Association of the HLA locus and TNF with type I autoimmune hepatitis susceptibility in New Zealand Caucasians

Autoimmune hepatitis (AIH) is a chronic progressive inflammatory liver disease of unknown etiology. It has a presumed autoimmune basis to its pathogenesis that is likely to involve a complex interaction of genetic and environmental factors (Czaja et al. 2010). Identification of these factors can potentially provide novel targets for treatment, diagnosis and prevention. AIH can be classified into two types according to circulating auto antibodies present in the patients’ sera. Type 1 AIH is characterized by the presence of anti-nuclear antibodies (ANA) and/or anti smooth muscle antibodies (SMA), whereas type 2 AIH is defined by the presence of anti-liver-kidney microsome antibodies (LKM1) (Czaja et al. 1995).

Type 1 AIH constitutes the vast majority of the disease in Caucasian populations. Genetic predisposition in type 1 AIH has been linked especially to the human leukocyte antigen (HLA) class II haplotype HLA-A1-B8, and the alleles HLA-DRB1*03:01 and HLA-DRB1*04:01 in North America and Europe (Doherty et al. 1994; Strettell et al. 1997). However, distinct differences in the pattern of association between genetic susceptibility and type 1 AIH are seen in populations from various geographical regions. For example, type 1 AIH is associated with HLA-DRB1*04:05 in Japan and Argentina (Seki et al. 1992; Pando et al. 1999), with HLA-DRB1*13:01 in Brazil and Venezuela (Fainboim et al. 1994; Czaja et al. 2002; Fortes Mdel et al. 2007) and with HLA-DRB1*04:04 in Mexico (Vazquez-Garcia et al. 1998). Polymorphisms in genes outside the HLA locus have also been associated with type 1 AIH. An A to G base-exchange polymorphism in exon 1 of the cytotoxic T-lymphocyte antigen 4 (CLTLA4) is associated with increased incidence of AIH in white North American and northern European patients (Agarwal et al. 2000; Djilali-Saiah et al. 2001), but not in South American (Bittencourt et al. 2003) and Japanese patients (Umemura et al. 2008). In young white AIH patients, the tumor necrosis factor (TNF) polymorphism -308G > A is associated with a poorer response to corticosteroid therapy. A Fas gene promoter polymorphism was found to influence susceptibility to AIH and its progression (Hiraide et al. 2005; Agarwal et al. 2007), leading to a more aggressive disease with an early development of cirrhosis (Hiraide et al. 2005). Polymorphisms in the vitamin D receptor gene have also been associated with AIH in German and Chinese patients (Vogel et al. 2002; Fan et al. 2005).

In New Zealand, only 2% of the AIH cases have a positive LKM1 antibody, indicating that type 1 AIH is the predominant disease in this population (Ngu et al. 2010). However, genetic loci previously shown to alter susceptibility to type 1 AIH have yet to be investigated in the New Zealand population. Our study had two aims. The first aim was to test for the association of six candidate risk loci (HLA-A1-B8, HLA-DRB1*03:01, HLA-DRB1*04:01, CTLA4, FAS, and TNF) in a population-based AIH New Zealand Caucasian cohort. The second aim was to examine whether these loci are associated with specific AIH phenotypes such as severe liver fibrosis stage at diagnosis and response to initial immunosuppression.

AIH is a multi-factorial disease that is believed to develop when a genetically susceptible individual is exposed to an environmental factor that triggers the loss of immune tolerance towards hepatocyte antigens. The number and nature of the genes that play a role in AIH development is, at present, poorly defined. To date, the strongest genetic associations for type 1 AIH have been observed with genes located in the major histocompatibility complex (MHC), although several non-MHC genes have also been implicated in type 1 disease including CTLA4, FAS, and TNF. Our study is the first to investigate type 1 AIH susceptibility loci in New Zealand Caucasians.

In our cohort, the most significant association was observed between the HLA-DRB1*03:01 tagging SNP rs2187668 and overall AIH susceptibility, with the minor allele of this SNP conferring a strong risk effect. This finding is consistent with previous reports in European and North American Caucasian populations which have found the HLA-DRB1*03:01 allele to be the principle genetic risk factor for type 1 AIH in these populations. The HLA-DRB1*04:01 allele has also associated, albeit to a lesser degree, with the type 1 AIH in European and North American Caucasians (Czaja 2008). In our cohort, we did not find an association of HLA-DRB1*04:01 with overall susceptibility, despite having 100% power to detect an effect size of 5.97 at α=0.05 (Strettell, Donaldson et al. 1997). We also found that the minor allele of SNP rs1800629, which tags the extended haplotype HLA-A1-B8-DR8, was significantly associated with overall disease risk (p = 0.004, OR = 1.89, 95% CI 1.21-2.94) in New Zealand Caucasians, consistent with findings of previous studies (Donaldson et al. 1991, Al-Chalabi et al. 2008).

In addition to the HLA locus, we tested SNPs within the CTLA4, FAS and TNF genes for association with type I AIH. TNF is a key cytokine in the inflammatory response and variations in expression of TNF have been implicated in many autoimmune diseases. There is evidence suggesting that the promoter polymorphism TNF -308G > A (rs1800629) increases expression of TNF, and an association of this polymorphism with type 1 AIH has been reported in a number of cohorts (Tang et al. 2012). In our cohort of New Zealand Caucasians, the minor allele TNF -308A conferred susceptibility to AIH (OR = 2.06, 95% CI [1.41-3.01], p = 0.0001). In contrast to overall susceptibility, we found no evidence of association of TNF -308G > A with either severe fibrosis at diagnosis or incomplete normalization of ALT at 6 months.

CLTLA4 codes for a T cell surface molecule that plays a key role as a negative regulator of T cell responses by inducing T cell apoptosis on binding to B7 molecules on antigen-presenting cells. CTLA4 49G/G genotype changes the sequence of CTLA4 (threonine to alanine) and leads to diminished inhibitory effects on T cell proliferation, and hence, hyperactivity of T cells (Kouki et al. 2000). Several studies have examined CTLA4 49A > G (rs231775) in AIH although nearly all them have failed to show a definitive association between rs231775 polymorphism and disease susceptibility (Agarwal, Czaja et al. 2000; Bittencourt et al. 2003; Fan et al. 2004; Schott et al. 2007; Umemura et al. 2008). Nevertheless, a recent combined analysis of these studies comprising 526 patients with type 1 AIH and 631 matched controls reported the CLTLA4 49A/A genotype conferred protection against type 1 AIH (OR = 0.66, 95% CI 0.50-0.86) (Miyake et al. 2011). In our cohort we detected no association of the exon 1 SNP CLTLA4 49A > G (rs231775) with overall disease susceptibility. However, we did find an increased risk of severe liver fibrosis at presentation in patients who were homozygous for minor allele of rs231775 (OR = 8.57, 95% CI [1.04-70.89], p = 0.046), although this association was not significant after correction for multiple testing.

FAS is a member of the TNF receptor superfamily of membrane-bound molecules. In vitro studies have shown that FAS plays a prominent role in the induction of hepatocyte apoptosis and tissue destruction in AIH (Fox et al. 2001). Moreover the promoter polymorphism FAS -670 T > C results in higher expression of FAS on the surface of activated T cells (Sun et al. 2005) and, although this polymorphism has not been associated with overall susceptibility to type 1 AIH, the minor allele FAS -670C was found to confer protection against early onset cirrhosis in Caucasian type 1 AIH patients (Agarwal, Czaja et al. 2007). In our study we found that whilst FAS -670A > C was not associated with overall disease susceptibility in our New Zealand Caucasian patients, the FAS -670C allele confers risk of severe fibrosis at diagnosis (OR = 2.03, 95% CI 1.05-3.93, p = 0.035). In addition, the FAS -670C allele was significantly associated with failure to achieve complete normalization of ALT at 6 months post-diagnosis compared to the FAS -670 T allele (OR = 3.94, 95% CI 1.62-9.54, p = 0.0015). Our earlier study had demonstrated that incomplete normalization of ALT at 6 months is an independent predictor of liver related death (Ngu, Gearry et al. 2013). These observations suggest that the presence of the FAS -670C allele is associated with a more aggressive disease phenotype.

Multivariate analysis showed that rs2187668 (HLA-DRB1*03:01) was the only independent risk allele significantly associated with the overall susceptibility to type 1 AIH. This result is of interest as we have shown for the first time that while HLA-A1-B8-DR3 and TNF were associated with type 1 AIH susceptibility, they were also significantly associated with HLA-DRB1*03:01. Our study demonstrates that, under the multivariate model, HLA-DRB1*03:01 was the dominant risk allele.

The strength of the present study is its population-based nature leading to results that are representative of the spectrum of AIH, as opposed to clinic-based studies where patients with more severe phenotypes are often over-represented. The weakness of this study is its small study size, and therefore it may be underpowered in confirming weaker associations. Nevertheless, it is important to note that despite the relatively small sample size, the results presented in this study did show highly significant associations.

Our study is the first population-based study to test for association of selected genetic loci with type 1 AIH and also the first study performed in New Zealand Caucasians. We report significant independent association of HLA-DRB1*03:01 with overall susceptibility to type 1 AIH. FAS, whilst not associated with type 1 AIH susceptibility, was significantly associated with a more aggressive disease phenotype. These findings are an important first step in increasing our understanding of the genetic basis of type 1 AIH, and may help guide the selection of candidates for future functional studies.