert Planta 1986 168 586. 108 M. Teusch Planta 1986 169 559. 109 L. Varin D. Barron and R. Ibrahim Z. Naturforsch. Sect. C 1986 41 813. 110 L. Crombie I. Holden N. van Bruggen and D. A. Whiting J. Chem. Soc. Chem. Commun. 1986 1063. 111 G. Kochs and H. Grisebach Eur. J. Biochem. 1986 155 311. 112 L. Crombie J. Rossiter and D. A. Whiting J. Chem. Soc. Chem. Commun. 1986 352. 113 M. J. Begley L. Crombie J. Rossiter M. Sanders and D. A. Whiting J. Chem. Soc. Chem. Commun. 1986 353. 114 J.A. Sweigard D. E. Matthews and H. D. VanEtten Plant Physiol. 1986 80 277. 115 A. Bonhoff R. Loyal J. Ebel and H. Grisebach Arch. Biochem. Biophys. 1986 246 149. 116 W. Hinderer J. Koster and W. Barz Arch. Biochem. Biophys. 1986 248 570. 117 D. A. Smith and S. W. Banks Phytochemistry 1986 25 979. 118 M. G. Marley R. Meganathan and R. Bentley Biochemistry 1986 25 1304. 119 K. Yazaki H. Fukui and M. Tabata Chem. Pharm. Bull. 1986 34 2290. 120 K. Yazaki H. Fukui and M. Tabata Phytochemistry 1986 25 1629. 121 E. Leistner in ref. 2 p. 243. 122 E. Leistner H. J. Bauch A. Boos U. Igbavboa R. Kolkmann and A. Weische Dtsch. Apoth. Ztg. 1986 126 2000. 123 A. Kawaguchi N. Uemura and S. Okuda J. Biochem. (Tokyo) 1986 99 1735. 124 J.Furukawa T. Tsuyuki N. Morisaka N. Uemura Y. Koiso B. Umezawa A. Kawaguchi S. Iwasaki and S. Okuda Chem. Pharm. Bull. 1986 34 5176. 125 A. K. Prahl M. Rittscher and R. Wiermann in ‘Biotechnology and Ecology of Pollen’ ed. D. L. Mulcahy G. B. Mulcahy and E. Ottaviano Springer New York 1986 p. 313 (Chem. Abstr. 1986 105 112 122).