Bile salt hydrolases: Structure and function, substrate preference, and inhibitor development

Abstract

The worldwide trend of limiting the use of antibiotic growth promoters (AGPs) in animal production creates challenges for the animal feed industry, thus necessitating the development of effective non‐antibiotic alternatives to improve animal performance. Increasing evidence has shown that the growth‐promoting effect of AGPs is highly correlated with the reduced activity of bile salt hydrolase (BSH, EC 3.5.1.24), an intestinal bacteria‐producing enzyme that has a negative impact on host fat digestion and energy harvest. Therefore, BSH inhibitors may become novel, attractive alternatives to AGPs. Detailed knowledge of BSH substrate preferences and the wealth of structural data on BSHs provide a solid foundation for rationally tailored BSH inhibitor design. This review focuses on the relationship between structure and function of BSHs based on the crystal structure, kinetic data, molecular docking and comparative structural analyses. The molecular basis for BSH substrate recognition is also discussed. Finally, recent advances and future prospectives in the development of potent, safe, and cost‐effective BSH inhibitors are described.

Abbreviations

AGP

antibiotic growth promoter

BSH

bile salt hydrolase

CBAH

conjugated bile acid hydrolase

CBSH

conjugated bile salt hydrolase

CGH

cholylglycine hydrolase

GC

preference for glyco‐conjugated bile acids

GCA

glycocholic acid

GCDCA

glycochenodeoxycholic acid

GDCA

glycodeoxycholic acid

MW

molecular weight

Ntn

N‐terminal nucleophilic

SDR

substrate‐preference determining residue

TC

preferential hydrolysis of tauro‐conjugated bile acids

TC/GC

equal hydrolysis of both tauro‐ and glyco‐conjugated bile acids

TCA

taurocholic acid

TCDCA

taurochenodeoxycholic acid

TDCA

taurodeoxycholic acid

Introduction

Antibiotic growth promoters (AGPs), a group of antibiotics used in subtherapeutic quantities, have been used as feed additives to improve average daily weight gain and feed conversion ratio in food animals since the 1950s.1, 2 Although the benefits of this long‐established technique to growth promotion are still evident, recent epidemiological studies strongly indicate that AGPs exert selection pressure for the emergence, persistence, and transmission of antibiotic resistant bacteria, raising food safety, and public health concerns.2, 3 For this reason, there is a global trend to restrict the use of AGPs as additives in the feed industry,1, 2, 4 which necessitates the need to develop effective alternatives to AGPs to maintain productivity and sustainability of food animals.

Understanding the mechanism of how AGPs exert their growth‐promoting effects is critical for developing novel and effective AGP alternatives. It is widely accepted that AGPs mediate enhanced growth performance by altering intestinal microbial diversity and creating bacterial shifts.5 Although the definitive gut microbial community required for AGP‐mediated optimal growth promotion is still largely unknown, previous animal studies6, 7, 8, 9 have shown that the growth‐promoting effects of AGPs in pigs and chickens are highly and inversely correlated with the decreased activity of bile salt hydrolase (BSH, EC 3.5.1.24), an enzyme produced by various commensal bacteria in the intestine.1, 10

BSH, also called conjugated BSH (CBSH),11 conjugated bile acid hydrolase (CBAH)12 and cholylglycine hydrolase (CGH),13 is an enzyme catalyzing the hydrolysis of glycine‐ and/or taurine‐conjugated bile acids into deconjugated bile acids and amino acid residues [Fig. 1(A)].14 It belongs to N‐terminal nucleophilic (Ntn) hydrolase family. The microbial role of BSH activity and its impact on the host have been reviewed extensively.10, 15, 16 Although BSH enzymes confer BSH‐inactive bacterial resistance to bile acids,17, 18, 19 help fight against giardiasis,20, 21 and significantly reduce the serum cholesterol levels in pigs,22 dogs,23 mice24, 25 as well as humans,26, 27 they also have many negative effects on the host, such as dysfunctional lipid metabolism. Conjugated bile acids, synthesized from dietary cholesterol in the liver, are composed of a hydrophobic steroid core that is conjugated with either glycine or taurine [Fig. 1(A)].28 Their amphipathic characteristics make them more efficient “biological detergents” than unconjugated bile acids to emulsify and solubilize dietary lipids for fat digestion in the small intestine.10 Deconjugation of bile acids by BSHs from gut bacteria, therefore, leads to lipid malabsorption, which may cause weight loss of the host.10, 29 On the contrary, inhibiting the activities of intestinal BSHs will enhance lipid metabolism and energy harvest, thus improving growth performance and feed efficiency in food animals.

Figure 1.

Structures of a variety of substrates and inhibitors of BSHs. (A) General structures of bile acids. As an example, the structure of glycocholic acid is shown. The bond hydrolyzed by BSH is indicated by an arrow; rings A to D are highlighted in blue; OH* and OH** represent hydroxyl groups absent in deoxycholic and chenodeoxycholic acids, respectively. (B) Structures of BSH inhibitors riboflavin and phenethyl caffeate.

Recently, many BSH inhibitors have been developed to enhance animal growth performance and production. Copper and zinc which inhibit the activity of BSH are proved to improve the growth performance of swine9, 30 and poultry31, 32 as well as feed efficiency. Several BSH inhibitory compounds with potential as alternatives to AGPs, for example, two outstanding compounds riboflavin and phenethyl caffeate [Fig. 1(B)], have also been identified by an efficient high‐throughput screening system.5 Substrate‐guided rational development of inhibitors is another promising strategy.33, 34, 35 To facilitate the successful discovery and design of potent, safe and cost‐effective BSH inhibitors, however, a prerequisite is the detailed understanding of the molecular basis of BSH catalysis and substrate preference. In this review, genetic and biochemical properties of various BSHs, especially substrate preferences, are discussed. The current understanding of substrate recognition mechanism and structural basis for the substrate preferences of BSHs are then described. Finally, current progress and future prospects in BSH inhibitors development are highlighted.

Distribution of Bile Acids and BSHs

There are six main conjugated bile acids in bile, including taurocholic acid (TCA), taurodeoxycholic acid (TDCA), taurochenodeoxycholic acid (TCDCA), glycocholic acid (GCA), glycodeoxycholic acid (GDCA), and glycochenodeoxycholic acid (GCDCA) [Fig. 1(A)]. Their abundance differs in various animals and humans. In chicken, the dominant bile acids present in the intestine are taurine conjugates TCDCA and TCA.6, 36 However, pigs mainly contain GCDCA, TCDCA, GCA, TCA, hyodeoxycholic, and hyocholic acids.36 The contents of conjugated bile acids in human are as follows: 9% TCA, 9% TDCA, 8% TCDCA, 26% GCA, 22% GDCA, and 26% GCDCA.37 Thus, the ratio of glycine to taurine conjugates is approximately 3:1.10, 38

BSH enzymes have been identified from commensal gut microbiota, pathogen, and free‐living bacteria.39 Most of them are detected in several genera of gastrointestinal autochthonous micro‐organisms in animals, including Bifidobacterium,40 Lactobacillus,41, 42 Enterococcus, 43 and Streptococcus.44 Among these bacteria, lactobacilli constitute an important aspect of the animal intestinal microbiota and contribute to the majority of the total BSH activity in vivo.45 Species of human intestinal archaea such as Methanosphaera stadmanae and Methanobrevibacter smithii are shown to produce BSHs.39 Certain pathogens like Brucella abortus,46 Listeria monocytogenes, 47 and Clostridium perfringens 48 have also been reported to encode BSHs. Brucella abortus 46 and opportunistic pathogens from Xanthomonas 49, 50 and Bacteroides genera51, 52 are the only Gram‐negative bacteria reported to exhibit BSH activity. Interestingly, free‐living bacteria isolated from hot water springs (Brevibacillus sp.),53, 54 Antarctic lakes (Planococcus antarticus),55 and marine sediments (Methanosarcina acetivorans)55 are also BSH positive.

Biochemical Properties and Substrate Preferences of BSHs

Nair et al.13 partially purified, for the first time, the BSH from C. perfringens ATCC 19574 in 1967, which is now commercially available (Sigma‐Aldrich Co., Chicago, IL). Since then, many BSH enzymes have been genetically and biochemically characterized. Among them, a total of 33 BSHs whose amino acid sequences and substrate preferences have been simultaneously reported are summarized in Table 1. As shown in this table, BSH enzymes from various sources differ in the number of amino acids, optimal pHs and temperatures, molecular weights (MWs), and substrate preferences. These BSHs are mainly intracellular enzymes56 encoded by 314–338 aa, with optimum pHs ranging from 3.8 to 7.0. Except for LjPF01_BSHC, whose optimum temperature is 70°C, most BSH enzymes identified act optimally at temperatures of 30–55°C. MWs of BSH subunits range from 34 to 42 kDa, while the native enzymes have MWs of 80–250 kDa. Most BSHs are homotetramers, with LaCRL1098_BSH, BlBB536_BSH, and BSH from B. fragilis ssp. fragilis ATCC 2528552 existing in homodimeric, homohexameric, and homooctameric forms, respectively. In addition, the four BSHs from Lactobacillus johnsonii 100–100 are homo‐ or heterotrimers.57, 58, 59 The occurrence of multiple forms of BSHs has also been observed in other strains such as two BSH homologs in L. salivarius LGM1447660 and L. acidophilus NCFM,28 three and four homologs in L. johnsonii PF0161, 62 and L. plantarum,41, 42, 63 respectively.

Table 1.

Biochemical Properties of All BSHs with Reported Amino Acid Sequences and Substrate Preferences from Various Micro‐organisms

							MW (kDa)
BSH enzymea	Strain	Amino acids	GenBank accession no.	Optimum pH	Optimum temperature (°C)	Substrate preference b	Subunit	Native	References
LsUCC118_BSH1	L. salivarius UCC118	316	ACL98201.1	6.5	—c	GC	35.7	—	100
LsJCM1046_BSH1	L. salivarius JCM1046	324	ACL98194.1	5.5	—	TC	36.5	—	100
LsLGM14476_BSH1	L. salivarius LGM14476	324	ACL98197.1	5.5–7.0	—	TC	36.0	140	60
LsLGM14476_BSH2	L. salivarius LGM14476	325	ACL98205.1	5.5–6.0	—	TC/GC	36.0	142	60
LsB‐30514_BSH1	L. salivarius B‐30514	324	AFP87505.1	5.5	41	GC	37.0		85
LpBBE7_BSH	L. plantarum BBE7	324		6.0	37	GC	37.0	140–150	101, 102
LpST‐III_BSH1	L. plantarum subsp. plantarum ST‐III	324	ADO00098.1	—	—	GC	37.0	—	42
Lp80_BSH	L. plantarum 80	324	AAB24746.1	4.7–5.5	30–45	GC	37.0		103
LpWCFS1_BSH1	L. plantarum WCFS1	324	CAD65617.1	—	—	GC	37.0	—	41
LpCGMCC8198_BSH2	L. plantarum CGMCC 8198	338	AGG13403.1	—	—	GC	37.5	—	63
LpCGMCC8198_BSH3	L. plantarum CGMCC 8198	328	AGG13404.1	—	—	TC/GC	36.1	—	63
LpCGMCC8198_BSH4	L. plantarum CGMCC 8198	317	AGG13405.1	—	—	TC	35.7	—	63
LgAM1_BSH	L. gasseri Am1	325	FJ439777.1	—	—	GC	36.2	—	37
LgFR4_BSH	L. gasseri FR4	326	WP_020806888.1	5.5	52	GC	37.0	—	82
LaNCFM_BSHA	L. acidophilus NCFM	325	AAV42751.1	—	—	GC	37.1	—	28
LaNCFM_BSHB	L. acidophilus NCFM	325	AAV42923.1	—	—	TC/GC	37.0	—	28
LrCRL1098_BSH	L. reuteri CRL 1098	325	WP_035157795.1	5.2	37–45	GC	36.1	80	65, 104
LjPF01_BSHA	L. johnsonii PF01	326	EGP12224.1	5.0	55	TC	36.6	—	61
LjPF01_BSHB	L. johnsonii PF01	316	EF536029.1	6.0	40	TC	34.0	—	61, 62
LjPF01_BSHC	L. johnsonii PF01	325	EGP12391.1	5.0	70	GC	36.4	—	61
Lj100–100_CBSHα	L. johnsonii 100–100	326	AAG22541.1	3.8–4.5	—	TC/GC	42.0	115	57, 58, 59
Lj100–100_CBSHβ	L. johnsonii 100–100	316	AAC34381.1	3.8–4.5	—	TC/GC	38.0	105	57, 58, 59
LfNCDO394_BSH	L. fermentum NCDO394	325	AEZ06356.1	6.0	37	GC	36.5	—	105
LrE9_BSH	L. rhamnosus E9	338	ANQ47241.1	—	—	GC	37.1	—	106
BlSBT2928_BSH	B. longum SBT2928	317	AAF67801.1	5.0–7.0	40	GC	35.0	125–130	72, 75
BlBB536_BSH	B. longum BB536	317		5.5–6.5	42	TC/GC	40.0	250	14, 107
BlLMG21814_BSH	B. longum subsp. suis LMG 21814	317	KFI71781.1	5.0	37	GC	35.0	107–124	108
BbATCC11863_BSH	B. bifidum ATCC 11863	316	AAR39435.1	—	—	GC	35.0	140–150	40, 76
BaBi30_BSH	B. animalis subsp. lactis Bi30	314	AEK27050.1	4.7–6.5	50	GC	35.0	120–140	109
BaKL612_BSH	B. animalis subsp. Lactis KL612	314	AAS98803.1	6.0	37	GC	35.0	—	110, 111
BpDSM20438_BSH	B. pseudocatenulatum DSM 20438	316	KFI75916.1	5.0	37	TC/GC	35.0	123–154	108
Cp13_CBAH1	C. perfringens 13	329	P54965.3	4.5	—	TC	36.1	147	48, 71
EfNCIM2403_BSH	Enterococcus faecalis NCIM 2403	324	EET97240.1	5.0	50	TC	37.0	140	77, 83

^a BSH, bile salt hydrolase; CBSH, conjugated bile salt hydrolase; CBAH, conjugated bile acid hydrolase.

^b TC, preferential hydrolysis of tauro‐conjugated bile acids; GC, preference for glyco‐conjugated bile acids; TC/GC, equal hydrolysis of both tauro‐ and glyco‐conjugated bile acids.

^c Not available.

Substrate preferences of BSHs listed in Table 1 were mostly determined by their kinetic parameters and specific activities toward different substrates. Most BSH enzymes characterized prefer to hydrolyze glyco‐conjugated bile acids (Table 1), which can be mainly ascribed to the steric hindrance caused by the sulfur atom in tauro‐conjugated bile acids [Fig. 1(A)].64 Because glyco‐conjugated bile acids are far more toxic for bacteria than the tauro‐conjugates, the higher affinity of BSHs for glyco‐conjugates may be of great importance in the ecology of gut microbe.37, 65 Seven BSH enzymes preferentially hydrolyze tauro‐conjugates, whereas other seven BSHs hydrolyze both glyco‐ and tauro‐conjugated bile acids, displaying a broad spectrum of specificity.

Most BSHs from L. johnsonii are more efficient at hydrolyzing tauro‐conjugated bile acids compared with glyco‐conjugates, although some exceptions are found. But the majority of BSH enzymes from L. plantarum and Bifidobacteria display preferential hydrolysis of glyco‐conjugated bile acids. Thus, the substrate preferences of BSHs may be strain dependent. In addition, multiple BSH homologues from the same strain may show different preferential activities such as LjPF01_BSHA, LjPF01_BSHB, and LjPF01_BSHC, exhibiting specific affinities for tauro‐, tauro‐, and glyco‐conjugated bile acids, respectively.61, 62

Potential Mechanism of Substrate Recognition

Despite the remarkable progress in identification and characterization of new BSHs, the molecular basis by which BSHs distinguish and recognize the two kinds of substrates remains unknown. Substrates of BSH enzymes consist of a cholate steroid nucleus and amino acid groups (glycine/taurine) [Fig. 1(A)]. Huijghebaert and Hofmann66 demonstrated that modification of amide bond or loss of a negative charge on the terminal group of an amino‐acid moiety of the bile acids greatly reduced the hydrolysis rate compared with glycine and taurine conjugates. Batta et al.67 showed that cholylglycine hydrolase was more efficient at hydrolyzing conjugates prepared from unconjugated bile acids and analogs of glycine and taurine with one or two methylene groups, but conjugates containing analogs of glycine and taurine with a tertiary amide group completely resisted hydrolysis. Most kinetic data available in the literature also indicate that substrates are predominantly recognized by BSHs at the amino acid moieties, and most BSHs prefer glyco‐conjugated bile acids over tauro‐conjugated ones.68, 69 But there is some evidence to suggest that modifications in the steroid ring system and the side chains affect the substrate preferences of BSH enzymes.28, 67, 70 It is, therefore, possible that BSHs have evolved to recognize bile acids on both the amino acid groups and cholate steroid nucleus. However, Batta et al.67 and Rossocha et al.71 reported that hydroxyl or epimerized hydroxyl substituents of cholate had no effect on the affinities of BSH enzymes and that the cholate was bound via hydrophobic interactions instead of hydrogen bonding, which complicates the identification of recognition sites. Thus, it appears that there are perhaps no specific binding conditions, except that the amino acids and cholate moieties should fit properly and are complementary to the substrate‐binding pockets.72

Structural Basis for the Substrate Preference of BSHs

Phylogenetic clustering using Clustal X2 and MEGA 4.0 (Fig. 2) was employed to uncover the evolutionary distance of BSH enzymes listed in Table 1. This phylogenetic tree reveals the divergence among genetic characteristics of these BSHs. BSH enzymes from various bacteria are clearly divided into four groups (Fig. 2). Members of each group share a recent ancestor and have shorter genetic distances. However, in each group, BSHs exhibit different substrate preferences, which may be because evolution tends to broaden substrate selectivity.37 Multiple sequence alignment shows that the pairwise sequence identity of these BSHs ranges from 20.33% to 99.69%, with an average of 40.88%. BSH enzymes from L. johnsonii PF01 and L. johnsonii 100–100 have a high degree of similarities to those from other species, such as L. acidophilus and L. plantarum, because they are acquired through horizontal gene transfer.10, 28, 57 However, LjPF01_BSHC seems to be common only among L. johnsonii strains.61

Figure 2.

Family tree of BSH members with characterized substrate preference. BSHs exhibiting a preference for glyco‐ and tauro‐conjugated bile salts are highlighted by red and blue, respectively; while BSHs equally hydrolyzing both tauro‐ and glyco‐conjugated bile salts are colored in orange.

BSH enzymes catalyze amide bond hydrolysis through their N‐terminal nucleophile residue, Cys2 (unless otherwise stated, all the residues are numbered from methionine in BlSBT2928_BSH in this review), which is activated by autocatalytic endoproteolytic process to form a free α‐amino group serving as a base in the catalytic reaction.73 Aside from Cys2, other residues involved in the catalytic reaction, such as Arg18, Asp21, Asn82, Asn173, and Arg226, are also strictly conserved in BSHs (Fig. 3),62, 72 except the lack of Asn173 in LsUCC118_BSH1. The catalytic mechanism of BSHs involves an initial nucleophilic attack of Cys2 on the amide bond, followed by the formation of a tetrahedral intermediate which is stabilized by an oxyanion hole formed by the peptide NH of Asn82 and Asn173 Nδ2.71, 73, 74 In this process, Arg18 stabilizes the negatively charged sulfhydryl group of Cys2, while the N‐terminal Cys2 forms the catalytic Ntn‐diad with Asp21.64, 71 Asn173 (corresponding to Asn175 in Cp13_CBAH1) participates in substrate recognition via hydrogen bonding,74 whereas Arg226 (equivalent to Arg228 in Cp13_CBAH1) helps in transition‐state stabilization.74 Among these six active sites, however, Cys2 and Asn82 (corresponding to Asn79 in EfNCIM2403_BSH) are the only two residues whose essential roles in BSH activity have been confirmed by site‐directed mutagenesis.72, 75, 76, 77

Figure 3.

Multiple sequence alignment of BSHs listed in Table 1. BSHs with substrate preference for GC, TC/GC, and TC are colored in red, orange, and blue, respectively. Active sides are highlighted by yellow. Key amino acid residues potentially involved in substrate selectivity of Cp13_CBAH1, BlSBT2928_BSH, and LgFR4_BSH are highlighted by green, purple, and pink, respectively. Residues at Position 68 are indicated by a five‐pointed star. Secondary structure elements of BlSBT2928_BSH (above) and Cp13_CBAH1 (below) are also shown.

As enzymes possessing similar residues in their binding pockets may have similar mechanisms of action,78 substrate preference of BSHs may rely on specific amino acids (also called substrate‐preference determining residues, SDRs) and secondary structural elements.79 Recently, some key amino acid residues and secondary structural elements that are potentially involved in substrate binding have been identified in BSHs by binding site similarity‐based scoring system, crystal structure, kinetic data, molecular docking, and comparative structural analyses and are summarized in Table 2 and Figure 3. Except for Trp143, the 14 residues involved in substrate binding are located in three loops: Loop1 (20–26), Loop2 (59–68), and Loop3 (131–142). But no key residues have been reported in Loop4 (261–269). In the BSHs included in Figure 3, their substrate‐binding residues are not strongly conserved, except for Leu141 and Trp143, consistent with the previous findings.10, 72 Besides, Trp143 has been shown to be crucial for substrate binding in BlSBT2928_BSH.72, 80 However, most substitutions are conservative. The four loops of BSHs are predominantly hydrophobic. As compared with BSHs with substrate preference for TC/GC and TC, glyco‐conjugating‐specific BSHs have the lowest percentage of hydrophobic amino acids in their four loops, while the percentage of hydrophilic amino acids or charged amino acids in the four loops of these BSHs is the highest. These findings indicate that BSH enzymes may recognize their substrates through hydrophobic interactions with the hydrophobic steroid moiety.12, 71

Table 2.

Key Amino‐Acid Residues and Secondary Structural Elements that are Potentially Involved in Determining Substrate Preference

BSH enzymea	Residue	Secondary structural element	Method	References
Cp13_CBAH1	Met20, Phe26, Phe61, Ala68, Ile133, Ile137, Leu142, and W144	β6 (54–61), β7 (64–72), and β11 (142–147)	Crystal structure analysis	61, 71, 72, 112
BlSBT2928_BSH	Trp21	Loops20–27, 35–50, 60–65, and 125–144	Crystal structure and kinetic data analyses	72
BlSBT2928_BSH	Leu20, Trp22, Phe24, Tyr26, Val59, Met61, Met66, Phe68, Val133, Q137, Ser139, and Leu141	Loop1 (20–26), Loop2 (59–68), and Loop3 (133–142)	Comparative structural analysis	12
BlSBT2928_BSH	— b	Loop1 (22–27), Loop2 (58–65), Loop3 (129–139), and Loop4 (261–269)	Binding site similarity‐based scoring system	55
LsB‐30514_BSH1	Leu18, Phe22, Tyr24, Ile 56, Leu63, Phe65, Phe130, and Leu134	Loop1 (20–27) and Loop2 (125–139)	Crystal structure analysis	81
LgFR4_BSH	Cys2, Phe25, Lys59, and Gln135	—	Molecular docking analysis	82
EfNCIM2403_BSH	Phe18, Tyr20, and Tyr65	Loop1 (21–25), Loop2 (57–64), Loop3 (125–137), and Loop4 (255–269)	Molecular docking analysis	77
BSH of L. plantarum PYPR1	Leu6, Ile8, Asn12, Lys88, and Asp126	—	Molecular docking analysis	84

^a BSH, bile salt hydrolase; CBAH, conjugated bile acid hydrolase.

^b —, not available

Position 68 has been shown to be a potential substrate preference determining site in the crystal structure of Cp13_CBAH1,71 and it is noted due to its conservation of a non‐polar residue (alanine, phenylalanine, and valine) in BSHs with preference for tauro‐conjugated bile acids or a polar residue (tyrosine, cysteine, threonine, and serine) for glyco‐conjugating‐specific BSHs (Fig. 3).61 Because Ala68 in Cp13_CBAH1 (equivalent to Phe68 in BlSBT2928_BSH) is located within the valeric acid side chain of deoxycholate which is near the hydroxyl substituents, more hydrophilic amino acids added here can, therefore, alter the affinities toward the corresponding substrates.71 Ala68 also appears to have close interaction with the β11 sheet, which is significant for substrate binding and catalysis in Ntn‐hydrolases owing to its position within the αββα core structure.73 As shown in Figure 3, however, there are still 11 BSHs, for example, LsUCC118_BSH1, LsB‐30514_BSH1, LrCRL1098_BSH, LgFR4_BSH, and LaNCFM_BSHA, that have nonpolar residue phenylalanine at Position 68 but preferentially hydrolyze glyco‐conjugated bile acids. This may be because the residue at Position 68 is just one of many SDRs that determine the substrate preferences of BSH enzymes.

Other potential SDRs have also been identified in different BSHs (Table 2). In Cp13_CBAH1, it is found that deoxycholate is sandwiched by Phe61 and Ile137 on the side of the deoxycholyl ring A [Fig. 1(A)] and by Met20, Phe26, and Ala68 on the side of valeric acid side chain of deoxycholate. Additional important hydrophobic interactions are observed between deoxycholyl Ring B and Ile133 and between Ring D and Leu142.71 By substrate binding studies and sequence alignment, Trp22 in BlSBT2928_BSH is shown to play a selective role in the binding of bile acids and penicillin V.72 Based on comparative structural analysis, Phe130 and Leu134 are found to condense the substrate‐binding pocket entrance of LsB‐30514_BSH1 to restraint the spatial configuration, while Tyr24 and Phe65 may force the substrate to sit deeply in its binding pocket.81 Being located at the bottom of the binding pocket, Ile56 and Leu63 may determine the differing substrate preferences of LsB‐30514_BSH1. According to the results of molecular docking studies, Cys2 and Lys59 in the substrate‐binding pocket of LgFR4_BSH are found to form hydrogen bonds with GCA.82 Similarly, GDCA forms hydrogen‐bonding interactions with Cys2, Phe25, and Lys59, but Lys59 and Gln135 interact with TCA via hydrogen bonding. The three aromatic residues (Phe18, Tyr20, and Tyr65) of the binding pocket in EfNCIM2403_BSH, which are different from the corresponding residues in other BSHs, are shown to interact with the cholate part of GCA.77 Besides, they may also contribute to the higher activity and unique allosteric behavior of EfNCIM2403_BSH.83 In BSH from L. plantarum PYPR1, Leu6, Ile8, and Asn12 are also shown by molecular docking analysis to interact with TCA through hydrogen bonding, while Lys88 and Asp126 form hydrogen bonds with GCA.84 Since the amino acid sequence of L. plantarum PYPR1 BSH is not available now, these important residues are not highlighted in Figure 3.

To date, little information is available regarding the structural basis for BSHs function. Three‐dimensional structures of the BSH enzymes from only four species, C. perfringens (2bjf, Cp13_CBAH1),71 B. longum (2hez, BlSBT2928_BSH),72 L. salivarius (5hke, LsB‐30514_BSH),81 and E. faecalis (4wl3, EfNCIM2403_BSH),77 have been reported. A superimposition of the three‐dimensional structures of these four BSHs shows that they possess the distinctly conserved αββα tetra‐lamellar tertiary structures of Ntn hydrolase superfamily and that they share the conserved catalytic active center containing the cysteine nucleophile (Cys2) and its co‐ordinated neighboring amino acids (Fig. 4).71, 72 However, the substrate‐binding loops flanking the active site (Loops1–4) are variable. These loops not only determine the mode of substrate binding but also define the volume of the site. As compared with the corresponding loops of penicillin V acylase, the loops in BSHs are generally shorter to accommodate the bulky steroid nucleus in the active site.64 Subtle changes in these loop structures significantly affect the catalytic specificity of BSHs.12 Out of the four loops in BSHs, Loop3 (131–142) and Loop4 (261–269) show significant differences in terms of their conformation and folding (Fig. 4). Loop3 which interacts with the cholate backbone and contains more polar residues is also dynamic in BSHs, with different sizes and orientations.64 Besides, the third loops of Cp13_CBAH1 and EfNCIM2403_BSH which have higher affinities for tauro‐conjugates are more close to deoxycholic acid than those of BlSBT2928_BSH and LsB‐30514_BSH1, consequently reducing the size of the substrate‐binding pocket. As compared with other BSHs, the Loop2 of EfNCIM2403_BSH, part of the adduct binding site (Site A), is more extended, which might increase its dynamicity and flexibility and facilitate fast shuttling of the substrate and quick adduct release.77

Figure 4.

Comparison of the binding pocket loops of Cp13_CBAH1 (green; PDB ID: 2bjf), BlSBT2928_BSH (red; PDB ID: 2hez), LsB‐30514_BSH (yellow; PDB ID: 5hke), and EfNCIM2403_BSH (blue; PDB ID: 4wl3). TAU, taurine; DXC, deoxycholic acid. The six residues (Cys2, Arg18, Asp21, Asn82, Asn173, and Arg226) involved in the catalytic reaction in BlSBT2928_BSH are shown.

Polar complementarity toward the conjugated bile acids also constitutes an important factor influencing substrate preferences and catalytic framework of BSHs.55, 67 In BlSBT2928_BSH and EfNCIM2403_BSH, polar complementarity, along with good hydrogen bonding interactions, was observed for all three hydroxyl groups of GCA. As compared with BlSBT2928_BSH, the lower polar complementarity of Cp13_CBAH1 contributes to its lower enzyme activity.55, 72 A more in‐depth understanding of the structural changes and polar complementarity that occur during substrate binding would provide a new insight into the varied substrate preferences and functions of BSHs.

BSH Inhibitor Development

As noted above, BSH has been proposed as a key mechanistic microbiome target for developing novel alternatives to AGPs, and several BSH inhibitors have recently been developed. Copper (CuCl₂) and zinc (ZnSO₄) which can inhibit 98.1% and 89.5% of the activities of glyco‐conjugating‐specific LsB‐30514_BSH1, respectively, have been identified as potent BSH inhibitors.85 When used at high concentrations, they can improve the growth performance and feed efficiency in swine9, 30 and poultry.31, 32 Some currently used AGPs such as penicillin V, ampicillin, doxycycline hydrochloride,82 and tetracycline5 also show significant inhibitory effects on BSH activities. However, long‐term use of these metal ions and antibiotics affects animal health, and increases the accumulation of metal ions in the treated animals and the environment as well as the risk of acquiring antibiotic‐resistant bacteria, respectively. Recently, Smith et al. identified several promising BSH inhibitors from 2240 diverse compounds by an efficient high‐throughput system. Among them, riboflavin and phenethyl caffeate [Fig. 1(B)] not only display relatively higher inhibitory effects on LsB‐30514_BSH15 but can also inhibit the activities of LgFR4_BSH82 and LjPF01_BSHB86 which show a preference for glycol‐ and tauro‐conjugated bile acids, respectively. The mechanism underlying the inhibitory effect of riboflavin was further unraveled by molecular docking analysis, which showed that both substrate and inhibitors (riboflavin and penicillin V) could bind to the substrate‐binding pocket of LgFR4_BSH with almost the same binding affinity/energy. It has also been shown that dietary supplementation with riboflavin (20 mg/kg feed) significantly enhances the feed efficiency and body weight in pigs.87 Additionally, riboflavin along with the probiotic L. gasseri FR4 can also serve as a BSH inhibitor.82 Because of its positive effects on animal physiology, riboflavin, a non‐toxic available reagent, has been approved by the Food and Drug Administration (FDA) of United States of America for the use as a feed additive to overcome vitamin B₂ deficiencies in animals.5, 88

Although BSH inhibitors have been shown to improve the feed efficiency and growth performance of food animals, the precise mechanism behind their enhancing effects is yet to be delineated. Aside from their significant roles in lipid digestion, bile acids also function as versatile signaling molecules regulating cellular and metabolic activities such as their own synthesis and circulation, as well as glucose, triglyceride, cholesterol, and energy homeostasis.38, 64 Besides, as conjugated bile acids are toxic to bacteria, alteration of bile acid profiles caused by BSH hydrolysis may change gastrointestinal microbiota, which has been associated with metabolic diseases, for example, obesity, diabetes, and hypercholesterolemia in animals and Type 2 diabetes mellitus and inflammatory bowel disease in humans.38, 64, 89 Therefore, further comprehensive animal trials are still needed to determine the precise mechanism underlying the enhancing effects of BSH inhibitors, especially the potential unexpected effects of BSH inhibitors such as excess fat deposition, disturbance of digestive microbiota, and changes in metabolism.

Concluding Remarks and Perspectives

Microbiome studies have shown that usage of AGPs does coincide with reduced BSH activity and decreased the population of some Lactobacillus species, the major BSH producer, in the intestine of food animals. Consequently, BSH has become a promising mechanistic microbiome target for developing novel alternatives to AGPs to enhance the productivity and sustainability of food animals. Although several BSH inhibitors have been identified by biochemical characterization and high‐throughput screening, the more detailed understanding of the structural basis for substrate preferences of BSHs would lay a solid foundation for the rationally tailoring of potent, safe, and cost‐effective BSH inhibitors.

As summarized in this review, structure analyses of BSH enzymes from various species reveal critical residues in catalysis and provide key information on substrate selectivity of BSHs. However, research on the substrate preference determinants of BSHs is still in its infancy. Future in‐depth structural analysis of BSHs in complex with specific substrates and inhibitors, along with comprehensive amino acid substitution mutagenesis would help to discover SDRs and understand the structural basis for the substrate preferences of BSHs.90 As both BSH and penicillin V acylase belong to Ntn hydrolases superfamily, the SDRs identified in penicillin V acylase may also affect the substrate preferences of BSH. Recently, based on the concept that the binding specificity would covary with SDRs, some structure‐driven computational and theoretical approaches have been developed for SDRs identification, including likelihood methods,91 independence tests,92 information theoretic approaches,93 Bayesian methods,94 and statistical coupling analysis.95 The basic idea is to search two sites in a multiple sequence alignment, or a pair of multiple sequence alignments that exhibit reciprocal changes when studying protein–protein interactions. Among these methods, the information theoretic approach which uses sequence information only without known crystal structure has been successfully used for the identification of SDRs in many protein recognition domains such as PDZ,96 SH3,97 and kinase domains.98 This makes information theoretic approach a good candidate for the precise identification of preference determinants of BSH enzymes in further studies.

As a close relative of BSHs, β‐lactamase (also known as penicillinase, EC 3.5.2.6) also acts on amide bonds, lessons learned from discovering or designing β‐lactamase inhibitors can also be applied to the development of BSH inhibitors. The important factors that must be taken into consideration when designing BSH inhibitors may include99 (i) exhibiting high affinity for the active site of the target enzyme; (ii) mimicking the “natural substrate” (substrate analogues); (iii) stabilizing interactions in the active site; (iv) rapid cell penetration. As BSHs are mainly in the intracellular space of bacteria, the rapidity with which inhibitors can access their target is a requisite for successful inhibition; (v) low propensity to induce BSH production; (vi) the risk of bacterial adaption to BSH inhibitors.

With the increased knowledge on the substrate preferences of BSHs and the role of BSH‐producing bacteria, it should be possible to tailor BSH inhibitors‐based feed additives to replace AGPs in the future. Such inhibitors could also be used in conjunction with probiotics to design rationally tailored dietary supplements that will enhance animal health and performance. Large animal trials are also highly warranted to determine the effects of these feed additives on the growth performance of food animals.

Conflict of Interest

The authors declare that there is no conflict of interest.