Skip to main content
NCBI home page
JHEP Reports logoLink to JHEP Reports
. 2023 Oct 5;6(1):100926. doi: 10.1016/j.jhepr.2023.100926

Fine mapping identifies independent HLA associations in autoimmune hepatitis type 1

You Li 1,†, Lu Zhou 2,†, Zuxiong Huang 3,†, Yue Yang 4,5,†, Jiming Zhang 6,†, Ling Yang 7,†, Yun Xu 8,†, Junping Shi 9,†, Shanhong Tang 10, Xiaoling Yuan 11, Jie Xu 11, Yiling Li 12, Xu Han 13, Jia Li 13, Yanmin Liu 14, Ying Sun 15, Xiaozhi Jin 16, Xiao Xiao 1, Bangmao Wang 2, Qiuxiang Lin 3, Yang Zhou 3, Xuejiao Song 4, Yong Cui 4, Lilin Hu 7, Yuhu Song 7, Jie Bao 8, Ling Gong 9, M Eric Gershwin 17, Xianbo Zuo 4,5,∗,‡, Huiping Yan 14,‡,∗, Zhengsheng Zou 15,‡,∗, Ruqi Tang 1,‡,∗, Xiong Ma 1,18,‡,∗; the Chinese AIH Consortium, on behalf of
PMCID: PMC10711477  PMID: 38089552

Abstract

Background & Aims

Association studies have greatly refined the important role of the major histocompatibility complex (MHC) region in autoimmune hepatitis (AIH). However, the effects of human leucocyte antigen (HLA) polymorphisms on AIH are not well established. The aim of this study is to systematically characterise the association of MHC variants with AIH in our well-defined cohort of patients.

Methods

We performed an imputation-based analysis on the extensive association observed within the MHC region using the Han-MHC reference panel, and tested the comprehensive associations of HLA polymorphisms with AIH in 1622 Chinese AIH type 1 patients and 10,466 population controls.

Results

A total of 588 HLA variants were significantly associated with AIH, with HLA-B∗35:01 (p = 8.17 × 10−304; odds ratio [OR] = 7.32) contributing the strongest signal. Stepwise conditional analysis revealed additional independent signals at HLA-B∗08:01 (p = 1.35 × 10−33; OR = 4.26) and rs7765379 (p = 5.08 × 10−18; OR = 1.66). A strong link between the lead HLA variant and clinical phenotypes of AIH was observed: patients with HLA-B∗35:01 were less frequently positive for ANA and tended to have higher serum AST and ALT levels at diagnosis, but lower serum IgG levels.

Conclusions

Our study reveals three novel and independent variants at HLA-B∗35:01, HLA-B∗08:01, and rs7765379 associated with AIH across the whole MHC region in the Han Chinese population. The findings illustrate the value of the MHC region in AIH and provide a new perspective for the immunogenetics of AIH.

Impact and implications

This study revealed three novel and independent variants associated with autoimmune hepatitis across the whole major histocompatibility complex region in the Han Chinese population. These findings are significant in identifying autoantigens, providing insights into the activation of the autoimmune processes, and further advancing our understanding of the immunogenetic basis underlying autoimmune hepatitis.

Keywords: Autoimmune hepatitis, Human leucocyte antigen, HLA-B∗35:01

Graphical abstract

graphic file with name ga1.jpg

Open in a new tab

Highlights

  • •

    HLA-B∗35:01 conferred the strongest relative genetic risk of autoimmune hepatitis among the Han Chinese population within the MHC region.

  • •

    A significant association between HLA-B35:01 and clinical characteristics of autoimmune hepatitis was identified.

  • •

    Stepwise conditional analysis revealed additional independent signals at HLA-B∗08:01 and rs7765379.

Introduction

Autoimmune hepatitis (AIH) is a rare immune-mediated inflammatory liver disease, characterised by female preponderance, seropositivity of autoantibodies, and interface hepatitis.1 Although the aetiology of AIH remains enigmatic, AIH is presumed to be a multifactorial polygenic condition triggered by environmental exposures in genetically susceptible individuals.2 Candidate gene association studies and large-scale genome-wide association studies (GWASs) have pointed to a strong genetic predisposition. Similar to other autoimmune diseases, the major histocompatibility complex (MHC) region is consistently and strongly associated with AIH in cohorts across different ethnicities. Among the risk human leucocyte antigen (HLA) alleles, HLA-DRB1∗03:01 and HLA-DRB1∗04:01, two of the most frequently described alleles, account for the vast majority of association between the HLA and AIH in European and American populations.3,4 In addition to affecting AIH susceptibility, the HLA alleles are linked to the clinical phenotype of the disease. Patients who are HLA-DRB1∗03:01-positive have higher immunoglobulin G (IgG) levels and are more likely to receive immunosuppressive therapy and liver transplantation; whereas individuals who are HLA-DRB1∗04:01-positive are late-onset, female predominance, and manifest less severe disease.3,5

Although there is extensive literature highlighting the genetic association between the HLA region and AIH, it remains particularly challenging to elucidate the precise role of the HLA region in the pathogenesis of AIH because of several reasons. Firstly, the findings mentioned were primarily derived from candidate gene association studies and did not comprehensively examine polymorphic HLA genes.6 Secondly, most studies only included a limited number of patients, and may lack sufficient statistical power to detect associations. Thirdly, the distribution and frequency of HLA alleles are highly variable across different ethnic groups, whereas trans-ethnic MHC association analyses are limited. Apart from this, the high gene density, extraordinary genetic polymorphism, and striking linkage disequilibrium in the MHC region, make it particularly challenging to define the causative variants and extrapolate to function.7

Recently, we performed a GWAS meta-analysis in 1,622 participants with AIH and 10,466 population controls of Han Chinese origin and confirmed the prominent role of the HLA locus.8 To use these findings to systematically investigate the role of MHC variants in AIH, we conducted MHC imputation using our previous GWAS data. We also performed stepwise regression analyses to explore MHC-independent determinants conferring protective or risk effects on AIH. Furthermore, the associations of the clinical and serological phenotypes of AIH with HLA variants were also investigated.

Patients and method

Study participants and genotyping

In this study, a total of 1,622 patients with AIH type 1 and 10,466 geographically and ethnically matched healthy individuals from 15 clinical centres in China were enrolled. Diagnostic scores were determined according to the 1999 revised International Autoimmune Hepatitis Group diagnostic criteria. Only patients with a pretreatment score of 10 points or more were classified as having AIH and included. Patients were excluded from this study if they had any other cause of chronic liver disease, including primary biliary cholangitis, primary sclerosing cholangitis, Wilson's disease, haemochromatosis, alcoholic liver disease, and chronic viral hepatitis B or C. Patients with a history of recent use of known hepatotoxic drugs were excluded. In addition, patients with positive LKM-1 and/or LC-1 antibodies (AIH type 2) were excluded. The 6,256 genotyped single nucleotide polymorphisms (SNPs) located in the entire MHC region (in the genomic region at 25–35 Mb on chromosome 6, NCBI build 37) were extracted from our previously published paper, as performed on the Infinium Global Screening Array (Illumina, San Diego, CA, USA).8 Details of the cohorts, genotyping methods, and quality control (QC) for samples and genotyped data have been described.8 The study was approved by the institution’s ethics committees and was carried out following the guidelines of the Declaration of Helsinki (2008).

HLA imputation

We defined the HLA variants as SNPs in the MHC region, classical two-digit and four-digit biallelic HLA alleles, biallelic HLA amino acid (AA) polymorphisms corresponding to the respective residues, and multiallelic HLA AA polymorphisms for each AA position. SNPs, classical HLA alleles, and AA variants, were all imputed for each case–control dataset separately in the extended MHC region in chromosome 6. The SNP2HLA software was used for imputation using a Han-MHC reference panel consisting of 29,948 variants for 10,689 Han Chinese individuals.9,10

We applied post-imputation QC criteria of minor allele frequency >0.01, and imputation quality (r2 >0.3). Individuals and SNPs with call rates of <90%, and/or Hardy–Weinberg equilibrium p <1.0 × 10−4 were excluded. After filtering alleles with poor imputation quality, 8,245 high-quality variants, including 6,173 SNPs, 273 alleles (94 two-digit and 179 four-digit) and 1,799 AA polymorphisms for eight HLA genes (HLA-A, HLA-B, HLA-C, HLA-DQA1, HLA-DQB1, HLA-DPA1, HLA-DPB1, and HLA-DRB1) remained.

Association and stepwise regression analysis

For the association study, the biallelic variants were encoded as allele 1 and allele 2, and the multiallelic variants including multiresidue positions and HLA alleles were encoded as the presence or absence of an individual allele. We assumed additive effects of the allele dosages on the log-odds scales for susceptibility to AIH, and evaluated associations of the HLA variants with the risk of AIH using a logistic regression model. To evaluate independent signals among the HLA variants, we conducted a forward-type stepwise conditional regression analysis. Namely, the top variant as a covariate was conditioned in the regression model to find the second variant, and then repeated to find the subsequent variants until no variants satisfied the significance threshold. All analyses were conducted using PLINK v1.9, which was already corrected by sex and the first 10 ancestral principal components.

Transethnic meta-analysis

We utilised 10 previously published association studies that investigated the relationship between HLA and type 1 AIH and had conducted HLA typing at the four-digit level: (1) a UK study comprised 119 cases with AIH and 177 controls in 1994;3 (2) an America study comprised 86 cases with AIH and 102 controls in 1997;4 (3) a Mexico study comprised 30 cases with AIH and 175 controls in 1998;11 (4) an Argentina study comprised 84 cases with AIH and 208 controls in 1999;12 (5) a Brazil study comprised 39 cases with AIH and 22 controls in 2001;13 (6) an India study comprised 20 cases with AIH and 120 controls in 2005;14 (7) a Venezuela study comprised 41 cases with AIH and 111 controls in 2007;15 (8) a Korea study comprised 62 cases with AIH and 154 controls in 2008;16 (9) a Japan study comprised 156 cases with AIH and 201 controls in 2014;17 (10) a Japan study comprised 360 cases with AIH and 1,026 controls in 2017.18 Summary-level statistics from each dataset were combined using fixed-effects meta-analysis, weighting the contribution of each population with the inverse variance method. The analysis was performed using the meta package and visualised with the R software version 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria).

Analysis of clinical phenotypes

Continuous variables were reported as median and IQR, and categorical data were presented as frequencies and percentages. Differences in continuous variables were compared using a two-tailed Student t test or Mann–Whitney U test, where applicable. Categorical variables were assessed using the Χ2 test or Fisher’s exact test as appropriate. Additionally, we used regression analysis to test the association between HLA variants and different clinical phenotypes after multiple corrections. All statistical analyses were performed using the statistical package SPSS 22.0 (SPSS Inc, Chicago, IL, USA) or R software version 4.1.3.

Results

Imputation and association analyses in the MHC region

To gain further insight and reveal extensive association within the MHC region with AIH, we performed imputation and association analyses across the whole MHC region. Association studies revealed that a total of 588 variants from HLA class I and II regions exceeded the genome-wide significance level (p <5 × 10−8). Notably, the most significant association was located in the HLA I region and the strongest allele was in the HLA-B gene (HLA-B∗35:01, p = 8.17 × 10−304) (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Overview of genetic effects in the MHC region influencing risk for AIH.

The analysis was conducted using logistic regression models. The horizontal axis shows genomic position and the vertical axis shows negative log10-transformed p values for association. The horizontal dashed line corresponds to the significance threshold of p = 5 × 10−8. The colour of the circles indicates the type of the marker: grey, SNPs; blue, amino acids; and red, classical HLA alleles. AIH, autoimmune hepatitis; HLA, human leucocyte antigen; MHC, major histocompatibility complex; SNP, single nucleotide polymorphism.

In terms of classical HLA alleles, 13 four-digit HLA alleles exceeded genome-wide significance, including the HLA-A allele, HLA-B alleles, HLA-C alleles, HLA-DQB1 alleles, and HLA-DRB1 alleles (Table 1). As mentioned, the strongest signal was observed at HLA-B∗35:01 (p = 8.17 × 10−304; OR = 7.32; 95% CI, 6.49–8.24). The subsequent significant alleles were HLA-C∗03:03 (p = 5.18 × 10−89; OR = 2.82; 95% CI, 2.54–3.13), HLA-B∗08:01 (p = 1.35 × 10−33; OR = 4.26; 95% CI, 3.30–5.49) and HLA-C∗04:01 (p = 9.75 × 10−17; OR = 1.76; 95% CI, 1.54–2.01). We also identified HLA-DQB1∗05:02, HLA-C∗03:04, HLA-C∗01:02, and HLA-B∗46:01 as the protective HLA alleles; while HLA-DRB1∗03:01, HLA-DQB1∗04:01, HLA-DRB1∗04:05, HLA-A∗03:01, and HLA-DQB1∗06:02 were AIH predisposing. However, no obvious correlations were observed for classical alleles in the HLA-DQA1, DPA1, or DPB1 locus.

Table 1.

Summary of association results between HLA alleles and AIH in a Han Chinese population.

Variant Position (bp)∗ A1/A2 Case Freq_A1 Control Freq_A1 p value OR (95% CI)
HLA-B∗35:01 31431272 P/A 0.184 0.030 8.17 × 10−304 7.32 (6.49–8.24)
HLA-C∗03:03 31346171 P/A 0.174 0.069 5.18 × 10−89 2.82 (2.54–3.13)
HLA-B∗08:01 31431272 P/A 0.031 0.007 1.35 × 10−33 4.26 (3.30–5.49)
HLA-C∗04:01 31346171 P/A 0.089 0.053 9.75 × 10−17 1.76 (1.54–2.01)
HLA-DRB1∗03:01 32660042 P/A 0.087 0.054 2.53 × 10−14 1.69 (1.47–1.94)
HLA-DQB1∗05:02 32739039 P/A 0.041 0.077 2.81 × 10−13 0.52 (0.43–0.62)
HLA-DQB1∗04:01 32739039 P/A 0.075 0.046 2.76 × 10−12 1.68 (1.45–1.94)
HLA-DRB1∗04:05 32660042 P/A 0.083 0.054 5.03 × 10−11 1.59 (1.38–1.82)
HLA-A∗03:01 30019970 P/A 0.059 0.036 8.89 × 10−11 1.70 (1.45–2.00)
HLA-C∗03:04 31346171 P/A 0.072 0.107 7.21 × 10−10 0.65 (0.56–0.74)
HLA-C∗01:02 31346171 P/A 0.128 0.171 9.54 × 10−10 0.71 (0.64–0.79)
HLA-B∗46:01 31431272 P/A 0.078 0.112 2.08 × 10−9 0.66 (0.58–0.76)
HLA-DQB1∗06:02 32739039 P/A 0.108 0.079 2.84 × 10−8 1.41 (1.25–1.59)
Open in a new tab

A, absent; A1, effective allele; A2, alternative allele; AIH, autoimmune hepatitis; OR, odds ratio; P, present.

∗

According to GRCh37/hg19 assembly.

In terms of SNPs, 456 SNPs within the MHC region were associated with AIH (Table S1). The most significant association was found at rs2243621 (p = 9.21 × 10−163; OR = 3.76; 95% CI, 3.40–4.17), which was located within the gene of HCP5 and 106 kb downstream of HLA-B.

At the AA level, 108 AA polymorphisms exceeded study-wide significance (Table S2). The most significant association was observed at position 67 with the AA phenylalanine of the HLA-B molecule (p = 3.39 × 10−139; OR = 2.95; 95% CI, 2.70–3.22). This specific location, known as Phe67, is located at the peptide binding groove of HLA-B, indicating its functional contributions to antigen-presentation ability. The strong predisposing effect was also observed in AA residues of HLA-C, DQA1, DQB1, and DRB1.

Transethnic meta-analysis of the reported HLA alleles

To further extend the above findings and explore the ethnic heterogeneity of genetic association, we performed trans-ethnic meta-analyses of the reported classical HLA alleles including published datasets from different ethnicities (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Transethnic meta-analysis of the effects of the reported HLA alleles.

Forest plot shows the estimated odds ratio (OR) with 95% CI of each cohort. The relative weight of each study in the meta-analysis is represented by the size of the squares. The solid black line represents OR = 1 and the red dashed line indicates OR from the fixed-effect meta-analysis.

In the previous studies, HLA-DRB1∗03:01 and HLA-DRB1∗04:01 alleles accounted for the vast majority of associations in European and American populations, with additional effects from HLA-DRB1∗04:05, HLA-DRB1∗13:01, HLA-DRB1∗13:02, HLA-DQB1∗04:01, and HLA-DQB1∗06:02.3,4,[11], [12], [13], [14], [15], [16], [17], [18] Conversely, the HLA-DRB1∗15:01 allele exhibited a protective effect.14,17 In the current study, HLA-DRB1∗03:01 (p = 2.53 × 10−14), HLA-DRB1∗04:05 (p = 5.03 × 10−11), HLA-DQB1∗04:01 (p = 2.76 × 10−12) and HLA-DQB1∗06:02 (p = 2.84 × 10−8) reached the genome-wide association threshold, whereas HLA-DRB1∗13:02 (p = 3.48 × 10−6) and HLA-DRB1∗04:01 (p = 7.68 × 10−5) only demonstrated moderate evidence for association with AIH in our Han Chinese cohorts. In the meta-analysis, the signal of HLA-DRB1∗03:01 (p <0.0001; OR = 2.13; 95% CI, 1.82–2.51), HLA-DRB1∗04:01 (p <0.0001; OR = 2.56; 95% CI, 1.97–3.34), HLA-DRB1∗04:05 (p <0.0001; OR = 2.23; 95% CI, 1.93–2.58), HLA-DRB1∗13:01 (p <0.0001; OR = 1.94; 95% CI, 1.39–2.71), HLA-DRB1∗13:02 (p <0.0001; OR = 0.48; 95% CI, 0.37–0.64) and HLA-DQB1∗04:01 (p <0.0001; OR = 1.90; 95% CI, 1.58–2.29) were sustained, whereas the associations for HLA-DRB1∗15:01 (p = 0.9901; OR = 1.00; 95% CI, 0.87–1.15) and HLA-DQB1∗06:02 (p = 0.2345; OR=1.10; 95% CI, 0.94–1.29) were mainly abolished.

In addition to MHC II, HLA-C∗01:02 was the only reported HLA allele in the MHC I genes. This allele demonstrated a strong association in Indian and Japanese populations, which was confirmed in our study (p = 9.54 × 10−10). In the meta-analysis, HLA-C∗01:02 still demonstrated an association signal (p = 0.0005; OR = 0.77; 95% CI, 0.67–0.89).

Stepwise logistic regression of HLA variants

Considering the tight linkage disequilibrium (LD) between HLA alleles, we performed reciprocal conditional analyses using logistic regression in the combined dataset, aiming to elucidate the variants in the MHC region exerting an independent effect on the disease.

We first conditioned on the top associated-variant HLA-B∗35:01 and the signals in the MHC region were attenuated (Fig. 3A); and HLA-C∗03:03, the second strongest allele in the association analysis, no longer demonstrated significant association (p = 2.99 × 10−7). Nevertheless, we identified a significant independent association at HLA-B∗08:01 (p = 7.15 × 10−37). When we conditioned on these two risk variants (HLA-B∗35:01 and HLA-B∗08:01), we observed another significant independent AIH-associated signal rs7765379 (p = 5.88 × 10−23), located between HLA-DQB1 and HLA-DQA2 (Fig. 3B). Conditioning on HLA-B∗35:01, HLA-B∗08:01 and rs7765379 revealed the fourth independent associated signal at position 73 with the AA glycine of the HLA-DRB1 molecule (p = 1.16 × 10−11) (Fig. 3C). However, HLA-DRB1-73G did not reach genome-wide significance in the association stage (p = 1.49 × 10−6). After conditioning on all the above variants, no variants in the MHC region satisfied the genome-wide significance (Fig. 3D). In combination, three independent signals (HLA-B∗35:01, HLA-B∗08:01, and rs7765379) accounted for most of the associations with AIH.

Fig. 3.

Fig. 3

Open in a new tab

Stepwise logistic regression of the variants for AIH in the MHC region.

(A) Conditioned on HLA-B∗35:01, (B) conditioned on HLA-B∗35:01 and HLA-B∗08:01, (C) conditioned on HLA-B∗35:01, HLA-B∗08:01 and rs7765379, (D) conditioned on HLA-B∗35:01, HLA-B∗08:01, rs7765379, and HLA-DRB1-73G. The analysis was conducted using logistic regression models. For each plot, the horizontal axis shows genomic position and the vertical axis shows negative log10-transformed p values for association. The horizontal dashed line corresponds to the significance threshold of p = 5 × 10−8. The colour of the circles indicates the type of the marker: grey, SNPs; blue, amino acids; and red, classical HLA alleles. HLA, human leucocyte antigen; SNP, single nucleotide polymorphism.

Association with clinical phenotypes

We next explored whether HLA variants influenced the clinical features of AIH (Table S3). First, a total of 8,425 HLA variants were analysed between 1,396 ANA-positive and 226 ANA-negative subjects. No variants achieved genome-wide significance (Fig. S1A), and the most significant association was found at rs16899646 (p = 2.35 × 10−5; OR = 0.63; 95% CI, 0.50–0.78). As for HLA alleles, HLA-B∗35:01 (p = 7.44 × 10−5; OR = 0.60; 95% CI, 0.47–0.77) and HLA-DPB1∗04 (p = 7.52 × 10−5; OR = 1.96; 95% CI, 1.42–2.77) demonstrated the strongest signal. Then we explored the association between HLA variants and smooth muscle antibody (SMA) positivity (Fig. S1B). The strongest association was found at rs2523534 (p = 5.03 × 10−5; OR = 0.72; 95% CI, 0.61–0.84) and HLA-DRB1∗04:05 (p = 9.80 × 10−4; OR = 1.59; 95% CI, 1.20–2.08) demonstrated the strongest signal among HLA alleles. We also tested HLA variants for association with serum levels of aspartate transaminase (AST), alanine transaminase (ALT), and IgG. Two HLA variants were associated with AST levels. HLA-B∗35:01 (p = 8.88 × 10−9; estimate = 104.11) demonstrated the strongest association, and the subsequent significant variant was HLA-B-95I (p = 1.37 × 10−8; estimate = 78.52) (Fig. 4A). As for ALT levels, only HLA-B∗35:01 reached the genome-wide significance (p = 1.59 × 10−8; estimate = 121.57) (Fig. 4B). Notably, a total of 74 variants were found to be associated with IgG levels, including 53 SNPs, 4 HLA alleles (2 two-digit and 2 four-digit) and 17 AA polymorphisms (Fig. 4C). Among them, the most significant association was found at HLA-B∗35:01 (p = 8.93 × 10−25; estimate = -4.01).

Fig. 4.

Fig. 4

Open in a new tab

Summary of association results between HLA variants and clinical phenotypes in the AIH cohort.

The analysis was conducted using linear regression models. Plots of the association between HLA variants with serum levels of AST (A), ALT (B) and IgG (C). For each plot, the horizontal axis shows genomic position and the vertical axis shows negative log10-transformed p values for association. The horizontal dashed line corresponds to the significance threshold of p = 5 × 10−8. The colour of the circles indicates the type of the marker: grey, SNPs; blue, amino acids; and red, classical HLA alleles. AIH, autoimmune hepatitis; ALT, alanine transaminase; AST, aspartate transaminase; HLA, human leucocyte antigen; IgG, immunoglobulin G; SNP, single nucleotide polymorphism.

We further reviewed the charts of 545 consecutive patients who had received the same treatment regimen. These patients were subsequently categorised into two groups based on their response to immunosuppressive therapy and 449 patients (82.4%) achieved complete biochemical response. We then performed logistic regression analysis to explore the potential relationship between HLA variants and treatment response. Intriguingly, our analysis discovered that HLA-B∗35:01 (p = 2.35 × 10−5; OR = 2.92; 95% CI, 1.87–4.73) exhibited the strongest correlation (Fig. S1C).

We divided AIH patients according to positivity for HLA-B∗35:01 to further evaluate the impact of the HLA-B∗35:01 on clinical features. Comparisons of clinical and laboratory features of each group of patients are illustrated in Table 2. Compared with patients without HLA-B∗35:01, patients with HLA-B∗35:01 were less frequently positive for ANA (80.9% vs. 88.9%, p <0.001). Strikingly, patients with HLA-B∗35:01 had significantly higher serum AST (279 vs. 156 U/L, p <0.001) and ALT (208 vs. 141 U/L, p = 0.007) at diagnosis, whereas significantly lower serum IgG levels were observed in these patients (18.5 vs. 22.0 U/L, p <0.001).

Table 2.

Comparison of HLA-B∗35:01 positive and negative AIH patients.

Feature HLA-B∗35:01 positive (n = 566) HLA-B∗35:01 negative (n = 1056) p value
Female (%) 501 (88.5) 890 (84.3) 0.020
Age at diagnosis (years) 54 (47–61) 56 (45–65) 0.476
ANA positive (%) 458 (80.9) 939 (88.9) <0.001
SMA positive (%) 166 (29.3) 284 (26.9) 0.296
AST (U/L) 279 (117–545) 156 (85–348) <0.001
ALT (U/L) 208 (100–498) 141 (67–318) 0.007
ALP (U/L) 141 (106–184) 141 (107–201) 0.065
γ-GT(U/L) 117 (64–207) 121 (64–208) 0.260
Bilirubin (μmol/L) 29 (17–79) 33 (19–87) 0.515
IgG (g/L) 18.5 (15.3–22.0) 22.0 (18.4–28.1) <0.001
Open in a new tab

Values are expressed as median (interquartile range) unless otherwise indicated.

γ-GT, gamma-glutamyl transpeptidase; AIH, autoimmune hepatitis; ALP, alkaline phosphatase; ALT, alanine transaminase; ANA, antinuclear antibody; AST, aspartate transaminase; IgG, immunoglobulin G; SMA, smooth muscle antibody.

We also investigated the impact of the HLA-B∗08:01 on clinical features (Table S4). HLA-B∗08:01-positive patients were more likely to have older age at disease onset (60 vs. 56 years, p = 0.007). In addition, patients with HLA-B∗08:01 tended to have higher serum IgG (23.0 vs. 20.6 U/L, p = 0.005).

Discussion

In the present study, we revealed three novel and independent variants at HLA-B∗35:01, HLA-B∗08:01, and rs7765379 associated with AIH across the whole MHC region in the Han Chinese population.

As the most polymorphic region of the human genome, HLA genes encode molecules that form stable complexes on binding to foreign peptides, playing a major evolutionary role in the regulation of both cellular and innate immunity.19 It is increasingly recognised that genetic variations in the HLA region alter the shape of the peptide-binding pocket in HLA molecules, potentially leading to dysregulation of antigen presentation and T-cell activation.20 Thus, understanding the role of HLA in AIH may help to expand our understanding of the immune genetic pathogenesis of the disease.

Our data showed that the HLA-B locus conferred the strongest relative genetic risk of AIH in the Han Chinese population, different from previous studies which reported MHC class-II haplotypes conferring the major HLA risk factor to AIH.3,4,21,22 These observations reflect the pathogenesis of patients with AIH among different ethnic populations may be heterogeneous. In general, AIH is a chronic inflammatory disease characterised by a T cell-mediated immune response targeting liver autoantigens.23 Given that human MHC class I molecules are expressed by almost all cells and ligands for the αβ receptors of CD8+ T cells, our findings underscore the critical role of CD8+ T cells in the pathogenesis of AIH.24 Interestingly, we have previously demonstrated the elevation of tissue-resident CD8+ T (CD8+ TRM) cells in the liver of patients with AIH, which correlated with biochemical and histological activity.25 Therefore, it would be helpful to examine the interaction between these risk HLA variants and CD8+ TRM cells in AIH.

HLA-B∗35:01 has been reported to present peptides derived from Epstein–Barr virus (EBV), cytomegalovirus (CMV), HIV and HCV.26 Consistently, candidate environmental factors, like infection by several viruses, including HCV, CMV, and EBV, are thought to contribute to predisposition to AIH.27 Interestingly, earlier studies have found that HLA-B∗35:01 is associated with drug-induced liver injury (DILI), including Polygonum multiflorum-induced liver injury, trimethoprim sulfamethoxazole-induced liver injury and green tea-induced liver injury.[28], [29], [30] These data indicated overlapping genetic and pathogenetic mechanisms between AIH and DILI, possibly involving specific similarities in HLA-restricted antigen recognition and activation of autoreactive T cells. In addition, Geng et al.31 found that peptide-deficient HLA-B∗35:01 tetramers could directly enhance CD8+ T-cell activation, implying a strong T-cell activation ability of HLA-B∗35:01.31 Further investigation is needed to identify and validate this hypothesis.

Our study also demonstrated a strong link between HLA-B∗35:01 and clinical phenotypes of AIH. We observed a significantly lower frequency of HLA-B∗35:01 allele in ANA-positive patients with AIH in our cohorts. Zhang et al. also provided evidence that the frequency of HLA-B∗35:01 allele was significantly higher in anti-soluble liver antigen/liver pancreas (anti-SLA/LP)-positive AIH patients than normal patients.32 Given that ANA are detected in 80% of White North American adults with AIH at presentation, SMA are present in 63%, and anti-SLA/LP are present in 7–22%, more extensive studies will be necessary to identify autoantibodies with a higher sensitivity and specificity than the conventional autoantibodies used for the diagnosis of AIH.33,34 Based on this attribution, definition of the specific peptides and their molecular interactions with HLA-B∗35:01 in AIH warrants further studies. In addition, patients who were HLA-B∗35:01-positive had higher levels of AST and ALT, but a lower level of IgG, highlighting that T cell-mediated immune responses may predominate in these patients.

In addition to HLA-B∗35:01, HLA-B∗08:01 and rs7765379 were identified as independent signals for AIH. HLA-B∗08:01 is related to several autoimmune diseases, such as psoriatic arthritis, myasthenia gravis, and polymyositis.[35], [36], [37] Rs7765379 is located in the HLA class II region, in proximity to HLA-DQB1 and HLA-DQA2. This SNP has been associated with Crohn’s disease, enteric fever, and ulcerative colitis.[38], [39], [40] Our data also revealed a strong association between AIH and the presence of a Phe residue at position 67 of HLA-B alleles. Notably, Phe67 is situated within the B pocket and accommodates the P2 anchor residue of HLA-bound peptides, which is shared by almost all B∗35 and B∗08 alleles.41,42 In this context, analysis of peptide pools eluted from patients with AIH with HLA-B∗35:01 or B∗08:01 alleles might provide valuable insights into the nature of the triggering antigen. In conclusion, as the largest fine-mapping analysis of the MHC region, the current study revealed three independent HLA signals implicated in the susceptibility to AIH. Our data advance our understanding of the genetic basis of AIH by identifying SNPs, alleles and amino acid residues conferring disease susceptibility in the MHC region in the Han Chinese population, and provides strong support for the postulate that specific AAs in specific antigen-binding pockets are critical to the aetiopathogenesis of this disease. Further study in independent samples will be needed to confirm these multiple associations within the MHC/HLA region.

Financial support

This work was supported by the National Natural Science Foundation of China grants (#82130017, 81830016, and 81771732 to XM, #81922010, 81873561, and 81570469 to RT, 82273523 to XZ, 82070583 to XX), Innovative research team of high-level local universities in Shanghai (#SHSMU-ZDCX20210301 to XM), Shanghai Municipal Science and Technology Committee of Shanghai outstanding academic leaders plan (#20XD1422500 to RT), Shanghai Municipal Education Commission-Gaofeng Clinical Medicine Grant Support (#20161311 to RT), Shanghai Sailing Program (#23YF1423300 to YL), China Postdoctoral Science Foundation (#2023M732296 to YL), China National Key R&D Program of China (#2022YFC3602002 to XZ), Capital’s Funds for Health Improvement and Research (#CFH2020-1-2031 to XZ) and National High Level Hospital Clinical Research Funding (#2022-NHLHCRF-LX-03 and 2023-NHLHCRF-YXHZ-ZRZD-06 to XZ).

Authors’ contributions

Conceptualised and supervised the study: XM, RT, ZZ, HY, XZ. Acquired funding: XM, RT, XZ, XX, YL. Collected samples and clinical information: ST, XY, JX, YL, XH, JL, YL, YS, XJ, BW, QL, YZ, XS, YC, LH, SY, JB, LG. Performed experiments: YL, LZ, ZH, YY, JZ, LY, YX, JS. Carried out the analyses: YL, XZ. Drafted the manuscript: YL, RT. Critically reviewed and revised the manuscript: XM, XZ, MEG. Approved the final version of the manuscript: all authors.

Data availability statement

The genotyping data have been deposited in the China National Genomics Data Center under accession code OMIX004314 (https://ngdc.cncb.ac.cn/). All other remaining data are available from the corresponding author upon reasonable request.

Conflicts of interest

The authors have no conflicting financial interests.

Please refer to the accompanying ICMJE disclosure forms for further details.

Acknowledgements

We are grateful to all participating members of Chinese AIH Consortium for providing samples and clinical information. We are also grateful to all the patients who took part in this study.

Footnotes

Author names in bold designate shared co-first authorship

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jhepr.2023.100926.

Contributor Information

Xianbo Zuo, Email: zuoxianbo@qq.com.

Huiping Yan, Email: bjyhp503@ccmu.edu.cn.

Zhengsheng Zou, Email: zszou302@163.com.

Ruqi Tang, Email: ruqi.tang@gmail.com.

Xiong Ma, Email: maxiongmd@hotmail.com.

Supplementary data

The following are the supplementary data to this article:

Multimedia component 1
mmc1.pdf (754.9KB, pdf)
Multimedia component 2

:

mmc2.docx (31KB, docx)
Multimedia component 3
mmc3.pdf (4MB, pdf)
Multimedia component 4
mmc4.xlsx (56.6KB, xlsx)
Multimedia component 5
mmc5.pdf (3.2MB, pdf)

References

  • 1.Webb G.J., Hirschfield G.M., Krawitt E.L., et al. Cellular and molecular mechanisms of autoimmune hepatitis. Annu Rev Pathol. 2018;13:247–292. doi: 10.1146/annurev-pathol-020117-043534. [DOI] [PubMed] [Google Scholar]
  • 2.Terziroli Beretta-Piccoli B., Mieli-Vergani G., Vergani D. Autoimmmune hepatitis. Cell Mol Immunol. 2022;19:158–176. doi: 10.1038/s41423-021-00768-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Doherty D.G., Donaldson P.T., Underhill J.A., et al. Allelic sequence variation in the HLA class II genes and proteins in patients with autoimmune hepatitis. Hepatology. 1994;19:609–615. doi: 10.1002/hep.1840190311. [DOI] [PubMed] [Google Scholar]
  • 4.Strettell M.D., Donaldson P.T., Thomson L.J., et al. Allelic basis for HLA-encoded susceptibility to type 1 autoimmune hepatitis. Gastroenterology. 1997;112:2028–2035. doi: 10.1053/gast.1997.v112.pm9178696. [DOI] [PubMed] [Google Scholar]
  • 5.van Gerven N.M., de Boer Y.S., Zwiers A., et al. HLA-DRB1∗03:01 and HLA-DRB1∗04:01 modify the presentation and outcome in autoimmune hepatitis type-1. Genes Immun. 2015;16:247–252. doi: 10.1038/gene.2014.82. [DOI] [PubMed] [Google Scholar]
  • 6.Higuchi T., Oka S., Furukawa H., et al. Genetic risk factors for autoimmune hepatitis: implications for phenotypic heterogeneity and biomarkers for drug response. Hum Genomics. 2021;15:6. doi: 10.1186/s40246-020-00301-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.de Bakker P.I.W., McVean G., Sabeti P.C., et al. A high-resolution HLA and SNP haplotype map for disease association studies in the extended human MHC. Nat Genet. 2006;38:1166–1172. doi: 10.1038/ng1885. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Li Y., Sun Y., Liu Y., et al. Genome-wide meta-analysis identifies susceptibility loci for autoimmune hepatitis type 1. Hepatology. 2022;76:564–575. doi: 10.1002/hep.32417. [DOI] [PubMed] [Google Scholar]
  • 9.Zhou F., Cao H., Zuo X., et al. Deep sequencing of the MHC region in the Chinese population contributes to studies of complex disease. Nat Genet. 2016;48:740–746. doi: 10.1038/ng.3576. [DOI] [PubMed] [Google Scholar]
  • 10.Jia X., Han B., Onengut-Gumuscu S., et al. Imputing amino acid polymorphisms in human leukocyte antigens. PLoS One. 2013;8 doi: 10.1371/journal.pone.0064683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Vazquez-Garcia M.N., Alaez C., Olivo A., et al. MHC class II sequences of susceptibility and protection in Mexicans with autoimmune hepatitis. J Hepatol. 1998;28:985–990. doi: 10.1016/s0168-8278(98)80347-4. [DOI] [PubMed] [Google Scholar]
  • 12.Pando M., Larriba J., Fernandez G.C., et al. Pediatric and adult forms of type I autoimmune hepatitis in Argentina: evidence for differential genetic predisposition. Hepatology. 1999;30:1374–1380. doi: 10.1002/hep.510300611. [DOI] [PubMed] [Google Scholar]
  • 13.Goldberg A.C., Bittencourt P.L., Mougin B., et al. Analysis of HLA haplotypes in autoimmune hepatitis type 1: identifying the major susceptibility locus. Hum Immunol. 2001;62:165–169. doi: 10.1016/s0198-8859(00)00234-2. [DOI] [PubMed] [Google Scholar]
  • 14.Shankarkumar U., Amarapurkar D.N., Kankonkar S. Human leukocyte antigen allele associations in type-1 autoimmune hepatitis patients from western India. J Gastroenterol Hepatol. 2005;20:193–197. doi: 10.1111/j.1440-1746.2004.03608.x. [DOI] [PubMed] [Google Scholar]
  • 15.del Pilar Fortes M., Machado I.V., Gil G., et al. Genetic contribution of major histocompatibility complex class II region to type 1 autoimmune hepatitis susceptibility in Venezuela. Liver Int. 2007;27:1409–1416. doi: 10.1111/j.1478-3231.2007.01581.x. [DOI] [PubMed] [Google Scholar]
  • 16.Lim Y.S., Oh H.B., Choi S.E., et al. Susceptibility to type 1 autoimmune hepatitis is associated with shared amino acid sequences at positions 70-74 of the HLA-DRB1 molecule. J Hepatol. 2008;48:133–139. doi: 10.1016/j.jhep.2007.08.019. [DOI] [PubMed] [Google Scholar]
  • 17.Umemura T., Katsuyama Y., Yoshizawa K., et al. Human leukocyte antigen class II haplotypes affect clinical characteristics and progression of type 1 autoimmune hepatitis in Japan. PLoS One. 2014;9 doi: 10.1371/journal.pone.0100565. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Oka S., Furukawa H., Yasunami M., et al. HLA-DRB1 and DQB1 alleles in Japanese type 1 autoimmune hepatitis: the predisposing role of the DR4/DR8 heterozygous genotype. PLoS One. 2017;12 doi: 10.1371/journal.pone.0187325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Dendrou C.A., Petersen J., Rossjohn J., et al. HLA variation and disease. Nat RevImmunol. 2018;18:325–339. doi: 10.1038/nri.2017.143. [DOI] [PubMed] [Google Scholar]
  • 20.Matzaraki V., Kumar V., Wijmenga C., et al. The MHC locus and genetic susceptibility to autoimmune and infectious diseases. Genome Biol. 2017;18:76. doi: 10.1186/s13059-017-1207-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Czaja A.J., Souto E.O., Bittencourt P.L., et al. Clinical distinctions and pathogenic implications of type 1 autoimmune hepatitis in Brazil and the United States. J Hepatol. 2002;37:302–308. doi: 10.1016/s0168-8278(02)00182-4. [DOI] [PubMed] [Google Scholar]
  • 22.Yoshizawa K., Ota M., Katsuyama Y., et al. Genetic analysis of the HLA region of Japanese patients with type 1 autoimmune hepatitis. J Hepatol. 2005;42:578–584. doi: 10.1016/j.jhep.2004.12.019. [DOI] [PubMed] [Google Scholar]
  • 23.Liberal R., Krawitt E.L., Vierling J.M., et al. Cutting edge issues in autoimmune hepatitis. J Autoimmun. 2016;75:6–19. doi: 10.1016/j.jaut.2016.07.005. [DOI] [PubMed] [Google Scholar]
  • 24.Pishesha N., Harmand T.J., Ploegh H.L. A guide to antigen processing and presentation. Nat Rev Immunol. 2022;22:751–764. doi: 10.1038/s41577-022-00707-2. [DOI] [PubMed] [Google Scholar]
  • 25.You Z., Li Y., Wang Q., et al. The clinical significance of hepatic CD69(+) CD103(+) CD8(+) resident-memory T cells in autoimmune hepatitis. Hepatology. 2021;74:847–863. doi: 10.1002/hep.31739. [DOI] [PubMed] [Google Scholar]
  • 26.Vita R., Mahajan S., Overton J.A., et al. The immune Epitope database (IEDB): 2018 update. Nucleic Acids Res. 2019;47:D339–D343. doi: 10.1093/nar/gky1006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Heneghan M.A., Yeoman A.D., Verma S., et al. Autoimmune Hepatitis Lancet. 2013;382:1433–1444. doi: 10.1016/S0140-6736(12)62163-1. [DOI] [PubMed] [Google Scholar]
  • 28.Li C., Rao T., Chen X., et al. HLA-B∗35:01 allele is a potential biomarker for predicting Polygonum multiflorum-induced liver injury in humans. Hepatology. 2019;70:346–357. doi: 10.1002/hep.30660. [DOI] [PubMed] [Google Scholar]
  • 29.Li Y.-J., Phillips E.J., Dellinger A., et al. Human leukocyte antigen B∗14:01 and B∗35:01 are associated with trimethoprim-sulfamethoxazole induced liver injury. Hepatology. 2021;73:268–281. doi: 10.1002/hep.31258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Hoofnagle J.H., Bonkovsky H.L., Phillips E.J., et al. HLA-B∗35:01 and green tea-induced liver injury. Hepatology. 2021;73:2484–2493. doi: 10.1002/hep.31538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Geng J., Altman J.D., Krishnakumar S., et al. Empty conformers of HLA-B preferentially bind CD8 and regulate CD8 T cell function. ELife. 2018;7 doi: 10.7554/eLife.36341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Zhang H.P., Liu Y.M., Li Z., et al. Clinical characteristics and HLA genotypes in Chinese patients with anti-SLA/LP-positive autoimmune hepatitis. Ann Transl Med. 2021;9:153. doi: 10.21037/atm-20-8036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Mack C.L., Adams D., Assis D.N., et al. Diagnosis and management of autoimmune hepatitis in adults and children: 2019 practice guidance and guidelines from the American Association for the Study of Liver Diseases. Hepatology. 2020;72:671–722. doi: 10.1002/hep.31065. [DOI] [PubMed] [Google Scholar]
  • 34.Zhang W.-C., Zhao F.-R., Chen J., et al. Meta-analysis: diagnostic accuracy of antinuclear antibodies, smooth muscle antibodies and antibodies to a soluble liver antigen/liver pancreas in autoimmune hepatitis. PLoS One. 2014;9 doi: 10.1371/journal.pone.0092267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Winchester R., Giles J., Jadon D., et al. Implications of the diversity of class I HLA associations in psoriatic arthritis. Clin Immunol. 2016;172:29–33. doi: 10.1016/j.clim.2016.07.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Saruhan-Direskeneli G., Hughes T., Yilmaz V., et al. Genetic heterogeneity within the HLA region in three distinct clinical subgroups of myasthenia gravis. Clin Immunol. 2016;166–167:81–88. doi: 10.1016/j.clim.2016.05.003. [DOI] [PubMed] [Google Scholar]
  • 37.Rothwell S., Cooper R.G., Lundberg I.E., et al. Dense genotyping of immune-related loci in idiopathic inflammatory myopathies confirms HLA alleles as the strongest genetic risk factor and suggests different genetic background for major clinical subgroups. Ann Rheum Dis. 2016;75:1558–1566. doi: 10.1136/annrheumdis-2015-208119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Yamazaki K., Umeno J., Takahashi A., et al. A genome-wide association study identifies 2 susceptibility Loci for Crohn’s disease in a Japanese population. Gastroenterology. 2013;144:781–788. doi: 10.1053/j.gastro.2012.12.021. [DOI] [PubMed] [Google Scholar]
  • 39.Dunstan S.J., Hue N.T., Han B., et al. Variation at HLA-DRB1 is associated with resistance to enteric fever. Nat Genet. 2014;46:1333–1336. doi: 10.1038/ng.3143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Okada S., Hata K., Shinagawa T., et al. A polymorphism in interleukin-1β gene is associated with the development of pouchitis in Japanese patients with ulcerative colitis. Digestion. 2021;102:489–498. doi: 10.1159/000503283. [DOI] [PubMed] [Google Scholar]
  • 41.Hildebrand W.H., Domena J.D., Shen S.Y., et al. HLA-B15: a widespread and diverse family of HLA-B alleles. Tissue Antigens. 1994;43:209–218. doi: 10.1111/j.1399-0039.1994.tb02327.x. [DOI] [PubMed] [Google Scholar]
  • 42.Chelvanayagam G. A roadmap for HLA-A, HLA-B, and HLA-C peptide binding specificities. Immunogenetics. 1996;45:15–26. doi: 10.1007/s002510050162. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.pdf (754.9KB, pdf)
Multimedia component 2

:

mmc2.docx (31KB, docx)
Multimedia component 3
mmc3.pdf (4MB, pdf)
Multimedia component 4
mmc4.xlsx (56.6KB, xlsx)
Multimedia component 5
mmc5.pdf (3.2MB, pdf)

Data Availability Statement

The genotyping data have been deposited in the China National Genomics Data Center under accession code OMIX004314 (https://ngdc.cncb.ac.cn/). All other remaining data are available from the corresponding author upon reasonable request.

Articles from JHEP Reports are provided here courtesy of Elsevier

RESOURCES