The development and applications of NMR metabolomics analysis of bacterial metabolomes

Metabolomics is a relatively new field that involves the study of metabolic responses that are occurring within a biological system. Metabolite profiles of an organism, tissue extract, and biofluids are important indicators to determine the physiological state of a biological profile. Comparison of such profiles from different phenotypes can be used to identify specific metabolic changes leading to the understanding of metabolic pathways, disease progression, drug toxicity and efficacy, and cellular responses to different intracellular and extracellular conditions. Metabolomics investigations often use sophisticated analytical techniques such as NMR spectroscopy to provide an unbiased and comprehensive approach to evaluate metabolic perturbation in different cell lysates. This dissertation will focus on the development and applications of NMR-based metabolomics methodologies to generate reliable and reproducible results. The protocol has been expanded greatly, optimizing all aspects of the metabolomics process including cell growth, sample preparation, sample handling, data collection, data processing, and data analysis. There are two main approaches in the protocol to decipher NMR metabolomics data: pattern recognition, such as PCA and OPLS-DA, comparing numerous 1-dimensional 1H NMR datasets to analyze biofluids at a global scale, and quantitative profiling of 13C-labeled metabolites using 2-dimensional 1H- 13C HSQC. As a result, our protocol provides a comprehensive analysis, describing unique characteristics and relationships between various samples that differ in their source or treatment. We applied our protocol to predict the in vivo mechanism of action for drug leads from NMR metabolomics data. The NMR analysis resulted in distinct clustering which would be classified by an in vivo mechanism. Also, we demonstrated the similarity of Staphylococcus epidermidis metabolomes resulting from exposure to divergent environmental stressors that are known to facilitate biofilm formation. Our results suggest that the tricarboxlic acid cycle acts as a metabolic signaling mechanism for the activation of biofilm formation. Also investigated was the mechanism of action of D-cycloserine in M. smegmatis and M. tuberculosis. Our findings proved that D-alanine-D-alanine ligase is the primary target as cell growth is inhibited when the production of D-alanyl-D-alanine is halted. Furthermore, we were able to identify an alternate path for the production of D-alanine via a possible transaminase mechanism