Analysis of the enzymatic domains in the modular portion of the coronafacic acid polyketide synthase

Citation
S. Jiralerspong et al., Analysis of the enzymatic domains in the modular portion of the coronafacic acid polyketide synthase, GENE, 270(1-2), 2001, pp. 191-200
Citations number
29
Categorie Soggetti
Molecular Biology & Genetics
Journal title
GENE
ISSN journal
03781119 → ACNP
Volume
270
Issue
1-2
Year of publication
2001
Pages
191 - 200
Database
ISI
SICI code
0378-1119(20010530)270:1-2<191:AOTEDI>2.0.ZU;2-6
Abstract
Coronafacic acid (CFA) is the polyketide component of coronatine (COR), a p hytotoxin produced by the plant pathogen Pseudomonas syringae. The CFA poly ketide synthase (PKS) consists of two open reading frames (ORFs) that encod e type I multifunctional proteins and several ORFs that encode monofunction al proteins. Sequence comparisons of the modular portions of the CFA PKS wi th other prokaryotic, modular PKSs elucidated the boundaries of the domains that are involved in the individual stages of polyketide: assembly. The tw o beta -keroacyl:acyl carrier protein synthase (KS) domains in the modular portion of the CFA PKS exhibit a high degree of similarity to each other (5 3%), but are even more similar to the KS domains of DEBS, RAPS, and RIF. Cf a6 possesses two acyltransferases- AT0, which is associated with a loading domain, and AT1, which uses ethylmalonyl-CoA (eMCoA) as a substrate for cha in extension. Cfa7 contains an AT that uses malonyl-CoA as a substrate for chain extension, The Cfa6 AT0 shows 35 and 32% similarity to the DEBS1 and NidA1 AT0s, respectively, and 32 and 36% similarity to the Cfa6 and Cfa7 AT 1s. Sequence motifs have previously been identified that correlate with AT substrates. The motifs in Cfa6 AT1 were found to correlate reasonably well with those predicted for methylmalonyl-CoA (mMCoA) ATs, The motifs in the A T of Cfa7 correlated mon poorly with those predicted for MCoA ATs. Three AC P domains occur in the modular proteins of the COR PKS. The Loading domain- associated ACP0 showed 38% similarity to the loading domain ACP0s of DEBS1 and NidA1 and 32-36% similarity to the two module-associated ACPs of the CO R PKS, It exhibited a higher degree of similarity to the module-associated ACPs of RAPS. The two module-associated ACPs show 39% similarity to each ot her, but appear more closely related ro module-associated ACP domains in RA PS and RIFS. Furthermore, the DH and KR domains of Cfa6 and Cfa7 show great er similarity to DH and KR domains in RAPS and RIFS than to each other. The CFA PKS includes a thioesterase domain (TE I) that resides at the C termin us of Cfa7 and a second thioesterase, which exists as a separate ORF (Cfa9, a TE II). Analysis of a Cfa7 thioesterase mutant demonstrated that the TE domain is required for the production of CFA, The co-existence of TE domain s, within modular PKSs along with physically separated, monofunctional TEs (TE IIs) has been reported for a number of modular polyketide and non-ribos omal peptide synthases (NRPS). An analysis of the two types of thioesterase s using Clustal X yielded a dendrogram showing that TE Iis from PKSs and NR PSs are more closely related to each other than to domain TEs from either P KSs or NRPSs. Furthermore, rhs dendrogram indicates that both types of TE I Is are more closely related to TE domains associated with PKSs than to TE d omains in NRPSs. Finally, the overall % G + C content and the % G + C conte nt at the third codon for all of the PKS genes in the COR cluster suggest t hat these genes may have been recruited from a gram-positive bacterium, (C) 2001 Elsevier Science B,V. All rights reserved.