Career: Investigation and Redesign of Nonribosomal Peptide Synthetases

The engineering of natural product biosynthetic pathways is a promising direction for the development and production of new structurally diverse biologically active compounds. However, the lack of mechanistic information on many enzymes of complex biosynthetic pathways continues to hinder our ability to manipulate these machineries. To address this fundamental problem, we propose to study the thiocoraline biosynthetic pathway through in vitro reconstitution and manipulation of its individual steps. Thiocoraline is a bisintercalator natural product composed of two unusual heteroaromatic units, the 3-hydroxyquinaldic acids (3HQA) that are appended to a nonribosomally biosynthesized peptidic core.1, 2 Quinaldic, indolic, and quinoxaline-2-carboxylic appendages (Fig. 1A) that are derived from L-Trp are found in various natural products, including bisintercalators3 and thiopeptide antibiotics.4 In Aim 1, we propose to delineate the sequential enzymatic steps required to convert L-Trp into 3HQA (Fig. 1B). The detailed biochemical characterization of the unique enzymatic transformations of this pathway will (i) make important contribution to decipher novel mechanistic features, and (ii) establish the groundwork for the genetic manipulation of L-Trp derivatization in other systems that could be used in producing diverse new compounds through combinatorial biosynthesis. The peptidic core of nonribosomal peptide (NRP) natural products such as thiocoraline is biosynthesized by large multifunctional enzyme complexes, the nonribosomal peptide synthetases (NRPSs). During NRP formation, the highly substrate specific NRPS adenylation (A) domains dictate the identity of the proteogenic and non-proteogenic amino acid building blocks to be incorporated into the growing natural product. Recently, it was shown that the small MbtH-like proteins5 found in hundreds of bacterial NRPS gene clusters contribute to stimulation of amino acid adenylation by A domains that are otherwise inefficient.6-8 However, the mechanism of this stimulation is still unknown. Unusual A domains interrupted by insertion of part of methyltransferase, oxidase, or ketoreducase enzymes have also been reported and proposed to be involved in NRP biosynthesis, but only one of them has been briefly biochemically investigated. Thus, to develop A domains as biosynthetic tools (Aim 2) we need to (a) understand the mechanism of the stimulation of their activity by MbtH-like proteins and explore the potential of cognate and noncognate MbtH-like protein/A domain pairs, (b) gain insight into the function of interrupted A domains, and (c) develop A domains with broad or altered substrate specificity. Aim 1: In vitro reconstitution of the 3HQA biosynthetic pathway. There were originally two proposed routes for the biosynthesis of 3HQA from L-Trp.3 Using both biochemical and gene deletion studies, we recently obtained sufficient evidence to suggest that 3HQA arises from a series of rearrangements carried out by eight enzymes (TioK, T, I, Q, F, L or M, G, and H) (Fig. 1B). Briefly, in the proposed pathway, L-Trp is first converted to â-hydroxykynurenine by consecutive action of the TioK/TioT pair of enzymes, TioI, TioQ, TioF, and TioL or TioM. The transformation of â-hydroxykynurenine into 3HQA is then proposed to occur by cyclization by TioG and final elimination of the 4-hydroxy moiety by TioH. We have already preliminarily biochemically characterized the TioF,9 the T domain of TioK, TioQ, and TioP enzymes and showed that TioP is likely not involved in 3HQA production.10 We now propose to biochemically study the remaining enzymes in the pathway: the cytochrome P450 TioI, the potential deformylase TioL, the kynurenine aminotransferase TioG, and the oxidoreductase TioH. Aim 2: Biochemical, engineering, and structural studies of adenylation (A) domains and their cognate MbtH-like proteins. Our specific objectives are: (a) biochemical and structural studies of cognate and non-cognate MbtH-like protein/A domain pairs, (b) biochemical characterization of interrupted A domains, and (c) engineering of A domain with relaxed or altered substrate specificity. Our biochemical, engineering, and structural studies will clarify the key structural features required for A domain activity and specificity. The studies of this proposal will likely lead to new tools for rational engineering and manipulation of NRPS assembly-lines.