Gut Microbiome and Bile Acid Metabolism in Liver Disease: Mechanisms, Clinical Implications, and Therapeutic Opportunities

Abstract

The intricate interplay between the gut microbiome and bile acid metabolism via the gut-liver axis is fundamental to hepatic homeostasis. Perturbations in this axis are increasingly implicated in the pathogenesis of diverse liver diseases, including metabolic dysfunction-associated steatotic liver disease, alcohol-associated liver disease, cholestatic liver diseases, and hepatocellular carcinoma. This review integrates current understanding of hepatic bile acid synthesis, enterohepatic circulation, and gut microbial bile acid transformations, detailing how bile acids function as signaling molecules through nuclear receptors including farnesoid X receptor (FXR), pregnane X receptor, vitamin D receptor, constitutive androstane receptor, and G-protein-coupled receptors; G protein-coupled bile acid receptor 1 (GPBAR1 also known as TGR5), and sphingosine-1-phosphate receptor 2. We explore disease-specific alterations in gut microbiota composition and bile acid profiles in MASLD, ALD, cholestatic liver diseases, and liver cancers, focusing on mechanisms linking gut dysbiosis, impaired intestinal barrier function, altered bile acid signaling, inflammation, and immune modulation to liver injury and progression. Furthermore, we discuss the clinical implications, highlighting the potential of microbiome signatures and bile acid profiles as diagnostic and prognostic biomarkers. Therapeutic strategies targeting the gut-liver axis, including probiotics, fecal microbiota transplantation, FXR agonists, and FGF19 analogs, are reviewed. Finally, we address current challenges and future directions, emphasizing the need for multi-omics integration, functional studies, and personalized medicine approaches to leverage the gut-liver axis for improved liver disease management.

I. Introduction

The liver, as the central metabolic organ, plays a critical role in regulating various metabolic processes, including glucose, lipid, energy, and xenobiotic metabolism.^1–3 It is also the primary site for vitamin storage and detoxification.^3,4 Liver diseases are becoming a serious global health concern, affecting millions of lives, and contributing significantly to morbidity, mortality, and healthcare burden worldwide. As of 2023, liver diseases account for approximately two million deaths annually, with a rising incidence linked to lifestyle-related metabolic dysfunction, such as obesity, sedentary lifestyles, type 2 diabetes, and metabolic syndrome, as well as alcohol consumption.^5,6 Among these, metabolic dysfunction-associated steatotic liver disease (MASLD), previously known as non-alcoholic fatty liver disease, and its progressive inflammatory form, metabolic dysfunction-associated steatohepatitis (MASH), have become the most prevalent causes of chronic liver disease and the main risk factors for liver cancer.⁷ Simultaneously, alcohol-associated liver disease (ALD) continues to pose a significant threat, particularly during the COVID-19 pandemic.^8–10 Compared to MASLD and ALD, cholestatic disorders, such as primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC), are much less common, but carry high individual disease burden due to lack of treatment and high risk of hepatobiliary malignancies.¹¹Furthermore, liver cancers, including hepatocellular carcinoma (HCC) and cholangiocarcinoma, frequently arise as complications of chronic liver diseases and are associated with poor prognosis due to late diagnosis and limited therapeutic options.⁶

Accumulating evidence indicates that different liver diseases share a common mechanistic axis the gut-liver axis, despite differences in etiology.¹² Gut microbiome emerges as a central regulator of this axis. In human gut, there are trillions of microorganisms with vast metabolic potential and profound impact on host physiology.¹³ Gut microbiome plays important role in regulating immune homeostasis, nutrient metabolism, and xenobiotic processing.¹⁴ Under normal physiological conditions, the gut and liver communicate dynamically via the portal circulation, which transports microbial metabolites, bile acids, and immune mediators. This bidirectional interaction, known as the gut–liver axis, maintains homeostasis through tightly regulated enterohepatic circulation of bile acids and microbial cross-talk.¹⁵ Disruption of this axis—particularly due to gut dysbiosis, characterized by reduced microbial diversity, functional shifts, and overgrowth of opportunistic organisms—can disturb bile acid metabolism and intestinal barrier integrity, leading to microbial translocation, systemic inflammation, and hepatocellular injury.^15–18 Conversely, liver dysfunction disrupts bile acid synthesis and secretion, resulting in changes in gut microbial composition and function. These bidirectional disturbances form a self-amplifying cycle that contributes to the development and progression of MASLD/MASH, ALD, PBC, PSC, and liver cancers.^19,20

This comprehensive review highlights recent advances in our understanding of the gut microbiome–bile acid–liver axis and their roles in maintaining liver health and driving liver pathologies. We begin by outlining the physiological processes of bile acid synthesis and microbial transformation, followed by an overview of gut microbiota composition in health and disease. We then explore the mechanistic links between dysbiosis, altered bile acid signaling, and liver pathology, focusing on inflammation, immune modulation, and metabolic reprogramming. Finally, we discuss the clinical implications of gut–liver axis disruption, including emerging diagnostic biomarkers and therapeutic interventions such as probiotics, bile acid modulators, and fecal microbiota transplantation. The review concludes with an outlook on current challenges and future directions, emphasizing the integration of multi-omics data and precision medicine approaches for liver disease management.

II. Gut Microbiome and Bile Acid Metabolism

Bile acids are amphipathic molecules synthesized in the liver from cholesterol and play a central role in the emulsification and absorption of dietary lipids.²¹ Beyond their classical function as digestive detergents, bile acids have emerged as potent signaling molecules that regulate glucose, lipid, and energy homeostasis. Increasing evidence highlights the critical role of the gut microbiome in shaping the bile acid pool through enzymatic transformations, thereby influencing both host metabolism and liver health. The reciprocal interactions between the gut microbiota and bile acid metabolism—often referred to as the “microbiome–bile acid axis”—are now recognized as key mediators of the gut–liver axis and central to the pathogenesis of various liver and metabolic diseases.

A. Hepatic bile acid synthesis and enterohepatic circulation

A.1. Hepatic bile acid synthesis

Bile acids are synthesized exclusively in the liver from cholesterol, serving as both detergents for lipid absorption and signaling molecules that regulate lipid, glucose, and energy metabolism. Hepatic bile acid synthesis represents the major catabolic route for cholesterol elimination, accounting for approximately half of daily cholesterol turnover in humans. This process is tightly regulated at multiple levels to maintain bile acid, lipid, and cholesterol homeostasis, and its dysregulation contributes to a spectrum of metabolic and cholestatic liver diseases. The liver synthesizes bile acids through two major pathways: the classic (neutral) pathway and the alternative (acidic) pathway.

In humans, approximately 75–90% of bile acids are produced through the classic pathway, which is initiated in the endoplasmic reticulum by cholesterol 7α-hydroxylase (cytochrome P450 (CYP) 7A1). This enzyme catalyzes the rate-limiting step, converting cholesterol into 7α-hydroxycholesterol. This intermediate is then oxidized at the 3β-hydroxy position by 3β-hydroxy-Δ⁵-C27-steroid oxidoreductase (HSD3B7) to yield 7α-hydroxy-4-cholesten-3-one (C4). C4 serves as a key metabolic intermediate and diagnostic biomarker for rates of bile acid synthesis, representing a branch-point precursor that determines the composition of primary bile acids. Sterol 12α-hydroxylase (CYP8B1) directs the synthesis of 12α,7α-dihydroxy-4-cholesten-3-one, which undergoes side-chain oxidation and peroxisomal β-oxidation to produce cholic acid (CA). In the absence or reduced activity of CYP8B1, 12α-hydroxylation does not occur, and C4 proceeds through the CYP27A1-dependent mitochondrial oxidation pathway, leading to the formation of chenodeoxycholic acid (CDCA), a dihydroxy bile acid. Thus, CYP8B1 activity determines the CA/CDCA ratio, which has important implications for bile acid hydrophobicity, signaling properties, and metabolic regulation (Fig. 1).

Fig.1. Bile acid synthesis pathways in human.

In humans, 75–90% of bile acids are synthesized via the classic pathway in the endoplasmic reticulum, initiated by CYP7A1, which converts cholesterol to 7 α-hydroxycholesterol. Subsequent oxidation by HSD3B7 forms 7α-hydroxy-4-cholesten-3-one (C4), a key intermediate and biomarker of bile acid synthesis. Sterol 12α-hydroxylase (CYP8B1) directs C4 toward CA formation through 12α-hydroxylation, side-chain oxidation, and peroxisomal β-oxidation. In the absence of CYP8B1, C4 is processed via CYP27A1-dependent mitochondrial oxidation to yield CDCA. CYP8B1 activity thus determines the CA/CDCA ratio, influencing bile acid hydrophobicity, receptor signaling, and metabolic regulation. CYP7A1, cholesterol 7α-hydroxylase; CA, cholic acid; CDCA, chenodeoxycholic acid.

The alternative (acidic) pathway represents a complementary route for bile acid synthesis, accounting for approximately 10–25% of total production under normal physiological conditions. Unlike the classic pathway, which is initiated in the endoplasmic reticulum, the alternative pathway begins in the mitochondria. Cholesterol transport to mitochondria by StAR-related lipid transfer domain protein 1 (StarD1) is the rate-limiting step of this pathway.^22,23 The mitochondrial sterol 27-hydroxylase (CYP27A1) catalyzes the oxidation of the cholesterol side chain to generate (25R)-26-hydroxycholesterol (commonly referred to as 27-hydroxycholesterol). This mitochondrial reaction is pivotal, as it introduces a hydroxyl group at the terminal side chain, enabling the subsequent oxidative and shortening reactions that define this pathway. The resulting oxysterol is transported to the ER, where oxysterol 7α-hydroxylase (CYP7B1) introduces a 7α-hydroxyl group, forming 3β,7α,27-trihydroxycholest-5-ene. This intermediate is further oxidized at the 3β-hydroxy position and the Δ⁵ double bond is reduced, producing 3α,7α,27-trihydroxycholestanoic acid, the principal C27 bile acid intermediate. Subsequently, this intermediate enters the peroxisome, where it undergoes β-oxidation of the side chain catalyzed by acyl-CoA oxidase 2, D-bifunctional protein, and sterol carrier protein X.^24,25 These reactions progressively shorten the side chain from C27 to C24, ultimately yielding CDCA—the predominant primary bile acid produced through this pathway.²⁶ Because sterol 12α-hydroxylase (CYP8B1) is not involved, 12α-hydroxylation does not occur, and CA is not synthesized via this route (Fig. 1). Bile acid synthesis is not uniform across all hepatocytes. In mice, the pericentral hepatocytes (those surrounding the central vein) show the highest expression of Cyp7a1, the rate-limiting enzyme for the classic pathway, along with its partner 3β-hydroxy-Δ⁵-C27-steroid oxidoreductase. Cyp8b1 is highly expressed in mid-lobular hepatocytes. By contrast, Cyp27a1, a key enzyme of the alternative pathway, exhibits a broader and less sharply defined zonation pattern across the lobule. This spatial organization is thought to optimize cholesterol catabolism and bile-acid homeostasis under varying metabolic conditions.^27–29

Although bile acid synthesis in rodents follows the same pathways as in humans, important species-specific differences exist in enzyme expression and bile acid composition. In rodents, CDCA is further converted into muricholic acids (MCAs)—primarily α-MCA and β-MCA—through 6β-hydroxylation catalyzed by CYP2C70, a rodent-specific enzyme highly expressed in hepatocytes.³⁰ This reaction introduces an additional hydroxyl group at the 6β-position, increasing bile acid hydrophilicity and reducing cytotoxicity. As a result, the murine primary bile acid pool is dominated by CA, α-MCA, and β-MCA (Fig. 1). ³¹

A.2. Hepatic bile acid conjugation

Conjugation occurs in hepatocytes through two sequential enzymatic reactions that are spatially compartmentalized between the cytosol and peroxisomes. In the first step, bile acids are activated at the cytosolic face of the endoplasmic reticulum to their CoA thioester derivatives by bile acid–CoA synthetase (BACS; solute carrier family (SLC) 27A5) in an ATP- and Mg²⁺-dependent reaction, generating bile acid–CoA intermediates. These activated intermediates are then transported into peroxisomes, where bile acid–CoA:amino acid N-acyltransferase (BAAT) catalyzes the second step by conjugating bile acids to glycine or taurine. This reaction forms an amide bond between the carboxyl group of the bile acid and the amino group of glycine or taurine, producing glycine-conjugated bile acids (e.g., glycocholic acid [GCA], glycochenodeoxycholic acid [GCDCA]) and taurine-conjugated bile acids (e.g., taurocholic acid [TCA], taurochenodeoxycholic acid [TCDCA]) (Fig. 3). ³²

Fig.3. Bile acid conjugations.

In hepatocytes, bile acid conjugation occurs in the cytosol and peroxisomes. BACS activates bile acids to CoA intermediates, which BAAT then conjugates with glycine or taurine, forming glyco- and tauro-conjugated bile acids such as TCA and GCA. BACS, bile-acid-CoA synthase; BAAT, Bile acid–CoA:amino acid N-acyltransferase; TCA, Taurocholic acid; GCA, glycocholic acid.

In rodents, bile acid conjugation differs substantially from that in humans, both in substrate preference and physiological outcome. While humans conjugate bile acids predominantly with glycine (~75%) and to a lesser extent with taurine (~25%), rodents almost exclusively conjugate with taurine. This species-specific preference is driven by differences in hepatic amino acid availability, intracellular pH, and enzyme substrate specificity of BAAT. The murine BAAT enzyme displays higher catalytic efficiency toward taurine, resulting in a bile acid pool composed largely of tauro-conjugated species such as TCA, tauromuricholic acids (TαMCA, TβMCA), and TCDCA.³³

Conjugation significantly lowers the pKa of bile acids (from ~5 to <2 for taurine conjugates and ~4 for glycine conjugates), ensuring that they remain fully ionized and water-soluble at physiological pH. This transformation not only minimizes hepatocellular membrane damage but also enhances their recognition and transport by canalicular export systems, particularly the bile salt export pump [BSEP; ATP-binding cassette subfamily B member 11(ABCB11)]. Once secreted into bile, these conjugated bile salts play essential roles in micelle formation, lipid emulsification, and fat-soluble vitamin absorption in the intestine.³⁴ In addition to conjugation with taurine and glycine, a small portion of bile acids also can be metabolized by phase I cytochrome P450 enzymes such as CYP3A4 and phase II sulfation enzymes such as SULT2A as well as glucuronidation enzymes such as UDP-glucuronosyltransferase (UGT) 1A3 and UGT2B7.

A.3. Bile acid transformation

Following their secretion into the intestinal lumen, bile acids undergo extensive biotransformation by the gut microbiota, which markedly alters their physicochemical properties and biological activities. These microbial modifications expand the bile acid pool and generate metabolites that act as signaling molecules regulating host lipid, glucose, energy metabolism and immune response. Several classes of bacterial enzymes that mediate these transformations have been well-characterized. 1) Bile Salt Hydrolases (BSHs): These enzymes catalyze the hydrolytic deconjugation of glycine- and taurine-conjugated bile acids, releasing unconjugated bile acids and free amino acids. Moreover, some BSHs have transferase activity, catalyzing a “transpeptidation” reaction, replacing glycine or taurine with either tyrosine, phenylalanine, leucine or tryptophan referred to as microbially conjugated bile acids.^35–38 While these microbial-conjugated bile acids significantly expand the structural diversity of the bile acid pool, recent quantitative analyses indicate that their absolute abundance in the general circulation is markedly lower (often present at trace levels) compared to the millimolar concentrations of traditional primary bile acids ³⁵ ; therefore, they differentially regulate bile acid receptors [e.g., FXR, pregnane X receptor (PXR) and MRGPRE] locally in the gut rather than systemically.³⁸ Mechanistically, it is also worth noting that cysteamine (the precursor for bile acid–methylcysteamine (BA-MCY)) and taurine are both metabolically derived from cysteine and share significant structural similarities, differing primarily in their sulfur oxidation state. Deconjugation is a critical initiating step that facilitates further microbial transformations and modulates bile acid solubility, intestinal reabsorption, and antimicrobial activity.^39,40 2) 7α-Dehydroxylases: Members of the Clostridium and Eubacterium genera express 7α-dehydroxylating enzymes that convert the primary bile acids—CA and CDCA—into the secondary bile acids deoxycholic acid (DCA) and lithocholic acid (LCA) in human, respectively.⁴¹ The 7α-dehydroxylation pathway is encoded by the bile acid–inducible (bai) operon, which comprises multiple genes (typically baiB, baiCD, baiE, baiF, baiH, baiI, etc.) (Fig. 4). In rodents, α-MCA and β-MCA are metabolized by gut bacteria into murideoxycholic acid (MDCA) and ω-MCA (Fig. 5).^42,43 3) Hydroxysteroid Dehydrogenases, Epimerases, and Oxidoreductases: These enzymes catalyze site-specific oxidation, reduction, and epimerization reactions at the C3, C7, and C12 positions of the bile acid steroid nucleus, generating a structurally diverse repertoire of oxo- and iso-bile acids.^41,44 The efficiency and diversity of these transformations depend heavily on the composition of the gut microbiome.

Fig.4. Bile acid transformation by gut bacteria in human.

After secretion into the intestinal lumen, bile acids are extensively modified by gut microbes, altering their solubility and signaling properties. BSHs deconjugate glycine- and taurine-conjugated bile acids, enabling further transformations. 7α-dehydroxylases from Clostridium and Eubacterium convert primary bile acids (CA, CDCA) into secondary bile acids (DCA, LCA) via the bai operon. BSH, bile-salt hydrolase; DCA, deoxycholic acid; CDCA, chenodeoxycholic acid; LCA, lithocholic acid.

Fig.5. Bile acid transformation by gut bacteria in rodents.

In rodents, α-MCA and β-MCA are metabolized by gut bacteria by 7α-dehydroxylation into MDCA and ω-MCA. MCA, muricholic acid; MDCA, murideoxycholic acid.

A.4. Enterohepatic circulation of bile acids

The enterohepatic circulation is a highly efficient recycling system that maintains the size and composition of the bile acid pool, linking the liver, intestine, and portal circulation in a tightly regulated feedback loop. This process ensures that bile acids are reused multiple times per day, maximizing their functional roles in lipid digestion and metabolic signaling while minimizing de novo synthesis from cholesterol. As shown in Fig. 6A–B., after synthesis and conjugation with glycine or taurine, bile acids are actively transported across the canalicular membrane of hepatocytes into bile via the BSEP/ABCB11, an ATP-dependent transporter. They are secreted together with phospholipids (via MDR3; ABCB4) and cholesterol (via ATP binding cassette subfamily G member 5/8 (ABCG5/ABCG8)) to form mixed micelles and phospholipid-cholesterol vehicles that prevent detergent-induced membrane damage. The bile is then concentrated and stored in the gallbladder, where bile acids remain conjugated and solubilized. Upon meal ingestion, cholecystokinin is released, stimulating gallbladder contraction and bile release into the duodenum. In the small intestine, bile acids emulsify dietary fats and facilitate lipid absorption by forming micelles that incorporate long chain fatty acids, cholesterol, and fat-soluble vitamins (A, D, E, K).

Fig.6. Enterohepatic circulation.

(A) Overview. Bile acids are synthesized in the liver, secreted into the intestine, with 95% reabsorbed at the terminal ileum and returned to the liver via the portal vein, while 5% is newly synthesized to replace losses in feces. (B) After synthesis and conjugation with glycine or taurine, bile acids are secreted from hepatocytes into bile canaliculi via the BSEP, together with phospholipids (MDR3) and cholesterol (ABCG5/8). Conjugated bile acids are reabsorbed into hepatocytes mainly by the NTCP, while unconjugated and sulfated forms use OATPs. Basolateral efflux transporters, including MRP3 and MRP4, mediate export of bile acids and other organic solutes into systemic circulation. Under pathological conditions, the OSTα/OSTβ heterodimer—normally restricted to enterocytes—can be upregulated in hepatocytes, contributing to alternative bile acid efflux. (C) Bile acids in bile ducts are reabsorbed by cholangiocytes via the apical ASBT or passive diffusion, then exported basolaterally through t-ASBT, OSTα/OSTβ, and MDR3 into the peribiliary plexus. These are recaptured by hepatocytes, an intrahepatic recycling route maintaining bile acid homeostasis. (D). About 95% of bile acids are reabsorbed in the terminal ileum via the apical ASBT. Inside enterocytes, they bind to IBABP for cytosolic transport and are then exported across the basolateral membrane into the portal vein by the OSTα/OSTβ heterodimer. BSEP, bile salt export pump; MDR3, multidrug resistance protein3; NTCP, sodium taurocholate cotransporting polypeptide; OATP, organic anion-transporting polypeptide; MRP, multidrug resistance–associated protein; OST, organic solute transporter; ASBT, apical sodium-dependent bile acid transporter; t-ASBT, truncated ASBT; IBABP, ileal bile acid-binding protein.

Bile acids, within the bile ducts, can be reabsorbed across the apical membrane of cholangiocytes—particularly in large intrahepatic ducts—and subsequently exported across the basolateral membrane into the peribiliary capillary plexus. On the apical membrane, which faces the bile duct lumen, the primary transporter is the Apical Sodium-dependent Bile Acid Transporter (ASBT, also known as SLC10A2). ASBT is a highly efficient transporter which actively moves bile acids into the cholangiocytes. In addition to active transport, unconjugated bile acids in their protonated form can enter the cholangiocytes via passive diffusion across the apical membrane. Once inside the cholangiocyte, bile acids are exported across the basolateral membrane into the peribiliary capillary plexus by truncated ASBT (t-ASBT), organic solute transporter α/β (OSTα/β) heterodimer, and MDR3. From there, they are taken up again by hepatocytes through the hepatic sinusoids. This recycling pathway, known as cholehepatic shunting, represents an important yet relatively underappreciated route of bile acid recirculation that contributes to the maintenance of the intrahepatic bile acid pool (Fig. 6C).

Approximately 95% of secreted bile acids are reabsorbed in the terminal ileum, primarily via ASBT/SLC10A2 on the apical membrane of ileal enterocytes. Once inside the cell, bile acids are bound by ileal bile acid-binding protein (IBABP), which facilitates their intracellular transport to the basolateral membrane, where they are exported into the portal circulation by the OSTα/β heterodimer (Fig. 6D). Reabsorbed bile acids return to the liver through the portal vein and are efficiently taken up by hepatocytes through specialized sinusoidal transporters. Conjugated bile acids are taken predominantly by the sodium taurocholate cotransporting polypeptide (NTCP; SLC10A1), whereas unconjugated and sulfated bile acids are taken up mainly by organic anion-transporting polypeptides (Fig. 6B). Once inside hepatocytes, both newly synthesized bile acids and deconjugated bile acids returning from the intestine undergo efficient, though not complete, reconjugation before being secreted into bile, thereby completing the enterohepatic cycle, which cycles approximately 3-6 times per day. In hepatocytes, there are also transporters in the basolateral membrane to export bile acids and other organic solutes into the systemic circulation. Both multidrug resistance-associated protein (MRP) 3 [ATP-binding cassette subfamily C members (ABCC) 3] and MRP4 (ABCC4) can transport conjugated bile acids including sulfated forms into the systemic circulation. OSTα/OSTβ is normally expressed in enterocytes, but can be upregulated in hepatocytes under pathological conditions. A small fraction (~5%) of bile acids escapes reabsorption and enters the colon, where gut bacteria modify them through deconjugation, dehydroxylation, and oxidation/isomerization to generate secondary bile acids, such as DCA and LCA. These secondary bile acids can be passively absorbed across the colonic epithelium and transported back to the liver, where they may be conjugated or sulfated to reduce toxicity before secretion.

A.5. Feedback and feedforward regulation of bile acid synthesis

Bile acid synthesis, uptake, and transport are tightly regulated through integrated feedback and feed-forward mechanisms that maintain hepatic cholesterol and bile acid homeostasis. In the classic pathway, the transcription of CYP7A1, the rate-limiting enzyme, is under a dual feedback control centered on Farnesoid X Receptor (FXR/NR1H4) signaling in the ileum and liver. In hepatocytes, bile acids directly activate FXR, inducing the expression of small heterodimer partner (SHP/NR0B2), which represses transcription factors such as liver receptor homolog-1 (LRH-1/NR5A1) and hepatocyte nuclear factor 4α, two positive regulators of CYP7A1 and CYP8B1 expression. In the ileum, bile acids activate FXR in enterocytes, inducing fibroblast growth factor 19 (FGF19/FGF15 in mice) and transport genes (OSTα/OSTβ) that facilitate basolateral efflux to portal blood.^45–50 FGF19 enters the portal circulation and binds to the fibroblast growth factor receptor 4 (FGFR4)/β-Klotho receptor complex in hepatocytes, suppressing CYP7A1 expression.⁵¹ This gut-liver endocrine axis provides a systemic level of feedback control of bile acid homeostasis (Fig. 7). ⁵²

Fig.7. Feedback regulation of Cyp7a1.

In the classic pathway, bile acids activate FXR in hepatocytes, inducing SHP, which suppresses LRH-1 and HNF4α, thereby downregulating CYP7A1 and CYP8B1 transcription. In enterocytes, FXR activation stimulates FGF19(FGF15 in mice) and transporters such as IBABP and OSTα/OSTβ, promoting bile acid efflux to portal vein. Circulating FGF19 then binds FGFR4/β-Klotho in hepatocytes, inhibiting CYP7A1 expression. This gut–liver FXR–FGF19 axis ensures systemic feedback regulation of bile acid homeostasis. FXR, farnesoid X receptor; SHP, small heterodimer partner; LRH-1, liver receptor homolog-1; HNF4α, hepatocyte nuclear factor 4-alpha; CYP, cytochrome P450; FGF, fibroblast growth factor; IBABP, ileal bile acid-binding protein; OST, organic solute transporter.

In the alternative (acidic) pathway of bile acid synthesis, CYP27A1 activity is not strongly inhibited by bile acid accumulation. Instead, this pathway operates through a feed-forward mechanism that responds to cholesterol availability. Cholesterol is first transported into mitochondria by StarD1. Within mitochondria, CYP27A1 hydroxylates cholesterol to form oxysterols.²⁶ The resulting oxysterols can enter the classical bile acid synthesis pathway or act as endogenous ligands for the liver X receptor (LXR). ⁵³ In rodents, LXRα directly regulates the classical bile acid synthesis pathway by activating Cyp7a1. A specific LXR response element (LXRE) exists in the mouse and rat Cyp7a1 promoters, but not in human. Simultaneously, LXR represses Cyp7b1, the enzyme that normally degrades oxysterols (LXR ligands). Activated LXR also promotes the transcription of the gene encoding StarD1. Therefore, when cholesterol accumulates, increased oxyterol formation activates LXR, which in turn increases bile acid synthesis. This feed-forward loop promotes cholesterol clearance and bile acid synthesis when cholesterol levels are high (Fig. 8).

Fig.8. Feed-forward regulation of bile acid synthesis.

In the alternative bile acid synthesis pathway, cholesterol is transported into mitochondria by StarD1, where CYP27A1 converts it to oxysterols. These oxysterols either enter the classical pathway or activate LXR. In rodents, LXR α upregulates Cyp7a1 and StarD1 while repressing Cyp7b1, enhancing bile acid synthesis. Because CYP27A1 is not strongly feedback-inhibited by bile acids, this pathway functions through a cholesterol-sensitive feed-forward loop, where rising cholesterol increases oxysterol and LXR activation, thereby promoting cholesterol clearance via enhanced bile acid synthesis. StarD1, StAR-related lipid transfer domain protein 1; CYP, cytochrome P450; LXR, liver X receptor.

B. Gut microbiota composition in healthy

The human gut harbors a highly diverse and dynamic microbial ecosystem composed primarily of bacteria, with additional contributions from archaea, viruses, and fungi. In healthy individuals, this microbiota is dominated by bacteria from the Firmicutes and Bacteroidetes phyla, along with smaller proportions of Actinobacteria, Proteobacteria, and Verrucomicrobia. This microbial community performs a range of critical functions, including fermentation of dietary fibers, production of short chain fatty acids (SCFAs), synthesis of vitamins, modulation of immune responses, and transformation of bile acids.

B.1. Regional distribution and dominant taxa

The lumen of the gastrointestinal tract (GI tract) is generally an anaerobic environment, especially in the ileum and colon. Consequently, more than 98% of bacteria colonizing the GI tract are obligate anaerobes, with smaller populations of facultative anaerobes. Distinct regions of the GI tract, including both the lumen and the mucosal surfaces, provide unique ecological niches for bacterial colonization. Moreover, there is a marked gradient in bacterial density and diversity, increasing from the duodenum (~10³/mL) to the cecum and colon (~10¹¹/g wet weight). The duodenum is primarily colonized by Lactobacillus and Streptococcus species, whereas the jejunum (~10⁴/mL) harbors Lactobacillus, Streptococcus, Staphylococcus, and Veillonella. The ileum (~10⁶–10⁸/mL) supports even denser microbial communities. In the cecum and colon (~10¹¹/g wet weight), more than 200 operational taxonomic units or “species” are typically detected under healthy conditions. At the phylum level, Firmicutes (Gram-positive) and Bacteroidetes (Gram-negative) dominate, with smaller populations of Proteobacteria (facultative Gram-negative) (Fig. 9).⁵⁴

Fig.9. Bacterial in the gastrointestinal tract.

The lumen of the GI tract is largely anaerobic, fostering a steep, distal-to-proximal gradient in microbial density and diversity. Obligate anaerobes comprise >98 % of the community. Counts ascend from ~10³−10⁴ CFU/mL in the duodenum (Lactobacillus, Streptococcus) to ~10⁴−10⁵ CFU/mL in the jejunum (additional Staphylococcus, Veillonella), ~10⁷–10⁸ CFU/mL in the ileum, and ~10¹¹ CFU/mL in the cecum and colon, where Firmicutes and Bacteroidetes predominate alongside minor Proteobacteria, yielding >200 OTUs under healthy conditions. GI, gastrointestinal; CFU, colony forming unit; OUT, operational taxonomic unit.

B.2. Functional diversity and metabolite production

Bacteria residing in the GI tract utilize diverse endogenous substrates for growth and energy production. Epithelial cells in the small intestine are sloughed approximately every three days, providing more than 100 g of cellular material daily to the large intestine, which serves as a key nutrient source for resident microbes. Host-specific factors can also influence microbial populations; for instance, individuals with blood group B may harbor higher levels of bacteria capable of utilizing corresponding glycoproteins compared to those with blood groups A or O.⁵⁵ In addition, gut bacteria metabolize bile components, including bilirubin conjugates, phosphatidylcholine, and cholesterol. Under anaerobic conditions, gut microbes rely primarily on fermentative metabolism, generating SCFAs—acetate, propionate, and butyrate—as principal end product.^56–58 These SCFAs serve as major energy sources for colonocytes and exert systemic effects by modulating immune responses, lipid metabolism, and glucose homeostasis via G-protein–coupled receptors (GPCRs) and epigenetic pathways.

In addition to SCFAs, the microbiota produces a broad range of metabolites that influence host physiology. Tryptophan-derived compounds such as indole, indole-3-propionic acid, and indole-3-pyruvate maintain epithelial integrity and modulate mucosal immunity. Bile acid derivatives— such as ursodeoxycholic acid (UDCA), glycoursodeoxycholic acid (GUDCA), α-, β-, and γ-muricholic acids, and 3-succinylated cholic acid—act as intestinal FXR antagonists, contributing to metabolic regulation. Conversely, intestinal FXR agonists such as tyrosocholic acid, phenylalanocholic acid, and leucocholic acid are associated with metabolic dysfunction.³⁷

Gut microbes also influence the secretion of glucagon-like peptide-1 (GLP-1) from intestinal L-cells through the production of bile acid–derived FXR agonists and antagonists, linking the gut microbiome to glucose metabolism and energy balance.^43,59 Additionally, microbial metabolites such as trimethylamine, produced from dietary phosphatidylcholine, L-carnitine, and choline, are converted in the liver to trimethylamine-N-oxide, which contributes to atherosclerosis and cardiovascular disease.^60,61

Microbial interactions further shape community structure. For example, Bacteroides uniformis secretes 3-succinylated cholic acid, which promotes the growth of Akkermansia muciniphila, a bacterium associated with metabolic health in both mice and humans.^62,63 In contrast, Clostridium scindens produces secondary bile acids and antimicrobial compounds that suppress Clostridioides difficile growth, illustrating how commensal bacteria can protect against pathogenic species.⁶⁴

B.3. Factors shaping gut microbiome composition

Multiple host and environmental factors influence gut microbiota composition, including diet, age, bile acids, gastrointestinal and hepatic diseases, antibiotic exposure, genetics, sex, and intestinal transit time. Diet is among the most potent modulators of microbial ecology. Short-term dietary interventions can rapidly—within approximately two days—alter microbial composition, metabolic end products, and gene expression.⁶⁵ Western-style diets high in fat and animal protein are associated with increased incidence of colon cancer, inflammatory bowel disease, MASLD, type 2 diabetes, and atherosclerosis. In contrast, plant-based diets enriched in fiber promote microbial fermentation and SCFAs production, supporting gut and metabolic health. O’Keefe et al. demonstrated this principle in a dietary exchange study between rural South Africans consuming a plant-based diet and African Americans consuming a Western diet. After two weeks, reciprocal changes were observed in colonic mucosal biomarkers of cancer risk, gut microbiota composition, secondary bile acid levels, and fermentation profiles—highlighting the rapid adaptability of the gut microbiome to dietary inputs.⁶⁶

Defining a “healthy” gut microbiome remains challenging, as microbial composition is dynamic and influenced by numerous factors. Nevertheless, high microbial diversity is a consistent feature of gut health. Enrichment of bacterial families such as Akkermansiaceae, Christensenellaceae, and Ruminococcaceae, along with SCFA-producing genera including Prevotella and Faecalibacterium, is linked to metabolic resilience and reduced systemic inflammation.^67–70 Conversely, expansion of lipopolysaccharide (LPS)-producing bacteria, particularly Enterobacteriaceae, contributes to gut dysbiosis, endotoxemia, and systemic inflammation.⁷¹ Humans are especially sensitive to LPS-induced immune activation via Toll-like receptor 4 (TLR4) signaling.⁷² Moreover, an increased Bacteroidetes-to-Firmicutes ratio is generally associated with metabolic dysregulation and microbial imbalance (Fig. 10).⁷³

Fig.10. Normal gut microbiome composition at phylum level.

Firmicutes (51%) and Bacteroidetes (41%) are dominant taxa, together accounting for ~92 % of sequences. Minor constituents include Proteobacteria (5%), Actinobacteria (1%), and other phyla (2%), collectively forming a community configuration typically observed in stool from healthy human subjects.

In summary, the human gut microbiome represents one of the most complex microbial ecosystems known. Through metabolic specialization and interspecies interactions, gut microorganisms degrade host- and diet-derived substrates to produce nutrients and signaling molecules essential for host metabolism, immunity, and intestinal homeostasis. The human body, therefore, functions as a symbiotic ecosystem of interacting prokaryotic and eukaryotic cells, jointly maintaining health and responding to environmental perturbations.

III. Bile Acids as Signaling Molecules

Bile acids were historically merely regarded as end-products of cholesterol catabolism and as amphipathic detergents facilitating dietary lipid solubilization and absorption in the intestine. This traditional perspective was revolutionized by the discovery of the first bile acid–activated nuclear receptor, the FXR (NR1H4), in 1999.^45,46,74,75 In addition to FXR, several other bile acid– responsive receptors have been identified, including the Takeda G protein–coupled receptor 5 [TGR5; G protein–coupled bile acid receptor 1 (GPBAR1)], the first membrane receptor activated by bile acids,^76,77 the PXR(NR1I2), which functions as xenobiotic sensors and modulate bile acid detoxification pathways,⁷⁸ and the vitamin D receptor (VDR; NR1I1), which recognizes LCA as a ligand and contributes to enterohepatic protection against bile acid toxicity.⁷⁹ More recently, the sphingosine-1-phosphate receptor 2 (S1PR2) has been shown to be activated by conjugated primary bile acids, linking bile acid signaling to sphingolipid metabolism and inflammatory regulation in hepatocytes and cholangiocytes.^80,81 These discoveries have redefined bile acids as endocrine and paracrine signaling molecules that exert broad regulatory effects extending far beyond their traditional roles in lipid digestion. Through coordinated activation of nuclear and membrane receptors, bile acids integrate hepatic and intestinal metabolic circuits, modulate glucose and lipid homeostasis, and fine-tune immune and inflammatory responses (Fig. 11).

Fig.11. Bile acid receptor signaling and metabolic regulation.

Bile acids act as signaling molecules that regulate metabolism, inflammation, and energy homeostasis through nuclear receptors (FXR, PXR, VDR) and G protein-coupled receptors (TGR5, S1PR2, MRGPRX4). FXR controls bile acid synthesis and transport via SHP and promotes metabolic and anti-inflammatory effects through FGF15/19 signaling. PXR and VDR respond to lithocholic acid derivatives to induce detoxifying enzymes and suppress inflammation. TGR5 activation stimulates cAMP-PKA signaling in immune and enteroendocrine cells, enhancing GLP-1 release and energy expenditure. S1PR2 mediates conjugated bile acid–induced ERK and Ca²⁺ signaling, while MRGPRX4 triggers bile acid–evoked itch in sensory neurons. The gut microbiota modifies bile acid composition through deconjugation and transformation, shaping receptor activity and systemic endocrine responses. FXR, farnesoid X receptor; PXP, pregnane X receptor; VDR, vitamin D receptor; TGR5, Takeda G protein-coupled receptor 5; S1PR2, sphingosine-1-phosphate receptor 2; MRGPRX4, mas-related G-protein–coupled receptor X4; FGF, fibroblast growth factor; cAMP-PKA, cyclic AMP- protein kinase A; GLP-1, glucagon-like peptide-1; ERK, extracellular signal-regulated kinase.

A. Bile acid receptors

The identification of bile acid–responsive receptors has significantly expanded our understanding of bile acids as signaling molecules that coordinate metabolic, inflammatory, and homeostatic processes across multiple organ systems.^82,83 These receptors can be broadly categorized into nuclear receptors, which act as transcriptional regulators of gene expression, and membrane-bound GPCRs, which mediate rapid, non-genomic responses.^46,76 Together, these pathways integrate bile acid synthesis, transport, and detoxification with systemic energy metabolism, immune regulation, and cell survival.^82–84

A.1. Farnesoid X Receptor (FXR; NR1H4)

FXR is the prototypical bile acid–activated nuclear receptor and serves as the master regulator of bile acid homeostasis. Highly expressed in hepatocytes, enterocytes, and renal tubular epithelial cells, FXR is primarily activated by CDCA, with lesser potency for CA, DCA, and LCA. Upon ligand binding, FXR forms a heterodimer with retinoid X receptor and binds to FXR response elements in the promoters of target genes to modulate transcription. It plays an important role in the feedback regulation of bile acid synthesis in the classical pathway. As illustrated in Fig. 6, hepatic FXR activation represses bile acid synthesis through induction of SHP, which inhibits the trans-activating activities of liver receptor homolog-1 and hepatocyte nuclear factor 4α on CYP7A1 and CYP8B1, the key enzymes of the classical bile acid biosynthetic pathway. FXR also promotes bile acid efflux via upregulation of BSEP/ABCB11 and enhances conjugation and detoxification through induction of SULT2A1 and UGT2B4. In the intestine, FXR activation induces FGF19 (Fgf15) in mice), which enters portal circulation and binds to the FGFR4/β-Klotho complex to further suppress CYP7A1 expression. This establishes an enterohepatic feedback loop that maintains bile acid pool size and composition. Beyond bile acid metabolism, FXR regulates triglyceride turnover, glucose utilization, and anti-inflammatory gene programs, making it a key therapeutic target in metabolic and cholestatic liver diseases. In contrast, several bile acid species act as natural FXR antagonists, including tauro-β-muricholic acid (TβMCA) and GUDCA.⁸⁵ A recent study using untargeted metabolomics, identified a group of BA-MCY conjugates highly abundant in the intestine. Their formation depends on the enzyme vanin-1, a pantetheinase expressed in intestinal tissues. BA-MCYs act as potent natural FXR antagonists, which promote bile acid synthesis. Multiple BA-MCY species were detected in human serum, suggesting that this regulatory mechanism is conserved in humans.⁸⁶

A.2. Other nuclear receptors: PXR and VDR

In addition to FXR, several nuclear receptors are also activated by bile acids, including PXR and VDR ^79,87. PXR is highly expressed in the liver, particularly in the periportal hepatocytes, and regulates genes involved in phase I and II detoxification, such as CYP3A4, SULT2A1, UGT2B4, and transporters (MRP2, OSTα/β). PXR plays central roles in xenobiotic detoxification, bile acid homeostasis, and drug–drug interactions. PXR can be activated by a broad range of ligands (rifampicin, hyperforin, dexamethasone, bile acids). PXR is also highly expressed in the intestine.

Activation of PXR reduces the production of proinflammatory mediators, such as TNFα and IL1β, and helps maintain intestinal barrier integrity. LCA and its derivatives (3-keto-LCA) are the most potent bile acid activators of PXR.⁷⁸VDR is widely expressed across many tissues, particularly in the intestine, kidney, and immune cells. In addition to 1,25-dihydroxyvitamin D₃ (calcitriol), LCA and its derivatives can activate VDR, promoting detoxification pathways and protecting intestinal epithelial cells.⁷⁹Together, these nuclear receptors function as a safety network to maintain bile acid homeostasis.

A.3. Takeda G protein–coupled receptor 5 (TGR5; GPBAR1)

TGR5 is the first membrane receptor identified for bile acids and serves as a key mediator of their non-genomic signaling effects.⁷⁶ It is broadly expressed in cholangiocytes, Kupffer cells, liver sinusoidal endothelial cells, brown adipose tissue, skeletal muscle, and enteroendocrine L cells.^88–95 TGR5 is preferentially activated by secondary bile acids, particularly LCA and DCA, as well as their taurine-conjugates. Upon ligand binding, TGR5 couples to Gs proteins, leading to adenylate cyclase activation and elevated intracellular cyclic adenosine monophosphate (cAMP) levels, which subsequently trigger protein kinase A signaling and downstream metabolic and immunomodulatory pathways. Through these signaling cascades, TGR5 exerts diverse physiological functions. In macrophages, TGR5 activation inhibits nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling and cytokine production, exerting potent anti-inflammatory effects. In enteroendocrine L cells, TGR5-mediated cAMP signaling promotes secretion of GLP-1, thereby enhancing insulin sensitivity and glucose homeostasis. In brown adipose and skeletal muscle, TGR5 activation increases energy expenditure by inducing type 2 iodothyronine deiodinase, which converts thyroxine to the active thyroid hormone triiodothyronine. These metabolic and immunomodulatory effects position TGR5 as a potential therapeutic target for obesity, insulin resistance, and inflammatory liver disease. The potency of endogenous bile acids activate TGR5 is (T)LCA>DCA>CDCA>CA.

A.4. Sphingosine-1-phosphate receptor 2 (S1PR2)

Sphingosine-1-phosphate receptor 2 (S1PR2), originally known as endothelial differentiation gene 5, is one of five high-affinity GPCRs (S1PR1–S1PR5) that mediate the pleiotropic actions of sphingosine-1-phosphate (S1P), a potent bioactive sphingolipid metabolite. The EDG family of S1P receptors was identified in the mid-1990s through molecular cloning of endothelial differentiation genes involved in vascular development.^96,97 S1PR2 was officially renamed following the functional characterization of S1P as its endogenous ligand.^98–100 S1PR2 activates diverse intracellular cascades via coupling to multiple heterotrimeric G proteins—Gαi/o, Gαq, and Gα12/13.¹⁰¹ The major downstream pathways include: 1) Extracellular signal-regulated kinase (ERK)1/2 (MAPK) and phosphatidylinositol 3-kinase (PI3K) /AKT pathways, which control cell proliferation, survival, and metabolism; 2) RhoA/Rho kinase (ROCK) signaling, which regulates cytoskeletal dynamics, cell shape, and motility; 3) Phospholipase C/Inositol 1,4,5-triphosphate/Ca²⁺ signaling, which modulates calcium homeostasis and inflammatory responses; and 4) Mammalian target of rapamycin complex 1 (mTORC1) and NF-κB activation, which link S1PR2 to lipid synthesis, insulin signaling, and cytokine expression.^102–105

Among the five S1P receptor subtypes, S1PR2 is the only one directly activated by bile acids. The seminal work of Studer and colleagues first identified S1PR2 as a bile acid–responsive GPCR, revealing a mechanistic link between bile acid signaling and sphingolipid metabolism. Conjugated primary bile acids—particularly TCA and GCA—serve as endogenous agonists of S1PR2 in hepatocytes and cholangiocytes, activating ERK1/2 and AKT signaling to regulate lipid and glucose metabolism, cell proliferation, survival, and inflammatory responses.^80,106–109

Bile acid–activated S1PR2 is also linked to the sphingosine kinase 2 (SphK2)–nuclear S1P axis. Nuclear S1P, generated by SphK2, serves as a potent endogenous inhibitor of histone deacetylases, thereby promoting histone acetylation and transcription of genes involved in lipid oxidation, mitochondrial function, and metabolic homeostasis.^110–112 This pathway represents a unique intersection between lipid signaling and epigenetic regulation within hepatocytes.

In summary, S1PR2 serves as a key regulatory hub integrating bile acid signaling, sphingolipid metabolism, and hepatic inflammation and metabolic responses. Its activation modulates not only cytoplasmic signaling, but also nuclear gene regulation. The S1PR2-SphK2 axis represents a promising therapeutic target for different types of liver diseases and cancers that express S1PR2.

A.5. Mas-related G-protein–coupled receptor X4 [MRGPRX4 (hX4)]

MRGPRX4, also called hX4, is a human-specific class A GPCR belonging to the Mas-related GPCR family, which is expressed primarily in a subset of sensory neurons of the dorsal root and trigeminal ganglia, and to a lesser extent in cholangiocytes and immune cells. It couples predominantly to the Gq/11 signaling pathway, leading to phospholipase C activation, Ca²⁺ mobilization, and neuronal excitation.¹¹³ Unlike other bile acid receptors involved in metabolic regulation, MRGPRX4 mediates neuro-sensory responses, particularly pruritus (itch). It has been reported that both unconjugated and conjugated bile acids can activate MRGPRX4. In humanized MRGPRX4 mice, bile acid stimulation triggered strong neuronal activation and scratching behavior under both acute and chronic cholestasis conditions. Although UDCA is widely used as an antipruritic agent in cholestatic liver diseases such as PBC and intrahepatic cholestasis of pregnancy, this study indicates that, along with DCA, UDCA itself can activate MRGPRX4, potentially promoting itch under certain conditions. This apparent contradiction may reflect UDCA’s complex pharmacodynamics—it improves hepatocellular function, reduces inflammation and fibrosis, and enhances bile flow, which collectively outweigh its receptor agonism and result in net antipruritic effects in patients.^113,114

In summary, the bile acid receptor network establishes a highly coordinated system that regulates enterohepatic circulation, energy homeostasis, and immune balance. The interplay between nuclear and membrane receptors ensures rapid and long-term regulation of bile acid synthesis, transport, and detoxification, while extending bile acid influence to systemic metabolic and inflammatory pathways. Dysregulation of these receptors contributes to diverse pathological states, including MASLD, cholestasis, fibrosis, and hepatocarcinogenesis. Consequently, pharmacological targeting of bile acid receptors represents a promising strategy for treating metabolic and hepatobiliary disorders.

B. Crosstalk between bile acids, gut hormones (GLP-1, FGF15/19), and metabolism

Beyond their roles as detergents and signaling molecules, bile acids serve as endocrine regulators that coordinate energy and glucose metabolism through interactions with gut-derived hormones such as FGF15/19 and GLP-1. These hormones form part of an intricate gut–liver–metabolic axis, linking bile acid flux to nutrient absorption, insulin sensitivity, and energy expenditure.

B.1. Intestinal FXR–FGF15/19 axis

Activation of FXR in ileal enterocytes induces transcription of FGF15 (in mice; FGF19 in humans), which is secreted into the portal circulation and acts on the hepatic FGFR4/β-Klotho complex. This signaling cascade inhibits CYP7A1, the rate-limiting enzyme in bile acid synthesis, thereby maintaining bile acid pool homeostasis.^115,116 Beyond its classical feedback role, FGF15/19 exerts metabolic effects, including suppression of gluconeogenic genes (PEPCK, G6Pase), stimulation of glycogen and protein synthesis, and reduction of hepatic triglyceride accumulation.¹¹⁷ Intestinal FXR–FGF15/19 signaling functions as an endocrine relay between the gut and liver, integrating bile acid and glucose metabolism to preserve postprandial energy balance.¹¹⁸

B.2. TGR5–GLP-1 axis

TGR5 is expressed in intestinal enteroendocrine L cells and mediates bile acid–induced secretion of GLP-1, an incretin hormone that enhances glucose-stimulated insulin secretion, delays gastric emptying, and reduces appetite.¹¹⁹ TGR5 activation by secondary bile acids such as LCA and DCA increases intracellular cAMP, stimulating protein kinase A and subsequent GLP-1 release. Through this mechanism, bile acids act as metabolic signals that directly influence insulin sensitivity and glucose tolerance. Moreover, TGR5 activation in brown adipose tissue and skeletal muscle induces type 2 iodothyronine deiodinase, which converts thyroxine into triiodothyronine, thereby promoting energy expenditure and thermogenesis.¹²⁰

B.3. Integration of FXR and TGR5 signaling

The FXR and TGR5 signaling pathways function synergistically to maintain systemic metabolic homeostasis. FXR–FGF15/19 signaling serves as a central regulator of bile acid synthesis, hepatic gluconeogenesis, and lipid metabolism, whereas TGR5–GLP-1 signaling enhances pancreatic insulin secretion, energy expenditure, and glucose tolerance. Crosstalk between these receptors ensures coordinated control of postprandial nutrient metabolism and energy balance. In metabolic diseases, impaired bile acid receptor signaling leads to diminished FGF19 and GLP-1 responses, promoting insulin resistance, dyslipidemia, and hepatic steatosis. Understanding the FXR–FGF19–TGR5–GLP-1 network provides critical insight into therapeutic strategies targeting metabolic and hepatobiliary disorders.

C. Microbial bile acid derivatives as receptor modulators

The gut microbiota plays an indispensable role in shaping the bile acid signaling landscape. Through a diverse repertoire of microbial enzymes—including BSHs that mediate deconjugation ,⁴¹ 7α/7β-dehydroxylases that form secondary bile acids, and enzymes responsible for oxidation and epimerization of hydroxyl groups.^121,122 These microbial transformations dramatically alter bile acid hydrophobicity, receptor affinity, and signaling potency, thereby influencing host metabolism, immunity, and inflammation.^43,123

Microbially derived bile acids serve as endogenous ligands for multiple bile acid receptors, including FXR, TGR5, PXR, VDR, and MRGPRX4.^{46,77,79,114,124} Among them, secondary bile acids such as LCA and its derivatives 3-oxo-LCA and iso-LCA function as potent VDR agonists, promoting mucosal immune tolerance, Treg differentiation, and antimicrobial peptide expression.^125,126 Conversely, DCA and LCA sulfates activate PXR, inducing CYP3A4 and other detoxification pathways that enhance bile acid clearance and reduce hepatic toxicity.^127–130 Most recently, 3-sulfated deoxycholic acid (DCA-3S) was identified as an endogenous ligand for the human-specific receptor MRGPRX4, linking bile acid signaling to cholestatic pruritus.¹³¹ Collectively, these microbially derived metabolites modulate receptor cross-talk between hepatic, intestinal, and immune compartments, maintaining metabolic and immune balance—but when disrupted, they contribute to metabolic, inflammatory, and cholestatic diseases.^124,132

D. Microbiota–bile pool composition and receptor regulation

Microbial remodeling of the bile acid pool establishes a dynamic reciprocal feedback loop with host receptors. The levels and composition of the circulating bile acid pool are determined by intestinal transit time, levels and composition of specific bacterial species that metabolize bile acids, which, in turn, determine receptor activation patterns. Conversely, receptor signaling through FXR modifies bile acid synthesis, secretion, and gut ecology. For example, intestinal FXR activation induces FGF15/19, which represses CYP7A1 transcription in hepatocytes, thereby limiting bile acid synthesis.^115,133 TGR5 activation stimulates GLP-1 secretion, gut motility, and antimicrobial peptide production, shaping the intestinal microbial environment.^{119,134–136} Perturbations of microbiota-host homeostasis, such as antibiotic exposure, a high-fat diet, or other environmental toxins, can alter bile acid composition and impair receptor signaling. This imbalance compromises gut barrier integrity, promotes systemic endotoxemia, and drives hepatic inflammation and metabolic dysfunction.^43,137,138

IV. Impact of Gut Microbiome and Bile Acids in Liver Diseases

The gut–liver axis represents a dynamic and bidirectional communication network that plays a crucial role in maintaining metabolic homeostasis and immune balance. Increasing evidence indicates that alterations in the gut microbiome and bile acid metabolism are key factors contributing to the development and progression of various liver diseases, including MASLD, ALD, cholestatic liver disease and liver cancer.^17,139 The gut microbiota regulates bile acid synthesis and composition through microbial enzymes, while bile acids, in turn, modulate microbial diversity and function via signaling pathways such as the FXR and TGR5.^86,140 Disruption of gut-liver axis can lead to intestinal barrier dysfunction, bacterial translocation, and hepatic inflammation.³¹ Understanding the molecular and microbial mechanisms underlying the gut microbiome–bile acid–liver axis not only provides insights into disease pathogenesis but also opens new therapeutic avenues, including the use of probiotics, prebiotics, and bile acid modulators for liver disease management.¹⁵

A. Gut microbiome dysbiosis in MASLD and MASH

MASLD is a chronic liver disease defined by the presence of hepatic steatosis in individuals with one or more metabolic risk factors, such as obesity, insulin resistance, type 2 diabetes mellitus, or dyslipidemia.¹⁴¹ This terminology replaces the former non-alcoholic fatty liver disease to better emphasize the metabolic origins of the disease rather than the absence of alcohol consumption. A subset of patients progresses to a more severe inflammatory form, known as MASH, involves hepatocyte injury, immune cell infiltration, and fibrogenesis, which may ultimately lead to cirrhosis, hepatic decompensation, or HCC.^142,143 MASLD and MASH are rapidly increasing worldwide and have become leading causes of chronic liver disease, liver cancer, and liver transplantation. Accumulating evidence indicates that gut microbiome dysbiosis is a major mechanistic contributor to disease initiation and progression, influencing hepatic metabolism, immune responses, and fibrogenesis through alterations in microbial metabolites, intestinal barrier integrity, and gut–liver signaling pathways.^7,144

A.1. Composition shifts of gut microbiome in MASLD and MASH

The occurrence and progression of MASLD are closely linked to dysbiosis in gut microbial composition and abundances. The gut microbiota exhibits stage-specific alterations throughout the MASLD spectrum, suggesting its potential value for both diagnostic biomarker discovery and for identifying mechanisms underlying disease progression. At the phylum level, enrichment of Proteobacteria—a phylum containing numerous opportunistic pathogens—represents one of the most consistent microbial signatures in MASLD, particularly in advanced fibrosis and cirrhosis.^145,146 Firmicutes, a dominant phylum in the healthy gut, shows variable trends in patients with MASLD across studies, likely due to changes in multiple constituent genera. Nonetheless, most evidence suggests a decrease in the relative abundance of Firmicutes, which tends to decline further as MASLD progresses to the fibrosis stage.^145,147,148 Bacteroidetes and Fusobacteria have also been reported to be enriched in the gut microbiota of MASLD patients; however, the findings are not entirely consistent across studies, likely due to differences in ethnicity, dietary patterns, and lifestyle factors.¹⁴⁵ Some alterations in gut bacterial profiles observed in MASLD patients mirror those seen in metabolic disorders, providing additional evidence for the shared pathological features between these conditions.¹⁴⁹ Proteobacteria, which includes multiple opportunistic pathogens such as Escherichia, shows an increased relative abundance in MASLD, particularly in MASLD-associated liver fibrosis and cirrhosis.^146,150,151 At the genus level, some bacterial changes have been repeatedly confirmed, and their functions have been further elucidated. Prevotella is enriched in MASLD patients and shows a positive correlation with fibrosis,^148,152 which has been reported in both adult and pediatric patients, ¹⁵³ Nevertheless, other studies have found a reduced abundance of Prevotella in the gut microbiota of MASLD with simple steatosis and MASH ^145,154. The Firmicutes contain many major SCFA-producing bacteria in the gut. Faecalibacterium, Eubacterium, and Lachnospira have been observed to decrease in the feces of MASLD and MASH patients; whereas the abundance of bacteria in Bacteroides, Acidaminococcus, Lactobacillus, and Clostridium is increased. However, the changes in some bacteria cannot be explained solely by their SCFA-producing capacity.¹⁵⁵ A simple classification of gut microbes into “beneficial” and “harmful” groups is inadequate for describing their roles in MASLD. The enrichment of Lactobacillus in MASLD highlights a potential shift from its traditional probiotic role, suggesting that strain-specific metabolic effects—such as lactate production and bile acid modification—may support disease progression under altered host conditions.

Gut microbiota alterations pattern provides insights into both the shared and distinct mechanisms associated with stages and metabolic conditions of MASLD. For example, although obese and lean MASLD patients share similar clinical and histological characteristics, distinct microbial signatures reflect differences in latent metabolic profiles. Independent of short-term caloric intake, non-obese MASLD is characterized primarily by a reduced of Faecalibacterium and Ruminococcus, whereas obese MASH patients show an increased abundance of Lactobacillus and overweight MASH patients exhibit a decrease in Bifidobacterium. Moreover, a reduction in butyrate-producing Eubacterium has been found to correlate with the stage of MASLD-associated liver fibrosis, and this association appears to be specific to non-obese individuals. These findings suggest that distinct phenotypes of gut microbiota may help explain the unique pathogenic mechanisms underlying MASLD in lean individuals (Table 1).

Table 1.

Gut microbiota alterations across the clinical spectrum of MASLD

Microbial taxa	Subjects	Findings	Reference
Actinobacteria Bacteroidetes	Fecal samples from MAFL, MASH, or obesity and healthy controls	MASLD: Actinobacteria ↑ Bacteroidetes ↓(Bradyrhizobium, Anaerococcus, Peptoniphilus, Propionibacterium acnes, Dorea, and Ruminococcus ↑, Oscillospira and Rikenellaceae ↓ )	PMID: 22241860 526
Actinobacteria Bacteroidetes Firmicutes Proteobacteria Verrucomicrobia	Fecal samples from MASLD-related cirrhosis and HCC, MASLD-related cirrhosis without HCC and healthy controls	Cirrhosis: Enterobacteriaceae ↑, Streptococcus ↑, Akkermansia ↓ HCC: Bacteroides ↑ Ruminococcaceae ↓, Bifidobacterium ↓	PMID: 29665135 146
Actinobacteria Firmicutes Proteobacteria	Fecal samples from obese with different degrees of MASLD	Steatosis and MASH: Enterobacteriaceae (ethanol-producing bacteria), Acidaminococcus and Megasphaera ↑ , Eggerthellaceae and Ruminococcaceae (SCFA-producing bacteria) ↓ Escherichia coli is related to steatosis and necroinflammatory activity, whilst Escherichia-shigella was associated with fibrosis and necroinflammatory.	PMID: 38265508 527
Bacteroidetes Proteobacteria	Children with biopsy-proven MASLD and children with obesity without MASLD	A high abundance of Prevotella copri was associated with more severe liver fibrosis, and genes involved in lipopolysaccharide biosynthesis were significantly enriched in MASH children.	PMID: 31255652 153
Bacteroides Firmicutes	Fecal samples from MASH, obese, and healthy children	Obese and MASH groups: Bacteroides ↑ , Firmicutes ↓ Proteobacteria gradually increased from healthy individuals to obese and then to MASH MASH patients exhibited elevated blood ethanol levels	PMID: 23055155 369
Bacteroidetes Firmicutes	Fecal samples from MASLD, F0/1 fibrosis or F ≥2 fibrosis (with or without MASH)	MASH and F≥2: Bacteroides ↑ Prevotella ↓ F≥2: Ruminococcus ↑ Bacteroides abundance was independently associated with MASH and Ruminococcus with F≥2 fibrosis.	PMID: 26600078 156
Bacteroidetes Firmicutes	Fecal samples from NAFL, MASH, healthy controls	MASLD: Bacteroidetes ↑ Firmicutes ↓ (Ruminococcaceae UCG-010, family Ruminococcaceae, order Clostridiales, and class Clostridia ↓)	PMID: 32204538 154
Bacteroidetes Firmicutes	Population-based cohort including steatosis and liver fibrosis	Immunogenic commensals (e.g. Prevotella copri, Holdemanella, Clostridiaceae) increased in liver fibrosis. Bacteroides caccae, Parabacteroides distasonis, Enterobacter and Marinifilaceae decreased. Alcohol consumption was associated with increasing Prevotella abundance.	PMID: 34633706 528
Bacteroidetes Firmicutes Proteobacteria	MASLD patients and healthy control	MASLD: Lachnospiraceae, Enterobacteriaceae, Erysipelotrichaceae, Streptococcaceae, Escherichia_Shigella, Lachnospiraceae_Incertae_Sedis, Blautia ↑ Prevotella ↓	PMID: 28823367 150
Bacteroidetes Firmicutes Fusobacteria	Fecal samples from healthy control group MAFL group, and MASH group/ database validation methods	MASLD with hepatic steatosis and MASH: Bacteroidetes ↑ , Fusobacteria ↑ , Firmicutes ↓ MASLD with hepatic steatosis: Prevotella ↓ MASH: Megamonas and Fusobacterium ↑	PMID: 38638907 145
Firmicutes	Fecal samples from obese MASLD patients with healthy controls	MASLD: Lactobacillus ↑ , Firmicutes (Lachnospiraceae; genera, Dorea, Robinsoniella, and Roseburia) ↑ , Ruminococcaceae; genus, Oscillibacter ↓ Fecal ester compounds are increased in MASLD	PMID: 23454028 529
Firmicutes	MASLD and non-MASLD (including MASLD with hepatic steatosis , MASH, fibrosis and non-fibrosis)	Ruminococcaceae and Veillonellaceae are the main microbiota associated with fibrosis severity in non-obese subjects.	PMID: 33020474 147
Firmicutes	Fecal samples from MASH and healthy people	EtOH-producing Enterocloster bolteae and Limosilactobacillus fermentum are enriched in MASH, accompanied by an increase in fecal EtOH and glucose.	PMID: 38035329 530
Firmicutes	Non-MASLD controls, MASLD-cirrhosis patients, and their first-degree relatives	MASLD-cirrhosis: Veillonella parvula (most discriminatory), Veillonella atypica, Ruminococcus gnavus, Clostridium bolteae, Acidaminococcus sp. D21 ↑ Eubacterium eligens, Eubacterium rectale, Faecalibacterium prausnitzii ↓108 microbial species showed significant associations with clinical phenotypes, and a gut microbiome signature of 19 discriminatory species accurately discriminated MASLD-cirrhosis.	PMID: 32610095 531
Firmicutes Proteobacteria	Fecal samples from MASLD patients and healthy subjects	MASLD:Escherichia, Anaerobacter, Lacto bacillus, Streptococcus ↑	PMID: 25644696 532
Firmicutes Proteobacteria	Fecal samples from MASLD-cirrhosis, MASLD without advanced fibrosis, non-MASLD controls	MASLD-cirrhosis/MASLD without advanced fibrosis: Streptococcus ↑ MASLD-cirrhosis : Megasphaera ↑ Enterobacteriaceae ↑ Non-MASLD control/MASLD without advanced fibrosis: Bacillus ↑ Lactococcus ↑	PMID: 30926798 151
Firmicutes Proteobacteria	Fecal samples from MASLD patients with different fibrosis stage	Mild/moderate MASLD: Firmicutes ↑ , Eubacterium rectale and Bacteroides vulgatus ↑ Advanced fibrosis: Proteobacteria ↑ R. obeum and E. rectale ↓	PMID: 28467925 148
Firmicutes Verrucomicrobia	Fecal and blood samples from child MASLD patients	MASLD: Lachnospira ↓, Faecalibacterium ↓ , Oscillospira ↓ Akkermansia ↓ Lachnospira is related to flavone and flavonol biosynthesis which is negatively related to insulin levels, insulin resistance and inflammation.	PMID: 38612453 533
Fusobacteria	Fecal samples from MASLD, MASH and healthy control	Fusobacteria/ Fusobacteriaceae were specifically increased. Compared with healthy controls, Prevotellaceae were elevated in MASH.	PMID: 30574320 157
Proteobacteria	Fecal samples from morbidly obese women	Steatosis patients have an enrichment of gut microbial genes related to dietary lipid processing and endotoxin biosynthesis (notably from Proteobacteria), hepatic inflammation and dysregulation of aromatic and branched-chain amino acid metabolism. Fecal microbiota transplantation and chronic phenylacetic acid (PAA) treatment were sufficient to induce experimental hepatic steatosis.	PMID: 29942096 534
Verrucomicrobia (Akkermansia)	Fecal samples from patients with/without recurrent MASLD following liver transplatation	MASLD recurrence: Akkermansia ↓, Fusobacterium ↑	PMID: 33204823 535

Abbreviations: MASLD: metabolic dysfunction-associated steatotic liver disease, MASH: metabolic dysfunction-associated steatohepatitis

A.2. Functional alterations of gut microbiome in MASLD

In MASLD, disruption of gut microbial balance modifies the production of key microbial metabolites that are absorbed and transported to the liver through the portal vein, exacerbating hepatic steatosis, inflammatory signaling, and fibrosis through the gut–liver axis. Combined profiling of gut microbial communities and intestinal metabolites in MASLD demonstrates that the microbiota contributes to disease progression not only through the production of bioactive metabolites but also via intrinsic structural components that activate hepatic immune and metabolic pathways. Gut microbiota–related metabolic alterations in MASLD primarily include reduced SCFA production, disrupted bile acid composition, LPS-induced immune activation, increased bacterially derived ethanol production, and enhanced generation of trimethylamine (TMA). Multi-omics studies further support these microbial–metabolite interactions in disease progression. For example, Lee et al. reported that non-obese MASLD patients with fibrosis exhibited significantly elevated total fecal bile acids, including both conjugated and unconjugated species. Levels of CA, CDCA, UDCA, GCDCA, and GUDCA were all increased, along with a progressive rise in propionate correlating with fibrosis stage. In contrast, obese patients demonstrated a decline in total conjugated bile acids with increasing fibrosis severity.^147,156 SCFA levels also show stage- and phenotype-specific alterations. Rau et al. observed that both MASLD and MASH patients had elevated fecal acetate and propionate, while butyrate levels were particularly increased in MASH. Correspondingly, SCFA-related taxa—including Prevotella copri, Megasphaera, Fusobacterium, Ruminococcus torques, and Eubacterium biforme—were enriched in MASH.¹⁵⁷ However, other studies have reported reduced fecal acetate and butyrate in MASLD, with further depletion in more advanced disease stages, whereas levels of other SCFAs remained unchanged.¹⁵⁸ These inconsistencies highlight the complexity of SCFA regulation, including dietary factors, regional and ethnic differences, and strain-specific microbial metabolism.

Endogenous ethanol production also contributes to hepatic metabolic stress. MASH-associated enrichment of Klebsiella pneumoniae, Lactobacillaceae, and other Proteobacteria and Enterobacteriaceae correlates with increased microbial ethanol generation.^159,160 In addition, altered abundance of TMA-producing bacteria—including members of the Gammaproteobacteria and Erysipelotrichi classes—have been linked to choline deficiency and hepatic lipid accumulation, supporting a role for TMA/trimethylamine N-oxide (TMAO) signaling in MASLD pathogenesis.^161,162

A.3. Mechanisms linking gut dysbiosis to hepatic injury in MASLD

Gut dysbiosis plays a central role in MASLD pathogenesis by altering microbial metabolism and compromising intestinal barrier function, leading to increased hepatic exposure to harmful microbial products and metabolic stress.

Metabolite remodeling represents a critical early consequence of dysbiosis. SCFAs, including acetate, propionate, and butyrate, normally sustain epithelial energy metabolism, strengthen tight junctions, enhance GLP-1 secretion, and support hepatic fatty acid oxidation while dampening lipogenesis and inflammation.^163–165 Accordingly, SCFAs exert important hepatoprotective effects. However, excessive SCFA—particularly acetate—production under high-fat dietary conditions can fuel de novo lipogenesis and insulin resistance, suggesting stage- and dose-dependent impacts of SCFAs on MASLD.¹⁶⁶ Other metabolic pathways are similarly disrupted. In MASLD patients, tryptophan metabolism shifts away from the beneficial production of indole derivatives such as indole-3-propionic acid and indole-3-acetic acid, which normally reinforce epithelial barrier integrity via aryl hydrocarbon receptor signaling.¹⁶⁷ Meanwhile, overgrowth of high-alcohol-producing Klebsiella pneumoniae and related Proteobacteria results in endogenous ethanol accumulation, mitochondrial dysfunction, and CYP2E1-driven oxidative injury, mimicking ethanol-related steatohepatitis.^159,160,168 Increased levels of TMAO have also been observed in MASLD and are associated with gut barrier impairment, hepatocellular lipid accumulation, endothelial dysfunction, and higher mortality risk, although some experimental findings suggest metabolic effects may depend on dietary and microbial context.^169–172

A second major pathophysiological process involves gut barrier dysfunction. Human cohort studies and animal models demonstrate that intestinal permeability increases early in MASLD, even preceding advanced fibrosis.^173,174 Dysbiosis reduces abundance of butyrate-producing taxa needed to maintain epithelial turnover, mucus production, and tight junction assembly, while Western-style diets further compromise barrier integrity.¹⁷⁵ As epithelial protection declines, microbial-associated molecular patterns—including lipopolysaccharide, bacterial DNA, and endogenous ethanol—enter the portal circulation and activate TLR4, inflammasome pathways, and other innate immune sensors in Kupffer cells and hepatocytes.^176,177 This stimulates pro-inflammatory cytokine release, oxidative stress, and hepatic stellate cell activation, thereby driving steatohepatitis and fibrosis.

Beyond metabolite and barrier alterations, MASLD is characterized by significant changes in bile acid composition and bile acid receptor signaling. In MASLD and MASH, circulating and fecal bile acid levels are increased, with a shift toward hydrophobic and conjugated bile acids that exert greater cytotoxicity and pro-inflammatory potential.^178–181 Disruption of intestinal FGF19 signaling can attenuate negative feedback regulation of hepatic bile acid synthesis, leading to inappropriate or dysregulated activation of CYP7A1 and altered bile acid homeostasis in certain metabolic disease contexts.^179,182,183 Although persistent upregulation of bile acid synthesis is not uniformly observed across all stages of MASLD, disease-associated changes in bile acid composition and pool size have been linked to impaired epithelial barrier function, gut microbial dysbiosis, and hepatocellular inflammatory stress signaling.¹⁸⁴ Changes in the bile acid pool also alter activation of bile acid receptors, including FXR, TGR5, S1PR2, PXR, and VDR, which collectively regulate bile acid synthesis, lipid homeostasis, glucose metabolism, energy expenditure, inflammatory tone, and fibrogenesis.¹⁸⁵ Notably, both intestinal FXR activation and intestinal FXR deficiency have been shown to improve metabolic homeostasis in distinct experimental settings, reflecting differences in dietary composition, microbiota, and downstream FXR–FGF15/19 and ceramide signaling pathways.^186–188 Dysregulated FXR signaling is particularly important; hepatic FXR activation normally suppresses de novo lipogenesis, enhances β-oxidation, and limits bile acid toxicity, whereas impaired hepatic FXR signaling in MASLD exacerbates steatosis and bile acid overload.¹⁸⁹ In contrast, the role of intestinal FXR in MASLD is complex, which point to the multifaceted role of the gut–liver axis. Excessive intestinal FXR activation promotes ceramide synthesis and insulin resistance, indicating compartment-specific roles for FXR.^190,191 Conversely, intestinal FXR activation can also limit intestinal lipid absorption, reducing the flux of dietary lipids delivered to the liver and alleviating hepatic metabolic burden under certain conditions.¹⁹² These apparently contradictory findings underscore that FXR signaling exerts tissue-, ligand-, microbiota-, and diet-dependent effects, rather than uniform metabolic outcomes. Beyond FXR, bile acid signaling through TGR5, particularly in enteroendocrine cells, stimulates GLP-1 secretion and energy expenditure, but may exert divergent effects in hepatocytes depending on inflammatory and lipotoxic context.^193,194 In addition, activation of S1PR2 by conjugated primary bile acids, such as TCA and GCA, promotes pro-inflammatory signaling, suppresses peroxisome proliferator–activated receptor (PPAR) α-dependent lipid catabolism, and enhances fibrotic and tumorigenic pathways, particularly in MASLD-associated HCC.^104,110,195 While S1PR2 signaling contributes to inflammatory and tumor-promoting pathways in chronic liver disease, its physiological roles in bile acid sensing and hepatic adaptation under normal conditions remain incompletely defined. Dysregulated PXR and VDR signaling further influences detoxification, mucosal immunity, and hepatic substrate handling, contributing to disease heterogeneity.^196,197

Together, these interconnected mechanisms demonstrate that gut dysbiosis and bile acid remodeling function in a bidirectional feedback loop: dysbiosis alters the production and signaling potency of microbial metabolites and bile acids, while bile acid signaling defects further remodel the microbiota and weaken epithelial defenses. This vicious cycle strengthens maladaptive gut–liver communication and accelerates the transition to steatohepatitis and fibrosis. The integration of microbial metabolite regulation, intestinal barrier failure, bile acid compositional shifts, and disrupted bile acid receptor signaling forms a comprehensive molecular framework explaining how MASLD progresses toward more severe liver injury. This conceptual model also highlights multiple therapeutic opportunities, including microbiota modification, restoration of epithelial integrity, restructuring of the bile acid pool, and tissue-selective targeting of FXR, TGR5, S1PR2, and other bile acid–responsive signaling pathways.

B. Alcohol-associated liver disease (ALD)

ALD remains one of the leading causes of liver-related morbidity and mortality worldwide.^198,199 Its clinical spectrum ranges from simple alcoholic fatty liver to alcoholic steatohepatitis, and can further progress to fibrosis, cirrhosis, hepatic decompensation, and severe, often fatal, alcoholic hepatitis (AH).^200,201 Despite its high global burden, effective therapies for ALD remain limited. While direct ethanol toxicity and alcohol-induced immune responses are central drivers of ALD pathogenesis, accumulating evidence indicates that bile acid homeostasis is markedly disrupted, suggesting that alternations in the gut microbiome and bile acids may contribute to disease progression. These pathways therefore represent emerging, but not exclusive, avenues for improved diagnostic and therapeutic strategies.

B.1. Alterations of gut microbiome and bile acid profiles in ALD

Chronic alcohol consumption profoundly alters gut microbial ecology, leading to dysbiosis that contributes to disease progression.²⁰² High-throughput 16S rRNA gene sequencing and shotgun metagenomics from patients with ALD consistently reveal decreased microbial richness and evenness, along with distinct shifts in β-diversity profiles (Table 2). Alcohol-induced dysbiosis is characterized by a loss of beneficial commensals and an expansion of potentially pathogenic taxa, including members of the phyla Firmicutes, Verrucomicrobiota, and Bacteroidetes.^203,204 These microbial shifts occur in parallel with alterations in bile acid composition and metabolism, further exacerbating gut barrier dysfunction and hepatic inflammation.

Table 2.

Alterations of gut microbiome composition in ALD

Microbial taxa	Subjects	Findings	Reference
Bacteria
Phylum Bacteroidetes Phylum Proteobacteria	Colonic biopsy samples from patients with ALD	Bacteroidetes decreased and Proteobacteria increased	PMID: 22241860 536
Phylum Actinomycetota: Bifidob acteria, Atopobium Phylum Firmicutes: Streptococci	Fecal samples from patients with ALD and ethanol-fed mice	Bifidobacteria and Streptococci increased, Atopobium decreased in AH	PMID: 26642859 207
Phylum Actinomycetota: Bifidob acterium Phylum Firmicutes: Lactobacillus	Fecal samples from alcohol dependance syndrome patients with and without liver cirrhosis	Alcoholic dependence was inversely associated with the levels of butyrate-producing species from the Clostridiales order. Bifidobacterium and Lactobacillus increased in alcohol dependance syndrome patients.	PMID: 29041989 537
Phylum Verrucomicrobiota: Akkermansia muciniphila	Cecal contents from ethanol-fed mice; Fecal samples from AH patients	Reduced Akkermansia muciniphila in ALD correlated with disease severity, and oral supplementation improved gut barrier integrity and alleviates liver injury in experimental models.	PMID: 28550049 247
Phylum Actinobacteria, Phylum Bacteroidetes	Fecal samples from cirrhosis patients	An increased abundance of Actinobacteria and a reduction in Bacteroidetes were observed cirrhosis patients with severe AH	PMID: 30144108 211
Phylum Firmicutes Phylum Bacteroidetes	Whole blood samples from moderate and severe AH patients and controls	Bacteroidetes decreased, Fusobacteria increased in alcohol consumption control and AH patients	PMID: 29083504 204
Phylum Firmicutes: Enterococcus faecalis	Fecal samples from patients with AH and controls; ethanol-fed mice	The exotoxin-secreting gut bacterium Enterococcus faecalis is a critical contributor to AH, which can be eliminated with a bacteriophage	PMID: 31723265 226
Phylum Verrucomicrobiota: Akkermansia Phylum Bacteroidetes	Fecal and serum samples from patients with AUD	Microbial alpha diversity decreased in patients with AUD. Akkermansia decreased and Bacteroides increased. An increased expression of gamma-aminobutyric acid metabolic pathways and energy metabolism in AUD-associated gut microbiota.	PMID: 31951082 538
Phylum Verrucomicrobiota: Akkermansia Phylum Firmicutes: Veillonella	Fecal samples from patients with AH	Akkermansia decreased, Veillonella increased in severe AH patients	PMID: 32684075 242
Signatures	Fecal samples from healthy and heavy drinker controls, moderate and severe AH patients	A unique microbiome composition distinguished AH from heavy-drinking controls, with differential representation of Ruminococcaceae, Veillonellaceae, Lachnospiraceae, Porphyromonadaceae, and Rikenellaceae families. Severe AH was associated with increased Proteobacteria. Firmicutes abundance declined from HDC to MAH to SAH. SCFA-producing bacteria decreased in ALD patients.	PMID: 32056227 202
Phylum Firmicutes	Fecal samples from ethanol-fed mice	Butyrate-producing bacteria that utilize the acetyl-CoA pathway (predominantly Lachnospiraceae family) were decreased in ALD	PMID: 34369304 203
Phylum Firmicutes: Lachnospiraceae, Oscillospiraceae, Ruminococcaceae Phylum Proteobacteria	Integrated published data in patients with ALD	Multiple butyrate-producing families, including Ruminococcaceae, Lachnospiraceae, and Oscillospiraceae decreased, while endotoxin-producing Proteobacteria increased in AUD.	PMID: 36768785 206
Phylum Proteobacteria Phylum Bacteroides Phylum Fusobacteria	Fecal samples from patients with ALD in different stages	The abundance of Proteobacteria increased and Bacteroides decreased in ALD. Fusobacteria levels were higher in AH samples.	PMID: 37000389 539
Phylum Proteobacteria: Escherichia Shigella	Fecal samples from AUD patients with and without AH	Abnormal level of tryptophan increased the abundance of pathogen Escherichia Shigella.	PMID: 39567509 540
Phylum Bacteroides	Intestinal contents from Gao-binge model mice	Bacteroides acidifaciens increased in ALD mice with fiber diet, which activates the FXR-FGF15 signaling pathway in the gut via unconjugated bile acids thereby alleviating liver injury in ALD.	PMID: 38959900 239
Phylum Proteobacteria: Escherichia coli	Fecal samples from multicenter patients with ALD, mice models	KpsM-positive E. coli is correlated with patient mortality. KpsM-positive E. coli exacerbated ALD by the translocation of capsular polysaccharides to the cell surface to evade phagocytosis by the scavenger receptor Marco on Kupffer cells in the liver.	PMID: 40441146 225
Phylum Firmicutes:Lachnospir aceae	Fecal samples from patients with ALD and ethanol-fed mice	Lachnospiraceae bacterium decreased. Oral supplementation with Lachnospiraceae bacterium mitigates alcohol-induced liver damage both in vivo and in vitro by inhibiting ferroptosis through N-Acetyl-glutamic acid (NAG)-mediated activation of the KEAP1-NRF2 pathway	PMID: 40511521 541
Fungi
Candida albicans	Fecal samples from patients with ALD	Fungal diversity decreased and Candida species overgrew in ALD patients	PMID: 28530644 PMID: 31606552 PMID: 31228214 227,487,542
Candida, Debaryomyces, Pichia, Kluyveromyces, and Issatchenkia,	Fecal samples from patients with AUD	The abundance of the genera Candida, Debaryomyces, Pichia, Kluyveromyces, and Issatchenkia increased, whereas Aspergillus decreased in ALD. Malassezia was higher in AUD patients with progressive liver disease compared with non-progressive liver disease.	PMID: 34349667 543
Fungal signatures	Fecal samples from patients with ALD and MASLD	The genera Saccharomyces, Kluyveromyces, Scopulariopsis, and the species Candida albicans, Malassezia restricta, Scopulariopsis cordiae were significantly increased in patients with ALD	PMID: 38298161 544
Viruses
Phages, mammalian viruses	Fecal samples from patients with AUD	Viral diversity increased in ALD. Escherichia-, Enterobacteria-, and Enterococcus phages were over-represented in patients with AH. Mammalian viruses such as Parvoviridae and Herpesviridae increased in patients with AH.	PMID: 32654263 545
Phages	Fecal samples from patients with AUD	The abundance of phages targeting Enterobacteria and Lactococcus species phages increased in patients with progressive liver diseases.	PMID: 35368152 546
Lactococcus phages	Fecal samples from patients with AUD and MASLD	Lactococcus phages differed between alcohol-consuming and non-alcohol-consuming patients	PMID: 36631002 361
Endogenous retroviruses	Duodenal and liver biopsies from patients with ALD and AUD	HERV expression increased in the duodenum and liver. Ethanol exploits ERV/Zbp1 pathway to promotie innate immune responses, resulting in gut barrier dysfunction and liver disease.	PMID: 40359032 547

Abbreviations: AH, alcoholic hepatitis; ALD, Alcohol-associated liver disease; AUD, alcohol use disorder; MASLD, metabolic dysfunction-associated steatotic liver disease; KEAP1, Kelch-like ECH-Associating protein1; NRF2, NF-E2-related factor 2.

Functionally, chronic and heavy alcohol consumption leads to gut microbial communities with reduced metabolic capacity, as reflected by decreased microbial genes involved in essential metabolic pathways in metagenomic analyses.²⁰⁴ A notable feature of this functional impairment is the significant loss of SCFA–producing bacteria, including key butyrate-producing taxa, observed in both experimental models and patients with ALD.^{202,203,205,206} SCFAs are critical for preserving epithelial integrity, modulating immune responses, and maintaining gut microbial homeostasis; thus, their depletion contributes to gut barrier dysfunction and heightened inflammation. Supporting a causal role for dysbiosis in ALD progression, fecal microbiota from patients with severe AH induce pronounced liver inflammation and injury when transferred into humanized mouse models.²⁰⁷

Bile acids are major gut microbiome–dependent metabolites that exert critical regulatory functions in liver physiology, immunity, and metabolic homeostasis. In patients with AH, serum levels of total and conjugated bile acids are significantly increased,²⁰⁸ alongside elevated secondary bile acid concentrations that closely correlate with disease severity.^207,209 In contrast, urinary secondary bile acids are reduced in severe AH patients,²¹⁰ consistent with other studies demonstrating a shift in the bile acid pool toward more hydrophobic and hepatotoxic species in advanced ALD and cirrhosis. ²¹¹ Further analysis of urinary bile acid profiles in patients with ALD showed that progression of ALD is accompanied by increased total urinary bile acid levels, an elevated primary-to-secondary bile acid ratio, and decreased conjugated-to-unconjugated bile acid ratios.²¹⁰ These metabolic alterations strongly correlate with Model for End-Stage Liver Disease scores, highlighting their potential as noninvasive biomarkers for monitoring disease severity and progression. A summary of bile acid profiling studies in ALD is provided in Table 3.

Table 3.

Alterations of bile acid profiles in ALD

Subjects	Alterations of bile acids in patients/ mice models	Reference
Serum samples from AH patients	Deoxycholate and glycodeoxycholate decreased	PMID: 25461442 548
Fecal and serum samples from patients	Endotoxin levels, serum conjugated bile acids (including conjugated DCA), and stool total bile acids were highest in current drinkers. In contrast, secondary bile acid formation was impaired, resulting in a reduced secondary-to-primary bile acid ratio. These bile acid alterations were associated with increased circulating inflammatory cytokines.	PMID: 24699327 209
Caecal samples from AH mice	Primary bile acid CDCA decreased and secondary bile acid UDCA decreased.	PMID: 26642859207
Serum from AH patients	Total and conjugated bile acids increased and correlated with disease severity	PMID: 29654817 208
Plasma and fecal samples from cirrhosis patientsS	Total bile acids and UDCA increased in serum, and total bile acids decreased in feces for those with severe AH	PMID: 30144108 211
Liver, gallbladder, intestine and plasma and colon samples from ethanol-fed mice	Free-bile acids, taurine-bile acids, glycine-bile acids, and total bile acids increased	PMID: 32232424 549
Urine samples from AH patients	CA, CDCA, and apocholic acid are increased in patients with alcoholic hepatitis (AH). Overall, primary bile acids are increased whereas secondary bile acids are decreased. Total bile acid levels and the primary-to-secondary bile acid ratio are elevated, while the ratio of unconjugated to conjugated bile acids decreases with increasing ALD severity.	PMID: 34027270 210
Plasma and fecal samples from patients with AH	Conjugated CA and CDCA increased in plasma, whereas deoxycholic acid decreased. Plasma TCDCA and TUDCA were related to disease severity, whereas fecal UDCA was inversely related.	PMID: 34984859 236
Fecal samples from ALD patients in different stages	Lithocholic acids decreased in alcoholic cirrhosis patients compared to healthy control	PMID: 37000389 539
Serum and fecal samples from AH patients	The ratio of glycine- to taurine-conjugated bile acids decreased after corticosteroid therapy, reflecting a relative enrichment of taurine-conjugated bile acids.	PMID: 39078043 550
Serum samples from ethanol-fed mice	Elevated serum total bile acid levels are associated with increasing disease severity in ALD.	PMID: 39398234 551
Serum samples from progressive and nonprogressive ALD patients	Glycine-conjugated bile acids are increased with disease progression in ALD	PMID: 40358071 552

Abbreviations: AH, alcoholic hepatitis; ALD, Alcohol-associated liver disease; CA, cholic acid; DCA, deoxycholic acid; CDCA, chenodeoxycholic acid; UDCA, ursodeoxycholic acid; TUDCA, tauroursodeoxycholic acid.

B.2. Mechanisms of gut dysbiosis in ALD

Alcohol consumption disrupts intestinal homeostasis at multiple levels, contributing to dysbiosis and disease progression in ALD. Ethanol impairs epithelial integrity, alters immune and metabolic signaling, and promotes translocation of microbial products to the liver, driving inflammation and fibrosis.

B.2.1. Impaired intestinal barrier integrity

Alcohol directly injures intestinal epithelial cells and dysregulates tight junction proteins, weakens the mucus layer, and reduces antimicrobial defenses collectively increasing gut permeability. ^18,212,213 Following alcohol consumption, approximately 70% of ethanol is absorbed in the proximal intestine and subsequently metabolized predominantly in the liver, where it is oxidized to acetaldehyde by alcohol dehydrogenase (ADH) in the cytosol of hepatocytes, cytochrome P450 2E1 (CYP2E1) in microsomes, and catalase in peroxisomes. Acetaldehyde is then further oxidized to acetate mainly by aldehyde dehydrogenase. Both oxidative and non-oxidative ethanol metabolites contribute to impairment of intestinal epithelial barrier integrity.²¹⁴

Ethanol disrupts intestinal barrier function through multiple mechanisms, including direct epithelial cell injury and destabilization of tight and adherent junctions. Chronic ethanol consumption has been shown to cause ultrastructural and histological alterations in the intestinal mucosa, including epithelial cell loss and a reduction in overall mucosal surface area. At the molecular level, ethanol and its metabolites impair barrier integrity by perturbing junctional organization and cytoskeletal dynamics through oxidative stress–related and post-translational mechanisms. Notably, acetaldehyde exhibits marked cytotoxicity by inducing chromosomal instability and forming adducts, which in turn promote inflammation and lipid accumulation.^214,215

These alcohol-induced permeability changes are closely associated with gut dysbiosis, partly mediated by reduced activity and expression of intestinal hypoxia-inducible factor 1α.^216–218 Surprisingly, mice lacking intestinal mucin-2 displayed reduced bacterial overgrowth, lower levels of translocated microbial products, and attenuated liver injury following alcohol feeding, suggesting that the mucus layer—while normally protective—may facilitate the expansion of pathogenic bacteria in ALD.²¹⁹ Alcohol also lowers indole-3-acetic acid and suppresses aryl hydrocarbon receptor (AHR) signaling, reducing IL-22–dependent antimicrobial regenerating islet–derived protein 3 lectins, facilitating bacterial translocation and hepatic inflammation.^220,221 These changes may additionally contribute to neuropsychiatric manifestations of alcohol dependence.²²²

B.2.2. Endotoxins and exotoxins

Alcohol-induced dysbiosis and permeability increase systemic exposure to bacterial products such as LPS. Elevated endotoxemia in severe AH activates hepatic Kupffer and stellate cells, promoting fibrosis and inflammation.^223–225 Beyond LPS, exotoxins add a pathogenic burden. Enterococcus faecalis–derived cytolysin induces hepatocyte death; bacteriophage-based eradication prevents ethanol-induced injury.²²⁶ Candidalysin, produced by Candida albicans, exacerbates liver injury via C-type lectin domain family 7 member A (CLEC7A) signaling and increases mortality without altering permeability.²²⁷ Thus, both bacteria and fungi contribute to toxic inflammatory signaling in ALD.

B.2.3. Immunological modulation

Ethanol and its metabolites disrupt the intestinal homeostasis, enabling translocation of pathogen-associated molecular patterns to the liver, where they trigger immune activation and inflammation through pattern recognition receptors. The resulting immune cells and their inflammatory mediators contribute significantly to the progression of ALD.^228,229 Key immunologic changes include: 1) expansion of Candida albicans-specific Th17 cells promoting liver damage via IL-17; antifungal treatment reverses this effect.²³⁰ 2) Loss of duodenal CD8⁺ T cells, correlating with increased bacterial translocation; restoring CD8⁺ T cell survival strengthens barrier function and reduces liver injury. ²³¹ 3) Loss of conventional type 1 dendritic cells worsens alcohol-induced liver disease. Conventional type 1 dendritic cells protect against ALD by sustaining antimicrobial peptide expression and maintaining Akkermansia abundance through IL-12–IFN-γ signaling.^232,233 Together, these immunologic shifts amplify hepatic inflammation and fibrosis.

B.3. Mechanisms of bile acids remodeling in ALD

Bile acids are synthesized in the liver, secreted into the duodenum, and then subsequently modified by the gut microbiota. Therefore, the alterations of bile acid profiles in ALD reflect coordinated disruptions in hepatic metabolism and microbe-mediated biotransformation. Multiple aspects of bile acid homeostasis—including synthesis, transport, biotransformation, enterohepatic circulation, and receptor signaling—are markedly altered in ALD.^200,234

Alcohol impairs hepatic bile acid synthesis, as evidenced by suppressed CYP7A1 expression and reduced circulating C4 levels in AH patients.²⁰⁸ Concurrent dysbiosis alters microbial enzymes such as BSH, reducing deconjugation and secondary bile acid generation, while the depletion of bacteria with 3α-, 3β-, 7α-, and 7β-hydroxysteroid dehydrogenase/epimerase activities, altering bile acid hydroxyl group epimerization.^235,236 Disruption of bile acid receptor signaling, including deficiency of TGR5, exacerbates ethanol-induced liver injury through enhanced macrophage recruitment and altered microbial and bile acid composition.²³⁷

As signaling molecules, bile acids regulate multiple metabolic and inflammatory processes in ALD, primarily through related transporters and receptors such as FXR. Alcohol-driven dysbiosis elevated intestinal miR-194 via the taurine-upregulated gene 1 (Tug1)–taurine axis to silence FXR transcription, thereby exacerbating ALD.²³⁸ Bacteroides acidifaciens produces unconjugated bile acids that activate intestinal FXR, leading to increased ornithine aminotransferase expression and thereby alleviating ALD via glutamine synthesis pathway.²³⁹ Administration of probiotic-derived nanoparticles from Lactobacillus rhamnosus GG restored FXR activity by modulating the intestinal FXR signaling axis and consequently ameliorated ALD. Consistently, patients with AH exhibit markedly elevated serum FGF15 and FGF19 levels.^208,235 Beyond their signaling functions, bile acids also play an essential role in alcohol metabolism and interorgan communication along the gut–liver axis. Biliary transport of acetaldehyde from the liver to the gut represents a key route for acetaldehyde clearance and metabolic turnover. Modulation of bile secretion and flow profoundly influence alcohol detoxification efficiency and drinking behavior.²⁴⁰

B.4. Therapeutic strategies and future perspectives in ALD

The gut microbiome and bile acid profiles are increasingly recognized as potential biomarkers for early detection, assessing severity and modifiable therapeutic targets for ALD. Early efforts to treat ALD have involved the use of antibiotics.²⁴¹ Subsequent studies have expanded this perspective, demonstrating that specific gut microbiome and metabolome signatures could reliably differentiate ALD and reflect disease severity.^202,207,242 With the identification of microbiota members that act as protective strains or pathobionts, microbiome-based interventions to restore gut microbial homeostasis are developed, including prebiotic - postbiotic administration, fecal microbiota transplantation (FMT).^{218,243–248} In addition, selectively targeting key pathobionts such as the application of bacteriophages also offers a strategy to modulate gut microbiota and eliminate pathogenic endogenous bacteria.²²⁶

Targeting bile acid signaling represents another promising approach for ALD therapy and will be discussed in greater detail in the following sections. In parallel, several novel pharmacological approaches for treating ALD are under investigation, including agents targeting the stress response, neuroinflammation, the aldosterone/mineralocorticoid receptor system, and neuroendocrine pathways such as ghrelin and GLP-1, highlighting the potential for individualized, mechanism-based interventions in managing ALD.²⁴⁹

Despite considerable progress, several challenges remain. For instance, fecal microbiota may not accurately reflect microbial composition within the lumen, especially in the small intestine. Considering the intricate interplay between the gut microbiome and bile acid metabolism, combined strategies targeting both systems may achieve synergistic therapeutic effects. Future research should continue to elucidate the mechanistic interplay between the gut microbiome and bile acids and develop precise microbiome- and bile acid-targeted therapeutic strategies to advance the prevention and treatment of ALD.

Taken together, alcohol disrupts microbial homeostasis, impairs intestinal barrier integrity, and alters intestinal and hepatic functions, especially bile acid metabolism, intricately related to the development and progression of ALD. Restoring microbial and metabolic homeostasis holds promise for future therapies.

C. Gut microbiome and bile acids in cholestatic liver diseases

Cholestatic liver diseases comprise a heterogeneous group of disorders defined by impaired bile flow, resulting in intrahepatic or extrahepatic cholestasis. Cholestasis may arise from defects at multiple levels of bile acid handling, including bile acid formation (hepatic synthesis and intracellular processing), canalicular secretion from hepatocytes into bile, or excretion/transport through the biliary tree into the intestine. Their etiologies include immune-mediated injury, genetic defects, environmental triggers, and drug toxicity. Major cholestatic liver diseases encompass PBC, PSC, and biliary obstruction–associated cholestasis, and inherited disorders such as progressive familial intrahepatic cholestasis, as well as pediatric and pregnancy-associated conditions including biliary atresia and intrahepatic cholestasis of pregnancy.^11,250,251 Defective bile acid synthesis or aberrant bile acid production may lead to excessive accumulation of bile acids, as seen in inborn errors of bile acid synthesis. Defects in bile acid secretion, such as BSEP or MDR3 deficiencies seen in progressive familial intrahepatic cholestasis, result in ineffective hepatocellular bile acid transport. Bile acid excretion impairment is most commonly caused by obstruction of large, medium, or small bile ducts from various etiologies, leading to bile stasis within the bile ducts.^252,253 Despite their heterogeneity, cholestatic liver diseases share a central pathological feature: reduced bile flow into the duodenum, leading to hepatic accumulation of toxic bile acids and progressive hepatocellular and biliary injury.²⁵⁴ Unlike other liver diseases in which metabolic or inflammatory insults dominate, dysregulation of bile acid homeostasis is a primary pathogenic driver in cholestatic liver diseases. Recent advances in bile acid metabolism and nuclear receptor signaling have revealed the critical contribution of the bile acid–gut–liver axis to cholestatic liver disease, wherein altered bile acid pools reshape the gut microbiome, and dysbiosis further exacerbates cholestatic injury.

C.1. Gut microbiome composition shifts in cholestatic liver diseases

Despite variations among studies, there is a clear consensus that cholestatic liver diseases are associated with significant gut microbiome dysregulation. Common features include: 1) reduced microbial richness and diversity, (2) expansion of potentially pathogenic or pro-inflammatory taxa, and (3) reduced abundance of beneficial bacteria and their hepatoprotective metabolites. These changes reflect a loss of microbial functional capacity to generate short-chain fatty acids and other protective metabolites, along with increased enrichment of harmful taxa such as Escherichia coli and Enterococcus faecalis. Decreased α-diversity has been consistently documented across cholestatic liver diseases cohorts.^255–257 Altered community structure frequently features enrichment of Streptococcus and Veillonella accompanied by reduced Faecalibacterium, which has been observed across multiple etiologies including biliary atresia, PSC, and PBC.²⁵⁸ In biliary atresia, pathogenic taxa such as Streptococcus and Klebsiella are markedly elevated and correlate with disease severity.^250,259 Notably, Klebsiella and Veillonella atypica positively correlate with serum transaminases, whereas Enterococcus faecium associates with elevated LCA derivatives.²⁵⁹ Patients with recurrent cholangitis show enrichment of Lachnospiraceae and Ruminococcaceae, while postoperative jaundice resolution is linked with increases in Campylobacter and Rikenellaceae.²⁶⁰

Most adult-focused research has centered on PSC and PBC. One of the most reproducible findings is increased abundance of Veillonella species across PSC/PBC cohorts.^261,262 Additional PSC-associated shifts include increases of Enterococcus, Streptococcus, Veillonella and decreased of Adlercreutzia equolifaciens, Prevotella copri.²⁶³ Across PSC cohorts, reduced Firmicutes (other than Veillonella) and increased Proteobacteria are frequently observed. Health-associated genera such as Alistipes, Blautia, Coprococcus, Roseburia, and Eubacterium hallii inversely correlate with liver injury markers.

In PBC, gut microbial composition is linked to disease severity and treatment response. Klebsiella enrichment correlates with elevated serum bilirubin and may serve as a prognostic indicator.²⁶⁴ Depletions of Faecalibacterium, Bacteroides, Sutterella, and Oscillospira are reported, with lower Faecalibacterium abundance in gp210-positive patients. UDCA therapy partially restores gut microbiota balance.^255,262

C.2. Functional alterations of gut microbiome in cholestatic liver diseases

The pathophysiology of cholestatic liver diseases arises not only from compositional shifts in the gut microbiota but also from impaired microbial function, including altered metabolite production, compromised barrier integrity, and dysregulated host–microbe immune signaling. These changes contribute directly to hepatic inflammation and cholestatic injury.²⁶⁵

FMT from patients with PBC into germ-free mice induces PBC-like liver injury—including elevated serum alkaline phosphatase (ALP) and cholangitis—or exacerbates disease in susceptible hosts, demonstrating causality between gut dysbiosis and PBC progression.²⁶⁶ Microbiota depletion prevents primary-to-secondary bile acid conversion, leading to accumulation of β-muricholic acid—an FXR antagonist—that increases bile acid synthesis and worsens hepatic injury in Mdr2^−/− mice.^85,267 These findings highlight the need to account for species-specific differences in bile acid profiles when interpreting experimental cholestatic liver diseases models.

The increased intestinal mucosal permeability in Mdr2^−/− mice allows Lactobacillus gasseri to translocate to the liver, where it induces IL-17 production. This bacterial translocation further activates the NLRP3 inflammasome, ultimately driving the progression of inflammation in PSC.^268,269 Similarly, Lactococcus garvieae disrupts epithelial tight junctions (reduced occludin and zonula occludens-1) and enhances bile acid reabsorption by upregulating ASBT, further aggravating cholestasis.²⁷⁰ Clinical and experimental studies show consistent enrichment of Klebsiella pneumoniae, Proteus mirabilis, and Enterococcus gallinarum in PSC patients. K. pneumoniae is particularly potent in compromising mucosal defenses; together these pathobionts breach the gut barrier and activate liver-infiltrating Th17 cells, exacerbating DDC-induced liver injury.²⁷¹ Direct evidence of bacterial translocation into the liver has also been demonstrated in PBC, further confirming gut–liver immune crosstalk in cholestatic liver diseases.²⁷²

C.3. Mechanisms linking gut dysbiosis to hepatic Injury in cholestatic liver diseases

The gut dysbiosis contributes to cholestatic liver diseases progression by impairing bile acid biotransformation, disrupting intestinal barrier function, and promoting inflammatory immune activation along the gut–liver axis. Altered microbiota composition weakens BSH and other bile-acid transforming activities. For example, Bacteroides fragilis is enriched in intrahepatic cholestasis of pregnancy and suppresses FXR signaling by altering BSH activity, thereby disrupting bile acid metabolism.²⁷³ Microbial depletion in CYP2C70^−/− mice, which have a human-like bile acid profile, increases circulating hydrophobic bile acids and exacerbates cholestatic liver injury.²⁷⁴ Beyond bile acids, reduced short- SCFAs production — especially butyrate — is observed in PSC and PBC andcorrelates with fibrosis severity.^255,275 Butyrate supplementation suppresses cholangitis through myeloid-derived suppressor cell activation.²⁷⁶ Metabolomic profiling in cholestatic liver diseases also reveals reduced branched-chain amino acids influenced by dysbiotic metabolism and impaired vitamin B6 synthesis, resulting in lower pyridoxal-5′-phosphate linked to poor prognosis.^277–279 Dysbiosis-induced barrier dysfunction enables bacterial translocation and hepatic inflammation. In PSC models, Lactobacillus gasseri, Klebsiella pneumoniae, Proteus mirabilis, and Enterococcus gallinarum penetrate the liver, activate Th17 responses, and worsen cholangiopathy.²⁷¹ Direct microbial translocation is also evident in PBC. ²⁷²

C.4. Altered bile acid composition in cholestatic liver diseases

Bile flow obstruction reduces intestinal bile acid delivery and results in elevated serum and hepatic bile acid levels, increased primary: secondary bile acid ratios, and higher conjugated bile acids due to impaired deconjugation.²⁸⁰ UDCA therapy partially normalizes these patterns by lowering taurine-conjugated bile acids.²⁸¹ In PSC, CA/CDCA and (CA+DCA)/(CDCA+LCA) ratios are decreased.²⁸² Low fecal bile acids indicate reduced substrates for microbial conversion.^283,284 Mouse models demonstrate that bile acid profile drives toxicity—mice lacking mouse-specific MCAs or expressing humanized bile acid pools develop more severe cholestasis.²⁸⁵ Adaptive responses to bile acid overload include increased urinary and fecal sulfated/glucuronidated bile acids, enhancing elimination of toxic hydrophobic species. ²⁸⁶

C.5. Bile acid receptors as key modulators of liver injury in cholestatic liver diseases

Bile acid receptors integrate metabolic, immune, and epithelial responses to cholestatic stress. FXR activation reduces bile acid accumulation by inhibiting bile acid synthesis, increasing efflux and reducing hepatocyte reuptake.²⁸⁷ FXR agonists [e.g., obeticholic acid] restore intestinal integrity, suppress NF-κB inflammation, and reduce bacterial translocation.^288–290 The non-steroidal agonist tropifexor improves FGF19-mediated feedback and enhances barrier maintenance.²⁹¹ Prevotella copri also activates intestinal FXR and reduces cholestasis.²⁹² Importantly, microbiota-mediated bile acid synthesis feedback regulation depends on the FXR pathway. Upon depletion of the gut microbiota, FXR is suppressed in Mdr2^−/− mice, resulting in remodeling of hepatic bile acids.²⁹³ Inhibiting ileal bile acid transport in Mdr2^−/− mice attenuates the bile acid dysregulation and hepatotoxicity caused by microbiota depletion, highlighting the essential role of intestinal FXR signaling in maintaining bile acid homeostasis through gut microbiota–mediated feedback regulation.²⁹⁴

TGR5 knockout in cholestatic mice leads to more severe liver injury, necrosis, inflammation, bile duct reaction and increased hydrophobicity of the bile acid pool.²⁹⁵ TGR5 activation in Kupffer cells suppresses LPS-induced cytokine production through a cAMP-dependent pathway and inhibits TLR4–NF-κB signaling, thereby limiting pro-inflammatory M1 macrophage polarization.^89,296,297 In cholangiocytes, TGR5 promotes proliferation, inhibits apoptosis, and preserves epithelial integrity, reducing bile leakage-mediated hepatocyte damage.^298–300 Through cAMP signaling, TGR5 also stimulates chloride and bicarbonate secretion to enhance biliary output and protect against the toxicity of hydrophobic bile acids.^301,302 Additionally, bile acid–dependent activation of TGR5 supports liver regeneration by promoting the differentiation of mesenchymal stem cells into hepatocytes.³⁰³ While these survival and proliferative signals are beneficial for cholestatic repair, they may also contribute to cholangiocarcinoma risk when chronically activated.

S1PR2 contributes to cholestasis-induced cholangiocyte proliferation by activating ERK1/2–AKT signaling.³⁰⁴ In macrophages, S1PR2 activation promotes NLRP3 inflammasome signaling and necroptosis, thereby amplifying hepatic inflammation.^305–307 TCA-induced activation of S1PR2 serves as a signal that triggers hepatic stellate cells activation and extracellular matrix (ECM) production, thereby promoting the fibrosis progression.³⁰⁸ Interestingly, recent work demonstrates that S1PR2 modulation improves gut microbial diversity and increases the abundance of short-chain fatty acid–producing taxa such as Romboutsia, indicating a regulatory role for S1PR2 in gut–liver axis homeostasis during cholestasis.³⁰⁹

PXR, constitutive androstane receptor (CAR), and VDR also help mitigate cholestatic injury by coordinating bile acid detoxification and efflux.³¹⁰ These receptors induce basolateral transporters MRP3 and MRP4 to enhance hepatocellular bile acid export.³¹¹ PXR/CAR activation upregulates detoxifying enzymes including CYP3A, SULT, UGT, and PAPSS2, which hydroxylate, conjugate, and promote urinary elimination of toxic bile acids; loss of PXR or CAR worsens bile acid accumulation and liver damage.³¹² VDR additionally suppresses NF-κB activation and NLRP3 inflammasome signaling, reducing inflammation in cholestatic injury.³¹³ Across diverse etiologies, cholestatic liver disease share convergent pathogenic mechanisms centered on bile acid overload, receptor dysfunction, and gut–liver axis breakdown. Reduced bile flow not only increases hepatic exposure to toxic bile acids but also perturbs gut microbial composition and intestinal permeability, facilitating translocation of inflammatory microbial metabolites to the liver. The interplay among dysbiosis, impaired bile acid signaling, and chronic inflammation establishes a reinforcing cycle that drives the progression of cholestatic liver disease

D. Liver Cancers

Liver cancer is one of the leading causes of cancer-related mortality worldwide. HCC accounts for approximately 80–90% and cholangiocarcinoma for 10–15% of primary cases.³¹⁴ In children, hepatoblastoma is the most common primary hepatic malignancy.³¹⁵ Importantly, liver metastases arising from extrahepatic tumors are 18–40 times more common than primary liver cancers.³¹⁴ Growing evidence demonstrates that gut microbiota dysbiosis and bile acid remodeling are key drivers of hepatocarcinogenesis, tumor progression, and immune evasion in liver cancers. Through the gut–liver axis, these systems establish an intricate biochemical and immunological network that shapes tumor metabolism and the tumor microenvironment (TME).

D.1. Alterations of gut microbiome and bile acid profiles in liver cancers

D.1.1. Alterations of gut microbiome in HCC

The increasing incidence of MASLD-associated HCC underscores the critical role of gut dysbiosis in metabolic liver carcinogenesis. HCC patients consistently exhibit decreased beneficial SCFA-producing taxa and increased pro-inflammatory/pathogenic bacteria. Moreover, distinct microbial signatures differentiate MASLD-HCC, viral HCC, and alcohol-related HCC (Table 4). Microbiome alterations also vary by disease stage. Early-stage HCC shows enrichment of Gemmiger and Parabacteroides, while advanced or metastatic HCC displays a marked loss of beneficial bacteria such as Anaerotruncus colihominis.³¹⁶ In contrast, advanced or metastatic HCC shows pronounced depletion of beneficial taxa, including Anaerotruncus colihominis and Dysosmobacter welbionis.³¹⁷ In HBV-related HCC, higher tumor burden has been linked to increased abundances of Bacteroides, Lachnospiraceae incertae sedis, and members of Clostridium cluster XIVa.³¹⁸ Functionally, the gut microbiome in MASLD and HCC is enriched in bacterial pathways responsible for the generation of SCFAs from dietary fiber, highlighting a shift in microbial metabolic activity during disease progression.³¹⁹ Experimental evidence further underscores the pathological consequences of altered microbial metabolism: soluble dietary fiber induces cholestasis and liver tumorigenesis in dysbiotic mice, while depletion of secondary bile acid– producing bacteria or administration of bile acid sequestrants such as cholestyramine effectively prevents HCC development.³²⁰ In addition, the gut microbiota also contributes to treatment responsiveness. Distinct microbial signatures differentiate responders from non-responders to immune checkpoint inhibitors (ICIs) in patients with unresectable HCC, suggesting that microbiome composition may influence antitumor immunity and immunotherapy outcomes.³²¹

Table 4.

Alterations of gut microbiome composition in primary liver cancers

Microbial taxa	Subjects	Findings	Reference
HCC
Phylum Actinobacteria	Fecal samples from patients with HCC and cirrhosis	Actinobacteria increased in early HCC versus cirrhosis.	PMID: 30045880 316
Phylum Bacteroidetes Phylum Firmicutes Phylum Actinomycetota: Bifidobacterium	Fecal samples from patients with MASLD-HCC, MASLD-cirrhosis and HC	Bacteroides and Ruminococcaceae increased in HCC, while Bifidobacterium decreased.	PMID: 29665135 146
Phylum Bacteroidetes Phylum Firmicutes	Fecal samples from patients with HBV-related HCC	Bacteroides, Lachnospiracea incertae sedis, and Clostridium XIVa enriched in HCC patients with a high tumor burden	PMID: 33225985 318
Phylum Proteobacteria: Enterobacteriaceae	Fecal samples from patients with MASLD-HCC, MASLD-cirrhosis and HC	At the family level, Enterobacteriaceae increased, while at species level, Bacteroides caecimuris and Veillonella parvula enriched in MASLD-HCC compared to MASLD-cirrhosis. Bacterial genes involved in SCFAs synthesis from dietary fibre characterised the microbiome of MASLD-HCC.	PMID: 33420074 319
Phylum Bacteroidetes Phylum Firmicutes	Fecal samples from patients with unresectable HCC	Prevotella was enriched in patients with progressive disease. Lachnoclostridium, Lachnospiraceae, and Veillonella were predominant in patients with objective responses.	PMID: 35738801 321
Phylum Actinobacteria	Fecal and liver samples MASLD-HCC mice	Bifidobacterium pseudolongum suppressed MASLD-HCC progression	PMID: 37459922 350
Phylum Proteobacteria	Fecal and liver samples from patients with HCC	Klebsiella pneumoniae enriched in patients with HCC and FMT mice	PMID: 39747695 349
Phylum Firmicutes	Fecal samples from patients with intrahepatic metastatic HCC and non-metastatic HCC	Beneficial species such as Anaerotruncus colihominis and Dysosmobacter welbionis decreased in patients with intrahepatic metastatic HCC	PMID: 40099491 317
CCA
Phylum Proteobacteria: H. pylori	Epithelial cells from the biliary duct	Fusobacterium, Prevotella, Actinomyces, Novosphingobium, and H. pylori increased in patients with CCAs.	PMID: 26493848 553
Phylum Firmicutes Phylum Actinobacteria	Fecal and serum samples from patients with iCCA, HCC and cirrhosis and HC	Lactobacillus, Actinomyces, Peptostreptococcaceae, and Alloscardovia were increased in patients with ICC. A greater abundance of the family Ruminococcaceae was in patients with vascular invasion than those without.	PMID: 31298745 324
Phylum Firmicutes Phylum Bacteroidetes	Bile samples from patients with CCA	Firmicutes decreased and Bacteroidetes increased in CCA. The most abundant genera were Enterococcus, Streptococcus, Bacteroides, Klebsiella, and Pyramidobacter in CCA’s biliary microbiota.	PMID: 33690612 554
Phylum Firmicutes Phylum Fusobacteria Phylum Actinobacteria	Fecal and bile samples from patients with CCA, PSC and HC	Firmicutes, Fusobacteria, and Actinobacteria increased in bile of CCA compared with HC	PMID: 35565248 555
Phylum Proteobacteria Phylum Firmicutes	Fecal and bile samples from patients with BTC and BBD	Enterobacteriaceae increased while Clostridia (Faecalibacterium and Coprococcus) decresed in BTC	PMID: 36358797 556
Phylum Verrucomicrobiota: Akkermansia	Saliva, gastric and pancreatic juice, bile, feces, tumor and nontumor tissue from patients with pancreatic or BTC	Akkermansia was only detected in the bile of patients with BTC	PMID: 35610303 557
/	Fecal and plasma samples from patients with iCCA and HC	Richness index and diversity index decreased in iCCA patients.	PMID: 38582012 359

Abbreviations: HCC, hepatocellular carcinoma; MASLD, metabolic dysfunction-associated steatotic liver disease; HC, healthy control; HBV, hepatitis B virus; CCA, cholangiocarcinoma; iCCA, intrahepatic cholangiocarcinoma; PSC, primary sclerosing cholangitis; BBD, benign biliary disease; BTC, biliary tract cancer.

D.1.2. Alterations of gut microbiome in cholangiocarcinoma and other liver cancers

Cholangiocarcinoma, the second most common primary liver cancer arising from cholangiocytes, carries a dismal prognosis with a 5-year overall survival rate below 10%.³²² Based on its anatomical location, cholangiocarcinoma is classified into intrahepatic, perihilar, and distal subtypes, each with distinct clinical and biological characteristics.³²³ Recent evidence suggests that gut microbiome dysbiosis is implicated in cholangiocarcinoma pathogenesis. In particular, patients with intrahepatic cholangiocarcinoma display elevated α- and β-diversity compared to healthy controls and those with cirrhosis or HCC. Notably, the combined abundance of Lactobacillus and Alloscardovia shows strong discriminatory power for identifying intrahepatic cholangiocarcinoma across disease states.³²⁴ Furthermore, enrichment of Ruminococcaceae has been associated with vascular invasion and worse prognosis, implicating specific microbial taxa in tumor aggressiveness.³²⁴

The liver is also a frequent metastatic site for gastrointestinal malignancies, especially colorectal cancer, because of direct venous drainage through the portal circulation. Microbial drivers of metastasis are emerging as key contributors in this process. For instance, viable Fusobacterium species derived from primary colorectal tumors has been detected in liver metastases, suggesting that this pathobiont may facilitate metastatic colonization and tumor progression.³²⁵

In addition to luminal dysbiosis, several independent studies have demonstrated the presence of intratumoral microbiota in liver cancers.^326–328 Tumor-resident microbial communities differ from the surrounding host microbiota, and these differences may represent actionable therapeutic vulnerabilities. Moreover, intratumoral microbial heterogeneity contributes to tumor subclassification and correlates with clinical outcomes, highlighting the potential utility of microbial signatures in diagnostic and prognostic stratification.³²⁹

D.1.3. Alterations of bile acid profiles in HCC

Patients with HCC exhibit marked alterations in systemic bile acid composition, characterized predominantly by elevated total bile acids driven by increased levels of conjugated primary species such as GCA, TCA, GCDCA, and TCDCA. Sulfated and glucuronidated bile acids are also significantly enriched in the circulation of HCC patients.³³⁰ Among these, GCA shows the most consistent and substantial increase, demonstrating a strong positive association with HCC incidence and progression.³³¹ Similarly, both total and specific taurine- and glycine-conjugated bile acids (TCDCA, TCA, and GCDCA) are significantly elevated in HCC liver tissue.³³²

Bile acid profiles are also linked to risk stratification in HCC. Elevated levels of primary bile acids and a predominance of taurine- over glycine- conjugation are positively associated with HCC risk, whereas higher secondary-to-primary bile acid ratios appeared to confer a protective effect.³³⁰ Additionally, microbiome–bile acid interactions contribute to treatment response: fecal UDCA and UCA levels are enriched in immunotherapy responders relative to non-responders, in association with increased abundance of Lachnoclostridium.³²¹

D.1.4. Alterations of bile acid profiles in cholangiocarcinoma and other liver cancers

In cholangiocarcinoma, both total and conjugated bile acids are elevated relative to benign biliary disease.³³³ Major bile acid alterations in cholangiocarcinoma include increased abundance of GCA, reduced levels of TCDCA, LCA, and DCA, and notable shifts in the ratio of glycoconjugated versus tauroconjugated bile acids, as well as the distribution between free and conjugated bile acid pools.³³⁴ Consistent with these trends, multiple conjugated bile acid species have been reported to be elevated specifically in intrahepatic cholangiocarcinoma.³²⁴ Beyond cholangiocarcinoma, studies examining bile acid dysregulation in other primary and metastatic liver cancers remain scarce and warrant further investigation.

D.2. Bile acid metabolism in liver cancers

Bile acids represent a mechanistic bridge between gut microbial activity and host physiology, and their metabolism is broadly dysregulated across liver cancers.

D.2.1. Bile acid synthesis and conjugation in liver cancers

Multiple genes regulating bile acid synthesis and conjugation are altered in hepatocarcinogenesis. In HCC, CYP7A1 mRNA is frequently upregulated, whereas CYP8B1 and key bile acid– conjugating enzymes are down-regulated.³³⁵ Dysregulated bile acid conjugation contributes to impaired immunosurveillance. Through weighted gene co-expression network analysis, SLC27A5 was identified as a tumor-suppressive biomarker whose downregulation promotes HCC progression.³³⁶ Mechanistically, SLC27A5 exerts tumor-suppressive effects by inhibiting thioredoxin reductase 1 expression through the Kelch-like ECH-Associating protein1/ NF-E2-related factor 2 pathways.³³⁷ In addition, inhibiting the bile acid-conjugating enzyme BAAT in hepatocytes was shown to enhance tumor-specific T cell responses and sensitize tumors to anti-programmed cell death protein 1 (PD-1) immunotherapy.³³² These findings suggest that modulating bile acid metabolism or supplementing specific bile acids may improve immunotherapeutic efficacy. Mechanistically, certain primary bile acids, such as TCDCA, induce oxidative stress in intratumoral CD8⁺ T cells, while the secondary bile acid, LCA, triggers ER stress, collectively impairing antitumor activity.³³² Conversely, UDCA mitigates toxic bile acid accumulation, restores CD8⁺T cell effector function, and improves immunotherapy responsiveness.³³²

D.2.2. Bile acid transporters in liver cancers

Dysregulated bile acid transport contributes to altered circulating bile acid profiles and clinical outcomes in HCC.³³¹ The canalicular efflux transporter BSEP (ABCB11), responsible for exporting bile acids into bile, is downregulated in HCC tumors. Supporting this tumor-suppressive role, Abcb11^−/− mice spontaneously develop HCC and cholangiocarcinoma with age.³³⁸ Hepatic uptake mechanisms are also impaired. NTCP expression is reduced in HCC and correlates with poor survival,³³⁹ potentially driven by promoter hypermethylation. Organic anion-transporting polypeptides 1B3, typically restricted to hepatocytes, is aberrantly expressed across heterogeneous liver cancer cell populations and correlates with Wnt/β-catenin activation.^340,341 These changes collectively result in abnormal bile acid accumulation in blood and reduced enterohepatic cycling. In parallel with bile acids uptake, disruption of basolateral export contributes further to bile acid remodeling. MRP4 (ABCC4) is upregulated in HCC and functions as a tumor-promoting efflux pump in the context of p53 haploinsufficiency and Tsc1 loss, effects that can be blocked by mTOR inhibition.³⁴²

D.2.3. Bile acid receptors in liver cancers

The Multiple bile acid–responsive receptors are altered in liver cancers. FXR and its downstream target SHP are downregulated in both human HCC and cholangiocarcinoma tissues, consistent with their antitumor functions.^343–346 In contrast, TGR5 (GPBAR1) is overexpressed in cholangiocarcinoma tumors.^299,347 Functionally, TGR5 promotes anti-apoptotic signaling via serine phosphorylation of the CD95 death receptor in cholangiocarcinoma cell lines, suggesting a potential role in promoting cholangiocarcinoma progression.²⁹⁹ Consistently, excessive bile acids activate TGR5 on cancer-associated fibroblasts, inducing CXCL10-mediated epithelial-mesenchymal transition, metastasis, and an immunosuppressive environment in cholangiocarcinoma through neutrophil recruitment.³⁴⁸

D.3. Mechanisms linking gut dysbiosis and bile acid remodeling to liver cancers

The TME in liver cancers is shaped by chronic inflammation, immune suppression, metabolic reprogramming, and oncogenic signaling—processes that are increasingly recognized as microbiome- and bile acid–regulated.

D.3.1. Inflammation in TME

Gut dysbiosis promotes proinflammatory hepatic milieu that facilitates tumor initiation and progression. FMT from HCC patients induces hepatic inflammation, fibrosis, and dysplasia in mice by disrupting gut barrier integrity and enabling bacterial translocation—particularly Klebsiella pneumoniae, which activates penicillin-binding protein 1B–TLR4 signaling to promote tumorigenesis.³⁴⁹ Dysbiosis also aggravates intrahepatic metastasis through excessive formation of neutrophil extracellular traps, which enhance angiogenesis and tissue necrosis to support metastatic expansion.³¹⁷ In contrast, certain commensals exert antitumor effects. Bifidobacterium pseudolongum–derived acetate activates GPR43 [also known as FFAR2 (Free Fatty Acid Receptor 2)] and suppresses IL-6/JAK/STAT3, reducing MASLD-HCC development.³⁵⁰.

Primary and secondary bile acids differentially regulate hepatic inflammation. The primary bile acid CDCA positively correlates with CXCL16 expression in HCC and cholangiocarcinoma tissues, facilitating recruitment of anti-tumor natural killer T (NKT) cells, while the secondary bile acid glycolithocholic acid is associated with reduced CXCL16 and diminished immune surveillance.³⁵¹ Additionally, the DCA–Senescence-associated secretory phenotype axis in hepatic stellate cells enhances chronic inflammation and promotes hepatocarcinogenesis in dietary and carcinogen models—a process that can be disrupted to prevent tumor formation in high-fat diet settings.^352,353 These findings demonstrate how dysbiotic shifts in gut microbiota and bile acid composition converge to impair intestinal barrier function, amplify hepatic inflammation, and create a tumor-permissive microenvironment.

D.3.2. Immune cells in TME

Gut microbiota profoundly modulate antitumor immune responses by shaping the composition, activation state, and spatial distribution of hepatic immune cells. Various immune populations—including T cells, natural killer (NK) cells, and monocytes/macrophages—play key roles in hepatocarcinogenesis and immunosurveillance. Disruption of commensal microbial communities using broad-spectrum antibiotics induces liver-specific antitumor immunity in HCC models, characterized by expansion of hepatic CXCR6⁺ NKT cells and enhanced IFN-γ production upon stimulation. This response is orchestrated through the bile acid/CXCL16/CXCR6 axis and linked to Clostridium species–derived bile acid metabolism.³⁵¹ A similar mechanism is observed in colorectal cancer liver metastasis, where long-term capsaicin intake alters microbial bile acid metabolism and increases NKT cell recruitment to the hepatic pre-metastatic niche, thereby facilitating metastatic seeding.³⁵⁴

Specific commensal bacteria also influence immunotherapy outcomes. In metabolic–dysfunction– associated HCC, CD8⁺ T cell exhaustion contributes to poor responsiveness to PD-1 blockade.³⁵⁵ Supporting this, bacterial extracts from MASLD-HCC patients promote expansion of IL-10⁺ regulatory T cells and suppress CD8⁺ T cells in peripheral immunity, indicating a dysbiosis-driven immunosuppressive milieu.³¹⁹ In contrast, Enterococcus faecium enhances sorafenib efficacy by increasing CD8⁺ T-cell IFN-γ production and promoting tumor ferroptosis in advanced HCC models.³⁵⁶

D.3.3. Oncogenic and tumor suppressor pathways

Accumulating evidence demonstrates that the gut microbiome directly modulates oncogenic signaling cascades in liver cancers. In HCC, Clostridium mitsuokai disrupts intestinal barrier integrity and translocates to the liver, where its metabolite quinolinic acid activates tyrosine kinase with immunoglobulin and EGF-like domain 2 signaling in hepatocytes, subsequently triggering the PI3K/AKT oncogenic pathway and accelerating tumor progression.³⁵⁷ Beyond canonical genetic mutations, dysregulated interactions between bile acid metabolism and transcriptional programs contribute significantly to hepatocarcinogenesis. Yes-associated protein (YAP) acts as a transcriptional repressor of FXR, impairing bile acid homeostasis and promoting tumor growth through a YAP–TEAD–FXR feed-forward oncogenic regulatory loop.³⁵⁸

D.3.4. Proliferation and cell death

Gut microbiome also influences tumor cell survival mechanisms. Gut dysbiosis promotes cholangiocarcinoma progression by reprogramming glutamine metabolism, which suppresses ferroptosis via the ALK5/NOX1 axis.³⁵⁹ Additionally, conjugated bile acids support extrahepatic cholangiocarcinoma growth. For example, TCA enhances spheroidal and duct-like structure formation in cholangiocarcinoma cells, promoting malignant proliferation.^109,360

Together, microbiome dysbiosis and bile acid remodeling exert multifaceted effects on liver cancer biology—including oncogenic signaling activation, impaired tumor suppression, altered metabolic fitness, and resistance to regulated cell death. While distinct microbial and bile acid signatures are observed across liver cancer subtypes, most current studies rely heavily on fecal profiling, which may not accurately reflect luminal and tumor-associated microbiota. Moreover, the considerable heterogeneity of liver tumors underscores the importance of careful patient stratification in future studies.

V. Clinical Implications

Advances in the pathophysiology of liver disease have profoundly underscored the integral relationship between the gut microbiota, bile acid metabolism, and hepatic health.^15,361 Across major liver diseases—including MASLD, ALD, cholestatic liver diseases, HCC and cholangiocarcinoma-disruption of the gut–bile acid axis is a common pathogenic hallmark.^19,20 These alterations at the molecular and microbial levels present a rich source of biomarkers for clinical diagnosis, prognostic assessment, and therapeutic stratification.

A. Diagnostic potential of the gut–bile acid axis

A.1. Gut microbiome signatures as biomarkers

The intricate relationship between the gut microbiota and the progression of chronic liver disease to cirrhosis, HCC and cholangiocarcinoma has driven the development of non-invasive diagnostic and risk prediction models using machine learning.^316,362 Cirrhosis, in particular, is widely associated with intestinal dysbiosis, characterized by decreased microbial diversity, a depletion of beneficial butyrate-producing bacteria (phylum Firmicutes), and an enrichment of potentially pathogenic bacteria (phylum Proteobacteria).³⁶³ Numerous clinical studies have identified specific microbial genera significantly enriched in cirrhotic patients, with a consistent finding being the enrichment of Veillonella and Megasphaera. The former, as a common oral resident, suggests oral microbial translocation, while the latter is found at higher abundance in individuals with alcoholism and cirrhosis.^364–366

The application of machine learning to microbiome data has yielded models with superior diagnostic accuracy over single-taxon markers. A meta-analysis found that gut microbiome-based ML models were highly accurate for predicting liver cirrhosis (SROC AUC 0.91), with performance for specific etiologies like ALD cirrhosis reaching a diagnostic AUC of 0.97.^362,367 Given that most HCC cases arise from cirrhosis, such models have been extended to cancer detection, demonstrating high potential in distinguishing HCC from healthy controls (AUROC 0.94) and even non-HCC cirrhotic controls (AUROC 0.82).^316,367,368

Beyond broad signatures for cirrhosis, research has identified microbial features directly linked to the specific pathophysiology of different liver diseases. For example, the pathogenesis of some MASLD patients is linked to endogenous ethanol production by the gut microbiota, particularly high-alcohol-producing Klebsiella pneumoniae strains, whose ethanol can induce mitochondrial dysfunction.^160,369,370 Furthermore, the translocation of specific K. pneumoniae strains is increasingly implicated in the progression to HCC; emerging evidence suggests their surface proteins can directly activate oncogenic signaling in hepatocytes via innate immune receptors like TLR4, thereby linking gut dysbiosis directly to hepatocarcinogenesis.³⁷¹

In severe AH, a distinct pathogenic signature was identified, with a profound enrichment of Enterococcus faecalis. Critically, strains carrying the virulence factor cytolysin are strongly associated with disease severity and mortality in hospitalized AH patients, establishing this microbial feature as a powerful prognostic marker for risk stratification in this specific population. ^226,372

Analyzing microbial functional pathways, such as the depletion of butyrate biosynthesis or changes in LPS biosynthesis pathways, provides a more stable and mechanistically relevant strategy than taxonomy alone.^373,374 The biological activity of LPS, a key driver of hepatic inflammation via TLR4, is critically dependent on the chemical structure of its lipid A portion.^364,375 Analysis of LPS biosynthesis pathways, rather than taxonomy, has been shown to better segregate responders from non-responders in cancer immunotherapy, highlighting the importance of functional analysis.^374,376

A.2. Bile acid profiles as biomarkers

Alterations in circulating and fecal bile acid profiles are a direct readout of a dysfunctional gut-liver axis in liver disease, with specific bile acid species and their ratios demonstrating significant diagnostic and prognostic value.³⁷⁷ In MASLD and associated liver fibrosis, elevations in serum conjugated bile acids, particularly GCA and TCA, serve as sensitive indicators of liver injury. Their concentrations increase in a stepwise manner with disease progression (from steatosis to MASH to cirrhosis) and correlate with key histological features such as lobular inflammation (GCA) and hepatocyte ballooning (TCA).^{151,181,378–381} However, the application of bile acid profiles as standalone diagnostic markers remains challenged by their significant variability, which is influenced by circadian rhythms and dietary status, contributing to heterogeneity across studies. The fecal ratio of secondary to primary bile acids (e.g., [DCA+LCA]/[CA+CDCA]) serves as a non-invasive functional index of the gut microbiota’s 7α-dehydroxylation capacity, primarily performed by Clostridiales.^41,138 This ratio is significantly decreased in patients with cirrhosis, particularly advanced disease, and in MASLD with significant fibrosis (F≥2), reflecting impaired microbial metabolic function.^{41,138,180,379,382}

Specific bile acid ratios and modifications serve as more advanced prognostic markers, particularly for HCC risk. Prospective studies have found that elevated pre-diagnostic serum levels of GCA, GCDCA, and specific conjugation ratios (e.g., TCDCA/GCDCA) are associated with an increased future risk of HCC.^{330,383–386} The ratio of 12α-hydroxylated to non-12α-hydroxylated bile acids also provides prognostic information, as this ratio typically decreases with liver disease progression.^179,387

In cholestatic liver diseases, such as PBC and PSC, alterations in bile acid profiles are particularly pronounced and directly reflect their core pathology. Due to obstructed biliary excretion, total serum bile acid levels, especially conjugated bile acids, are dramatically elevated and serve as key indicators for diagnosis and monitoring disease activity.³⁸⁸ Furthermore, the liver activates compensatory detoxification pathways, leading to a disproportionate increase in non-classical sulfated and glucuronidated bile acids.³⁸⁰ The levels of these non-classical bile acids, along with the ratio of specific hydrophobic to hydrophilic bile acids, not only help assess the severity of cholestasis but have also been shown to be powerful prognostic markers for disease progression (e.g., predicting transplant-free survival) .^388,389

A.3. Integrated multi-omics models for enhanced diagnostic accuracy

To overcome the limitations of conventional clinical scores (e.g., FIB-4, APRI) in diagnosing MASH and moderate fibrosis (F≥2), integrated multi-omics models are being developed. These models combine clinical indices with multi-omics data, notably, not only gut microbiome taxonomy and bile acid profiles, but also other key microbial-derived metabolites such as SCFAs and tryptophan derivatives, demonstrating superior diagnostic performance.^{371,390–392}

For example, the ABD-LTyG model, which integrates clinical parameters, a metabolic index, and proteomic biomarkers (Table 5), showed superior efficacy (AUROC 0.807) for diagnosing high-risk MASH (NAS ≥ 4 and F ≥ 2) compared to traditional scores.³⁹³ Similarly, the Malnutrition–Alcohol–Sarcopenia–Ejection Fraction score, incorporating 12 lipids with clinical indicators, demonstrated excellent sensitivity and negative predictive value for identifying “at-risk MASH” (MASH with F ≥ 2) (AUROC 0.789).³⁹⁴ For the specific clinical need of detecting MASH within the F2-F3 fibrosis stages (the target population for emerging therapeutics), the “Floating” machine learning model, which integrates clinical variables and two specific metabolites (3-ureidopropionate and alpha-ketoglutarate), has shown outstanding accuracy (AUROC 0.94).³⁹⁵ These clinically validated multi-omics models significantly enhance diagnostic precision and hold considerable promise for the non-invasive, accurate staging of liver disease.

Table 5.

Summary of biomarkers derived from the gut–bile acid axis

Biomarker Category	Example / Concept	Key Clinical Application / Finding	Evidence Level	Potential Application
Microbiome Signatures	Microbial signatures (e.g., enriched Veillonella, Megasphaera)	Associated with cirrhosis. Machine learning models show high accuracy for cirrhosis diagnosis (AUC 0.91)	Observational cohort / Meta-analysis	Cirrhosis diagnosis / Risk stratification
	Specific pathobionts (e.g., Cytolytic Enterococcus faecalis)	Strongly associated with mortality in severe AH, serving as a prognostic marker	Observational cohort	Prognosis in severe AH
	Microbial functional pathways (e.g., LPS biosynthesis)	Provides mechanistic link to hepatic inflammation via TLR4. May be more stable than taxonomic markers	Mechanistic / Translational study	(Pre-clinical / Early translation)
Bile Acid Profiles	Serum conjugated bile acids (e.g., GCA, TCA)	Increase in a stepwise manner with MASLD progression (NAFL -> MASH -> cirrhosis)	Observational / Cross-sectional cohort	MASLD progression monitoring
	Fecal secondary-to-primary bile acid ratio	A functional index of microbial 7α-dehydroxylation. Ratio is decreased in cirrhosis and MASLD with significant fibrosis (F≥2)	Observational cohort	Index of microbial function / Cirrhosis status
	Specific bile acids Ratios (e.g., TCDCA/GCDCA) & Sulfated bile acids	Associated with increased future HCC risk. Also used for diagnosis and prognosis in cholestatic diseases (PBC/PSC)	Prospective cohort / Case-control	HCC risk prediction; Cholestatic liver diseases diagnosis & Prognosis
Integrated Models	Multi-omics + clinical scores (e.g., MASEF, “Floating” ML model)	Demonstrate superior diagnostic performance over conventional scores (FIB-4) for identifying “at-risk MASH” (MASH with F≥2)	(Internal) Validation cohort	“;At-risk MASH” (F≥2) Diagnosis

Abbreviations: AUC, area under the curve; AH, alcoholic hepatitis; FIB-4, Fibrosis-4 Index; GCA, glycocholic acid; GCDCA, glycochenodeoxycholic acid; HCC, hepatocellular carcinoma; MASH, metabolic dysfunction-associated steatohepatitis; TCA, taurocholic acid; TCDCA, taurochenodeoxycholic acid.

B. Therapeutic strategies targeting the gut–bile acid axis

Given the central role of the gut-bile acid axis in the pathophysiology of liver diseases, interventions targeting this axis have emerged as a highly promising therapeutic frontier.^396,397 These strategies aim to halt or reverse disease progression by remodeling the gut microbiome, modulating host bile acid signaling pathways, or combining both approaches (Table 6).

Table 6.

Overview of therapeutic strategies targeting the gut–bile acid axis

Therapeutic Target / Strategy	Interventio n Example	Mechanism of Action	Potential Beneficiary (Trial Population)	Key Monitoring Endpoints (in Trials)	Observed Mechanism-Related Side Effects (in Trials)
Microbiota Modulation	Probiotics,Prebiotics,Synbiotics	Remodel gut microbiota, enhance barrier integrity, reduce endotoxemia	MASLD; HE; ALD	Liver enzymes, steatosis; HE prevention	Not specified
	Fecal microbiota transplantati on	Systemic reset of the gut ecosystem; restores microbial functions (e.g., bile acids 7α- dehydroxylation)	Recurrent C. difficile (rCDI); SAH; Recurrent HE; MASLD/MASH	1-year survival (SAH); Cognition, dysbiosis (HE); Insulin sensitivity (MASLD)	Not specified
Host Signaling Targets	FXR agonists (e.g., Obeticholic acid, Cilofexor)	Master regulator; suppresses bile acids synthesis, anti- inflammatory, anti-fibrotic	PBC (inadequate UDCA response); MASH; PSC	Fibrosis improvement, MASH resolution (endpoints often not met)	Pruritus, elevated LDL-C
	FGF19 analogs (e.g.,Aldafermin)	Non-tumorigenic FGF19; potently suppresses bile acids synthesis via gut-liver axis	MASH	Liver fat reduction, anti-fibrotic potential	Elevated LDL-C
	ASBT inhibitors (e.g., Maralixibat)	Blocks ileal bile acids reabsorption, reducing the total BA pool and toxic BA load	Cholestatic pruritus (e.g., Alagille syndrome, PBC)	Pruritus scores	Not specified
HCC- Specific Strategies	Microbiome modulation (as adjuvant)	Gut microbiome composition modulates host response to ICIs	(ICIs non-responders)	(Tumor response to ICIs)	Not specified
	FGFR4 inhibitors (e.g., Fisogatinib)	Selectively blocks the oncogenic FGF19/FGFR4 signaling axis, a driver in some HCCs	Molecularly defined HCC subtypes	(Tumor response)	Not specified

Abbreviations: FGF19, fibroblast growth factor 19; FXR, farnesoid X receptor; HE, hepatic encephalopathy; SAH, severe alcoholic hepatitis; ICIs, immune checkpoint inhibitors; MASH, metabolic dysfunction-associated steatohepatitis; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis; UDCA, ursodeoxycholic acid.

B.1. An integrated view of therapeutic targets: common and divergent mechanisms

Therapeutic strategies targeting the gut-bile acid axis in liver disease are based on a nuanced understanding of their underlying pathophysiology. Despite distinct triggers, major chronic liver diseases such as MASLD, ALD, and cholestatic liver diseases surprisingly converge on a set of common pathogenic hubs, including gut barrier dysfunction (“leaky gut”) and a dysregulated gut-bile acid signaling axis.^15,398,399 Leaky gut permits the translocation of microbial products (e.g., LPS), perpetuating hepatic inflammation and fibrogenesis.^18,400 The dysregulated bile acid signaling axis is profoundly bidirectional: in cholestatic liver diseases, failed bile flow remodels the microbiota; conversely, in MASLD/ALD, dysbiosis produces an abnormal secondary bile acid profile that is hepatotoxic and induces a state of functional FXR resistance, further weakening the barrier in a vicious cycle.^{11,15,43,258,401,402} These common hubs provide the rationale for broad, pan-disease strategies such as microbiota modulation and FXR agonists.

However, the distinct primary insults of each liver disease also create unique therapeutic vulnerabilities. MASLD is defined by systemic metabolic overload, with specific strains such as high-alcohol-producing Klebsiella pneumoniae even producing endogenous ethanol, providing a rationale for targeting specific microbial metabolic functions.^160,289 ALD is initiated by the direct toxicity of ethanol, which favors specific pathobionts like cytolysin-producing Enterococcus faecalis, creating a clear rationale for precision therapies such as bacteriophages.^16,226,403 Finally, cholestatic liver diseases (e.g., PBC) are fundamentally driven by cholangiocyte injury and cholestasis (the accumulation of toxic, hydrophobic bile acids), which explains the efficacy of therapies like UDCA and ASBT inhibitors that aim to improve bile flow or reduce the bile acid load ). ^11,404

B.2. Modulating the gut microbiota

Interventions with probiotics, prebiotics, and synbiotics (collectively, PPS) aim to exert therapeutic effects by remodeling the gut microbiota composition and restoring intestinal barrier integrity.^405,406 Mechanistically, they counteract dysbiosis by enhancing the production of beneficial metabolites like SCFAs, improving mucosal barrier function, and reducing endotoxin translocation, thereby mitigating hepatic inflammation.^407,408 Numerous randomized controlled trials (RCTs) and meta-analyses have confirmed that PPS supplementation can significantly improve clinical parameters in patients with MASLD, including hepatic steatosis, liver enzymes, and systemic inflammation markers.^409–412 In patients with ALD (non-HE), clinical studies also suggest they can reduce liver enzymes and inflammatory marker.^413,414 For cholestatic liver diseases (specifically PBC), small studies suggest synbiotics may alter the microbiome and bile acid profiles, but robust data on key clinical endpoints are limited.^{262,281,415,416} Additionally, in the context of hepatic encephalopathy, RCTs have demonstrated that probiotics can effectively prevent its development in patients with cirrhosis.³⁶³ However, it must be noted that most current clinical trials for PPS in MASLD, while promising, are limited by small sample sizes and significant trial heterogeneity.³⁷¹

FMT represents a “whole gut microbiome replacement” strategy designed to systemically reset the gut ecosystem.^417,418 It restores a healthy microbial community, re-establishing colonization resistance and normalizing metabolic functions such as bile acid 7α-dehydroxylation.⁴¹⁹ FMT has achieved remarkable success (approx. 90% cure rate) in treating recurrent C. difficile infection and is FDA-approved for this indication.^420–422 A pilot study in patients with severe alcoholic hepatitis found FMT was associated with markedly improved 1-year survival.⁴²³ For recurrent HE, an RCT demonstrated that FMT improved both cognition and dysbiosis.⁴²⁴ In MASLD/MASH, several pilot RCTs have shown that FMT from lean donors can improve insulin sensitivity.^425,426 In the cholestatic liver diseases space, a pilot study for PSC suggested FMT may be safe, but data remain very limited.⁴²⁷ Despite its promise, the broader application of FMT is hampered by significant challenges related to safety, donor screening, and standardization, and similar to PPS studies, current clinical trials are often constrained by small sample sizes.⁴²⁸

B.3. Disease-specific therapeutic strategies targeting bile acid signaling and metabolism

The therapeutic landscape for liver diseases has undergone a paradigm shift following recent FDA approvals of targeted molecular therapies. While the gut–bile acid axis remains a convergent therapeutic target, the mechanistic rationales and clinical endpoints diverge significantly between metabolic and cholestatic liver diseases.

B.3.1. Therapeutic strategies for MASLD and MASH

In MASH, therapeutic strategies prioritize the resolution of steatohepatitis and attenuation of fibrosis progression through restoration of metabolic homeostasis.

Thyroid Hormone Receptor-β (THR-β) Agonists:

The FDA approval of resmetirom in 2024 marked the first authorized pharmacologic therapy for MASH with moderate-to-advanced fibrosis. By selectively activating hepatic THR-β, resmetirom enhances mitochondrial β-oxidation, improves lipid handling, and reduces hepatocellular lipotoxicity. Emerging evidence further suggests that THR-β agonism modulates bile acid homeostasis, thereby functionally linking systemic lipid metabolism with bile acid signaling⁴²⁹.

FXR Agonists:

The steroidal FXR agonist obeticholic acid demonstrated anti-fibrotic efficacy in the REGENERATE trial; however, its regulatory approval was precluded by an unfavorable benefit-risk profile, including dose-limiting pruritus and low-density lipoprotein-cholesterol elevation.⁴³⁰ Consequently, therapeutic development has shifted toward to non-steroidal FXR agonists (e.g., cilofexor, tropifexor), which aim to preserve anti-fibrotic efficacy while minimizing adverse metabolic effects, particularly within combination therapies.⁴³¹

FGF19 Analogs:

Aldafermin, an engineered FGF19 analog, functions as an endocrine regulator to suppress hepatic bile acid synthesis and mitigate lipotoxic stress. Clinical studies, including the ALPINE-4 trial, demonstrate that aldafermin significantly reduces hepatic fat content and improves non-invasive markers of fibrosis (e.g., Enhanced Liver Fibrosis score) in patients with advanced MASH, supporting that the gut-liver endocrine axis as a druggable target for advanced disease.⁴³²

B.3.2. Therapeutic strategies for cholestatic liver diseases

In cholestatic liver diseases, including PBC and PSC, therapeutic strategies focus on reducing the cytotoxic bile acid burden, enhancing choleresis, and attenuating bile acid-induced inflammatory and immune-mediated injury.

PPAR Agonists (New Standard of Care in PBC):

The management of PBC was transformed by the 2024 FDA approvals of elafibranor, a dual PPARα/δ/agonist, and seladelpar, a selective PPARδ agonist. In patients with inadequate biochemical response or intolerance to UDCA, both agents demonstrated superiority over placebo in achieving alkaline phosphatase normalization and meaningful improvements in cholestatic pruritus.^433,434 Through coordinated regulation of bile acid transport, lipid metabolism, and inflammatory signaling, PPAR agonism exerts combined choleretic and immunomodulatory effects, firmly establishing PPAR-targeted therapy as the preferred second-line pharmacologic strategy in PBC.

Norucholic Acid (NorUDCA) for PSC:

A major therapeutic advance for PSC has emerged with norUDCA, a side-chain–shortened bile acid that induces bicarbonate-rich choleresis via cholehepatic shunting. In a pivotal Phase 3 trial, norUDCA significantly reduced serum alkaline phosphatase compared with placebo.⁴³⁵ Importantly, exploratory analyses suggested attenuation of fibrosis progression in treated patients, providing early evidence that augmentation of protective choleresis may favorably modify disease trajectory. While longer-term outcome data are still required, these findings position norUDCA as the most promising disease-modifying candidate to date for PSC, a disorder historically lacking effective medical therapy.

ASBT/IBAT Inhibitors:

Pharmacologic interruption of enterohepatic bile acid circulation through inhibition of ASBT/IBAT represents a complementary therapeutic strategy. Agents such as maralixibat and odevixibat are now established therapies for pediatric cholestatic pruritus, including Alagille syndrome, and are undergoing active clinical evaluation in adult cholestatic liver diseases, where their primary benefit is anticipated to be symptomatic relief rather than structural disease modification.⁴³⁶

B.4. Therapeutic applications in HCC

Accumulating evidence implicates gut dysbiosis and aberrant bile acid signaling as key contributors to hepatocarcinogenesis, positioning the gut–bile acid axis as a promising therapeutic target for both HCC chemoprevention and treatment.^437,438 One of the most compelling emerging paradigms is the role of the gut microbiome as a determinant of therapeutic responsiveness to ICIs. Multiple studies have demonstrated that gut microbiome composition strongly correlates with clinical response to anti–PD-1therapy.⁴³⁹ Building on these observations, microbiome-modulating strategies—including FMT) from ICI responders to non-responders and the use of defined probiotic consortia such as Akkermansia muciniphila—are currently under clinical investigation as adjuvant approaches to enhance immunotherapy efficacy.⁴⁰⁶

Beyond immunomodulation, direct targeting of bile acid signaling pathways represents an additional avenue for therapeutic intervention in HCC. While the nuclear bile acid receptor FXR exerts tumor-suppressive functions in early hepatocarcinogenesis, its therapeutic exploitation in established HCC has been complicated by concerns that FXR activation may promote tumor growth through induction of the FGF19 signaling cascade.^440–443 As a result, attention has shifted toward targeting downstream oncogenic effectors rather than FXR itself. Among these, the FGF19–FGFR4 signaling axis has been validated as a dominant oncogenic driver in a molecularly defined subset of HCC.⁴⁴⁴ This insight has catalyzed two complementary therapeutic strategies currently under clinical evaluation: (i) the use of engineered, non-tumorigenic FGF19 analogs (e.g., aldafermin) that retain metabolic benefits without activating oncogenic signaling, and (ii) the development of highly selective FGFR4 inhibitors, such as fisogatinib and roblitinib, designed to directly suppress FGF19-driven tumor proliferation.^445–447 Together, these approaches exemplify a precision-oncology framework that leverages bile acid–associated signaling vulnerabilities in HCC.

B.5. Berberine as a modulator of gut microbiome and bile acid metabolism

Berberine, a plant-derived isoquinoline alkaloid, has emerged as a promising microbiota-directed therapeutic with relevance to chronic liver diseases. In contrast to conventional systemically absorbed small-molecule drugs, berberine exhibits low oral bioavailability and exerts the majority of its biological effects within the intestinal lumen, where it potently reshapes gut microbial ecology. Mechanistically, berberine selectively suppresses pathogenic Proteobacteria while enriching beneficial taxa, including short-chain fatty acid–producing bacteria such as Akkermansia muciniphila and Bifidobacterium species. These microbiome shifts enhance intestinal barrier integrity, reduce endotoxemia, and improve host metabolic homeostasis through coordinated modulation of bile acid signaling pathways and downstream lipid and ceramide metabolism.^448–450 In the context of MASLD and MASH, recent randomized controlled trials demonstrate that berberine significantly reduces hepatic fat content, improves insulin sensitivity, and lowers serum transaminases, effects that are accompanied by restoration of beneficial microbial bile acid transformations and normalization of bile acid pool composition.^451,452 These findings support a model in which berberine indirectly regulates host bile acid receptor signaling—particularly along the gut–liver axis—via microbiome-dependent mechanisms rather than direct nuclear receptor agonism. Emerging preclinical studies in cholestatic liver diseases, including PSC, further suggest that berberine confers protection against biliary injury by limiting accumulation of hydrophobic, hepatotoxic bile acids and attenuating peribiliary inflammation through microbiome remodeling.⁴⁵³ While these data remain preliminary, they raise the possibility that berberine may complement existing bile acid–targeted therapies by simultaneously addressing microbial dysbiosis and bile acid toxicity. Collectively, current evidence positions berberine as a dual-acting therapeutic candidate that integrates modulation of the gut microbiome with correction of bile acid dysregulation. However, larger, rigorously controlled clinical trials will be required to define optimal dosing strategies, formulations, patient selection, and long-term safety before its broader clinical implementation.⁴⁴⁸

VI. Challenges and Future Directions

Systematic investigation of the bidirectional interplay between the gut microbiome and bile acid metabolism has established the gut–liver axis as a central determinant of hepatic homeostasis and a critical contributor to the pathogenesis of diverse liver diseases.^15,17 As detailed throughout this review, bile acids function not merely as digestive detergents but as pleiotropic signaling molecules and metabolic integrators that couple microbial activity to host hepatic function. Disruption of this communication network—through microbial dysbiosis, altered bile acid composition, or dysregulated receptor signaling—has been mechanistically linked to a broad spectrum of liver disorders, including MASLD, MASH, ALD, cholestatic diseases, and HCC.^19,20 Despite rapid advances, major conceptual and translational gaps remain. Much of the existing literature is dominated by associative studies, with limited ability to establish causality or delineate disease-stage–specific mechanisms. Heterogeneity in microbiome profiling methodologies, bile acid quantification platforms, host genetic backgrounds, and environmental exposures continues to impede cross-study comparability and reproducibility.^454,455 Moreover, current interventional strategies—ranging from probiotics and bile acid–targeted agents to microbiome modulation— often yield variable or context-dependent outcomes, underscoring an incomplete understanding of patient stratification and mechanism-based target engagement. To clarify the path forward, this section first delineates fundamental conceptual, methodological, and translational challenges that constrain progress in the field, and then outlines emerging opportunities likely to shape the next decade of gut–liver axis research and clinical translation.

A. Fundamental challenges in understanding and targeting the gut-liver axis

A.1. Conceptual hurdles: deciphering system complexity and heterogeneity

The inherent biological complexity of the gut-liver axis presents substantial challenges to a comprehensive understanding. Many investigations reveal associations between alterations in the gut microbiome or bile acid profiles and liver disease states.^149,456 However, determining the direction of causality remains a significant hurdle. Discerning whether observed changes are a cause or a consequence of the disease process is often difficult, a common challenge in microbiome research.⁴⁵⁷ Distinguishing whether observed changes are drivers or consequences of disease progression requires study designs capable of inferring causality, including longitudinal cohort studies, well-controlled interventions, and genetic approaches such as Mendelian randomization.⁴⁵⁸

The system is also marked by substantial heterogeneity, both between individuals and within specific disease categories. Factors such as host genetics, dietary habits, lifestyle, medication use, and existing comorbidities contribute to considerable variation in gut microbiome composition and bile acid metabolism among different people.^459,460 Even within a single disease entity such as MASLD, substantial variability exists across disease stages (simple steatosis, MASH, fibrosis) and clinical subtypes (e.g., lean vs. obese MASLD).^147,430 This heterogeneity complicates identification of consistent biomarkers and underscores the need to move beyond population averages toward functional endotyping—stratifying patients based on mechanistic features rather than phenotypic labels.⁴⁶¹

Limitations of experimental models further constrain interpretation. While animal models remain indispensable, species-specific differences in bile acid composition and conjugation—such as muricholic acids and predominant taurine conjugation in mice—limit direct translation to human disease.^33,462 In addition, reliance on fecal sampling provides only a partial representation of gut–liver axis activity, potentially overlooking critical spatial dynamics within the small intestine or portal circulation.^463,464

A.2. Methodological bottlenecks: measuring quantity, quality, flow, and function

Progress in the field is fundamentally limited by challenges in accurately and reproducibly measuring key components of the gut–liver axis. Lack of methodological standardization across microbiome studies—including differences in sample collection, sequencing platforms, and bioinformatic pipelines—remains a major barrier to reproducibility.⁴⁶⁵ These challenges are particularly acute for bile acid profiling, where variability in sample preparation, analytical platforms (LC-MS versus GC-MS), and reporting standards complicates cross-study comparison. Critically, many studies fail to resolve bile acid isomers and conjugation states (glycine-, taurine-, sulfated-, or glucuronidated forms), despite their profound influence on solubility, receptor affinity, and biological activity.^380,437 As summarized in Table 7, ongoing community efforts (e.g., MIxS and STORMS initiatives) are working to establish consensus guidelines, and broader adoption of such standards is crucial for the field to generate more robust and comparable data.

Table 7:

Recommended Reporting Standards Checklist for Gut-Liver Axis Multi-Omics Studies

Category	Reporting Item	Rationale & Key Standards
Study Design	Pre-registration of protocol and analysis plan (e.g., ClinicalTrials.gov) Clear definition of cohorts, inclusion/exclusion criteria Detailed sample collection & storage (time, temp, stabilizers)	Reduces publication bias and clarifies primary vs. exploratory endpoints. Ensures generalizability and allows for accurate cross-study comparison. Sample integrity is critical; variations in collection are a known source of inconsistency.
Microbiome (16S / Metagenomics)	DNA extraction method and kit details Sequencing platform, target region (16S), sequencing depth Bioinformatics pipeline (QC parameters, ASV/OTU method, database version) Adherence to MIxS (Minimum Information about any (x) Sequence) standards	A major source of known bias; different kits have varying lytic efficiencies for different bacterial cell walls. Methodological variations in sequencing impact taxonomic resolution and sensitivity. Crucial for reproducibility; different pipelines can significantly alter results and hinder comparison. The consensus standard for reporting microbial sequence data.
Metabolomics (Bile Acids)	Sample preparation (e.g., extraction solvents, derivatization) Analytical platform (LC-MS/GC-MS), column, key parameters Quantification method (targeted/untargeted), use of internal standards Clear reporting of specific isomers and conjugates (Glycine, Taurine, Sulfated) QC procedures (blanks, pooled samples, CV% reporting) Adherence to STORMS (Standards for Reporting Metabolomics Studies) initiatives	Impacts recovery rates for different bile acids classes; variations are a critical barrier to consensus. Defines the sensitivity and specificity; platform differences (e.g., LC-MS vs. GC-MS) hinder comparison. Essential for accurate quantification; variations in methods are a key source of inconsistency. Functionally distinct molecules are often grouped; inconsistent reporting obscures mechanistic insights. Monitors instrument stability and analytical variability to ensure data quality. The consensus standard for reporting metabolomics data, ensuring transparency.
Data Analysis & Availability	Method for addressing batch effects Statistical methods, rationale, and multiple-testing correction Analysis code availability Data availability statement (raw and processed data in public repository)	A common source of technical noise and false signals in large multi-omics studies. Ensures statistical rigor and reduces the rate of false discoveries. Ensures full reproducibility of the statistical analysis and data visualization. Ensures full transparency and enables re-analysis or meta-analysis by the community.

Abbreviations: CV, coefficient of variation; GC-MS, gas chromatography-mass spectrometry; LC-MS, liquid chromatography-mass spectrometry; OTUs, operational taxonomic units; QC, quality control; STORMS, Standards for Reporting Metabolomics Studies.

Furthermore, there is an urgent need to shift research focus towards assessing function over composition. Inferring activity solely from metagenomics is often insufficient.⁴⁶⁶ Future research must explore methodologies beyond genomics, such as metatranscriptomics, metaproteomics, and activity-based probes,⁴⁶⁷ to better approximate the in situ enzymatic activity of key pathways (e.g., BSH, Hydroxysteroid dehydrogenase, bai gene cluster) .³⁸² Moreover, techniques such as stable isotope tracing to measure bile acid flux, or dynamic assessments like standardized bile acid challenge tests, represent important conceptual avenues to capture the functional state of the axis, moving beyond static, single time-point fecal analysis. Capturing spatial-temporal dynamics also remains difficult, requiring more sophisticated sampling approaches to access different compartments (small intestine vs. colon, portal vs. systemic blood).⁴⁶⁴

A.3. Translational implementation barriers: The path to clinical utility

Even with improved mechanistic understanding, translating discoveries into clinical practice faces systemic obstacles. A key hurdle is rigorously defining and demonstrating “clinical utility” (proving tangible improvement in patient management) versus “clinical validity” (association with outcome), consistent with FDA–NIH BEST definitions and frameworks established by the Evaluation of Genomic Applications in Practice and Prevention working group and the National Academies of Sciences, Engineering, and Medicine.^{454,468–470} For example, while the novel multi-omics models discussed in section V.A.3 demonstrate high AUROCs, their translational path is challenged by a lack of robust external validation in large, diverse cohorts. A critical research gap remains in demonstrating whether these models offer significant reclassification value (e.g., via net reclassification improvement, or decision curve analysis) over existing non-invasive scores like FIB-4, which is essential to justify their utility.^471,472

Developing therapies targeting the axis also encounters inherent scientific challenges. Balancing efficacy with mechanism-related side effects is a common problem, exemplified by the clinical trial experiences with FXR agonists (pruritus, lipid changes) ^430,473 and ASBT inhibitors (potential vitamin malabsorption, diarrhea).⁴⁷⁴ Similarly, as noted in section V.B.2, even relatively established microbiome interventions like probiotics and FMT face translational hurdles due to limitations in current clinical evidence, such as small sample sizes and trial heterogeneity, complicating the assessment of their true clinical efficacy in liver diseases.^411,475,476 Significant inter-individual variability in response underscores a fundamental lack of understanding of the patient-specific factors (host or microbial) that govern drug response.^50,477 Achieving sufficient tissue or pathway specificity to mitigate off-target effects remains difficult.⁴⁷⁸ Finally, navigating complex regulatory pathways for novel modalities like live biotherapeutic products and FMT represents a significant barrier that separates basic research from eventual application.^479–481 Systemic hurdles like establishing robust manufacturing standards, achieving reimbursement, and integrating new diagnostics into clinical workflows also impede adoption.⁴⁸²

B. Future directions and emerging opportunities

While remarkable progress has illuminated the gut–liver axis, several key frontiers remain open. Future research will need to bridge mechanistic insights with translational tools, refine diagnostics and therapeutics, and ultimately enable personalized interventions. The following sections outline emerging opportunities and directions likely to shape the next decade of research in this field.

B.1. Deepening mechanistic insight: function, integration, and new dimensions

Gaining deeper mechanistic insight requires moving beyond correlation towards integrated, functional understanding. Systems biology approaches integrating multi-omics data (genomics, transcriptomics, proteomics, metabolomics, metagenomics) hold promise for revealing complex interaction networks and key regulatory nodes.^483–485 A critical shift is towards function-centric approaches.⁴⁶⁶ Focusing on the activity of key metabolic pathways (e.g., specific bile acid transformations like BSH, hydroxysteroid dehydrogenase, bai gene cluster activity; SCFA production; endogenous ethanol) is essential.^160,382,486 For example, deeply dissecting the molecular mechanisms by which specific pathobionts (like high-alcohol producing Klebsiella pneumoniae or cytolysin-producing Enterococcus faecalis) drive liver injury, or how key beneficial microbes (like Akkermansia muciniphila or Bifidobacterium pseudolongum) exert hepatoprotective effects, will be crucial.^{160,226,247,350,486} Similarly, elucidating how specific bile acids (such as DCA, LCA, or UDCA and its derivatives) precisely modulate the hepatic immune microenvironment is a key area for future investigation.^332,351 Exploring the “unknown metabolic space” in untargeted metabolomics may also reveal novel bioactive molecules. Understanding must also expand to the wider microbial ecosystem, including the mycobiome, virome (phages), and archaea, and their interplay with bacteria, bile acid metabolism, and host immunity.^487–491 Incorporating chronobiology (circadian rhythms of bile acids/microbiome) and the exposome (impact of diet, drugs, surgery) will provide a more holistic view of the axis regulation.^43,65,489 Building on these mechanistic foundations, the next challenge lies in translating functional signatures into clinically meaningful diagnostic tools.

B.2. Next-generation diagnostics: towards actionable phenotyping

Future diagnostic research should aim for actionable phenotyping that reflects underlying mechanisms.^15,396 The goal is to move beyond simple staging towards identifying functional endotypes (e.g., high endogenous ethanol producers, impaired 7α-dehydroxylation) that may predict therapeutic response.^316,395 This necessitates the development of dynamic and functional readouts reflecting pathway activity or flux (e.g., fecal secondary/primary bile acid ratio, bile acid sulfation indices, challenge tests) rather than just static concentrations.^138,377

B.3. Innovative therapeutic strategies: precision targeting of the axis

Future therapeutics may achieve greater precision by leveraging deeper mechanistic understanding. To overcome the limitations in standardization and targeting efficacy associated with traditional probiotics and FMT, next-generation microbiome therapies are emerging. Precision microbiome engineering could involve rationally designed consortia targeting defined functions⁴⁹², engineered microbes (”living therapeutics”) designed for in situ actions such as bile acid detoxification or delivery of therapeutic molecules (e.g., IL-22),⁴⁹³ or phage therapy specifically targeting pathobionts like cytolysin-producing Enterococcus faecalis.^226,406 Supplementing specific microbial metabolites or components, known as postbiotics, offers another controllable intervention strategy. However, these innovative approaches face significant hurdles including ensuring efficient delivery, in vivo stability, ecological safety, and navigating regulatory pathways.⁴⁹⁴ Novel host target strategies may include gut-restricted/biased bile acid receptor modulators (FXR, TGR5),^{137,495–497} or targeting specific bile acid transporters (ASBT, OSTβ) or metabolic enzymes.⁴⁹⁸ Rationale-driven combination therapies, pairing agents with complementary mechanisms guided by patient endotyping (e.g., FXR agonist + bile acid sequestrant, GLP-1RA + FXR modulator), represent a significant future research direction.^499–502 However, precision therapies will reach their full potential only when embedded within a broader systems framework.

B.4. Expanding the framework: the gut-liver axis in a multi-organ network

The gut-liver axis is part of a larger network critical for systemic homeostasis, and understanding these broader connections is essential.¹⁴ The gut-liver-brain axis is particularly significant due to the high prevalence of neurological and psychiatric comorbidities in chronic liver disease.^503,504 Dysfunction in the gut-liver unit propagates signals via neural, humoral, and immune pathways, impacting brain function. This contributes not only to overt hepatic encephalopathy, but also to associated symptoms such as fatigue, anxiety, and depression,^505–507 Gut dysbiosis,⁵⁰⁸ increased permeability,⁵⁰³ and impaired detoxification ⁵⁰⁹ allow gut-derived neurotoxins (ammonia) ⁵¹⁰ and inflammatory mediators (LPS) ⁵¹¹ into the circulation. These cross a compromised blood-brain barrier,⁵¹² triggering neuroinflammation,⁵¹³ astrocyte swelling,⁵¹⁴ microglial priming,⁵¹⁵ and disruption of neurotransmission (e.g., glutamate and GABA signaling). ⁵¹⁶ Altered bile acid profiles also contribute, with specific BAs crossing the BBB ⁵¹⁷ to modulate neuroinflammation, likely via receptors like TGR5 and S1PR2 expressed on glial cells and neurons.^518,519 Understanding these intricate mechanisms opens avenues for mechanism-based therapeutic strategies beyond traditional ammonia-lowering agents. For instance, targeting gut dysbiosis (e.g., with specific antibiotics like rifaximin or FMT), restoring intestinal barrier integrity, modulating neuroinflammation (e.g., cytokine blockade like anti-TNF-α), or regulating aberrant bile acid signaling within the CNS (e.g., TGR5 agonists or S1PR2 antagonists) represent potential future approaches derived directly from the pathophysiology.^{503,508,511,519,520} Beyond the brain, systemic connections link gut-liver dysfunction to other organs, including hepato-renal syndrome ¹⁵ and cardiovascular disease risk associated with metabolites like TMAO.^{19,43,393,521}

B.5. Towards personalized medicine: from endotypes to individual care

Precision management tailored to individual patients represents a central goal of future research in n hepatometabolic disease.⁵²² Advances in multi-omics diagnostics—integrating genomics, metabolomics, metagenomics, and lipidomics—offer the potential to characterize individual variations in gut microbial function and bile acid metabolism. Such analyses may enable the identification of “endotypes”, or molecular subtypes defined by distinct underlying pathophysiology.⁵²³ For example, multi-omics profiling might identify patient subgroups with unique functional signatures, such as a ‘high endogenous ethanol producer’ endotype (potentially linked to specific Klebsiella enrichment), a ‘severe gut barrier dysfunction’ endotype (marked by high circulating LPS), or a ‘deficient bile acid 7α-dehydroxylation’ endotype (indicated by a low secondary/primary bile acid ratio). These molecularly defined subtypes may help predict disease progression or therapeutic responsiveness to interventions targeting the gut–liver axis.^524,525 As illustrated conceptually in Fig. 12, this approach supports a paradigm shift toward biomarker-guided patient stratification, enabling mechanism-based interventions.^454,455 In this framework, therapeutic strategies could be tailored to each endotype: Patients with a “high ethanol” endotype may benefit from interventions that reduce carbohydrate fermentation or selectively inhibit ethanol-producing bacteria ¹⁶⁰; Those with “barrier dysfunction” may respond better to barrier-protective therapies that restore gut integrity.⁵⁰³ The “low 7α-dehydroxylation” subgroup may benefit from targeted microbial modulation or supplementation with secondary bile acids.¹³⁸ Ultimately, realizing personalized medicine requires integrating these gut-liver axis endotypes with broader host context—genetics, diet, comorbidities, lifestyle, medications—which all influence the axis and disease outcome.^65,149 Developing computational algorithms capable of synthesizing these multi-dimensional data layers will be critical for guiding individualized prevention and treatment.^522,523 However, it is crucial to maintain a grounded perspective. Nearterm research goals likely involve developing better predictive biomarkers for trial enrichment, while the long-term vision of widely accessible personalized pathways requires overcoming significant remaining research and implementation hurdles.

Fig.12. Conceptual Framework for Personalized, Endotype-Driven Therapy.

This radar chart illustrates how functional endotypes (e.g., Endotype A, B, C), defined by multi-dimensional readouts (e.g., bile acid synthesis rate, intestinal permeability, endogenous ethanol production), could guide mechanism-based therapeutic matches. Evidence Level Key: [C] = Clinical (Existing clinical data), [P] = Preclinical (Animal or in vitro data), [H] = Hypothetical (Mechanism-based hypothesis).

In summary, the field is transitioning from descriptive association to predictive, mechanism-guided precision science. Integrating systems biology, computational modeling, and rigorously designed clinical studies will be essential to translate gut–liver axis biology into durable clinical benefit. Achieving this vision will require interdisciplinary collaboration, standardized methodologies, and sustained focus on clinical utility to move from discovery to individualized care (Fig. 13). Collectively, these advances position the gut microbiome–bile acid–liver axis not only as a pathogenic framework, but as a therapeutically tractable system for precision intervention across the spectrum of liver diseases

Fig.13.

This review summarizes how bile acid synthesis, enterohepatic circulation, and microbial BA transformations regulate liver function through receptors including FXR, TGR5, PXR, VDR, S1PR2 and MRGPRX4. Gut dysbiosis, altered BA signaling, and compromised intestinal barriers contribute to MASLD, ALD, chronic liver diseases, and liver cancers. Microbiome and bile acid profiles offer diagnostic and prognostic insights, while interventions targeting the gut–liver axis—such as probiotics, fecal microbiota transplantation, FXR agonists, and FGF19 analogs—represent emerging therapeutic strategies. FXR, farnesoid X receptor; TGR5, Takeda G protein-coupled receptor 5; PXR, pregnane X receptor; VDR, vitamin D receptor; S1PR2, sphingosine-1-phosphate receptor 2; MRGPRX4, mas-related G-protein–coupled receptor X4; MASLD, metabolic dysfunction-associated steatotic liver disease; ALD, Alcohol-associated liver disease; FGF19, fibroblast growth factor 19.

Fig.2. Bile acid synthesis pathways in rodents.

Rodent bile acid synthesis parallels humans but differs in enzyme specificity and bile acid composition. In rodents, CDCA and UDCA are converted to α- and β-MCAs by the rodent-specific enzyme Cyp2c70. This reaction increases bile acid hydrophilicity, yielding a primary pool dominated by CA, α-MCA, and β-MCA. CDCA, chenodeoxycholic acid; UDCA, ursodeoxycholic acid; MCA, muricholic acid.

Significance Statement:

Disruption of the gut microbiome–bile acid–liver axis is now recognized as a unifying mechanism driving multiple liver diseases, including metabolic dysfunction-associated steatotic liver disease, alcohol-associated liver disease, cholestatic liver diseases, and hepatocellular carcinoma. Unraveling the molecular and microbial interactions within this axis offers fundamental insights into disease pathogenesis and reveals novel therapeutic opportunities. Integrating multi-omics technologies with artificial intelligence–based analytics will accelerate the discovery of predictive biomarkers and personalized interventions, advancing the field toward precision-based liver disease treatment protocols.

Abbreviations

ABCB11 (BSEP)

ATP-binding cassette subfamily B member 11 (bile salt export pump)

ABCB4 (MDR3)

ATP-binding cassette subfamily B member 4

ABCC2/3/4 (MRP2/3/4)

ATP-binding cassette subfamily C members 2, 3, and 4

ABCG5/8

ATP-binding cassette subfamily G members 5 and 8

ADH

Alcohol dehydrogenase

AH

Alcoholic hepatitis

AHR

Aryl hydrocarbon receptor

AKT

Protein kinase B

ALDH

Aldehyde dehydrogenase

ALD

Alcohol-associated liver disease

AMPK

AMP-activated protein kinase

ASBT (IBAT; SLC10A2)

Apical sodium-dependent bile acid transporter

ATP

Adenosine triphosphate

AUROC

Area under the receiver operating characteristic curve

BAAT

Bile acid–CoA:amino acid N-acyltransferase

BA-MCY

Bile acid–methylcysteamine

BACS (SLC27A5)

Bile acid–CoA synthetase

bai operon

Bile acid–inducible operon encoding bacterial 7α-dehydroxylation enzymes

BSEP

Bile salt export pump

BSH

Bile salt hydrolase

CA

Cholic acid

CAR (NR1I3)

Constitutive androstane receptor

CDCA

Chenodeoxycholic acid

CYP

Cytochrome P450

CYP7A1

Cholesterol 7α-hydroxylase

CYP8B1

Sterol 12α-hydroxylase

CYP27A1

Sterol 27-hydroxylase

DCA

Deoxycholic acid

FDA

The Food and Drug Administration

ERK

Extracellular signal–regulated kinase

FGF15/19

Fibroblastgrowth factor 15 (mouse) / 19 (human)

FGFR4

Fibroblast growth factor receptor 4

FMT

Fecal microbiota transplantation

FXR (NR1H4)

Farnesoid X receptor

GCA

Glycocholic acid

GCDCA

Glycochenodeoxycholic acid

GI

Gastrointestinal

GLP-1

Glucagon-like peptide-1

GPBAR1 (TGR5)

G protein–coupled bile acid receptor 1

GPCR

G-protein–coupled receptors

GUDCA

Glycoursodeoxycholic acid

HCC

Hepatocellular carcinoma

HSD3B7

3β-hydroxy-Δ⁵-C27-steroid oxidoreductase

IBABP

Ileal bile acid–binding protein

ICIs

Immune checkpoint inhibitors

IFN-γ

Interferon-gamma

IL

Interleukin

JNK

c-Jun N-terminal kinase

LCA

Lithocholic acid

LPS

Lipopolysaccharide

LXR

Liver X receptor

MAPK

Mitogen-activated protein kinase

MASLD

Metabolic dysfunction–associated steatotic liver disease

MASH

Metabolic dysfunction–associated steatohepatitis

MCA

Muricholic acid

MDCA

Murideoxycholic acid

MRGPRX4

Mas-related G-protein–coupled receptor X4

NF-κB

Nuclear factor kappa-light-chain-enhancer of activated B cells

NKT cells

Natural killer T cells

NTCP (SLC10A1)

Sodium taurocholate cotransporting polypeptide

OSTα/β (SLC51A/B)

Organic solute transporter alpha/beta

PBC

Primary biliary cholangitis

PD-1

Programmed cell death protein 1

PI3K

Phosphatidylinositol 3-kinase

PPAR

Peroxisome proliferator–activated receptor

PPS

Probiotics, prebiotics, and synbiotics

PSC

Primary sclerosing cholangitis

PXR (NR1I2)

Pregnane X receptor

RCT

Randomized controlled trial

SAH

Severe alcoholic hepatitis

SCFA

Short-chain fatty acid

SHP (NR0B2)

Small heterodimer partner

S1P

Sphingosine-1-phosphate

S1PR2

Sphingosine-1-phosphate receptor 2

SphK2

Sphingosine kinase 2

TαMCA

Tauro-α-muricholic acid

TβMCA

Tauro-β-muricholic acid

TCA

Taurocholic acid

TCDCA

Taurochenodeoxycholic acid

THR-β

Thyroid hormone receptor-β

TGR5

Takeda G protein–coupled receptor 5

TLR4

Toll-like receptor 4

TMA

Trimethylamine

TMAO

Trimethylamine N-oxide

TME

Tumor microenvironment

TNF-α

Tumor necrosis factor alpha

UDCA

Ursodeoxycholic acid

UGT

UDP-glucuronosyltransferase

VDR (NR1I1)

Vitamin D receptor

Data availability:

This review does not contain any data