Analysis of the cryptic biosynthetic gene cluster encoding the RiPP curacozole reveals a phenylalanine-specific peptide hydroxylase

Abstract

Curacozole belongs to an unusual cyanobactin-like sub-family of ribosomally synthesized and post-translationally modified peptides (RiPPs). The molecule is distinguished by its small, macrocyclic structure, a poly-azole sequence that includes a phenyloxazole moiety, and a D-allo-Ile residue. The enzymatic steps required for its formation are not well understood. The predicted biosynthetic gene cluster (BGC) for curacozole in Streptomyces curacoi is cryptic, but is shown to be potently activated upon constitutive expression of the bldA specified Leu-tRNA(UUA) molecule. Heterologous expression and gene deletion studies have defined the minimum BGC as consisting of seven genes, czlA, D, E, B1, C1, F, and BC. The biosynthetic pathway is highly substrate tolerant, accepting six variants of the precursor peptide CzlA to form new curacozole derivatives. This includes replacing the phenyloxazole moiety of curacozole with indole and p-hydroxyphenyloxazole groups by conversion of the corresponding CzlA Phe18Trp and Phe18Tyr variants. In vitro experiments with purified enzymes demonstrate that CzlD and CzlBC perform cyclodehydration and dehydrogenation reactions, respectively, to form a single oxazole from Ser 22 of CzlA. The curacozole BGC is flanked czlI, a non-essential but conserved gene of unknown function. In vitro studies demonstrate CzlI to be a non-heme iron / 2-oxoglutarate dependent dioxygenase, catalyzing the hydroxylation of Phe18 on CzlA to form the CzlA Phe18Tyr variant, which is then processed to form the p-hydroxyphenyloxazole derivative of curacozole. Overall, this work highlights the amenability of RiPP biosynthesis for engineering the production of new compounds and adds to the repertoire of known RiPP enzymes.