Biosynthesis of Polyketides

BY T. MONEY

1 Introduction

The polyketides are a structurally diverse group of naturally occurring compounds produced by the acyl–polymalonate biosynthetic route. A description of the essential features of this route (Scheme 1) was given in a previous Report which covered the literature to the end of 1972. The various compounds considered in this Report have been classified into groups (tetraketide, pentaketide, etc.) according to the number of C2-units (acetate and malonate) involved in their biosynthesis, and the literature has been covered for 1973 and 1974.

There has been a dramatic increase in the number of studies in which 13C-labelled acetate has been used as a precursor, including elegant and complementary variations based on the observation of 13C-13C couplings in products labelled alternatively from [1, 2-13C] acetate or from a 1:1 mixture of [1-13C]- and [2-13C]-acetate; examples are noted in this Chapter and are fully described in Chapter 7.

A new text dealing with the biosynthesis of acetate-derived compounds contains an excellent chapter on polyketide biosynthesis.

2 Aromatic Polyketides and Derivatives (Acetate a Chain-initiating Unit)

Tetraketides. — A full account of recent investigations on the biosynthesis of patulin (8) and co-metabolites describes further experimental evidence in support of the metabolic grid shown below. Using replacement cultures of Penicillium patulum and appropriate precursors labelled with 2H, 3H, and 14C it was established that the co-metabolites m-cresol, m-hydroxybenzyl alcohoL m-hydroxybenzaldehyde, and gentis-aldehyde (6) [but not gentisic acid (7)] are intermediates in the biosynthetic sequence leading from 6-methylsalicylic acid (1) to patulin (8). Toluquinol (3) is not an intermediate in the sequence but can be converted into desoxyepoxydon (4) by the fungus The mechanism of cleavage of the aromatic ring in late stages of patulin biosynthesis remains to be elucidated. (For 13C studies on the ring cleavage in penicillic acid biosynthesis, see p. 214).

Further studies on the purification of 6-methylsalicylic acid (6-MSA) synthase and its susceptibility to inhibitors have been reported. The results support the view that inhibition of enzymic activity by acetylenic thiol esters (e.g.3-pentynoyl-NAC* and 2-hexynoyl-NAC) is similar to that previously demonstrated for unsaturated fatty acid synthase in E. coli, namely specific inhibition of the dehydration step in the biosynthetic route (Scheme 2).

Additional studies have shown that concentrations of acetylenic inhibitor which completely inhibit 6-MSA synthesis still allow NADPH oxidation to continue at a significant rate. This result supports the proposal that acetylenic inhibitors act after the reduction step and that the latter process occurs on a 6-carbon rather than an 8-carbon intermediate.

In view of the important anti-cancer properties of mycophenolic acid (10; X = Me) considerable effort has been expended on the synthesis of analogues which could display enhanced or modified biological activity. A recent report has shown that the enzymic system of P. brevicompactum can convert the halogenated phthalides (9a) and (9b) into the corresponding analogues of mycophenolic acid.

Incorporation studies using intact Eupatorium rugosum plants have shown that the aromatic ring in dehydrotremetone (12) is derived from acetate and the furan ring from mevalonate. Further investigations using a cell-free homogenate of E. rugosum leaves demonstrated that isopentenyl pyrophosphate and tremetone (11) are efficient precursors. In contrast, dimethylallyl pyrophosphate and the acetophenones (13) and (14) were poorly utilized by the cell-free system and it has been suggested that isoprenylation of the enzyme-bound β-triketo-thiol ester intermediate occurs prior to aromatization.

Conflicting proposals concerning the biosynthesis of ring B in nidulin (15) and trisdechloronornidulin (16) (syn. yasmin, unguinol) have been resolved by studies which demonstrate that the incorporation of acetate and methionine is consistent with biosynthetic Scheme 3.

Pentaketides. — One of the first examples of the use of 13C – 13C coupling in polyketide biosynthesis is described in a recent report. [1,2-13C]Acetate was used as precursor and the new technique was used to elucidate the biosynthetic route to mollisin (17). Two possible pathways [paths a and b in Scheme 4] to this compound have previously been considerable, but the use of [1,2-13C]acetate has shown that neither is correct and that the biosynthesis of mollisin follows a third route [path c in Scheme 4].

Administration of [1-13C]acetate to Phialophora lagerbergii and the subsequent enhancement of signals due to C(1) C(3), C(4a) C(6) and C(8) in the 13C n.m.r. spectrum of skytalone (18) has demonstrated that this compound is derived by the linear condensation of five acetate units; 14 cf. p. 228.

Hexaketides. — Incorporation studies using[1-13C]-, [2-13C]-, and [1,2-13C]-acetate have shown that the biosynthesis of multicolic acid (21) and multicolosic acid (22) in P. multicolor involves the intermediate formation of 6-pentylresorcylic acid (19) followed by cleavage of the C(4) — C(5) bond. The 13C–13C couplings in the 13C n.m.r. spectrum of multicolic acid derived from doubly labelled acetate showed that intact C2 units were arranged as shown in formula (21), thus excluding the possibility of cleavage at C(1)–C(2) in (19) and the intermediacy of a symmetrical intermediate such as 5-pentylresorcinol (20); a fuller account of the n.m.r. aspects is given on p. 216.

Heptaketides. — The pattern of radioactivity in aloenin (23) after administration of [1-13C]acetate, [2-14C]malonate, and [methyl-14C]methionine to Aloe arborescens has been determined and the results are consistent with the expected biosynthetic route.

Octaketides. — The labelling pattern in asperentin 8-methyl ether (24) after administration of [2–14C]malonate to Aspergillus jlavus has shown that the C(6') — C(7') 'acetate-derived' portion of the molecule is significantly less radioactive than the malonate-derived two-carbon units. Observations of this type have been recorded previously and are to be expected if reconversion of malonate into acetate only occurs to a limited extent The 13C n.m.r. spectra of tajixanthone (29) (Aspergillus variecolor) derived from [1-13C]- and [2-13C]-acetate show enrichment patterns consistent with a biosynthetic pathway (Scheme 5) in which the xanthone nucleus is derived by ring cleavage of chrysophanol (26) or chrysophanol anthone (25). A mevalonoid origin for the C- and O-prenyl groups was also established during these investigations.

Although ring cleavage of anthraquinone intermediates seems to be an established biosynthetic process it has been suggested that in the biosynthesis of tajixanthone (29) the actual intermediate undergoing ring cleavage is the anthrone (25). The product (27) of this reaction has also been proposed as an intermediate in the bio-synthesis of arugosin A (30), B(31), and C(32), which co-occur with tajixanthone (29) in A. variecolor. For a full discussion, see p. 222.

Evidence for the direct involvement of emodinanthrone (33) in the biosynthesis of the range of anthraquinonoids [skyrin (36) iridoskyrin (37), (+)-rugulosin (38), and (-)-rubroskyrin (39)] of P. brunneum and P. islandicum has been reported. The low incorporation of emodin (34) has also led to the suggestion that this compound is not itself on the main biosynthetic route. Normally, the absence of oxygen functionality at the expected positions in acetate-derived phenolic compounds is thought to arise by reduction and dehydration of the β-polyketo-thiol ester intermediate. However, the efficient incorporation of emodinanthrone (33) into islandicin (35 iridoskyrin (37), and rubroskyrin (39) seems to indicate that in this system direct reduction of the aromatic ring can also occur (Scheme 6).

The biosynthesis of ochrephilone (40), a new metabolite of P. multicolor, has been elucidated by incorporation experiments using [1-13C]-, [2-13C]-, and [1,2-13C]- acetate as precursor. In the latter case 13C-13C coupling between C(1)-C(2), C(3)- C(4) etc. was observed, while a 1:1 mixture of [1-13C]- and [2-13C]-acetate resulted in coupling between C(2)-C(3), C(4)-C(5) etc. The results obtained (cf. p. 224) are consistent with the proposal that ochrephilone (40) is formed by the condensation of two acetate-derived chains and the introduction of a C1 unit (presumably from methionine) at C(4).

Decaketides. — Previous studies on the biosynthesis of the aflatoxins have shown that the entire carbon skeleton is derived from acetate. Recent investigations provide support for the proposal that averufin (41) and sterigmatocystin (42) are intermediates in the biosynthetic route (Scheme 7). Radioactive averufin (41) was prepared by adding [1-14C]acetate to a mutant of Aspergillus parasiticus which was impaired in its ability to produce aflatoxins. When administered to wild-type mycelium the efficient conversion of averufin (41) into aflatoxins B1(44), G1 (45 and B2 (15,16-dihydro-B1) was observed. 6-Hydroxydihydrosterigmatocystin (43)* has previously been implicated in the biosynthesis of aflatoxin B2 (15,16-dihydro-B 1) and G2 (15,16-dihydro G1) and a recent report has shown that sterigmatocystin (42) can be efficiently converted into aflatoxin B1(44). Related studies using [l,2-13C]acetate as precursor (see p. 214) have shown that sterigmatocystin (42) obtained from cultures of A. versicolor displays 13C–13C couplings consistent with the biosynthetic route shown in Scheme 7. A distinction between pathways (a) and (b) was made possible by the presence of 13C–13C-coupling between C(4)–C(5) and C(6) — C(7) (but see p. 247).

3 Aromatic Polyketides (Cinnamate as Chain-initiating Unit)

Polyketides whose biosynthesis involves the condensation of a C6 — C3 unit (cinna-mate and derivatives) with two or three acetate units are characteristic metabolites of higher plants and are rarely found in fungi. Recent tracer studies have, however, established both hispidin (47) (Polyporus hispidus) and chlorflavonin (48) (Aspergillus candidus) as true fungal metabolites. In the case of hispidin (47) the cumulative efforts of two research groups have shown that phenylalanine, tyrosine, cinnamate, p-coumarate, caffeate, and acetate are specifically incorporated into hispidin (47) by cultures of Pol yporus hispidus and Polyporus schweinitzii. Evidence for the presence of bis-noryangonin (46) in the mycelial extract was also obtained and an enzyme preparation was able to convert this compound into hispidin. One of the most interesting aspects of recent studies in this area is the discovery that light (380 and 440 nm) stimulates p-coumarate hydroxylase activity and styrylpyrone biosynthesis in cultures of P. hispidus. The suggestion has been made that hispidin could be derived by several alternative routes and these are summarized in Scheme 8. In a similar study the specific incorporation of (3-14C)-phenylalanine into chlorflavonin (48) and the co-metabolite (49) has clearly demonstrated that these compounds are biosynthesized de novo by Aspergillus candidus; there are structural parallels for the terphenyl (49) but compound (48) is probably the only authenticated flavonoid from a fungus.

The labelling patterns found in eucomin (51) after feeding aqueous solutions of radioactive phenylalanine, acetate, and methionine to the roots of Eucomis bicolor are consistent with a biosynthetic route involving a 2-methoxychalcone (50) intermediate. The involvement of the 2-methoxy-group in the formation of the hetero-cyclic ring in eucomin (51) is reminiscent of the reaction which converts isoflavonoids into rotenoids.

An authoritative account of recent major developments in isoflavone and rotenoid biosynthesis is provided in an excellent paper which specifically deals with the bio-synthesis of amorphigenin (52) by Amor pha fruticosa seedlings. Preliminary accounts of this work were published previously and an outline of the proposed biosynthetic scheme and the evidence supporting it has been given in a previous Report (see Vol. 2, p. 206).

Administration of [3-3H, 3-14C]cinnamic acid to the embryo fruit of Aesculus carnea results in the formation of (-)-epicatechin (55), procyanidin B-2 (58), and proanthyocyanidin A-2 (59) with radioactivity at the expected positions. However, the level of radioactivity in each ha1f of (58) and (59) was different, and this experimental finding has prompted new mechanistic proposals to explain the biosynthesis of these dimeric compounds.

A review describing the importance of α-hydroxychalcones in flavonoid biosynthesis has been published; flavonoid biosynthesis is dealt with by Harbome in Chapter 4.

4 Non-aromatic Polyketides*

The first illustration of the use of [1,2-13C]-labelled acetate in polyketide biosyn-thesis was provided in a report in which the known penta-acetate origin of dihydrolatumcidin (60) was confirmed. In accordance with the established bio-synthetic route, coupling was observed between C(2) — C(3), C(4) — C(4a), C(5) — C(6), C(7) — C(7a), and C(8) — C(9). Complementary results were obtained when a 1:1 mixture of [1-13C]- and [2-13C]-acetate was used as precursor. In this case only one of the four combinations of acetate units (... CH213CO ... 13CH2CO ...) gives the observed coupling, which is between C(3) — C(4), C(4a) — C(7a), C(5) — C(8), and C(6) — C(7).

Incorporation studies using 14C- and 13C-labelled acetate have shown that nigrifactin (65) is produced in Streptomyces nigrifaciens by linear condensation of six acetate units (presumably acetate + 5 malonates). However, as with other acetate-derived piperidine alkaloids. the biosynthesis of nigrifactin (65) presumably involves oxidation of the appropriate straight-chain saturated fatty acid. Support for this biosynthetic route (Scheme 9) has been provided by the specific incorporation of 5-oxododecanal (62), 2-n-heptylpiperideine (63), and 2-n-heptylpiperidine (64) into nigrifactin (65). A similar biosynthetic route (Scheme 10) has been established for avenaciolide (69), a metabolite of Aspergillus avenaceus. In this case the postulated dodecanoyl intermediate (66) is presumably converted into a 3-oxododecanoyl derivative (67) which subsequently condenses with succinate. The proposed scheme (Scheme 10) is based on experiments in which the precursor activity of [1-13C]- acetate and [2-13C]acetate was separately determined However, it should be noted that there is no direct evidence for the intermediacy of the postulated intermediates (67) and (68).

Cytochalasin B (phomin) (74) and cytochalasin D (zygosporin A) (75) are structurally related compounds isolated from Phoma sp. and Zygosporium masonii respectively. Incomplete results obtained from studies using 14C-labelled precursors led to the proposed biosynthetic scheme outlined in Scheme 11. Complementary results using 13C-labelled acetate have provided strong support for these proposals and have added information on the biosynthetic origin of regions in the structures [e.g. C(1), C(9)] which were inaccessible by degradation.

The biosynthesis of thermozymocidin (76) from nine acetate units and serine has been confirmed by recent incorporation experiments using [1-13C]acetate. Analyses of the 13C n.m.r. spectra of prodigiosin (79) after incorporation of 13C-labelled acetate, alanine, proline, methionine, and serine have resulted in the proposed biosynthetic scheme shown in Scheme 12 (For similar data on some pro-digiosin analogues, see p. 211.)