Natural products, such as fatty acids, polyketides, and non-ribosomal peptides, exhibit diverse biological functions toward human health and are constructed via many different pathways; however, these pathways often share the same synthetic logic. Central to these pathways is a carrier protein (CP). The role of a CP is to transfer elongating intermediates between catalytic domains and is mediated by protein-protein interactions between the CP and its partner enzymes. These interactions are transient, making it difficult to understand how they communicate with each other in recognizing a cargo or how the CP meets the right partner enzymes. The Burkart laboratory has been developing fluorescent and mechanistic probes to understand these CP-partner protein interactions.
The CP requires post-translational modification to become an active form through the action of 4’-phosphopantetheinyl transferase (PPTase) loading 4’phosphopantetheine prosthetic (PPant) arm at the serine residue. The first part of this dissertation focuses on utilizing the PPTase function to search for (1) the minimum peptide substrate required for CP recognition via machine learning and (2) the biosynthetic pathway from unculturable microorganisms. Using the ability of some PPTases to recognize and transfer an unnatural PPant arm to a CP, we identified orthogonal peptide substrates that can be labeled by two different classes of PPTases. These peptide substrates can be appended to a protein and used as a peptide tag. Furthermore, we employed a PPTase to fluorescently label CPs in previously uncultured microorganisms. This enabled us to sequence single cells for the identification of a biosynthetic cluster with active PPTase-CP pairs.
In the second part of this dissertation, we developed two mechanistic probes to study protein-protein interactions between epimerization (E) domain and peptidyl carrier protein in non-ribosomal peptide synthetases. D-amino acids are incorporated into non-ribosomal peptides, which contribute to their unique conformation and bioactivity. The E domains convert L- to D-amino acids by deprotonating/reprotonating Cα-H. Despite the past research on the E domain, the mechanistic details remain unclear. Herein, with the help of molecular dynamics simulations and mutational studies, our research reveals more detailed evidence on which catalytic residues work as an acid/base in this mechanism.