Non-ribosomal peptide synthetases (NRPSs) are responsible for the biosynthesis of many pharmaceutically relavant compounds. Type II NRPSs are an emerging subfamily of NRPSs that form hybrid pathways with type I fatty acid synthases (FAS), polyketide synthases (PKS), type I NRPSs, or others. The type II NRPSs commonly contain tailoring enzymes that generate unique substrate modifications, such as dehydrogenations and halogenation. Unlike type I NRPSs, the type II systems consists of standalone enzymes, an ideal feature for combintarial biosynthesis and metabolic engineering. Unfortunately, engineering efforts have been met with limited success due to lack of understanding of protein-protein interactions inherent to these pathways.

My dissertation work focuses on using structural biology to investigate type II pyrrole containing natural product pathways, specifically, the antifungal agent pyoluteorin and two prodiginine antitumor agents prodigiosin and undecylprodigiosin. Important to pyrrole formation are the peptidyl carrier protein (PCP) and the adenylation (A) domain. The PCP is post-translationally modified by a 4’-phosphopanetetheine group (holo-PCP) at a conserved serine residue, and the terminal thiol serves as the point of attachment for all NRPS intermediates. The A domain facilitates covalent attachment of a specific amino acid to the holo-PCP. The PCP then shuttles the cargo from one tailoring enzyme to the next in an organized fashion (Fig. 1) The two proteins are vital to precursor incorporation into pathways and substrate alteration. In many pathways (including pyoluteorin), a FADH2-dependent halogenase introduces chlorines to the pyrrole. Halogenation is essential for the biological activity of many natural products. Structural and chemo-enzymatic investigation of these three enzymes will aid in future engineering efforts in NRPS pathways.

In FAS and PKS pathways, the acyl carrier protein sequesters tethered substrates in a hydrophobic cleft between helix II and III for protection from undesirable reactions. Substrate sequestration in NRPS PCPs has not been demonstrated. To investigate the phenomena, we determined solution NMR structures of the type II PCP PltL, the peptidyl carrier protein from the pyoluteorin pathway (Fig. 2). Naturally, PCP and substrate are covalently attached through a thioester bond, a labile bond known to hydrolyze in aqueous environments. Chemoenzymatic methods were used to stabilize the pyrrolyl-PltL intermediate for protein NMR studies. The structures of both the holo-PltL and pyrrolyl-PltL intermediates were determined as the first functionally characterized type II PCP.

The recognition between PCP and A domain is specific in NRPSs. In fact, the homologous pairs from pyoluteorin and undecylprodigiosin pathways are only active with the cognate partner. We analyzed the homologous PCP and A domain from the prodigiosin pathway and, surprisingly, the PCP PigG was a promisicuous substrate for A domains from all three pathways. We decided to structurally investigate the specificity differences between the pyoluteorin PltL and prodigiosin PigG. The solution NMR structure of holo-PigG was determined and compared to the structure of holo-PltL. The structural features of the two proteins are similar, as expected due to the distinct pyrrole PCP family. Although, dynamic simulations revealed significantly more flexibility in holo-PigG. NMR titration experiments revealed the loop 1 region of both PCPs that was significantly perturbed when the A domain partners were introduced (Fig. 4). Mutations to the loop 1 region of PltL and PigG significantly altered the loading activity of the A domains compared to mutations in other regions. The mutant studies further confirmed the importance of loop 1 in PCPs.