Review on the Function, Substrate Affinity, and Potential Application of Bile Salt Hydrolase Originated from Probiotic Strains of Lactobacillus, Bifidobacterium, and Enterococcus

Bile salt and bile acid are distinguished if glycine or taurine is conjugated or not. If no glycine or taurine is attached, the substance is referred to as bile acid; otherwise, it is called bile salt (Daly et al., 2021). Primary bile acids are synthesized in the liver and denatured by bacteria to form secondary bile acids (Daly et al., 2021). Haslewood (1967) reported that the primary bile acids, cholic acid (CA) and chenodeoxycholic acid (CDCA), are found in the bile of vertebrates. Similarly, secondary bile acids, including deoxycholic acid (DCA), lithocholic acid (LCA), and ursodeoxycholic acid, are found in this fluid (Haslewood, 1967). Notably, DCA is produced by modifying CA, and LCA is derived from CDCA (Bachmann et al., 2015; Evangelakos et al., 2021; García-Cañaveras et al., 2012; Perez and Briz, 2009). Tagliacozzi et al. (2003) and Thakare et al. (2018) reported that CDCA predominated in human plasma, followed by deoxycholic acid (DCA), and glycochenodeoxycholic acid was three times higher than taurochenodeoxycholic acid.

A wide range of bile acids are distributed across different species (Kuhajda et al., 2006; Li and Dawson, 2019). Bile salts conjugated with taurine are dominant in most cases, excluding those in humans and other animals. In addition, Karakus et al. (2024) reported that glycine-conjugated bile salts are dominant in humans, and taurine-conjugated bile salts are most common in dogs (García-Cañaveras et al., 2012; Kakimoto et al., 2017; Rabin et al., 1976; Vessey, 1978).

Bile salts primarily aid in food digestion (Maldonado-Valderrama et al., 2011). de Buy Wenniger and Beuers (2010) and Redinger (2003) reported that bile salts are synthesized in the liver, stored in the gallbladder, and secreted along with pancreatic enzymes in the duodenum upon ingestion of food. Notably, these salts have amphipathic characteristics (Daly et al., 2021). The hydrophobic part attaches to ingested lipid droplets and divides them into smaller particles. These fine lipid particles help lipolytic enzymes, such as lipase, to work better, and the bile salt is reabsorbed 95% near the ileum when the process is complete. It enters the portal vein along the capillaries and re-enters the liver thereafter. The 5% of bile salt that was not absorbed from the ileum was fermented or deconjugated by the gut microbiome. Most of the affected bile salt is excreted with feces but some of it is reabsorbed. The whole circulation and enzyme effect for bile circulation is shown in Fig. 1 (Daly et al., 2021; de Buy Wenniger and Beuers, 2010; Redinger, 2003).

Fig. 1. Bile salt circulation (copyright by Bio-Render^@). PCBA, primary conjugated bile acid; PDBA, primary deconjugated bile acid; SCBA, secondary conjugated bile acid; SDBA, secondary deconjugated bile acid; FXR, farnesoid X receptor; GCA, glycocholic acid; TCA, taurocholic acid; GCDCA, glycochenodeoxycholic acid; TCDCA, taurochenodeoxycholic acid; GDCA, glycodeoxycholic acid; TDCA, taurodeoxycholic acid; CA, cholic acid; DCA, deoxycholic acid; CDCA, chenodeoxycholic acid.

Download Original Figure

Several study reports revealed that high blood cholesterol levels (LDL-C>190 mg/dL; Bittencourt et al., 2020), accompanied by a sedentary lifestyle, are increasingly prevalent across generations (Evangelakos et al., 2021; Perez and Briz, 2009; Sun et al., 2023). Individuals with high plasma LDL-C levels have a high risk of developing CVD and a shorter life expectancy. Prescription of statins or ezetimibe to patients with high-cholesterol syndrome may relieve symptoms. However, they cause side effects, such as myopathy, acute pancreatitis, and liver toxicity (Florentin et al., 2008; Kashani et al., 2008). Furthermore, patients with liver or CVDs are particularly vulnerable to these side effects (Begley et al., 2006; Schuster, 2004).

Several studies have reported that probiotics can lower blood cholesterol levels with almost no side effects and have positive effects on various conditions, including the treatment of atopic dermatitis, colon cancer, Crohn’s disease, diarrhea, and constipation (Ishimwe et al., 2015; Ooi and Liong, 2010; Sivamaruthi et al., 2019). However, their respective mechanisms remain unclear (Gill and Guarner, 2004; Mercenier et al., 2003; Reid et al., 2003; Sanders and Klaenhammer, 2001; Tuohy et al., 2003; Woo et al., 2023). Ahn et al. (2003) reported that cholesterol was reduced following 4 weeks of consuming milk containing Lactobacillus acidophilus SNUL01, and Fuentes et al. (2013) highlighted that cholesterol was lowered by consuming the capsule form of Lactobacillus plantarum (CECT 7527, CECT 7528, and CECT 7529). The hypothesis supporting the reduction of blood cholesterol levels as a function of BSH revealed so far can be explained as follows: more than 95% of bile salts are reabsorbed in the human ileum (Li and Chiang, 2020; Naumann et al., 2020), and the remaining 5% passes through the ileum. Colonic bile salts are deconjugated by BSH activity, forming deconjugated bile salts. BSH has a specific active site (Fig. 2), especially the cys-2 (or 22) site, which is essential for BSH catalysis (Begley et al., 2006). In bile salt deconjugation, cys-2 attacks the carbonyl carbon of the excision amide bond in bile salt, followed by the removal of glycine or taurine by hydrolysis (Chand et al., 2018). Deconjugated bile salts are water-soluble in a colonic pH environment of 7–8 (Trivedi and Puranik, 2017; Yamamura et al., 2023). The metabolic activity of intestinal microorganisms, particularly lactic acid and short-chain fatty acid production, further lowers the pH, causing the precipitation of deconjugated bile salt (Begley et al., 2006). Therefore, colon enterocytes no longer absorb it, leading to its excretion in the feces, which in turn lowers blood cholesterol.

Fig. 2. Comparative analysis of BSH and PVA active site. The genes used for the phylogenetic tree were translated and aligned using the MEGA 11software. The black arrows indicate the amino acid residues involved in the active site for BSH while red arrows pertain to PVA. The black and red numbers next to each amino acid indicate the location of the residue along the length of the peptide sequence. BSH, bile salt hydrolase; PVA, penicillin V acylase.

Download Original Figure

Nonetheless, confirming whether BSH is a factor remains difficult because microorganisms that reduce blood cholesterol levels exist despite the absence of the BSH gene, such as Streptococcus thermophilus MCC0200 (Kapse et al., 2024). In addition, Choi et al. (2015) reported that deconjugated bile salt has a stronger affinity for the farnesoid X receptor that regulates bile synthesis, reducing hepatic bile acid synthesis; by this result, the effect of BSH does not alter the blood cholesterol concentration. Consequently, current experimental results do not identify BSH activity as a major factor for the LDL-C-lowering effect in the presence of probiotics.

Even if probiotics do not have an LDL-C-lowering effect, maintaining high survival rates in the intestine is critical to elicit other health benefits to the host. This concept originated from the study of Fuller (1995). According to Dobson et al. (2012), probiotics are resistant to acid and produce antibacterial substances, including bacteriocins, hydrogen peroxide, and organic acids. Furthermore, probiotics are highly resistant to bile salts compared with other bacteria (da Silva et al., 2024; Gu et al., 2024; Horackova et al., 2020; Spínello et al., 2024; Urdaneta and Casadesús, 2017; Yang et al., 2019). The study by De Smet et al. (1995) suggested that BSH-positive probiotics would have stronger bile resistance than negative. However, subsequent studies showed no correlation between probiotic bile tolerance and BSH levels by enzyme knockout experiments (Begley et al., 2005; Moser and Savage, 2001). In addition, genomic analyses suggest no relationship between bile concentration and BSH gene expression (Horackova et al., 2020; Yang et al., 2019). Recently, Jarocki et al. (2014) hypothesized that deconjugated bile salt reacts with other organic substances, producing a hydrogel that can promote the colonization of intestinal microorganisms (Jarocki et al., 2014; Sobotka and Czeczowiczka, 1958). If these experiments can be replicated in vivo, new insights between BSH and probiotics can be processed.

BSH and PVA may be considered moonlighting proteins because of similar structures. Moonlighting proteins are defined as the same enzyme that performs more than one distinct action (Jeffery, 2018). However, the results of the experiment by Kumar et al. (2006) show that the enzymes presumed to be BSH or PVA from Bacillus sphaericus, Clostridium perfringens, and Bifidobacterium longum have only about 30% peptide similarity (not moonlighting protein). Otherwise, BSH and PVA are classified as choloylglycine hydrolases (CGH) within the N-terminal nucleophilic (Ntn) hydrolase enzyme superfamily (Daly et al., 2021).

These two enzymes, which appear to be similar only in structure, can hydrolyze each other’s substrates. There are genomic analyses for this phenomenon. O’Flaherty et al. (2018) reported that Lactobacillus gorillae, Lactobacillus frumenti, Lactobacillus vaginalis, Lactobacillus panis, Lactobacillus antri, Lactobacillus agilis, Lactobacillus salivarius, and L. plantarum strains are simultaneously active against bile acids and penicillin. For in vitro tests, Lambert et al. (2008) reported that L. plantarum WCFS1 has four bsh genes, including bsh-1, bsh-3, and bsh-4, which possess BSH activity. In contrast, bsh-2, bsh-3, and bsh-4 showed PVA activity, with bsh-3 showing the strongest activity. Furthermore, Kusada et al. (2022a) reported that L. paragasseri JCM 5343 bsh-A showed common substrate specificity for PVA.

However, because Lactobacillus was the primary focus of these results, it was necessary to compare strains belonging to Bifidobacterium or Enterococcus. For advanced data, phylogenetic analysis based on BSH peptide sequences obtained from the NCBI or ATCC databases was performed to determine the conserved domains of BSH. A phylogenetic tree was constructed using the data described in Table 2. The bacteria information source is based on Table 1 and searched against ATCC and NCBI databases. A Neighbor-Joining tree was constructed using the Jukes-Cantor model with uniform rates and bootstrap replications of 1,000 datasets using MEGA-11 software. Nodes farther apart are genetically distant, while genes on the same bridge are phylogenetically closer (Fig. 3).

Fig. 3. BSH phylogenetic tree of Lactobacillus and 29 other species based on the BSH gene. The orange line represents the gene of Lactobacillus, the blue line represents Bifidobacterium, the green line represents Enterococcus, the purple line represents Bacteroides, and the black line represents Closteridium. The orange boxes indicate genes that simultaneously or preferentially have PVA activity, and black boxes indicate genes from strains that are judged to be highly likely also to exhibit PVA activity. BSH, bile salt hydrolase; PVA, penicillin V acylase.

Download Original Figure

To explain Fig. 3, genes from Lactobacillus, Bifidobacterium, Enterococcus, Bacteroides, and Clostridium were represented using orange and blue lines. BSH genes that simultaneously or preferentially exhibit PVA activity were indicated using orange boxes, and BSH genes from strains that are highly likely to exhibit PVA activity were denoted using black boxes because they share the same bridge.

The reconstructed phylogeny suggests that Lactobacillus may have originated the BSH genes of Enterococcus and Bifidobacterium, in line with reports from Morinaga et al. (2022). Several clusters showed highly similar genetic distances from each other: Cluster 1 (Lactobacillus casei ASCC 1521, L. rhamnosus ASCC 290, and L.rhamnosus GG), and Cluster 2 (L. plantarum WCFS1-3, L. plantarum LP91, and L.plantarum 299V), and Cluster 3 (L.acidophilus ATCC 4356-A, L.acidophilus ATCC 4357-A, and L. acidophilus LA1-A), and Cluster 4 (L.acidophilus ATCC 4356-B, L.acidophilus ATCC 4357-B, and L.acidophilus LA1-B). The sequence identity percentage of the BSH gene was identical in each group. There is likely no affinity difference and the enzymes function similarly.

Table 1 shows differences in bile acid affinity within the same cluster, as revealed by the analysis. Within cluster 1, L. rhamnosus ASCC 290 and L. rhamnosus GG exhibited a strong affinity for GCA, whereas L. casei ASCC 1521 demonstrates a strong affinity for TCA. L. plantarum LP91, L. plantarum 299V and L. plantarum WCFS1-3 exhibited a strong affinity for GCA within cluster 2. However, L. plantarum 299V, and L. plantarum WCFS1-3 exhibited a strong affinity for GDCA, whereas L. plantarum LP91 demonstrates a strong affinity for TDCA. The BSH genes of L. acidophilus ATCC 4356, L. acidophilus ATCC 4357, and L. acidophilus LA1 in Cluster 3 exhibited a completely identical gene sequence with 100% similarity. Cluster 4 exhibited the same characteristics. Therefore, in either Cluster 3 or Cluster 4, the BSH should display identical affinities. However, comparing clusters 3 and 4 was challenging because an established BSH affinity for each substrate was lacking, as indicated in Table 1.

We conducted a comprehensive analysis to draw more conclusive interpretations of BSH and PVA activities from various LAB. Particularly with bsh genes that encode enzymes with PVA, such as L. paragasseri JCM 5343T-A and L. plantarum WCFS1-3, are important (Kusada et al., 2022a; Lambert et al., 2008). In addition, in Cluster 2, L. plantarum WCFS1-3, L. plantarum LP91, and L.plantarum 299V had highly similar nucleotide sequences. Therefore, it was necessary to determine whether LP91 and 299V can also metabolize penicillin.

The phylogenetic tree significantly correlated the BSH across the same species, as demonstrated in the previous chapter. It is important to analyze genetically connected, but the active site of these enzymes is also important. The active site is predicted using point mutations as explained in Chand et al. (2018). Most active sites reported for BSH appear to be highly conserved.

To determine whether this fact appears not only in Lactobacillus but also in Bifidobacterium or Enterococcus, sequence alignment was performed based on the active site. Chand et al. (2018) confirmed the active site of BSH by using point mutations in a predicted region. Most active sites reported for BSH appear to be highly conserved. In B. bifidum, Cys-2, Arg-18, Asp-21, Asn-72, Asn-173, and Arg-226 are predicted to be the residues involved in active sites (Kim et al., 2004a; Song et al., 2019). Regarding this, the BSH gene sequence was aligned and compared with the peptides described above using MEGA. Except for Bacteroides fragilis ATCC 25285-B, which changed Asp-21 to Glu-43 as shown in Fig. 2, all samples shared the same active site.

These results confirm the homogeneity of almost all active sites, making it difficult to distinguish between PVA and BSH based on this aspect. Finding the difference between the two is very important for understanding the identity and mechanism of BSH in the future. Meanwhile, Avinash et al. (2016) reported that two Trp residues (at positions 23 and 87, based on B. bifidum) of PVA were important for interactions having the benzene ring of penicillin. In addition to Trp, Phe, and Tyr (benzene ring amino acids) were discovered in PVA’s peptide sequence of PVA (Chand et al., 2018; Daly et al., 2021; Suresh et al., 1999).

Based on this theory, Fig. 2 is analyzed additionally. The bsh-A gene of L. paragasseri strain JCM 5343^T and the second, third, and fourth bsh genes of L. plantarum WCFS1 showed experimentally verified PVA activity. The 2^nd, 3^rd, and 4^thbsh genes in L. plantarum WCFS1 have conserved Phe-23 and Tyr-87 residues. In contrast, the first bsh gene in L. plantarum WCFS1 did not exhibit PVA activity and contained only one benzene ring amino acid (Tyr-22 and Asn-87). However, the bsh-A gene of L. paragasseri strain JCM 5343^T contains Val-23 and Asn-87 except for the benzene ring amino acid, with an affinity for ampicillin (Kusada et al., 2022a). These results make it difficult to conclude that the amino acids at positions 23 and 87 of the benzene ring are critical for PVA activity. Therefore, further research is vital and required to determine which amino acid sequence produces PVA activity.

If BSH is intended for use in medication, postbiotics (cell-free supernatants and soluble factors secreted by live bacteria; Martyniak et al., 2021) or an overexpression protocol can be used without the genetic manipulation of probiotics. However, postbiotics safety has not yet been verified (Zhong et al., 2023). Also, BSH has unknown risks proved by in vitro and in vivo analyses.

Several studies reported that DCA, in reaction to BSH, can be a causal factor for CRC. Analysis revealed that patients with CRC had high levels of secondary bile acids in their large intestine (Aguirre et al., 2022; Choi et al., 2015; Sun et al., 2023). Ajouz et al. (2014) reported that excessive concentration of secondary bile acids that pass into the large intestine may cause inflammatory responses, cell membrane destruction, and DNA damage. This affects the intestinal cells, ultimately leading to CRC.

Another potential risk is that BSH may act as an antibiotic resistance factor, given that both BSH and PVA belong to the CGH family. In this regard, investigation of the active site or peptide sequence holds little relevance in distinguishing BSH with PVA activity. The PVA enzyme inhibits penicillin activity, allowing bacteria to survive in the presence of the said antibiotics (Lambert et al., 2008; Sunder et al., 2017). Kusada et al. (2022a) reported that L. paragasseri JCM 5343^Tbsh-A has AMR by BSH activity. Furthermore, Lambert et al. (2008) reported that L. plantarum WCFS1 bsh-2, bsh-3, and bsh-4 showed PVA activity, with bsh-3 showing the strongest activity. However, these phenomena were observed mostly in vitro.

Until now, only the negative effects of secondary bile acids produced by BSH have been highlighted. Studies have indicated that certain intestinal diseases are caused by an imbalance of secondary bile acids. Diversity of intestinal microorganisms is needed for a healthy BSH pool, which therefore balances the secondary bile acids. In several studies, patients with inflammatory bowel disease (IBD) had significantly reduced amounts of secondary bile acids, DCA and LCA (Fiorucci et al., 2021; Heinken et al., 2019; Larabi et al., 2023). Ultimately, the key is to prevent excessive formation of secondary bile acids.

Complete inhibition of BSH activity would eliminate its cholesterol-lowering effect in the blood, which would be a disadvantage in various aspects. Instead, preventing the conversion of primary deconjugated bile salt to secondary deconjugated bile salt could effectively maintain an appropriate amount of primary and secondary bile acids. Bustos et al. (2018) reported that the 7α-dehydroxylase of gut bacteria removes the 7α-hydroxy group and converts primary deconjugated bile salts to secondary deconjugated bile salts (Fig. 1). However, LAB do not have this function. According to Takahashi and Morotomi (1994), bacterial genera used as probiotics (bifidobacteria and lactobacilli) cannot dehydrogenate primary deconjugated bile salts. Thus, if we can control the 7α-dehydroxylation pathway, we will also be able to suppress the formation of excessive secondary bile acids.

To summarize, the effects of BSH need to be studied further for safe use. While the suppression of BSH generally has positive effects, it can also lead to various side effects, underscoring the importance of mitigating methods to control the 7-dehydroxylation pathway.

Multiple BSH genes and their surrounding regions showed minimal sequence similarity, indicating that BSH is the product of horizontal gene transfer. Furthermore, insertion into similar regions and the existence of mobile genetic markers support this theory (Daly et al., 2021). The PVA enzyme inactivates penicillin activity, allowing bacteria to survive in the presence of antibiotics (Lambert et al., 2008; Sunder et al., 2017). BSH and PVA are difficult to distinguish based on structural or peptide sequence differences because BSH and PVA share about 30% sequence similarity with each other, but in fact, about 30% similarity is also found among BSHs across different species (Kumar et al., 2006). Also show different substrate specificities for each strain (Lambert et al., 2008). Understanding the causes of these characteristics is important for future gut microbiome research.

Many studies have examined the correlation between IBD and CRC, the amount of BSH, and the proportion of bile acid (Evangelakos et al., 2021; Fiorucci et al., 2021; Heinken et al., 2019; Larabi et al., 2023; Perez and Briz, 2009; Sun et al., 2023). However, it was only found that representative microorganisms produced active BSH and did not measure the colonic pH of healthy people and patients or sufficiently investigate the composition of LAB. A recent study measured the real-time colonic pH and showed that the normal cecal pH is approximately 5.5, and the large intestine is approximately pH 5.5 to pH 7 (31, 32). If the production of deconjugated bile salt occurs at the beginning of the large intestine, the pH at which it precipitates sufficiently matches this. Therefore, further research is needed on changes in the ratio of intestinal LAB, colonic pH, and the toxicity of deconjugated bile salt.

Recently, a published paper showed that BSH with different substrate specificities towards glycine- and taurine-bile acids has differences in how the microorganism regulates cholesterol (Zhao et al., 2024). This study explains the reason for the difference in bile salt affinity, which was previously difficult to interpret solely based on the BSH sequence. However, the lack of relevant papers necessitates sufficient verification, allowing us to directly design cholesterol control mechanisms in patients or healthy individuals using probiotics.

To date, most studies on BSH have focused on those involving LAB, especially Lactobacillus. However, understanding the relationship between BSH and PVA, the evolutionary history of BSH and PVA, and the biological flow of genes requires a deeper understanding of the relationship between BSH and various microorganisms. Therefore, in addition to Lactobacillus, Bifidobacterium, and Enterococcus, research on the BSH of Listeria, Clostridium, and other microorganisms is required.