Bile Salt Hydrolase-Competent Probiotics in the Management of IBD: Unlocking the “Bile Acid Code”

Bile acid (BA) species and the gut microbiota (GM) contribute to intestinal mucosa homeostasis. BAs shape the GM and, conversely, intestinal bacteria with bile salt hydrolase (BSH) activity modulate the BA pool composition. The mutual interaction between BAs and intestinal microorganisms also influences mucosal barrier integrity, which is important for inflammatory bowel disease (IBD) pathogenesis, prevention and therapy. High levels of secondary BAs are detrimental for the intestinal barrier and increase the intestinal inflammatory response and dysbiosis. Additionally, a lack of BSH-active bacteria plays a role in intestinal inflammation and BA dysmetabolism. Thus, BSH-competent bacteria in probiotic formulations are being actively studied in IBD. At the same time, studies exploring the modulation of the master regulator of BA homeostasis, the Farnesoid X Receptor (FXR), in intestinal inflammation and how this impacts the GM are gaining significant momentum. Overall, the choice of probiotic supplementation should be a peculiar issue of personalized medicine, considering not only the disease but also the specific BA and metabolic signatures of a given patient.


Introduction
The entire gastrointestinal tract hosts the gut microbiota (GM), currently considered as an additional essential organ for living organisms. The GM's physiological functions range from fermentation of non-digestible dietary fibers, to production of critical metabolites and immune-modulatory functions. The whole GM consists of different types of microorganisms, including bacteria, virus, fungi and even parasites. In the last two decades, studies have consistently shown the involvement of a dysbalanced gut bacterial composition in the pathogenesis of several diseases, while the role of virus and fungi is just starting to emerge. Intestinal bacteria could be categorized in six main phyla: Bacteroidota and Firmicutes make up to 90% of the total bacterial composition, while the remaining part is composed of Proteobacteria, Fusobacteria, Actinobacteria and Verrucomicrobia [1]. The host organism and its resident microbiota are in a constant mutual communication, regulated by sophisticated rules. A disproportionate number, reduced diversity or unbalanced communities of microorganisms (namely dysbiosis) decrease the ability of the host to resist infections, absorb nutrients, synthesize vitamins, and/or carry out peculiar metabolic activities [2], thereby contributing to the onset and progression of disease of the gut-liver axis [3], inflammatory [4], metabolic or autoimmune conditions [5] and even neurodegenerative disorders [6][7][8]. Despite the 'big unknown' in the GM field, it is universally accepted that the microbiome is unique for each individual and substantial differences are present even in identical twins [9]. In fact, the GM composition is not only determined genetically, but it is profoundly impacted by multiple factors including epigenetics, nutrition, drug administration, physical exercise, and general lifestyle.
In terms of metabolic functions, the GM produces short-chain fatty acids (SCFAs) [10] and synthesize vitamins, e.g., vitamin K and constituents of the vitamin B [11][12][13]. SCFAs, beta (KLB), initiating a phosphorylation signaling cascade, ultimately resulting in CYP7A1 inhibition [31][32][33], hence suppressing de novo hepatic BA synthesis. In addition to this well-characterized pathway, BA-activated FXR also modulates BA cytotoxicity, aiming at creating a balanced BA pool that in physiological conditions and under proper nutritional cues does not harm the intestinal cell physiology. BA-activated FXR has also been shown to have anti-inflammatory [34][35][36] and antitumorigenic properties [37][38][39], and actively participates in the maintenance of intestinal barrier integrity [34,40]. Additionally, FGF19 was shown to have similar functions in mice and induce a reshape of the gut microbial community in the presence of an inflammatory trigger [41]. The ability of FXR and FGF19 to shape the circulating BA pool size and composition is critical for the relation BAs have with the GM community.

Bile Acid Deconjugationand BSH-Competent Microorganisms
The integrated metabolism of the BA pool and GM is among the most complex transgenomic biochemical interactions between the host and symbiotic microorganisms inhabiting it [42]. If one considers BA modulatory activities on cholesterol, triglycerides, glucose and energy homeostasis (reviewed in [43]), it is intuitive that gut bacteria-dependent BA biotransformation have crucial biological implications. Primary BAs synthesized and conjugated to glycine or taurine before secretion into the bile canaliculi are then transformed into secondary BAs, mainly deoxycholic acid (DCA), lithocolic acid (LCA) and β-murideoxycholic acid (β-MDCA; in mice) via bacteria with a peculiar activity: the bile salt hydrolase (BSH) activity [22,44,45]. The BSH catalyzes the "gateway" reaction in the bacterial metabolism of conjugated BAs hydrolyzing the C-24 N-acyl bond conjugating BA to glycine or taurine and is a prerequisite for the subsequent 7α/β-dehydroxylation [44,[46][47][48] and 7α-dehydrogenation, ultimately resulting in BAs deconjugation and epimerization. GMdependent BA metabolization mostly occurs in the terminal ileum and proximal colon, the focal colonization site of bacterial strains with BSH activity [22].
In addition to the metabolic activities due to BSH-competent bacteria, there are other enzymatic biotransformations performed by gut microbes, such as the 7α/β-dehydroxylation, pyridine nucleotide-dependent hydroxysteroid dehydrogenization (operated by HSDHs enzymes), and conjugation of the cholate backbone with the aminoacids tyrosine, leucine and phenylalanine. The 7α/β-dehydroxylation is a multistep biotransformation carried out on free BAs and performed by a range of BA-inducible (bai) and bai-like genes [68]. Bai genes convert CA and CDCA to DCA and LCA, respectively, and have been identified in multiple species, mostly in Faecali catena contorta S122, Clostridium hiranonis and Clostridium scinden in humans [69] and Eggerthella lenta [70,71]. HSDHs performs the oxidation/reduction of hydroxy groups at the carbon atoms in positions 3, 7, and 12 of BAs [72], while the amide conjugation of cholate with aminoacids was recently found to be associated with Clostridium bolteae in mice fed a high-fat diet [73]; however, it is not currently clear what the function of such modified BA is. FXR is differentially activated by the different BA species and their metabolites. Chenodeoxycholic Acid (CDCA) is the major FXR activator, followed by deoxycholic acid (DCA), lithocolic acid (LCA) and cholic acid (CA) [74]. Therefore, intestinal microorganisms not only act directly on BA species, but they also indirectly influence FXR transcriptional activity via the modulation of BA metabolism [75]. As a consequence, BA transformation by BSH-competent bacteria shapes the magnitude of BA-induced FXR activation in the gut [75][76][77][78]. Specifically, the major metabolic action of this specific group of bacteria is therefore to deconjugate intraluminal BAs, thus reducing the amount of conjugated BA (T/GCA and T/GCDCA) that can be absorbed by the enterocytes via the intestinal bile acid transporter (IBAT). These events translate into a de-activation of intestinal FXR (due to lower amounts of intracellular ligands) coupled with a significant increased fecal BA loss (up to 3 fold) ( Figure 1). In this respect, one could speculate that the role of BSH-competent enteric bacterial would be similar to that of cholestyramine, which is able to bind intraluminal BA and increase their fecal release. Intriguingly, resins such ascholestyramine are nowadays used in clinic in BA-induced diarrhea in IBD and eventually BSH probiotics could complement this therapeutic strategy.
performed by a range of BA-inducible (bai) and bai-like genes [68]. Bai genes convert CA and CDCA to DCA and LCA, respectively, and have been identified in multiple species, mostly in Faecali catena contorta S122, Clostridium hiranonis and Clostridium scinden in humans [69] and Eggerthella lenta [70,71]. HSDHs performs the oxidation/reduction of hydroxy groups at the carbon atoms in positions 3, 7, and 12 of BAs [72], while the amide conjugation of cholate with aminoacids was recently found to be associated with Clostridium bolteae in mice fed a high-fat diet [73]; however, it is not currently clear what the function of such modified BA is.
FXR is differentially activated by the different BA species and their metabolites. Chenodeoxycholic Acid (CDCA) is the major FXR activator, followed by deoxycholic acid (DCA), lithocolic acid (LCA) and cholic acid (CA) [74]. Therefore, intestinal microorganisms not only act directly on BA species, but they also indirectly influence FXR transcriptional activity via the modulation of BA metabolism [75]. As a consequence, BA transformation by BSH-competent bacteria shapes the magnitude of BA-induced FXR activation in the gut [75][76][77][78]. Specifically, the major metabolic action of this specific group of bacteria is therefore to deconjugate intraluminal BAs, thus reducing the amount of conjugated BA (T/GCA and T/GCDCA) that can be absorbed by the enterocytes via the intestinal bile acid transporter (IBAT). These events translate into a de-activation of intestinal FXR (due to lower amounts of intracellular ligands) coupled with a significant increased fecal BA loss (up to 3 fold) ( Figure 1). In this respect, one could speculate that the role of BSH-competent enteric bacterial would be similar to that of cholestyramine, which is able to bind intraluminal BA and increase their fecal release. Intriguingly, resins such ascholestyramine are nowadays used in clinic in BA-induced diarrhea in IBD and eventually BSH probiotics could complement this therapeutic strategy. We have previously shown that enrichment of the gut milieu with BSH-competent species, such as Lactobacilli and Bifidobacteria, with VSL#3 probiotic administration in mice We have previously shown that enrichment of the gut milieu with BSH-competent species, such as Lactobacilli and Bifidobacteria, with VSL#3 probiotic administration in mice limits the availability of efficient FXR ligands and produces significant alterations in BA homeostasis that translates into enhanced fecal BA excretion, hepatic BA synthesis and biliary output [79], similar to what happens in mice receiving the BA sequestrant coleselavam [80]. Intriguingly, this phenomenon is abrogated in both Fxr-and Fgf15-null mice [79], indicating that the activation of the Fxr-Fgf15 duo is required for probiotic BSHcompetent bacteria to promote fecal unconjugated BA loss and hepatic BA synthesis. For all these reasons, understanding how bacterial species affects BA metabolism and the FXR-FGF15/19 duo in the gut liver axis is a very relevant question gaining significant momentum from studies demonstrating that dysbiosis and BA dysmetabolism are associated with intestinal diseases.

Bile Acid-Dependent Shaping of Intestinal Microbial Community
The mutual interaction between the bileome and intestinal microorganisms has a direct influence on the integrity of the intestinal epithelial barrier. In fact, BAs flowing along the intestine control bacterial overgrowth and protect against intestinal mucosal damage. In particular, high levels of hydrophobic BAs have a direct antimicrobial activity mainly causing damage to the bacterial membrane [81,82]. On the one hand, Gram-negative bacteria, especially those belonging to Bacteroides, have been found to be more sensitive to free BAs [70,83]; on the other hand, Gram-positive bacteria seem to be more sensitive to tauro-and glyco-conjugated and primary BAs [59].
The tight interconnection between the GM and BA metabolism goes beyond BA intrinsic antimicrobial properties, as BAs indirectly shape the composition of GM by promoting the expression of genes fostering innate defense via FXR activation [40]. A lower level of BAs as a consequence of clinically and/or experimentally induced liver injuries can trigger intestinal bacterial overgrowth [84][85][86]. For example, surgical closure of the common bile duct in rodents results in bacterial overgrowth and translocation across the intestinal epithelial barrier, disruption of its integrity, and systematic infection that can be reversed by BA administration [87,88]. In the presence of an obstruction of bile flow, abnormal hepatic BA accumulation occurs and intestinal FXR signaling is abrogated [40]. In addition, devoiding the intestinal lumen of BA has a detrimental effect on the intestinal microbial community and causes mucosal injury [89]. Administration of BAs to both rodents and humans in conditions of biliary obstruction [88,[90][91][92], as well as FXR activation [40] or FGF19 administration [41] in the presence of intestinal inflammatory triggers, counteracts intestinal mucosal damage and bacterial overgrowth and translocation. Rcent data from our group show that in a murine model of obstructed BA flow, FXR activation by its novel specific ligand TC-100 prevents intestinal mucosal damage, goblet cells and mucus depletion, and increases the expression of tight junctions by the mutual modulation of the GM and BA pool size and composition [93]. Both the peculiar enrichment of Akkermansia muciniphila and associated reduction in CA levels contribute to these beneficial effects [93]. BA-dependent activation of the Fxr-Fgf15/19 axis and the consequent inhibition of Cyp7a1, hence BA synthesis, is affected in rodents treated with antibiotics, germ-free and gnotobiotic animals [75,[94][95][96][97][98] and mice lacking Fxr display ileal bacterial overgrowth [40] and a compromised epithelial barrier already at basal state [34,40].
Administration of the FXR-specific agonist obeticholic acid (OCA) in healthy volunteers was shown to inhibit endogenous BA synthesis while concomitantly enriching the GM with a number of Gram-positive strains, namely Streptococcus thermophilus, Lactobacillus casei and paracasei, Bifidobacterium breve, and Lactococcus lactis, boosting the notion that changes in the endogenous BA pool size and composition result in an FXR-dependent reshaping of the gut microbial composition [99]. In a mouse model of spontaneous colorectal cancer onset, the Apc min/+ mouse model, administration of a diet supplemented with CA increased the abundance of opportunistic pathogens such as Prevotella and Desulfovibrio while decreasing the presence of beneficial bacteria, including Ruminococcus, Lactobacillus, and Roseburia [100]. A similar diet administration in rats, induced alterations in the composition of the GM; the phylum of Firmicutes predominated at the expense of Bacteroidetes and an expansion of several strains in the classes Clostridia and Erysipelotrichi occurred [76]. Concomitantly, the administered CA was efficiently transformed into DCA by a bacterial 7α-dehydroxylation reaction [76], suggesting that DCA could boost the reshaping of the GM composition due to its powerful antimicrobial activity. In line with this, mice fed with a diet supplemented with DCA display a lower abundance of BSH-competent Lactobacillus, Clostridium XI, and Clostridium XIV and a concomitant enriched presence of Parabacteroides and Bacteroides compared to controls [101]. Taken together, these data demonstrate that BA supplementation reshapes the composition of gut microbiota.

Bile Salt Hydrolase Activity and Probiotics
Dietary supplementation with probiotics was demonstrated to modify not only hosts' BA pool but also their metabolism. BSH-competent bacteria in probiotic formulations are being actively studied for their BA composition-modifying properties and their ability to reduce serum cholesterol levels [102]. In fact, BSH-dependent BA deconjugation decreases the recycling of BAs promoting their de novo synthesis, thus reducing serum cholesterol levels [102]. Additionally, bacteria with BSH activity metabolize aminoacids produced from deconjugation as carbon, nitrogen and energy sources [44]. Using germ-free and conventionally raised mouse models, Joyce et al. demonstrated that gastrointestinal expression of BSH promotes BA deconjugation with concomitant alteration of the host's immune homeostasis, energy metabolism and circadian rhythm [103]. In conventionally raised mice, BSH activity reduced weight gain, serum cholesterol, and hepatic triglyceride levels [103]. Expression of cloned BSH enzymes in the gastrointestinal tract of gnotobiotic or conventionally raised mice exhibited a gene expression profile characterized by Clock and Arntl induction, genes involved in circadian rhythm regulation and energy expenditure [103]. Mukherji et al. observed that the GM, Toll-like receptor expression on epithelial cells and circadian rhythm are tightly connected and perturbations of this interaction promote the development of a pre-diabetic syndrome [104]. Administration of the BSH-competent Lactobacillus reuteri NCIMB 30242 ® in human healthy volunteers improves unconjugated plasma BA profile highlighting the ability of this specific probiotic to modulate BA metabolism in humans [102]. Several studies conducted in mice and humans demonstrated the involvement of probiotics with Lactobacillus species such as Lactobacillus fermentum K73 and Lactobacillus plantarum 299v in the reduction of body weight and cholesterol levels due to their BSH competence [102,105]. Interestingly, it was shown that the yeast strain Saccharomyces Boulardii has BSH activity in vitro [106]. Daily administration of this yeast in hamsters fed a high-cholesterol diet decreased total plasma cholesterol levels and modified the GM composition [107]. In humans, administration of Saccaromices Boulardii for 8 weeks reduced remnant lipoprotein concentration, a predictive biomarker and potential therapeutic target in the prevention/treatment of coronary artery disease [108]. Additionally, mice administered with the probiotic VSL#3, containing the BSH-competent Lactobacillus acidophilus and Bifidobacterium infantis strains, modulates the BA profile by upregulating hepatic de novo BA synthesis [79] and their fecal excretion [79].

BSH-Competent Bacteriaand Probiotics Administrationin Inflammatory Bowel Disease
Patients affected by inflammatory bowel diseases (IBD) display low levels of BSH activity in the gut microbiota [109,110]. IBD is a chronic inflammatory condition of the gastrointestinal tract characterized by a dysregulation of the gut mucosal immune functions and a dysbiotic GM occurring in genetically susceptible hosts. IBD encompasses two major phenotypes: ulcerative colitis (UC) and Crohn's disease (CD), which present with overlapping symptoms, despite being two distinct pathologies.
Previously published data have shown how secondary BAs are detrimental and coresponsible for the disruption of intestinal barrier integrity, increase of the inflammatory response and dysbiosis [101,111,112]. Interestingly, rectal instillation of DCA/LCA in different murine models resulted in anti-inflammatory benefits, partially due to TGR5 signaling [113], indicating a dual pro-and anti-inflammatory effect of DCA in different experimental contexts and administration routes. This strongly suggests that caution should be taken when translating mouse data to IBD patients' bedside.
Analysis of data collected on an IBD multiomics database (The Integrative Human Microbiome Project), revealed that BAs are a prominent component of the network in relation to changes in the microbiome [114]. In particular, primary BAs (specifically CA, TCA and GCA) levels were higher, while secondary BAs levels were lower [114]. This was associated with an enrichment of Roseburia and strongly indicated a lack of BSH-competent bacteria IBD-related dysbiosis [114]. Franzosa et al. also demonstrated that levels of primary BAs were higher in IBD patients and potentially associated with the coherent microbial groups, while secondary BA levels were lower in CD patients [115]. In line with this, patients with UC pouch displayed low levels of secondary BAs and high levels of CDCA and this was consistently associated with lower levels of Ruminococcaceae and bai genes [113], indicating once again a lack of BSH-competent bacteria, hence a dysbiosisdependent disturbance of BA metabolization. Beside unravelling the intertwined relation between BA species and BSH-competent bacteria in IBD pathogenesis, its clinical and therapeutic relevance is also being actively studied. Omics data revealed that specific GM composition in IBD patients predicts early remission response to anti-cytokine and anti-integrin therapies, via gut bacterial clusterization associated with secondary BAs enrichment [71]. Probiotic supplementation in these patients may represent a useful tool for the prevention, treatment and remission of the acute phase. In IBD, probiotics have a beneficial impact when are able to modulate the immune response, gut barrier integrity and GM composition [116]. Several studies have been conducted in animal models and IBD patients to find suitable probiotics in IBD, but a universal consensus in not available yet.
Data on the GM composition of IBD patients indicated a lower level of Firmicutes and a higher level of Proteobacteria [117]. Consistent with this, recent data have shown that BSH-competent Proteobacteria are increased, while BSH-competent Firmicutes are decreased in IBD patients [118]. In vitro studies and mouse models have shown that Faecalibacterium prausnitzii and Lactobacillus rhamnosus CNCM I-3690 ameliorate gut barrier integrity by increasing the expression of occludin and cadherin proteins [119,120]. Experimental colitis caused by Trinitrobenezenesulphonic acid (TNBS), increased Escherichia coli and Clostridium strains and reduced Bifidobacterium and Lactobacillus populations [121]. On the contrary, Bifidobacterium supplementation in murine experimental colitis, seems to have beneficial effects and decrease inflammation [122][123][124][125]. Bifidobacterium infantis administration reduced symptoms of colitis such as colonic edema and weight-loss, protected the epithelial cell layer, prevented goblet cell loss and modulated cytokine levels, specifically increasing the anti-inflammatory cytokine IL-10 and decreasing the pro-inflammatory one IL-1β [123,125,126]. In Sodium Dextran Sulphate (DSS)-induced murine colitis, administration of Bifidobacterium animalis and Bifidobacterium longum ameliorated inflammatory symptoms, preserved the colonic structure and decreased intestinal epithelial cell apoptosis and TNF-α levels [122,124]. Furthermore, in a TNBS-induced colitis mouse model, the influence of Bifidobacterium animalis and Lactobacillus reuteri administration on the immune system was studied [127]. Authors observed that Bifidobacterium animalis beneficial effects on the host were linked to the bone marrow-derived dendritic cell development and IL-17A secretion, whereas Lactobacillus reuteri administration promoted tolerogenic dendritic cells and Tregs population development [127]. Interestingly, two different formulations of VSL#3 have shown divergent metabolic and anti-inflammatory activities in mice subjected to DSS and TNBS colitis [128].
Clinical trials are ongoing in IBD patients with the aim to evaluate the effects of probiotics on the disease. In UC patients, several studies indicated that different Bifidobacterium strains are able to reduce pro-inflammatory interleukins and C-reactive protein (CRP) levels, sustaining the remission phase [129][130][131][132][133]. On the other hand, in CD patients, the administration of Saccaromyces boulardii promoted the maintenance of the remission phase [134]. Probio-Tec AB25, a cocktail of Lactobacillus acidophilus strain LA-5 and Bifidobacteriumanimalis subsp. lactis BB12, was administered to 32 UC patients, with no beneficial effects in maintaining remission [135]. VSL#3 is among the most common probiotic mixtures of proven efficiency on human IBD, and it contains different Lactobacilli, Bifidobacteria and Streptococcus thermophilus strains [136,137]. In a double-blind placebo-controlled trial, patients with the active form of UC received VSL#3 twice daily for 12 weeks. Authors observed remission characterized by the reduction of stool frequency and rectal bleeding [136]. In another clinical trial, VSL#3 was used as an adjuvant added to patients' pharmaceutical treatment for 8 weeks. A reduction in rectal bleeding was observed, although no differences were found in endoscopic inflammatory scores, highlighting the importance of long-term use of probiotics [137]. Furthermore, the administration of VSL#3 in children with UC determined a significant remission of patients due to a reduction of IFN-γ and TNF-α levels together with a modulation of the GM [138]. Conversely, VSL#3 treatment in CD patients does not seem to have clear benefits [139], although further research is warranted. In these patients, supplementation with Faecalibacterium prausnitzii, instead, exhibited anti-inflammatory effects [140]. Tables 1 and 2 summarize the ongoing and/or completed clinical trials in IBD patients administered with either probiotics or microbial metabolites.

Conclusions and Discussion
BA species and intestinal microorganisms both contribute to intestinal fitness, are tightly interconnected and communicate directly with each other and indirectly via mainly FXR and related signaling pathways. Despite the evident association between BSHcompetent bacteria and BA pool composition, metagenomic analysis of IBD patients' GM revealed that not all BSH-competent species are created equal and the fact that any GM metabolic activity is strainspecific should always be remembered. Therefore, further studies on probiotics administration and their effect on gut microbial reshaping and consequent BA metabolism modulation are needed. At the same time, these studies should be paralleled by the generation of data on how microbial reshaping can be achieved dueto the modulation of the BA pool size and composition, as it occurs when FXR-or FGF19-based drugs are used. Clinical use of probiotics containing BSH-competent species is currently ongoing in patients with IBD, pouchitis and diarrhea and could be also relevant in patients with leaky gut secondary to liver disease (such as intrahepatic cholestasis) or colorectal cancer. On the other hand, a few considerations about limitations and potential adverse effects related to the indiscriminate use of probiotics resulting in high levels of unconjugated bile acids, lipid malabsorption and steatorrhea should be made. Elevated concentrations of DCA and LCA, produced from unconjugated BAs, are considered a risk factor for colorectal cancer [149]. Experimental evidence hasdemonstrated that BSH activity promotes gut colonization by pathogenic bacterial strains such as Listeria monocytogenes and Brucella abortus [63]. Moreover, mice and humans display different BA profiles characterized by high levels of tauro-conjugated BAs and high levels of glyco-conjugated BAs, respectively [59,74]. In this scenario, previous data on the species-specific difference affecting the translation of data from mice to humans and confounding the selection of probiotics are actively being validated in clinical trials. The field is new and a greatereffort should be made in integrating the enormous amounts of data generated by all the available omic platforms. Overall, the choice of probiotic supplementation should be considered a peculiar issue of the personalized medicine, considering not only the disease but also the specific BA and metabolic signature of a given patient.