(295g) Characterizing and Predicting Xenobiotic Transformations Through Human and Bacterial Cyps

Ingested environmental chemicals can be metabolized by enzymes expressed in the human host as well as the microbiota of the gastrointestinal (GI) tract, leading to derivatives that are potentially more bioactive.  In humans and other mammals, a superfamily of oxidoreductases, known as cytochrome P450 (CYP) enzymes, comprise the major class of enzymes responsible for the transformation of nonnative (xenobiotic) chemicals.  As the bacteria of the GI tract microbiota also express CYP enzymes, the number of CYP enzymes that could be active in the human body is large. To date, more than 11,500 distinct CYP proteins have been identified.  

Metabolomics has emerged as a powerful approach to explore the metabolite profiles of the GI tract, and to compare these profiles under different host conditions and perturbations.   To date, most studies have utilized an untargeted approach to obtain a comprehensive profile of the altered metabolites. While this approach has discovery potential, pure experimental exploration of the fate of an ingested xenobiotic chemical is impractical due to the large range of metabolites that could be derived from the chemical and the sensitivity limits associated with simultaneous analysis of very many compounds.  On other hand, targeted methods require a priori knowledge of the metabolites to be analyzed, limiting the potential to identify new compounds. In this regard, computational identification of possible CYP transformations of xenobiotic chemicals could upgrade the discovery potential of targeted metabolomics, while affording sensitive detection and quantitative analysis.  One challenge in computationally identifying transformation outcomes is the flexibility of CYP enzymes. A given CYP enzyme can convert multiple substrates to different products.  Moreover, the currently available catalogue of CYP enzymes is dispersed across multiple databases (e.g. KEGG and Uniprot). Moreover, these databases have limited coverage of environmental chemicals and their derivatives.

In this work, we characterize CYP transformations for both human and bacteria and provide a comparative summary.   Additionally, we present a method, PxT, for predicting how xenobiotics are transformed using both human and microbial CYPs.  The PxT method consists of two steps that identify how a metabolite is transformed via xenobiotic transformation enzymes of the host.  The first step is to catalogue the CYP enzymes based on their ability to operate on molecular fragments rather than the chemical reactions they catalyze. The second step is to apply a select set of transformations from the lookup table to the microbiota metabolites.   We provide results to show that the transformations predicted by the PxT method for bisphenol A (BPA) analogs, a class of widely used industrial chemicals implicated as endocrine disrupting chemicals, are similar to those confirmed in the literature.