Introduction

Primary biliary cholangitis (PBC) is a chronic and slowly progressive liver-specific autoimmune disease characterized by female preponderance and destruction of intrahepatic bile ducts that eventually leads to cirrhosis and hepatic failure [1,2,3,4]. The prevalence of PBC ranges from 20−40 cases per 100,000 individuals [2, 3, 5], although the number of patients with PBC, specifically asymptomatic PBC, is on the rise owing mainly to increased awareness and earlier detection by disease-specific antimitochondrial antibodies. The prognosis of patients with PBC has improved greatly over the past two decades thanks to earlier diagnosis and the widespread use of ursodeoxycholic acid (UDCA), which is currently the most effective treatment for PBC [4, 6] according to most guidelines [7, 8]. The terminology of PBC was changed from primary biliary cirrhosis to more accurately reflect the disorder and its natural history [9,10,11,12,13]. However, some cases still exhibit disease progression via completely unknown mechanisms.

The precise cause of PBC is unclear, but as with other autoimmune diseases, is likely related to both genetic susceptibility and environmental factors. Genetic factors are considered to play a prominent role in disease onset as higher concordance rates in monozygotic twins than in dizygotic twins and familial clustering of patients with PBC have been demonstrated in numerous family and population studies [14,15,16,17,18,19]. PBC is therefore presumed to be a multifactorial polygenic condition caused by allelic triggers and environmental influences in genetically susceptible individuals along with other epigenetic mechanisms [20, 21].

In the present review, we summarize the literature on (1) human leukocyte antigen (HLA) involvement in PBC onset, (2) genome-wide association study (GWAS) findings from North American, European, Japanese, and Chinese populations, and (3) epigenetics to explore the disease pathways involved in PBC pathogenesis.

Microsatellites and single nucleotide polymorphisms (SNPs)

Microsatellites ranging in length from 2- to 13-base-pair nucleotides are subcategories of typically 5–50 tandem repeats that make up genomic repetitive regions. Tandem repeats are evolutionarily relevant due to their instability since they mutate at rates between 103 and 106 per cell generation, i.e., up to 10 orders of magnitude greater than point mutations [22]. Microsatellites are used in genetic linkage analyses to locate genes or mutations responsible for a wide variety of diseases. Several reports have described an association between microsatellites and PBC susceptibility [23,24,25,26]. However, there are differences between the candidate genes indicated by microsatellite markers and those reported by SNP-based studies or by GWASs, which is likely attributable the intermarker distances among microsatellites that contain new or previously unidentified genes responsible for the pathogenesis of PBC.

The modern unit of genetic variation is now the SNP. SNPs are single base-pair changes in the DNA sequence that occur at high frequencies in the human genome [27]. SNPs may fall within the coding sequences of genes, non-coding regions of genes, or in intergenic regions. Although a single SNP may cause a Mendelian disease, SNPs do not usually function individually and work in coordination with other SNPs to produce clinical manifestations in complex diseases. In this aspect, association studies can determine whether a genetic variant is associated with PBC or not. To date, many significant susceptibility SNPs, such as CTLA4, TNF-a, STAT4, PTPN22, VDR, Clec16A, SOCS1, and SPIB, have been identified over the past two decades using candidate gene methods [28,29,30,31,32,33].

Associations between HLA and PBC susceptibility

Among the SNPs previously reported to be associated with susceptibility to PBC, only HLA has been consistently linked to the disease in distinct patient cohorts across ethnicities as depicted by many significant dots on genome-wide Manhattan plots of GWASs [34,35,36,37,38,39,40,41,42,43].

The HLA complex on chromosomal position 6p21 contains the most polymorphic genes in the human genome [44]. HLA genes play an essential role in numerous infectious and autoimmune diseases, such as autoimmune hepatitis as previously reviewed [45]. The products of the classical HLA class I (A, B, and C) and class II (DR, DQ, and DP) genes include cell-surface glycoproteins involved in the binding and presentation of self or non-self peptides to T-cell receptors. Class I molecules are recognized by CD8+ cytotoxic T cells presenting endogenous peptides, while class II molecules are recognized by CD4+ helper T cells presenting processed peptides from exogenous pathogens. The extent of endogenous and exogenous peptide binding to HLA molecules depends on allelic polymorphisms. Additionally, HLA class I and II molecules have functional roles in protein interactions, transcription regulation involved in the inflammatory response, and natural killer cell-cytokine interactions as part of acquired and innate immunities.

HLA polymorphisms have been extensively studied in immune-mediated diseases to reveal associations among specific alleles and ankylosing spondylitis, Behçet’s disease, psoriasis, multiple sclerosis, insulin-dependent diabetes mellitus, systemic lupus erythematosus, inflammatory bowel disease, rheumatoid arthritis, narcolepsy, autoimmune hepatitis, and autoimmune pancreatitis among others. Early investigations on associations between HLA polymorphisms and PBC were carried out >25 years ago [46]. Based on their findings, subsequent cumulative studies have provided evidence that PBC is associated with DRB1*08 as a predisposing allele and DRB1*11 and DRB1*13 as protective alleles [47, 48]. Li et al. conducted a meta-analysis to assess for relationships between HLA class II and disease susceptibility to PBC and demonstrated that HLA DR*07 and *08 alleles were risk factors for PBC in certain populations, while DR*11, *12, *13, and *15 alleles were protective factors [49].

Five key reports [47, 50,51,52,53] on the association between HLA haplotype and PBC susceptibility or resistance are summarized in Table 1. It was noteworthy that HLA DR*08 alleles were related to disease susceptibility and HLA DRB1*11 and *13 alleles conferred disease protection in haplotype analyses across all ethnicities apart from a Sardinian population [52]. Four major differences were indicated to have functional importance in the amino acid residues encoded by the different associated alleles, including the encoding of glycine-13, tyrosine-16, and leucine-74 by HLA DRB1*0801 versus that of serine-13 and phenylalanine-47 by most HLA DRB1*11 and HLA DRB1*13 alleles, respectively [47]. Moreover, HLA DRB1*11 alleles have been identified in several infectious diseases, such as hepatitis C virus [54], human papilloma virus [55], and human immunodeficiency virus [56], while HLA DRB1*13 alleles have been associated with hepatitis B virus [57] and human papilloma virus [58]. Molecular mimicry has been widely suspected to initiate autoimmunity in patients with PBC. Since protective HLA class II alleles also play a functional role in blocking virus infection, one of the pathogenic mechanisms in PBC may be viral insult. Retrovirus infection was reportedly associated with PBC, although these findings have not yet been replicated [59]. Lastly, it should be noted that individuals harboring the above haplotypes constitute only a minority of patients with PBC, suggesting that other candidate genes and environmental factors evoke PBC pathogenesis as well.

Table 1 HLA haplotype associations with PBC
Full size table

GWASs on PBC

A gene-specific, candidate-driven approach that specifically tests a small number of pre-specified genetic regions has limitations in finding pathogenic genes of such multi-factorial diseases as PBC. GWASs typically focus on associations between SNPs and traits like major human diseases in investigations of the entire genome. There have been extensive GWASs aiming to clarify genetic susceptibility in patients with PBC. To date, seven GWASs [34,35,36,37,38, 42, 43], two Illumina immunochip studies [39, 40], and one genome-wide meta-analysis (GWMA) [41] on PBC have been documented in well-characterized cohorts in North American, European, Japanese, and recently Chinese populations (Table 2). All studies confirmed that the HLA class II domain showed the strongest association with PBC susceptibility, particularly at the HLA-DRB1, -DQA1, and -DQB1 loci, and indicated the involvement of HLA class II polymorphisms in the pathogenesis of PBC across ethnicities. However, HLA alone does not explain the entire genetic predisposition to PBC, mainly since 80–90% of patients with the disease do not carry the most common HLA susceptibility alleles. In this regard, other genes apart from HLA loci are suggested to contribute to disease development. At present, GWASs have identified 44 non-HLA PBC predisposition loci at a genome-wide level of significance (Table 3).

Table 2 GWASs on PBC
Full size table
Table 3 Non-HLA risk loci identified through GWASs to be associated with PBC at the genome-wide level of significance
Full size table

The first GWAS [34] in a North American cohort revealed a significant association between PBC and genetic variants at interleukin (IL)12A and IL12RB2, respectively encoding IL-12 p35 and IL-12 receptor β2. The second GWAS [36] confirmed the existence of additional risk loci, including interferon regulatory factor 5 (IRF5), transportin 3 (TNPO3), and transcription factor Spi-B (SPIB) encoding a transcription factor involved in B-cell receptor signaling and T-cell lineage decisions. A subsequent GWAS from Japan showed that tumor necrosis factor superfamily, member 15 (TNFSF15) and POU domain class 2-associating factor 1 (POU2AF1) genes also constituted novel risk loci in Japanese patients with PBC, and a recent GWAS of a Han Chinese population found the highest association was between TNFSF15 and PBC. The IL12A and IL12RB2 loci were not significantly associated with PBC in the Asian cohorts. This discrepancy among ethnicities indicated important differences in the pathogenesis of PBC despite several common key molecules and pathways, such as the IL-12 pathway to induce Th1 polarization of CD4+ T cells. TNFSF15 interacts with death receptor 3 (DR3) not only to promote T-cell expansion, but also to induce apoptosis in cells that overexpress DR3, which can synergistically interact with IL-12 and IL-18 to promote interferon-gamma production [60]. Thus, the IL-12 signaling pathway appears to be a key player in the pathogenesis of PBC worldwide. To support this hypothesis, STAT4, which factors prominently in IL-12 signal transduction polymorphisms, was associated with susceptibility in Japanese and Han Chinese PBC patients [38, 43].

The IL-12 pathway has also been associated with animal models of PBC and a pediatric case of a congenital IL-12 deficiency that developed PBC [61]. Although antibodies or drugs targeting the IL-12 pathway would seem to be effective, clinical trials using ustekinumab, a human monoclonal antibody directed against IL-12 and IL-23, failed to produce clinical effects in phase II trials [62]. One reason explaining this disconnection between GWAS and clinical testing results may be that clinicians typically encounter patients who have already become complicated with cholestasis; in fact, the immunological destruction of cholangiocytes occurs in the very early stages of PBC. Accordingly, the mechanisms of disease progression should also be addressed to halt the deterioration of disease status and afford PBC patients an improved prognosis. We are now at the next stage of how genetic findings can be translated into the clinical setting.

It is particularly interesting that many genes implicated in PBC pathogenesis by GWASs have also been reported in other immune-mediated disorders, such as systemic lupus erythematosus, systemic sclerosis, and Sjögren’s syndrome (Table 3), suggesting genetic overlap among autoimmune disorders and the existence of substantial pleiotropy, i.e., one gene leading to two or more seemingly unrelated disease phenotypes. Therefore, on a larger scale, understanding the mechanisms involved in the onset and progression of certain other autoimmune diseases may correspondingly shed light on those in PBC.

Lastly, it is noteworthy that Juran et al. identified risk-conferring epistatic interactions between IL12RB2 and IRF5 loci [39] as well as between CTLA4 and TNFα loci [63] in the pre-GWAS era. Epistatic interactions between genes in the pathogenesis of PBC revealed by GWASs also merit exploration in future studies.

Epigenetics on PBC

Epigenetics is the study of stable heritable changes in gene function that do not involve alterations in DNA sequence. Several molecular mechanisms are involved in epigenetic modification, including DNA methylation, histone modification, and modification via noncoding RNA. Although GWASs have identified several candidate genes implicated in PBC pathogenesis, they are thought to explain <20% of heritability in PBC; indeed, a relatively low concordance of PBC was found in an earlier study of monozygotic and dizygotic twins [19]. Epigenetics may also have an important role in symptoms, disease progress rate, biochemical response to UDCA therapy, and other disease phenotypes, implying that patients at genetic risk for PBC may manifest the disease via epigenetic factors as well. Case–control studies have identified potential exposure sources in PBC, such as urinary tract infection, smoking and cigarette smoke, and exogenous estrogens [15]. How these factors are molecularly involved in PBC onset remains unclear. Elsewhere, it was reported that X-chromosome gene expression instability might also participate in female predominance in the disease [64, 65], and that monozygotic twins and sisters discordant for PBC exhibited particular epigenetic differences [66]. These collectively support the notion that epigenetic factors influence the onset of PBC. Epigenetics has numerous and varied potential medical applications and may be able to uncover novel pathways for therapeutic discovery and intervention in PBC.

Conclusions and future directions

The pathogenesis of PBC is incompletely understood but appears to involve genetic susceptibility and resistance alleles in HLA and other gene loci, with a possible overlap with several autoimmune diseases. It is also probable that genetically susceptible individuals develop PBC following exposure to environmental factors that trigger DNA methylation, histone modification, modification via noncoding RNA, and other epigenetic modifications. Consequently, both adaptive and innate immune responses result in portal inflammation and bile duct epithelial damage. In addition to susceptibility, the precise mechanisms of PBC progression should be addressed to improve patient prognosis and quality of life.