The biosynthesis of shikimate metabolites Paul M. Dewick School of Pharmaceutical Sciences, University of Nottingham, Nottingham, UK NG7 2RD Covering: 1995 and 1996 Previous review: 1995, 12, 579 1 Introduction 2 The shikimate pathway 2.1 DAHP synthase 2.2 3-Dehydroquinate synthase 2.3 3-Dehydroquinase 2.4 The quinate utilization pathway: quinate dehydrogenase and 3-dehydroshikimate dehydratase 2.5 Shikimate kinase 2.6 EPSP synthase 2.7 Chorismate synthase 2.8 Chorismate mutase 2.9 Phenylalanine and tyrosine 2.10 Aromatic amino acid hydroxylases 2.11 Tyrosinase and melanin 2.12 p-Aminobenzoic acid, anthranilic acid and related compounds 2.13 Isochorismate synthase 3 Tryptophan and related compounds 3.1 Tryptophan 3.2 Tryptophanase 3.3 Indole-3-acetic acid and related metabolites 3.4 Indigo and indirubin 4 Phenols and phenolic acids 4.1 Phenol and tyrosine phenol-lyase 4.2 4-Hydroxybenzoic acid 4.3 Salicylic acid 4.4 Homogentisic acid 5 Phenylpropanoids 5.1 General 5.2 Phenylalanine ammonia-lyase 5.3 Hydroxycinnamic acids and esters 5.4 Coumarins 5.5 Lignins 5.6 Lignans 5.7 Tropic acid 5.8 Phenylpropenes 6 Flavonoids 6.1 Chalcones 6.2 Flavanones, flavonols, anthocyanidins and related flavonoids 6.3 Flavonoid sulfates 6.4 Carthamin 6.5 Diels–Alder-type adducts 6.6 Isoflavonoids 7 Stilbenes and dihydrophenanthrenes 8 Xanthones 9 Quinones 9.1 Naphthoquinones 9.2 Ubiquinones 9.3 Shikonin 9.4 Tocopherols 10 Cyanogenic glycosides, glucosinolates and related compounds 11 Miscellaneous shikimate metabolites 11.1 Sphagnum acid 11.2 4-(4-Hydroxyphenyl)butan-2-one 11.3 Phenylacetylenes from Asparagus 11.4 Diarylheptanoids and phenylphenalenones 11.5 Antibiotic LL-C10037· 11.6 Brominated tyrosine metabolites from sponges 11.7 Cyclohexanecarboxylic acid 11.8 3-Amino-5-hydroxybenzoic acid and mC7N units 11.9 3-Amino-4-hydroxybenzoic acid 11.10 Betalains 12 References 1 Introduction This report reviews the literature that was published during 1995 and 1996 on the biosynthesis of compounds, mainly non-nitrogenous, that are derived wholly or partly from shikimate, and continues the coverage described in Volume 12 of Natural Product Reports1 and earlier reports.2, 3 This review includes descriptions of the biosynthetic pathways, the enzymes and enzyme mechanisms involved, and brief information about the genes encoding for these enzymes.An authoritative review on the enzymology of the shikimate pathway has been published.4 Other recent reviews describe early steps in the shikimate pathway5, 6 and the molecular organization of the pathway in higher plants.7 A natural isomer of shikimic acid, namely 3-epishikimic acid 1, has been isolated along with shikimic acid from Sequoiadendron giganteum.8 The synthesis of (6R)- 2 and (6S)-fluoroshikimic acids 3 from the corresponding isomers of 3-fluorophosphoenolpyruvate has been achieved by the use of a mixture of the recombinant enzymes DAHP synthase, dehydroquinate synthase, dehydroquinase and shikimate dehydrogenase, together with the appropriate cofactors.9 2 The shikimate pathway 2.1 DAHP synthase The first reaction in the shikimate pathway involves the aldol-like condensation of phosphoenolpyruvate (PEP) 4 with D-erythrose 4-phosphate 5 giving 3-deoxy-D-arabinoheptulosonate- 7-phosphate (DAHP) 6 (Scheme 1), a reaction catalyzed by the enzyme DAHP synthase (phospho-2-dehydro- 3-deoxyheptonate aldolase).DAHP synthase enzymes fall into two distinct classes according to whether they are plantderived or microbial in origin. In microorganisms, isozymes of DAHP synthase are characterized by diVerent sensitivities towards feedback inhibition by the aromatic amino acids L-phenylalanine, L-tyrosine and L-tryptophan. In the methylotrophic actinomycete Amycolatopsis methanolica, a single DAHP synthase activity identified in the wild-type strain is inhibited by all three aromatic amino acids.10 However, a leaky phenylalanine-requiring auxotroph possessed an additional isozyme which was strongly inhibited OH OH HO CO2H OH OH HO CO2H R2 R1 1 2 R1 = F; R2 = H 3 R1 = H; R2 = F Dewick: The biosynthesis of shikimate metabolites 17by tyrosine. Tryptophan-sensitive DAHP synthase enzymes from Streptomyces coelicolor, S.rimosus and Neurospora crassa have been purified to homogeneity and their amino acid sequences have been determined.11 From the sequence data, these three enzymes appear unrelated to other microbial DAHP synthases, but are actually similar to plant enzymes.Thus, the distinction between plant and microbial enzymes now becomes less clear, with examples from bacterial and eukaryotic organisms falling into the plant class. The phenylalanine-regulated isozyme from Escherichia coli has been crystallized and preliminary crystallographic data have been reported for binary and tertiary complexes with substrates and/or phenylalanine.12 Isozymes of DAHP synthase in plants are diVerentiated by their requirements for the divalent cations Mn2+ or Co2+.The former isozyme type is typically located in the chloroplast, the latter in the cytosol. Cytosolic DAHP synthase–Co from cell cultures of carrot (Daucus carota) has been isolated, purified to homogeneity, and characterized.13 This isozyme has previously only been obtained in a partially purified form, and its properties have not been fully described.An antisense DNA construct encoding part of the DAHP synthase sequence in potato (Solanum tuberosum) has been inserted into potato cells.14 Some, but not all, of the derived transgenic plants expressing antisense RNA showed reduced levels of woundinduced DAHP synthase activity. 2.2 3-Dehydroquinate synthase 3-Dehydroquinate synthase catalyzes the conversion of DAHP 6 into 3-dehydroquinate 7 by a sequence of reactions requiring an oxidation, a ‚-elimination, a reduction and an intramolecular aldol condensation.Some of these steps undoubtedly occur non-enzymically, and current thinking is that dehydroquinate synthase is probably responsible only for the oxidation and reduction steps, whilst the other changes are brought about by spontaneous reactions occurring on the intermediates generated (see ref. 1). In recent studies,15 a cDNA clone encoding for dehydroquinate synthase has been isolated from tomato (Lycopersicon esculentum) and identified by complementation of a dehydroquinate synthase-deficient strain of E.coli. The deduced amino acid sequence resembles those of bacterial enzymes more than it does those of fungal systems, and it contains a putative N-terminal plastid-specific transit peptide. A close analogy to the dehydroquinate synthase reaction is found in the biosynthesis of 2-deoxy-scyllo-inosose, an intermediate between glucose and 2-deoxystreptamine in the formation of aminoglycoside antibiotics (see ref. 2). The full paper describing these studies has been published.16 2.3 3-Dehydroquinase The dehydration of 3-dehydroquinate 7 to 3-dehydroshikimate 8 (Scheme 1) is catalyzed by 3-dehydroquinase (3-dehydroquinate dehydratase). This enzyme activity has been identified as a monofunctional activity in bacteria, as part of a bifunctional system with shikimate dehydrogenase in plants, as one of the five activities of the AROM pentafunctional enzyme complex in fungi and yeasts, and as part of an inducible multifunctional system in fungi which is produced to metabolize quinate 10.Two diVerent classes of dehydroquinase enzymes having quite diVerent physical and biochemical properties are now recognized (Scheme 2). Type I enzymes have a biosynthetic role and use a SchiV base mechanism between the substrate and an active site lysine (see ref. 1). On the other hand, Type II enzymes can function as either biosynthetic or catabolic enzymes and do not use a SchiV base mechanism.They also carry out the elimination with opposite stereochemistry to Type I enzymes, involving an anti dehydration with loss of the axial pro-S hydrogen from C-2. With the use of electrospray mass spectrometry, an essential arginine residue in Type II dehydroquinases of Streptomyces coelicolor and Aspergillus nidulans has been identified.17 The technique was used to characterize inactivated protein modi- fied with the Arg-specific reagent phenylglyoxal, and demonstrated that reaction had occurred predominantly with a single conserved arginine residue, Arg-23 in the S.coelicolor enzyme, and Arg-19 in the case of A. nidulans. There was no evidence for reaction with a further conserved arginine during inactivation of the enzyme. Mutant enzymes derived by replacing Arg-23 in the S. coelicolor protein with Lys, Gln or Ala all had drastically reduced catalytic activities, though with stronger CO2 – PO O PO HO OH CO2 – PO HO OH OH O OH OH O CO2 – HO OH OH O CO2 – OH OH HO CO2 – OH OH HO CO2 – HO 4 PEP 5 Erythrose 4-phosphate 6 DAHP 7 3-Dehydroquinate NAD+ ii 8 3-Dehydroshikimate 10 Quinate iv iii v i 9 Shikimate NADH NADH 2 P = phosphate 6 Scheme 1 Enzymes: i, DAHP synthase; ii, 3-dehydroquinate synthase; iii, 3-dehydroquinase; iv, shikimate dehydrogenase; v, quinate dehydrogenase HO OH O HR HS OH –O2C HO OH O H H OH –O2C HO OH O H OH –O2C H A HO OH –O2C HN H H HO HO OH –O2C HN H HO H B OH OH O CO2 – OH OH O CO2 – H H B B 8 3-Dehydroshikimate 7 3-Dehydroquinate · * * * * * · · · · Type I Type II 11 + + Scheme 2 Enzyme: 3-dehydroquinase 18 Natural Product Reports, 1998substrate binding.18 A role for Arg-23 in stabilizing a carbanion intermediate was proposed.Comparison of the amino acid sequences around the essential arginine residues in both Type II and Type I (from E. coli) enzymes indicates a conserved structural motif that may reflect common substrate binding at the active centre.Evidence from a series of kinetic isotope studies supports a proposal for an enolate intermediate and an E1cb mechanism for Type II dehydroquinases (Scheme 2).19 Labelled (2R)- and (2S)-[2-2H]dehydroquinates in the presence of H2O or 2H2O were employed with Type I enzyme from E. coli and Type II enzymes from A. nidulans and Mycobacterium tuberculosis, and appropriate isotope eVects were measured. For all three enzymes, proton abstraction was partially rate-limiting.Substrate and solvent isotope eVects for the Type I enzyme were small, consistent with a complex multistep mechanism. Significant substrate isotope eVects were observed for the two Type II enzymes, but only the A. nidulans enzyme showed a large solvent isotope eVect. All of the results for the Type II enzymes could be rationalized by the action of a stepwise E1cb mechanism involving an enolate intermediate 11, formation of which is partially rate-limiting. It is proposed that there is an electrostatic interaction of the dehydroquinate with the essential arginine residue, and that this residue may also moderate the basicity of a histidine general base responsible for deprotonation (Scheme 3).There is no evidence that Type II dehydroquinases require any metal cofactors for activity.20 The dodecameric enzyme from A. nidulans adopts an arrangement in which two hexameric ring-like structures stack on top of each other. However, there are data to indicate that a ligand-induced conformational change occurs which increases modification of the active site arginine by phenylglyoxal.Binding also partially protects against proteolysis in the highly conserved N-terminal region which contains this arginine residue. The gene encoding Type II dehydroquinase in the stomach pathogen Helicobacter pylori has been isolated and sequenced.21 The gene, designated aroQ, has strong sequence identity to other members of the Type II family.The 5-deoxy 12 and 4,5-dideoxy 13 analogues of dehydroquinate have been tested as substrates of Type I (E. coli) and Type II (M. tuberculosis) enzymes.22 Results suggest that the C-4 hydroxy group is essential for imine formation on the Type I enzyme, only the 5-deoxy compound acting as a modest substrate through an imine-bound species. Both compounds were poor substrates for the Type II enzyme, but since the dideoxy compound was a competitive inhibitor, the 4-hydroxy group does not contribute significantly to specificity. 2.4 The quinate utilization pathway: quinate dehydrogenase and 3-dehydroshikimate dehydratase An inducible quinate-utilizing (QUT) catabolic pathway is observed in certain fungi, and a Type II dehydroquinase functions in this pathway. Quinate 10 enters the pathway by the action of quinate dehydrogenase (quinate:oxidoreductase) giving 3-dehydroquinate 7, and is subsquently transformed to protocatechuate (3,4-dihydroxybenzoate) by the action of 3- dehydroshikimate dehydratase on dehydroshikimate.All three QUT pathway enzymes, quinate dehydrogenase, dehydroquinase and dehydroshikimate dehydratase from Aspergillus nidulans can now be purified in bulk via overexpression in E. coli.23 The isolation and purification of quinate dehydrogenase from Phaseolus mungo sprouts24 and dehydroshikimate dehydratase from Vigna mungo cells cultured in the presence of shikimate25 have been described. A cofactor-independent quinate hydrolyase activity capable of converting quinate directly into shikimate has been reported in pea (Pisum sativum) roots.26 This activity may function in channelling quinate into the shikimate pathway. 2.5 Shikimate kinase Phosphorylation of shikimate 9 to shikimate 3-phosphate 14 is brought about by shikimate kinase in the presence of ATP (Scheme 4). Recent sequence data for the aroK-encoded isozyme of shikimate kinase from E. coli have indicated that sequences published earlier contained a number of errors, and now establish that the two isozymes aroK and aroL are of comparable length and share 30% identity.27 2.6 EPSP synthase EPSP synthase (3-phosphoshikimate 1-carboxyvinyltransferase) catalyzes the condensation of shikimate 3-phosphate 14 with PEP to produce the enol ether 5-enolpyruvylshikimate HO OH O H H OH –O2C N HN H N NH2 H2 N 7 3-Dehydroquinate + Scheme 3 Enzyme: 3-dehydroquinase (Type II) OH O CO2 – HO O CO2 – HO 12 13 4 5 5 OH OH HO CO2 – OH OH PO CO2 – OH O PO CO2 – CO2 – OH O CO2 – CO2 – 9 Shikimate 15 EPSP 14 Shikimate 3-phosphate 16 Chorismate i ii iii iv ATP PEP OH –O2C 17 Prephenate CO2 – O Scheme 4 Enzymes: i, shikimate kinase; ii, EPSP synthase; iii, chorismate synthase; iv, chorismate mutase Dewick: The biosynthesis of shikimate metabolites 193-phosphate (EPSP) 15 (Scheme 4).Gene cloning and expression has allowed the enzyme from Bacillus subtilis to be studied and modified through site-directed mutagenesis.28 The native Bacillus enzyme exhibited allosteric behaviour, a property not yet reported amongst other bacterial and plant EPSP synthases so far investigated.It was established via the mutant enzymes that both shikimate 3-phosphate and PEP have multiple interaction sites. There are two sites for PEP binding, a catalytic one and a non-catalytic one. A shikimate 3-phosphate interacting site has been located in the N-terminal domain.29 A mutant strain of E.coli EPSP synthase in which His-385 is replaced with asparagine is much less catalytically competent (6% activity) than the corresponding glutamine mutant, previously shown to retain some 25% activity.30 Accordingly, the Â-nitrogen of histidine is suggested to play a role in catalysis, perhaps by hydrogen-bonding to another residue that complexes the substrates or products. The broad spectrum herbicide glyphosate [N-(phosphonomethyl) glycine] 18 is a powerful inhibitor of EPSP synthase, and is generally considered to bind to the PEP binding site as an analogue of the PEP oxonium ion 19 formed transiently during catalysis.A stable ternary complex of the enzyme with glyphosate and shikimate 3-phosphate is produced. Recent studies now question the concept that glyphosate acts as a transition state analogue inhibitor.31 Thus, in the reverse reaction involving EPSP and phosphate, there should be little chance of glyphosate interacting with the enzyme or enzyme– substrate complex, but it was found that glyphosate can also be trapped on the enzyme–EPSP to give a ternary complex.Therefore the carboxyvinyl group of EPSP does not prevent glyphosate binding, and it is actually found to strongly facilitate binding. Results of rotational-echo double-resonance NMR and molecular dynamics simulations also support these results, and suggest glyphosate is unlikely to bind in the same fashion as PEP.32 A new, very potent inhibitor 20 of EPSP synthase from E.coli has been designed by eVectively joining together shikimate 3-phosphate with glyphosate.33 Although this should have a specific interaction with the shikimate 3-phosphate binding site, there is likely to be incomplete overlap at the PEP/ glyphosate subsite. The highly substituted shikimate ring can be replaced by structurally simple pyrrole systems to produce shikimate 3-phosphate analogue inhibitors of EPSP synthase. Thus, compound 21 is a surprisingly good inhibitor of the enzyme.34 2.7 Chorismate synthase The elimination of phosphoric acid from EPSP 15 yields chorismate 16 and is catalyzed by the enzyme chorismate synthase (5-enolpyruvylshikimate 3-phosphate-lyase) (Scheme 4).The enzyme achieves a stereochemically ambiguous anti 1,4-elimination, and has a requirement for a reduced flavin for activity, despite the reaction involving no overall change in oxidation state (Scheme 5). Some enzymes, e.g.those from Neurospora crassa and Bacillus subtilis, are bifunctional and have an associated flavin reductase that utilizes NADPH to produce the reduced flavin (Scheme 5). Others, e.g. enzymes from E. coli and the plants Pisum sativum and Corydalis sempervirens, are monofunctional and must be supplied with a reduced flavin. Using a recombinant chorismate synthase from E. coli, only a small isotope eVect was noted when (6R)-[6-2H]EPSP was employed as substrate and no isotope eVect occurred with the (6S)-[6-2H]-labelled substrate.35 Similar isotope eVects were observed for the decay of a spectroscopically characterized flavin semiquinone intermediate.This is consistent with the main rate-limiting step being distinct from the loss of the 6-pro-R hydrogen, and indicates that the flavin intermediate forms before C–H bond cleavage. (6S)-6-FluoroEPSP 22 is a competetive inhibitor of chorismate synthase, but is also converted into 6-fluorochorismate by the E.coli enzyme and at a rate some two orders of magnitude slower than EPSP.36 Of various possible mechanisms for the conversion, the decreased rate of reaction is most consistent with destabilization of an allylic cationic intermediate 23 (Scheme 6). However, this mechanism assigns no role for the reduced flavin cofactor, unless its electron rich character could play a role in stabilizing a cationic intermediate. Alternatively, an allylic radical intermediate 24 may also be destabilized by a fluorine substituent, and its formation and decay would provide a direct role for the cofactor. Extensive kinetic studies following substrate consumption, product formation, phosphate dissociation, and the formation and decay of the flavin intermediate also support these proposals.37 The overall reaction is non-concerted, and the flavin intermediate forms before the substrate is consumed so it is not simply associated with the conversion of substrate into product.A cDNA from Neurospora crassa encoding bifunctional chorismate synthase/flavin reductase has been isolated, sequenced, and expressed in E. coli.38 Its deduced amino acid sequence was found to be 79% similar to that of the Saccharomyces cerevisiae enzyme. Two regions unique to the N. crassa protein were deleted, but the protein was still able to 18 Glyphosate + 19 + CO2 – N P O O– O– H H CO2 – H H O H P O O– O– 21 20 OH NH PO CO2 – P O O– O– CO2 – NH CO2 – P O O– –O OH OH O PO CO2 – CO2 – OH O CO2 – CO2 – H H6 R H6 S 15 EPSP 16 Chorismate i · · NADPH NADP+ FMN FMNH2 Scheme 5 Enzyme: i, chorismate synthase OH O PO CO2 – CO2 – H F 22 6 20 Natural Product Reports, 1998complement a chorismate synthase-deficient mutant of E.coli. Domains responsible for the flavin reductase activity appear to lie within regions of homology with monofunctional chorismate synthases. The cDNA responsible for chorismate synthase in the plant Corydalis sempervirens had earlier been shown to encode a precursor polypeptide containing an N-terminal plastid transit peptide domain.Coding regions of a cDNA for the precursor peptide, and for the mature chorismate synthase have been expressed in E. coli.39 Only the mature chorismate synthase, and not the precursor protein, was enzymically active. Three plastidic isozymes were identified via their cDNAs in tomato (Lycopersicon esculentum), though these were derived from only two genes.40 Two of these isozymes were expressed in E.coli (the third was unstable in E. coli) both as precursor proteins with transit peptides, and as mature proteins. Again, only mature proteins were enzymically active. 2.8 Chorismate mutase Chorismate mutase catalyzes the Claisen-like rearrangement of chorismate 16 to prephenate 17, during which the PEP-derived side-chain becomes directly bonded to the carbocycle (Scheme 4). This is achieved by binding the unfavoured pseudoaxial conformer 25 of chorismate, achieving a chair-like transition state (Scheme 7).The uncatalyzed thermal rearrangement of chorismate to prephenate is also possible, but chorismate mutase increases the rate of reaction by a factor of 2#106. All of the experimental data favour a simple pericyclic process for the enzyme-catalyzed reaction. From a combined quantum mechanical–molecular mechanical study of chorismate mutase from Bacillus subtilis it was concluded that catalysis may be rationalized in terms of a combination of substrate strain and transition state stabilization, and the enzyme is not functioning by a chemical catalysis mechanism.41 Consideration of several simple model systems has indicated that selective transition state binding is achieved by appropriately positioned H-bond donors in the protein.42 Determination of activation barriers for the B.subtilis enzyme shows that the entropy barrier is nearly as unfavourable as for the uncatalyzed reaction so that factors other than entropy are important.43 The enzyme exerts considerable conformational control over the molecule and is highly complementary to the pseudoaxial conformer via multiple hydrogen-bonding and electrostatic interactions.The active site of yeast (Saccharomyces cerevisiae) chorismate mutase has been located by comparison with the mutase domain of the bifunctional chorismate mutase–prephenate dehydratase (the P-protein) from E. coli.44 The active site domains of the two enzymes are very similar, and of the seven active site residues, four are conserved.There are two Arg residues which bind to carboxylates, Lys which binds to the ether oxygen, and Glu which binds to the hydroxy group (see Scheme 8 which shows binding to the transition state analogue and inhibitor 26). Crystal structures of the yeast enzyme bound to the allosteric inhibitor tyrosine and the allosteric activator tryptophan have also been established and interpreted.45 The OH O PO CO2 – CO2 – OH O CO2 – CO2 – OH O CO2 – CO2 – OH O CO2 – CO2 – 15 EPSP 16 Chorismate FMNred FMNsq Pi FMNred 24 FMNsq 23 Pi H+ + Scheme 6 Enzyme: chorismate synthase CO2 – HO O CO2 – OH CO2 – O –O2C CO2 – –O2C O OH 17 Prephenate 16 Chorismate 25 Scheme 7 Enzyme: chorismate mutase OH CO2H O HO2C 26 – – + + + + – – + – – O O H O O O O H2N HN N OH O O H O O NH O NH3 NH NH2 H2N O O H O O O O NH2 HN NH2 O NH2 H2N NH O O S H H Glu-198 (Glu-52) Asn-194 (Asp-48) Glu-246 (Gln-88) Arg-16 (Arg-11) Lys-168 (Lys-39) Thr-242 (Ser-84) Arg-7 Arg-28 (Arg-157) Amino acid residues as shown: Saccharomyces cerevisiae chorismate mutase Amino acid residues in brackets: E.coli chorismate mutase Arg-90 Glu-78 Cys-75 Tyr-108 Bacillus subtilis chorismate mutase Scheme 8 Dewick: The biosynthesis of shikimate metabolites 21crystal structure of the E. coli enzyme shows it to be a homodimer with two equivalent active sites, each having contributions from both monomers.46 Site-directed mutagenesis of a fully functional chorismate mutase engineered from the E.coli P-protein has been employed to demonstrate that Lys-39 and Gln-88 play critical catalytic roles, in that mutant proteins with these residues replaced suVered a major loss of catalytic activity compared to the wild type enzyme.47 In parallel studies,48 Lys-39 and Gln-88 were shown to be critical for enzyme activity, Arg-11 and Arg-28 to be important, whilst Glu-52 played only a minor role.These results stress the importance of hydrogen bonding to the enolpyruvyl oxygen and carboxylate. Thermodynamic parameters for the E. coli enzyme and a Gln88Glu mutant enzyme have been determined and used to provide a direct mechanistic link between the chorismate mutase enzymes of E. coli and S. cerevisiae.49 The active site in Bacillus subtilis chorismate mutase (Scheme 8) has rather low similarity to those in E. coli and S. cerevisiae. Again, site-directed mutagenesis has been used to establish the relative importance of the various residues.50 Critical roles have been identified for Arg-7 and Arg-90, whilst Tyr-108, Arg-116, Phe-57, and Cys-75 do not feature significantly in catalysis.Glu-78 positioned near the 4-hydroxy group contributes to binding, and its replacement reduces activity. In other studies,51 a positively charged residue either at position 88 (Lys) or 90 (Arg or Lys) is essential for activity, strongly supporting the hypothesis that a positive charge is required to stabilize a dipolar transition state resulting from scission of the C–O bond.Essential active site residues in the bifunctional chorismate mutase–prephenate dehydrogenase (the T-protein) from E. coli have been identified by residue-specific chemical reactions.52 A single lysine (Lys-37) is essential for the mutase activity, whilst His-131 is essential for dehydrogenase activity. A cDNA clone encoding a cytosolic chorismate mutase isozyme has been identified in the plant Arabidopsis thaliana and shown to encode a protein with a deduced amino acid sequence 50% identical to that of the plastidic isozyme previously isolated.53 The plastidic isozyme expressed in E.coli was activated by tryptophan, and inhibited by phenylalanine and tyrosine, whilst the cytosolic enzyme was insensitive to these amino acids. Transformed roots of Datura stramonium also contain two chorismate mutase isozymes, only one of which is present in intact plastids.54 At least two chorismate mutase activities were identified in the methylotrophic actinomycete Amycolatopsis methanolica.10 One of these was a bifunctional protein which also carried DAHP synthase activity.Both chorismate mutase activities were inhibited by the amino acids phenylalanine and tyrosine. Reviews describing chorismate mutase in microorganisms and plants,55 and in Escherichia coli,56 have been published. 2.9 Phenylalanine and tyrosine Several pathways for the conversion of chorismate 16 into the aromatic amino acids L-phenylalanine 28 and L-tyrosine 31 are known to exist (Scheme 9) and the pathway employed is dependent on the organism.Intermediates phenylpyruvate 27, 4-hydroxyphenylpyruvate 30, or L-arogenate 29 may be involved, and often, more than one route may operate in a particular species as a result of the enzyme activities available. Phenylalanine- and tyrosine-requiring auxotrophs of the actinomycete Amycolatopsis methanolica have been identified and shown to be deficient in prephenate dehydratase or arogenate dehydrogenase, respectively.57 These strains thus have single pathways for the production of phenylalanine and tyrosine.Extracts of the organism yielded two fractions with aminotransferase activity towards phenylalanine and tyrosine, and three fractions acting on prephenate. Mutant and biochemical studies identified two aminotransferase activities that were dominant in phenylalanine and tyrosine biosynthesis.The prephenate aminotransferase has been partially purified.58 Prephenate dehydratase from this organism has also been purified, and shown to be a homotetramer.59 Tyrosine aminotransferase in humans is concerned primarily with tyrosine catabolism; a cDNA encoding this enzyme in the liver has been cloned and expressed in HeLa cells.60 A recombinant tyrosine aminotransferase from Trypanosoma cruzi has been shown to also possess alanine aminotransferase activity.61 Purification of one of the two aromatic amino acid aminotransferases found in Azospirillum brasiliense showed it was able to utilize all three amino acids.62 Enzymes of the shikimate pathway involved in aromatic amino acid biosynthesis in the unicellular chlorophyte alga Chlorella sorokiniana have been characterized and shown to be typical of higher plants rather than photosynthetic prokaryotes such as cyanobacteria.63 Thus, the following distinctive enzymes were identified: aMn2+-stimulated isozyme of DAHP synthase, a bifunctional dehydroquinase-shikimate dehydrogenase, an allosterically controlled isozyme of chorismate mutase, a thermotolerant prephenate aminotransferase, tyrosine-inhibited arogenate dehydrogenase, and arogenate dehydratase.The biocatalytic conversion of glucose into aromatic compounds in E. coli can be enhanced considerably by modification of the chromosome.64 Insertion of a multigene cassette carrying the genes aroA, aroC and aroB (encoding for EPSP synthase, chorismate synthase and dehydroquinate synthase, respectively) into an E.coli strain where transcriptional repression of aroL (shikimate kinase) is not operative resulted in a doubling of the yields of L-phenylalanine, prephenic acid and phenyllactic acid. The enzyme phenylalanine dehydrogenase catalyzes the reversible NAD+-dependent oxidative deamination of Lphenylalanine to phenylpyruvate 27 and ammonia.The gene encoding the enzyme in Bacillus badius has been sequenced.65 The deduced amino acid sequence for the enzyme was found to be similar to those reported earlier from B. sphaericus and Thermoactinomyces intermedius. For the enzyme from 16 Chorismate 17 Prephenate i 29 L-Arogenate 28 L-Phenylalanine 27 Phenylpyruvate 30 4-Hydroxyphenylpyruvate 31 L-Tyrosine iii PLP ii iv PLP v vii NAD+ vi NAD+ viii PLP CO2 – O CO2 – O OH OH O CO2 – CO2 – OH –O2C CO2 – O OH –O2C CO2 – NH2 CO2 – NH2 CO2 – NH2 OH Scheme 9 Enzymes: i, chorismate mutase; ii, prephenate dehydratase; iii, phenylpyruvate aminotransferase; iv, prephenate aminotransferase; v, arogenate dehydratase; vi, arogenate dehydrogenase; vii, prephenate dehydrogenase; viii, 4-hydroxyphenylpyruvate aminotransferase 22 Natural Product Reports, 1998B. sphaericus, amino acid residues Gly-124 and Leu-307 have been suggested to play an important role, probably contributing towards the substrate specificity of the enzyme.These residues have been altered by site-specific mutagenesis to Ala and Val, respectively, which are corresponding residues in the related enzyme leucine dehydrogenase.66 Single and double mutants displayed lower activity towards phenylalanine and enhanced activity towards aliphatic amino acid substrates. 2.10 Aromatic amino acid hydroxylases Direct hydroxylation of L-phenylalanine to L-tyrosine and of L-tyrosine to L-(3,4-dihydroxyphenyl)alanine (L-DOPA) may be achieved in certain organisms via aromatic amino acid hydroxylases.This route is particularly important in mammals, which lack the shikimate pathway. These enzymes are members of a group of tetrahydropterin-dependent hydroxylases and include phenylalanine hydroxylase, tyrosine hydroxylase and tryptophan hydroxylase. The mixed oxidation reaction consumes molecular oxygen and one equivalent of reduced pterin cofactor to hydroxylate the substrate. Most of the enzymes contain a metal at the active site which is reduced by the tetrahydropterin cofactor, though an exception has been found in phenylalanine hydroxylase from Chromobacterium violaceum.To help identify the function of metals in mammalian phenylalanine hydroxylases, the Fe-dependent rat liver enzyme has been compared with the metal-independent C. violaceum enzyme using 2H-labelled phenylalanine substrates. 67 Both enzymes displayed similar kinetic isotope eVects with [4-2H]phenylalanine as substrate, and an almost identical retention of 2H (85%) via NIH shift to position 3.NIH shifts with [2,3,5,6-2H4]phenylalanine were also identical (84% 2H4). This suggests that the NIH shift is probably not directly mediated by the enzyme. Results with other phenylalanine substrates and analogues indicated that both rat liver and C. violaceum phenylalanine hydroxylases must utilize a very similar oxygenating intermediate. Recombinant human phenylalanine hydroxylase has been produced in E. coli and purified to homogeneity.68 The recombinant enzyme exists as a mixture of tetramers and dimers.By expression as a fusion protein, proteolytic degradation can be circumvented.69 Rat liver phenylalanine hydroxylase can be activated by cAMP-dependent protein kinase-mediated phosphorylation of Ser-16, and it has been suggested that this is a consequence of the introduction of the negatively charged phosphate group. By site-directed mutagenesis, Ser-16 has been replaced with a variety of other amino acids.70 Activation was only noted when Ser-16 was replaced with the negatively charged amino acids glutamine or aspartic acid.The mechanism of action of recombinant rat liver tyrosine hydroxylase, an Fe-containing enzyme, has been studied using a variety of 4-substituted phenylalanines as substrates.71 The enzyme accepts phenylalanine, 4-halogenated phenylalanines, and several other compounds, giving multiple hydroxylated products in most cases (Scheme 10).It was found that as the size of the substituent increases, the site of hydroxylation switched from position 4 to position 3 in the aromatic ring. The substituent at position 4 was sometimes eliminated, yielding tyrosine as product, or it could be migrated via an NIH shift, producing 3-substituted tyrosine. From the observations it was concluded that the transition state for hydroxylation must be highly electron-deficient or cationic in nature.Iron– oxo type intermediates 32 and 33 may be involved (Scheme 10). Purified recombinant rat liver tyrosine hydroxylase contains 0.5–0.7 Fe3+ atoms per subunit.72 During turnover, this is reduced by the tetrahydropterin to the Fe2+ form, though a fraction becomes reoxidized by O2. Five histidine residues are highly conserved in all tyrosine hydroxylase proteins that have been sequenced, and also in related phenylalanine hydroxylase and tryptophan hydroxylase.By site-directed mutagenesis, these residues in tyrosine hydroxylase have been replaced separately.73 Mutant proteins without histidine at positions 331 or 336 contained no iron and had no detectable activity; others contained wild-type levels of iron, and somewhat lower enzymic activity. This is consistent with His-331 and His-336 being ligands for the active site iron atom. Sequence data for the human tryptophan hydroxylase gene show considerable similarity to those for tyrosine hydroxylase and phenylalanine hydroxylase.74 Purification techniques for tryptophan hydroxylase from rat brain have been reported,75 and the tryptophan hydroxylase gene has been expressed in a human cell line.76 Deletion of the regulatory domains from the full-length sequence has allowed cloning and expression of the catalytic core, which displays high levels of trytophan hydroxylase activity.77 The structure and function of aromatic amino acid hydroxylases, 78 and human tyrosine hydroxylase79 have been reviewed. An extensive review on tyrosine hydroxylase has been published.80 2.11 Tyrosinase and melanin Melanin, the main pigment in mammalian skin, hair and eyes, is a heterogeneous polymer derived by oxidation of L-tyrosine (Scheme 11). The principal enzyme involved is tyrosinase, which catalyzes both the hydroxylation of L-tyrosine to L-DOPA 34, and the subsequent oxidation of L-DOPA to DOPAquinone 35.Several of the steps beyond DOPAquinone are apparently spontaneous chemical transformations.In recent studies, crystalline tyrosinase has been obtained from the skin of white silky fowl (Gallina lanigera),81 and partially purified enzyme has been isolated from potato (Lycopersicon esculentum).82 Tyrosinase isoforms have been identified in fruit bodies of the fungus Agaricus bisporus,83 and a cDNA encoding the protein has also been isolated from this source.84 The deduced amino acid sequence for the protein has significant homology to tyrosinase from Neurospora crassa.Six isozymes have been reported in the browned gills of Lentinus edodes,85 each isozyme consisting of two types of polypeptide. In the later stages of melanin biosynthesis, L-DOPAchrome 37 is transformed into either 5,6-dihydroxyindole-2-carboxylic acid (DHICA) 39 or 5,6-dihydroxyindole (DHI) 38 (Scheme 11). Polymerization of DHICA to melanin by a glycoprotein from mouse melanoma cells has been reported.86 An enzyme D-DOPAchrome tautomerase which converts the non-natural D-isomer of DOPAchrome into DHI has been isolated from human blood erythrocytes.87 A cDNA encoding this enzyme in rat liver has been cloned and sequenced.88 A comprehensive review of the mechanism of tyrosinase has been published.89 Other reviews cover its role in mammalian pigmentation,90 physiological implications of kinetic properties91 and molecular structure.92 The chemistry of melanins and melanogenesis has also been reviewed.93 R = 32 33 + + X R OFe X R OH R OH R X X OFe H R X OH R CO2H NH2 Scheme 10 Enzyme: tyrosine hydroxylase Dewick: The biosynthesis of shikimate metabolites 232.12 p-Aminobenzoic acid, anthranilic acid and related compounds The biosynthetic pathways to p-aminobenzoate 41 and anthranilate (o-aminobenzoate) 43 share many features (Scheme 12).A glutamine amidotransferase activity provides ammonia from glutamine, and this then reacts with chorismate either at the 4 or 2 positions, yielding 4-amino-4-deoxychorismate 40 or 2-amino-2-deoxyisochorismate 42.Lyase enzymes then release p-aminobenzoate or anthranilate from these respective intermediates. The two subunits PabA and PabB required for 4-amino-4-deoxychorismate synthesis in E. coli appear to act as a complex, but only recently has the successful isolation of the intact complex been reported.94 The association of the two subunits is enhanced by the presence of glutamine.Mutations in PabB can aVect its ability to associate with PabA, as well as its ability to aminate chorismate. Both PabA and PabB have cysteine residues essential for catalytic activity and/or subunit interaction.95 Anthranilate synthase from Ruta graveolens plants and cell cultures has been purified.96 Glutamine- and ammoniadependent activities copurified in all steps. cDNA cloning and complementation of an E. coli deletion mutant defective in anthranilate synthase showed young Ruta plants express two genes for functional ammonia-dependent AS· subunits.An AS‚ subunit is required for the glutamine-dependent reaction. The two AS· isozymes vary in function, one being elicitorinducible and involved in alkaloid biosynthesis, whilst the other is constitutive and strongly feedback inhibited by tryptophan. 97 The genes are diVerentially expressed according to the plant’s requirements for tryptophan or for defence-related alkaloid biosynthesis. Two anthranilate synthase isozymes have been detected in cell cultures of Ailanthus altissima.98 A glutamine-dependent isozyme was purified and characterized.An Arabidopsis thaliana mutant containing anthranilate synthase which shows enhanced feedback-resistance towards tryptophan also accumulates levels of tryptophan three-fold higher than the wild type plant.99 Tryptophan resistance was traced back to the single amino acid substitution Asp-341] Asn in the mutant. The enzymes anthranilate synthase, p-aminobenzoate synthase and isochorismate synthase (see Section 2.13) exhibit significant amino acid sequence homology and may thus share common mechanistic features.To probe this hypothesis, three compounds 44–46 designed to mimic, in their all-axial conformers 47, the putative transition state 25 have been synthesized and tested against the three E. coli-derived enzymes.100 All three were competitive inhibitors, binding HO CO2H NH2 HO CO2H NH2 O CO2H NH2 HO O HO HO NH CO2H HO O N CO2H HO HO NH O O NH HO HO NH CO2H O O NH CO2H 34 L-DOPA Melanins 31 L-Tyrosine 35 DOPAquinone 37 DOPAchrome 38 DHI 39 DHICA O2 O2 O2 –CO2 i i ii iii i 36 CycloDOPA Scheme 11 Enzymes: i, tyrosinase; ii, DOPAchrome tautomerase; iii, DHICA oxidase OH O CO2 – CO2 – NH2 O CO2 – CO2 – CO2 – NH2 O CO2 – 16 Chorismate OH O CO2 – CO2 – 16 Chorismate 40 4-Amino-4-deoxychorismate 41 p-Aminobenzoate Glu O CO2 – CO2 – 42 2-Amino-2-deoxyisochorismate Gln + CO2 – NH3 43 Anthranilate O CO2 – ii iii + Glu Gln + NH3 iv v vi NH2 NH2 + i Scheme 12 Enzymes: i, glutamine amidotransferase (PabA); ii, 4-amino-4-deoxychorismate synthase (PabB); iii, 4-amino-4- deoxychorismate lyase (PabC); iv, glutamine amidotransferase (anthranilate synthase II, AS‚, TrpG); v, 2-amino-2- deoxyisochorismate synthase; vi, 2-amino-2-deoxyisochorismate lyase; [v and vi/anthranilate synthase I (AS·, TrpE)] º 44 X = Y = OH 45 X = NH3 +; Y = OH 46 X = OH; Y = NH3 + 47 Y O CO2 – CO2 – X O CO2 – Y X –O2 C 24 Natural Product Reports, 1998strongly to anthranilate synthase and isochorismate synthase, but weakly to p-aminobenzoate synthase.The aYnity of the 6-amino-4-hydroxy isomer 45 was some ten times greater than the 4-amino-6-hydroxy isomer 46, though this is largely due to their conformational equilibria and the relative proportion of the all-axial conformer. The similarity between anthranilate synthase and isochorismate synthase is extended by the observation that both enzymes catalyze the formation of 2-amino- 2-deoxyisochorismate 42 in the presence of ammonia.The findings support a mechanism of direct substitution of the C-4 hydroxy group with a C-6 substituent via the proposed Mg-coordinated transition state 48 in the case of these two enzymes (Scheme 13). For p-aminobenzoate synthase, the mechanism must be somewhat diVerent, perhaps by a sequential mechanism (Scheme 14) rather than by direct substitution.The remarkable hydrolysis of an amine to an alcohol in the reverse reactions involving anthranilate synthase and p-aminobenzoate synthase has been studied and discussed.101 The enzymology of anthranilate synthase in microorganisms and plants has been reviewed.102 In the pathway to anthranilate by oxidative degradation of L-tryptophan 50 (Scheme 15), the first reaction is catalyzed by tryptophan 2,3-dioxygenase, producing N-formylkynurenine 51. The reaction is believed to involve addition of oxygen across the C-2–C-3 bond to form a dioxetane with subsequent decomposition. In a recent report,103 tryptophan 2,3-dioxygenase from yeast (Saccharomyces cerevisiae) has been purified.Hydrolytic cleavage of L-kynurenine 52 and L-3-hydroxykynurenine 53 to anthranilate 43 and 3- hydroxyanthranilate 54, respectively, is brought about by kynureninase, a PLP-dependent enzyme. The amino acid sequence of kynureninase from rat liver has been established, 104 and a cDNA clone encoding human kynureninase has been isolated.105 The predicted amino acid sequence of the human enzyme has high similarity to that of the rat liver enzyme and also to a S.cerevisiae gene product tentatively ascribed to kynureninase. A PLP-binding site could be defined. 4-Aminobenzoate hydroxylase is an FAD-, O2- and NAD(P)H-dependent monooxygenase responsible for the decarboxylative hydroxylation of p-aminobenzoate 41 to produce the 4-hydroxyaniline 55 moiety of N-(„-L-glutamyl)-4- hydroxyaniline 56 in the mushroom Agaricus bisporus (Scheme 16).A cDNA encoding this enzyme has been cloned and sequenced.106 Two regions on the protein share homology with other flavoproteins such as salicylate hydroxylase and 4-hydroxybenzoate hydroxylase. 48 25 42 2-Amino-2-deoxyisochorismate Y = N 49 Isochorismate Y = O O CO2 – OH –O2C O CO2 – HnO –O2C YHn Mg Enz O CO2 – –O2C YHn Scheme 13 Enzymes: anthranilate synthase/isochorismate synthase 40 4-Amino-4-deoxychorismate 16 Chorismate NH3 + OH O CO2 – CO2 – O CO2 – CO2 – Nu Enz NH2 O CO2 – CO2 – Enz Nu Scheme 14 Enzyme: p-aminobenzoate synthase NH CO2 – NH2 CO2 – O NH NH2 CHO CO2 – O NH2 NH2 CO2 – O NH2 NH2 OH CO2 – NH2 OH 50 L-Tryptophan CO2 – NH2 52 L-Kynurenine 51 N-Formylkynurenine 53 3-Hydroxykynurenine 43 Anthranilate i O2 ii iv Ala iii Ala iii 54 3-Hydroxyanthranilate Scheme 15 Enzymes: i, tryptophan 2,3-dioxygenase; ii, kynurenine formamidase; iii, kynureninase; iv, kynurenine 3-hydroxylase HO2C NH2 HO NH2 HO N H O CO2H NH2 56 i NAD(P)H O2 41 55 ii Glu Scheme 16 Enzymes: i, 4-aminobenzoate hydroxylase; ii, „-glutamyltransferase Dewick: The biosynthesis of shikimate metabolites 25Some cyclic hydroxamic acids based on 2,4-dihydroxy-2H- 1,4-benzoxazin-3(4H)-one (DIBOA) 58 found in maize, wheat, rye and other members of the Graminae, serve as major defence compounds against bacteria, fungi and insects.It has been established that part of the tryptophan biosynthetic pathway from anthranilate is used in their biosynthesis, though tryptophan itself is not a precursor (see ref. 1). A suggested pathway is via oxidative decarboxylation of 1-(2- carboxyanilino)-1-deoxyribulose 5-phosphate 61 (see Scheme 19) and subsequent cleavage of glyceraldehyde 3-phosphate. In more recent studies, it has been shown that maize (Zea mays) shoots grown in the presence of labelled indole 57 accumulate labelled DIMBOA 59 (Scheme 17).107 In particular, retention of label from [2-14C]indole suggested this compound was on the direct pathway; incorporation via anthranilic acid derived by cleavage of indole would involve loss of this label.Also, feeding of [2H6]indole gave DIMBOA with equal deuterium labelling at the three aromatic positions and at the acetal. The mechanism of benzoxazinone formation from indole has yet to be established. The chemistry of biologically active benzoxazinoids has been reviewed.108 2.13 Isochorismate synthase The formation of isochorismate 49 from chorismate 16 (Scheme 18, see also Scheme 13) creates a branchpoint away from the main shikimate pathway leading to the aromatic amino acids, and is utilized for the biosynthesis of some quinones (see Section 9) and simple phenolic acid derivatives such as 2,3-dihydroxybenzoic acid, a precursor of the siderophore enterobactin. The reaction is catalyzed by isochorismate synthase (isochorismate hydroxymutase) and has been shown to proceed via a double SN2* sequence.In independent studies109–111 it has been demonstrated that diVerent genes encode for isochorismate synthase according to its involvement in menaquinone (menF) or in enterobactin (entC) biosynthesis. An entC mutant of E. coli deficient in enterobactin biosynthesis grows normally and synthesizes wild-type levels of menaquinone under anaerobic conditions in iron-suYcient media.109 The gene responsible (menF) has been identified and sequenced, and shown to encode a protein 23.5% identical and 57.8% similar to EntC.110 The involvement of menF in menaquinone biosynthesis was proved via selective mutation studies and complementation of entC deletion mutants.Duplicate genes menF and dhbC for isochorismate synthase have also been implicated in Bacillus subtilis.111 MenF deletion mutants are able to produce wild-type levels of menaquinone and 2,3-dihydroxybenzoate, via utilization of the dhbC gene product.On the other hand, deletion of dhbC results in synthesis of only menaquinone, and menF is unable to compensate for the loss of 2,3-dihydroxybenzoate. It is shown that expression of dhbC, but not of menF, is regulated by iron concentration, and this repression can be abolished by mutations within the dhb promoter. Transgenic root cultures of Rubia peregrina containing the entC gene from E. coli have been established.112 This leads to increased isochorismate synthase activity and higher anthraquinone production. 3 Tryptophan and related compounds 3.1 Tryptophan L-Tryptophan 50 is formed from chorismate 16 via anthranilate 43 as shown in Scheme 19. A feature of this pathway is the Amadori rearrangement on phosphoribosylanthranilate 60 catalyzed by the enzyme phosphoribosylanthranilate isomerase. It has been shown using the enzyme from Saccharomyces cerevisiae that the primary product of a single turnover of 60 is NH NH O O NH O O OH NH O O OH MeO N O O OH N O O OH MeO OH OH Anthranilate 57 Indole Tryptophan 59 DIMBOA 58 DIBOA O2, NADPH O2, NA DPH O2, NADPH Scheme 17 OH O CO2 – CO2 – O CO2 – CO2 – OH OH2 16 Chorismate 49 Isochorismate Scheme 18 Enzyme: isochorismate synthase CO2 – NH2 CO2 – NH CH2OP OH HO O CO2 – NH O OP OH OH NH OH HO OP NH CO2 – NH2 NH 61 1-(2-carboxyanilino)-1-deoxyribulose 5-phosphate ii 50 L-Tryptophan 62 Indole-3-glycerol phosphate iii iv 43 Anthranilate v b2 Ser PLP Ser PLP 60 Phosphoribosylanthranilate Chorismate 57 Indole i v a v a2b2 Scheme 19 Enzymes: i, anthranilate synthase (TrpE, TrpG); ii, anthranilate phosphoribosyltransferase (TrpD); iii, phosphoribosylanthranilate isomerase (TrpF); iv, indole-3-glycerol phosphate synthase (TrpC); tryptophan synthase (TrpA, TrpB), · and ‚ refer to subunits of tryptophan synthase 26 Natural Product Reports, 1998fluorescent, but it slowly isomerizes to a stable non-fluorescent product, which is then the competent substrate for the next enzyme, indole-3-glycerol phosphate synthase.113 The fluorescent product is tentatively assigned as the enol amine 63, whilst the non-fluorescent product is the keto amine 1-(2- carboxyanilino)-1-deoxyribulose 5-phosphate 61 (Scheme 20).Conversion of 63 into 61 is not eVected by any of the three enzymes anthranilate phosphoribosyltransferase, phosphoribosylanthranilate isomerase or indole-3-glycerol phosphate synthase, and it appears that tryptophan biosynthesis is thus rate-limited by an uncatalyzed enol–keto tautomerism.Escherichia coli contains a bifunctional enzyme phosphoribosylanthranilate isomerase–indole-3-glycerol phosphate synthase. The individual monofunctional proteins have now been obtained through genetic engineering as stable monomeric proteins possessing full catalytic activity.114 The two monofunctional enzymes did not associate in vitro, nor in vivo by coexpression of the domains in the same cells. There appear to be no obvious selective advantages for the bifunctional enzyme over the monofunctional enzymes.In Arabidopsis thaliana, three genes encoding phosphoribosylanthranilate isomerase have been characterized.115 Two genes were virtually identical, whilst the third shared about 90% identity. The phosphoribosylanthranilate isomerase protein is not associated with indole-3-glycerol phosphate synthase as in most microorganisms. Transgenic plants expressing antisense phosphoribosylanthranilate isomerase RNA showed a significantly reduced level of protein and enzymic activity.A cDNA encoding indole-3-glycerol phosphate synthase has been isolated from A. thaliana, and deduced sequences for the encoded protein showed 22–38% identity to reported microbial sequences.116 Crystal structures for this enzyme isolated from the hyperthermophilic archaeon Sulfolobus solfataricus have been reported.117, 118 This monomeric enzyme shows only 30% sequence identity with the indole-3-glycerol phosphate synthase domain of the bifunctional enzyme from E.coli, but there is a high degree of structural similarity. Tryptophan synthase catalyzes the final reaction in the sequence, transformation of indole 3-glycerol phosphate 62 plus L-serine into L-tryptophan. It is a multienzyme complex (the ·2‚2 complex), comprising two · subunits and a ‚2 dimeric subunit. The · subunits also catalyze aldolytic cleavage of indole 3-glycerol phosphate, and, by use of the cofactor PLP, the ‚ subunits catalyze reaction of L-serine and indole 57, giving L-tryptophan.The overall reaction is catalyzed only by the ·2‚2 complex, and indole is not normally released as an intermediate, but is channelled between the active sites via a hydrophobic tunnel. In a single enzyme turnover of the ·2‚2 reaction, indole cannot be detected, but by site-directed mutagenesis, the size of the tunnel can be restricted thus impairing eYcient channelling of indole and allowing the direct observation of indole.119 This has been achieved in the Salmonella typhimurium enzyme by replacing Cys-170 in the ‚ subunit with tryptophan or phenylalanine.Evidence for an ·-amino acrylate SchiV base intermediate 65 in the ‚-reaction (Scheme 21) has been available only through absorption and fluorescence spectroscopy. Direct solid state NMR observation of 65 has now been obtained, using L-[3-13C]serine with the ·2‚2 complex from S. typhimurium.120 The ·-amino acrylate SchiV base is in equilibrium with its methyl keto amine tautomer 66 at the enzyme active site in the ‚ subunit.Conformational changes associated with formation of the ·-amino acrylate SchiV base 65 are believed to play a part in the mechanism by which the tryptophan synthase ‚ subunit accelerates cleavage of indole 3-glycerol phosphate catalyzed by the · subunit. This allosteric activation has been demonstrated by producing a mutant enzyme in which the ‚-active site lysine (Lys-87) was replaced with threonine.121 This enzyme failed to catalyze the ‚-reaction, though there was normal activity for the ·-reaction.A SchiV base between PLP and serine could be formed at the ‚-active site, and using a chemical rescue method, this was converted into an ·-amino acrylate type intermediate using a high concentration of ammonia. This led to a significant enhancement in ·-reaction activity, showing the allosteric interaction in the absence of tryptophan formation. From a series of kinetic studies and isotope eVects it is concluded that the chemical reaction in which indole attacks the enzyme-bound ·-amino acrylate SchiV base is best described as occurring in two steps, a binding step followed by a Michael addition onto the conjugated double bond.122 Again, conversion of the serine external aldimine 64 into the ·-amino acrylate SchiV base 65 is implicated as the ‚-site process which activates the ·-site.cDNAs encoding the ·-protein of tryptophan synthase have been isolated from Arabidopsis thaliana123 and maize (Zea mays).124 In both cases, the cDNA clone complemented an E.coli trpA mutant. Several lines of evidence suggest the active tryptophan synthase enzyme from A. thaliana is a heterosubunit complex, probably analogous to the prokaryotic ·2‚2 complex. The tryptophan biosynthetic enzymes phosphoribosylanthranilate isomerase and the · subunit of tryptophan synthase from A.thaliana are synthesized as higher molecular weight precursors, are then imported into chloroplasts and processed into their mature forms.125 Two DNA fragments from Buchnera, the prokaryotic symbiont of the aphid Schlechtendalia chinensis have been cloned and sequenced and shown to contain the trpEG genes in the smaller fragment, and trpDCFBA genes in the larger.126 Both gene clusters are present as a single copy and probably constitute two transcriptional units. The structure, function and genetic aspects of tryptophan synthase from Salmonella typhimurium have been covered in recent reviews.127, 128 The genetics of tryptophan biosynthesis in plants has also been reviewed.129 Water-stressed leaves of tomato (Lycopersicon esculentum) accumulate increased amounts of L-tryptophan and N-malonyl-L-tryptophan.130 Crude leaf fractions were able to malonylate both L- and D-tryptophan.A partially purified tryptophan N-malonyltransferase was obtained which catalyzed malonyl CoA-dependent malonylation of both enantiomers, 131 the same protein being responsible for utilization of both L- and D-forms of the amino acid.A tryptophan 2*,3*-oxidase from Chromobacterium violaceum catalyzes double bond formation in the side-chain of L-tryptophan and related substrates.132 The product from L-tryptophan has been established to have the Z configuration CO2 – NH POCH2 HO OH O CO2 – PO OH HO OH NH CO2 – PO O HO OH NH 63 (enol amine) 60 Phosphoribosylanthranilate 61 (keto amine) 62 Indole-3-glycerol phosphate i ii Scheme 20 Enzymes: i, phosphoribosylanthranilate isomerase; ii, indole-3-glycerol phosphate synthase Dewick: The biosynthesis of shikimate metabolites 2767.The reaction, for which no cofactor was required, could also occur if tryptophan was part of a short peptide (5–24 residues) irrespective of its position in the sequence. 3.2 Tryptophanase Tryptophanase [tryptophan indole-lyase (deaminating)] catalyzes the synthesis of tryptophan from indole, pyruvate and ammonia in a PLP-dependent reaction, though its physiological role is cleavage of tryptophan rather than its synthesis. The gene for tryptophanase from Enterobacter aerogenes has been highly expressed in E.coli, giving cells capable of synthesizing L-tryptophan from indole, pyruvate and ammonia.133 The enzyme was subsequently purified and shown to have higher stability towards heat than tryptophanase from E.coli. A kinetic study of tryptophanase from E. coli using ·-deuteriated L-tryptophan has suggested that proton transfer to carbon to form the indoleninium intermediate 68 (Scheme 22) is relatively slow, and probably at least relatively ratelimiting. 134 With aza and thia analogues of L-tryptophan, the rate-determining step changes according to the position and type of heteroatom. Structural, spectral and catalytic aspects of tryptophanase have been reviewed.135 3.3 Indole-3-acetic acid and related metabolites The plant growth hormone indole-3-acetic acid (IAA) 69 can be produced from tryptophan by several diVerent pathways according to species.Intermediates indole-3-pyruvic acid, tryptamine, indole-3-acetamide or indole-3-acetaldoxime may be involved (Scheme 23). Further metabolism may then yield a variety of indole or oxindole derivatives. The indole-3-acetamide pathway is not usually found in plants, but is typical of many plant pathogenic and symbiotic bacteria.Plants usually gain the capacity for converting indole- 3-acetamide into IAA after transformation with organisms such as Agrobacterium tumefaciens and incorporation of the gene encoding indoleacetamide hydrolase. Results of experiments with suspension cultures of scorzonera (Scorzonera hispanica) and carrot (Daucus carota) have demonstrated that this enzyme is absent from normal cells, but is present after transformation with A. tumefaciens.136 Conversion of tryptophan into indole-3-acetamide is catalyzed by the flavoprotein tryptophan 2-monooxygenase.The gene encoding this enzyme in Pseudomonas savastanoi has been isolated and expressed in E. coli.137 The purified enzyme contained tightly bound indole- 3-acetamide which could be removed by dialysis. The enzyme would accept phenylalanine and methionine, amino acids having large hydrophobic side-chains, as alternative substrates. Kinetic studies138, 139 on the enzyme have indicated that many features of the reaction resemble those of D-amino acid oxidase, and a mechanism involving the same initial steps is proposed (Scheme 24).Cleavage of the C–H bond is likely to produce a carbanion 70 which interacts with the flavin. The tryptophan 2-monooxygenase reaction then diVers in that a rapid decarboxylation of the imino acid occurs. Indole-3-acetonitrile features in the pathway to IAA in Arabidopsis thaliana. This is converted into IAA by the hydrolytic enzyme nitrilase, which is known to exist in at least four NH PO O– N+ Lys H Enz NH PO O– N+ H CO2 – HO NH PO O– N+ H CO2 – HO NH PO O– N+ H CO2 – NH PO O N+ H CO2 – NH PO O– N+ H CO2 – HN + NH PO O– N+ H CO2 – NH NH PO O– N+ H CO2 – NH Ser Enz-Lys H+ H+ Indole H+ Enz-Lys Trp Internal aldimine 64 External aldimine 66 Methyl keto amine tautomer 65 a-Amino acrylate Schiff base + + + + + H H H Scheme 21 Enzyme: tryptophan synthase NH CO2H NH2 67 28 Natural Product Reports, 1998isoforms.The nucleotide sequence of the nitrilase I gene has been reported.140 The gene encoding nitrilase II has been integrated into the genome of Nicotiana tabacum, and catalytic activity of the expressed protein demonstrated.141 Hydrolysis of indole-3-acetonitrile to IAA has also been reported for the first time in the plant-associated bacteria Agrobacterium and Rhizobium, and the enzyme has been purified.142 The indole-3-pyruvate route rather than the indole-3- acetamide pathway is operative in a plant-associated strain of Erwinia herbicola.143 Indole-3-pyruvate and indole-3- acetaldehyde, but not indole-3-acetamide or other potential intermediates could be converted into IAA, and these compounds were also detected in the cultures.Similar conclusions are reported for Bradyrhizobium elkanii,144 and some species of Rhizoctonia.145 The enzyme activity indolepyruvate decarboxylase was detected in cell-free extracts of B. elkanii, and indole-3-ethanol (tryptophol) was a further metabolite from both tryptophan and indole-3-acetaldehyde in R.solani. The soil bacterium Azospirillum brasiliense normally employs the indole-3-acetamide pathway for IAA synthesis, but under microaerophilic conditions, it can convert indole-3-pyruvic acid, indole-3-lactic acid and indole to give IAA.146 Two potential indole-3-acetaldehyde dehydrogenases catalyzing the NAD+-dependent conversion of indole-3-acetaldehyde into IAA have been identified in the phytopathogenic fungus Ustilago maydis.147 However, one activity appeared to be primarily involved in oxidation of ethanol to acetic acid.The second activity was concerned with conversion of indole-3- acetaldehyde derived from tryptamine, and was induced by arabinose, but repressed in the presence of glucose. A soluble auxin-binding protein from mung bean (Vigna radiata) was shown to be an indole-3-acetaldehyde reductase activity catalyzing NADPH-dependent conversion of indole-3- acetaldehyde into indole-3-ethanol.148 It was found to have high amino acid sequence homology with alcohol dehydrogenase enzymes, and indole-3-acetaldehyde was actually a rather poor substrate compared with several other aldehydes.An indole-3-ethanol oxidase enzyme has been isolated from Internal aldimine External aldimine B1 HB1 B1 68 Indoleninium intermediate B1 gem-Diamine iminoacrylate Internal aldimine + + B1 B1 Amino acrylate Schiff base + HN CO2 – H NH2 Lys-270 HN PLP N CO2 – H +NH PLP N CO2 – +NH PLP H B2 H B2 N+ CO2 – +NH PLP H B2 H N H B2 CO2 – NH PLP H2 N Lys-270 N H CO2 – +NH2 PLP HN Lys-270 N CO2 – +NH PLP H B2 Lys-270 NH2 Scheme 22 Enzyme: tryptophanase NH CO2H Indole-3-acetaldoxime Tryptophan Indole-3-acetonitrile Indole-3-acetamide Tryptamine Indole-3-pyruvic acid Indole-3-lactic acid Indole-3-acetaldehyde Indole-3-ethanol (Tryptophol) 69 IAA Scheme 23 Dewick: The biosynthesis of shikimate metabolites 29seeds of bean (Phaseolus vulgaris).149 This enzyme required molecular oxygen and produced H2O2. An alternative pathway to IAA via indole but not tryptophan has also been suggested, particularly in maize (Zea mays) (see ref. 1). Clear-cut confirmation of this has yet to be achieved. Tryptophan was a precursor in dark grown maize seedlings,150 and results from labelling experiments on seedlings grown in 2H2O were complicated by exchange reactions and the randomization of label during MS analysis.151 Although indole appeared to be converted into IAA without the intermediacy of tryptophan in maize endosperm, there was evidence to suggest that several diVerent pathways to IAA coexisted.152 Catabolism of IAA by some strains of Bradyrhizobium japonicum leads to oxindole derivatives 71–73, isatinic acid 74 and eventual formation of anthranilic acid (Scheme 25).153 The homologue of IAA, indole-3-butyric acid (IBA) 75, is found naturally in some plants, and has been thought to arise from IAA and acetyl CoA via a typical acetate chain extension process.Recent investigations on the biosynthesis of IBA in maize (Zea mays) suggest a two-step reaction.154 A microsomal membrane preparation from dark grown shoots and roots converted 14C-labelled IAA and acetyl CoA in the presence of ATP into an unknown product, but not into IBA. The labelled product was then eYciently transformed into IBA when supplied to other maize cell fractions along with NADPH.On the other hand, microsomal membranes from light grown shoots and roots converted IAA into IBA when supplied with acetyl CoA, and the reaction was enhanced by the addition of ATP and NADPH. In this system, the unknown product did not accumulate. The product hydrolyzes to IAA, is not formed in the absence of acetyl CoA and is not IAA CoA. It is believed to be an intermediate in the reaction, possibly a conjugate 76 of IAA with ADP. In further studies,155 the indole-3-butyric acid synthetase activity has been partially purified.When obtained from dark grown seedlings, this consisted of two proteins, one of which was responsible for producing the intermediate. A single protein obtained from light grown seedlings appeared to be analogous to the combined dark grown proteins, and light may induce combination of the proteins to form a single enzyme which catalyzes the conversion without release of an intermediate. Recent reviews discuss IAA biosynthesis and metabolism, 156–158 the synthesis of phytohormones by plantassociated bacteria,159 and the structure and function of indolepyruvate decarboxylase.160 3.4 Indigo and indirubin There is growing evidence that the pigment indigo 77 is derived from indole by a sequence not involving tryptophan (compare DIMBOA, Section 2.12). In recent studies,161 2H-labelled indole supplied to shoot cultures of Polygonum tinctorium was incorporated specifically into both indigo and indirubin 78. In N N NH N O O R R CO2 – NH3 + H N N NH N O O R R CO2 – NH3 + N N X N O O R R CO2 – NH2 + H NH N NH N O O R R CO2 – NH2 + R O NH2 NH – – – B B H+ CO2 O2 R = 70 Scheme 24 Enzyme: tryptophan 2-monooxygenase NH CO2H NH CO2H HO O NH OH O NH O O NH2 CO2H O CO2H NH2 69 IAA 72 Dioxindole 71 Dioxindole-3-acetic acid 73 Isatin 43 Anthranilic acid 74 Isatinic acid Scheme 25 NH CO2H NH O P OH O P O OH O CO.SCoA OH O O OH OH Ad 75 IBA 76 NH HN O O HN NH O O 77 Indigo 78 Indirubin 30 Natural Product Reports, 1998suspension and root cultures, however, the feeding of indole suppressed indigo biosynthesis, and only indirubin accumulated.Feeding of tryptophan did not lead to the production of either labelled product. 4 Phenols and phenolic acids 4.1 Phenol and tyrosine phenol-lyase Tyrosine phenol-lyase is found in various bacteria and catalyzes the PLP-dependent elimination of phenol from L-tyrosine, giving pyruvate and ammonia. The crystal structure of the enzyme from Erwinia herbicola has recently been established.162 The enzyme exists as a tetramer.Site-directed mutagenesis of the Citrobacter freundii enzyme has been used to establish that Tyr-71, an invariant residue in all known tyrosine phenol-lyase sequences, has a dual role.163 It acts as a general acid catalyst in the reaction mechanism (Scheme 26), and is also involved in binding the cofactor PLP. All enzymic activity is lost when Tyr-71 is replaced with phenylalanine.Histidine-343 is also conserved in all known sequences of both tyrosine phenol-lyase and the related tryptophan indole-lyase (tryptophanase, see Section 3.2). However, site-directed mutagenesis experiments have established that this residue is not an essential basic group in the C. freundii enzyme, although it does appear to play an important function in catalysis, perhaps facilitating a conformational change when the substrates bind.164 Replacement of His-343 with Ala resulted in lower rates of elimination. 4.2 4-Hydroxybenzoic acid 4-Hydroxybenzoate 79 is produced in bacteria from chorismate via the enzyme chorismate lyase, but in plants it is synthesized by side-chain degradation of cinnamic acids (Scheme 27). Chorismate lyase is not normally present in plants; in E. coli it is encoded by the gene ubiC and provides entry to ubiquinone biosynthesis (see Section 9.2). Expression of ubiC in tobacco (Nicotiana tabacum) has led to transgenic plants showing high chorismate-lyase activity, and in which 4-hydroxybenzoate was accumulated as a glucoside.165 4-Hydroxybenzoate content was some 1000 times higher than in untransformed plants.Feeding experiments with [1,7- 13C2]shikimic acid yielded double-labelled 4-hydroxybenzoate and therefore proved that the chorismate-lyase pathway was now operative rather than the route via cinnamic acids. 4-O-Glucosylbenzoic acid 80 is characteristically accumulated in shikonin-free cell cultures of Lithospermum erythrorhizon, and is located mainly in the vacuoles. Enzymes involved in its synthesis and subsequent hydrolysis, namely 4-hydroxybenzoate glucosyltransferase and ‚-glucosidase, respectively, are located exclusively in the cytosol.166 This suggests the glucoside is synthesized in the cytosol, but is then transported and stored in the vacuole.It is then again transported to be hydrolyzed and subsequently utilized in shikonin biosynthesis (see also Section 9.3). 4.3 Salicylic acid Salicylic (2-hydroxybenzoic) acid has long been known as a plant constituent, but in recent years it has been identified to have plant hormone activity as a natural inducer of disease resistance. Its biosynthetic origins appear to be from cinnamic acid via a chain shortening mechanism, but two routes by way of the intermediates 2-coumaric acid or benzoic acid may be involved, depending on the sequence of ortho-hydroxylation and chain shortening.In infected cucumber (Cucumis sativus) plants, the pathway cinnamic acid]benzoic acid]salicyclic acid is implicated.167 Both labelled phenylalanine and benzoic acid were incorporated into salicyclic acid, and 2-aminoindane- 2-phosphonic acid, a specific inhibitor of PAL (see Section 5.2), completely inhibited the incorporation of phenylalanine but not that of benzoic acid. A soluble benzoic acid 2-hydroxylase has been obtained from tobacco (Nicotiana tabacum) and has been partially purified.168 The enzyme had characteristics of a P450-dependent system, and was strongly induced in tobacco leaves by inoculation with tobacco mosaic virus, or by infiltration with benzoic acid.Shoots of rice (Oryza sativa) also use the same pathway, readily converting cinnamic acid and benzoic acid into salicylic acid.169 Although salicyclic acid is present largely in the free form in rice shoots, endogenously supplied salicyclic acid is conjugated as a glucoside by the action of an inducible glucosyltransferase.An analogous enzyme activity has been reported in salicylic acid-treated tobacco leaves.170 This enzyme may play an important role in regulating the levels of free salicylic acid. The isolation of the isomeric mixture 2-carbomethoxyoxepin 81 and 1-carbomethoxybenzene 1,2-oxide 82 from cultures of the fungus Phellinus tremulae has suggested that arene oxides CO2 – H +NH PLP O H CO2 – +NH PLP O H HO Tyr-71 CO2 – +NH PLP O –O Tyr-71 H B2 H CO2 – +NH PLP O HO Tyr-71 B2 H CO2 – H +NH PLP O H B1 B2 HB1 B2 HB1 B1 L-Tyrosine + PLP Scheme 26 Enzyme: tyrosine phenol-lyase OH O CO2 – CO2 – OH CO2 – CO2 – OH CO2 – NH2 16 Chorismate 79 i OGlc CO2 – L-Phe 80 Scheme 27 Enzyme: i, chorismate-lyase Dewick: The biosynthesis of shikimate metabolites 31may be intermediates in the biosynthesis of methyl salicylate 83 and salicyclic acid in this organism.171 [7-13C]Benzoic acid was eYciently incorporated into methyl salicylate, methyl benzoate and 2-carbomethoxyoxepin, supporting a sequence benzoic acid]methyl benzoate]oxepin 81, then rearrangement to methyl salicylate (Scheme 28).The acid-catalyzed rearrangement of 82 to methyl salicylate has been shown to occur readily and it involves an NIH shift with migration of the carbomethoxy group. The biosynthesis and metabolism of salicylic acid has been reviewed.172 4.4 Homogentisic acid Homogentisic acid 85 is a tyrosine-derived phenylacetic acid derivative formed by the action of 4-hydroxyphenylpyruvate dioxygenase on 4-hydroxyphenylpyruvate 30 in the presence of oxygen.Molecular oxygen is incorporated into the carboxy and new hydroxy groups, and a mechanism involving oxidative decarboxylation to a peroxy acid 84, which then oxygenates the ring and allows migration of the side-chain has been proposed (Scheme 29). Homogentisic acid is of importance as a source of quinonoid derivatives, e.g. tocopherols and plastoquinones (see Section 9.4).The enzyme has also been implicated in the biosynthesis of melanin-like pigments (pyomelanin) in microorganisms.173 Thus, pigment formation correlated with homogentisic acid production and expression of the enzyme 4-hydroxyphenylpyruvate dioxygenase in three disparate marine species, Vibrio cholerae, a Hyphomonas strain, and Shewanella colwelliana. 4-Hydroxyphenylpyruvate dioxygenase isolated from rat liver also displays a related oxidative decarboxylation reaction on ·-ketoisocaproate 86 giving ‚-hydroxyisovalerate 87 (Scheme 30).174 In fact, isolation and sequence determination of the enzyme catalyzing the reaction of Scheme 30 indicated the enzyme to be analogous to liver-specific rat F antigen, which is believed to be a species variant of 4-hydroxyphenylpyruvate dioxygenase.Both activities are provided by a single enzyme. The enzyme required the cofactors molecular oxygen, Fe2+, ascorbate and dithiothreitol for activity.Molecular cloning and expression of the protein in E. coli has allowed further studies of its substrate specificity.175 The C-terminal 14 amino acid portion is indispensible for catalytic activity; truncation results in complete loss of both activities.176 5 Phenylpropanoids 5.1 General Stress-induced phenylpropanoid metabolism has been reviewed.177 5.2 Phenylalanine ammonia-lyase Phenylalanine ammonia-lyase (PAL) catalyzes the stereospecific elimination of ammonia from L-phenylalanine to give (E)-cinnamic acid 89 (Scheme 31).Recent sources from which the enzyme has been isolated and purified include the plant basil (Ocimum basilicum),178 and microorganisms Rhodosporidium toruloides,179 Rhodotorula glutinis,180, 181 Sporidiobolus pararoseus182 and Ustilago maydis.183 The biological function of the enzyme in yeasts and fungi is still somewhat unclear. Gene sequences encoding PAL in Arabidopsis thaliana,184 O CO2Me O CO2Me CO2Me 82 81 H+ 83 CO2H CO2 Me OH Scheme 28 85 Homogentisate 30 4-Hydroxyphenylpyruvate CO2 O2 i, Fe2+ 84 O CO2 – OH O O OH O H O O– OH O O CO2 – OH OH CO2 – OH Scheme 29 Enzyme: i, 4-hydroxyphenylpyruvate dioxygenase 86 87 O2, Fe2+ i CO2 – O CO2 – HO Scheme 30 Enzyme: i, 4-hydroxyphenylpyruvate dioxygenase L-Phe 88 89 Cinnamate + + + + + + + + HN NH OH CO2 – NH3 HR HS CO2 – NH3 HR HS OH HN NH B Enz HN NH OH CO2 – H B Enz CO2 – NH3 HO HN NH H B Enz Scheme 31 Enzyme: phenylalanine ammonia-lyase (PAL) 32 Natural Product Reports, 1998Bromheadia finlaysoniana,185 lemon (Citrus limon),186 tobacco (Nicotiana tabacum),187 rice (Oryza sativa),188 parsley (Petroselinum crispum),189 hybrid aspen (Populus kitakamiensis), 190, 191 Stylosanthes humulis192 and wheat (Triticum aestivum)193 have been reported, and shown to be highly similar to those known for other plants.Small gene families are frequently reported, e.g. in Arabidopsis (three PAL genes),184 tobacco,187 rice,188 parsley (four genes),189 aspen (four genes),190, 191 and wheat (two genes).193 A PAL gene from poplar (Populus trichocarpa#deltoides) has been expressed in cell cultures from the insect Spodoptera frugiperda.194 Sitedirected mutagenesis and expression of a mutant enzyme demonstrated that Ser-202 was essential for catalytic action.Transgenic tobacco plants containing a PAL gene from bean overexpress PAL, and synthesize increased levels of chlorogenic (5-O-caVeoylquinic) acid and 4-coumaric acid glucoside.195 The active site of PAL is known to contain a dehydroalanine residue which participates in the elimination of ammonia.This dehydroalanine is produced post-translationally by modification of a serine residue, Ser-202. By site-directed mutagenesis, Ser-202 in parsley PAL has been replaced with alanine or threonine.196 Mutant proteins and recombinant wild-type enzyme were then treated with NaBH4. All three proteins were virtually inactive with L-phenylalanine, but catalyzed deamination of L-(4-nitrophenyl)alanine.L-Tyrosine was a poor substrate, whilst DL-m-tyrosine was converted at a rate comparable to phenylalanine. These findings are consistent with a mechanism in which the crucial step is an electrophilic attack of the dehydroalanine prosthetic group at position 2 or 6 of the phenyl group (Scheme 31). In the resulting carbocation 88, the ‚-HSi atom is activated in a similar way as it is in the nitro analogue 90.This mechanism contrasts with that previously proposed which invokes attack of the phenylalanine nitrogen onto dehydroalanine, then removal of the ‚-HSi to generate a carbanion intermediate (Scheme 32). 5.3 Hydroxycinnamic acids and esters Hydroxycinnamic acids are usually obtained by further aromatic substitution of the cinnamic acid formed by the action of PAL, sequentially building up the oxygenation and methylation pattern. Cinnamic acid 4-hydroxylase (CA4H) is a cytochrome P450-dependent enzyme transforming cinnamic acid into 4-coumaric acid.Genes encoding for CA4H in Populus tremuloides,197 P. kitakamiensis198 and Catharanthus roseus,199 have recently been isolated and sequenced. A small gene family is present in hydrid aspen (P. kitakamiensis).198 Translational fusions of CA4H with cytochrome P450 reductase allow production of an enzyme that requires no additional cofactors.199 Methods for the isolation of CA4H from plant microsomes have been described.200 Ferulate 5-hydroxylase from Arabidopsis thaliana has a maximum 34% sequence identity with other P450 enzymes, much less with CA4H, and is considered to belong to a new subfamily of P450 monooxygenases.201 A cinnamic acid O-methyltransferase activity in wheat (Triticum aestivum) seedlings showed a preference for caVeic acid as substrate (giving ferulic acid) in the early stages of development prior to lignin deposition.202 After lignin deposition began, the O-methyltransferase activity then showed a preference for 5-hydroxyferulic acid (giving sinapic acid).The substrate specificity of a bispecific caVeic acid–5- hydroxyferulic acid O-methyltransferase from aspen (Populus tremuloides) has been investigated.203 Although a wide range of phenolic substrates were accepted, none was as eVective as the natural substrates. Site-directed mutagenesis experiments demonstrated that conserved cysteine residues played no important role in the catalytic mechanism.Nucleotide sequences for the aspen gene,204 and a caVeic acid O-methyltransferase gene from almond (Prunus amygdalus)205 have been established. There is growing evidence that in many plants, methylation may occur with a cinnamoyl CoA substrate rather than on the free cinnamic acid. Both types of enzyme are found in Zinnia elegans, and they are diVerentially expressed.206 Nucleotide sequences for caVeoyl CoA 3-Omethyltransferase genes in long-stalked chickweed (Stellaria longipes)207 and aspen (Populus tremuloides)208 have been established.Activation of cinnamic acids to coenzyme A esters is accomplished by ligase enzymes. Gene sequences for 4- coumarate:CoA ligase from Loblolly pine (Pinus taeda),209 Arabidopsis thaliana,210 tobacco (Nicotiana tabacum)211 and Lithospermum erythrorhizon212 and for a ferulate–sinapatespecific CoA ligase from aspen (Populus tremuloides)213 have been reported. Most of the genes display high homology.Multiple genes were isolated from tobacco and Lithospermum. The 4-coumarate:CoA ligase proteins typically accept other hydroxycinnamic acids, e.g. ferulate and caVeate, as substrates. Coenzyme A esters are involved in the acylation of shikimic acid in young green fruits of date (Phoenix dactylifera) and of quinic acid in leaves of chicory (Cichorium endivia).214 The enzymes were partially purified, then treated with amino acid-specific reagents to demonstrate that histidine, cysteine and lysine residues were important for catalytic activity.Four hydroxycinnamoyltransferases in gametophytes and sporophytes of horsetail (Equisetum arvense) have been partially purified and studied.215 These enzymes utilize caVeoyl CoA 91 to acylate meso-tartaric acid (giving mono- then di-esters 94 and 95), shikimic acid (giving the 5-ester dactylifric acid 93) and quinic acid (giving the 5-ester chlorogenic acid 92) (Scheme 33). The enzymes varied according to their ability to accept caVeoyl CoA and 4-coumaroyl CoA as substrates; cinnamoyl CoA and feruloyl CoA were less eVective donors.In the case of tartaric acid as an acceptor, only the meso isomer was utilized. An enzyme from tobacco (Nicotiana tabacum) 90 + + + + + CO2 – NH2 N –O O– CO2 – NH2 N O O– O NH H2N HS CO2 – H Ph HR B+ H OH NH B H2N HS CO2 – H Ph HR B B OH NH B H2N CO2 – H B H Ph H O NH B NH2 B H Ph CO2 – O NH NH3 B+ H H B H Ph CO2 – H+ + + – + + + Scheme 32 Enzyme: phenylalanine ammonia-lyase (PAL) Dewick: The biosynthesis of shikimate metabolites 33that transfers hydroxycinnamoyl groups from their CoA esters to the ¢-hydroxy group of 16-hydroxypalmitic acid has been purified.216 Feruloyl CoA was the most eVective donor, though sinapoyl CoA and 4-coumaroyl CoA could also be utilized; a number of long chain 1-alcohols and ¢-hydroxyfatty acids would function as acceptors.This type of enzyme activity has been detected in a wide range of plant systems.217 Acylation of the amine tyramine is catalyzed by a pathogen elicitorinducible enzyme preparation from cell suspension cultures of potato (Solanum tuberosum).218 Feruloyl CoA was the preferred acyl donor, and tyramine the acceptor, though octopamine and dopamine were also utilized less eVectively.Purification of the enzyme showed it to be a heterodimer with the subunits non-covalently associated.219 Cell-free extracts of Aphelandra tetragona stem are able to acylate the polyamines spermine 96 and spermidine, but not putrescine, with 4-coumaroyl CoA acting as a better donor than feruloyl CoA.220 The product from reaction of spermine with 4-coumaroyl CoA is a monoamide, though the position of acylation has not been confirmed.Roots of Aphelandra contain the macrocyclic polyamine alkaloid aphelandrine 97 which is composed of spermine with two 4-coumaroyl residues, and the demonstrated enzyme activity may be involved in the early stages of its biosynthesis. A variety of hydroxycinnamic acid conjugates are known to be formed by processes involving glucose esters rather than the coenzyme A esters as the activated form of the acid.An enzyme purified from Brassica napus has been shown to transfer glucose from UDPglucose to sinapic acid in the synthesis of 1-O-sinapoyl glucose.221 A short review describes the biosynthesis of rosmarinic acid, the ester of caVeic acid with 3,4-dihydroxyphenyllactic acid, in Coleus blumei.222 5.4 Coumarins The isolation of a glucosyltransferase activity from Duboisia myoporoides converting scopoletin 98 into scopolin 99 has been reported.223 5.5 Lignins Lignins are natural phenolic polymers found in plant cells and are generally believed to arise by phenolic oxidative coupling of hydroxycinnamyl alcohol monomers (monolignols) brought about by peroxidase enzymes. The most important of these monolignols are 4-hydroxycinnamyl (4-coumaryl) alcohol, coniferyl alcohol and sinapyl alcohol which are derived, respectively, from 4-coumaric acid, ferulic acid and sinapic acid via the corresponding coenzyme A esters and aldehydes.However, recent observations have identified an additional pathway in which methylation occurs at the coenzyme A level rather than on the free cinnamic acids (see Section 5.3). Altered lignin composition in transgenic tobacco (Nicotiana tabacum)224 and Populus tremula x P. alba225 has been achieved by the expression of antisense gene constructs for the caVeic acid–5-hydroxyferulic acid bispecific O-methyltransferase.In certain lines, this led to a reduced O-methyltransferase activity, 92 Chlorogenic acid 93 Dactylifric acid meso-Tartaric acid 95 Quinic acid 94 Caffeoyl-CoA Shikimic acid 91 Caffeoyl-CoA O CO2H HO HO OH O OH OH O HO OH O OH OH CO2H COSCoA HO HO HO HO O CO2H H O HO CO2H H HO HO O CO2H H O CO2H H OH HO O O Scheme 33 97 Aphelandrine 96 Spermine O OH H H O N NH NH O HN H H2N N H H N NH2 98 Scopoletin R = H 99 Scopolin R = Glc O O MeO RO 34 Natural Product Reports, 1998a reduction in the syringyl (4-hydroxy-3,5-dimethoxyphenyl) to guaiacyl (4-hydroxy-3-methoxyphenyl) ratio in the lignin, and the appearance of a new component, namely 5-hydroxyguaiacyl residues.Introduction of chimeric sense– antisense gene constructs for 4-coumarate CoA ligase into tobacco plants also yielded lower enzyme activity, and a reduction in lignin content.226 Reduction of cinnamoyl CoA precursors to cinnamyl alcohols is catalyzed by two NADPH-dependent enzymes cinnamoyl CoA:NADPH oxidoreductase and cinnamyl alcohol dehydrogenase (CAD). Gene sequences encoding for CAD have recently been reported for alfalfa (Medicago sativa),227 poplar (Populus trichocarpa x P.deltoides),227 Arabidopsis thaliana,228, 229 Pinus radiata230 and Pinus taeda.231 Deduced amino acid sequences for the proteins were 50–80% identical to those established in other plants. Site-directed mutagenesis has been used to demonstrate the involvement of Ser-212 in Eucalyptus gunni CAD.232 This residue is conserved in the enzyme from several species, and it also corresponds to Asp-223 in horse liver alcohol dehydrogenase, an NADHdependent enzyme.A protein in which Ser-212 was replaced with aspartic acid was overexpressed in E. coli, and shown to have significantly decreased catalytic eYciency with NADP+, but an unchanged low activity with NAD+ (about 4% of the activity of wild-type enzyme using NADP+).This implicates Ser-212 in recognition of the phosphate grouping in NADPH. Introduction of an antisense construct for the cinnamyl alcohol dehydrogenase gene from Aralia cordata into tobacco has resulted in transgenic plants with much lower CAD enzyme activity.233 Lignin content was not significantly aVected, but there was an observed increase in the levels of 4-hydroxycinnamaldehyde groups incorporated into the lignin.Monolignols such as coniferyl alcohol and sinapyl alcohol may be synthesized and stored as glycosides, and then released enzymically when required for lignin biosynthesis. Coniferin 100 typically accumulates to high levels in gymnosperms prior to lignin deposition during the spring. Two ‚-glucosidase enzymes have been isolated from lodgepole pine (Pinus contorta var. latifolia) and shown to hydrolyze coniferin and syringin 101.234 One of these enzymes was purified, shown to be a homodimer, and had an N-terminal sequence with high homology to other known plant ‚-glucosidases.Coniferin labelled with 3H in its side-chain and 14C in the glucose residue was transformed into syringin in Magnolia kobus with loss of all 14C label.235 Thus, conversion of guaiacyl to syringyl substitution must occur by preliminary hydrolysis to the monolignol. EPR spectroscopy has been used to study the eVects of methoxy substitution on the unpaired electron distribution in lignin precursor radicals.236 It was deduced that methoxy substitution increases the unpaired electron density on the phenolic oxygen, and that this subsequently determines the nature of bond formation during polymerization, explaining the high degree of selectivity observed in the incorporation of phenylpropanoids into lignin.Thus, it is predicted that for 4-coumaryl alcohol, coupling is primarily 8–8 with a small amount of 8–5, and for coniferyl alcohol mainly 8–8 and 8–5 with some 8–O–4 (see Scheme 34). 8–O–4 Coupling becomes more important for sinapyl alcohol. These predictions are borne out by the couplings observed with in vitro dehydrogenation polymers produced from the precursors. The oxidizing abilities of a fungal laccase and horse radish peroxidase have been compared.237 The laccase enzyme oxidized sinapic acid better than ferulic or 4-coumaric acids, and sinapyl alcohol better than coniferyl alcohol. Horse radish peroxidase oxidized 4-coumaric acid better than ferulic or sinapic acids, and coniferyl alcohol better than sinapyl alcohol.It was also demonstrated that with both enzymes, ferulic acid was predominantly oxidized from a mixture of 4-coumaric acid and ferulic acid, sinapic acid from a mixture of 4-coumaric acid and sinapic acid or from a mixture of ferulic acid and sinapic acid, and sinapyl alcohol from a mixture of sinapyl alcohol and coniferyl alcohol. To explain these observations it was suggested that the 4-hydroxyphenyl radical can oxidize guaiacyl and syringyl groups to produce their respective radicals, and that the guaiacyl radical can oxidize the syringyl group.These proposals are supported by results using peroxidase isozymes from Vigna angularis.238 This peroxidase oxidized sinapic acid and sinapyl alcohol only very slowly, but rapidly transformed compounds with a 4-hydroxyphenyl or guaiacyl group. However, the oxidation of sinapyl alcohol could be greatly enhanced by adding 4-coumaric acid or ferulic acid.This indicates that there is indirect oxidation of sinapyl alcohol by the radicals generated from 4-coumaric acid or ferulic acid. Reviews on lignin biosynthesis have been published,239–241 including discussion of molecular biology240 and mechanisms of control.241 5.6 Lignans Lignans are essentially cinnamyl alcohol dimers (dilignols), though further cyclization and other modifications create a wide range of structural types.The coupling process is an example of phenolic oxidative coupling, analogous to that proposed for the biosynthesis of lignins. However, this coupling is not the result of H2O2-requiring peroxidases, and, in contrast to lignin biosynthesis, stereochemically controlled couplings operate leading to enantiomerically pure products. Oxidizing enzymes (laccases and peroxidases) from Forsythia intermedia convert coniferyl alcohol 102 into (&)-pinoresinol together with racemic lignans (&)-dehydrodiconiferyl alcohol and (&)-guaiacylglycerol-‚-O-coniferyl ether.242 However, in the presence of an additional 78 kDa protein, the only product is (+)-pinoresinol 103.This protein may thus function as a switching mechanism between lignin and lignan pathways. In the lignan pathway (Scheme 35), (+)-pinoresinol is successively reduced by NADPH-dependent reductase activities to (+)-lariciresinol 104 and (")-secoisolariciresinol 105. Two isoforms of a (+)-pinoresinol/(+)lariciresinol reductase have been purified from F.intermedia, both of which catalyze the sequential stereospecific reductions and have comparable physical properties.243 The stereospecificity was confirmed by incubating a preparation containing both isoforms with (&)-pinoresinol and [(4R)-2H]NADPH and observing the 100 Coniferin R = H 101 Syringin R = OMe OH OGlc MeO R Coupling sites: 2 3 5 4 6 7 9 8 OH OH OMe Scheme 34 Dewick: The biosynthesis of shikimate metabolites 35formation of (+)-[(7*R)-2H]lariciresinol (Scheme 36).Similarly, (+)-lariciresinol was transformed into (")- secoisolariciresinol. The gene encoding this enzyme activity has been cloned and expressed in E. coli as a functional ‚-galactosidase fusion protein. The deduced amino acid sequence for the protein revealed a strong homology (41–44% identity, 61–64% similarity) with the enzyme isoflavone reductase (see Section 6.6). Cell-free extracts obtained from Arctium lappa catalyze the enantioselective formation (20% enantiomeric excess) of (+)-secoisolariciresinol from coniferyl alcohol in the presence of NADPH and H2O2.244 Cell cultures of Pinus taeda under appropriate conditions can be induced to undergo cell wall lignification and production of an extracellular lignin precipitate.245 Addition of KI, an H2O2 scavenger, inhibits lignification, but results in the build-up of lignans derived from coniferyl alcohol, probably as part of the lignin assembly process.These include dehydrodiconiferyl alcohol 107 and dihydrodehydrodiconiferyl alcohol 108, guaiacylglycerol-‚-O-coniferyl ether 106, pinoresinol 103, shonanin 109 and isoshonanin 110. Some of these are probably related biosynthetically as shown in Scheme 37, but cannot be O O H Ar H H Ar H O H H Ar H D Ar H HO H H Ar D D Ar H HO H OH N CONH2 R H D N CONH2 R H D 7¢ 7 103 (+)-Pinoresinol 104 (+)-Lariciresinol 105 (–)-Secoisolariciresinol Scheme 36 OH OH MeO OH O MeO O O H OH OMe HO MeO H H O OMe O MeO H HO OH O H OH OMe HO MeO H OH H OH OMe HO MeO H OH HO 102 Coniferyl alcohol x 2 103 (+)-Pinoresinol 104 (+)-Lariciresinol O 105 (–)-Secoisolariciresinol NADPH NADPH Scheme 35 O OH OH OMe HO MeO O OH OH OMe HO MeO O O H OH OMe HO MeO H O MeO HO OH OMe O MeO HO OH OMe OH OMe OH OH OMe HO 107 Dehydrodiconiferyl alcohol 103 (+)-Pinoresinol 108 Dihydrodehydrodiconiferyl alcohol 106 Guaiacylglycerol-b- O-coniferyl ether 110 Isoshonanin 109 Shonanin Scheme 37 36 Natural Product Reports, 1998explained on the basis of simply phenolic oxidative coupling, requiring other processes, e.g.reduction. The biosynthesis of lignans is covered in a recent review.246 5.7 Tropic acid The biosynthetic origin of tropic acid 111 by a rearrangement process from phenylalanine has been appreciated for many years, but the mechanism of the rearrangement and the nature of intermediates still challenge investigators.From earlier labelling studies, an intramolecular migration of the carboxy group with retention of configuration at the benzylic centre of phenylalanine and a back migration of the 3-pro-S hydrogen to C-2 [Scheme 38(a)] has been the basis for speculation and experimentation. Phenyllactic acid 114 is now well established as an intermediate between phenylalanine and tropic acid, though the rearrangement takes place after esterification to tropine, the component heterocycle of the tropane alkaloids.However, the use of phenyllactic acid has now allowed the stereochemistry associated with the rearrangement to be reassessed. Incubation of (R)-D-3-phenyl[2-13C,2H]lactic acid with transformed root cultures of Datura stramonium gave hyoscyamine 113 with retention of the 13C–2H bonding.247 2H-Label was lost from a similar feeding of the enantiomeric substrate. This establishes that the R enantiomer of phenyllactic acid in littorine 112 is the one processed, and rules out phenylpyruvate as a precursor, since this would result in loss of 2H.A similar conclusion was reached based on feeding of the separate enantiomers of phenyllactic acid to the roots of Datura stramonium plants, though in the aerial parts, substantial incorporations of the S enantiomer were recorded.248 In further experiments,249 (RS)-DL-3-phenyl[2-3H]lactic acid was used as precursor, giving hyoscyamine with 3H located at C-3 of tropic acid. The chirality at the 3-hydroxymethyl was then established as S by reductive conversion to a chiral methyl and subsequent oxidation to acetate, indicating that any back migration of hydrogen from the benzylic position must involve inversion. However, by using several diVerent chirally labelled (R)-phenyllactate precursors with 13C at C-2 and 2H at C-2 or C-3, it was firmly demonstrated that the true fate of the C-2 and C-3 protons is as shown in Scheme 38(b).250 Thus, the 3-pro-S hydrogen is retained during the rearrangement whilst the 3-pro-R hydrogen is lost.The rearrangement proceeds with stereochemical inversion at the migration terminus and there is no migration of hydrogen from C-3 to C-2. A plausible mechanism for the rearrangement of the phenyllactate moiety in littorine 112 to the tropate residue in hyoscyamine 113 has been proposed as a result of feeding experiments with labelled phenyllactic acid in transformed root cultures of Datura stramonium and Brugmansia candida x B.aurea.251 (RS)-3-Phenyl[1,3-13C2]lactic acid was incorporated into hyoscyamine with the expected retention of doublelabelling as measured by MS, the incorporation being una Vected by concomitant feedings of tropic acid. In addition, the tropane alkaloids 3·-(2*-phenylacetoxy)tropane 119 and 3·-(2*-hydroxyacetoxy)tropane 118 were also labelled to a similar extent, 119 being double-labelled and 118 being singlelabelled (Scheme 39). This indicated that the phenylacetic acid group of 119 cannot be derived from free phenyllactic acid since decarboxylation would give a single-labelled species, but probably arises by loss of the hydroxymethyl from tropic acid.In confirmation, 3-phenyl[2-13C, 2-2H]lactic acid did not label 3·-(2*-phenylacetoxy)tropane but did produce labelled hyoscyamine and littorine. Similarly, it is suggested that the 2-hydroxyacetate residue is also produced from phenyllactate, by cleavage of the phenylmethyl group.Littorine labelled with 2H in the tropane methyl group and 1,3-13C2 in the phenyllactate system gave the predicted M+5 molecular ions for hyoscyamine and 119, and M+4 for 118. A mechanism for tropic acid biosynthesis allowing formation of these alternative acid residues is shown in Scheme 39. This begins with abstraction of a C-3 hydrogen to generate a free radical 115, and rearrangement occurring through the cyclopropane alkoxyl radical 116, via cleavage of the C-1–C-2 bond.The alternative acid groups are suggested to arise from radicals 115 and 117 by appropriate cleavage processes as indicated. 5.8 Phenylpropenes The overall pathway to epoxyisoeugenol 2-methylbutyrate (EPB) 125 in Pimpinella anisum has been delineated via feeding CO2H HS H HR NH2 H OH H HO2C CO2H HR H HS NH2 CO2R HR HO HS H H CO2R OH H H N Me O 28 L-Phenylalanine · · 111 Tropic acid ( a) ( b) R = tropyl: * · · ° * 2 3 2 3 112 Littorine 113 Hyoscyamine Scheme 38 CO2H HO H 114 ( R)-Phenyllactic acid 2 3 OH R O OH R O O OR OH OH O OR OH O OR O OR OH R O N Me O H H 1¢ 2¢ 3¢ 1¢ 3¢ Cleavage C-2¢–C-3¢ Cleavage C-2¢–C-3¢ 2¢ R = tropyl: 118 119 112 Littorine 113 Hyoscyamine 117 115 116 = 13C • • • • •• • • • • • • • • Scheme 39 Dewick: The biosynthesis of shikimate metabolites 37experiments (Scheme 40).Although superficially a phenylpropanoid, EPB contains the rare 2,5-dioxygenation and has been shown to arise from anethole 123 through pseudoisoeugenol 124 via a migration of the side-chain, initiated by a hydroxylation-induced NIH shift.On the basis of incorporation data, (4-methoxy)cinnamyl alcohol 122 rather than anol 121 was considered an intermediate. Recent experiments have reversed this conclusion, and anol is now implicated as an obligatory intermediate.252, 253 Although both 121 and 122 are incorporated into EPB in callus cultures of P. anisum, the incorporation of anol 121 is severely depressed by its rapid modification by peroxidases.When fed along with sodium azide, an inhibitor of peroxidases, incorporations were dramatically increased, and anol was substantially better than (4-methoxy)cinnamyl alcohol as a precursor. Furthermore, an anol O-methyltransferase activity was detected in a callus homogenate; this also methylated allyl phenols eugenol and chavicol, but not 4-coumaryl alcohol 120, coniferyl alcohol or hydroxycinnamic acids. There was also evidence for the presence of a 4-coumaryl alcohol reductase activity in the extracts.In further studies, the enzymes PAL (see Section 5.2) and cinnamic acid 4-hydroxylase (see Section 5.3) were also shown to be active in the cell-free extracts.254 Addition of the PAL inhibitor L-2-aminooxy-3-phenylpropionic acid inhibited incorporation of phenylalanine into EPB. 6 Flavonoids 6.1 Chalcones The condensation of 4-coumaroyl CoA 126 with three molecules of malonyl CoA to produce the chalcone naringeninchalcone 127 is catalyzed by the enzyme chalcone synthase [naringenin-chalcone synthase; malonyl CoA:4-coumaroyl CoA malonyltransferase (cyclizing)] (Scheme 41).Flavonoids containing a resorcinol substitution pattern in the acetate/ malonate-derived ring, rather than the phloroglucinol pattern as in naringenin-chalcone, are derived from the chalcone isoliquiritigenin 129, which arises as a result of a reductase acting concomitantly with chalcone synthase (Scheme 41) (see also Section 7).Multiple genes encoding chalcone synthase have now been reported, and sequenced, from several plant species. These include subterraneum clover (Trifolium subterraneum) (at least nine copies, of which six have been sequenced),255 Pueraria lobata (four genes)256 and soybean (Glycine max) (at least seven genes in gene clusters).257 Two members of a gene family (at least six members) from potato (Solanum tuberosum) have been sequenced,258, 259 as have thirteen genes from seven species of Ipomoea.260 A single chalcone synthase gene in rice (Oryza sativa) has been reported.261, 262 There is little detailed knowledge available concerning flavonoid biosynthesis in the lower plants, and information has only recently been presented demonstrating chalcone synthase activity in a representative liverwort Marchantia polymorpha.263 This yielded naringenin when incubated with 4-coumaroyl CoA and malonyl CoA, and the enzyme crossreacted with antibodies raised against chalcone synthase derived from three higher plants (rye, parsley and spinach) suggesting a close similarity in the proteins. Chalcone reductase is also encoded by multiple genes in alfalfa (Medicago sativa), and seven clones have been isolated from a cDNA library isolated from alfalfa leaves infected with the fungus Pseudomonas syringae pv.pisi.264 Three of the genes have been sequenced.265 In parallel studies, sequence data for two chalcone reductase genes from alfalfa,266 and one gene from Glycyrrhiza echinata267 have been reported.Although stress conditions stimulate flavonoid biosynthesis in M. sativa, OH OH OH OMe OMe OH OH OMe O OMe O O Phe 122 121 Anol 123 Anethole 124 Pseudoisoeugenol 120 125 EPB Scheme 40 O CoAS OH O OH O O SCoA O O OH O SCoA O H OH OH OH O OH HO O OH O OH HO O OH O HO OH OH O HO 126 4-Coumaroyl-CoA Malonyl-CoA 127 Naringenin-chalcone 128 Naringenin 130 Liquiritigenin i 129 Isoliquiritigenin ii NADPH iii iii Scheme 41 Enzymes: i, chalcone synthase; ii, reductase, iii, chalcone isomerase 38 Natural Product Reports, 1998the expression of chalcone synthase and chalcone reductase activities appears to be controlled diVerentially according to the various conditions.264 Chalcone synthase-like genes active during corolla development in Gerbera hybrida have been identified, and two of these encode for typical chalcone synthase enzymes.268, 269 Their gene expression appears to correlate with the production of flavonol and anthocyanin pigments.However, a third gene is significantly diVerent, its expression being unrelated to pigment production, and its deduced amino acid sequence deviating from those of typical chalcone synthase or stilbene synthase proteins. Furthermore, the protein is unable to transform 4-coumaroyl CoA, instead utilizing eVectively only benzoyl CoA to give as yet unidentified products. Relationships between chalcone synthase and stilbene synthase are discussed further in Section 7. 6.2 Flavanones, flavonols, anthocyanidins and related flavonoids The cyclization of chalcones to (2S)-flavanones is catalyzed by the enzyme chalcone isomerase (chalcone-flavanone isomerase) (Scheme 41). A chalcone isomerase cDNA from Pueraria lobata has been cloned and overexpressed in E. coli.270 Sitespecific mutagenesis showed that Cys-119 was not an essential active site residue. Elicited cultures of old man cactus Cephalocereus senilis synthesize a group of flavonoids characterized by an unsubstituted B-ring, e.g.the flavone baicalein 131 and the aurone cephalocerone 132. Preliminary studies (see ref. 3) had indicated that chalcone synthase from these cultures was active with cinnamoyl CoA, and that chalcone isomerase accepted the product 2*,4*,6*-trihydroxychalcone. In recent studies,271 two isoforms of chalcone isomerase were identified, and these demonstrated activity towards both naringenin-chalcone and 2*,4*,6*-trihydroxychalcone.A single form of cinnamate CoA ligase was active with both cinnamate and 4-coumarate. Therefore, the cultures contain enzyme activities for the synthesis of both B-ring hydroxy- and deoxy-flavonoids, though elicitation results in synthesis of only B-ring deoxy compounds. It is suggested that metabolic compartmentation may allow bypassing of cinnamate-4-hydroxylase activity, and that cinnamoyl CoA is then utilized by the nondiscriminating chalcone synthase and chalcone isomerase activities.Cell-free extracts from cell suspension cultures of Cassia didymobotrya contain polyphenol oxidase enzymes capable of transforming 2-hydroxychalcones into aurones.272 Thus, chalcones 133 were converted into a mixture of (&)-flavanone, (E)- and (Z)-aurones 134, and auronol 135, the latter proven to occur by further oxidation of the aurone. A free radical coupling sequence (Scheme 42) is proposed to account for these products.The relationship of flavanones to dihydroflavonols, flavonols, flavan-3,4-diols (leucoanthocyanidins) and anthocyanidins is outlined in Scheme 43. Flavanone 3-hydroxylases are 2-oxoglutarate-dependent dioxygenases requiring the cofactors oxygen, Fe2+ and ascorbate. Genes encoding for flavanone 3-hydroxylase have been identified in a number of plants and sequence data presented. Recently reported sources include Arabidopsis thaliana,273, 274 Bromheadia finlaysoniana,275 maize (Zea mays),276 alfalfa (Medicago sativa)277 and carnation (Dianthus caryophyllus).278 Most of the deduced amino acid sequences share 80–90% identity.A microsomal flavonoid 3*-hydroxylase in sweet orange (Citrus sinensis) has been demonstrated to be a typical cytochrome P450 system.279 It hydroxylated naringenin 128 to eriodictyol 136, and also dihydrokaempferol 138 to dihydroquercetin 139, and kaempferol 141 to quercetin 142. Dihydroflavonol 4-reductase is an NADPH-dependent enzyme converting dihydroflavonols into flavan-3,4-diols (leucoanthocyanidins), precursors of anthocyanidins.Using a Petunia hybrida-derived probe, a full length cDNA clone from petals of Rosa hybrida has been isolated and sequenced.280 Sequence identity at the amino acid level with other reported dihydroflavonol 4-reductase genes was 59–67%. A Petunia cultivar, pale pink in colour due to a deficiency in flavonoid 3*-hydroxylase and flavonoid 3*,5*-hydroxylase, was transformed with the Rosa dihydroflavonol 4-reductase DNA to give transgenic plants salmon-pink in colour and containing pelargonidin 146, an anthocyanidin rarely found in Petunia.280 The insertion of the Rosa gene enabled the conversion of dihydrokaempferol 138 into leucopelargonidin 143 to be achieved, a conversion not catalyzed by the Petunia dihydro- flavonol 4-reductase. A gene encoding dihydroflavonol 4-reductase from Gentiana triflora has also been characterized, along with those for flavonoid 3*,5*-hydroxylase and flavonoid 3-glucosyltransferase.281 Sequence identities with counterparts from other plants were 57–71, 62–77 and 31–52%, respectively.The latter two genes were expressed in E. coli to give active enzymes. Flavonoid 3*,5*-hydroxylase was eVective in hydroxylating naringenin 128 (to 136), eridictyol 136 (to 137), dihydrokaempferol 138 (to 139), dihydroquercetin 139 (to 140) and the flavone apigenin 149. The anthocyanidins pelargonidin 146, cyanidin 147, and delphinidin 148 were the preferred O O HO OH HO HO O O O O 131 Baicalein 132 Cephalocerone OH R O HO OH OH R O HO O OH R O HO O OH R O HO O O R O HO O HO O O O H HO O O OH R R HO O O OH R OH 133 R = H, OMe 134 R = H, OMe – H (±)-Flavanone – H 135 R = H, OMe + H O Scheme 42 Dewick: The biosynthesis of shikimate metabolites 39substrates for the glucosyltransferase.An anthocyanidin 3-Oglucosyltransferase from cell suspension cultures of grape (Vitis vinifera) had highest activity with cyanidin as acceptor, though delphinidin and pelargonidin were also excellent substrates. 282 Anthocyanin biosynthetic genes are expressed in all tissues of grape, except for that encoding flavonoid 3-Oglucosyltransferase which is expressed only in the fruit skin. White grapes lack anthocyanins because they are deficient in flavonoid 3-O-glucosyltransferase.283 Cell-free extracts of developing leaves of lemon (Citrus limon) contain a hesperetin 7-O-glucosyltransferase activity.284 As well as glucosylating the flavanone hesperetin 151, the enzyme also accepted naringenin 128, the flavones apigenin 149 and crysin 150, and the flavonols morin 152 and kaempferol 141 as substrates.Wild-type Petunia pollen accumulates high levels of flavonol 3-O-glycosides, and pollen germination appears to be dependent on these flavonoids. Conditionally male-fertile pollen contains no flavonols, and is unable to germinate, though this function can be restored if kaempferol or quercetin is added.Two glycosyltransferase activities have been isolated from Petunia pollen, and shown to be highly specific for flavonols.285 Both kaempferol and quercetin are O OH O OH HO O OH O OH HO OH O OH O OH HO OH O OH O OH HO OH OH O OH O OH HO OH OH OH O OH OH OH HO OH O OH OH OH HO OH OH O OH OH OH HO OH OH OH O OH OH HO OH O OH OH HO OH OH O OH OH HO OH OH OH O OH O OH HO OH O OH O OH HO OH OH O OH O OH HO OH OH 138 Dihydrokaempferol 136 Eriodictyol 128 Naringenin 139 Dihydroquercetin 140 Dihydromyricetin 144 Leucocyanidin 145 Leucomyricetin 146 Pelargonidin 147 Cyanidin 148 Delphinidin 141 Kaempferol 142 Quercetin i iii ii ii iv iv iv v vi vi vi iii 137 v ii + + + 143 Leucopelargonidin Scheme 43 Enzymes: i, flavonoid 3*-hydroxylase; ii, flavanone 3-hydroxylase; iii, flavonoid 3*,5*-hydroxylase; iv, dihydroflavonol 4-reductase; v, flavonol synthase; vi, leucoanthocyanidin oxidase/dehydratase O R O OH HO O OMe O OH HO OH O OH O OH HO OH OH 149 Apigenin, R = OH 150 Crysin, R = H 151 Hesperetin 152 Morin 40 Natural Product Reports, 1998transformed by the action of the enzymes into their respective 3-O-(2+-O-glucosyl)galactosides 153 and 154, identical to materials found in the wild-type pollen.The major anthocyanins found in an Afghan cultivar of carrot (Daucus carota) are cyanidin 3-O-xylosylglucosylgalactoside acylated with sinapic or ferulic acid.Two of the enzymic glycosylations involved in the biosynthesis of these compounds have been detected in protein preparations.286 A galactosyltransferase utilizes UDPgalactose to glycosylate the 3-hydroxy group of cyanidin 147, and a xylosyltransferase then transfers the xylosyl group from UDPxylose to give the 3-xylosylgalactoside 155 (Scheme 44). The galactosyltransferase activity was subsequently purified. Anthocyanin acyltransferases have been isolated from cell cultures of Ajuga reptans.287 Of two separable activities displaying rather broad specificities, one catalyzed the transfer of a hydroxycinnamic acid from its CoA ester to 3-O-glycosides and 3,5-di-Oglycosides of cyanidin and delphidin.The second activity transferred the malonyl group from malonyl CoA preferably to anthocyanidin 3-O-coumaroylglucoside-5-O-glucosides and anthocyanidin 3-O-glucosides. Compounds such as the acylated diglucoside 156 of cyanidin are typical of the Ajuga anthocyanins.The main anthocyanins can contain two hydroxycinnamoyl groups. Purple leaves of lettuce (Lactuca sativa) contain cyanidin 3-O-(6+-malonyl)glucoside and enzyme extracts are able to malonylate the 6-hydroxy group of the glucose moiety in cyanidin 3-O-glucoside. The enzyme showed broad specificity and also eYciently malonylated 3-O-glucosides of pelargonidin, delphinidin, peonidin 157, petunidin 158 and malvidin 159.288 Succinyl CoA, but not acetyl CoA, was able to substitute for malonyl CoA at lower eYciency.Green leaves contained about 20% of the enzyme activity found in the purple leaves. Succinyl esters of anthocyanins are found naturally in the flowers of cornflower (Centaurea cyanus). In blue flowers the main derivative is cyanidin 3-succinylglucoside-5-glucoside, whilst in pink flowers the corresponding pelargonidin derivative predominates. An enzyme preparation from blue flowers catalyzed succinyl CoAdependent succinylation of the 3-O-glucosides of cyanidin and pelargonidin, but not of the 3,5-di-O-glucosides.289 Similar results were obtained using extracts from pink flowers, and both enzymes would also malonylate the anthocyanidin 3-Oglucosides equally eVectively.It may be deduced that acylation of the anthocyanidin 3-O-glucosides precedes the 5-Oglucosylation step in forming the cornflower pigments, in keeping with similar observations on the biosynthesis of acetylated anthocyanidin diglucosides in other plants.The plant Scutellaria baicalensis contains a wide range of flavone glucuronides, and a flavone-specific glucuronidase enzyme has now been purified from callus cultures.290 The enzyme is a homotetramer and accepts as substrates luteolin 3-O-glucuronide 160, bacailin 161, wogonin 7-O-glucuronide 162 and oroxylin 7-O-glucuronide 163. O OH O OH HO O O O OH HO HO OH HO HO OH O R 153 R = H 154 R = OH O OH OH HO OH O OH HO OH O HO O OH OH HO O OH OH HO OH OH O OH OH HO OH O OH HO O O HO 147 Cyanidin UDPXyl 155 UDPGal + + + Scheme 44 O OH HO OH O HO HO OH O O O HO O HO HO OH O HO O O O 156 + O OH OH HO R OH OMe 157 Peonidin R = H 158 Petunidin R = OH 159 Malvidin R = OMe + Dewick: The biosynthesis of shikimate metabolites 41The flavanone sakuranetin 164 is synthesized as a phytoalexin on UV irradiation of rice (Oryza sativa).A methyltransferase catalyzing SAM-dependent methylation of naringenin has been detected in extracts of irradiated leaves.291 The flavone apigenin 149 and flavonol kaempferol 141 were also eYciently methylated, but surprisingly, the most eVective substrate was the flavone luteolin 165.An eriodictyol 136 4*-O-methyltransferase has been partially purified from Citrus aurantium.292 A gene encoding flavonol 3*,5*-O-methyltransferase in Chrysosplenium americanum has been isolated.293, 294 The deduced amino acid sequence for the encoded protein showed 67–85% similarity to other plant O-methyltransferase sequences. Recombinant protein was produced which showed strict specificity for positions 3* and 5* of partially methylated flavonols based on quercetin 142 and quercetagetin 166, but not quercetin itself, in keeping with the earlier enzymic studies and the pattern of flavonols found in Chrysosplenium.294 The genetics and biochemistry of anthocyanin biosynthesis have been reviewed.295 6.3 Flavonoid sulfates Sulfation of the flavonol quercetin in Flaveria chloraefolia is catalyzed by two sulfotransferases, giving firstly the 3-sulfate, and then the 3,4*-disulfate.These enzymes show strict substrate and position specificities, yet cDNA sequencing has indicated that the enzymes share 69% identity in their amino acid sequences. In recent studies,296 chimeric flavonol sulfotransferases have been constructed by reciprocal exchange of DNA fragments then expression in E. coli. The chimeric proteins were active, though at a reduced level, and a specific region in the sequence was deduced to be responsible for determining substrate and position specificity.Other regions were shown to be highly conserved in other plant and animal sulfotransferase amino acid sequences. Region IV in the plant 3-sulfotransferase has been demonstrated to be required for the binding of the cofactor PAPS (3*-phosphoadenosine 5*-phosphosulfate) with Arg-276 being a critical residue.297 Region I was not required for PAPS binding, though Lys-59 was critical for catalysis and may help to stabilize an intermediate.A flavonol sulfotransferase-like cDNA clone in Flaveria bidentis has been identified and sequenced.298 6.4 Carthamin Carthamin 169 is a red pigment from the saZower (Carthamus tinctorius) and has a bichalcone-type structure. A likely biosynthetic precursor 168 which is converted into carthamin by oxidative decarboxylation on exposure to air or oxidizing agents has been isolated, and was reported earlier (see ref. 1). The same conclusions are reached by an independent research group who have also isolated the unstable yellow compound 168.299 In addition, this group has reported the isolation of a further possible intermediate in carthamin biosynthesis, namely anhydrosaZor yellow B 167.300 This dark orange compound is transformed into carthamin by allowing a solution in methanol–pyridine (NMR solvent) to stand for about a month. Of particular interest is the additional glucose moiety in 167, which may ultimately supply the extra carbon atom linking the two chalcone fragments in the carthamin structure (Scheme 45).A precarthamin material which can be converted enzymically into carthamin is released from a bound form of precarthamin by treatment of powdered florets with enzymes such as ‚-glucosidase, cellulase or pectinase.301 6.5 Diels–Alder-type adducts Flavonoid-based molecules that appear to be derived by Diels–Alder-type reactions between a chalcone dienophile and a flavonoid with a modified prenyl group as the diene found in plants of the Moraceae are reviewed in a recent article.302 From the high incorporations of label and the labelling O O GlcAO R2 OH R1 O OH HO OH OGlcA 161 Bacailin R1 = H; R2 = OH 162 R1 = H; R2 = OMe 163 R1 = OMe; R2 = H 160 O O O OH MeO OH O O OH HO OH OH O O OH HO OH OH 164 Sakuranetin OH 165 Luteolin HO 166 Quercetagetin O HO Glc HO O O OH O O HO O O HO OH Glc Glc HO OH OH HO CO2H OH O O HO O O HO OH Glc Glc HO OH OH O OH O O Glc HO OH OH H H HO OH H H OH CH2OH OH 167 Anhydrosafflor yellow B 168 169 Carthamin Scheme 45 42 Natural Product Reports, 1998patterns obtained, it has been suggested that both phenylalanine and tyrosine can be utilized in supplying the C6C3 portions of chalcomoracin 170 and kuwanon J 171 in Morus alba cell cultures (see ref. 1). Further evidence has now been presented to indicate that two independent pathways to 4-coumaroyl CoA exist in this system.303 Simultaneous administration of [1-13C]phenylalanine and [3-13C]tyrosine gave chalcomoracin with incorporation of label into all of the appropriate positions.Furthermore, [2-13C]cinnamic acid as its N-acetylcysteamine thioester was also specifically incorporated. Thus, the pathways phenylalanine]cinnamic acid]4- coumaric acid and tyrosine]4-coumaric acid are both shown to be operative. 6.6 Isoflavonoids Isoflavonoids are structurally modified flavonoids in which the shikimate-derived aromatic ring has migrated to the adjacent carbon atom of the heterocycle.Rearrangement of flavanone precursors such as liquiritigenin 130 and naringenin 128 is brought about by the actions of isoflavone synthase, a cytochrome P450-dependent monooxygenase, and then a dehydratase, and yields initially the corresponding isoflavone. These can then be transformed into a range of other iso- flavonoid variants. The pterocarpan medicarpin 177 is derived in alfalfa (Medicago sativa) from reduction of 2*-hydroxyformononetin 174 to the isoflavanone vestitone 175, followed by a further NADPH-dependent reduction to the isoflavanol 176.Subsequent cyclization is achieved via the enzyme 7,2*- dihydroxy-4*-methoxyisoflavanol (DMI) dehydratase (Scheme 46). The penultimate enzyme in this sequence, vestitone reductase, has now been produced by cDNA cloning and expression in E. coli.304 The enzyme had strict substrate specificity for (3R)-vestitone and not (3S)-vestitone as also observed for the enzyme isolated from alfalfa.The pterocarpan maackiain 180 contains a methylenedioxy bridge which is formed early in the biosynthetic sequence at the isoflavone oxidation level. A microsomal preparation from elicited cell cultures of chickpea (Cicer arietinum) has been obtained which transforms the isoflavone calycosin 178 into pseudobaptigenin 179 (Scheme 46).305 This enzyme required O2 and NADPH cofactors and was characterized as a cytochrome P450- dependent system.It would also catalyze methylenedioxy bridge formation in pratensein (5-hydroxycalycosin) giving the corresponding 5-hydroxypseudobaptigenin. A cDNA clone encoding an isoflavone reductase-like protein has been characterized in white lupin (Lupinus albus).306 The pea (Pisum sativum) phytoalexin (+)-pisatin 185 has the opposite absolute configuration to (")-(6aR,11aR)-maackiain 180 from chickpea, and is produced by 6a-hydroxylation then O-methylation of (+)-(6aS,11aS)-maackiain 184 (Scheme 47).However, the reduction product of pea isoflavone reductase is (")-(3R)-sophorol 182, stereochemically equivalent to (")- (6aR,11aR)-maackiain 180. This implicates a change of stereochemistry at a late stage in the pathway to pisatin (see ref. 1). In investigating this, it has been found that isoflavone 181 and (")-sophorol 182 were more eYciently incorporated into pisatin than (+)-sophorol 183 or (+)-maackiain 184.307 This O OH OH HO HO O HO OH OH O OH OH HO HO HO OH O OH OH 170 Chalcomoracin 171 Kuwanon J O O HO OH O O HO OMe O O HO OMe OH O O HO O O O O HO O O H H O O HO OMe HO O O HO OMe HO H O HO HO OMe HO H O O HO H H OMe 172 Daidzein 180 Maackiain 179 Pseudobaptigenin 173 Formononetin 130 Liquiritigenin 7 4¢ 174 178 Calycosin 175 (3 R)-Vestitone 176 v NADPH i iii 177 Medicarpin vii NADPH iv vi O OH O OH HO ii O2, NADPH O2, NADPH O2, NADPH O2, NADPH Scheme 46 Enzymes: i, isoflavone synthase; ii, dehydratase; iii, iso- flavone reductase; iv, vestitone reductase; v, DMI dehydratase; vi, pseudobaptigenin synthase Dewick: The biosynthesis of shikimate metabolites 43suggests that isomerization at the isoflavanone stage, the most obvious timing, is not significant. Detection of the corresponding isoflav-3-ene 186 in extracts leads to a suggestion that this may be a non-chiral intermediate in the conversion of (")- (3R)-sophorol 182 into (+)-pisatin 185.A discrepancy in the detailed understanding of isoflavone biosynthesis relates to the formation of formononetin 173 from daidzein 172 (Scheme 46).The isoflavone O-methyltransferase activity found in yeast elicitor-treated cell suspension cultures of Medicago sativa preferentially methylates the 7-hydroxy group of daidzein in vitro, whereas in vivo, stress-related isoflavonoid biosynthesis requires and involves 4*-Omethylation to formononetin. To investigate this anomaly further, the enzyme was purified by a substrate-based aYnity system, and some internal peptide sequences were determined. 308 Sequence similarities to regions in catechol O-methyltransferase from barley, and 6a-hydroxymaackiain 3-O-methyltransferase from pea were noted, though there was little similarity to other O-methyltransferase enzymes in alfalfa acting on caVeic acid or isoliquiritigenin. The purified iso- flavone O-methyltransferase had substrate specifity towards isoflavones with a free 7-hydroxy group, converting daidzein into isoformononetin (7-O-methyldaidzein) and genistein 187 into prunetin 188, but could also methylate the 5-hydroxy group of genistein giving a compound tentatively identi- fied as 4*-hydroxy-5,7-dimethoxyisoflavone. Isoflavones with a 4*-methoxy group, e.g.formononetin and biochanin A, were not susceptible to 7-O-methylation. By far the best substrate however was 6,7,4*-trihydroxyisoflavone, which has no role in the formation of alfalfa stress-induced isoflavonoids.Despite the in vitro characteristics of this enzyme, it is still believed that its in vivo role is to catalyze 4*-O-methylation of daidzein to produce formononetin. The isoflavonoid phytoalexin pathway has been reviewed with emphasis on enzymes, genes, and transcription factors.309 7 Stilbenes and dihydrophenanthrenes Stilbenes are produced from the same cinnamoyl CoA/malonyl CoA substrates as are chalcones, but the enzyme-bound polyketide is folded diVerently and a decarboxylation occurs during cyclization, leading to resveratrol 190 rather than naringenin-chalcone 127 for example (Scheme 48).Accordingly, it has been found that stilbene synthases and chalcone synthases are closely related homodimeric enzymes with high amino acid sequence identity. Using stilbene synthase from peanut (Arachis hypogaea) and chalcone synthase from white mustard (Sinapis alba), protein cross-linking and site-directed mutagenesis studies have identified a subunit contact site close to the active site Cys-169.310 By producing heterodimers in which one monomer was inactive, it was demonstrated that in both stilbene synthase and chalcone synthase, each subunit acts independently and is able to synthesize the end product.The active sites of the two subunits are close together in the dimer. In the presence of a monomeric reductase (see Section 6.1), chalcone synthase also synthesizes a 6*-deoxychalcone, with the 6*-hydroxychalcone as a second product when using plant preparations.Both products were also synthesized by E. coli expressing the individual proteins, and both were also obtained with a chalcone synthase heterodimer containing a single active site. This indicates that 6*-deoxychalcone synthesis requires no other plant factor apart from the reductase, O O HO O O HO O O HO O O HO H O O HO O O HO H O O HO O O H H O O HO O O H H O O MeO O O H OH 182 (–)-(3 R)-Sophorol 181 183 (+)-(3 S)-Sophorol 180 (–)-(6a R,11a R)-Maackiain 6a 11a 184 (+)-(6a S,11a S)-Maackiain 185 (+)-(6a R,11a R)-Pisatin i ? NADPH ? Scheme 47 Enzyme: i, isoflavone reductase O HO O O HO O O R OH OH 186 187 Genistein, R = OH 188 Prunetin, R = OMe 7 4¢ 5 R CoAS O R O O COSCoA O R O O SCoA O O R O OH HO OH R HO OH 126 R = OH 189 R = H º 127 Naringenin-chalcone, R = OH 190 Resveratrol, R = OH 191 Pinosylvin, R = H 3 x Malonyl-CoA CO2 i ii i, ii Scheme 48 Enzymes: i, stilbene synthase; ii, chalcone synthase 44 Natural Product Reports, 1998and that formation of the two products is probably an intrinsic property of the interaction between dimeric chalcone synthase and the monomeric reductase.cDNAs representing two closely related stilbene synthase genes have been isolated from Pinus strobus using a probe derived from P. sylvestris.311 The encoded proteins showed only five amino acid diVerences. The proteins expressed in E. coli both preferred cinnamoyl CoA 189 to dihydrocinnamoyl- CoA, though P.strobus accumulates pinosylvin 191 and dihydropinosylvin 192. Otherwise, the proteins demonstrated quite significant variations in properties. In particular, one enzyme (STS1) had only 3–5% of the activity shown by the other (STS2), and with cinnamoyl CoA yielded a second, unknown product. Site-directed mutagenesis showed that a single arginine to histidine exchange in STS1 was responsible for all the observed diVerences. Bibenzyls are intermediates on the way to 9,10- dihydrophenanthrenes, phytoalexins of several orchid species.Bibenzyl synthase is an enzyme closely related to stilbene synthase, though it utilizes CoA esters of a dihydrocinnamic acid as substrate rather than the ester of a cinnamic acid (Scheme 49). By exploiting the homology between bibenzyl synthase and stilbene synthases, two full-length cDNA clones were isolated from elicitor-induced plants of Phalaenopsis species, and subsequently expressed in E.coli.312 Both proteins were catalytically active, with virtually the same selectivity towards dihydro-3-coumaroyl CoA 193 as had the enzyme isolated from plant tissues. Thus, this substrate was preferred over phenylpropionyl CoA, and cinnamoyl CoA and 4-coumaroyl CoA were rather poor substrates. A cDNA clone for (S)-adenosylhomocysteine hydrolase was also isolated in these studies; this enzyme is involved in the supply of SAM utilized in the methylation of the bibenzyl 194 to give batatasin III 195 (Scheme 49).Oxidative coupling then leads to the dihydrophenanthrene hircinol 196. 8 Xanthones The xanthone skeleton is derived from a benzoyl CoA and three malonyl CoA units via a benzophenone. A benzophenone synthase activity has been demonstrated in cell-free extracts obtained from cultures of Centaurium erythraea that accumulate 1,3,5-trihydroxyxanthone 199.313 The extracts catalyzed the synthesis of 2*,3,4,6-benzophenone 198 from 3-hydroxybenzoyl CoA 197 and malonyl CoA (Scheme 50). 9 Quinones 9.1 Naphthoquinones A range of naphthoquinone derivatives, including phylloquinone (vitamin K1) and menaquinone (vitamin K2), are produced from chorismate via isochorismate 49, 2-succinyl- 6-hydroxycyclohexa-2,4-diene-1-carboxylate (SHCHC) 200, o-succinylbenzoate (OSB) 201, and 1,4-dihydroxynaphthoate 203 (Scheme 51). The enzyme o-succinylbenzoyl CoA synthetase which activates OSB by conversion into its ‘aliphatic’ CoA ester 202 in E.coli has been overexpressed and puri- fied.314 The enzyme is a homotetramer, requires ATP and coenzyme A, and its activity is stimulated by Mg2+ ions. Neither benzoyl CoA nor benzoylpropionic acid (succinylbenzene) acts as a substrate for the purified enzyme stressing the requirement for both the benzoyl carboxy group and the succinyl side-chain. 9.2 Ubiquinones Ubiquinones (coenzyme Q) 209 are also derived from chorismate, but via 4-hydroxybenzoate 79, and an isoprenoid sidechain of length varying according to the organism is attached to this substrate early in the biosynthetic sequence.The generally accepted pathway is shown in Scheme 52 though not all intermediates are fully proven. 3-Hexaprenyl-4- hydroxybenzoic acid 204 (n=6) has been isolated as the predominant labelled lipid material from cultures of yeast (Saccharomyces cerevisiae) supplied with 14C-labelled 4-hydroxybenzoic acid and grown under log-phase conditions. 315 In stationary phase cultures, labelled 204 was also produced, but the major labelled product was coenzyme Q 209 (n=6).A yeast mutant (coq7-1) deficient in ubiquinone synthesis was found to synthesize the intermediate demethoxyubiquinone 207 (n=6).316 The corresponding wild-type coq7 gene was isolated and sequenced, and shown to restore growth and ubiquinone biosynthesis to the mutant. Several coq7 HO OH 192 Dihydropinosylvin CoAS O HO HO HO OH HO MeO OH HO MeO OH 3 x Malonyl-CoA 193 i SAM Homocysteine 194 195 Batatasin III S-Adenosylhomocysteine 196 Hircinol ii Scheme 49 Enzymes: i, bibenzyl synthase; ii, (S)-adenosylhomocysteine hydrolase OH CoAS O OH HO OH O OH O HO OH O OH 3 x Malonyl-CoA 199 197 198 Scheme 50 Dewick: The biosynthesis of shikimate metabolites 45deletion mutants then generated were shown to accumulate 3-hexaprenyl-4-hydroxybenzoic acid 204 rather than 207.The coq7 gene may thus encode a protein involved in one or more monooxygenase/hydroxylase steps in the biosynthetic sequence.In eukaryotes, the first O-methylation step is carried out by the Coq3 protein on 3,4-dihydroxy-5-polyprenylbenzoate 205, whilst in E. coli, the predicted substrate is 2-octaprenyl-6-hydroxyphenol 206 (n=8), with the enzyme yet to be identified. The second O-methylation, conversion of demethylubiquinone 208 (n=8) into ubiquinone, involves UbiG, an enzyme which is 40% identical in amino acid sequence to yeast Coq3. It has now been demonstrated that UbiG probably catalyzes both O-methylation steps in E.coli.317 The ubiG gene was able to restore respiration and ubiquinone synthesis in yeast coq3 mutants, provided a mitochondrial leader sequence was also inserted to target the activity to the mitochondria. In vitro assays showed the UbiG enzyme was able to methylate synthetic analogues such as the ‘eukaryotic’ 3,4-dihydroxy-5-farnesylbenzoate 205 (n=3) and the ‘prokaryotic’ 2-farnesyl-6-hydroxyphenol 206 (n=3).Methylation of 3,4-dihydroxy-5-farnesylbenzoate 205 (n=3) was also accomplished using mitochondria from yeast containing the coq3 gene, showing that the chain length is not critical for substrate recognition.318 The ubiquinone biosynthetic pathway in the protozoon Leishmania major probably shows similar characteristics to those in mammalian and bacterial systems. Thus in preliminary studies,319 4-hydroxybenzoic acid, acetate and mevalonate were established as precursors of ubiquinone-9 209 (n=9). 9.3 Shikonin Shikonin 211 is an example of another group of naphthoquinones, which have their origins in 4-hydroxybenzoate rather than OSB. The remainder of the skeleton of shikonin is derived from geranyl diphosphate (Scheme 53). Deoxyshikonin 210 has recently been established as a late intermediate in the pathway.320 Labelled phenylalanine was rapidly incorporated into deoxyshikonin in cell cultures of Lithospermum erythrorhizon and then into fatty acid esters of shikonin, e.g.acetylshikonin 212 and ‚-hydroxyisovalerylshikonin 213. Deoxyshikonin was also incorporated into the esters of shikonin, though shikonin itself was only poorly utilized. Similar results were obtained using a cell-free system from L. erythrorhizon. A sequence of hydroxylation and esterification via shikonin is postulated, the poor utilization of shikonin being accounted for if a membrane-bound multienzyme system is operative. The biosynthesis of acetylshikonin, 4-hydroxybenzoic acid O-glucoside 80 and rosmarinic acid in L.erythrorhizon cultures is completely suppressed if 2-amino-2-phosphonic acid, an inhibitor of PAL, is added.321 Use of mevinolin, an inhibitor of HMGCoA reductase, specifically blocked acetylshikonin biosynthesis, thereby increasing the formation of 4- hydroxybenzoic acid O-glucoside (see also Section 4.2). Shikonin biosynthesis has been reviewed.322 CO2H OH O HO2C CO2H OH O CO2H O CO2H CO2H O COSCoA CO2H OH OH CO2H 202 (49) Isochorismate Chorismate 200 SHCHC 201 OSB 203 Vitamin K Anthraquinones 2-Oxoglutarate TPP HSCoA ATP i ii, iii iv v vi Scheme 51 Enzymes: i, isochorismate synthase (entC); ii, 2- oxoglutarate decarboxylase; iii, SHCHC synthase [ii and iii/menD]; iv, OSB synthase (menC); v, OSB:coenzyme A ligase (menE); vi, naphthoate synthase (menB) CO2H OH CO2H OH R CO2H OH R HO CO2H OH R MeO OH R OH R HO OH R MeO O R MeO O O R MeO O Me O R MeO O Me HO O MeO O Me MeO H H OPP Chorismate i 79 204 205 206 207 208 n 209 Ubiquinone n ii Eukaryotes Prokaryotes iii iv v vi vii vi Saccharomyces cerevisiae: n = 6 E.coli: n = 8 Man: n = 10 viii ix Scheme 52 Enzymes: i, chorismate lyase (ubiC); ii, 4-hydroxybenzoate polyprenyltransferase (UbiA); iii, Coq3; iv, UbiD; v, UbiB; vi, UbiG; vii, UbiH; viii, UbiE; ix, UbiF/Coq7 46 Natural Product Reports, 19989.4 Tocopherols Plant chloroplasts accumulate a range of polyisoprenoid quinones, quinols and chromanols, e.g.plastoquinones and tocopherols, that have their origins in homogentisic acid (see Section 4.4) and mevalonic acid. These products contain methyl substituents on the aromatic ring, one of which is produced by the decarboxylation of the homogentisate sidechain, the rest originating by C-methylation from SAM. An enzyme responsible for the cyclization step giving a chroman ring in the biosynthesis of ·-tocopherol 214 in the cyanobacterium Anabaena variabilis was recently reported and its reaction mechanism has been probed (Scheme 54) (see ref. 1). In further studies on this system,323 the use of a range of synthetic substrate analogues has established that three major recognition sites are necessary for the enzyme to exert its action. These are, the hydroxy group at C-1 of the hydroquinone, the E configuration of the double bond in the side-chain, and the length of the lipophilic side-chain (3 or 4 isoprene units). 10 Cyanogenic glycosides, glucosinolates and related compounds The biosynthetic pathways to cyanogenic glycosides and glucosinolates appear to share many features, though there are considerably more experimental data available for the former group.The basic pathway to cyanogenic glycosides as exemplified by dhurrin 221 is shown in Scheme 55. Tyrosine N-hydroxylase, an enzyme catalyzing the transformation of tyrosine into 4-hydroxyphenylacetaldoxime 218 and 219 in the biosynthesis of dhurrin in Sorghum bicolor, is shown to be a multifunctional cytochrome P450-dependent enzyme of a type not previously demonstrated in plant biosynthetic pathways.324 It is proposed that a double N-hydroxylation is achieved giving N-hydroxytyrosine 215 then N,Ndihydroxytyrosine 216, and that this is converted into the oxime via dehydration to the nitroso compound 217 and subsequent decarboxylation.The dehydration and decarboxylation appear to be non-enzymic. The E to Z ratio in the oxime product was 69:31.The multifunctional nature of the enzyme was demonstrated in that binding of either L-tyrosine or N-hydroxy-L-tyrosine mutually excludes binding of the other substrate. A cDNA clone encoding the enzyme has been obtained from S. bicolor, and sequence data presented.325 The sequence showed highest identity with that of the flavonoid 3*,5*-hydroxylase from Petunia hybrida (see Section 6.2) and a P450 protein of unknown function from avocado. Recombinant protein has been expressed in E.coli and purified.326 The enzyme showed high specificity in that it would also bind phenylalanine, but not transform this substrate. The hydrolysis of cyanogenic glycosides is brought about by the action of ‚-glucosidases liberating cyanohydrins, which are then decomposed into aldehyde and HCN by hydroxynitrile lyases (Scheme 56). cDNAs encoding for the ‚-glucosidases hydrolyzing dhurrin 221 in Sorghum bicolor (dhurrinase)327 and amygdalin 222 in Prunus serotina (amygdalin hydrolase) 328 have been cloned and sequenced.(R)-Mandelonitrile lyase from the fern Phlebodium aureum has been purified to homogeneity and shown to be a homomultimer existing in at least three isoforms.329 In contrast to other hydroxynitrile lyases, this enzyme is not inhibited by sulfhydryl- or hydroxymodifying agents and may therefore employ a diVerent catalytic mechanism. It is not a flavoprotein and therefore diVers from the (R)-mandelonitrile 223 lyases of the Rosaceae which CO2H OH CO2H OGlc CO2H OH OH OH O O OH OH O O OH OH OH 79 O O OH OH 210 Deoxyshikonin 80 212 R = Me 213 R = Me2C(OH)CH2 OCOR 211 Shikonin GeranylPP Scheme 53 HO OH O HO D i 214 a-Tocopherol D+ 1 Scheme 54 Enzyme: tocopherol cyclase CO2H NH2 CO2H HN OH CO2H N OH HO CO2H N O CO2H N+ O –O N OH N HO CN HO H CN GlcO H HO HO HO HO HO HO HO HO CN HO HO S 221 Dhurrin 219 ( Z)-Oxime O2 218 ( E)-Oxime 215 NADPH 31 L-Tyrosine 217 220 4-Hydroxymandelonitrile 216 O2 NADPH NADPH O2 Scheme 55 Dewick: The biosynthesis of shikimate metabolites 47are FAD-dependent. X-Ray crystallographic studies on (S)-4- hydroxymandelonitrile 220 lyase from Sorghum bicolor have been reported.330 Recent reviews discuss the biosynthesis of cyanogenic glycosides with emphasis on the P450 enzymes involved,331 and the functions and properties of hydroxynitrile lyases.332 The biosynthetic pathway to glucosinolates 226 is outlined in Scheme 57.Glucosinolate synthesis can be induced in seedlings of Sinapis alba by treatment with jasmonic acid, leading to an accumulation of 4-hydroxybenzylglucosinolate 227.333 An NADPH-dependent microsomal system from S.alba seedlings converting L-tyrosine into 4-hydroxyphenylacetaldoxime (224, compare 218, 219) has been characterized, demonstrating similarities with cyanogenic glycoside biosynthesis in the early part of the pathway. A similar enzyme activity producing phenylacetaldoxime from phenylalanine is present in jasmonic acid-treated seedlings of nasturtium (Tropaeolum majus).334 Both these systems diVer from enzymes in Brassica species which are not P450-dependent and appear to be flavin monooxygenases or peroxidase-type enzymes.Two distinct P450 independent monooxygenases in Brassica napus leaves had been shown to convert homophenylalanine 228 or dihomomethionine 229 into the corresponding aldoximes by NADPH- and O2-dependent oxidative decarboxylation. In further studies on these systems,335 the latter enzyme was shown to have no activity towards methionine, homomethionine, phenylalanine, tyrosine or tryptophan.Immature seeds of rape (Brassica napus) have been shown to contain enzymes capable of converting 35S-labelled desulfoglucosinolates 225, e.g. desulfoindol-3-ylmethylglucosinolate 230, into the corresponding glucosinolates.336 Glucosinolates undergo hydrolysis by the action of thioglucoside glucohydrolases (myrosinases), giving (initially) a thiohydroximate O-sulfonate 231 which is normally followed by a Lossen-type rearrangement to give isothiocyanates 232 (Scheme 58).Other types of product, e.g. nitriles and thiocyanates, may be formed, depending on substrate, pH, and other factors. Myrosinase from Sinapis alba is normally involved with the hydrolysis of sinigrin 233, but actually has better aYnity for glucotropaeolin 234, and the aromatic substrate often features in mechanistic studies of the enzyme.The stereochemistry of hydrolysis has been examined by NMR spectrometry, and shown to involve retention of configuration at the anomeric centre, suggesting a mechanism similar to related O-glucosidase hydrolysis [Scheme 59(a)].337 However, the ready inactivation of myrosinase with 2-deoxy-2- fluoroglucotropaeolin via accumulation of a long-life glucosylenzyme intermediate indicates that there is probably only one catalytic residue, a nucleophilic glutamate, whilst the acidic catalyst residue found in the corresponding O-glucosidase is missing [Scheme 59(b)].Using a variety of synthetic deoxyglucotropaeolins, the role of the individual hydroxy groups during catalysis has been probed.338 Loss of a hydroxy group from any position produced a strong decline in reaction rate, with the 2-hydroxy being especially important. It is suggested that the 2-hydroxy forms hydrogen bonds with the essential carboxylate (Glu or Asp) involved in the hydrolytic reaction. Myrosinases appear to be encoded by gene families according to species.Gene families MA and MB in Brassica napus contain about four and ten genes, respectively, and a third family MC containing three or four genes has now also been isolated.339 Three myrosinase genes in Arabidopsis thaliana show little similarity to the MA and MB gene families in the Brassicaceae and cannot be grouped with either family.340 Two isozymes of myrosinase from Raphanus sativus have been isolated and purified.341 The biosynthesis of glucosinolates in Brassicas,342 and the organization and biochemistry of the myrosinase– glucosinolate system,343 are the subjects of recent reviews.CN GlcGlcO H CN GlcO H CN HO H CHO 223 Mandelonitrile R + HCN 222 Amygdalin Prunasin i iii ii R Scheme 56 Enzymes: i, amygdalin hydrolase; ii, prunasin hydrolase; iii, mandelonitrile lyase R CO2 H NH2 R CO2 H HN OH R N OH R N OH SH R N OH SGlc R N OSO3 – SGlc O2 Thiohydroximic acid 224 225 Desulfoglucosinolate UDPGlc 226 Glucosinolate –SR¢ PAPS NADPH Scheme 57 CO2H NH2 MeS CO2H NH2 N OH SGlc NH N OSO3 – SGlc HO 228 Homophenylalanine 229 Dihomomethionine 227 4-Hydroxybenzylglucosinolate 230 R N OSO3 – SGlc R N OSO3 – SH R NCS 226 Glucosinolate 231 232 i Scheme 58 Enzyme: myrosinase N OSO3 – SGlc N OSO3 – SGlc 233 Sinigrin 234 Glucotropaeolin 48 Natural Product Reports, 199813C-Isotopic enrichments at various atomic positions in aromatic glucosinolates and cyanogenic glycosides can be correlated with the known biosynthetic origins of the individual carbons.344 These values may be used to distinguish between natural and synthetic materials.The cruciferous phytoalexins such as brassinin 235 and cyclobrassinin 236 found in pathogen-inoculated Japanese radish (Raphanus sativus var. hortensis) roots, are synthesized by a pathway closely related to glucosinolate biosynthesis (Scheme 60). Feeding experiments have demonstrated the conversion of brassinin 235 into cyclobrassinin 236, but not into methoxybrassinin 237, and 237 was not incorporated into 236.345 Camalexin 238 is the major phytoalexin of Arabidopsis thaliana, but previous studies (see ref. 1) have suggested tryptophan does not appear to be a precursor. Accumulation of camalexin is accompanied by induction of mRNAs and proteins for all the tryptophan pathway enzymes tested for, suggesting this pathway and phytoalexin synthesis are coordinately regulated.However, camalexin biosynthesis probably branches away from the main tryptophan pathway, e.g. at indole-3-glycerolphosphate.346 11 Miscellaneous shikimate metabolites 11.1 Sphagnum acid Results of feeding experiments with labelled L-phenylalanine, cinnamic acid and 4-coumaric acid have demonstrated their incorporation into trans-sphagnum acid 240 in cultures of the peat moss Sphagnum fallax.347 It is suggested that sphagnum acid thus arises from 4-coumaric acid 239 by further incorporation of a C2 fragment from acetate (Scheme 61).Other typical phenolics found in peat mosses, e.g. the hydroxybutenolide 241, 4-hydroxyacetophenone 242 and 4-hydroxybenzoic acid 79, may also be formed via sphagnum acid (Scheme 61). Light-grown cultures of S. fallax incorporate 4-coumaric acid into a glucosyl conjugate 244 of cis-4-coumaric acid 243 as well as into sphagnum acid.348 In the dark, this material is not formed, indicating light-dependent trans–cis isomerization of 4-coumaric acid.A 4-coumaric acid O-glucosyltransferase with strict specificity towards the cis isomer was subsequently isolated from a cytosolic fraction. O O R HO HO OH OH HO O –O O O HO HO OH OH –O O O O H O H O OH HO HO OH OH HO O –O O O S R HO HO OH OH –O O O HO HO OH OH O O H O H O OH HO HO OH OH –O O (a) b-Glucosidase ( b) Myrosinase Scheme 59 N OSO3 – SGlc NH NCS NH NH NH SMe S NH S N SMe 235 Brassinin 236 Cyclobrassinin L-Met Scheme 60 NH N SMe S OMe NH 237 Methoxybrassinin 238 Camalexin S N CO2H HO CO2H HO O HO O HO O CO2H HO HO CO2H GlcO CO2H CO2H HO Acetate 240 Sphagnum acid 242 79 241 243 UDPGlc 244 hn 239 Scheme 61 Dewick: The biosynthesis of shikimate metabolites 4911.2 4-(4-Hydroxyphenyl)butan-2-one The aroma of ripe raspberry (Rubus idaeus) fruits is due to raspberry ketone, 4-(4-hydroxyphenyl)butan-2-one 246.This compound is known to arise from condensation of 4-coumaroyl CoA 126 with malonyl CoA, giving the benzalacetone 4-(4-hydroxyphenyl)but-3-en-2-one 245, which is then reduced to 246 (Scheme 62).The condensation resembles the chalcone synthase and stilbene synthase reactions, but stops after the addition of the first malonyl unit, and the product is decarboxylated before release from the enzyme. A benzalacetone synthase enzyme preparation converting 4-coumaroyl CoA and malonyl CoA into 245 has now been obtained from ripe raspberry fruits.349 During purification, the enzyme preparation became preferentially enriched in benzalacetone synthase activity over chalcone synthase activity, but whether this was due to contamination with chalcone synthase, or whether benzalacetone synthase may possess a level of dual activity has not been clarified. 11.3 Phenylacetylenes from Asparagus Cultured cells of Asparagus oYcinalis produce 4-[5-(4- methoxyphenoxy)pent-3-en-1-ynyl]phenol 248. Preliminary feeding experiments using [1-13C]- and [U-13C]-glucose precursors gave labelling patterns in 248 which indicated that both aromatic rings were derived from shikimate.350 Labelling in the C5 chain suggested it might have its origins in a C6C2 intermediate plus a C3 glycolysis metabolite; [1,2-13C2]acetate was not incorporated.In further studies with 2H- and 13Clabelled phenylanine substrates,351 it was shown that all 17 skeletal carbons were derived from phenylalanine. One phenylalanine molecule provided the phenylacetylenic C6C2 unit, and the second the phenoxypropenyl unit.It was also shown that the sequence of labels in the three carbons C-9 to C-11 was reversed when compared to the side-chain of its phenylalanine precursor. To accommodate these findings, an intermediate 247 of the spirotetrahydrofuran type has been proposed (Scheme 63). 11.4 Diarylheptanoids and phenylphenalenones 9-Phenyl-1H-phenalenones are a group of plant pigments for which there is limited evidence that the ring system is derived from phenylalanine/tyrosine and that two C6C3 units may be incorporated via a diarylheptanoid.The simplest member of the series is anigorufone 250, which occurs in the roots of the Australian plant kangaroo paws (Anigozanthos preissii). To investigate its biosynthesis, [2-13C]cinnamic acid was fed to root cultures of the plant and shown to yield anigorufone labelled at C-5 and C-8 (Scheme 64).352 Further, 13C-labelled heptadienone 249 was well incorporated. Cyclization of 249 to the phenylphenalenone skeleton resembles an intramolecular [4+2] cycloaddition of the Diels–Alder type.Indeed, periodate oxidation of 249 is known to generate a 1,2-quinone which spontaneously cyclizes to lachnanthocarpone 252, and this, by reduction, may be the precursor of anigorufone. In further studies,353 the origin of all the carbon atoms in hydroxyanigorufone 251 was established using 13C-labelled precursors. In particular, the high incorporation of C-2 of acetate proved that the central carbon (C-6a) was derived from this precursor.The chemistry of natural diarylheptanoids, including their biosynthesis, has been reviewed.354 11.5 Antibiotic LL-C10037· Antibiotic LL-C10037· 255 is known to be formed in a species of Streptomyces from 3-hydroxyanthranilic acid 54 (see Section 2.12), and a pathway via dihydroxyacetanilide 253 (Scheme 65) has been established in earlier work. A COSCoA HO HO O HO O 126 245 246 i ii NADPH Malonyl-CoA CO2, HSCoA Scheme 62 Enzymes: i, benzalactone synthase; ii, benzalactone reductase HO O OMe HO2C NH2 O O O OH Phe C6C2 Phe 248 247 Ñ · * Ñ Ñ * * · · 7 8 9 10 11 Scheme 63 CO2H OH HO O O OH O O O Acetate O OH 252 Lachnanthocarpone 249 OH 250 Anigorufone R = H 251 Hydroxyanigorufone R = OH • = 13C 5 6a 8 O R Scheme 64 50 Natural Product Reports, 1998dihydroxyacetanilide epoxidase from Streptomyces oxidizes 253 to the epoxy quinone 254 utilizing molecular oxygen but without a requirement for any other cofactor.An NADPHdependent oxidoreductase then produces LL-C10037·. It has been shown that a full equivalent of 18O from 18O2 is incorporated at the epoxide.355 However, control reactions revealed a rapid exchange of oxygen with H2 18O at the C-4 carbonyl. By coupling the dihydroxyacetanilide epoxidase reaction with the epoxyquinone oxidoreductase from the same organism, LL-C10037· was produced with about 20% incorporation of a second 18O atom at the C-4 alcohol.This suggests that dihydroxyacetanilide epoxidase is a dioxygenase with an epoxidation mechanism (Scheme 66) which is essentially the same as that observed for dihydrovitamin K epoxidation during the mammalian vitamin K-dependent glutamate carboxylase reaction (see ref. 2). 11.6 Brominated tyrosine metabolites from sponges Feeding experiments have demonstrated low but significant incorporations of label from L-tyrosine, L-3-bromotyrosine 256 and L-3,5-dibromotyrosine 257 into dibromoverongiaquinol 261 and aeroplysinin-I 259 in the sponge Aplysina fistularis.356 [methyl-14C]Methionine was specifically incorporated into the O-methyl group of aeroplysinin-I, but no incorporations from the O-methyl ethers of tyrosine or 3,5-dibromotyrosine were recorded.A tentative pathway to 261 via successive bromination of tyrosine, and oxime 258 and phenylacetonitrile 260 intermediates has been proposed (Scheme 67). Unfortunately, mono- and di-brominated (4-hydroxyphenyl)acetonitriles, although present in A.fistularis, were not incorporated in feeding experiments, though this may be a consequence of the experimental diYculties encountered in using sponges for biosynthetic studies. The O-methylation required for production of 259 is thought to occur before formation of the nitrile, but potential brominated (methoxyphenyl)acetonitrile intermediates were not transformed. 11.7 Cyclohexanecarboxylic acid Cyclohexanecarboxylic acid is a component unit in several secondary metabolites, and it is derived from shikimic acid by a surprisingly complicated series of reactions, a combination of dehydrations and double bond reductions arranged so that the ring system is never prone to aromatization (Scheme 68) (see ref. 2). The gene encoding cyclohex-1-enylcarbonyl CoA reductase, the reductase catalyzing the final stage (264]265) in the formation of the cyclohexyl moiety of ansatrienein A 266 in Streptomyces collinus, has been cloned and characterized.357 The deduced amino acid sequence for the protein showed high similarity with members of short-chain alcohol dehydrogenase enzymes.The gene was overexpressed in E. coli to give a protein with comparable properties to the enzyme from S. collinus. However, it was able to catalyze in vitro three of the reductive steps involved in the formation of cyclohexanecarboxylic acid. In addition to the reduction of 264 into 265, this enzyme would also reduce 262 and 263 (Scheme 68).A mutant organism in which this gene had been disrupted lost this enzyme activity and the ability to synthesize cyclohexanecarboxylic acid and ansatrienin A. Since the three reductive steps all involve reduction of an ·,‚-conjugated double bond with the same stereochemical characteristics (shown in previous experiments to be anti, see ref. 2), it is possible that the single enzyme is responsible for catalyzing all three steps in the pathway. OH NH2 CO2H OH NH2 OH NH2 OH OH NHAc OH O NHAc O O O NHAc OH O – CO2 O2 NADPH 255 LL-C10037a 54 253 4 254 Scheme 65 OH NHAc OH O O Enz O OH O NHAc –O O– OH O NHAc O O NHAc O OH HO O NHAc O O O NHAc O O – H2O – H2O Scheme 66 CO2H NH2 HO Br CO2H NH2 HO Br Br CO2H NOH HO Br Br CN HO Br Br HO Br Br O NH2 O Br Br O NH2 OH CN MeO Br Br OH OH 259 Aeroplysinin-I 257 260 261 Dibromoverongiaquinol 256 L-Tyr 258 Scheme 67 Dewick: The biosynthesis of shikimate metabolites 51The potent immunosuppressants ascomycin (FK520) 270, FK506 269 and rapamycin 272 are macrocyclic polyketides that all contain a unit probably derived from 3,4-dihydroxycyclohexanecarboxylic acid.The biosynthetic pathway to this acid, while sharing features with that to cyclohexanecarboxylic acid, is known to diverge at the first intermediate beyond shikimate, namely the dihydroxy diene 262. The production of 4-hydroxy-3-methoxycyclohexanecarboxylic acid involves 3-O-methylation after assembly of the polyketide.The gene encoding 31-O-demethyl-FK506 methyltransferase has been isolated from two strains of Streptomyces species (FK506 producers) and S. hygroscopicus subsp. ascomyceticus (FK520 producer), sequenced and expressed in S. lividus.358 Disruption of the gene in one of the Streptomyces species yielded a mutant organism that produced 31-O-demethyl-FK506 267. The protein was established to methylate both 31-O-demethyl- FK506 267 and 31-O-demethyl-FK520 268.A second gene from these organisms was found to encode a protein with a strong sequence similarity to P450 hydroxylases. Disruption of this gene resulted in formation of 9-deoxo-31-O-demethyl- FK506 271, confirming that the gene indeed encoded a cytochrome P450 hydroxylase which was responsible for the 9-hydroxylation in FK506 and FK520. Clustered polyketide synthase genes in S. hygroscopicus responsible for the biosynthesis of rapamycin 272 have been characterized.359 A total of 70 constituent active sites makes this the most complex multienzyme system identified so far.The biosynthesis of the cyclohexanecarboxylic acidcontaining antibiotic asukamycin and other miscellaneous shikimate-derived antibiotics, e.g. reductiomycin, is discussed in a recent review.360 11.8 3-Amino-5-hydroxybenzoic acid and mC7N units The meta-C7N aminobenzoate units of several ansamycin antibiotics are provided by 3-amino-5-hydroxybenzoic acid 276 which is derived from a modification of the shikimate pathway.A DAHP synthase-like enzyme synthesizes amino- DAHP 273 via the introduction of nitrogen from glutamine, then the amino group is retained in what appears to be a normal pathway (Scheme 69) (see ref. 1). Cell-free extracts of the rifamycin B 277 producer Amycolatopsis mediterranei have been shown to transform the various intermediates in the pathway into 3-amino-5-hydroxybenzoic acid.361 Thus, amino- DAHP 273, aminoDHQ 274 and aminoDHS 275 were eY- ciently modified, but DAHP was not.The low, but significant, conversions of PEP and erythrose 4-phosphate into amino- DAHP and into 3-amino-5-hydroxybenzoic acid were also COSCoA HO OH OH COSCoA OH OH COSCoA OH OH COSCoA OH COSCoA OH COSCoA OH COSCoA COSCoA COSCoA O O NH OMe HO O O O NH O Shikimoyl-CoA 266 Ansatrienin A 265 262 Although all intermediates are shown as coenzyme A thioesters, the exact point of esterification is not known 264 263 Scheme 68 HO R1O O HO O R2 OMe O OMe OH O O N O HO HO O HO O OMe O OMe OH O N O HO MeO O O OMe OH O N 267 R1 = H; R2 = CH2CH=CH2 268 R1 = H; R2 = Et 269 FK506, R1 = Me; R2 = CH2CH=CH2 270 FK520, R1 = Me; R2 = Et O O O OMe OH O MeO 271 9 31 272 Rapamycin 52 Natural Product Reports, 1998noted.Cell-free extracts of the ansatrienin A 266 producer Streptomyces collinus would also occasionally catalyze transformations, but much less eYciently than A. mediterranei. Although a modified DAHP synthase is required for amino- DAHP synthesis, the next two reactions in the pathway could be normal pathway enzymes.Thus, 3-dehydroquinate synthase and shikimate dehydrogenase from E. coli can both catalyze transformation of the amino analogues, but 3-dehydroquinase (a type II enzyme, see Section 2.3) from A. mediterranei does not accept aminoDHQ. 11.9 3-Amino-4-hydroxybenzoic acid 3-Amino-4-hydroxybenzoic acid, a biosynthetic precursor of 4-hydroxy-3-nitrosobenzamide in Streptomyces murayamaensis is unusual in being the only aminohydroxybenzoic acid derivative so far reported which is not derived via shikimate.It has been established to arise from a C4 unit from the tricarboxylic acid cycle and a C3 unit, possibly PEP.362 11.10 Betalains Betalains are yellow to violet water-soluble nitrogenous pigments restricted to plants of the order Centrospermae, but found also in the caps of some larger fungi such as Amanita and Hygrocybe.Some structures, e.g. the betacyanin betanin 280 are derived from two molecules of DOPA 34 via cycloDOPA 36 (see Scheme 11) and the ring-cleaved betalamic acid 22, whilst others contain only the betalamic acid portion that is DOPA-derived. A tyrosinase enzyme (see Section 2.11) isolated from the betalain-producing coloured parts of the fly agaric (Amanita muscaria) is probably involved in betalain biosynthesis.363 This enzyme converted tyrosine into DOPA, but was not specific for tyrosine.In addition, it would oxidize a number of diphenols to o-quinones (diphenolase activity) and may therefore contribute towards melanin biosynthesis. Unusually for tyrosinases, the enzyme was a heterodimer. Betanidin O-glucosyltransferase activities have been isolated from cell suspension cultures of Dorotheanthus bellidiformis and shown to catalyze 5-O-glucosylation giving betanin 280 and 6-O-glucosylation giving gomphrenin I 281 (Scheme 70).364 Three isoforms of the 5-O-glucosyltransferase were identified.Kinetic data for the enzymes suggest that at low substrate concentrations, betanin formation would be favoured, whilst at high substrate concentration, gomphrenin I would be the preferred product. There is no such correlation between extractable O-glucosyltransferase activities and in vivo accumulation of products. 12 References 1 P. M. Dewick, Nat. Prod. Rep., 1995, 12, 579. 2 P. M. Dewick, Nat. Prod. Rep., 1995, 12, 101. 3 P. M. Dewick, Nat. Prod. Rep., 1994, 11, 173. 4 E. Haslam, Prog. Chem. Org. Nat. Prod., 1996, 69, 157. 5 K. M. Herrmann, Plant Cell, 1995, 7, 907. 6 K. M. Herrmann, Plant Physiol., 1995, 107, 7. 7 J. Schmid and N. Amrhein, Phytochemistry, 1995, 39, 737. 8 H. Geiger, S. El-Dessouki and T. Seeger, Phytochemistry, 1995, 40, 1705. 9 P. J. Duggan, E. Parker, J. Coggins and C. Abell, Bioorg. Med. Chem. Lett., 1995, 5, 2347. 10 G. J. W. Euverink, G. I. Hessels, C. Franke and C. Dijkhuizen, Appl.Environ. Microbiol., 1995, 61, 3796. 11 G. E. Walker, B. Dunbar, I. S. Hunter, H. G. Nimmo and J. R. Coggins, Microbiology, 1996, 142, 1973. 12 I. A. Shumilin, R. H. Kretsinger and R. Bauerle, Proteins: Struct., Funct., Genet., 1996, 24, 404. 13 N. Suzuki, M. Sakuta and S. Shimizu, J. Plant Physiol., 1996, 149, 19. 14 J. D. Jones, J. M. Henstrand, A. K. Handa, K. M. Herrmann and S. C. Weller, Plant Physiol., 1995, 108, 1413. 15 M. BischoV, J. Rösler, H.-R. Raesecke, J.Görlach, N. Amrhein and J. Schmid, Plant Mol. Biol., 1996, 31, 69. 16 N. Yamauchi and K. Katinuma, J. Org. Chem., 1995, 60, 5614. 17 T. Krell, A. R. Pitt and J. R. Coggins, FEBS Lett., 1995, 360, 93. 18 T. Krell, M. J. Horsburgh, A. Cooper, S. M. Kelly and J. R. Coggins, J. Biol. Chem., 1996, 271, 24 492. 19 J. M. Harris, C. Gonzalez-Bello, C. Kleanthous, A. R. Hawkins, J. R. Coggins and C. Abell, Biochem. J., 1996, 319, 333. 20 J. R. Bottomley, A. R. Hawkins and C. Kleanthous, Biochem.J., 1996, 319, 269. CO2H PO HO OH NH2 O OH NH2 O CO2H HO OH NH2 O CO2H HO NH2 CO2H AcO OH OH MeO NH O O HO HO O O O CO2H Gln 274 AminoDHQ PEP + Erythrose 4-phosphate 273 AminoDAHP 276 Glu 275 AminoDHS 277 Rifamycin B Scheme 69 CO2H NH2 HO HO HO2C CO2H NH2 HO O NH CO2H HO HO NH O CO2H HO2C NH CO2H HO2C N CO2 – HO HO NH CO2H HO2C N CO2 – R1O R2O 34 L-DOPA 36 CycloDOPA 278 Betalamic acid 279 Betanidin 280 Betanin R1 = Glc; R2 = H 281 Gomphrenin I R1 = H; R2 = Glc 6 UDPGlc 5 + + Scheme 70 Dewick: The biosynthesis of shikimate metabolites 5321 J.R. Bottomley, C. L. Clayton, P. A. Chalk and C. Kleanthous, Biochem. J., 1996, 319, 559. 22 J. M. Harris, W. J. Watkins, A. R. Hawkins, J. R. Coggins and C. Abell, J. Chem. Soc., Perkin Trans. 1, 1996, 2371. 23 K. A. Wheeler, H. K. Lamb and A. R. Hawkins, Biochem. J., 1996, 315, 195. 24 X. Kang, H. E. Neuhaus and R. Scheibe, Z. Naturforsch., C: Biosci., 1994, 49, 415. 25 T. Tateoka and I.Yasuda, Plant Cell Rep., 1995, 15, 212. 26 C. Leuschner, K. M. Herrmann and G. Schultz, Plant Physiol., 1995, 108, 319. 27 M. J. Whipp and A. J. Pittard, J. Bacteriol., 1995, 177, 1627. 28 K. Majumder, A. Selvapandiyan, F. A. Fattah, N. Arora, S. Ahmad and R. K. Bhatnagar, Eur. J. Biochem., 1995, 229, 99. 29 A. Selvapandiyan, S. Ahmad, K. Majumder, N. Arora and R. K. Bhatnagar, Biochem. Mol. Biol. Int., 1996, 40, 603. 30 W. A. Shuttleworth and J. N. S. Evans, Arch. Biochem. Biophys., 1996, 334, 37. 31 R. D. Sammons, K. J. Gruys, K. S. Anderson, K. A. Johnson and J. A. Sikorski, Biochemistry, 1995, 34, 6433. 32 L. M. McDowell, C. A. Klug, D. D. Beusen and J. Schaefer, Biochemistry, 1996, 35, 5395. 33 M. A. Marzabadi, K. J. Gruys, P. D. Pansegrau, M. C. Walker, H. K. Yuen and J. A. Sikorski, Biochemistry, 1996, 35, 4199. 34 M. L. Peterson, S. D. Corey, J. L. Font, M. C. Walker and J. A. Sikorski, Bioorg. Med. Chem. Lett., 1996, 6, 2853. 35 S. Bornemann, S.Balasubramanian, J. R. Coggins, C. Abell, D. J. Lowe and R. N. F. Thorneley, Biochem. J., 1995, 305, 707. 36 S. Bornemann, M. K. Ramjee, S. Balasubramanian, C. Abell, J. R. Coggins, D. J. Lowe and R. N. F. Thorneley, J. Biol. Chem., 1995, 270, 22 811. 37 S. Bornemann, D. J. Lowe and R. N. F. Thorneley, Biochemistry, 1996, 35, 9907. 38 J. M. Henstrand, N. Amrhein and J. Schmid, J. Biol. Chem., 1995, 270, 20 447. 39 J. M. Henstrand, N. Amrhein and J. Schmid, Plant Physiol., 1995, 108, 1127. 40 M. Braun, J. M. Henstrand, J. Görlach, N. Amrhein and J. Schmid, Planta, 1996, 200, 64. 41 P. D. Lyne, A. J. Mulholland and W. G. Richards, J. Am. Chem. Soc., 1995, 117, 11 345. 42 O. Wiest and K. N. Houk, J. Am. Chem. Soc., 1995, 117, 11 628. 43 P. Kast, M. Asif-Ullah and D. Hilvert, Tetrahedron Lett., 1996, 37, 2691. 44 Y. Xue and W. N. Lipscomb, Proc. Natl. Acad. Sci. USA, 1995, 92, 10 595. 45 N. Sträter, K. Håkansson, G. Schnappauf, G. Braus and W. N. Lipscomb, Proc.Natl. Acad. Sci. USA, 1996, 93, 3330. 46 A. Y. Lee, P. A. Karplus, B. Ganem and J. Clardy, J. Am. Chem. Soc., 1995, 117, 3627. 47 S. Zhang, P. Kongsaeree, J. Clardy, D. B. Wilson and B. Ganem, Biooorg. Med. Chem., 1996, 4, 1015. 48 D. R. Liu, S. T. Cload, R. M. Pastor and P. G. Schultz, J. Am. Chem. Soc., 1996, 118, 1789. 49 C. C. Galopin, S. Zhang, D. B. Wilson and B. Ganem, Tetrahedron Lett., 1996, 37, 8675. 50 S. T. Cload, D. R. Liu, R. M. Pastor and P. G. Schultz, J.Am. Chem. Soc., 1996, 118, 1787. 51 P. Kast, M. Asif-Ullah, N. Jiang and D. Hilvert, Proc. Natl. Acad. Sci. USA, 1996, 93, 5043. 52 D. Christendat and J. Turnbull, Biochemistry, 1996, 35, 4468. 53 J. Eberhard, T. T. Ehrler, P. Epple, G. Felix, H.-R. Raesecke, N. Amrhein and J. Schmid, Plant J., 1996, 10, 815. 54 C. Leuscher, N. J. Walton, G. Herzog and G. Schultz, Plant Physiol. Biochem. (Paris), 1995, 33, 367. 55 R. M. Romero, M. F. Roberts and J. D. Phillipson, Phytochemistry, 1995, 40, 1015. 56 S. Bornemann, D. J. Lowe and R. N. F. Thorneley, Biochem. Soc. Trans., 1996, 24, 84. 57 A. Abou-Zeid, G. J. W. Euverink, G. I. Hessels, R. A. Jensen and L. Dijkhuizen, Appl. Environ. Microbiol., 1995, 61, 1298. 58 A. M. Abou-Zeid, E. Elwy and A. R. El-Shanshoury, Microbios., 1995, 81, 213. 59 G. J. W. Euverink, D. J. Wolters and L. Dijkhuizen, Biochem. J., 1995, 308, 313. 60 G.-E. Séralini, V. Luu-Thé and F. Labrie, Biochim. Biophys. Acta, 1995, 1260, 97. 61 M.Montemartini, J. Bua, E. Bontempi, C. Zelada, A. M. Ruiz, J. A. Santome, J. J. Cazzulo and C. Nowicki, FEMS Microbiol. Lett., 1995, 133, 17. 62 L. Soto-Urzua, Y. G. Xochinua-Corona, M. Flores-Encarnacion and B. E. Bala, Can. J. Microbiol., 1996, 42, 294. 63 C. A. Bonner, R. S. Fischer, R. R. Schmidt, P. W. Miller and R. A. Jensen, Plant Cell Physiol., 1995, 36, 1013. 64 K. D. Snell, K. M. Draths and J. W. Frost, J. Am. Chem. Soc., 1996, 118, 5605. 65 A. Yamada, T. Dairi, Y.Ohno, X.-L. Huang and Y. Asano, Biosci. Biotech. Biochem., 1995, 59, 1994. 66 S. Y. K. Seah, K. L. Britton, P. J. Baker, D. W. Rice, Y. Asano and P. C. Engel, FEBS Lett., 1995, 370, 93. 67 R. T. Carr, S. Balusubramanian, P. C. D. Hawkins and S. J. Benkovic, Biochemistry, 1995, 34, 7525. 68 D. Kowlessur, B. A. Citron and S. Kaufmann, Arch. Biochem. Biophys., 1996, 333, 85. 69 A. Martinez, P. M. Knappskog, S. Olafsdottir, A. P. Døskeland, H. G. Eiken, R. M. Svebak, M. Bozzini, J.Apold and T. Flatmark, Biochem. J., 1995, 306, 589. 70 D. Kowlessur, X.-J. Yang and S. Kaufmann, Proc. Natl. Acad. Sci. USA, 1995, 92, 4743. 71 P. J. Hillas and P. F. Fitzpatrick, Biochemistry, 1996, 35, 6969. 72 A. J. Ramsey, P. J. Hillas and P. F. Fitzpatrick, J. Biol. Chem., 1996, 271, 24 395. 73 A. J. Ramsey, S. C. Daubner, J. I. Ehrlich and P. F. Fitzpatrick, Protein Sci., 1995, 4, 2082. 74 S. Boularand, M. C. Darmon and J. Mallet, J. Biol. Chem., 1995, 270, 3748. 75 C.M. D’Sa, R. E. Arthur, I. Jennings, R. G. H. Cotton and D. M. Kuhn, J. Neurosci. Methods, 1996, 69, 149. 76 C. M. D’Sa, R. E. Arthur, J. C. States and D. M. Kuhn, J. Neurochem., 1996, 67, 900. 77 C. M. D’Sa, R. E. Arthur and D. M. Kuhn, J. Neurochem., 1996, 67, 917. 78 S. E. Hufton, I. G. Jennings and R. G. H. Cotton, Biochem. J., 1995, 311, 353. 79 T. Nagatsu, Essays Biochem., 1995, 30, 15. 80 S. Kaufman, Adv. Enzymol. Relat. Areas Mol. Biol., 1995, 70, 103. 81 H. Wang, W.Liu and N. Ulbrich, Biochim. Biophys. Acta, 1995, 1243, 251. 82 J. Wang, R. Kou, H. Cheng, J. Yuan and Y. Zhou, Shanxi Daxue Xuebao, Ziran Kexueban, 1995, 18, 184; Chem. Abstr., 1995, 123, 136 928. 83 H. J. Wichers, Y. A. M. Gerritsen and C. G. J. Chapelon, Phytochemistry, 1996, 43, 333. 84 H. J. Wichers, T. van den Bosch, Y. A. M. Gerritsen, J. I. Oyevaar, C. E. M. Ebbelaar, K. Recourt and R. W. Kerrigan, Mushroom Sci., 1995, 14, 723; Chem. Abstr., 1996, 125, 136 306. 85 K. Kanda, T.Sato, S. Ishii, H. Enei and S.-I. Ejiri, Biosci. Biotech. Biochem., 1996, 60, 1273. 86 A. K. Chakraborty, J. T. Platt, K. K. Kim, B. S. Kwon, D. C. Bennett and J. M. Pawelek, Eur. J. Biochem., 1996, 236, 180. 87 P. Bjoerk, P. Aaman, A. Hindemith, G. Odh, L. Jacobsson, E. Rosengren and H. Rorsman, Eur. J. Haematol., 1996, 57, 254. 88 M. Zhang, P. Åman, A. Grubb, I. Panagopoulos, A. Hindemith, E. Rosengren and H. Rorsman, FEBS Lett., 1995, 373, 203. 89 A. Sanchez-Ferrer, J.N. Rodriguez-Lopez, F. Garcia-Canovas and F. Garcia-Carmona, Biochim. Biophys. Acta, 1995, 1247, 1. 90 V. del Marmol and F. Beermann, FEBS Lett., 1996, 381, 165. 91 A. Ramaiah, Indian J. Biochem. Biophys., 1996, 33, 349. 92 K. Lerch, in Enzymatic Browning and Its Prevention, ed. C. Y. Lee and J. R. Whitaker, ACS Symposium Series, Vol. 600, ACS, Washington, DC, 1995, p. 64. 93 G. Prota, Prog. Chem. Org. Nat. Prod., 1995, 64, 93. 94 E. A. Ray, J. M. Green and B. P. Nichols, Biochim.Biophys. Acta, 1996, 1295, 81. 95 V. K. Viswanathan, J. M. Green and B. P. Nichols, J. Bacteriol., 1995, 177, 5918. 96 J. Bohlmann, V. DeLuca, U. Eilert and W. Martin, Plant J., 1995, 7, 491. 97 J. Bohlmann, T. Lins, W. Martin and U. Eilert, Plant Physiol., 1996, 111, 507. 98 R. Romero and M. F. Roberts, Phytochemistry, 1996, 41, 395. 99 J. Li and R. L. Last, Plant Physiol., 1996, 110, 51. 100 M. C. Kozlowski, N. J. Tom, C. T. Seto, A. M. Sefler and P. A. Bartlett, J.Am. Chem. Soc., 1995, 117, 2128. 101 B. Ganem, Tetrahedron Lett., 1995, 36, 815. 102 R. Romero, M. F. Roberts and J. D. Phillipson, Phytochemistry, 1995, 39, 263. 103 Y. Iwamoto, I. S. M. Lee, M. Tsubaki and R. Kido, Can. J. Microbiol., 1995, 41, 19. 54 Natural Product Reports, 1998104 F. Takeuchi, R. Tsubouchi, M. Yoshino and Y. Shibata, Biochim. Biophys. Acta, 1995, 1252, 185. 105 D. Alberti-Giani, R. Buchli, P. Malherbe, C. Broger, G. Lang, C. Köhler, H.-W. Lahm and A.M. Cesura, Eur. J. Biochem., 1996, 239, 460. 106 H. Tsuji, T. Oka, M. Kimoto, Y.-M. Hong, Y. Natori and T. Ogawa, Biochim. Biophys. Acta, 1996, 1309, 31. 107 S. R. Desai, P. Kumar and W. S. Chilton, Chem. Commun., 1996, 1321. 108 Y. Hashimoto and K. Shudo, Phytochemistry, 1996, 43, 551. 109 O. Kwon, M. E. S. Hudspeth and R. Meganathan, J. Bacteriol., 1996, 178, 3252. 110 R. Müller, C. Dahm, G. Schulte and E. Leistner, FEBS Lett., 1996, 378, 131. 111 B. M. Rowland and H.W. Taber, J. Bacteriol., 1996, 178, 854. 112 A. H. Lodhi, R. J. M. Bongaerts, R. Verpoorte, S. A. Coomber and B. V. Charlwood, Plant Cell Rep., 1996, 16, 54. 113 U. Hommel, M. Eberhard and K. Kirschner, Biochemistry, 1995, 34, 5429. 114 M. Eberhard, M. Tsai-Pfugfelder, K. Bolewska, U. Hommel and K. Kirschner, Biochemistry, 1995, 34, 5419. 115 J. Li, J. Zhao, A. B. Rose, R. Schmidt and R. L. Last, Plant Cell, 1995, 7, 447. 116 J. Li, S. Chen, L. Zhu and R. L. Last, Plant Physiol., 1995, 108, 877. 117 M. Hennig, B. Darimont, R. Sterner, K. Kirschner and J. N. Jansonius, Structure (London), 1995, 3, 1295. 118 T. R. Knoechel, M. Hennig, A. Merz, B. Darimont, K. Kirschner and J. N. Jansonius, J. Mol. Biol., 1996, 262, 502. 119 K. S. Anderson, A. Y. Kim, J. M. Quillen, E. Sayers, X.-J. Yang and E. W. Miles, J. Biol. Chem., 1995, 270, 29 936. 120 L. M. McDowell, M. Lee, J. Schaefer and K. S. Anderson, J. Am. Chem. Soc., 1995, 117, 12 352. 121 U. Banik, D.-M. Zhu, P.B. Chock and E. W. Miles, Biochemistry, 1995, 34, 12 704. 122 C. A. Leja, E. U. Woehl and M. F. Dunn, Biochemistry, 1995, 34, 6552. 123 E. R. Radwanski, J. Zhao and R. L. Last, Mol. Gen. Genet., 1995, 248, 657. 124 V. C. Kramer and M. G. Koziel, Plant Mol. Biol., 1995, 27, 1183. 125 J. Zhao and R. L. Last, J. Biol. Chem., 1995, 270, 6081. 126 C.-Y. Lai, P. Baumann and A. N. Moran, Insect Mol. Biol., 1995, 4, 47. 127 E. W. Miles, in Proteins: Structure, Function and Engineering, eds.B. Biswas and S. Roy, Plenum, New York, 1995, p. 207. 128 E. W. Miles, S. A. Ahmed, C. C. Hyde, A. M. Kayastha, X.-J. Yang, S. B. Ruvinov and Z. Lu, in Molecular Aspects of Enzyme Catalysis, eds. T. Fukui and K. Soda, Kodansha, Tokyo, 1994, p. 127. 129 E. R. Radwanski and R. L. Last, Plant Cell, 1995, 7, 921. 130 Y. Liu, A. L. Silverstone, Y. M. Wu and S. F. Yang, Phytochemistry, 1995, 40, 691. 131 Y. M. Wu, A. L. Silverstone, Y. Liu and S. F. Yang, Phytochemistry, 1995, 40, 699. 132 R. Genet, P.-H. Bénetti, A. Hammadi and A. Ménez, J. Biol. Chem., 1995, 270, 23 540. 133 K. Kawasaki, A. Yokota and F. Tomita, Biosci. Biotech. Biochem., 1995, 59, 1938. 134 M. J. Sloan and R. S. Phillips, Biochemistry, 1996, 35, 16 165. 135 Y. M. Torchinsky and Y. Kawata, in Molecular Aspects of Enzyme Catalysis, eds. T. Fukui and K. Soda, Kodansha, Tokyo, 1994, p. 165. 136 T. A. Karkova, K. Z. Gamburg, L. V. Gamanets and A. G. Enikeev, Russ. J. Plant Physiol., 1995, 42, 585; Chem.Abstr., 1995, 123, 251 407. 137 J. J. Emanuele, C. J. Heasley and P. F. Fitzpatrick, Arch. Biochem. Biophys., 1995, 316, 241. 138 J. J. Emanuele and P. F. Fitzpatrick, Biochemistry, 1995, 34, 3710. 139 J. J. Emanuele and P. F. Fitzpatrick, Biochemistry, 1995, 34, 3716. 140 L. Zhou, B. Bartel and R. Thornburg, Plant Physiol., 1996, 110, 337. 141 R. C. Schmidt, A. Müller, R. Hain, D. Bartling and E. W. Weiler, Plant J., 1996, 9, 683. 142 M. Kobayashi, T. Suzuki, T.Fujita, M. Masuda and S. Shimizu, Proc. Natl. Acad. Sci. USA, 1995, 92, 714. 143 M. Brandl, E. M. Clark and S. E. Lindow, Can. J. Microbiol., 1996, 42, 586. 144 K. Minamisawa, K.-I. Ogawa, H. Fukuhara and J. Koga, Plant Cell Physiol., 1996, 37, 449. 145 T. Furukawa, J. Koga, T. Adachi, K. Kishi and K. Syono, Plant Cell Physiol., 1996, 37, 899. 146 T. Bar and Y. Okon, in Azospirillum VI and Related Microorganisms: Genetics, Physiology, Ecology, eds. I. Fendrick, M. del Gallo, J.Vanderleyden and M. de Zamaroczy, Springer, Berlin, 1995, p. 347. 147 C. W. Basse, F. Lottspeich, W. Steglich and R. Kahmann, Eur. J. Biochem., 1996, 242, 648. 148 S. Sugaya and S. Sakai, Physiol. Plant., 1996, 97, 433. 149 J. Zhu and G. K. Scott, Biochem. Mol. Biol. Int., 1995, 35, 423. 150 T. Koshiba, Y. Kamiya and M. Iino, Plant Cell Physiol., 1995, 36, 1503. 151 P. J. Jensen and R. S. Bandurski, J. Plant Physiol., 1996, 147, 697. 152 N. I. Rekoslavskaya, Russ.J. Plant Physiol., 1995, 42, 143; Chem. Abstr., 1995, 122, 235 369. 153 J. B. Jensen, H. Egsgaard, H. van Onckelen and B. U. Jochimsem, J. Bacteriol., 1995, 177, 5762. 154 J. Ludwig-Müller, W. Hilgenberg and E. Epstein, Phytochemistry, 1995, 40, 61. 155 J. Ludwig-Müller and W. Hilgenberg, Physiol. Plant., 1995, 94, 651. 156 J. Normanly, J. P. Slovin and J. D. Cohen, Plant Physiol., 1995, 107, 323. 157 R. S. Bandurski, J. D. Cohen, J. P. Slovin and D. M. Reinecke, in Plant Hormones, ed.P. J. Davies, Kluwer, Dordrecht, Netherlands, 2nd edn, 1995, p. 39. 158 M. Kawaguchi and K. Syono, Plant Cell Physiol., 1996, 37, 1043. 159 A. Costacurta and J. van der Leyden, Crit. Rev. Microbiol., 1995, 21, 1. 160 J. Koga, Biochim. Biophys. Acta, 1995, 1249, 1. 161 S.-U. Kim, K.-S. Song, D.-S. Jung, Y.-A. Chae and H. J. Lee, Planta Med., 1996, 62, 54. 162 S. V. Pletnev, M. N. Isupov, Z. Dauter, K. S. Wilson, N. G. Faleev, E. G. Harutyunyan and T. V. Demidkina, Biochem.Mol. Biol. Int., 1996, 38, 37. 163 H. Y. Chen, T. V. Demidkina and R. S. Phillips, Biochemistry, 1995, 34, 12 276. 164 H. Y. Chen, P. Gollnick and R. S. Phillips, Eur. J. Biochem., 1995, 229, 540. 165 M. Siebert, S. Sommer, S. Li, Z. Wang, K. Severn and L. Heide, Plant Physiol., 1996, 112, 811. 166 K. Yazaki, K. Inushima, M. Kataoka and M. Tabata, Phytochemistry, 1995, 38, 1127. 167 P. Meuwly, W. Mölders, A. Buchala and J.-P. Métraux, Plant Physiol., 1995, 109, 1107. 168 J.Leon, V. Shulaev, N. Yalpani, M. A. Lawton and I. Raskin, Proc. Natl. Acad. Sci. USA, 1995, 92, 10 413. 169 P. Silverman, M. Seskar, D. Kanter, P. Schweizer, J.-P. Métraux and I. Raskin, Plant Physiol., 1995, 108, 633. 170 S. Seo, K. Ishizuka and Y. Ohashi, Plant Cell Physiol., 1995, 36, 447. 171 W. A. Ayer and E. R. Cruz, J. Nat. Prod., 1995, 58, 622. 172 H.-I. Lee, J. Leon and I. Raskin, Proc. Natl. Acad. Sci. USA, 1995, 92, 4076. 173 S. I. Kotob, S. L. Coon, E. J. Quintero and R.M. Weiner, Appl. Environ. Microbiol., 1995, 61, 1620. 174 J. E. Baldwin, N. P. Crouch, Y. Fujishima, M. H. Lee, C. H. MacKinnon, J. P. N. Pitt and A. C. Willis, Bioorg. Med. Chem. Lett., 1995, 5, 1255. 175 N. P. Crouch, J. E. Baldwin, M.-H. Lee, C. H. MacKinnon and Z. H. Zhang, Bioorg. Med. Chem. Lett., 1996, 6, 1503. 176 M.-H. Lee, Z.-H. Zhang, C. H. MacKinnon, J. E. Baldwin and N. P. Crouch, FEBS Lett., 1996, 393, 269. 177 R. A. Dixon and N. L. Paiva, Plant Cell, 1995, 7, 1085. 178 Z. Hao, D. J. Charles, L. Yi and J. E. Simon, Phytochemistry, 1996, 43, 735. 179 G. D. Rees and D. H. Jones, Enzyme Microb. Technol., 1996, 19, 282. 180 S. Takac, B. Akay and T. H. Oezdamar, Enzyme Microb. Technol., 1995, 17, 445. 181 G. B. D’Cunha, V. Satyanarayan and P. M. Nair, Phytochemistry, 1996, 42, 17. 182 R. I. Monge, M. Lara and A. Lopez-Munguia, Biotechnol. Tech., 1995, 9, 423. 183 S. H. Kim, J. W. Kronstad and B. E. Ellis, Phytochemistry, 1996, 43, 351. 184 L. A. Wanner, G. Li, D. Ware, I. E. Somssich and K. R. Davis, Plant Mol. Biol., 1995, 27, 327. 185 L. Chye-Fong, G. Chong-Jin, L. Chiang-Shiong and L. Saw- Hoon, Plant Physiol., 1996, 112, 863. Dewick: The biosynthesis of shikimate metabolites 55186 D. Seelenfreund, M. Chiong, S. Lobos and L. M. Prez, Plant Physiol., 1996, 111, 348. 187 T. Fukasawa-Akada, S. Kung and J. C. Watson, Plant Mol. Biol., 1996, 30, 711. 188 Q. Zhu, T. Dabi, A. Beeche, R. Yamamoto, M. A. Lawton and C.Lamb, Plant Mol. Biol., 1995, 29, 535. 189 E. Logemann, M. Parniske and K. Hahlbrock, Proc. Natl. Acad. Sci. USA, 1995, 92, 5905. 190 Y. Osakabe, Y. Ohtsubo, S. Kawai, Y. Katayama and N. Morohoshi, Plant Sci., 1995, 105, 217. 191 Y. Osakabe, K. Osakabe, S. Kawai, Y. Katayama and N. Morohoshi, Plant Mol. Biol., 1995, 28, 1133. 192 J. M. Manners, C. L. McIntyre and J. P. Nourse, Plant Physiol., 1995, 108, 1301. 193 Y.-C. Liao, H.-P. Li, F. M. Kreuzaler and R. Fischer, Plant Physiol., 1996, 112, 1398. 194 G. R. McKegney, S. L. Butland, D. Theilmann and B. E. Ellis, Phytochemistry, 1996, 41, 1259. 195 P. A. Howles, V. J. H. Sewalt, N. L. Paiva, Y. Elkind, N. J. Bate, C. Lamb and R. A. Dixon, Plant Physiol., 1996, 112, 1617. 196 B. Schuster and J. Rétey, Proc. Natl. Acad. Sci. USA, 1995, 92, 8433. 197 L. Ge and V. L. Chiang, Plant Physiol., 1996, 112, 861. 198 S. Kawai, A. Mori, T. Shiokawa, S. Kajita, Y. Katayama and N. Morohoshi, Biosci. Biotech. Biochem., 1996, 60, 1586. 199 M. Hutze, G. Schröder and J. Schröder, FEBS Lett., 1995, 374, 345. 200 F. Durst, I. Benveniste, M. Schalk and D. Werck-Reichhart, Methods Enzymol., 1996, 272, 259. 201 K. Meyer, J. C. Cusumano, C. Somerville and C. C. S. Chapple, Proc. Natl. Acad. Sci. USA, 1996, 93, 6869. 202 T. B. T. Lam, K. Iiyama and B. A. Stone, Phytochemistry, 1996, 41, 1507. 203 H. Meng and W. H. Campbell, Arch. Biochem. Biophys., 1996, 330, 329. 204 C.-J. Tsai, G. K. Podila and V.L. Chiang, Plant Physiol., 1995, 107, 1459. 205 J. Garcia-Mas, R. Messeguer, P. Arus and P. Puigdomenech, Plant Physiol., 1995, 108, 1341. 206 Z.-H. Ye and J. E. Varner, Plant Physiol., 1995, 108, 459. 207 X.-H. Zhang, E. E. Dickson and C. C. Chinnappa, Plant Physiol., 1995, 108, 429. 208 H. Meng and W. H. Campbell, Plant Physiol., 1995, 108, 1749. 209 K. S. Voo, R. W. Whetten, D. M. O’Malley and R. R. SederoV, Plant Physiol., 1995, 108, 85. 210 D. Lee, M. Ellard, L. A. Wanner, K.R. Davis and C. J. Douglas, Plant Mol. Biol., 1995, 28, 871. 211 D. Lee and C. J. Douglas, Plant Physiol., 1996, 112, 193. 212 K. Yazaki, A. Ogawa and M. Tabata, Plant Cell Physiol., 1995, 36, 1319. 213 A. Kawaoka and V. L. Chiang, Biotechnol. Pulp Pap. Ind., 1995, 6, 315; Chem. Abstr., 1997, 126, 56 655. 214 S. Lofty, Plant Physiol. Biochem (Paris), 1995, 33, 423. 215 M. Hohlfeld, M. Veit and D. Strack, Plant Physiol., 1996, 111, 1153. 216 S. Lofty, F. Javelle and J. Negrel, Planta, 1996, 199, 475. 217 S. Lofty, F. Javelle and J. Negrel, Phytochemistry, 1995, 40, 389. 218 H. Hohlfeld, W. Schürmann, D. Scheel and D. Strack, Plant Physiol., 1995, 107, 545. 219 H. Hohlfeld, D. Scheel and D. Strack, Planta, 1996, 199, 166. 220 C. Hedberg, M. Hesse and C. Werner, Plant Sci., 1996, 113, 149. 221 S. X. Wang, T. Vogt and B. E. Ellis, in Phytochemicals and Health, eds. D. L. Gustine and H. E. Flores, American Society for Plant Physiology, Rockville, Maryland, 1995, p. 300. 222 M. Petersen, E. Szabo, J. Meinhard, B. Karwatzki, C. Gertlowski, B. Kempin and E. Fuß, Plant Cell, Tissue Organ Cult., 1995, 43, 89. 223 P. Betry, M. A. Fliniaux, M. Mackova, F. Gillet, T. Macek and A. Jacquin-Dubreuil, J. Plant Physiol., 1995, 146, 503. 224 R. Atanassova, N. Favet, F. Martz, B. Chabbert, M.-T. Tollier, B. Monties, B. Fritig and M. Legrand, Plant J., 1995, 8, 465. 225 J. van Doorsselaere, M. Baucher, E. Chognot, B. Chabbert, M.-T. Tollier, M. Petit-Conil, J.-C.Leplé, G. Pilate, D. Cornu, B. Monties, M. van Montagu, D. Inzé, W. Boerjan and L. Jouanin, Plant J., 1995, 8, 855. 226 S. Kajita, Y. Katayama and S. Omori, Plant Cell Physiol., 1996, 37, 957. 227 J. van Doorsselaere, M. Baucher, C. Feuillet, A. M. Boudet, M. van Montagu and D. Inzé, Plant Physiol. Biochem. (Paris), 1995, 33, 105. 228 M. Baucher, J. van Doorsselaere, J. Gielen, M. van Montagu, D. Inzé and W. Boerjan, Plant Physiol., 1995, 107, 285. 229 D. A.Somers, J. P. Nourse, J. M. Manners, S. Abrahams and J. M. Watson, Plant Physiol., 1995, 108, 1309. 230 A. Wagner, A. Walden and C. Walter, Plant Physiol., 1996, 112, 1397. 231 J. J. MacKay, W. Lu, R. Whetten, R. R. SederoV and D. M. O’Malley, Mol. Gen. Genet., 1995, 247, 537. 232 V. Lauvergeat, K. Kennedy, C. Feuillet, J. H. McKie, L. Gorrichon, M. Baltas, A. M. Boudet, J. Grima-Pettenati and K. T. Douglas, Biochemistry, 1995, 34, 12 426. 233 T. Hibino, K. Takabe, T. Kawazu, D.Shibata and T. Higuchi, Biosci. Biotech. Biochem., 1995, 59, 929. 234 D. P. Dharmawardhana, B. E. Ellis and J. E. Carlson, Plant Physiol., 1995, 107, 331. 235 N. Matsui, K. Fukushima, S. Yasuda and N. Terashima, Holzforschung, 1996, 50, 408. 236 W. R. Russell, A. R. Forrester, A. Chesson and M. J. Burkitt, Arch. Biochem. Biophys., 1996, 332, 357. 237 U. Takahama, Physiol. Plant., 1995, 93, 61. 238 U. Takahama, T. Oniki and H. Shimokawa, Plant Cell Physiol., 1996, 37, 499. 239 R.Whetten and R. SederoV, Plant Cell, 1995, 7, 1001. 240 A. M. Boudet, C. Lapierre and J. Grima-Pettenati, New Phytol., 1995, 129, 203. 241 M. M. Campbell and R. R. SederoV, Plant Physiol., 1996, 110, 3. 242 L. B. Davin and N. G. Lewis, An. Acad. Bras. Cienc., 1995, 67 (Suppl. 3), 363, Chem. Abstr., 1995, 125, 270 573. 243 A. T. Dinkova-Kostova, D. R. Gang, L. B. Davin, D. L. Bedgar, A. Chu and N. G. Lewis, J. Biol. Chem., 1996, 271, 29 473. 244 T. Umezawa and M. Shimada, Biosci.Biotech. Biochem., 1996, 60, 736. 245 M. Nose, M. A. Bernards, M. Furlan, J. Zajicek, T. L. Eberhardt and N. G. Lewis, Phytochemistry, 1995, 39, 71. 246 N. G. Lewis, M. J. Kato, N. Lopes and L. B. Davin, in Chemistry of the Amazon: Biodiversity, Natural Products and Environmental Issues, eds. P. R. Seidel, O. R. Gottlieb and M. A. C. Kaplan, ACS Symposium Series, Vol. 588, ACS, Washington, DC, 1995, p. 135. 247 N. C. J. E. Chesters, D. O’Hagan and R. J. Robins, J. Chem.Soc., Chem. Commun., 1995, 127. 248 M. Ansarin and J. G. Woolley, J. Chem. Soc., Perkin Trans. 1, 1995, 487. 249 N. C. J. E. Chesters, D. O’Hagan, R. J. Robins, A. Kastelle and H. G. Floss, J. Chem. Soc., Chem. Commun., 1995, 129. 250 N. C. J. E. Chesters, K. Walker, D. O’Hagan and H. G. Floss, J. Am. Chem. Soc., 1996, 118, 925. 251 R. J. Robins, N. C. J. E. Chesters, D. O’Hagan, A. J. Parr, N. J. Walton and J. G. Woolley, J. Chem. Soc., Perkin Trans. 1, 1995, 481. 252 J.Reichling, R. Martin and B. Kemmerer, Plant Cell, Tissue Organ Cult., 1995, 43, 131. 253 B. Kemmerer and J. Reichling, Phytochemistry, 1996, 42, 397. 254 J. Reichling, B. Kemmerer and H. Sauer-Guerth, Pharm. World Sci., 1995, 17, 113; Chem. Abstr., 1995, 123, 251 434. 255 P. A. Howles, T. Arioli and J. J. Weinman, Plant Physiol., 1995, 107, 1035. 256 O. Nakajima, M. Shibuya, T. Hakamatsuka, H. Noguchi, Y. Ebizuka and U. Sankawa, Biol. Pharm. Bull., 1996, 19, 71. 257 S. Akaba and S.K. Dube, Plant Mol. Biol., 1995, 29, 189. 258 J. H. Jeon, H. S. Kim, K. H. Choi, Y. H. Joung, H. Joung and S. M. Byun, Biosci. Biotech. Biochem., 1996, 60, 1907. 259 J.-H. Jeon, H.-S. Kim, K.-H. Choi, Y.-H. Joung, H. Joung and S.-M. Byun, Plant Physiol., 1996, 111, 348. 260 M. L. Durbin, G. H. Learn, G. A. Huttley and M. T. Clegg, Proc. Natl. Acad. Sci. USA, 1995, 92, 3338. 261 B. E. ScheZer, A. R. Reddy, I. HoVmann and U. Wienand, Plant Physiol., 1995, 109, 722. 262 A.R. Reddy, B. ScheZer, G. Madhuri, M. R. Srivastava, A. Kumar, P. V. Sathyanarayanan, S. Nair and M. Mohan, Plant Mol. Biol., 1996, 32, 735. 263 S. Fischer, U. Böttcher, S. Reuber, S. Anhalt and G. Weissenböck, Phytochemistry, 1995, 39, 1007. 264 C. Sallaud, J. El-Turk, C. Breda, D. BuVard, I. de Kozak, R. Esnault and A. Kondorosi, Plant Sci., 1995, 109, 179. 265 C. Sallaud, J. El-Turk, L. Bigarré, H. Sevin, R. Welle and R. Esnault, Plant Physiol., 1995, 108, 869. 266 G. M. Ballance and R.A. Dixon, Plant Physiol., 1995, 107, 1027. 267 T. Akashi, T. Furuno, K. Futami, M. Honda, T. Takahashi, R. Welle and S. Ayabe, Plant Physiol., 1996, 111, 347. 56 Natural Product Reports, 1998268 Y. Helariutta, P. Elomaa, M. Kotilainen, R. J. Griesbach, J. Schröder and T. H. Teeri, Plant Mol. Biol., 1995, 28, 47. 269 Y. Helariutta, M. Kotilainen, P. Elomaa, N. Kalkkinen, K. Bremer, T. H. Teeri and V. A. Albert, Proc. Natl. Acad. Sci. USA, 1996, 93, 9033. 270 Y.Terai, I. Fujii, S.-H. Byun, O. Nakajima, T. Hakamatsuka, Y. Ebizuka and U. Sankawa, Protein Expression Purif., 1996, 8, 183. 271 Q. Liu, M. S. Bonness, M. Liu, E. Seradge, R. A. Dixon and T. Mabry, Arch. Biochem. Biophys., 1995, 321, 397. 272 B. Botta, G. Delle Monache, M. C. De Rosa, R. Scurria, A. Vitali, V. Vinciguerra, P. Menendez and D. Misiti, Heterocycles, 1996, 43, 1415. 273 M. K. Pelletier and B. W. Shirley, Plant Physiol., 1995, 109, 1125. 274 M. K. Pelletier and B.W. Shirley, Plant Physiol., 1996, 111, 339. 275 C.-F. Liew, C.-J. Goh, C.-S. Loh and S.-H. Lim, Plant Physiol., 1995, 109, 339. 276 G. B. Deboo, M. C. Albertsen and L. P. Taylor, Plant J., 1995, 7, 703. 277 B. Charrier, C. Coronado, A. Kondorosi and P. Ratet, Plant Mol. Biol., 1995, 29, 773. 278 J. Dedio, H. Saedler and G. Forkmann, Theor. Appl. Genet., 1995, 90, 611. 279 H. Doostdar, J. P. Shapiro, R. Niedz, M. D. Burke, T. G. McCollum, R. E. McDonald and R. T. Mayer, Plant Cell Physiol., 1995, 36, 69. 280 Y. Tanaka, Y. Fukui, M. Fukuchi-Mizutani, T. A. Holton, E. Higgins and T. Kusumi, Plant Cell Physiol., 1995, 36, 1023. 281 Y. Tanaka, K. Yonekura, M. Fukuchi-Mizutani, Y. Fukui, H. Fujiwara, T. Ashikara and T. Kusumi, Plant Cell Physiol., 1996, 37, 711. 282 C. B. Do, F. Cormier and Y. Nicolas, Plant Sci., 1995, 112, 43. 283 P. K. Boss, C. Davies and S. P. Robinson, Plant Mol. Biol., 1996, 32, 565. 284 M. A. Berhow and D. Smolensky, Plant Sci., 1995, 112, 139. 285 T. Vogt and L. P. Taylor, Plant Physiol., 1995, 108, 903. 286 A. Rose, W. E. Gläßgen, W. Hopp and H. U. Seitz, Planta, 1996, 198, 397. 287 A. Callebaut, N. Terahara and M. Decleire, Plant Sci., 1996, 118, 109. 288 M.-A. Yamaguchi, S. Kawanobu, T. Maki and I. Ino, Phytochemistry, 1996, 42, 661. 289 M.-A. Yamaguchi, T. Maki, T. Ohishi and I. Ino, Phytochemistry, 1995, 39, 311. 290 S. Morimoto, T. Harioka and Y. Shoyama, Planta, 1995, 195, 535. 291 R. Rakwal, M. Hasegawa and O. Kodama, Biochem. Biophys. Res. Commun., 1996, 222, 732. 292 O. Benaventa-Garcia, J. Castillo, F. Sabater and J. A. Del Rio, Plant Physiol. Biochem. (Paris), 1995, 33, 263. 293 A. Gauthier, P. J. Gulick and R. K. Ibrahim, Plant Physiol., 1995, 108, 1341. 294 A. Gauthier, P. J. Gulick and R. K. Ibrahim, Plant Mol. Biol., 1996, 32, 1163. 295 T. A. Holton and E. C. Cornish, Plant Cell, 1995, 7, 1071. 296 L. Varin, F. Marsolais and N. Brisson, J. Biol. Chem., 1995, 270, 12 498. 297 F. Marsolais and L. Varin, J. Biol. Chem., 1995, 270, 30 458. 298 S. Ananvoranich, P. Gulick and R. K. Ibrahim, Plant Physiol., 1995, 107, 1019. 299 K. Kazuma, E. Shirai, M. Wada, K. Umeo, A. Sato, T. Matsumoto and T. Okuno, Biosci. Biotech. Biochem., 1995, 59, 1588. 300 K. Kazuma, K. Sato, T. Matsumoto and T. Okuno, Tennen Yuki Kagobutsu Toronkai Koen Yoshishu, 1996, 38, 7; Chem. Abstr., 1997, 126, 29 121. 301 K. Saito and K.-I. Miyakawa, Biologia (Bratislava), 1995, 50, 591; Chem. Abstr. 1996, 124, 284 406. 302 T. Nomura, Y. Hano and S. Ueno, Stud. Nat. Prod. Chem., 1995, 17, 451. 303 Y. Hano and T. Nomura, Tennen Yuki Kagobutsu Toronkai Koen Yoshishu, 1995, 37, 666; Chem. Abtsr., 1996, 124, 141 103. 304 L. Guo and N. C. Paiva, Arch. Biochem. Biophys., 1995, 320, 353. 305 S. Clemens and W. Barz, Phytochemistry, 1996, 41, 457. 306 S. Attucci, S. M. Aitken, R. K. Ibrahim and P. J. Gulick, Plant Physiol., 1996, 110, 1435. 307 G. DiCenzo and H. VanEtten, in Phytochemicals and Health, eds. D. L. Gustine and H. E. Flores, American Society for Plant Physiology, Rockville, Maryland, 1995, p. 263. 308 X.-Z. He and R. A. Dixon, Arch. Biochem. Biophys., 1996, 336, 121. 309 R. A. Dixon, M. J. Harrison and N. L. Paiva, Physiol. Plant., 1995, 93, 385. 310 S. Tropf, B. Kärcher, G. Schröder and J. Schröder, J. Biol. Chem., 1995, 270, 7922. 311 S. Raiber, G. Schröder and J. Schröder, FEBS Lett., 1995, 361, 299. 312 R. Preisig-Müller, P. Gnau and H. Kindl, Arch. Biochem. Biophys., 1995, 317, 201. 313 L. Beerhues, FEBS Lett., 1996, 383, 264. 314 O. Kwon, D. K. Bhattacharyya and R. Meganathan, J. Bacteriol., 1996, 178, 6778. 315 W. W. Poon, B. N. Marbois, K. F. Faull and C. F. Clarke, Arch. Biochem. Biophys., 1995, 320, 305. 316 B. N. Marbois and C. F. Clarke, J. Biol. Chem., 1996, 271, 2995. 317 A. Y. Hsu, W. W. Poon, J. A. Shepherd, D. C. Myles and C. F. Clarke, Biochemistry, 1996, 35, 9797. 318 J. A. Shepherd, W. W. Poon, D. C. Myles and C. F. Clarke, Tetrahedron Lett., 1996, 37, 2395. 319 G. Ranganathan and A. J. Mukkada, Ind. J. Parasitol., 1995, 25, 279; Chem. Abstr., 1995, 123, 79 190. 320 T. Okamoto, K. Yazaki and M. Tabata, Phytochemistry, 1995, 38, 83. 321 S. Gaisser and L. Heide, Phytochemistry, 1996, 41, 1065. 322 M. Tabata, Shokubutsu Soshiki Baiyo, 1996, 13, 117 (review in English); Chem. Abstr., 1996, 125, 243 054. 323 A. Stocker, H. Fretz, H. Frick, A. Rüttimann and W.-D. Woggan, Bioorg. Med. Chem., 1996, 4, 1129. 324 O. Sibbesen, B. Koch, B. A. Halkier and B. L. Møller, J. Biol. Chem., 1995, 270, 3506. 325 B. M. Koch, O. Sibbesen, B. A. Halkier, I. Svendsen and B. L. Møller, Arch. Biochem. Biophys., 1995, 323, 177. 326 B. A. Halkier, H. L. Nielsen, B. Koch and B. L. Møller, Arch. Biochem. Biophys., 1995, 322, 369. 327 M. Cicek and A. Esen, Plant Physiol., 1995, 109, 1497. 328 L. Zheng and J. E. Poulter, Plant Physiol., 1995, 109, 31. 329 N. Wajant, S. Förster, D. Selmar, F. EVenberger and K. Pfizenmaier, Plant Physiol., 1995, 109, 1231. 330 H. Lauble, S. Knödler, H. Schindelin, S. Förster, H. Wajant and F. EVenberger, Acta Crystallogr., Sect. D, 1996, 52, 887. 331 O. Sibbesen, B. M. Koch, P. Rouze, B. L. Møller and B. A. Halkier, in Amino Acids and their Derivatives in Higher Plants, ed. R. Wallsgrove, Cambridge University Press, Cambridge, 1995, p. 227. 332 A. Hickel, M. Hasslacher and H. Griengl, Physiol. Plant., 1996, 98, 891. 333 L. Du, J. Lykkesfeldt, C. E. Olsen and B. A. Halkier, Proc. Natl. Acad. Sci. USA, 1995, 92, 12505. 334 L. Du and B. A. Halkier, Plant Physiol., 1996, 111, 831. 335 R. N. Bennett, A. J. Hick, G. W. Dawson and R. M. Wallsgrove, Plant Physiol., 1995, 109, 299. 336 D. Toroser, H. GriYths, C. Wood and D. R. Thomas, J. Exp. Bot., 1995, 46, 1753. 337 S. Cottaz, B. Henrissat and H. Driguez, Biochemistry, 1996, 35, 15 256. 338 R. Iori, P. Rollin, H. Streicher, J. Thiem and S. Palmieri, FEBS Lett., 1996, 385, 87. 339 A. Falk, B. Ek and L. Rask, Plant Mol. Biol., 1995, 27, 863. 340 J. Xue, M. Jørgensen, U. Pihlgren and L. Rask, Plant Mol. Biol., 1995, 27, 911. 341 E. W. Jwanny, S. T. El-Sayed, M. M. Rashad, A. E. Mahmoud and N. M. Abdallah, Phytochemistry, 1995, 39, 1301. 342 R. M. Wallsgrove and R. N. Bennett, in Amino Acids and their Derivatives in Higher Plants, ed. R. Wallsgrove, Cambridge University Press, Cambridge, 1995, p. 243. 343 A. M. Bones and J. T. Rossiter, Physiol. Plant., 1996, 97, 194. 344 M. Butzenlechner, S. Thimet, K. Kempe, H. Kexel and H.-L. Schmidt, Phytochemistry, 1996, 43, 585. 345 K. Monde, K. Tamura and M. Takasugi, Phytochemistry, 1995, 39, 587. 346 J. Zhao and R. L. Last, Plant Cell, 1996, 8, 2235. 347 S. Rasmussen, G. Peters and H. Rudolph, Physiol. Plant., 1995, 95, 83. 348 S. Rasmussen, C. WolV and H. Rudolph, Phytochemistry, 1996, 42, 81. 349 W. Borejsza-Wysocki and G. Hrazdina, Plant Physiol., 1996, 110, 791. 350 K. Terada, C. Honda, S. Takeyama, K. Suwa and W. Kamisako, Biol. Pharm. Bull., 1995, 18, 1472. 351 K. Terada, K. Suwa, S. Takeyama, C. Honda and W. Kamisako, Biol. Pharm. Bull., 1996, 19, 748. Dewick: The biosynthesis of shikimate metabolites 57352 D. Hölscher and B. Schneider, J. Chem. Soc., Chem. Commun., 1995, 525. 353 D. Hölscher and B. Schneider, Nat. Prod. Lett., 1995, 7, 177. 354 G. M. Keseru and M. Nogradi, Stud. Nat. Prod. Chem., 1995, 17, 357. 355 S. J. Gould, M. J. Kirchmeier and R. E. LaFever, J. Am. Chem. Soc., 1996, 118, 7663. 356 J. R. Carney and K. L. Reinhart, J. Nat. Prod., 1995, 58, 971. 357 P. Wang, C. D. Denoya, M. R. Morgenstern, D. K. Skinner, K. K. Wallace, R. Digate, S. Patton, N. Banavali, G. Schuler, M. K. Speedie and K. A. Reynolds, J. Bacteriol., 1996, 178, 6873. 358 H. Motamedi, A. Shafiee, S.-J. Cai, S. L. Streicher, B. H. Arison and R. R. Miller, J. Bacteriol., 1996, 178, 5243. 359 T. Schwecke, J. F. Aparicio, I. Molnar, A. König, L. E. Khaw, S. F. Haydock, M. Oliynyk, P. CaVrey, J. Cortes, J. B. Lester, G. A. Böhm, J. Staunton and P. F. Leadlay, Proc. Natl. Acad. Sci. USA, 1995, 92, 7839. 360 A. Nakagawa and S. Omura, J. Antibiot., 1996, 49, 717. 361 C.-G. Kim, A. Kirschning, P. Bergon, P. Zhou, E. Su, B. Sauerbrei, S. Ning, Y. Ahn, M. Breuer, E. Leistner and H. G. Floss, J. Am. Chem. Soc., 1996, 118, 7486. 362 S. J. Gould, C. R. Melville and M. C. Cone, J. Am. Chem. Soc., 1996, 118, 9228. 363 L. A. Mueller, U. Hinz and J.-P. Zry�d, Phytochemistry, 1996, 42, 1511. 364 S. Heuer, T. Vogt, H. Böhm and D. Strack, Planta, 1996, 199, 244. 58 Natural Product Reports, 19