Organization of biosynthetic gene cluster for avermectin in Streptomyces avermitilis: analysis of enzymatic domains in four polyketide synthases.

The analysis of the incorporation of 13C-labeled precursors into avermectins indicates that the avermectin aglycons are synthesized by head-to-tail condensation of various acyl groups, which is similar to the biosynthesis of other polyketides. Polyketide synthases (PKS) use the appropriate CoA ester as a primer and add acetate units from malonyl-CoA and propionate units from methylmalonyl-CoA to assemble the polyketides. Avermectin aglycons are formed by addition to the starter unit (2-methylbutyrate or isobutyrate) of 12 acyl condensations in the order P-A-A-A-A-P-P-A-P-A-P-A (P, propionyl; A, acetyl). Within the 90-kb gene cluster for a vermectin biosynthesis, the central 65-kb segment was found to be required for aglycon biosynthesis by phenotypic analysis of strains containing deletion or insertion mutations in this region. A complete sequence analysis of the 65-kb segment indicated that this segment encodes avermectin PKS. The avermectin PKS genes are organized into two converging blocks of ORFs. From the results of sequencing analysis, a feature of the two regions, aveA1/aveA2 and avea3/aveA4, is that they encode four kinds of large multifunctional polypeptides containing 55 domains which possess putative fatty acid synthase-like activities. The avermectin PKS (AVES 1-4) appear to contain two, three, or four modules. AVES1 and 2 contain two and four modules, respectively, whereas AVES 3 and AVES 4 each contains three modules. The 12 modules correspond to the 12 cycles required for synthesis of the avermectin aglycon.