Mouse model of primary biliary cholangitis with a striking female predominance: A new powerful research tool

Potential conflict of interest: Nothing to report. See Article on Page 1189 Primary biliary cholangitis (PBC), hitherto called primary biliary cirrhosis, is a chronic cholestatic liver disease of unknown etiology, characterized by autoimmune‐mediated destruction of small‐ and medium‐size intrahepatic bile ducts. It affects mainly women in the fifth and sixth decade of life, with a female‐to‐male ratio of 1:10.1 A genetic basis for susceptibility has been demonstrated.3 The laboratory hallmarks of PBC are antimitochondrial antibody (AMA; a highly disease‐specific autoantibody present in 95% of cases), increased levels of alkaline phosphatase and immunoglobulin M, and, in a lower proportion of cases, PBC‐specific antinuclear antibodies. The major auto antigenic targets of AMA are the E2 subunits of pyruvate dehydrogenase (PDC‐E2), branched chain 2‐oxo acid dehydrogenase (BCOADC‐E2), and 2‐oxo‐glutarate dehydrogenase (OGDCE2).4 AMA can be detected by indirect immunofluorescence (IFL), immunoblotting, and enzyme‐linked immunosorbent assay (ELISA). IFL, using rodent tissue sections, has been the mainstay of AMA detection5 (Fig. 1A) and allows simultaneous identification of several other autoantibodies relevant to autoimmune liver disease, but is subjective and time‐consuming. Currently, commercial ELISAs are reliable,6 do not require special skills, and allow the evaluation of numerous sera in a single run.Figure 1: (A) AMA detected by IFL stains the mitochondria‐rich renal tubules (right) and gastric parietal cells (left) of rodent kidney and stomach. (B) Left: Chronic nonsuppurative destructive cholangitis (CNSDC) with focal granulomatous reaction in a patient with PBC (picture kindly provided by Dr. Yoh Zen, Kobe University Hospital, Japan). Right: CNSDC in the mouse model described in the present issue of Hepatology.11Affected patients usually complain of fatigue and pruritus, which occur independently of disease severity, and often have associated autoimmune disorders (e.g., scleroderma, Sjogren's syndrome, and systemic lupus erythematosus). Histologically, the disease is characterized by inflammation leading to destruction/loss of small‐ and medium‐size intrahepatic bile ducts (Fig. 1B) and development of fibrosis and biliary cirrhosis. Epithelioid granulomas with no caseous necrosis surrounding the damaged bile ducts, a characteristic lesion of PBC, can be observed at any stage. The definitive diagnosis of PBC requires the fulfilment of three criteria: presence of AMA, a cholestatic biochemistry (i.e., raised alkaline phosphatase) for over 6 months, and a compatible or diagnostic liver histology. When only two of these three criteria are fulfilled, a diagnosis of probable PBC is made. In the presence of a cholestatic liver profile and positive AMA, however, current clinical practice is to reach the diagnosis without histological assessment, in view of the disease specificity of AMA, a liver biopsy becoming essential when AMA is negative, when the biochemical profile shows a mixed cholestatic and hepatocellular pattern, or in the presence of comorbidities (e.g., nonalcoholic steatohepatitis). Curiously, despite the many features suggestive of an autoimmune pathogenesis (female preponderance, association with other autoimmune disorders, and positive autoantibodies), PBC does not respond to classic immunosuppressive treatment, whereas disease progression is slowed by treatment with ursodeoxycholic acid.2 It is possible that by the time the disease becomes symptomatic, the original pathogenic events leading to the autoimmune reaction against the intrahepatic bile ducts have long disappeared. Thus, AMA can be positive several years before the appearance of the clinical disease and predicts its onset, as shown by the Newcastle group: Patients with normal liver function tests and no symptoms suggestive of liver disease were "accidentally" found to be positive for AMA—as the result of an autoantibody profile testing. Liver histology was compatible with PBC in 83% of them.7 Ten years later, three quarters of these subjects became clinically symptomatic with persistently cholestatic liver biochemistry,8 the median follow‐up being 17.8 years, with 1 subject positive for AMA as long as 23 years before the diagnosis of PBC was made. It is therefore not surprising that understanding of the early events involved in the induction of tissue inflammation and autoimmunity in PBC remains elusive in humans. To overcome this, over the last decade, researchers have intensively investigated animal models attempting to recapitulate the characteristic clinical features of human PBC, with AMA and lymphocyte infiltration and ensuing biliary epithelial cell pathology.9 A novel animal model mimicking very closely the human disease is published in the current issue of Hepatology.11 The investigators have exploited a mouse strain with a deletion of the interferon (IFN) 3' untranslated region adenylate‐uridylate–rich element. This "designer" mouse has chronic expression and over production of IFNγ and develops a primary biliary cholangiopathy similar to PBC, characterized by up‐regulation of total bile acids, spontaneous production of AMA, and portal duct inflammation. Most important, in contrast to previous models, the disease has a sex‐differential bias with a clear female predominance, akin to human PBC. Female ARE‐Del–/– mice develop moderate‐to‐severe portal tract lymphoid cell infiltration at 20 weeks of age, whereas male ARE‐Del–/– mice have only mild‐to‐moderate inflammatory infiltration. Portal and lobular inflammation, small bile duct destruction, and granuloma formation are more severe in female than male ARE‐Del–/– mice. Moreover, though progression to severe fibrosis similar to human PBC is not observed, female ARE‐Del–/– mice do develop mild fibrotic changes that are not observed in male mice. Antibodies to PDC‐E2, BCOADC‐E2, and OGDC‐E2 become detectable in 8‐ to 10‐week‐old female ARE‐Del–/– mice and continue to be present at the age of 20 weeks, the dominant reactivity being against PDC‐E2. In contrast, 8‐ to 10‐week‐old male ARE‐Del–/– mice develop antibodies only to PDC‐E2, which disappear by 20 weeks of age. Some features of PBC are present also in heterozygote female mice. In particular, female ARE‐Del+/– mice have portal tract inflammatory infiltrates and biliary duct lesions similar to female ARE‐Del–/– mice, whereas male ARE‐Del+/– mice have only mild lesions. In addition, AMA and total bile acid blood levels are elevated in the heterozygote female, but not in the male. Expression signature studies indicate that major histocompatibility complex class II genes are highly expressed in both male and female ARE‐Del–/– mice, suggesting a critical role for these genes in the initiation of the cholangiopathy. In contrast, IFN signaling is predominant in female ARE‐Del–/– mice, which also have higher levels of serum IFNγ than male mice. Similarly, female ARE‐Del–/– mice have higher signaling of chemokines (monokine induced by IFNγ, IFNγ‐inducible protein 10, and macrophage inflammatory protein beta) and cytokines (tumor necrosis factor alpha, interleukin [IL]‐10, and IL‐13) than their male ARE‐Del–/– counterparts. Interestingly, this article shows that in the human disease, affected patients have a marked liver tissue up‐regulation of IFNγ signaling in areas of chronic nonsuppurative destructive cholangitis (CNSDC) and biliary epithelial cell lesions. Moreover, it shows that female ARE‐Del–/– mice have gene expression patterns of T‐helper‐cell differentiation, dendritic cell maturation, and B‐cell development similar to those identified in affected patients. Cell transfer experiments from ARE‐Del–/– mice to B6/Rag1–/– mice (an immune‐deficient strain that produces no mature T or B cells) show that CD4 T cells cause histological lesions similar to those observed upon transfer of whole spleen cells, indicating that CD4 T cells are critical in the induction of cholangitis. This article reports, for the first time, on an animal model of PBC that recapitulates faithfully the human disease, including its striking female preponderance. Convincing evidence is provided that differential gene expression in male and female mice underscores the differences in disease phenotype, and that IFNγ plays a pivotal role in the pathogenesis of the disease. This new model is ideal to unravel further the pathogenic mechanisms of PBC, investigate the effect of currently used drugs, and design novel, more‐effective therapeutic agents.