Update on the Genetics and Genomics of PBC

Abstract

Despite recent progress, the pathogenic mechanisms governing PBC development, treatment response and outcome remain unknown. This deficiency is in large part due to the complex nature of PBC, wherein various environmental factors may be capable of prompting disease, but only in the context of underlying genetic susceptibility. Identification of genomic loci containing these heritable risk factors has been slowed by the rarity and late onset of PBC, which has made difficult the collection of sufficient numbers of patients and family members for meaningful genetic analyses. Advancements in our ability to catalog the genetic variation in large numbers of individuals at a genome-wide scale, coupled with unprecedented efforts to recruit PBC patients for genetic study, positions us to generate data that could fundamentally change our understanding of PBC and lead to clinical innovation. Indeed, the first genome-wide association study for PBC has been published, in which multiple genes involved with IL12 signaling, a pathway that is being targeted in treatment of other inflammatory conditions, were implicated in disease. However, this study was relatively small in the genome-wide milieu and a significantly expanded effort will be necessary to truly elucidate the genetic architecture of PBC. Moving ahead, cooperation between the groups collecting biospecimens and generating genome-wide data from large numbers of patients with PBC will be essential, not only to increase power for fine mapping and future studies of rare variants and epistasis; but to streamline efforts to perform functional validation of novel discoveries. Here we provide a brief update of the current state of genetics in PBC to form a basis for understanding the considerable progress that is likely to be made in the coming years.

Introduction

Primary biliary cirrhosis (PBC) is a rare autoimmune liver disease typified by destruction of small to medium intrahepatic bile ducts causing chronic cholestasis and development of fibrosis, which often progresses to cirrhosis and end-stage liver disease [1]. As currently understood, PBC is initiated when tolerance to a ubiquitously expressed subunit of the pyruvate dehydrogenase complex (i.e. PDC-E2) is lost; a phenomenon that is often associated with the development of PDC-E2 specific anti-mitochondrial antibodies (AMAs) [2], which are detectable in some 90% of PBC patients and may be present for years prior to clinically recognizable PBC [3]. Recent findings indicate that biliary targeting of the autoimmune attack is due to the presence of intact immunoreactive PDC-E2 within apoptotic blebs of biliary epithelial cells (BECs); a feature that appears to be unique to BECs but not specific to PBC [4]. While progress has been made, the mechanisms underlying the development of autoimmunity in PBC as well as the factors contributing to its variable disease progression, response to treatment, and ultimately, outcome remain mysterious.

As with most autoimmune diseases, PBC is genetically complex. That is, individual alleles contributing to disease processes are not likely to be sufficient for PBC development, and instead simply act as modifiers of disease risk by modulation of biological processes contributing to pathogenesis. The overall impact of such risk (or protective) alleles on disease is likely to be mediated by interaction, possibly by means of environment or through epistasis with other functional genetic variants, adding to the apparent complexity.

Two models are regularly invoked to describe the genetic architecture of complex diseases; the common disease-common variant [5] and the common disease-rare variant [6] hypotheses. The common-variant premise assumes that genetic polymorphisms with some effect on functionality may become quite prevalent in the population (minor allele frequency [MAF] in the range of 5% or greater) and act as the primary drivers of complex disease. Evidence from genome-wide association studies (GWAS) supports the role of common genetic variants in many complex diseases [7]; however it is becoming clear that these variants are not highly penetrant, generally with disease odds ratios less than 2 and more often in the range of 1.1–1.4. In contrast, the rare-variant proposal suggests a highly diverse spectrum of disease loci wherein rare genetic variants (a loose term generally describing polymorphisms with a population frequency of less than 1% but also including short lived de novo variants present only in an individual or family) that are highly-penetrant, are the major cause of disease [6]. While these two hypotheses are at odds, current thinking suggests that both common and rare variants will have significant impact on most complex diseases, including PBC [7].

Evidence for a genetic contribution to PBC

As with the majority of autoimmune diseases, the genetic contribution to PBC risk and pathogenesis is thought to be quite important. Support for this assertion is provided by familial clustering of PBC, high disease concordance in monozygotic twins [8], and increased prevalence of other autoimmune conditions in PBC patients and their family members [9].

As more genetic material is shared among family members than with the general population, clustering of disease within families can, with some caveats, be used as evidence of a genetic influence on disease. Familial aggregation of PBC has been long appreciated, with past epidemiological studies showing familial PBC prevalence (i.e. multiple cases of PBC within a family) to range between 1.0% and 6.4% [10]; significantly higher than expected in the general population. Other familial evidence for genetic risk to PBC comes from a geographically based epidemiological study from the UK which found the relative risk to siblings (λ_s) of PBC patients to be 10.5 [11], on the same order of magnitude as other autoimmune diseases [9]. Moreover, a large questionnaire based study of 1032 US PBC patients identified familial history as the strongest risk factor for PBC development (5.9% PBC vs. 0.5% controls, OR 10.7, 95% CI 4.2–27.3, p<0.001) [12].

In addition to increased prevalence of PBC itself, the presence of AMA has also been shown to aggregate in first degree relatives of afflicted individuals, regardless of familial history of PBC [13]. As well, the increased incidence of other autoimmune disorders in PBC patients and their family members is well established [12] and often noted as further evidence for the genetic component of PBC. This is because the pathogenic mechanisms underlying autoimmune diseases are broadly shared, and thus, genetic propensity to their development is to some extent a general phenomenon [14]. Within this autoimmune-permissive genetic background, disease specific environmental factors and genetic susceptibility contributes to the development of specific autoimmune diseases like PBC [14].

Genetic variants effecting PBC susceptibility

The rarity and late onset of PBC has prohibited collection of adequate numbers of affected families to perform linkage analysis in PBC, and the majority of past efforts to identify PBC associated genetic variants have focused on candidate genes. Many of these projects were plagued by low power and insufficient coverage of the genes of interest; primarily due to the genetic naiveté of the times and not a reflection on the dedicated investigators. Recently, many of the groups performing the early genetics work have significantly expanded their collections, the first GWAS for PBC has been published [15] and more genome-wide efforts are expected in the coming years. Below we present an updated overview of the important genetic findings for PBC, whether identified through studies of candidate genes or by the GWAS, generated to date (see also Table 1).

Table 1.

Summary of genetic variants identified/studied in the larger PBC investigations

Gene	Variant	Location/Type	Population(s)	Affect	Reference
HLA DRB1	*0801	Haplotype	European, North American	Risk	16–18
	*0803	Haplotype	Japanese	Risk	19
	*11	Haplotype	Italian	Protective	17, 21
	*13	Haplotype	UK, Italian	Protective	17, 21
HLA DQB1	rs2856683		Canadian GWAS (peak signal)	Risk	15
TNF	−308A/G (rs1800629)	Promoter	Multiple populations	Unresolved	9, 23–27
IL12A (p35)	rs4679868	3’ Flanking	Canadian GWAS (fine mapping)	Risk	15
IL12RB2	rs6679356	Intronic	Canadian GWAS (fine mapping)	Risk	15
STAT4	rs3024921	Intronic	Canadian GWAS (replicated)	Risk	15
CTLA4	rs231725	3’ Flanking	US	Risk	42
PTPN22	R620W rs2476601	Coding	US, Italian, Canadian	No association	10, 41, 46
PDCD1	PD1.3 rs11568821	Intronic	US	No association	38
VDR	BsmI polymorphism	Intronic	Italian, Japanese	Risk	51
SLC4A2 (AE2)	12 Tag-SNPs	Across gene	US	No association	54

HLA

The human leukocyte antigen (HLA) genes are located in the gene-dense and highly polymorphic major histocompatibility complex (MHC) at chromosome 6p21. The connection between autoimmunity and HLA has been appreciated for quite some time and genetic variants in HLA genes have been associated with the majority of autoimmune diseases. Thus, it is not a surprise that numerous efforts at deciphering the genetics of PBC have focused on this region. To date, over 20 studies have been published on the HLA association with PBC [10].

The most commonly detected HLA association with PBC has been with the class II DRB1*08 allele family; specifically DRB1*0801 in European and North American Caucasians [16–18], and DRB1*0803 in the Japanese [19]. The DRB1*0801 association has recently been confirmed in larger studies from the UK and Italy [17, 20]. In addition to the DRB1*0801 allele, the DRB1*0801-DQA1*0401-DQB1*402 extended haplotype has been associated with PBC in studies of patients from the US [18], UK, and Italy [17]. Strong linkage disequilibrium (LD) in the region, the fact that this haplotype is somewhat less frequent than the DRB1*0801 allele, and data that the Japanese risk haplotype is DRB1*0803-DQA1*0103-DQB1*0601[19] points to the DRB1*08 allele family as the likely HLA determinant of PBC risk, although this conjecture is certainly far from definitive. Further support for the involvement of HLA genes with PBC comes from the recent Canadian GWAS in which strong association signals were found across the MHC locus encompassing HLA DQB1, DPB1, DRB1, DRA, c6orf10 and BTNL2 genes; peaking at rs2856683 in DQB1[15]. Haplotype analysis of these variants found that a four SNP haplotype (rs2395148, rs3135363, rs2856683 and rs9357152) accounted for all of the PBC risk [15]. Whether or not this observed association is simply indicative of underlying linkage with the previously recognized extended HLA DRB1*08 risk haplotypes remains to be explored.

In addition to alleles associated with an increased risk of PBC, some class II HLA alleles have been shown to protect against PBC including DQA1*0102 in US and Japanese studies [16, 19] and DQB1*0602 in a US study [16], neither of which has been confirmed. More convincingly, DRB1*13 was found to be protective against PBC in both UK and Italian PBC patients and DRB1*11 was protective only in the Italian patients [17, 21]. A large follow-up study by the Italians confirmed the protective nature of the DRB1*11 and DRB1*13 haplotypes, and they noted a dosage effect wherein possession of two protective haplotypes (i.e. *11/*11 or *11/*13) demonstrated lower odds ratios than carriage of a single protective haplotype [20]. Moreover, they found that DRB1 *08/*11 heterozygosity was neither a risk for or protective from PBC [20] suggesting these haplotypes offset each other in their effects on disease risk. Variation in the hydrophobicity or size of four DRB1 amino acids differing between the DRB1*08 risk allele and DRB1*11/*13 protective alleles has been hypothesized as a potential mechanism behind the observed associations, perhaps through alteration of the binding properties of the MHC molecule [17]. Evidence supporting this assertion would be of great importance.

Compared to HLA genes of class II MHC little effort has been made to assess the impact of class I HLA alleles with PBC and a handful of small studies identified no significant associations. A few rare alleles of HLA–B were reported to be associated with PBC in an Italian study of 112 PBC patients and 558 controls [21]; however, these findings have not been confirmed. While there is currently no evidence suggestive of a major role for variant alleles of class I MHC genes in PBC, they remain interesting candidates for future investigation.

TNF

TNF has long been long appreciated as an important inflammatory mediator with potential involvement in autoimmunity, and remains among the few class III MHC genes to have been studied in PBC. After it was demonstrated that the A allele of a G/A polymorphism at position −308 in the TNF promoter resulted in increased TNF expression [22]; a number of PBC association studies focused on this variant were performed [9, 23–27]. The first study found carriage of the less common −308A allele to be reduced among PBC patients compared to controls [24], suggesting the allele expressing high levels of TNF was protective against PBC. Three of the studies reported no association between the −308 polymorphism and PBC, but alluded to effects on disease severity [23, 27, 28]. However, the findings of these studies were at odds as the first found heterozygous patients (i.e. −308AG) to have significantly higher Mayo PBC Risk Scores than −308G/G patients, suggesting that −308A contributed to disease severity [23]; the second found homozygosity for −308G/G to be elevated in patients with an advanced histological stage, suggesting that the −308G allele was to blame [28]; and the third found carriage of the −308A allele to be moderately increased in patients who had received OLT compared to those who had not [27]. Two other studies showed no association of the −308 allele with disease or severity [25, 26]. To date, the possible association between PBC and the TNF −308 allele remains unresolved. However, considering the evidence for the role of −308A in autoimmunity, and on development of hepatic fibrosis, further study of this polymorphism in PBC is necessary.

IL12

IL12 is a major cytokine for development of Th1 responses, playing a role in both innate and adaptive immunity [29]. Functional IL12 is comprised of 2 subunits, p35 (encoded by the IL12A gene) and p40 (encoded by the IL12B gene), which are secreted by antigen presenting cells (APCs) following their activation [29]. This heterodimer signals through a cell surface receptor on naïve CD4+ T-cells composed of two chains, IL12Rβ1 and IL12Rβ2 coded by the IL12RB1 and IL12RB2 genes, leading to commitment and activation of a Th1 response after T cell receptor engagement, a process that is largely dependant on STAT4 [29]. In addition to their participation in IL12 signaling, the p40 subunit and the IL12RB1 chain are also components of IL23, the major driver of Th17 polarization and proliferation [30]. An overview of IL12 and IL23 is provided in FIGURE 1.

Figure 1. Structure and characteristics of IL12 and IL23.

IL12 and IL23 share in common the p40 subunit coded by the IL12B gene as well as the β1 subunit of the receptor coded by the IL12RB1 gene. IL12 signaling through STAT4 contributes to activation and maintenance of Th1 cells whereas IL23 signaling also invokes polarization and proliferation of the Th17 response. The only genome-wide association study in PBC to date identified genetic variants in the IL12 specific IL12A (p35) and IL12RB2 (β2 receptor subunit) genes as well as in STAT4.

Genetic variants in two genes of the IL12 immunomodulatory signaling pathway were found to be strongly associated with disease in the first PBC GWAS, second only to the HLA association [15]. Specifically, IL12A, encoding the p35 subunit; and IL12RB2, encoding the β2 subunit of the IL12 receptor were identified. For IL12A the strongest association in combined analysis was with rs6441286 in the 3’ flanking region of the gene (OR=1.54, 95% CI 1.38–1.72, p=2.42×10⁻¹⁴) with fine mapping efforts in the Canadian population giving a peak signal at rs4679868 (some 4.7 KB from 6441286 and in very tight linkage). For IL12RB2 the strongest association in combined analysis was with rs3790567 in intron 9 (OR=1.51, 95% CI 1.33–1.70, p=2.76×10–11) with fine mapping yielding a peak signal at rs6679356, also in intron 9 [15]. Moreover, association with the SNP rs3024921 in intron 3 of STAT4, an important downstream component of IL12 signaling, was also significant in the combined analysis (OR=1.81, 95% CI 1.45–2.26, p=5.76×10-8).

Increased levels of IL23, relating to an imbalance of Th17/Treg (i.e. increased Th17, decreased Treg), have been found in PBC patients compared to hepatitis b patients and healthy controls [31]. Other evidence for the involvement of Th17 cells in PBC has been posed in both human cells and the IL2Ra KO mouse model of PBC [32]. Moreover, histologically detectable autoimmune cholangitis in the TGFβR2 PBC mouse model was alleviated by deletion of IL12p40, suggesting this IL12 subunit is essential to development of cholangitis in the model, lending further evidence for the importance of IL12 in PBC pathogenesis [33]. In addition to its role in IL12 and IL23, IL12p40 can form a homodimer that antagonizes IL12 through competitive binding of the IL12 receptor complex. As well, p40 homodimer was recently shown to induce the production of nitric oxide (NO) by APCs via signaling through the IL12Rβ1, but not IL12Rβ2 receptor [34], which subsequently has a suppressive effect on regulatory T cells [34].

While the evidence for the role of altered IL12 signaling in PBC is strong, determination of the functional consequence of the disease associated genetic variants in IL12A (p35) and IL12RB2 will be necessary to better understand the mechanisms contributing to development of autoimmunity in PBC.

CTLA4

The cytotoxic T-lymphocyte antigen 4 (CTLA4) gene encodes an immunoreceptor that plays a key role in the maintenance of tolerance and prevention of autoimmunity, in part through inhibition of T-cell activation upon MHC presentation of antigen to the T-cell receptor (TCR) [35]. Genetic variants of CTLA4, most often 49AG (rs231775) and/or CT60 (rs3087243), have demonstrated association with a variety of autoimmune diseases[10]. Primary studies of CTLA4 in PBC identified an association with the G allele of 49AG in UK [36], Chinese [25] and French [37] populations, that was not confirmed by our large US study [38] or by a smaller study from Brazil [39]. A follow-up effort by the UK group, which investigated additional CTLA4 SNPs in a larger patient population, did not replicate their initial findings of the 49AG association with PBC, calling into question this SNPs relevance to disease [40]. However, Italian [41] and US [38] studies did provide evidence for the potential involvement of CTLA4 variants with AMA positivity among PBC patients.

In a follow-up to our earlier study, we utilized an LD-based approach to select 8 SNPs across CTLA4 (and flanking by 5KB) providing quite good coverage (MAF 1% and r2=0.9) of the common genetic variation. These SNPs were genotyped in 402 PBC patients (351 AMA positive) and 279 controls [42]. From the analysis of this data we identified a novel association of AMA positive PBC with the SNP rs231725 in the 3’ flanking region of CTLA4 (approximately 2kb downstream of the 3’UTR) under a recessive model (“AA” genotype 17% AMA+ PBC vs. 9% controls, OR=2.22, 95% CI 1.34–3.68, Pcorrected=0.003). Moreover, homozygosity for this allele was found in only 2 of 45 AMA negative PBC patients (4%) suggesting the potential functional consequence of this CTLA4 association may strongly contribute to the AMA positivity classically found in disease, especially considering that a clinical assay was used for AMA assessment that is likely not as sensitive as some of the newer assays. Interestingly, haplotype analysis showed that the implicated A allele of rs231725 is isolated on a single haplotype that also contains the 49AG and CT60 alleles. A later study of CTLA4 in PBC patients and controls from Canada also found association between PBC and a very similar haplotype to that identified in our study [43]. Of note, remapping of the CTLA4 association with type I diabetes using a large number of multiply affected nuclear families recently reported the association peak to be at rs231727 [44], a SNP that is strongly linked (i.e. r²=1.0) with rs231725 identified for PBC in our study. In light of these recent findings, the continued investigation of CTLA4 genetic variation in PBC is clearly warranted.

Other PBC candidate genes

PTPN22

The protein tyrosine phosphatase 22 (PTPN22) gene encodes lymphoid tyrosine phosphatase (LYP), a molecule which has an inhibitory effect on T-cell activation following engagement of the T-cell receptor. A coding (SNP) in this gene, C1858T (rs2476601), results in an arginine to tryptophan substitution at codon 620 (R620W) which apparently alter the function of both T- and B-lymphocytes [45] and has been associated with many autoimmune diseases. Investigation of this SNP in Italian [41], Canadian [46] and US [10] studies found no association with PBC. Together, these studies suggest that the C1858T variant of PDCD1 does not directly confer risk for PBC.

PDCD1

The programmed cell-death 1 (PDCD1) gene encodes a receptor with T-cell inhibitory function similar to CTLA4. Recent reports have suggested that PDCD1 and CTLA4 work synergistically through distinct mechanisms in order to inhibit T-cell activation and promote tolerance [47]. PDCD1 and its ligands PDL1 and PDL2 were shown to be up-regulated in the livers of PBC, autoimmune hepatitis, and chronic hepatitis C patients, but not in normal controls [48] demonstrating a potential role in liver inflammation and by extension PBC. The PD1.3 SNP (rs11568821) in PDCD1 has been implicated in multiple autoimmune disorders and is thought to alter PDCD1 regulation [49]. To date, the only published study of this variant in PBC found no association in US patients [38].

Keratins

Genetic variation in exonic (coding) sequence of three keratin (K) genes (K8/K18/K19), which are expressed by biliary epithelial cells, was examined in 201 Italian patients and 200 blood bank controls in response to the observation that mutations in K8 and K18 are overrepresented in patients with end-stage liver disease and K8 null mice develop serum antimitochondrial antibodies [50]. This was accomplished through the use of PCR followed by denaturing high-performance liquid chromatography (dHPLC). Analysis showed an increase in carriage of pathogenic variants across the three genes (17/201 PBC patients vs. 4/200 controls; OR 4.53, 95% CI 1.5–13.7, p=0.0004) [50]. Possible deficit in the cytoprotective function of these genes is postulated; however this is certainly just conjecture. Perhaps more importantly, this finding illustrates the challenge of detecting and evaluating rare genetic variants with a possibly large disease effect. For instance, K18 contributed 1 newly discovered pathogenic variant to the analysis (in a patient) and K19 contributed 4 (3 patients with 1 variant, 1 control with another). Evidence for the role of two common K8 variants (G62C and R341H) was stronger (total 12/201 PBC and 3/200 controls) [50]. Functional studies of these variants would be quite interesting.

Vitamin D receptor (VDR)

Three polymorphisms in the Vitamin D receptor (VDR) were genotyped in Japanese and Italian PBC patients and controls by PCR-RFLP (polymerase chain reaction restriction fragment length polymorphism) [51] because of the known immuno-modulatory effects of VDR agonists and previous smaller scale studies demonstrating association with PBC [52, 53]. Homozygosity for the minor allele of the intronic BsmI polymorphism (SNP rsID not given) was significantly increased in PBC patients compared to controls, overall (21% PBC vs. 13% controls, OR 1.80, 95% CI 1.19–2.73, p=0.005), as well as in the separate populations, despite a substantial difference in variant frequencies between Japanese and Italians (Japanese 7% PBC vs. 1% controls, OR 13.8, 95%CI 1.79–105.81, p=0.001; Italians 40% PBC vs. 27% controls, OR 1.83, 95% CI 1.12–2.99, p=0.019). Follow-up studies would be of interest.

AE2 (SLC4A2)

Twelve SNPs tagging the common genetic variation (r²=0.85, MAF 5%) in the Cl−/HCO₃ − anion exchanger 2 (AE2; officially SLC4A2) were genotyped in 409 PBC patients and 300 controls [54] as its expression has been shown to be decreased in PBC patients [55] and Ae2 deficient mice develop a phenotype resembling PBC [56]. Allele and genotype frequencies were quite significant between patients and controls with no significant associations noted [54]. However, homozygosity for the minor allele of the coding SNP rs2303929 was significantly increased in AMA negative compared to AMA positive patients (17% vs. 5%, OR 3.79, 95%CI 1.49–9.09, Pcorrected=0.03) suggesting a possible contribution of AE2 to the development of AMA in PBC [54]. While this study did not find genetic evidence to support the observed reduction of AE2 expression in PBC patients, potential regulatory sequence outside of a 5KB window flanking the gene and possible epigenetic changes to the AE2 promoter were not explored. Moreover, the affect of these variants on disease progression were only cursorily studied leaving open the possibility that AE2 polymorphisms could play a role in disease severity. Considering the evidence provided by the Ae2 knock-out mice and the observed association with AMA negativity, this gene remains of considerable interest to the study of PBC.

Other genes

Polymorphisms in genes influencing xenobiotic metabolism and transport were assessed in Italian PBC patients to address the hypothesis that halogenated xenobiotics modify self-molecules and facilitate the breakdown of tolerance to mitochondrial antigens [57]. This study genotyped a variant in the multidrug resistance 1 gene (MDR1) that is associated with decreased expression in the intestine, three SNPs in the PXR (pregnane X receptor) gene which regulates expression of MDR1 in the liver and intestine, and polymorphisms of the drug metabolizing CYP2D6 and CYP2E1 cytochrome P450 genes [57]. None of these polymorphisms was associated with PBC development; however, an allele of CYP2E1 was increased in advanced-stage PBC patients compared to those with earlier stage PBC [57]. A total of 9 out of the 10 patients carrying this allele were in the advanced disease group, suggesting this variant may contribute to rapid disease progression [57]. The frequency of this allele was low in the patient population (3% of CYP2E1 alleles), but may prove to be a useful prognostic indicator. Further study in a much larger patient population and more complete evaluation of variation in the CYP2E1 gene would be of great interest.

Genetic variants in genes with a role in T-cell proliferation were investigated in another Italian study, which included single SNPs in the FOXP3, ICOS and IL2Rα (CD25) genes [41]. No positive associations were noted. However, the variation in these genes was not comprehensively addressed, and so they remain interesting candidates for PBC because of the direct role in immune function.

The X chromosome and PBC

The preponderance of women in the PBC affected population obviously implicates gender derived differences in PBC etiology, but what these may be remains unclear. The traditional concept for the female predominance of autoimmunity is based on a differential effect of sex hormones on the immune system, with androgens favoring the activation of CD8 cells and development of a Th1 response and estrogens promoting a Th2 response through the activation of B-lymphocytes and subsequent induction of antibody production. Microchimerism induced by pregnancy has also been suggested as a possible mechanism contributing to female predominant autoimmunity. This fetal-maternal chimerism has been studied in PBC, but no compelling evidence for a significant involvement with disease has been found [58]. However, fetal microchimerism could influence PBC development in a subset of individuals, possibly in the context of environmental exposure and/or other genetically encoded autoimmune propensity, and thus should not be summarily dismissed.

While differential sex chromosome activity and increased microchimeric exposure provide clues to the reasons that autoimmunity predominates among females, these justifications offer relatively little insight into why some females develop autoimmunity when most do not. To address the frequency of X chromosome monosomy in women with PBC was investigated as immune function genes can be found on the X chromosome, and women with Turner’s syndrome, the congenital loss of one X, often display autoimmune features and cholestatic manifestation [59]. Of interest, PBC patients were found to have an increased frequency of X chromosome monosomy in peripheral blood cells in all tested white blood cell subpopulations, suggesting that instabilty of the X chromosome could be involved in PBC [59]. Further work by this group using quantitative flourescent PCR found X chromosome loss to be preferential in PBC [60]. This study also assessed X chromosome inactivation (XCI), and no significant association between PBC and XCI skewing was identified [60]. Together, these findings suggest that particular X-linked alleles or haplotypes, likely in genes escaping XCI, may predispose to the development of PBC as the result of haploinsufficiency due to the acquired monosomy, providing us with novel candidate regions to search and address in PBC.

The future of PBC genetic research

Rapid advancement in genotyping technologies coupled with decreasing cost has led to a flood of genome-wide studies in a multitude of disease phenotypes, including the first GWAS for PBC. This accomplishment, while certainly of great importance, should not be seen as the pinnacle of genetic investigations into PBC, but as the beginning. Indeed, genome-wide data from thousands of PBC patients will be required in order to truly begin understanding the scope of the genetic impact on and architecture of this enigmatic disease.

We face many challenges if we are to gain insights from endeavors into the genetic architecture of PBC that will lead to clinical benefit for affected patients. Key among these is the rarity and late onset of PBC, which renders the recruitment of patients and their family members into genetic studies a daunting task, and thus the preliminary genome-wide efforts of individual programs will likely lack power to reliably detect risk factors of weaker effect (i.e. OR <1.4). While these initial efforts will undoubtedly be quite informative, a singular international collaboration will eventually be required to maximize power and impact. And even if such cooperation is achieved, we are not likely to amass the tens of thousands of patients necessary to detect the numerous low-impact variants (i.e. OR ~ 1.05–1.15) likely to contribute to PBC. Careful selection of control populations and utilization of publically accessible repositories of genotype and phenotype data will be important to curtail statistical artifacts while maximizing power. And it may prove beneficial to use multiple control groups, including familial controls, considering the limited number of patients ultimately available for study. Functional approaches to elucidate the mechanisms and impact of well-mapped polymorphisms will in the end be required to confirm their relevance. However, in many instances new approaches will be required considering the potential subtlety and complexity of the resulting effects, and early efforts will likely be poorly conceived.

Also challenging is the apparent latency of PBC, which has posed limits on our ability to assess its natural history and monitor the pivotal events early in disease that could form the basis for stratification of the patients for genetic analysis. Indeed, sophisticated sub-classification methods would benefit our efforts to identify the common genetic polymorphisms contributing to variability of disease among patients PBC, and will be vital to future efforts seeking rare genetic contributors to disease, and in the evaluation of pathologic environmental interaction and genetic epistasis. To surmount this challenge, inclusion of at-risk populations, particularly the adult children of PBC patients, in continuing PBC registries will provide a base upon which to draw for future at-risk cohort studies of PBC development. Combined with genotyping, such studies would allow us to develop and test predictive algorithms, initially for PBC onset, but subsequently for treatment response; as well as allowing us to better assess the influence of environmental exposure.

Our understanding of human genetics and genomics has been radically increased over the past decade and continues to expand at an unprecedented pace. Such knowledge holds promise to better understand the etiological and pathogenic bases of PBC leading to improved prognostication, prevention and therapy of this disease.

I am pleased to contribute this work in recognition of Professor Harry Moutsopoulos’ lifetime contributions in the field of autoimmunity and as part of the series that recognizes great contributions to autoimmunologists [61–63]. I am indebted to my beloved medical school Professor Harry Moutsopoulos, for inspiring me to become a clinician-investigator. His unparallel dedication to patient care, teaching and research and his unbridled passion for Immunology, led him to establish and advance the discipline of autoimmunity in Greece, despite scarce resources, for the benefit of patients, students and medical community. Greatly done Professor and well-merited honor. - Kostas Lazaridis.