Biochemical evidence for an editing role of thioesterase II in the biosynthesis of the polyketide pikromycin

The pikromycin biosynthetic gene cluster contains the pikAV gene encoding a type II thioesterase (TEII). TEII is not responsible for polyketide termination and cyclization, and its biosynthetic role has been unclear. During polyketide biosynthesis, extender units such as methylmalonyl acyl carrier protein (ACP) may prematurely decarboxylate to generate the corresponding acyl-ACP, which cannot be used as a substrate in the condensing reaction by the corresponding ketosynthase domain, rendering the polyketide synthase module inactive. It has been proposed that TEII may serve as an "editing" enzyme and reactivate these modules by removing acyl moieties attached to ACP domains. Using a purified recombinant TEII we have tested this hypothesis by using in vitro enzyme assays and a range of acyl-ACP, malonyl-ACP, and methylmalonyl-ACP substrates derived from either PikAIII or the loading didomain of DEBS1 (6-deoxyerythronolide B synthase; AT(L)-ACP(L)). The pikromycin TEII exhibited high K-m values (> 100 muM) with all substrates and no apparent ACP specificity, catalyzing cleavage of methylmalonyl-ACP from both AT(L)-ACP(L) (k(cat)/K-m 3.3 +/- 1.1 M-1 s(-1)) and PikAIII (k(cat)/K-m 2.9 +/- 0.9 M-1 s(-1)). The TEII exhibited some acyl-group specificity, catalyzing hydrolysis of propionyl (k(cat)/K-m 15.8 +/- 1.8 M-1 s(-1)) and butyryl (k(cat)/K-m 17.5 +/- 2.1 M-1 s(-1)) derivatives of AT(L)-ACP(L) faster than acetyl (k(cat)/K-m 4.9 +/- 0.7 M-1 s(-1)), malonyl (k(cat)/K-m 3.9 +/- 0.5 M-1 s(-1)), or methylmalonyl derivatives. PikAIV containing a TEI domain catalyzed cleavage of propionyl derivative of ATL-ACPL at a dramatically lower rate than TEII. These results provide the first unequivocal in vitro evidence that TEII can hydrolyze acyl-ACP thioesters and a model for the action of TEII in which the enzyme remains primarily dissociated from the polyketide synthase, preferentially removing aberrant acyl-ACP species with long half-lives. The lack of rigorous substrate specificity for TEII may explain the surprising observation that high level expression of the protein in Streptomyces venezuelae leads to significant (>50%) titer decreases.