Figure 1. The polyacetate hypothesis. In general, in all figures, methyl groups are represented by a line without the addition of  CH₃. The figure shows the possible cyclizations of tetraacetic acid assumed to be formed by condensation of four acetyl units. Aldol type condensations,  –CH₂– + –CO–, are shown as A, leading to orsellinic acid [I], or 6-methylsalicylic acid (6-MSA) [II]. Acyl-type condensations, –CH₂– + –CO–OH, are shown as B, leading to phloroacetophenone [III]. Acyl cyclizations  –CO– + –CO–, are shown as C leading to a pyrone [IV] or as D leading to tetraacetic acid lactone [V]. Note that orsellinic acid formation requires only loss of one  H₂O, but 6-MSA formation requires the reduction,  –CO– –CHOH–, and loss of two  H₂O. With the possible exception of [IV] all structures are natural products. Phloroacetophenone [III] is an intermediate in usnic acid biosynthesis and its 2,4-dimethyl ether is xanthoxylin. The solid circles (•) represent  ¹⁴C introduced from [1-¹⁴C]acetate and for simplicity are only shown for 6-MSA (four labelled positions). With [1,3-¹⁴C₂]malonate the ring carbon attached to  CH₃ is not labelled so there are only three labelled positions. Similar labelling patterns (not shown) would be obtained for the other metabolites.

Figure 2. Diketide biosynthesis. AC = acetyl-CoA carboxylase, AT = acyltransferase, KS = ketosynthase, CoA = coenzyme A, ACP = acyl carrier protein. For description, see text. The diketide is oxobutyryl-ACP, CH₃–CO–CH₂–CO–S–ACP.

Figure 3. Orsellinic acid biosynthesis. The top line shows the further elongation of a diketide to a triketide by the mechanism of Figure 2. Further elongation leads to the tetraketide-ACP structure. Aromatization to orsellinic acid requires loss of a single  H₂O (see arrows) and a tautomeric rearrangement. ACP = acyl carrier protein.

Figure 4. Biosynthesis of 4-(2-butenyl)-4-methyl-l-threonine [VI]. KS = ketosynthase, KR = ketoreductase, DH = dehydratase, ER = enoyl reductase, SAM implies methylation from S-adenosylmethionine (SAM). The dotted and wedge shaped bonds are the usual stereochemical conventions for below and above the paper plane respectively. An 8-carbon chain is produced from one acetate and three malonate units by three ‘cycles’ of elongation; an additional carbon is added by C-methylation. The activities used in the three cycles are as follows: cycle 1, KS, KR, DH; cycle 2, KS, C-methylation, KR, DH, ER; Cycle 3, KS, KR. At some point the  C₂ amino group is introduced, presumably by transamination to a carbonyl structure.

Figure 5. Tylactone biosynthesis. The molecule shown is tylactone itself. A = acetate, B = butyrate, P = propionate; the carbon skeletons of these units are shown by thickened bonds with the carboxyl terminus indicated as •. To form the tylosin aglycone, the original  CH₃ from butyrate becomes CHO. The indicated  CH₃, from propionate becomes  CH₂OH, and adds a monosaccharide unit. The OH marked as adds a disaccharide unit.

Figure 6. Actinorhodin biosynthesis. In the description following: ARO = aromatase; CYC = cyclase; CLF = chain length factor; KS = ketosynthase; E = enzyme; PKS = polyketide synthase; ACP = acyl carrier protein; KR = ketoreductase. Genes and activities as follows: (1), actI, minimal PKS; three open reading frames in actI constitute the so-called ‘minimal PKS’ (KS, CLF, ACP) directing the first ring cyclization. Note the loss of seven CO₂ (not shown) and that (1) is assigned to two reaction steps. (2) = actIII, KR (a reducing agent is required, probably NADPH). (3) = act VII, ARO; loss of two  H₂O converts first ring to an aromatic structure. (4) = actIV, CYC; a second ring is formed. (5) = actVI, actVA, actVB; these genes produce ‘tailoring’ enzymes to complete formation of the dimeric actinorhodin molecule.

Figure 7. 6-Deoxyerythronolide B biosynthesis. DEBS = deoxyerythronolide B synthase, M = module, TE = terminal thioesterase; other abbreviations as before – see earlier figures. The acyl carrier protein (ACP) domain is represented by S for its thiol group. This metabolite is an example of a polypropionate structure. The process begins with the loading of a single propionyl unit and the figure shows the structure at the end of the reactions catalysed by the components of each module. In the final product, all of the  CH₃ and OH substituents have a defined chirality.