Microbiome-immune interactions in autoimmune liver diseases

Abstract

The impact of gut microbiota on immune regulation has received increasing attention in recent years, particularly its role in immune-mediated liver diseases such as autoimmune hepatitis (AIH), primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), and immunoglobulin G4 (IgG4)-related sclerosing cholangitis (IgG4-SC). Disturbances in gut microbiota composition (gut dysbiosis) compromise the intestinal barrier and trigger abnormal immune activity, which foster persistent inflammation, autoimmunity and subsequent liver damage. An increase in Veillonella, Streptococcus, and Lactobacillus were found in most studies for all four diseases. A decrease in Clostridium was detected in AIH and PBC, but an increase in these bacteria is observed in PSC and IgG4-SC. Other common changes in AIH and PBC were an increase in Klebsiella and a decrease in Ruminococcus. In contrast, Ruminococcus is often found to be elevated in PSC. Common features of PBC and PSC are an increase in Enterococcus and a decrease in Faecalibacterium, Enterobacterium, and Coprococcus. Gut microbiome changes in IgG4-SC are most similar to PSC. Probiotics have positive effects in AIH and cholestatic liver diseases in small randomized controlled trial (RCTs). Positive effects of butyrate were in mice models of AIH and PBC. A small study found a positive effect of butyrate on patients with PBC. Indole-3-carboxaldehyde reduced the inflammatory activity of T-cells in patients with AIH in vitro and reduced inflammation in a PSC mouse model. Fecal microbiota transplantation showed beneficial effects in mouse models of AIH and in patients with PSC. Many studies demonstrated the effectiveness of microbiota-targeted therapy on mouse models of autoimmune liver diseases. However, these results remain to be verified in RCTs involving humans.

Key Words: Microbiota; Microbiome; Bacterial translocation; Dysbiosis; Gut-liver axis; Gut-immunity axis

Core Tip: The impact of gut microbiota on immune regulation has received increasing attention in recent years, particularly its role in the onset and progression of immune-mediated liver disorders. Disturbances in the microbial balance compromise the intestinal barrier and trigger abnormal immune activity, which foster persistent inflammation, autoimmunity and subsequent liver damage. A deeper understanding of the gut-liver axis is expected to support the creation of more effective approaches aimed at improving prognosis and enhancing patients’ quality of life.

INTRODUCTION

Autoimmune liver diseases, such as autoimmune hepatitis (AIH), primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), and immunoglobulin G4 (IgG4)-sclerosing cholangitis (IgG4-SC), are conditions, where the immune system mistakenly attacks liver cells or bile ducts. These diseases induce chronic liver damage and can lead to cirrhosis and liver failure[1-3]. Certain aspects of the etiology and pathogenesis of these diseases remain not well-understood; however, genetic predisposition, environmental triggers, and immune dysregulation are believed to be the main contributors to these diseases[2,4-6]. Recent research suggests that the gut microbiota also plays a crucial role in the development and progression of autoimmune conditions by influencing immune responses and inflammation in the liver[7-9]. Dysbiosis, or imbalance in the microbiota composition, has been observed in patients with autoimmune liver conditions, suggesting that alterations in microbial communities may contribute to immune dysregulation and hepatic inflammation[9-11].

Unfortunately, treatment for these diseases is not always effective[1-3]. Therefore, there is a need to identify new targets for therapy. This target could be the gut microbiota. The aim of the review is to provide the most up-to-date information on the relationship between the gut microbiota and immune liver diseases and to describe the results of using microbiota-targeted therapies in the treatment of these patients.

IMMUNOBIOLOGY OF AIH

AIH is characterized by chronic liver inflammation and the production of autoantibodies against hepatocytes[1]. There are two types of AIH: Type 1 (AIH-1) is characterized by positive antinuclear antibodies (ANA) and/or anti-smooth muscle antibodies (SMA); Type 2 (AIH-2) is characterized by positive liver kidney microsomal antibody type 1 (LKM1) and/or by anti-liver cytosol type 1 antibody (anti-LC1)[12,13].

Immunogenetics play an important role in AIH predisposition. The strongest correlation was found between human leukocyte antigen (HLA)-DRB1 1301, 0301, 0401, HLA-DRB3 0101, and HLA-DQB1 0201 alleles and the risk of AIH-1, whereas HLA-DRB1 1302, HLA-DR5, HLA-DQB1 0301, and HLA-DQ3 play a protective role[4,14]. HLA-DRB1 0701, 0301, and 0201 alleles are associated with AIH type II. Non-HLA genes are also associated with AIH (Fas-670a/g, VDR, GATA-2, and others)[4,15].

The immune response in AIH is likely initiated by the presentation of autoantigens to naive T helpers (Th0) via antigen-presenting cells (APC), including dendritic cells (DCs), macrophages, B-lymphocytes, Kupffer cells, liver sinusoidal endothelial cells, hepatic stellate cells, and hepatocytes[12]. Known AIH-related autoantigens include human O-phosphoseryl-tRNA selenium transferase, also known as soluble liver antigen (SLA) in both AIH-1 and AIH-2, and cytochrome P2D6 and formimidoyltransferase-cyclodeaminase (targets of anti-LKM1 and anti-LC1, respectively) in AIH-2[12,16]. Anti-SLA antibodies are the only disease-specific autoantibodies in AIH, and their presence is correlated with a worse prognosis[12].

After antigen stimulation, Th0 transforms into Th1 in the presence of interleukin (IL)-12, into Th2 in the presence of IL-4, or into Th17 in the presence of IL-1β, IL-6, and transforming growth factor-β (TGF-β). Th1 starts to produce IL-2, interferon (IFN)-γ, and tumor necrosis factor (TNF)-α, leading to T-killers’ activation, which causes hepatocellular damage and promotes antigen expression by hepatocytes[12,17]. Th2 produces IL-4, and Th17 produces TNF-α, IL-17, and IL-22 and increases IL-6 secretion by hepatocytes. Recently, the role of Th17 in AIH has been investigated. Levels of IL-17 and Th17 master transcription factor retinoic acid receptor-related orphan receptor C2 are correlated with disease activity, but more data are needed[12]. Th1, Th2, and Th17 infiltrate the liver; moreover, Th liver infiltration is twice as many as T-killers infiltration in AIH patients[14].

Regulatory T cells (Tregs), responsible for immunological self-tolerance and prevention of autoimmune disorders, play an important role in AIH pathogenesis[18]. AIH mouse models can be treated with adoptive Tregs transfer, suggesting that Tregs in AIH are dysfunctional[19]. In AIH patients, the number of Tregs is decreased within the liver and peripheral blood and correlates with disease activity[18,19]. Tregs in AIH have a diminished responsiveness to IL-2, resulting in the impaired production of IL-10 and the impaired ability to suppress IL-17 production by cluster of differentiation (CD) 4⁺ T cells[18,20]. Tregs from patients with active AIH are less capable of CD4⁺ control compared to Tregs from patients in remission and healthy controls. One of the mechanisms is decreased expression of galectin-9, which binds with the Tim-3 receptor of CD4⁺ cells, inducing their death. Moreover, Tim-3 levels are also decreased in AIH patients[12,18].

Tregs from AIH patients are also impaired in their ability to control CD8⁺ cells and monocytes. Tregs in AIH demonstrate defective expression of the ectonucleotidase CD39, which also leads to impaired control of IL-17 production by CD4⁺ cells[12]. These alterations can derive from altered aryl hydrocarbon receptor signaling[21]. In a normal state, this pathway plays a critical role in the regulation of the immune response and metabolism, but it was found to be inhibited in AIH[22].

CD8⁺ T cells promote hepatocytes’ apoptosis through their receptor to the specific class I major histocompatibility complex molecule antigen. In AIH patients, upregulation of CD8⁺ mediators, perforin and granzyme B, is observed: These mediators promote cellular death via apoptosis[23].

Gamma delta T (γδT) cells, a population of unconventional T-lymphocytes, are present in the gut mucosa, liver, and peripheral blood. They can produce various cytokines and chemokines, recruit innate cells, regulate immunoglobulin production by B cells, and present antigens to CD4⁺ and CD8⁺ T cells[24]. γδT cells are increased in peripheral blood and liver of AIH patients. γδT cells in the peripheral blood were shown to produce IFN-γ and granzyme B. In an AIH mouse model, γδT cells were shown to produce anti-inflammatory IL-10 and granzymes and prevent autoimmune disease. Other studies demonstrated a pathogenic effect of IL-17, produced by these cells; therefore, these cells are thought to have both pro- and anti-inflammatory activity[23,24].

Natural killer T (NKT) cells have been found to produce proinflammatory cytokines, such as IFN-γ, IL-4, IL-13, and others, in AIH patients. In concanavalin A (ConA) induced AIH mice, NKT cells play a significant role in immunological and inflammatory responses via IFN-γ, IL-4, TNF-α, and IL-10 production[25].

The role of NKT cells in humans has been less investigated, however, some studies showed an increased expression of NKT cells markers (Vα24) in the liver and peripheral blood of AIH patients[26], an increased number of NKT cells in the liver of AIH patients[27], and a reduced number of circulating NKT cells in peripheral blood, especially during the active phase of AIH[25]. One study showed a prevalence of type II NKT in the liver and peripheral blood of AIH patients and increased production of TNF-α by type II NKT[28].

T follicular helper (or follicular B helper T cells) cells (Tfh), a subtype of CD4⁺ T cells, can induce B-cell activation and differentiation into plasma cells and produce cytokines, including IL-6, IL-10, IL-21, the signal transcription factor B-cell lymphoma 6, and others. These cells can be divided into three subsets (Tfh1, Tfh2, and Tfh17) depending on the cytokines produced[29]. Tfh cells also play role in AIH pathogenesis via elevated IL-21 secretion, which induces the differentiation of memory B cells and plasma cells producing autoantibodies, whereas the secretion of anti-inflammatory cytokines (IL-10 and TGF-β1) by Tfh is inhibited[12,29]. Furthermore, Tfh cells in AIH patients were shown to express surface-programmed cell death protein-1, which inhibits the adaptive immune response[29].

B cells after stimulation with IL-4 and IL-21 mature into plasma cells that produce autoantibodies. B cells can also present antigens and produce cytokines[14,30]. B-cell activating factor is increased in AIH patients and positively correlates with liver inflammation[31]. B cells from AIH patients were shown to frequently express tetraspanin 1, a membrane protein involved in cell proliferation, adhesion, and migration. These B cells produce higher amounts of proinflammatory cytokines (INF-γ and TNF-α) and granzyme B, have superior antigen-presentation ability compared to healthy controls, and can play a significant role in AIH progression[32,33].

Thus, pathogenesis of AIH involves a genetic predisposition and results in hepatocyte damage (Figure 1); however, certain mechanisms remain unclear, and the role of the gut microbiota is an emerging area of research.

Figure 1 Immunopathogenesis of autoimmune hepatitis. Hepatic autoantigens interact with antigen-presenting cells, which lead to the activation of cluster of differentiation (CD) 4⁺ T helper (Th) 0 with differentiation into Th1, Th2, Th17 cells, secreting different cytokines [various interleukin (IL), tumor necrosis factor-α, and interferon-γ]. Activation of cytotoxic CD8⁺ T cells with Th1 cytokines inducing hepatocyte apoptosis. CD4⁺ Th2, natural killer T cells and T follicular helper activate B cells with different cytokines (IL-4, IL-6, IL-21 and etc.) and promote their differentiation into plasma cells, producing autoantibodies. Regulatory T cells are defective in autoimmune hepatitis patients and cannot suppress CD8⁺ and CD4⁺ T cells, but can transform into cytotoxic T cells with further hepatocyte apoptosis. Gamma delta T cells can act as pro- and anti-inflammatory cells. Proinflammatory cytokines induce inflammation, cellular infiltration and liver damage. Tregs: Regulatory T cells; Th: T helper; IL: Interleukin; IFN: Interferon; TNF: Tumor necrosis factor; APC: Antigen-presenting cells; TGF: Transforming growth factor; DC: Dendritic cell; NKT: Natural killer T cell; Tfh: T follicular helper cell; γδT: Gamma delta T cell.

GUT MICROBIOTA AND AIH

The gut microbiota can play an important role in AIH pathogenesis, as dysbiosis has been shown in both experimental mouse models and patients (Figure 2). A marked heterogeneity of results was observed (Table 1)[34-44]. In most studies 16s rRNA sequencing was used and stool collection was prior to therapy (treatment-naive AIH). Patients with AIH had decreased abundance of Bifidobacterium and increased abundances of Veillonella and Streptococcus in the gut microbiota. Interestingly, Lactobacillus abundance was increased in some studies and decreased in others. Decreased Bifidobacteria was associated with treatment resistance among AIH patients[33].

Figure 2 Microbiota-associated pathogenesis of autoimmune hepatitis. Dysbiosis induce decreased tight junction (TJ) proteins expression and increased intestinal permeability, which leads to elevated serum lipopolysaccharide (LPS), infiltration of lamina propria and enhanced cytokines production. Increased LPS, bacteria and inflammatory cells migrate to the liver and initiate inflammation. Dysbiosis also leads to decreased production of short-chain fatty acids, branched-chain amino acids, secondary bile acids, indole, sphingolipids and etc., which play a protective role in the gut and liver. Decreased level of polymeric immunoglobulin receptor leads to reduced immunoglobulin A secretion and dysbiosis, inhibited TJ proteins expression and increased intestinal permeability. secBAs: Secondary bile acids; BCAA: Branched-chain amino acid; LPS: Lipopolysaccharide; SCFA: Short-chain fatty acid; IgA: Immunoglobulin A; TJ: Tight junction; PIGR: Polymeric immunoglobulin receptor; NKT: Natural killer T cell.

Table 1 Gut dysbiosis in experimental models of autoimmune hepatitis and in patients with this disease.

Ref.	Model/patients	Sample	Sequencing method	Decreased gut microbiota taxa	Increased gut microbiota taxa	Correlations
Experimental models of autoimmune hepatitis
Wang et al[34]	MRL+/+ mice treated with 0.5 mg/mL trichloroethene	Feces	16S rRNA	Lactobacillus, Bifidobacterium, Rikenellaceae	Akkermansiaceae, Lachospiraceae	Dysbiosis associated with increased gut permeability
Centa et al[35]	Traf6ΔTEC mice	Feces	16S rRNA	Ruminococcaceae, Clostridiales	Sutterella, Parabacteroides distasonis, Bacteroides acidifaciens, Parabacteroides-like S24-7	Dysbiosis associated with increased Foxp3+ T cells and hepatic inflammation
Wang et al[36]	ConA-induced hepatitis in GSDMD-/- mice	Feces	16S rRNA	Lactobacillus, Roseburia	Allobaculum, Dubosiella	Lactobacillus negatively correlated with AST, hepatic expression of IL-17, serum IFN-γ, TNF-α, IL-17A and positive correlation with tight junction proteins expression; Allobaculum and Dubosiella positively correlated with ALT, AST, hepatic pathological score, serum IFN-γ, TNF-α, IL-17A and negatively correlated with tight junction protein expression
Kang et al[37]	S100/complete Freund’s adjuvant-induced AIH mouse model	Feces	16S rRNA	Ruminococcaceae, Clostridia UCG-014, Rikenella, Dubosiella, Alistipes, Bifidobacterium	Lactobacillaceae, Enterobacteriaceae, Escherichia-Shigella Alloprevotella, Proteobacteria, Bacteroides	Bacteroides vulgatus, Alloprevotella, Lactobacillus murinus, and Prevotellaceae UCG-001 positively correlated with bacterial translocation, inflammatory cytokine expression and negatively correlated with tight junction proteins expression
Lin et al[38]	S100-induced mice model in Pigr-/- mice	Feces	qRT-PCR	Akkermansia	Lactobacillus, Anaeromassilibacillus, Oscillospiraceae	Depletion of Pigr increased intestinal dysbiosis and permeability, but this effect was reduced after antibiotics treatment
Patients with autoimmune hepatitis
Lin et al[39]	Naive AIH, n = 24; HCs, n = 8	Feces	16S rDNA quantitative PCR	Bifidobacterium, Lactobacillus		Association of serum LPS, decreased expression of ZO-1 and occludin with advanced stages of the disease
Abe et al[40]	AIH, n = 17; PBC, n = 39; HCs, n = 15	Saliva	16S rRNA	Streptococcus, Fusobacterium	Veillonella, Neisseria	Veillonella positively correlated with the salivary levels of IL-1β, IL-6, IL-8, IL-12p70 and IgA
		Feces	16S rDNA	Clostridium cluster XIVa	Lactobacillales
Lou et al[41]	Naive AIH, n = 37; HCs, n = 78	Feces	16S rRNA	Pseudobutyrivibrio, Blautia, Lachnospira, Alistipes, Phascolarctobacterium, Erysipelotrichaceae incertae sedis, Ruminococcaceae incertae sedis	Veillonella, Faecalibacterium, Klebsiella, Gemella, Akkermansia, Lactobacillus	Veillonella, Lactobacillus, Megasphaera, Klebsiella, Faecalibacterium correlated with AST, ALT, GGT and total bilirubin; Akkermansia correlated with AST and ALT
Liwinski et al[33]	AIH, n = 72; HCs, n = 95; PBC, n = 99; UC, n = 81	Feces	16S rRNA	Faecalibacterium, Bifidobacterium	Streptococcus, Veillonella, Lactobacillus	Decreased Bifidobacterium associated with an increase of serum IgG; Veillonella abundance associated with ALT
Elsherbiny et al[42]	Naive AIH, n = 15; HCs, n = 10	Feces	16S rRNA	Prevotella, Parabacteroides, Dilaster	Faecalibacterium, Blautia, Streptococcus, Haemophilus, Bacteroides, Veillonella, Eubacterium, Lachnospiraceae, Butyricicoccus
Wei et al[10]	Naive AIH, n = 91; HCs, n = 98	Feces	16S rRNA	Clostridiales, RF39, Ruminococcaceae, Rikenellaceae, Oscillospira, Parabacteroides, Coprococcus	Veillonella, Klebsiella, Streptococcus, Lactobacillus	Veillonella positively correlated with AST and hepatic inflammation
Liang et al[43]	Naive AIH, n = 32; NAFLD, n = 20; HCs, n = 20	Feces	qRT-PCR, 16S rRNA	Bifidobacterium, Lactobacillus, Bacteroides, Clostridium leptum	Escherichia coli
Zhang et al[44]	Naive AIH, n = 68; HCs, n = 15	Feces	16S rRNA	Lactobacillus, Ruminiclostridium	Bacteroides, Prevotellaceae UCG-001

ConA: Concanavalin A; GSDMD: Gasdermin D; Pigr: Polymeric immunoglobulin receptor; AIH: Autoimmune hepatitis; HC: Healthy controls; qRT-PCR: Quantitative real-time polymerase chain reaction; PBC: Primary biliary cholangitis; NAFLD: Non-alcoholic fatty liver disease; UC: Ulcerative colitis; IL: Interleukin; IFN: Interferon; TNF: Tumor necrosis factor; AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; LPS: Lipopolysaccharide; ZO-1: Zonula occludens-1; IgA: Immunoglobulin A; IgG: Immunoglobulin G; GGT: Gamma-glutamyl transpeptidase.

The study with the largest number of patients was conducted by Wei et al[10] and included 91 patients with AIH. 16s rRNA sequencing was used for microbiome analysis. The authors found an increase in Veillonella, Klebsiella, Streptococcus, Lactobacillus, and a decrease in Clostridiales, Ruminococcaceae, and others. Additionally, Veillonella was positively correlated with aspartate aminotransferase (AST) and hepatic inflammation[10].

Abe et al[40] examined the saliva of 17 AIH patients and found increased Veillonella and Neisseria and decreased Streptococcus and Fusobacterium. Moreover, the abundance of Veillonella positively correlated with IL-1β, IL-6, IL-8, IL-12p70, and immunoglobulin A (IgA) in saliva. The gut microbiota of these patients was characterized by increased abundance in Lactobacillales and decreased abundance in Clostridium subcluster XIVa. The relative abundance of Lactobacillales in feces positively correlated with the relative abundance of Veillonella in saliva, whereas salivary Streptococcus negatively correlated with fecal Bifidobacterium.

Zhang et al[45], using the mendelian randomization (MR) method, found a positive correlation between AIH and Actinomycetales and the super pathway of L-tryptophan biosynthesis and L-arginine degradation II (AST pathway) and a negative correlation between AIH and the super pathway of heme biosynthesis from glutamate, the super pathway of polyamine biosynthesis I, and Roseburia unclassified. Fu et al[46], using the weighted median method, provided results that support the aggressive role of Victivallaceae and the protective role of Anaerostipes in AIH pathogenesis[46].

Recently, many studies have focused on the mechanism of microbiota-dependent AIH development and progression. These studies often included interventions targeting the gut microbiota, the results of which are described in Table 2[47-57].

Table 2 Gut microbiota-targeted therapy in the treatment of autoimmune hepatitis and its experimental models.

Ref.	Model/patients	Sequencing method	Intervention	Taxa increased in gut microbiota after the treatment	Taxa decreased in gut microbiota after the treatment	Changes in gut after the treatment	Changes in immune system after the treatment	Changes in liver after the treatment	Other notes
Experimental models of autoimmune hepatitis
Sheng et al[47]	FXR knockout male mice on a western diet	16S rRNA, qPCR	Butyrate supplementation			Increased expression of tight junction proteins and SCFA receptor genes	Reduced hepatic TNF-α, TIMP1, IL-6, ACTA2, CCNA2	Reduced hepatic lymphocyte infiltration and steatosis
Wu et al[48]	S100/complete Freund’s adjuvant-induced AIH mouse model		Sodium butyrate			Decreased disruption of intestinal villi, increased expression of tight junction proteins	Decreased IL-6, TNF-α	Decreased serum AST, ALT, ameliorated lymphocyte infiltration; decreased Escherichia coli translocation to the liver	Decreased blood LPS
Hu et al[49]	S100-induced AIH mouse model		High fiber diet and sodium butyrate			Decreased intestinal lesions, normalization of villus height to crypt depth ratio, increased expression of ZO-1	Decreased Th17, IL-17A, IL-6, increased Treg/Th17 and Foxp3/RORγt ratios, IL-10, TGF-β	Improved liver histology, decreased serum transaminase	Elevated total fecal SCFAs
Zhang et al[50]	S100-induced AIH mouse model	16S rRNA	Bifidobacterium animalis ssp. lactis 420 for 4 weeks	Lactobacillus, Clostridia	Bacteroides, Ruminococcus	Normalized ratio of villus height to crypt depth, increased ZO-1 and occludin expression	Decreased hepatic macrophage/monocyte activation, TNF-α, IL-6, CCL2 expression, decreased hepatic Th17 cells	Decreased serum ALT, AST, attenuated portal inflammation and infiltration	Increased fecal SCFAs, decreased serum LPS
Liang et al[43]	S100/complete Freund’s adjuvant-induced AIH mouse model, treated with broad-spectrum antibiotics	qRT-PCR, 16S rRNA	Fecal microbiota transplantation	Bifidobacterium, Lactobacillus	Escherichia coli		Decreased serum IgG, hepatic IL-21, increased Tfr cells and Tfr/Tfh ratio, hepatic IL-10, TGF-β, Foxp3-mRNA	Decreased serum ALT, AST, improved histological score
Yamaguchi et al[51]	ConA-induced hepatitis mouse model	qRT-PCR, 16S rRNA	6.8% inulin	Akkermansia, Allobaculum		No difference	Decreased IFN-γ and TNF-α expression	Increased hepatic PPARγ and UCP2 expression	Increased fecal SCFAs
Liu et al[52]	S100/complete Freund’s adjuvant-induced AIH mouse model	16S rRNA	6 strains of Bifidobacterium, 8 strains of Lactobacillus, Streptococcus thermophilus St21	Bifidobacterium, Bacteroides, Clostridium, Ruminococcus, Anaerostipes, Blautia	Faecalibacterium, Helicobacter, Staphylococcus	Recovered ileum structure, elevated ZO-1 and occludin expression	Decreased Th1, Th17 and increased Tregs in the spleen; decreased hepatic IL-17A, IFN-γ and increased TGF-β	Alleviated liver injury and infiltration, decreased serum transaminases	Inhibited LPS translocation, increased fecal SCFAs
Ma et al[53]	S100-induced AIH mouse model	RT-qPCR, 16S rRNA	Prednisone + Lactobacillus treatment for 7 days	Bacteroides fragilis, Clostridium, Clostridium leptum, Bifidobacterium, Lactobacillus		Decreased intestinal damage	Reduced serum Tfh cells, IL-6, IL-21, TNF-α, reduced hepatic IL-21, Bcl-6, CXCR5, increased serum TGF-β, KGF-1, KGF-2, IL-10	Attenuated histological injury	Reduced serum endotoxin
Yang et al[54]	ConA-induced hepatitis mouse model		Methyl butyrate				Inhibited CXCL9, CXCL10, CXCL11, decreased hepatic T cells and neutrophils, hepatic IFN-γ, TNF-α and IL-6, decreased serum IFN-γ	Attenuated inflammatory infiltration and necrosis
Kang et al[37]	S100/complete Freund’s adjuvant-induced AIH mouse model	16S rRNA	Lactobacillus acidophilus, Bifidobacterium infantis and prebiotic KGMO for 42 days	Patescibacteria, Deferribacterota, Clostridia UCG-014, Oscillospiraceae, Enterobacteriaceae, Rikenella, Alistipes, Ruminococcaceae	Proteobacteria, Escherichia-Shigella, Akkermansia	Normalized ratio of villus height to crypt depth, increased ZO-1 and occludin	Decreased Tregs in spleen, hepatic IFN-γ, IL-6, IL-17A, IL-1β, reduced NF-κB and NLRP3 inflammasome expression	Alleviated liver inflammation, decreased liver weight, decreased AST, ALT	Decreased serum LPS
Liu et al[55]	ConA-induced hepatitis mouse model	16S rRNA	15 strains of probiotics, complex prebiotics and synbiotic groups treatment for 7 days	Muribaculaceae, Bifidobacterium, Akkermansia, Alloprevotella	Lachnospiraceae NK4A136, Alloprevotella, Escherichia-Shigella	Recovered ileum structure, increased expression of ZO-1 and occludin	Decreased hepatic injury, hepatic F4/80 + macrophage infiltration, decreased Th17 and increased Tregs in spleen, decreased IL-1β, IL-17A, IFN-γ, TNF-α expression	Decreased serum AST, ALT, total BAs	Decreased serum LPS
Ma et al[56]	S100/complete Freund’s adjuvant-induced AIH mouse model	RT-qPCR	Fecal microbiota transplantation	Lactobacillus, Bifidobacterium, Clostridium leptum	Escherichia coli		Elevated Tfh cells, IL-10, TGF-β, decreased IL-21	Attenuated liver injury, decreased serum AST, ALT, total bilirubin	Inhibited activation of TLR/MyD88 signaling
Song et al[57]	S100/complete Freund’s adjuvant-induced AIH mouse model	RT-qPCR	Bifidobacterium, Lactobacillus, Enterococcus capsules	Lactobacillus, Bifidobacterium, Clostridium leptum, Bacteroides fragilis	Escherichia coli		Decreased IgG, IgA, IgM, increased hepatic Tregs and reduced Th17, decreased hepatic IL-33	Attenuated liver injury and infiltration, decreased serum AST, ALT	Increased total SCFAs
Patients with autoimmune hepatitis
Ma et al[53]	Naive AIH, n = 25; naive AIH controls, n = 25	RT-qPCR, 16S rRNA	Prednisone + Lactobacillus for 7 days vs prednisone (RCT)	Bacteroides fragilis, Clostridium, Clostridium leptum, Bifidobacterium, Lactobacillus		Decreased diamine oxidase	Decreased serum SMA, ANA, IgG, IgA, IgM	Decreased serum AST, ALT, total bilirubin	Decreased LPS
Song et al[57]	AIH, n = 10; AIH controls, n = 10	RT-qPCR	Bifidobacterium, Lactobacillus, Enterococcus capsules + conventional treatment	Lactobacillus, Bifidobacterium, Clostridium leptum, Bacteroides fragilis	Escherichia coli		Decreased IgG, IgA, IgM, serum IL-33	Decreased serum AST, ALT	Increased total fecal SCFAs

ConA: Concanavalin A; AIH: Autoimmune hepatitis; FXR: Farnesoid X receptor; qRT-PCR: Quantitative real-time polymerase chain reaction; RT-qPCR: Reverse transcription quantitative polymerase chain reaction; RCT: Randomized controlled trial; ZO-1: Zonula occludens-1; SCFAs: Short-chain fatty acids; IL: Interleukin; TGF: Transforming growth factor; Th: T helper; IFN: Interferon; TNF: Tumor necrosis factor; Tregs: Regulatory T cells; Tfr: T follicular regulatory cell; Tfh: T follicular helper cell; IgG: Immunoglobulin G; KGF: Keratinocyte growth factor; NF-κB: Nuclear factor kappa-B; IgA: Immunoglobulin A; IgM: Immunoglobulin M; SMA: Smooth muscle antibodies; ANA: Antinuclear antibodies; AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; LPS: Lipopolysaccharide; BAs: Bile acids; TLR: Toll like receptor; PPAR: Peroxisome proliferator-activated receptor; KGMO: Konjac glucomannan oligosaccharides.

One study on Traf6ΔTEC mice (mice with medullary thymic epithelial cell depletion) found that these mice generated autoreactive T cells and developed AIH, while germ-free Traf6ΔTEC mice did not develop AIH, suggesting that the microbiota plays an important role in AIH pathogenesis. The authors revealed that the aberrant selection of thymic T cells leads to an impaired gut microbiota, which leads to the development of AIH via Toll like receptor (TLR)-mediated microbial sensing and increased hepatic Foxp3⁺ Tregs and hepatic inflammation[35].

Li et al[58] demonstrated that γδT17 cells, which produce IL-17A, are controlled by gut microbiota. The researchers used germ-free mice (GFM) and antibiotic-treated (ABx) mice and determined that GFM had a significantly decreased number of hepatic γδT17 cells compared to the ABx mice; moreover, providing ABx and GFM with commensal microbes led to normalization of the hepatic γδT17 level. The study also found that the microbiota promoted proliferation of hepatic γδT17 cells[58].

Gut microbes can also induce NKT cell activation and composition[59]. An et al’s experiment on mice showed that, early in life, Bacteroides fragilis produces inhibitory sphingolipids, which modify iNKT cells and protect against iNKT-mediated colitis[60]. Moreover, treatment with Bacteroides fragilis glycosphingolipids (GSL-Bf717) reduced iNKT infiltration and protected against colitis in adulthood. The authors suggested that sphingolipids may be a potential therapy in autoimmune disorders[60].

The gut microbiota was also shown to induce the secretion of autoantibodies, IgA, and cytokines by B cells in different autoimmune models[61].

Another significant pathogenetic mechanism is impaired intestinal permeability, which leads to increased bacterial translocation and inflammation. Different studies have demonstrated impairment of the intestinal barrier in AIH patients and mouse models. Zhang et al[44] found increased intestinal permeability, decreased zonula occludens-1 (ZO-1) and occludin levels, increased serum zonulin, TNF-α, IL-6, and IL-1β levels, and an increased number of resident and infiltrating macrophages in the liver in AIH patients. Research by Lin et al[39] showed a disrupted gut barrier and infiltration of lamina propria with inflammatory cells in AIH patients’ duodenums. The research also revealed reduced amounts of anaerobes and unchanged aerobes, decreased Bifidobacteria/Escherichia coli, and significantly increased lipopolysaccharide (LPS) plasma levels in AIH patients[39].

A study by Manfredo Vieira et al[62] found that tight junction (TJ) protein expression was decreased in AIH patients, whereas LPS levels were increased. They also found a predominance of Enterococcus, especially Enterococcus gallinarum (E. gallinarum), in liver tissues of patients with AIH. Human hepatocytes stimulated with E. gallinarum induced autoimmune-promoting factors (type I IFNs and others)[62]. E. gallinarum was also found to induce IFN-γ + Th17 differentiation in patients with AIH and systemic lupus erythematosus. In a mouse model of lupus, translocation of E. gallinarum led to increased IgG3 anti-RNA autoantibody titers and was correlated with disease activity[63].

Polymeric immunoglobulin receptor (Pigr), a central component of mucosal immunity, was shown to be reduced in an AIH mouse model. The deletion of Pigr led to increased liver damage, reduced IgA levels, destruction of TJ proteins, and increased intestinal permeability. The authors also revealed an increased abundance in the Oscillospiraceae, particularly Anaeromassilibacillus, in Pigr-deficient mice. Moreover, treatment with Lactobacillus rhamnosus GG supernatant was not effective in Pigr-deficient mice, but it reversed liver damage in mice with Pigr[38]. Gasdermin D (GSDMD), a regulator of pyroptosis, also plays a role in AIH. GSDMD-/- mice had a downregulated expression of TJ proteins (ZO-1, occludin, and claudin 4); thus, these mice had increased intestinal permeability, decreased abundance of Lactobacillus and Roseburia, and increased abundance of Allobaculum and Dubosiella[64].

Another pathogenetic mechanism is the production of bacterial metabolites such as short-chain fatty acids (SCFAs), branched-chain amino acids (BCAAs), bile acids (BAs), etc., that can also trigger immune dysregulation[7]. In animal models and patients with AIH, SCFAs were shown to be reduced due to dysbiosis[65]. In an S100-induced mouse model, a high-fiber diet and sodium butyrate attenuated hepatitis, normalized the fecal SCFAs’ level, increased the Treg/Th17 and Foxp3/RORγt ratios, IL-10, and TGF-β, and decreased IL-17A and IL-6. Moreover, normalization of the ratio of the intestinal villus height to crypt depth and improved intestinal permeability were seen[49].

In farnesoid-X-receptor (FXR)-knockout mice, butyrate supplementation led to a reduction in liver lymphocyte infiltration and normalization of BA synthesis[47]. Wu et al[48] found that sodium butyrate ameliorated S100/Freund’s complete adjuvant-induced hepatitis via inhibition of the TLR4 signaling pathway, resulting in improved intestinal permeability and reduced IL-6 and TNF-α levels[48]. Another study on a ConA mouse model treated with methyl butyrate confirmed the protective role of butyrate. In the methyl butyrate group, lower levels of AST, alanine aminotransferase (ALT), and IFN-γ and decreased hepatic inflammatory infiltration and necrosis were observed[54].

In a ConA mouse model pretreated with prebiotic (inulin) for 2 weeks, a lower level of adenosine triphosphate, which limited neutrophil infiltration, was observed. The protective role of inulin may be due to increased SCFA-producing bacteria (Akkermansia, Allobacullum) and increased levels of SCFAs in the gut and portal plasma. It was shown that SCFA can induce the hepatic γ-type peroxisome proliferator-activated receptor, which activates uncoupling protein 2 reducing production of harm reactive oxygen species by liver mitochondria[51]. Therefore, SCFAs acts as a potential therapeutic target in AIH.

The levels of secondary BAs and secondary BA-producing gut bacteria are also decreased in AIH patients[65,66]. In a ConA mouse model, NKT cells were shown to express GPBAR1, also known as TGR5, a receptor that is activated by secondary BAs. Activation of this receptor led to the attenuation of liver inflammation in mice, as GPBAR1 agonism redirects the NKT cells’ polarization toward regulatory NKT10 cells that secrete anti-inflammatory IL-10[67].

Another study on a ConA mouse model found that activation of FXR can ameliorate portal inflammation and hepatocyte apoptosis via the suppression of proinflammatory cytokine (TNF-α, IFN-γ, IL-4, and IL-2) secretion and apoptotic signaling pathways (Fas/FasL, TRAIL and caspase-3)[68]. Therefore, BAs may play a significant role in AIH pathogenesis, and further research is needed.

BCAAs (valine, leucine, and isoleucine) can suppress liver stellate cells and promote Tregs-dependent immune tolerance of via the mammalian target of rapamycin (mTOR) signaling pathway[7]. The bacterial metabolite tryptophan can be broken down to indole, which can activate the aryl hydrocarbon receptor and induce production of IL-10, IL-17, and IL-22 by Th cells[69]. AIH patients were shown to have lower levels of fecal indole-3-carboxaldehyde (ICA) compared to controls. Moreover, fecal ICA level was negatively correlated with serum ALT, AST, and alkaline phosphatase (ALP) activity. Peripheral T cells from AIH patients supplemented with ICA had downregulated expression of early activation markers (CD25, CD69) and inhibited proliferation and production of proinflammatory cytokines. ICA also inhibited phosphatidylinositol 3-kinase/protein kinase B/mTOR signaling in T cells, which restricted their activation. The authors also used a murine ConA model to explore the role of ICA and found attenuated hepatitis, lower serum AST, ALT, and IFN-γ, and inhibited activation of hepatic T cells in an aryl hydrocarbon receptor-dependent manner[70].

It was also shown that vesicles derived from mesenchymal stem cells (MSC-EVs) protected mice liver in a ConA model, and their effect depended on the gut microbiota. Yang et al[71] demonstrated that MSC-EVs ameliorated hepatocyte necrosis and collagen deposition, but this effect was diminished in mice after antibiotic treatment. The authors conclude that MSC-EVs increase the abundance of norank_f_Muribaculaceae and its metabolite acetyl-DL-valine, which have a protective effect[71].

Moreover, probiotics were shown to have a positive effect in an AIH mouse model. The intervention of probiotic Bifidobacterium animalis ssp. lactis 420 alleviated liver inflammation, lowered AST and ALT levels, increased the level of fecal SCFAs, increased the expression of TJ proteins (ZO-1 and occluding) in an S100-induced mouse model[50]. Another study demonstrated the positive effect of combination of Bifidobacterium, Lactobacillus, and Enterococcus on AIH in a mouse model, including reduced liver injury, Treg/Th17 rebalance, suppressed IL-33 expression, and an inhibited TLR2/4 signaling pathway[57].

Ma et al[53] used a murine model of AIH to reveal the effect of Lactobacillus treatment in AIH and found that Lactobacillus enhanced the suppressive role of prednisone via Tfh cells’ regulation, which led to decreased IL-21 and IL-6 and increased IL-10 secretion.

Liu et al[55] revealed that in a ConA mouse model, probiotics, prebiotics, and synbiotics decreased the level of serum transaminases, hepatic macrophage cells, and hepatocellular apoptosis, maintained intestinal integrity, decreased proinflammatory cytokines, and increased anti-inflammatory cytokines. The anti-inflammatory effect in the synbiotic group was superior to that produced by prebiotics and probiotics alone. Another study also confirmed that symbiotic treatment in mouse models of AIH led to alleviated liver inflammation, decreased ALT and AST, and a decreased number of Tregs and inflammatory cytokines (IFN-γ, IL-6, IL-17A, IL-1β). In the synbiotic group, there was a decreased level of LPS and its receptor TLR4 and, therefore, decreased activation of the proinflammatory nuclear factor kappa-B (NF-κB) pathway. Synbiotics improved the diversity and richness of the gut microbiota, restored the intestinal mucosa, and increased ZO-1 and occludin level in gut[37].

Liang et al[43] used an AIH mouse model and performed fecal microbiota transplantation (FMT), which led to decreased serum ALT and AST levels, improved intestinal permeability, and elevated anti-inflammatory levels of IL-10, TGF-β, and Foxp3-messenger RNA[43]. Ma et al[56] also performed FMT to S100/CFA mice model. The authors found attenuated liver injury, decreased serum ALT, AST, total bilirubin, IL-21 levels and increased IL-10, TGF-β and Tfh cells via TLR4/11/MyD88 pathway[56].

All of the above-described interventions targeting the gut microbiota were conducted in experimental models of AIH. However, two studies have already been conducted in patients with AIH who have undergone such complementary therapy (Table 2). 25 AIH patients treated with prednisone and Lactobacillus capsules in randomized controlled trial (RCT) had reduced serum levels of ALT, AST, total bilirubin, SMA, ANA, IgG, IgA, and immunoglobulin M (IgM) when compared with patients who received only prednisolone[53]. A complex probiotic (Bifidobacterium, Lactobacillus, Enterococcus) added to standard AIH therapy in 10 patients produced a similar effect[57].

Thus, gut dysbiosis promotes bacterial translocation, which can trigger immune activation and mislead immune responses toward inflammation. These processes collectively facilitate immune-mediated hepatocyte injury, playing a significant role in AIH pathogenesis. Targeting the gut microbiota offers the potential for novel treatment approaches.

IMMUNOBIOLOGY OF PBC

PBC is an autoimmune liver disease characterized by the destruction of small and medium intrahepatic bile ducts[72]. Antimitochondrial antibodies (AMAs) targeting the E2 subunit of the pyruvate dehydrogenase complex (PDC-E2) of mitochondria are the most common antibodies in PBC and are seen in 90%-95% cases; some patients also have ANA antibodies (anti-sp100 and anti-gp210)[73,74]. Anti-gp210-positive patients have a worse response to ursodeoxycholic acid (UDCA), obeticholic acid, fibrates, glucocorticoids, and immunosuppressants and a worse prognosis[2,75].

Genetic predisposition in PBC includes HLA-associated genes (DQB1 0301, DRB1 08, DRB1 13, and others) and about 44 non-HLA genes[2,5,76-78]. These genetic factors can promote the presentation of autoantigen PDC-E2 to Th cells, which differentiate into effector T cells and attack cholangiocytes[79]. PDC-E2 is overexpressed on the luminal surface of the cholangiocytes of PBC patients and is recognized by AMA, which form antigen-antibody complexes and promote cell damage[2,80]. Cholangiocytes can also play a role of APC, secrete cytokines, and induce CD4⁺ cells[74].

Apart from the immune response, cholangiocytes’ apoptosis in PBC is also induced by the downregulation of chloridion (Cl^-)/bicarbonate ion (HCO₃^-) anion exchanger 2 (AE2), which leads to the loss of cholangiocytes’ protection from hydrophobic BAs[80-82].

BAs play an essential role in PBC pathogenesis and induce cholangiopathy. Their metabolism is also impaired by increased total serum BAs and the predominant accumulation of primary BAs[83]. In PBC, cholestasis causes the activation of TLR and induces immune response and chemokine production, restricting the activation of protective FXR and TGR5[74].

PAMPs (including LPS) in bile lead to cholangiocytes’ TLR activation and cell injury via the NF-κB pathway and chemokine release (IL-8 and CX3CL1). These chemokines also lead to CD4⁺ and CD8⁺ infiltration and cellular damage[80]. In naive PBC patients, significantly increased serum primary BA (cholic acid, glycocholic acid, taurocholic acid, and glycochenodeoxycholic acid) levels were detected, whereas serum secondary BAs (deoxycholic acid, lithocholic acid, and taurolithocholic acid) were decreased. Serum BA levels correlated with liver enzymes and IgM levels[84].

T-cell dysfunction plays a significant role in PBC pathogenesis. CD4⁺ cells (Th cells) are elevated in the portal area; CD8⁺ cells also infiltrate the liver and induce cholangiocyte injury[74]. Early-phase PBC is characterized by the IL-12/Th1 response, whereas the IL-23/Th17 response is seen in later stages[83]. A subset of innate-like T cells, mucosal-associated invariant T (MAIT) cells, known to protect cholangiocyte from microbial triggers, is decreased in patients with PBC; moreover, UDCA treatment does not lead to MAIT normalization[2].

Tregs are decreased in PBC patients[80]. However, Foxp3⁺ Tregs were shown to be elevated near the inflamed portal areas[74]. Studies on mice with deletion of Foxp3⁺ Tregs showed an increased production of AMAs, lymphocytic infiltration in the portal areas, and bile duct damage[85].

B cells in PBC play the role of APCs and produce AMA[2,86]. CD19⁺ B cells are highly increased in the liver and lead to the elevated production of IL-6, IL-10, IFN-γ, and TNF-α; moreover, treatment with UDCA decreases the level of CD19⁺ B cells in peripheral blood. The B-cell activating factor level is also increased in PBC. This protein stimulates B cell development, induces apoptosis of CD4⁺ CD25⁺ Tregs cells and suppresses IL-10 and TGF-β secretion[87].

The pathogenesis of PBC involves a complex interplay of genetic predisposition, environmental triggers, and immune-mediated injury. The gut microbiota is increasingly recognized as a pivotal factor in the development and progression of PBC[5,88].

GUT MICROBIOTA IN PBC

The gut microbiota can play an important role in PBC pathogenesis, as dysbiosis has been shown in both experimental models and patients (Table 3)[89-97]. Despite the pronounced heterogeneity of the results, patients with PBC had decreased abundance of Faecalibacterium, Lachnospiraceae, Ruminococcaceae and increased abundances of Veillonella, Lactobacillus and Streptococcus in the gut microbiota in most studies. Most of the studies used 16S rRNA sequencing for microbiota analysis and most patients were treates with UDCA prior to stool collection.

Table 3 Gut dysbiosis in experimental models of primary biliary cholangitis and in patients with this disease.

Ref.	Model/patients	Sample	Sequencing method	Decreased gut microbiota taxa	Increased gut microbiota taxa	Correlations
Experimental models of primary biliary cholangitis
Ma et al[89]	TLR2-deficient dnTGF-βR II mice	Feces	16S rRNA	S24-7, Ruminococcaceae, Rikenellaceae, Porphyromonadaceae	Lachnospiraceae, Bacteroidaceae	Treatment with antibiotics associated with significantly reduced hepatic mononuclear cells
Wang et al[90]	2-OA-OVA-induced murine cholangitis	Feces	16S rRNA		Clostridiaceae, Lachnospiraceae, Ruminococcaceae	TLR2 (receptor for microbial antigens) activation associated with cholangiocyte apoptosis
Liu et al[91]	DDC-diet mouse model	Liver	16S rRNA		Ligilactobacillus murinus, Lactococcus garvieae	Lactococcus garvieae aggravates DDC-induced cholangitis and was only found in PBC patients but not HCs
Patients with primary biliary cholangitis
Lv et al[92]	PBC, n = 42; HCs, n = 30	Feces	16S rRNA	Acidobacteria, Lachnobacterium, Bacteroides eggerthii, Ruminococcus bromii	γ-Proteobacteria, Enterobacteriaceae, Neisseriaceae, Spirochaetaceae, Veillonella, Streptococcus, Klebsiella, Actinobacillus pleuropneumoniae, Anaeroglobus geminatus, Enterobacter asburiae, Haemophilus parainfluenzae, Megasphaera micronuciformis and Paraprevotella clara	Fusicatenibacter and Roseburia faecis negatively associated with IgG; Megamonas negatively associated with ALT; Enterobacteriaceae positively correlated with ALT and direct bilirubin; Catenibacterium positively correlated with IL-16; Parasutterella secunda positively associated with GGT; Prevotella positively associated with IL-8
Tang et al[11]	PBC, n = 60; HCs, n = 80	Feces	16S rRNA	Sutterella, Oscillospira, Faecalibacterium, Bacteroides	Haemophilus, Veillonella, Clostridium, Lactobacillus, Streptococcus, Pseudomonas, Klebsiella, an unknown genus in the family of Enterobacteriaceae	Faecalibacterium negatively correlated with gp210; Lactobacillus positive; correlated with conjugated bilirubin; Klebsiella positively correlated with GGT, total and conjugated bilirubin
Abe et al[40]	Naive AIH, n = 17; naive PBC, n = 39; HCs, n = 15	Saliva	16S rRNA	Fusobacterium	Veillonella, Eubacterium	Veillonella positively correlated with the levels of salivary IL-1β, IL-8 and IgA
		Feces		Clostridium cluster XIVa	Lactobacillales	Fecal Lactobacillales positively correlated with Veillonella in saliva
Furukawa et al[93]	PBC, n = 76; HCs, n = 23	Feces	16S rRNA	Clostridiales, Lachnospiraceae, Ruminococcaceae, Faecalibacterium	Bifidobacterium, Streptococcus, Lactobacillus, Enterococcus, Prevotella	Faecalibacterium abundance positively correlated with UDCA response
Lammert et al[94]	PBC with non-advanced fibrosis, n = 15; PBC with advanced fibrosis, n = 8	Feces	16S rRNA	Weissella, Ruminococcaceae UCG-005, Ruminococcaceae UCG-004, Ruminococcaceae NK4A214, Christensenellaceae R-7, Lachnospiraceae in advanced fibrosis group	Veillonella, Ruminococcaceae UCG-010 in advanced fibrosis group	Collinsella positively correlated with total SCFAs within non-advanced fibrosis; Lachnospiraceae and Collinsella positively correlated with fecal acetate within non-advanced fibrosis; fecal acetate and SCFAs were higher in advanced fibrosis
Kitahata et al[95]	PBC, n = 34; HCs, n = 21	Small intestinal mucosa	16S rRNA	Leptotrichiaceae, Burkholderiaceae, Comamonadaceae	Sphingomonadaceae, Pseudomonadaceae, Methylobacteriaceae, Moraxellaceae	Sphingomonadaceae associated with chronic nonsuppurative destructive cholangitis
Huang et al[96]	PBC FHRAC-positive, n = 25; PBC FHRAC-negative, n = 18	Feces	16S rRNA	Lachnospiraceae incertae sedis, Anaerostipes, Clostridium XlVb, Ruminococcus, Faecalibacterium, Fusicatenibacter in FHRAC-positive patients	Campylobacter, Lachnospira, Desulfovibrio, Klebsiella, Barnesiella, Lactobacillus, Romboutsia, Turicibacter, Acidaminococcus in FHRAC-positive patients	Lachnospiraceae incertae sedis and Veillonella associated with FHRAC
Zhou et al[8]	Naive PBC, n = 25, HCs, n = 25	Feces	16S rRNA	Faecalibacterium, Ruminococcaceae, Sutterellaceae, Oscillospiraceae, Parasutterella, Clostridia, Coprococcus, Christensenellaceae	Acidimicrobiia, Yersiniaceae, Serratia	Oscillospiraceae negatively correlated with anti-gp210; Serratia and Yersiniaceae positively related to the IgG level
Shi et al[97]	PBC ALBI grade 1, n = 36; PBC ALBI grade 2-3, n = 39	Feces	16S rRNA	Clostridia, Lachnospira in ALBI 2-3 patients	Bacilli, Lactobacillales, Streptococcus in ALBI 2-3 patients

TLR: Toll like receptor; 2-OA-OVA: Ovalbumin conjugated with 2-octynoic acid; dnTGF-βR: Dominant negative form of transforming growth factor β receptor; DDC: 3,5-diethoxycarbonyl-1,4-dihydrocollidine; PBC: Primary biliary cholangitis; AIH: Autoimmune hepatitis; HC: Healthy controls; FHRAC: Patients with a combination of five HLA-DRB1 high-risk alleles; ALBI: Albumin-bilirubin; IgG: Immunoglobulin G; ALT: Alanine aminotransferase; GGT: Gamma-glutamyl transpeptidase; SCFAs: Short-chain fatty acids; IL: Interleukin; UDCA: Ursodeoxycholic acid.

Tang et al[11] revealed a significant reduction in microbial diversity in 60 patients with PBC; decreased abundance of Bacteroidetes and genera including Sutterella, Oscillospira, and Faecalibacterium; and increased abundance of Haemophilus, Veillonella, Clostridium, Lactobacillus, Streptococcus, Pseudomonas, Klebsiella, and an unknown genus in the family of Enterobacteriaceae compared to healthy controls. Moreover, the authors demonstrated normalization of the levels of Haemophilus, Streptococcus, and Pseudomonas and enrichment in Bacteroidetes, Sutterella, and Oscillospira after six months of UDCA treatment. In gp210-positive patients, Faecalibacterium was significantly reduced compared to anti-gp210-negative patients[11].

Systematic review carried out by Wang et al[98] demonstrated a significantly increased abundance of the Veillonella, Lactobacillus, Streptococcus, and Enterococcus and decreased Faecalibacterium, Eubacterium, and Coprococcus in PBC and PSC patients[98]. A recent meta-analysis using MR revealed that Acidaminococcaceae, Bifidobacteriaceae, Clostridiaceae, and Methanobacteriaceae play an obvious harmful role, but Lactobacillaceae and Lachnospiraceae play a protective role in PBC[99].

Abe et al[40] demonstrated increased the Veillonella and Eubacterium levels and decreased the Fusobacterium level in the saliva of treatment-naive PBC patients. The relative abundance of Streptococcus was negatively correlated with IL-1β, IL-4, IL-6, IL-7, IL-8, IL-12p70, IL-17, granulocyte colony-stimulating factor, IFN-γ, and TNF-α, while the relative abundance of Neisseria and Eubacterium/Filifactor was positively correlated with these salivary cytokines. Moreover, the relative abundance of Veillonella and Prevotella/Porphyromonas was positively correlated with IgA in saliva. Fecal microbiota in this study demonstrated increased Lactobacillales and decreased genus Clostridium cluster XIVa, moreover, fecal Bifidobacterium were negatively correlated with salivary Veillonella abundance[40].

Bacteroides, Escherichia, Faecalibacterium, Phascolarctobacterium, Prevotella, Lachnospiraceae incertae sedis, Megamonas, Veillonella, and Megasphaera were increased in patients with HLA alleles that have high risk of PBC[96].

The mucosa-associated microbiota was also shown to be impaired in PBC. Dysbiosis was observed in the ileal microbiota of patients with PBC, along with overgrowth of Sphingomonadaceae and Pseudomonas, and the abundance of Sphingomonadaceae was associated with destructive cholangitis[95].

The albumin-bilirubin (ALBI) score can be used to predict the liver transplant rate and death risk in PBC patients. A higher ALBI score related to worse prognosis. One study found a correlation between the ALBI score and microbiota changes. The ALBI grade 1 (mild disease) patients’ microbiota was richer and more diverse compared to that of ALBI grade 2-3 patients (severe disease). The abundance of Clostrdia and Lachnospira were decreased, whereas Bacilli, Lactobacillales and Streptococcus were expanded in ALBI 2-3 patients in compared with ALBI grade 1 patients. Functional analysis of the gut microbiota revealed a lower expression of the sulfur metabolism pathway in ALBI 2-3 patients in compared with ALBI grade 1 patients[97].

Bidirectional two-sample MR demonstrated the association of Selenomonadales, Bifidobacteriales, and Lachnospiraceae UCG 004 with a higher risk of PBC and a protective effect of Peptostreptococcaceae and Ruminococcaceae[100]. Another MR study found a decreased risk of PBC with the Bacillales, Peptostreptococcaceae, Ruminococcaceae, and Anaerotruncus and an increased risk with the Selenomonadales and Bifidobacteriaceae[101]. Cui et al[102] found a negative correlation between the risk of PBC and Clostridium innocuum group, Butyricicoccus, and Erysipelatoclostridium[102].

Luo and You[103] found that the class Coriobacteriia and the order Coriobacteriales were associated with a higher risk of PBC, whereas Deltaproteobacteria had a protective role. A positive correlation was found for Odoribacter splanchnicus and a negative correlation for Parabacteroides, Gordonibacter, and Gordonibacter pamelaeae[45]. Fu et al[46] also found that Deltaproteobacteria and Desulfovibrionaceae were associated with lower risk of PBC, whereas Coriobacteriia, Coriobacteriaceae, Coriobacteriales, and Ruminiclostridium increased the risk.

The microbiota-associated pathogenesis of PBC involves several interconnected mechanisms through which the gut microbiota influences disease development and progression. To demonstrate the role of the gut microbiota in PBC pathogenesis, the fecal microbiota of patients with PBC was transplanted into pseudo-germ-free mice and PBC mouse models, which led to increased serum ALP, total BA content, and liver injury in pseudo-germ-free mice and an increased serum glycylproline dipeptidyl aminopeptidase level and liver damage in the PBC mouse model. Microbiota transplantation also led to the upregulation of hepatic immune pathways[104].

In PBC mouse model induced with ovalbumin conjugated with 2-octynoic acid, a higher expression of TLR2, TLR7, and TLR9 was seen in the bile duct, and the activation of TLR2 after binding with bacteria products led to cholangiocytes apoptosis[90]. In mice with a diet of 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC), Ligilactobacillus murinus and Lactococcus garvieae were found in the liver. Administration of Lactococcus garvieae to these mice led to a lower expression of occludin and ZO-1 and, therefore, disrupted the intestinal barrier and increased serum BAs[91].

As mentioned above, BAs are critical for PBC pathogenesis, whereas the gut microbiota plays an essential role in BA metabolism. Serum secondary BAs were negatively correlated with PBC-enriched genera (Veillonella, Klebsiella). UDCA treatment of these patients changed BA proportion and increased the taurine-metabolizing bacteria Bilophila[84].

Gut microbiota may share structural similarities with host mitochondrial antigen PDC-E2, which are targeted in PBC; thus, molecular mimicry can lead to the enhanced production of autoantigens. For example, beta-galactosidase of Lactobacillus delbrueckii and PDC-E2 epitope from Novosphingobium aromaticivorans were shown to react with autoantibodies in PBC patients[105]. The Escherichia coli antigen shares a significant homology with human PDC-E2 and may trigger PBC[106]. Pseudomonas also contains proteins that are homologous with PDC-E2 and can cross-react with CD4⁺ and CD8⁺ cells, and Sphingomonadaceae may activate specific NKT cells[95].

Some studies suggested that the microbiota may also induce the production of anti-gp210 via molecular mimicry. For example, Escherichia coli mutY and Salmonella typhimurium mutB genes’ proteins are homologous to the epitope of gp210, and the antibody against the histone-like DNA-binding protein of Streptococcus intermedius can bind to synthetic gp210 peptide; moreover, Streptococcus intermedius can induce the production of anti-gp210 and infiltration in bile ducts of mice. Clinical studies also found that anti-gp210 may be produced 10 years after liver transplantation; thus, the antigen inducing the production might not be derived from the liver[75].

The primary treatment for PBC is UDCA; however, up to 40% of PBC patients have an insufficient response to this treatment for unknown reasons. Various studies have demonstrated the relationship between a poor response and the gut microbiota (Table 4)[107-109]. Han et al[107] found that the microbiota of 11 PBC patients with a poor response to UDCA was characterized by decreased Gemmiger qucibialis, Bariatricus comes, Faecalibacterium prausnitzii, Blautia obeum, CAG-41 sp900066215, and Prevotella sp900557255 and increased Ruminococcus gnavus[107].

Table 4 Association of the gut microbiota composition and function with the effectiveness of therapy for primary biliary cholangitis.

Ref.	Patients	Sequencing method	Treatment	Taxa increased in response group	Taxa decreased in response group	Taxa associated with poor response	Correlations	Other findings
Li et al[111]	PBC response, n = 14; PBC poor response, n = 14	Whole-genome shotgun	Cholestyramine	Lachnospiraceae	Roseburia intestinalis	Klebsiella pneumonia, Prevotella copri	Lachnospiraceae negatively correlated with total bilirubin, positively correlated with secondary BAs	Elevated fecal SCFAs in response group
Martinez-Gili et al[109]	PBC response, n = 212; PBC non response, n = 191	16S rRNA	UDCA	Desulfobacterota, Verrucomicrobiota, Actinobacteriota	Clostridia, Saccharimonadia, Veillonellaceae	Sellimonas, Coriobacteriia	Actinobacteriota positively correlated with a secondary BAs derived from cholic acid; Veillonellaceae correlated with urine levels of CDCA and UDCA	Responders had higher level of fecal secondary BAs and lower level of urinary BAs
Han et al[107]	PBC response, n = 14; PBC poor response, n = 11	Not specified	UDCA	Gemmiger qucibialis, Bariatricus comes, Faecalibacterium prausnitzii, Blautia obeum, and others	Ruminococcus, Clostridium	Ruminococcus gnavus	Gemmiger qucibialis negatively correlated with serum levels of ALT, total bilirubin and total BAs
Wang et al[108]	UDCA-naive patients with PBC, n = 51; HCs, n = 50		UDCA				Decreased fecal butyrate associated with poor response, ALT and total bilirubin	The suppressive function of circulating MDSCs was reduced in non-responders and restored by butyrate
Liu et al[115]	UDCA-naive patients with PBC, n = 132; HCs, n = 131	16S rRNA	UDCA	Bacteroides, Clostridia	Veillonella		Low abundance of fecal Clostridia was associated with poor response	Increased secondary BAs were associated with better response

PBC: Primary biliary cholangitis; HC: Healthy controls; UDCA: Ursodeoxycholic acid; BAs: Bile acids; CDCA: Chenodeoxycholic acid; ALT: Alanine aminotransferase; SCFAs: Short-chain fatty acids; MDSCs: Myeloid-derived suppressor cells.

Wang et al[108] demonstrated decreased butyrate level and defective myeloid-derived suppressor cells (MDSCs) in PBC patients resistant to UDCA. Moreover, butyrate supplementation led to alleviated cholangitis in mice; in patients, butyrate supplementation led to normalization of the suppression function of MDSCs[108]. Martinez-Gili et al[109] included 403 naive PBC patients and found that 191 patients with an insufficient response to UDCA treatment had a lower fecal BA profile, including secondary BAs (deoxycholic acid, lithocholic acid), and a lower abundance of fecal glycine-conjugated BAs. The authors also found a higher abundance of Sellimonas and Coriobacteriia and a lower abundance of Monoglobales, Tissierellales, and the Staphylococcales in the nonresponding group, compared to the responding group. The authors revealed that a lower abundance of bacteria with BA deconjugation capacity may be associated with an incomplete response to UDCA treatment[109].

Yu et al[110] determined that 25 UDCA-responders had less hepatic infiltration of CD4⁺ T cells and T-bet⁺ Th1 cells, whereas 9 nonresponders had more severe CD4⁺ T-cell infiltration after UDCA than before treatment. The CD3⁺ T-cell, CD8⁺ T-cell, Foxp3⁺ Treg, and CD20+ B-cell infiltration were not significantly different.

Li et al[111] demonstrated that 14 PBC patients with remission of cholestasis after cholestyramine treatment had enrichment in SCFA-producing bacteria (Lachnospiraceae) and elevated fecal SCFAs using metagenomics. However, another study by Lammert et al[94] using 16S rRNA sequencing found that the fecal total SCFAs and acetate, as well as the genus Weissella, were higher in the advanced fibrosis group (8 patients); there was also a positive correlation between the total SCFAs and abundance of Collinsella and between the fecal acetate and abundance of Lachnospiraceae and Collinsella[94].

Research on the role of microbiota-targeted therapy in the course of PBC is limited. In a study by Liu et al[112] using a mouse model of cholestasis [bile duct ligation (BDL)], probiotic Lactobacillus rhamnosus GG administration led to an increased expression of fibroblast growth factor (FGF) 15, decreased CYP7α1 and decreased BAs, as well as liver fibrosis degree reduction. In addition, after treatment, increased excretion of BAs in the urine and stool was observed. A decrease in liver fibrosis was also observed in the BDL model after administration of Lactococcus lactis D4, along with a significant increase in Ki-67 expression[113].

In 2024, the results of the randomized controlled clinical trial were published, in which the effect of taking Lactobacillus acidophilus (L. acidophilus) on the course of cholestasis was studied[114,115]. Unfortunately, the study did not identify the specific diseases of the patients. The criteria for cholestasis were an increase in ALP by more than 1.5 times and gamma-glutamyl transpeptidase (GGT) by more than 3 times. A total of 20 people were included in the study (10 people took UDCA and 10 people took UDCA + L. acidophilus). After 14 days of treatment, a significant reduction in markers of cholestasis, AST, and ALT was observed in the probiotic group. In an experiment on mice with cholestasis (BDL model), L. acidophilus attenuated hepatic injury. This effect is likely due to the inhibition of CYP7α1 and activation of FGF15 and Small heterodimer partner, which leads to a decrease in the primary BAs synthesis. Mice also showed elevated levels of bile salt hydrolase in the feces, resulting in increased BA excretion.

Therefore, gut microbiota dysbiosis may contribute to PBC pathogenesis by enhancing intestinal permeability, promoting immune activation via microbial translocation, inducing molecular mimicry against mitochondrial antigens, and altering BA metabolism (Figure 3). These factors collectively break down the immune tolerance and sustain biliary inflammation. Unfortunately, the impact of microbiota-targeted therapy on the course of PBC has been little studied, which opens up new research opportunities.

Figure 3 Gut microbiota in primary biliary cholangitis pathogenesis. A: Bacterial molecular mimicry for E2 subunit of the pyruvate dehydrogenase complex (PDC-E2) can activate cytotoxic T cells, which transfer to the liver and attack cholangiocytes. Dysbiosis leads to the increased intestinal permeability and endotoxinemia, and decreased secondary bile acids (secBAs) production despite the increased intake of primary bile acids into the intestine; B: Antigen-presenting cells present PDC-E2 or molecules similar to it to cluster of differentiation 4⁺ T cells, that transform into T helper (Th) 1, Th2, Th17 cells with production of different cytokines (various interleukin, tumor necrosis factor-α, and interferon-γ) and activate both cytotoxic T cells and B cells. This process leads to increased autoantibodies secretion and cholangiocyte damage; C: Activated cytotoxic T cells induce cholangiocytes apoptosis. Suppressed anion exchanger 2 reduces the protection of cholangiocytes from hydrophobic bile acids in bile. Decreased secBAs lead to insufficient farnesoid X receptor activation and enhanced inflammation, as well as excessive levels of lipopolysaccharide and pathogen-associated molecular patterns activate Toll-like receptors. secBAs: Secondary bile acids; prBAs: Primary bile acids; PDC-E2: E2 subunit of the pyruvate dehydrogenase complex; Tregs: Regulatory T cells; Th: T helper; IL: Interleukin; IFN: Interferon; TNF: Tumor necrosis factor; APC: Antigen-presenting cells; TLR: Toll like receptor; LPS: Lipopolysaccharide; PAMPs: Pathogen-associated molecular patterns; IgA: Immunoglobulin A; FXR: Farnesoid X receptor; AE2: Anion exchanger 2; BA: Bile acid.

IMMUNOBIOLOGY OF PSC

PSC is a chronic immune-mediated liver disease with extra- and intrahepatic bile duct strictures[116]. Inflammatory bowel disease (IBD) is seen in up to 80% of PSC patients, whereas 2%-7.5% of IBD patients have PSC[117].

The pathogenesis of PSC is mostly unclear; however, the main theory is the genetic sensitivity of PSC patients to unknown triggers; more than 20 susceptibility genes were found for PSC[118,119]. Quraishi et al[120] explored mucosal transcriptomic profiles in PSC with IBD (PSC-IBD), ulcerative colitis (UC), and healthy controls and found 1343 differently expressed genes, enrichment in 948 biological processes (824 were upregulated and associated with immune activity) in PSC-IBD compared to controls, and enrichment in 563 biological processes (104 were upregulated and associated with BA metabolism and transport, and fatty acid metabolic processes) in PSC-IBD compared with UC[120].

Positive ANAs, perinuclear anti-neutrophil nuclear antibody, and anti-SMAs are seen in PSC patients[121,122]. In PSC, there is liver infiltration by immune cells, including neutrophils, macrophages, NK, T cells, NKT, and MAIT cells[123]. A study revealed that the bile of some PSC patients contains MAIT cell antigens, leading to MAIT cell activation and further inflammation[124]. Neutrophils are increased in the bile ducts of Mdr2-/- mice (model of PSC) with elevated expression of CCL24 in cholangiocytes and liver macrophages. In patients with PSC, serum CCL24 was also found to be increased, which correlated with liver fibrosis; IL-8, which attracts neutrophils, was also elevated in the bile ducts[123].

Th17 dysregulation is seen in both human and murine studies of PSC. A dysregulated Th17 axis produces proinflammatory cytokines and diminished Tregs function[123]. Monocytes in PSC patients have an increased capacity to produce IL-1β and IL-6 and accumulate around bile ducts[125].

In PSC patients, the downregulation of cholangiocyte TGR5 protein has also been shown, which promotes the secretion of Cl^- and HCO₃^- and hepatic FGF19 overexpression and leads to the accumulation of toxic BAs in the liver[2]. Cholangiocytes in PSC were shown to be activated via TLRs and nucleotide-binding oligomerization domain-like receptors (NODs) and to secrete proinflammatory cytokines such as IL-2, TNF-α, TGF-β1, IL-1β, and IL-6[2]. Inflammation and fibrosis lead to cholestasis in PSC, and disturbances in BA metabolism are seen in patients, which has a direct toxic effect on cholangiocytes[119].

GUT MICROBIOTA IN PSC

Gut dysbiosis has been detected in patients with PSC; as with AIH and PBC, there is marked heterogeneity in the results of the studies (Table 5)[126-141]. Most studies used 16S rRNA sequencing for microbiota analysis and include patients on different medications (UDCA, sulfasalazine, prednisone and etc.).

Table 5 Changes in the microbiota in primary sclerosing cholangitis patients.

Ref.	Patients	Sample	Sequencing method	IBD	Decreased gut microbiota taxa	Increased gut microbiota taxa	Correlations
Sabino et al[126]	PSC, n = 13; PSC + IBD, n = 39; HCs, n = 52	Feces	16S rRNA	-	Anaerostipes	Enterococcus, Fusobacterium, Lactobacillus, Morganella, Streptococcus, Veillonella	Enterococcus and Lactobacillus positively correlated with GGT; Enterococcus positively correlated with serum ALP
				+		Enterococcus, Fusobacterium, Lactobacillus, Streptococcus, Veillonella
Torres et al[127]	PSC + IBD, n = 19; PSC, n = 1; IBD, n = 15; HCs, n = 9	Ileocolonic biopsies	16S rRNA	+/-	Clostridiales	Barnesiellaceae, Blautia
Kummen et al[128]	PSC, n = 85 (PSC + IBD, n = 55); UC, n = 36; HCs, n = 348	Feces	16S rRNA	+/-		Veillonella, Akkermansia, Clostridium	Veillonella associated with PSC
Bajer et al[129]	PSC, n = 43 (PSC + IBD, n = 32); UC, n = 32; HCs, n = 31	Feces	16S rRNA	-	Faecalibacterium prausnitzi, Coprococcus catus, Ruminococcus gnavus, Adlercreutzia equolifaciens, Prevotella copri	Rothia, Enterococcus, Streptococcus, Clostridium, Veillonella, Haemophilus, Lactobacillus, Staphylococcus, Escherichia	The abundance of Prevotella copri was negatively associated with PSC
				+	Faecalibacterium prausnitzi, Prevotella copri, Coprococcus catus, Ruminococcus gnavus, Phascolarctobacterium	Coprobacillus, Escherichia, Corynebacterium, Rothia, Enterococcus, Streptococcus, Clostridium, Veillonella, Haemophilus, Lactobacillus salivarius	Phascolarctobacterium negatively associated with the IBD; Coprobacillus, Escherichia, Corynebacterium, Lactobacillus associated with PSC-IBD
Pereira et al[130]	PSC, n = 80; HCs, n = 46	Bile	16S rRNA	-		Prevotella, Streptococcus, Veillonella, Fusobacterium, Haemophilus, Enterobacteriaceae	Streptococcus abundance positively correlated with an increase in disease severity
Torres et al[131]	PSC + IBD, n = 15; IBD, n = 15	Feces	16S rRNA	+	Dorea, Veillonella, Lachnospira, Blautia, Roseburia	Ruminococcus, Fusobacterium
Iwasawa et al[132]	Pediatric PSC + IBD, n = 24: UC, n = 16; HCs, n = 24	Saliva	16S rRNA	+	Pasteurellaceae, Rothia, Haemophilus	Lachnospiraceae, Bacteroidetes, Oribacterium
Rühlemann et al[117]	PSC, n = 137 (PSC + UC, n = 75); UC, n = 118; HCs, n = 133	Feces	16S rRNA		Holdemanella, Desulfovibrio, Faecalibacterium, Clostridium IV, Coprococcus	Veillonella, Streptococcus, Lactobacillus, Enterococcus, Bacteroides, Parabacteroides	Patients with PSC-IBD had decreased Bilophila
Rühlemann et al[133]	PSC, n = 65 (PSC + IBD, n = 32); UC, n = 38; HCs, n = 66	Feces	PCR, spacer 2		Saccharomyces	Candida, Humicola
Lemoinne et al[134]	PSC, n = 22; PSC + IBD, n = 27; IBD, n = 33; HCs, n = 30	Feces	16S rRNA	-	Saccharomyces	Exophiala, Sordariomycetes, Ruminococcaceae, Rikenellaceae, Filobasidium	PSC associated with increases in Exophiala and Sordariomycetes
				+		Lachnospiraceae, Veillonellaceae, Enterococcaceae, Fusarium, Enterobacteriaceae, Sordariomycetes, Monographella	No strong associations were found
Visseren et al[135]	PSC, n = 97 (rPSC, n = 14)	Colonic biopsies	16S rRNA	-	Gammaproteobacteria, Shigella in rPSC patients	Deinococcus Thermus, Fusobacteraceae, Listeriaceae, Desulfosarcina, Fastidiosipila in rPSC patients	Shigella was significantly higher in patients without rPSC
				+	Bacteroidetes, Rhodobacteraceae	Moraxellaceae, Acinetobacter
Quraishi et al[120]	PSC + IBD, n = 20; UC, n = 10; HCs, n = 10	Colonic biopsies	16S rRNA	+	Lachnospiraceae, Lentisphaerae, Gammaproteobacteria, Enterobacteriacea, Prevotellacae, Paraprevotellacae, Myxococcales	Bacilli, Pseudomonas, Streptoccocus, Haemophilus parainfluenzae, Staphylococcus, Parvimonas, Bacteroides fragilis
Cortez et al[136]	Pediatric PSC, n = 11; PSC + UC, n = 7; UC, n = 12; HCs, n = 23	Feces	16S rRNA	-		Veillonella, Megasphaera, Streptococcus	A higher abundance of Ruminoclostridium 5 and Ruminococcaceae UCG 002 associated with remission/controlled disease; a higher abundance of Veillonella associated with active disease
				UC	Lactobacillus	Prevotella, Veillonella
Denoth et al[137]	PSC + IBD, n = 7; IBD, n = 42; HCs, n = 28	Ileocolonic biopsy	16S rRNA	+	Bacteroides	Roseburia, Haemophilus, Fusobacterium, Bifidobacterium, Actinobacillus, Brachyspira
Lapidot et al[138]	PSC, n = 17; PSC + IBD, n = 18; HCs, n = 30	Feces	16S rRNA	-	Bacteroides thetaiotaomicron, Faecalibacterium prausnitzii, Bilophila, Clostridium, Ruminococcus, Lachnospiraceae	Clostridium XlVa, Clostridium symbiosum, Clostridium perfringens, Streptococcus salivarius, Veillonella dispar, Ruminococcus gnavus, Bacteroides fragilis, Enterococcus, Lactobacillus, Blautia, Butyricicoccus, Megasphaera	Salivary and fecal Veillonella, Scardovia and Streptococcus associated with PSC
				+	Bacteroides thetaiotaomicron, Faecalibacterium prausnitzii	Clostridium XlVa, Clostridium symbiosum, Clostridium perfringens, Streptococcus salivarius, Veillonella dispar, Ruminococcus gnavus, Bacteroides fragilis, Enterococcus, Lactobacillus, Blautia
		Saliva		+/-		Streptococcus salivarius, Prevotella histicola, Rothia mucilaginosa, Veillonella parvula, Actinomyces, Campylobacter concisus, Bifidobacterium stellenboschense, Phocaeicola
Hole et al[139]	PSC, n = 84; PSC-LT, n = 51; HCs, n = 40	Ileocolonic biopsies	16S rRNA	-	Faecalibacterium, Odoribacter, Lachnospiraceae, Bilophila, Akkermansia	Haemophilus, Veillonella, Roseburia, Streptococcus, Hungatella, Klebsiella	Klebsiella associated with reduced LT-free survival; reduced Akkermansia associated with concomitant IBD
Leibovitzh et al[140]	PSC + IBD, n = 54; IBD, n = 62	Feces	Not specified	+	Blautia obeum	Veillonella atypica, Veillonella dispar, Clostridium scindens	Decreased serum sulphated bile acids and increased serum conjugated secondary bile acids were associated with PSC-IBD and increased liver fibrosis
Del Chierico et al[141]	Pediatric PSC + UC, n = 26; UC, n = 27; HCs, n = 26	Feces	16S rRNA	+/-	Akkermansia, Bacteroides, Dialister, Parabacteroides, Oscillospira, Meyerozyma, Malassezia	Streptococcus, Saccharomyces, Sporobolomyces, Tilletiopsis, Debaryomyces	The abundance of Streptococcus associated with PSC-UC

IBD: Inflammatory bowel disease; PSC: Primary sclerosing cholangitis; UC: Ulcerative colitis; LT: Liver transplantation; rPSC: Recurrent primary sclerosing cholangitis after liver transplantation; GGT: Gamma-glutamyl transpeptidase; ALP: Alkaline phosphatase.

In PSC patients, bacterial diversity is decreased[117]. Ostadmohammadi et al[142] showed that Bacteroidetes was the most abundant phylum in PSC-IBD patients (but not in IBD patients), whereas Enterococcus, Lactobacillus and Bifidobacterium were decreased, and Firmicutes/Bacteroidetes ratio was lower in PSC-IBD patients as compared to IBD only patients[142]. The levels of SCFA-producing bacteria (Ruminococcus, Eubacterium, Ruminiclostridium, Faecalibacterium prausnitzii, Lachnoclostridium, Coprococcus, Phascolarctobacterium, Blautia, Desulfovibrio, Adlercreutzia, Succinivibrio) were decreased in PSC and PSC-IBD patients[143,144].

Enterococcus faecalis and its virulence factors, cytolysin and gelatinase, were recently demonstrated to be increased in PSC patients; moreover, high fecal cytolysin was associated with lower overall survival[145]. In another study, higher abundance of Enterococcus was associated with increased levels of serum ALP and disease severity in PSC patients[126].

MR analysis revealed that increased Eubacterium rectale was associated with an elevated risk of PSC[146]. Another MR analysis revealed three microbial taxa correlated with PSC: Ruminococcaceae was shown to increase the risk of PSC, whereas Betaproteobacteria and Ruminiclostridium had a protective role. The authors also found an increased abundance of the Dialister, Veillonella, Melainabacteria, and Gastranaerophilales in PSC patients[147]. Another MR analysis found a correlation between Selenomonadales, Rhodospirillaceae, and RuminococcaceaeUCG013 and PSC risk and a negative correlation with Actinomycetales, Actinomycetaceae, Actinomyces, Alloprevotella, Barnesiella, and Peptococcus[101].

There is currently no effective treatment for PSC. Liver transplantation is a treatment for end-stage PSC; however, up to 25% of patients experience recurrence of PSC (rPSC) after this procedure. A pilot study found that patients with rPSC had lower abundance in the Gammaproteobacteria and Shigella pretransplant compared with patients without recurrence[135].

Patients with PSC after liver transplantation (PSC-LT) were shown to have a similar reduction in alpha diversity, increased Haemophilus, Veillonella, and Roseburia, and decreased SCFAs producing bacteria (Faecalibacterium and Odoribacter) as patients with PSC without transplantation. Recurrent PSC-LT patients had a lower abundance of the Lachnospiraceae ND3007 group, Lachnospiraceae CAG-56, and Bilophila and increased Streptococcus and Hungatella compared to patients with PSC-LT without recurrence. High sigmoid dysbiosis index, a loge [(sum relative abundance of increased PSC genera)/(sum relative abundance of decreased PSC genera)], and the presence of Klebsiella in the ascending colon were associated with reduced mean transplantation-free survival[139].

In the mucosa-associated microbiota of patients with PSC, enrichment in Barnesiellaceae and Blautia was observed compared to controls and patients with PSC-IBD[127]. Another study also found differences in the mucosa microbiota of PSC patients, with higher Roseburia, Haemophilus, Fusobacterium, Bifidobacterium, and Actinobacillus and lower levels of Bacteroides[137].

The salivary microbiome of patients with PSC was also significantly altered, with decreased bacterial diversity and elevated Streptococcus salivarius, Prevotella histicola, Rothia mucilaginosa, Veillonella parvula, Actinomyces, Campylobacter concisus, Bifidobacterium stellenboschense, and Phocaeicola[138]. In a pediatric PSC cohort, the salivary microbiota had decreased Rothia and Haemophilus compared to healthy controls and decreased Haemophilus and increased Oribacterium compared to UC pediatric patients[132].

The bile microbiome was also shown to be altered in PSC patients. Liwinski et al[148] used samples from the oral cavity, duodenal fluid, mucosa, and ductal bile of patients with PSC. The bile microbiome diversity was different from other samples’ microbiomes and healthy controls microbiomes, with reduced diversity and increased Enterococcus faecalis, Veillonella dispar, Staphylococcus, and Neisseria. Enterococcus was also associated with an increase in the noxious and potentially carcinogenic taurolithocholic acid[148]. Pereira et al[130] also found biliary dysbiosis in PSC patients with increased Prevotella, Streptococcus, Veillonella, Fusobacterium, and Haemophilus; moreover, Streptococcus abundance was positively correlated with disease progression.

The fungal composition of the microbiota is also damaged in PSС. Lemoinne et al[134] demonstrated that patients with PSC had a relative increase in fungal biodiversity, an increased proportion of Exophiala, and decreased Saccharomyces cerevisiae[134]. Another study revealed increased Candida and Humicola (Humicola grisea, recently reclassified as Trichocladium griseum) in PSC patients’ stool[133].

Animal studies helped to reveal the pathogenesis of PSC and its close connection with gut dysbiosis. Nakamoto et al[149] used gnotobiotic mice and inoculated fecal samples from patients with PSC-associated UC (PSC/UC), patients with UC, and healthy controls and examined T-cell profiles in the livers and colons of mice. Mice with healthy microbiota showed induced Th17 differentiation in the colon, whereas PSC/UC mice showed Th17 differentiation in the liver and upregulated inflammatory genes in the colon and liver. PSC/UC mice did not have histological and serological liver damage, but a DDC diet resulted in an elevation in total bilirubin and ALP in them, whereas a germ-free condition led to mild hepatobiliary injury, and mice with microbiota from healthy donors showed liver inflammation, though it was milder than that of PSC/UC mice. The authors also revealed three bacteria (Klebsiella pneumoniae, Proteus mirabilis, and E. gallinarum) that translocated from the gut to mesenteric lymph nodes; moreover, these bacteria were enriched in the gut in PSC/UC mice and were shown to promote liver inflammatory response via Th17 induction. Klebsiella pneumonia was found to disrupt the intestinal barrier via the formation of epithelial pores[149].

Germ-free Mdr2 knockout mice (Mdr2-/-; a PSC model) had significantly higher ALP, AST, and bilirubin, absent secondary BAs, and more severe fibrosis and ductular reaction compared to conventionally housed Mdr2-/- mice, which means that microbiota also had protective effects against biliary injury[150]. Liao et al[151] demonstrated that Mdr2-/- mice showed pronounced NLRP3 inflammasome activation via increased bacterial translocation, whereas suppression of NLRP3 activation led to ameliorated liver injury. Moreover, microbiota transferred from Mdr2-/- mice to wildtype control mice led to liver injury in the recipient mice.

Tedesco et al[152] showed increased serum levels of IL-17 and an increased number of hepatic γδT cells [IL-17A⁺ Vγ6Jγ1 γδT cell receptor (TCR)⁺ cells] and Lactobacillus gasseri in the livers of Mdr2-/- mice. Isolated γδT cells were shown to produce IL-17 in response to Lactobacillus gasseri and induce liver infiltration. Moreover, anti-γδ TCRs reduced serum IL-17, infiltration of the liver and hepatic fibrosis[152].

In a previous case report, a 66-year-old woman with PSC and treatment-resistant UC underwent subtotal colectomy and proctectomy. Interestingly, her PSC achieved full remission after the proctectomy, with normalization of the ALP, GGT, and AST levels and clinical and histological remission, which indicated the role of gut pathogens, including in the rectum, in PSC pathogenesis in human[153].

Bacterial outer membrane vesicles (OMVs) were shown to promote liver inflammation and fibrosis in a mouse model and in PSC-IBD patients. In ductal organoids, OMVs activated TLR4 signaling and NLRP3-gasdermin-D[154]. Fusobacterium were found to be increased in PSC patients: One study found that Fusobacterium nucleatum (F. nucleatum) can produce OMVs, which stimulate secretion of proinflammatory IL-8 and TNF-α via colonic epithelial cells. In mice harboring a human microbiota pretreated with antibiotics, F. nucleatum led to the disruption of colonic architecture and increased proinflammatory cytokines; moreover, mice without antibiotic pretreatment did not develop inflammation, suggesting that an intact microbiome is protective against F. nucleatum[155].

A higher abundance of microbial genes involved in BA metabolism, a decreased stool BA pool, and increased serum conjugated BAs were found in IBD-PSC patients. Decreased serum sulphated BAs and increased serum conjugated secondary BAs were associated with liver fibrosis progression in the IBD-PSC group[140].

Schneider et al[156] demonstrated the role of BA changes in PSC pathogenesis. Mdr2-/- mice treated with antibiotics had reduced alpha diversity and bacterial abundance and higher concentrations of the fecal FXR-antagonistic primary conjugated BAs (tauro-α-muricholic acid and tauro-β-muricholic acid), and no FXR-activating BAs were detected in their stool. The mice were shown to have reduced ileal FGF15, higher hepatic CYP7A1 expression, increased hepatic concentrations of BAs, and liver injury. Patients with PSC demonstrated a significant decrease in the marker of BA synthesis, 7α-hydroxy-4-cholesten-3-one, and its levels and higher ratios of primary to secondary BAs were negatively associated with mortality. The authors concluded that PSC-related liver injury is associated with microbiota depletion and decreased FXR signaling[156].

One study revealed the elevated expression of GPBAR1 (TGR5), the receptor for secondary BAs, in macrophages, NK cells, sinusoidal cells, and cholangiocytes in liver samples of patients with PSC. An in vitro experiment demonstrated the anti-inflammatory effect of the selective GPBAR1 agonist BAR501 on the human cholangiocytes cell line. Treatment of Abcb4-/- mice with BAR501 reduced biliary inflammation, liver fibrosis, and gut permeability[157].

A recent study showed that patients with PSC had lower fecal unconjugated deoxycholic acid compared with controls; moreover, the fecal deoxycholic acid levels negatively correlated with the total bilirubin levels. Alcohol intake was associated with higher fecal deoxycholic acid and lower serum total bilirubin levels; Blautia and Lachnoclostridium were positively correlated with fecal deoxycholic acid and negatively correlated with Streptococcus[158]. Another study did not find differences in the fecal BA profile of patients with PSC-IBD, IBD alone, and controls, although this study only included seven patients with PSC, eight patients with IBD, and eight controls[159].

Microbiota-targeted therapy was shown to have a promising effect in ameliorating PSC. In Mdr2-/- mice, a germ-free condition and vancomycin treatment led to increased hepatobiliary inflammation and reduced SCFAs and Lachnospiraceae; administration of SCFAs and colon colonization with Lachnospiraceae resulted in reduced fibrosis and inflammation[160]. Administration of Pediococcus pentosaceus Li05 to a DDC model animals led to the attenuation of liver damage, inflammation, and fibrosis via the activation of the FXR signaling pathways and to improved gut microbiota profile with reduced endotoxin translocation[161].

Another bacterial metabolite of Lactobacillus reuteri, ICA, was shown to promote mucosal immune homeostasis in a DDC murine model of PSC, to prevent liver fibrosis, and to prevent bacterial translocation by activating the aryl hydrocarbon receptor-IL-22 axis[162].

Clostridium scindens was shown to protect against vancomycin-induced BAs accumulation and liver fibrosis in C57BL/6J mice known for their robust immune response. These effects were associated with the elevated production of secondary BAs by Clostridium scindens and the activation of intestinal FXR-FGF15/19 signaling[163]. Prevotella copri (P. copri) was found to be decreased in a PSC mouse model. Moreover, supplementation with P. copri can lead to a significant decrease in ALP, ALT, AST, and total bilirubin levels, as well as fibrotic factors such as collagen 1a1 and TIMP1. Further, the total BA level was improved via FXR-related signaling pathways. Thus, the authors revealed that P. copri can improve cholestasis and fibrosis[164].

Probiotic therapy has not been studied for patients with PSC. In 2008, the results of a pilot randomized controlled clinical trial involving 14 patients with PSC and IBD were published, which studied the effectiveness of probiotic therapy using four Lactobacillus and two Bifidobacillus strains. Taking the prebiotic for 3 months did not lead to a significant reduction in itching, weakness, and frequency of bowel movements, or a reduction in bilirubin, ALP, GGT, AST, or ALT[165].

The literature also describes one clinical case of treatment of a 13-year-old patient with PSC and IBD who received a combination of prednisolone (30 mg/day), salazosulfapyridine (3000 mg/day), and a probiotic (Lactobacillus casei Shirota, 3 g/day) for 14 days, which led to a significant reduction in ALP and GGT. Subsequently, prednisone was discontinued, but the patient continued to take UDCA, salazosulfapyridine and probiotics for 3 years of follow-up. After 30 months, a follow-up biopsy showed a reduction in inflammatory cell infiltration and periductal fibrosis and a decrease in fibrotic areas from 10.5% to 3.6% in each specimen[166].

Phase 2 of the randomized, double blind, placebo controlled, parallel study evaluating the safety and efficacy of LB P8 in patients with PSC (ClinicalTrials.gov, No. NCT06699121) has been underway since November 2025. The study plans to evaluate the safety and effect of the probiotic in a group of 75 people. It is planned to investigate the effectiveness of low and high doses and placebo control. The study is scheduled to be completed by 2029.

Oral vancomycin was shown to have a potential therapeutic role for colitis in PSC-IBD in very small trials[167,168]. One study demonstrated a positive effect of oral vancomycin on ALT and ALP levels and colitis in PSC-IBD patients. Increased Proteobacteria, Fusobacteria, Verrucomicrobia, Veillonella, and Klebsiella and decreased Faecalibacterium prausnitzii, Anaerostipes hadrus, Bifidobacterium longum, and multiple SCFA producing species belonging to Clostridium, Ruminococcus, Lachnospira, and Roseburia were observed during oral vancomycin treatment compared with the baseline. The authors also revealed that the treatment led to the loss of specific fecal SCFAs and secondary BAs[169].

A clinical case of a patient with PSC and recurrent acute bacterial cholangitis demonstrated the positive effect of FMT on PSC. Before the FMT, the patient had high abundance of Proteobacteria, Enterobacter, Catenibacterium, and Dialister. After the FMT, an increase in the abundance of Bacteroides, Megamonas, and Bifidobacterium was observed, lower levels of AST, ALT, and ALP were seen; however, after 1 year, cholangitis recurred[170]. In one study, FMT in PSC-IBD patients led to decreased ALP levels in 30% of patients and improved microbial diversity in 100% of patients[171]. A randomized multicenter study (FARGO) is ongoing[172]. The results of microbiota-targeted therapy for PSC are summarized in Table 6[161-164,169-171].

Table 6 Gut microbiota-targeted therapy in the treatment of primary sclerosing cholangitis and its experimental models.

Ref.	Model/patients	Sequencing method	Intervention	Taxa increased in gut microbiota after the treatment	Taxa decreased in gut microbiota after the treatment	Changes in gut after the treatment	Changes in immune system after the treatment	Changes in liver after the treatment	Other notes
Experimental models of primary sclerosing cholangitis
D’Onofrio et al[162]	DDC-diet induced mouse model	16S rRNA	Indole-3-carboxaldehyde-loaded enteric microparticles	Lactobacillus reuteri	Escherichia coli, Klebsiella pneumonia	Increased ZO-1 expression and improved intestinal permeability	Reduced serum IL-6, IL-17A, INF-γ, TNF-α, decreased hepatic CD11b+, F4/80+ cells	Decreased serum ALT, total bilirubin
Jiang et al[164]	DDC-diet induced mouse model	16S rRNA	Prevotella copri treatment	Prevotella copri, Muribaculaceae, Lactobacillus, Turicibacter, Bifidobacterium, Clostridium sensu stricto, Odoribacter, Alistipes	Staphylococcus, Akkermansia			Decreased serum ALT, AST, ALP, total bilirubin, reduced collagen-1a and TIMP1 hepatic expression	Activated FXR signalling pathway
Han et al[161]	DDC-diet induced mouse model	16S rRNA	Pediococcus pentosaceus Li05 treatment	Bacteroidaceae, Ruminococcaceae, Sutterellaceae, Rhodocyclaceae, Bacteroides, Ruminococcus, Pediococcus, Tuzzerella	Erysipelotrichaceae, Eggerthellaceae, Anaeromyxobacteraceae, Comamonadaceae, Faecalibaculums, Turicibacter, Enterorhabdus, Lactobacillus	Restored expression of ZO-1, occludin, claudin1, and MUC2	Decreased hepatic expression of IL-1β, IL-6, TNF-α, decreased serum IL-1β, IL-2, IL-6, INF-γ	Decreased serum ALT, AST, ALP, total and direct bilirubin, total BAs; reduced hepatic necrosis and infiltration	Increased FXR and FGF15 ileal expression, increased fecal BAs and SCFAs
Xiao et al[163]	C57BL/6J mice	16S rRNA	Oral vancomycin for 3 weeks + Clostridium scindens for 3 days					Decreased serum ALT, AST, decreased hepatic collagen synthesis and deposition, ACTA2, Col1a1 and TGF-β expression	Decreased primary/secondary bile acid ratio
Patients with primary sclerosing cholangitis
Philips et al[170]	38 years old male patient with PSC	16S rRNA	FMT	Bacteroides, Megamonas, Bifidobacterium	Proteobacteria, Clostridium, Veillonella, Fecalibacterium, Oscillopsira, Lachnospira			Decreased serum ALT, AST, ALP, GGT, total bilirubin	Decreased total BAs
Allegretti et al[171]	PSC + IBD, n = 10	16S rRNA	FMT	Odoribacter, Alistipes, Erysipelotrichaceae incertae sedis, Clostridia	Negativicutes, Verrucomicrobiae			Decreased ALP	BAs profile did not change
Quraishi et al[169]	PSC-IBD, n = 15	16S rRNA	Oral vancomycin for 4 weeks	Fusobacteria, Verrucomicrobia, Fusobacterium nucleatum, Enterobacter hormaechei, Escherichia coli, Veillonella, Klebsiella, Akkermansia muciniphila, Roseburia hominis, Prevotella buccae, Lachnospiraceae bacterium, Blautia hansenii, Bacteroides thetaiotaomicron	Faecalibacterium prausnitzii, Anaerostipes hadrus, Bifidobacterium longum, Clostridium, Ruminococcus, Lachnospira, Roseburia	Reduction in fecal calprotectin and mucosal inflammation		Decreased serum ALT, ALP, total bilirubin	Decreased secondary BAs, increased primary BAs, reduced fecal SCFAs

DDC: 3,5-diethoxycarbonyl-1,4-dihydrocollidine; IBD: Inflammatory bowel disease; PSC: Primary sclerosing cholangitis; FMT: Fecal microbiota transplantation; ZO-1: Zonula occludens-1; IL: Interleukin; TGF: Transforming growth factor; IFN: Interferon; TNF: Tumor necrosis factor; CD: Cluster of differentiation; AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; GGT: Gamma-glutamyl transpeptidase; ALP: Alkaline phosphatase; FXR: Farnesoid X receptor; FGF: Fibroblast growth factor; SCFAs: Short-chain fatty acids; BAs: Bile acids.

Gut microbiota dysbiosis contribute to chronic immune activation in PSC through the gut-liver axis. Therapeutic modulation of the microbiome holds promise but requires further investigation.

IMMUNOBIOLOGY OF IGG4-SC

IgG4-SC is a chronic fibro-inflammatory disease of the biliary system, which can be part of IgG4-related disease (IgG4-RD) or occur in an isolated form[6,173,174]. Moreover, IgG4-SC can mimic different diseases; therefore, diagnosis is usually complicated[175]. Serum IgG4 levels are increased; transmural lymphoplasmacytic infiltration and storiform fibrosis in the intra- and extrahepatic bile ducts are observed[3,176]. Infiltration of abundant IgG4-positive cells is a characteristic finding[177].

Genetic predisposition plays a role in the development of IgG4-SC. HLA serotypes DRB1 0405 and DQB1 0401 are more common in the Japanese population, and DRB1 0301 and DQB1 0201 are more common in the English population[178].

IgG4-SC was first described in 1991, and certain mechanisms of pathogenesis and etiology remain unclear[179]. The leading theory for the pathogenesis of IgG4-SC posits an aberrant interaction between different immune cells. Activation of NOD-2 and TLR on monocytes and basophils activates the B-cell-activation factor and enhanced IgG4 response in IgG4-autoimmune pancreatitis. Activation of TLR3 and TLR4 leads to immune-mediated cholangitis and pancreatitis in animal models[178]. Some patients have ANA and anti-annexin A11 IgG4 antibodies[178]. Th2 lymphocyte response dominates in IgG4-SC patients, with the release of IL-4, IL-5, and IL-13. Memory Tregs are increased in IgG4-SC patients with high concentrations of Foxp3⁺ CD4⁺ CD25⁺ Tregs, which produce IL-10, in the bile duct tissue[178,180]. Okazaki et al[181] developed a pathogenic theory of IgG4-SC, including a decrease in naïve Tregs and induction of Th response with the release of cytokines such as IL-4 and IL-10, which stimulate B cells to produce IgG4, whereas TGF-β from memory Tregs lead to fibrogenesis[181]. Patients with active IgG4-SC/autoimmune pancreatitis show enhanced surface secretion of programmed cell death protein 1 on CXCR5⁺ Tfh cells (Tfh1, Tfh2, and Tfh17 cells), with increased absolute numbers of programmed cell death protein 1⁺ Tfh2 cells. These cells are essential for B-cell selection and maturation into plasmoblasts. Circulating programmed cell death protein 1⁺ Tfh cells also correlated with IL-4 and IL-21 production, with serum IgG4 and immunoglobulin E (IgE) levels, and with disease activity[182]. Cytotoxic CD4⁺ T cells were also found to be increased and correlated with IgG4 levels in IgG4-RD[183].

Chemokines also play an important role in IgG4-RD, with increased CCR8 expression in Th2 and Foxp3 Tregs. CCR8-positive cells are found around bile ducts and peribiliary glands. CCL1 is also expressed in the ductal and glandular epithelia[178]. Some IgG4-RD patients were shown to have hypocomplementemia (low serum C3 and/or C4), which correlated with higher peripheral eosinophils, inflammatory markers, and IgG4 and IgE levels[184].

Stimulation of B cells with IL-4 and IL-10 led to increased IgG4 production[178,180]. Patients with IgG4-RD showed elevated plasmoblasts and B regulatory cells. Patients with IgG4-SC had three activated immunological cascades associated with B cells and immunoglobulins: Fc-gamma receptor-mediated phagocytosis (phagocytosis of different antigens, coated with antibodies), the B-cell receptor signaling pathway (a cascade that mediates B-cell activation, survival, proliferation, and differentiation), and the Fc-epsilon receptor I signaling pathway (primarily involved in allergic responses)[185]. Cargill et al[186] demonstrated that patients with IgG4-related pancreatobiliary disease had decreased complement receptor 2 expression on IgG1- and IgG4-producing B cells. Normally, the activation of this receptor leads to the inhibition of B-cell proliferation and antibody production. The authors also found a decreased Fc receptor for IgG (FcγRIIb) expression on IgG1-producing B cells, the activation of which led to suppressed antibody secretion. Another alteration was an increased Fc receptor for IgE (FcεRII) expression on IgG4-producing B cells, which activate mast cell sensitization and degranulation. Thus, in IgG4-RD patients, B cells are different from healthy controls, with elevated sensitivity to complement and IgE[186].

IgG4-SC is driven by an abnormal immune response involving Th2 and Treg cell activation, infiltration of IgG4-positive plasma cells, and cytokine-mediated fibrosis. The role of gut microbiota in this disease is not as well characterized as in other hepatobiliary diseases; however, gut dysbiosis was also demonstrated.

GUT MICROBIOTA IN IGG4-SC

Liu et al[9] revealed reduced alpha diversity and altered abundances of individual taxa in IgG4-SC patients; however, the dysbiosis was relatively modest compared with that in PSC patients. The authors also revealed several common features in microbiota and their metabolite profiles between IgG4-SC and PSC, for example, elevated L-palmitoylcarnitine, which may regulate T cells and promote inflammation. They explored distinct alterations for each diseases with 16s rRNA analysis: In IgG4-SC patients, they found depleted Blautia, which is associated with allergic diseases, and elevated succinic acid, which can modulate immune responses.

Metagenomic sequencing revealed that patients with IgG4-RD also had decreased alpha diversity, depleted Bacteroidetes, and increased Actinobacteria, including Eggerthella lenta, which was also elevated in patients with multiple sclerosis and rheumatoid arthritis. Patients with IgG4-RD also had an elevated abundance of Clostridium, Veillonella, and Streptococcus and decreased Faecalibacterium, which produces butyrate[187]. Salivary microbiome 16s rRNA analysis of IgG4-RD patients revealed significantly higher relative proportions of Mogibacterium, Solobacterium moorei, Slackia, and Moryella compared with those in patients with Sjögren’s syndrome[188]. One study also found a correlation between IgG4 levels and gut microbiome composition using 16s rRNA sequencing, with Anaerostipes, Lachnospiraceae, Megasphaera, and Eubacterium hallii association in women and Megasphaera, Eubacterium hallii group, Faecalibacterium, Ruminococcus, and Romboutsi association in men[189].

In autoimmune pancreatitis, another expression of IgG4-RD, gut dysbiosis was shown to trigger type I IFN production, IL-33, and IFN-α by activating plasmacytoid DCs, which play a significant role in driving fibro-inflammatory responses; A maintained intestinal barrier may prevent bacterial translocation and inflammation[190,191].

While the role of gut microbiota in IgG4-SC remains hypothetical and under investigation, existing evidence hints at possible immune-modulatory effects that warrant further study to clarify potential pathogenic or therapeutic implications.

COMMON AND SPECIFIC GUT-IMMUNITY PATHWAYS IN IMMUNE LIVER DISEASES

AIH, PBC, and PSC have similar features of pathogenesis (increased activation of Th cells by autoantigens, dysfunction of Treg cells and their reduced number, activation of B cells, and increased synthesis of autoantibodies). In the context of microbiota-mediated pathogenesis, common features include intestinal barrier dysfunction, SCFAs, secondary BAs, and other bacterial metabolites deficiency (the effects of disbiosis on immune cells are shown in Figure 4).

Figure 4 Microbiota-derived metabolites and immune system. Gut dysbiosis leads to a reduced production of short-chain fatty acids (SCFA), branched-chain amino acids, secondary bile acids, and indole metabolites. These changes result in enhanced activation of pro-inflammatory cells (T helper 17 and others) and in the production of pro-inflammatory cytokines. In immune liver diseases in both patients and animal models, there is a deficiency and dysfunction of regulatory T cells (Tregs). Dysbiosis-associated decrease in the production of SCFAs and indole metabolites leads to more significant decrease in anti-inflammatory Treg cells and the cytokines they secrete (interleukin-10, transforming growth factor-β). Additionally, dysbiosis increases intestinal permeability, leading to high concentrations of lipopolysaccharide in blood and activation of Toll like receptor 4, which also amplifies inflammation. SCFA can induce the hepatic gamma-type peroxisome proliferator-activated receptor; therefore, a decrease in the production of the former leads to reduced activation of the latter, which results in increased inflammation and fibrosis processes. TLR: Toll like receptor; LPS: Lipopolysaccharide; Tregs: Regulatory T cells; SCFAs: Short-chain fatty acids; IL: Interleukin; TGF: Transforming growth factor; PPAR: Peroxisome proliferator-activated receptor; Th: T helper; BCAA: Branched-chain amino acid; FXR: Farnesoid X receptor; secBAs: Secondary bile acids.

However, there are also important differences in the disease pathogenesis. For example, AE2 dysfunction plays an important role in PBC, while decreased TGR5 function plays an important role in PSC, both of which lead to cholangiocytes damage. MAIT cell dysfunction was also detected in PBC, and an excessive number of them in PSC. In both cases, these cells can be a potential therapeutic target. γδT cells play an important role in AIH, performing a dual function, both protective and pathological, and are activated by the microbiota, which also requires further study.

In diseases, general changes in the microbiota composition are also observed: A decrease in alpha diversity and an increase in Veillonella, Streptococcus, and Lactobacillus were found in most studies for all four diseases. In addition, a decrease in Clostridium was detected in AIH and PBC, but in PSC and IgG4-RD, on the contrary, an increase in these bacteria is usually observed. Another common feature of changes in the microbiota composition in AIH and PBC is the frequent detection of an increase in Klebsiella and a decrease in Ruminococcus. In contrast, in PSC, Ruminococcus is often found to be elevated. Common features of PBC and PSC are an increase in Enterococcus and a decrease in Faecalibacterium, Enterobacterium, and Coprococcus, as revealed in a systematic review with meta-analysis. No systematic review has been conducted for data on the microbiome changes in AIH. There is very limited data on IgG4-RD, but one study found an increase in Veillonella, Streptococcus, Clostridium, and a decrease in Faecalibacterium, which is most similar to PSC in terms of phenotype.

In addition, MR analysis was performed to identify microorganisms that increase the risk of disease development. In AIH, we only found two analyses that identified an increased risk for Actinomycetales and Victivallacea, and a reduced risk for Roseburia and Anaerostipes. For PBC, we found six MR analyses, where the high risk of disease most often associated with Selenomonadales, Bifidobacteriales, and Coriobacteriia, and the low risk associated with Peptostroptococcacea, Ruminococcacea, and Deltaproteobacteria. Three MR analyses were found for PSC; the most common risk factor is Ruminococcacea. As mentioned above, these microorganisms are usually elevated in PSC, and decreased in AIH and PBC.

It is also important to highlight the positive effect of probiotic therapy in AIH and cholestatic liver diseases. However, RCTs involving patients are limited by a small number of participants.

In addition, data were obtained on the effect of GPBAR1 agonist therapy in mice models of AIH and PSC, and in a human cholangiocyte cell line, which indicate a reduction in inflammation and may serve as a therapeutic target. Many studies in mice have also explored the positive effect of butyrate on models of AIH and PBC, and one study found a positive effect of butyrate on patients with PBC, although the study involved only a small number of patients. Therefore, further studies on the safety and efficacy of butyrate are required. Another potentially effective bacterial metabolite is ICA, which reduced the inflammatory activity of peripheral T cells in patients with AIH in vitro and reduced inflammation in a PSC mice model.

Thus, many studies have demonstrated the potential effectiveness of microbiota-associated therapy on the course of autoimmune liver diseases. However, most studies have been conducted on mice models, and the results may not be replicated in studies involving patients.

LIMITATIONS OF CURRENT EVIDENCE

Many studies include a small number of patients. Additionally, most studies of microbiota in PBC and PSC patients include patients on medication (most studies on AIH included treatment-naive patients), which can influence microbiota composition. The use of antibiotics and prebiotics was an exclusion criterion in almost all studies.

Moreover, an important limitation is that many studies were conducted on mouse models, which may not objectively reflect the processes occurring in the human body. Human RCTs using microbiota-targeted interventions are very few and include small numbers of patients, limiting the verification of translation of research from the laboratory to clinical practice. This may be due to the generally small number of such patients compared to, for example, patients with cirrhosis or viral hepatitis, the skepticism of clinicians and patients, and the novelty of this approach. Currently, according to the clinicaltrials.gov database, vancomycin (No. NCT02137668 and No. NCT05876182), vancomycin with amoxicillin (No. NCT06197308), and FMT (No. NCT06286709 and No. NCT06286709) are being tested in PSC. No active RCTs using gut microbiota-targeted interventions for other immune-mediated liver diseases were found in this database.

CONCLUSION

The gut microbiota plays an important role in the pathogenesis and progression of immune-mediated liver diseases, including AIH, PBC, PSC, and IgG4-SC. Dysbiosis and altered gut-liver axis interactions contribute to immune dysregulation, chronic inflammation, and tissue damage. While substantial evidence highlights significant microbiota alterations in AIH, PBC, and PSC, the role of gut microbiota in IgG4-associated liver disease remains less well-defined but is a promising area for future research. Insights into the microbiome composition and function and its immune-modulatory effects may pave the way for novel diagnostic biomarkers and therapeutic strategies, such as microbiota-targeted therapies, probiotics, and personalized interventions, aimed at restoring intestinal and hepatic immune homeostasis. Continued multidisciplinary investigations are essential for better understanding these complex host-microbiome interactions and translating findings into clinical practice.