Source: Dissertation Abstracts International, Volume: 79-05(E), Section: B.

Adviser: Jason Michael Crawford.

Sequencing of bacterial genomes has revealed that many species have the potential to produce more natural products than previously assumed. Despite this, however, over the past few decades, the number of new drugs arising from natural products has been in decline. This inability to translate biosynthetic potential into novel structures has been due to a variety of technical issues, including the inability to activate `'silent" gene clusters in the laboratory as well as the inability to dereplicate extracts and identify novel compounds for characterization. To address these issues, reverse genetic approaches using heterologous expression systems have sought to activate silent gene clusters in the laboratory, while new advances in mass spectrometry have assisted in rapidly identifying molecules that have been previously characterized or are derivatives of known molecules. The work described here ties together genetic and metabolomics techniques to identify molecules produced by the colibactin pathway. The colibactin pathway is a hybrid non-ribosomal peptide--polyketide synthase (NRPS--PKS) found in strains of E. coli that colonize the gut and in other related organisms. The pathway produces a family of genotoxic small molecules that have DNA alkylation activity, which in colitis mouse models results in colon cancer. The structures of various pathway intermediates have been characterized; however, a molecule that requires the function of every gene in the pathway has yet to be identified.

Chapter 1 uses the colibactin pathway as an example to discuss research focused on linking biosynthetic gene clusters to their metabolites. This chapter focuses on mass spectrometry approaches used to characterize new metabolites, particularly focusing on the recent development of pathway-targeted metabolomics. Other natural products that are produced through analogous biosynthetic logic and some current disconnects between bioinformatics predictions and experimental structural characterizations are also highlighted.

Chapter 2 describes the pathway-targeted analysis of the colibactin pathway. The metabolic analysis led to the isolation and full structural characterization of three precolibactin metabolites, as well as allowed us to propose the structures of six additional metabolites based off of these structures. Additionally, the two of these metabolites were synthesized for activity assays and to determine the stereochemistry of one of the natural metabolites.

Chapter 3 investigates the biosynthesis of previously characterized precolibactin metabolites. Individual domains of the biosynthetic enzymes in the colibactin pathway were inactivated with point mutations. A comparative analysis between the metabolomes of mutants and control strains enabled the characterization of "multidomain signatures," or functional readouts of NRPS-PKS domain contributions to the pathway-dependent metabolome. Multidomain signatures provided experimental support for individual domain contributions to colibactin biosynthesis and delineated the assembly line timing events of colibactin heterocycle formation. Additionally, the analysis led to the structural characterization of two reactive precolibactin metabolites. We demonstrate the fate of these reactive intermediates in the presence and absence of CIbP, which dictates the formation of distinct product groups resulting from alternative cyclization cascades.

Chapters 4-6 contains preliminary and unpublished work. Chapter 4 discusses the experiments to discover metabolites that require the function of every enzyme in the colibactin pathway for production. Specific focus is put on the pathway-dependent ions that require the function of the PKS CIbO and annotated peptidase C1bL. Comparative metabolomics between the mutant and control strains revealed the presence of a handful of metabolites that require both genes for production. Isotopic labeling with amino acids known to be incorporated the colibactin pathway provide evidence for a dimeric structure of these metabolites.

Chapter 5 details experiments performed to investigate the biosynthesis of 1aminocyclopropane carboxylate (ACC) in the colibactin pathway. ACC is produced by the NRPS CIbH. Proteomics attempts to characterize intermediates covalently attached to the thiolation domain of CIbH were unsuccessful, as the constructs produced in this experiment were inactive. However, an alternative approach is proposed to guide future experiments.

Chapter 6 describes preliminary investigations into the biosynthesis of the luciferin coelenterazine in the ctenophore Mnemiopsis leidyi. An efficient extraction procedure to detect coelenterazine from a single organism was developed. Transcriptomics datasets were mined for transcripts of small peptides that could be precursors to coelenterazine. Two peptides were identified that fit the criteria set to identify these peptides as putative ribosomally produced precursors to coelenterazine. Future experiments are proposed to investigate these hypothetical precursors.