Discovery and development of small molecule inhibitors of anaerobic microbial choline metabolism as potential therapeutics

Abstract

Anaerobic metabolism of dietary choline to trimethylamine (TMA) by the human gut microbiota is a disease-associated pathway. The host’s impaired ability to oxidize TMA into the downstream metabolite trimethylamine-N-oxide (TMAO) results in trimethylaminuria, while TMAO is positively correlated with various metabolic diseases. Small molecule inhibition of bacterial choline metabolism has been shown to attenuate the development of chronic kidney disease and cardiovascular disease in mice, highlighting the therapeutic potential of modulating this pathway. Inhibitors previously developed to target this pathway are often substrate mimics of choline trimethylamine-lyase (CutC), the key enzyme involved in this transformation that is encoded within the choline utilization (cut) gene cluster. The close structural similarity of these inhibitors to choline raises potential safety concerns of off-target inhibition of host choline and acetylcholine receptors. This dissertation describes our efforts to discover and develop structurally distinct inhibitors of anaerobic microbial choline metabolism as potential therapeutics, by utilizing a combination of high-throughput screening (HTS), medicinal chemistry, quantitative proteomics, and in vivo experiments. Chapter 2 describes the identification and optimization of structurally distinct inhibitors of anaerobic microbial choline metabolism, as well as preliminary investigations into their mode of action. Leveraging on our understanding of anaerobic choline metabolism in bacteria, we designed a growth-based phenotypic HTS that successfully identified two potential hits for further optimization. We used medicinal chemistry to improve upon the broad-spectrum TMA inhibitory activity and potency of the inhibitors, and reduced their antibacterial activity. Based on previous genetic studies of choline-metabolizing bacteria, we identified three potential modes of action in which inhibitors could exert their effects. After prioritizing the inhibitors based on their potency and toxicity, we assessed their ability to inhibit CutC activity in vitro, ability to repress the expression of the cut gene cluster, and their ability to inhibit choline uptake into the cell. These targeted mode of action assays pointed to a more complex mechanism of action for the inhibitors. We then turned to a more unbiased approach to identify the targets of our inhibitors. In Chapter 3, we utilized knowledge from our inhibitor optimization efforts to design photo-crosslinking probes that we then used for photo-affinity labeling and quantitative proteomics. These probes labeled specific proteins in the proteome of Escherichia coli MS 200-1, a choline-metabolizing human gut isolate. Identification of these proteins by mass spectrometry revealed that they were homologs of outer membrane protein PagN, an adhesion/invasion protein involved in the virulence of Salmonella. Validation of the protein hit by genetic knockout suggested that these proteins, though binders of the photo-crosslinking probe, were likely not involved in anaerobic choline metabolism. These data suggests that the biologically relevant target may only be present in low abundance or is inaccessible to our probe in a cellular setting. Finally, in Chapter 4, we evaluated the most promising inhibitors in vivo. In collaboration with the Rey Lab (University of Wisconsin–Madison), we confirmed that our inhibitor was able to reduce serum TMAO levels in both a gnotobiotic and conventional mouse model. We also characterized the shifts in gut microbial community composition of the gnotobiotic mice before and after treatment with our inhibitor and find that there was no significant change. This could imply that our inhibitor can selectively inhibit anaerobic choline metabolism without broader impacts on the gut microbiome. Lastly, we designed a panel of second-generation inhibitors aimed at increasing gut retention. Preliminary in vivo experiments with these inhibitors suggests that we may be able to modulate the bioavailability of the inhibitors by changing their physicochemical properties. Future studies will be focused on improving in vivo activity of these inhibitors and elucidating their mechanism of action in the gut microbiome.