Annamarie Bustion

Pharmacogenomics research revealed that host genetic variants lead to interindividual variation in drug response. From research in the recently minted — yet nearly century-old — field of pharmacomicrobiomics, however, we now know that the bacteria residing within human hosts also affect a therapeutic’s pharmacokinetics. So far, over 200 FDA-approved compounds are known to be altered when exposed to the human gut microbiome, but only thirty of these reactions have been linked to a characterized gut-associated bacterial enzyme. As high-throughput experimental research continues, this imbalance between drugs identified relative to their metabolite and enzyme characterizations will likely widen. This project narrows that gap by employing chemoinformatic and metagenomic methods to computationally predict the bacterial enzymes responsible for previously reported gut microbiome drug metabolism events. Via reaction fingerprinting and chemical similarity calculations, microbiome drug-metabolism events lacking a characterized enzyme are compared to all reactions in MetaCyc, a curated database of bacterial enzyme-driven reactions. Using profile Hidden Markov Models, the MetaCyc gene sequence annotations of these most similar reactions are then searched against human gut microbiome gene catalogs and metagenome repositories to yield candidate enzymes capable of the drug-metabolism query. When queried with thirty drugs metabolized by characterized gut bacterial enzymes, the pipeline accurately predicts the positive control enzymes. Furthermore, when queried with the as yet uncharacterized bacterial metabolism of methotrexate into DAMPA and glutamate, the pipeline predicts enzymes found in isolates previously shown to metabolize the anti-arthritic compound. Now, enzyme purification experiments are being employed to further validate the method’s ability to predict previously unknown gut bacterial enzymes responsible for drug metabolism. In sum, this project leverages the wealth of recent experimental evidence of drug-metabolism in the gut microbiome and expands prior work by determining the specific enzymes responsible.