The Biosynthesis of Polyketides

BY T.J. SIMPSON

1 Introduction

This chapter covers the literature appearing between January 1979 and December 1981 and follows the format of the previous report. It has been a particularly active period with a welcome, increasing trend towards studies aimed at elucidating the mechanisms of the intermediate steps in polyketide biosynthesis. The potential of 2H-labelling mentioned previously has been Yrealised and extended in the review period with several studies using both direct methods, ie 2H n.m.r., and indirect methods, ie 2H α-isotope shifts and 2H-13C couplings in 13C n.m.r. using doubly labelled [2H,13C] precursors. These methods have been reviewed. A potentially more useful technique than the α-isotope shift method has been described. In this, 2H is placed β to the reporter 13C nucleus in a doubly labelled precursor; an isotope shift is still observable but the unfavourable relaxation and nOe effects associated with 2H directly attached to 13C are avoided. This has been applied to only one study in the review period, but will clearly find much use. A related and also extremely useful technique makes use of 18O-induced isotope shifts in 13C n.m.r. to detect the biosynthetic origins of oxygen by incorporating doubly labelled [18O,13C precursors or by growing organisms in an atmosphere containing 18O2 and subsequent 13C n.m.r. analysis of the labelled metabolites. The number of studies using advanced intermediates continues to increase and 2H has great potential in this area. A number of books which cover aspects of polyketide biosynthesis have appeared, with Steyn's book on the biosynthesis of mycotoxins being particularly valuable.

2 Fatty Acids

The stereochemical mechanism of enoyl reductase, the enzyme catalysing the final reduction in the cycle of condensation-reduction-dehydration-reduction that lengthens the fatty acid chain by one -CH2CH2- unit at each turn of the cycle on fatty acid synthetase, appears to be species specific. As shown in Scheme 1, the enoyl reductase from yeast converts the enoyl thioester (1) to the acyl thioester (2) by an anti-addition of hydride from NADPH to the si face of the β-carbon with protonation of the α-carbon from thsi face. However with the reductase from both E. coli and Brevicterium ammoniagenes, a syn-addition of hydrogen via a 2-re, 3-si attack occurs, whereas the reductase from rat liver also carries out a-addition, but this time via 2-si, 3-re attack. The stereochemistry of hydride donation from NADPH is related to the stereochemistry of addition, with the pro-4S hydrogen being used for 3-si addition and the pro-4R hydrogen being used for 3-re addition.

Incorporation studies with [2-13C, 2H3] acetate and analysis of the 2H α-isotope shifts in the simultaneously 1H and 2H noise decoupled 13C n.m.r. spectrum has shown that palmitic acid (3) is biosynthesised in the alga Anacystis nidulans with a gradation of 2H retention along the acyl chain as shown in Scheme 2. The results are interpreted as being consistent with a 'post-malonate' exchange process, presumably associated with reversible transfer of the growing acyl chain from the acyl carrier protein to a cysteine residue of β-keto acyl ACP synthetase. Similar results have been obtained using [2H3] acetate in E. coli.

The incorporation of 2H from [2H3] acetate into lipoic acid (4) is consistent with its formation from octanoic acid with the loss of one 2H label from C-8. The 2H incorporated at C-6 of octanoic acid is retained, and since this 2H is incorporated with the L-configuration during fatty acid biosynthesis but is known to have the D-configuration in lipoic acid, an inversion of configuration must occur at C-6 during sulphur insertion. This suggested the involvement of hydroxylated octanoic acids as intermediates. However, feeding studies with 2H-labelled 6-hydroxy-, 8-hydroxy-, and 6,8-dihydroxyoctanoic acids gave negligible incorporations and so direct introduction of sulphur at the saturated carbons of octanoic acid seems likely.

An authoritative review of the structure of fatty acid synthetase has appeared.

3 Tetraketides

Both 2H n.m.r. spectroscopy and β-isotope shifts in 13C n.m.r. have been used to measure the incorporation of 2H from [2H3]- and [1-13C, 2H3]acetates into 6-methyl-salicylic acid (5) by Penicillium griseofulvum. Both methods show that there is a preferential in corporation into the methyl of the acetyl-CoA derived starter unit and significantly more 2H is retained at C-3 than at C-5. It is suggested that the non-uniform incorporation could arise from differing degrees of random exchange during the chain assembly process or, more interestingly, it could reflect the actual mechanisms of cyclisation and aromatisation of the precursor polyketide. More examples will be needed to test the validity of this observation.

Addition of 5-chloroorsellinic acid to growing cultures of Penicillium cyclopium inhibits the biosynthesis of penicillic acid (8) and results in the accumulation of the previously indicated intermediates orsellinic acid (6) and 3-methoxytoluquinol (7) and its corresponding quinone.

The post-gentisaldehyde part of the biosynthetic pathway to patulin (13) has been extensively investigated, using mutant strains of Penicillium urticae. A patulin-minus mutant, J1, accumulates phyllostine (10) and isoepoxydon (11). Another patulin-minus mutant, J2, which is blocked immediately after gentisaldehyde (9), converts both (10) and (11) to patulin in yields of 90% and 60% respectively. They are interconverted by a specific dehydrogenase and cell-free systems have been isolated from the mutants which carry out their interconversion and further transformation. A further mutant, S15, accumulates isopatulin (12) and immobilised cells of the wild-type strain convert phyllostine to isopatulin in good yield. Finally, a cell suspension of mutant P3, blocked between phyllostine and isopatulin, converts isopatulin to patulin. These results are summarised in Scheme 3. The changes in oxidation levels occurring during these transformations are somewhat puzzling and labelling studies would be useful. This and earlier works are summarised in a review which also compares patulin biosynthesis with the pathways leading to a number of other fungal lactones.

Elasnin (14), a novel inhibitor of human granulocyte elastase, has been isolated from Streptornyces noboritoensis. Incorporation of [13C2] acetate has shown it to be derived from twelve acetates but it would appear to be best regarded as a tetraketide, as the most plausible route is extension of a hexanoate starter by three 2-butylmalonate units as indicated in Scheme 4.

[13C2] Acetate is incorporated into zinniol (15) by cultures of Alternaria solani as shown.

4 Pentaketides

Incorporation studies with singly and doubly labelled [13C] acetates have confirmed that the dihydroisocournarin moiety of the important mycotoxin ochratoxin A (16) has a regular...