Electrostatic Interactions Dictate Bile Salt Hydrolase Substrate Preference

Abstract

The human intestines are colonized by trillions of microbes, comprising the gut microbiota, which produce diverse small molecule metabolites and modify host metabolites, such as bile acids, that regulate host physiology. Biosynthesized in the liver, bile acids are conjugated with glycine or taurine and secreted into the intestines, where gut microbial bile salt hydrolases (BSHs) deconjugate the amino acid to produce unconjugated bile acids that serve as precursors for secondary bile acid metabolites. Among these include a recently discovered class of microbially-conjugated bile acids (MCBAs), wherein alternative amino acids are conjugated onto bile acids. To elucidate the metabolic potential of MCBAs, we performed detailed kinetic studies to investigate the preference of BSHs for host- and microbially-conjugated bile acids. We identified a BSH that exhibits positive cooperativity uniquely for MCBAs containing an aromatic sidechain. Further molecular modeling and phylogenetic analyses indicated that BSH preference for aromatic MCBAs is due to a substrate-specific cation-π interaction and is predicted to be widespread among human gut microbial BSHs.

Figure 1.

(a) Gut microbial bile salt hydrolases (BSHs) catalyze the hydrolysis of bile acid conjugates, whereas amino acid conjugation is performed by liver or gut microbial enzymes. (b) Activity heat map of host- and microbially-conjugated bile acid hydrolysis by C. perfringens CGH (Cp_CGH) and L. plantarum BSH1 (Lp_BSH1). BSH (387.8 nM) was incubated with 1 mM bile acid in phosphate buffered saline, pH = 6.2, with 10 mM DTT for 16 h at 37 °C. Data are representative of three independent experiments.

Figure 2.

Kinetic characterization of C. perfringens CGH (a-e) and L. plantarum BSH1 (f-j). BSH (a-b) 2.7 nM, (c-e) 387.8 nM, (f) 9.7 nM and (g-j) 193.4 nM was incubated with the indicated amount of substrate in phosphate buffered saline, pH = 6.2, with 10 mM DTT under initial rate conditions (a,b: 5 min; c-e: 1 h; f-i: 3 min; j: 10 min) at 37 °C. Data points were fit to the Michaelis-Menten equation (a-b, f-g), Hill equation (h, i), or linear regression (c-e, j). Data are representative of three independent experiments, n=3, points = mean, error bars = standard deviation.

Figure 3.

Molecular modeling analysis of L. plantarum BSH1. (a) Modeled homotetramer structure. Monomeric chains are colored purple, yellow, cyan, and blue. Modeled complexes for (b) GCA and (c) TyrCA, with interactions with selectivity loop (ball and stick, yellow loop) shown. Atom coloring (ball and stick models): Grey (carbon), red (oxygen), blue (nitrogen), white (hydrogen). (d) Sequence alignment of C. perfringens CGH (CGH) and of L. plantarum BSH1 (BSH1). Neighboring loop residues of CGH and BSH1 are highlighted (black box). Conserved negatively- and positively-charged residues are highlighted with red and blue boxes, respectively. Conserved residues are indicated with asterisks, whereas conservatively substituted and semiconservatively substituted residues are indicated with colons and periods, respectively.

Figure 4.

Phylogenetic analysis of BSH protein sequences from the human gut containing an aromatic MCBA selectivity loop. (a) Neighbor-joining tree of BSHs from MCBA selectivity loop, with tips colored by phylum. Tree scale represents phylogenetic distance determined by BLOSUM62 score, which is based on sequence similarity where a distance of 1.0 means 100% similar. (b) Genus-level abundances for human gut microbial Bacillota (formerly Firmicutes) BSHs containing aromatic MCBA selectivity loop. Genera accounting for 1% or more of all sequences with assigned genera are shown.