Characterization of a widely distributed cardiac drug-inactivating enzyme from the human gut bacterium Eggerthella lenta

Abstract

The human body is colonized by trillions of microorganisms that are increasingly implicated in modulating human health and disease. In particular, the human gut microbiota is involved in the metabolism of over fifty pharmaceuticals, yielding metabolites with altered biological properties and toxicities. It has been known for decades that particular isolates of the human gut bacterium Eggerthella lenta transform the plant-derived toxin and cardiac drug digoxin into the inactive metabolite (20R)-dihydrodigoxin, leading to decreased efficacy in a considerable subset of patients. Recently, the Turnbaugh lab identified the cardiac glycoside reductase (cgr) operon, a digoxin-inducible gene cluster that was predicted to be responsible for digoxin metabolism. In this thesis, we sought to expand our mechanistic understanding of this clinically relevant transformation and investigate its broader implications for gut microbial and human health. Through heterologous expression and in vitro biochemical characterization, we discovered that the E. lenta enzyme Cgr2 is sufficient for digoxin reduction and inactivation. Having validated the cgr2 gene as a biomarker for digoxin reduction, we probed the distribution of this metabolism among E. lenta strains and in the general human population. Using culturing and sequencing analyses, we identified seven additional digoxin-metabolizing strains of E. lenta with remarkable sequence conservation of the cgr2 gene (>98% sequence identity). Metagenomic and qRT-PCR analyses confirmed the high sequence conservation of the cgr2 gene and revealed that cgr2+ E. lenta are widespread, but often low in abundance in the human gut microbiota. We next probed the biochemical mechanism of digoxin reduction by the prevalent Cgr2 enzyme. Using a combination of biochemical, bioinformatic, and spectroscopic techniques, we determined that Cgr2 is a unique reductase that requires an FAD cofactor, harbors oxygen-sensitive, redox-active [4Fe-4S] clusters, and may contain a divalent metal cation center. Although the presence of [4Fe-4S] clusters in Cgr2 was unexpected, these metalloclusters proved to be essential for Cgr2 stability and likely serve a catalytic, electron transfer role. We further identified six cysteine residues that are important for Cgr2 activity and may influence metallocofactor binding. We next explored the evolutionary origins and impacts of digoxin metabolism on E. lenta. Despite the high sequence conservation of the cgr operon, no obvious physiological benefit (e.g. growth advantage) could be linked to this metabolism. We thus investigated whether digoxin is the endogenous substrate of Cgr2 by assessing the activity of this enzyme toward a panel of alternative candidate substrates that may be relevant in the gut. However, Cgr2 metabolism appears to be restricted to digoxin and highly similar cardenolide toxins produced by plants. We thus propose that the cgr pathway may have evolved to protect humans from ingested toxins in an analogous manner to host xenobiotic-detoxifying enzymes. By maintaining host health, this pathway could thus provide a habitat for E. lenta colonization. The studies presented in this thesis shed light on a long-standing clinical problem, and provide a roadmap for the discovery and mechanistic characterization of additional gut microbial-xenobiotic interactions of medical importance. Such knowledge may allow for accurate prediction of xenobiotic metabolism in the clinic and enable the development of therapeutic interventions to modulate these activities and ultimately affect human health.