Fine-mapping analysis of the MHC region for vitiligo based on a new Han-MHC reference panel

Highlights

• This is the first MHC fine-mapping study for vitiligo in the Han Chinese population using a large-scale reference panel.

• HLA-DQβ1 amino acid position 135 and HLA-B amino acid positions 45–46 were prioritized for the risk of vitiligo.

• In total, 8.60% of phenotypic variance in vitiligo was explained by identified variants.

Abstract

Vitiligo is an immune-related disease with patchy depigmentation of skin and hair caused by selective destruction of melanocytes. In recent decades, many studies have shown the association between vitiligo and HLA genes; however, the results of Han Chinese are scarce. In this study, we performed a fine-mapping analysis of the MHC region in 2818 Han Chinese subjects through a widely used HLA imputation method with a newly built large-scale Han-MHC reference panel. Three new four-digit HLA alleles (HLA-DQB1 ∗ 02:02, HLA-DQA1 ∗ 02:01 and HLA-DPB1 ∗ 17:01) were identified to be associated with the risk of vitiligo, and four previously reported alleles were confirmed. Further conditional analysis revealed that two important variants, HLA-DQβ1 amino acid position 135 (OR = 1.79, P = 1.87 × 10⁻¹¹) and HLA-B amino acid positions 45–46 (OR = 1.44, P = 5.61 × 10⁻¹¹), conferred most of the MHC associations. Three-dimension ribbon models showed that the former is located within the β2 domain of the HLA-DQβ1 molecule, and the latter lies in the α1 domain of the HLA-B molecule, while both are involved in specific antigen presenting process. Finally, we summarized all significant signals in the MHC region to clarify their complex relationships, and 8.60% of phenotypic variance could be explained based on all reported variants in Han Chinese so far. Our findings highlight the complex genetic architecture of the MHC region for vitiligo in Han Chinese population and expand our understanding of the roles of HLA coding variants in the etiology of vitiligo.

Introduction

Vitiligo is an autoimmune disorder characterized by patchy depigmentation of skin and hair resulting from progressive loss of melanocytes, with an estimated prevalence ranging from 0.5% to 1% in most populations worldwide (Ezzedine et al., 2012). Various theories have been proposed for the explanation of the dysfunction and loss of melanocytes in vitiligo, such as the biochemical hypothesis, neural hypothesis, autoimmune hypothesis, oxidative hypothesis and genetic hypothesis (Ongenae et al., 2003; Schallreuter et al., 2008; Spritz, 2008). Among them, much evidence has indicated that autoimmune processes are the major explanation, and genetic data has already provided important insights (Alkhateeb et al., 2003; Ongenae et al., 2003; Sandoval-Cruz et al., 2011; Spritz, 2012).

The major histocompatibility complex (MHC) region on chromosome 6p21.3, which contains the human leukocyte antigen (HLA) genes, was thought to play an essential role in immune response. The association between vitiligo and HLA genes has been documented for a few decades. In the early stage, case-control studies with a small sample size based on serological typing or DNA typing of HLA have shown an association. However, only the associations of HLA-A ∗ 02, HLA-DRB1 ∗ 04 and HLA-DRB1 ∗ 07 with vitiligo were somehow consistent among different studies (Zhang et al., 2005; Liu et al., 2007; Spritz, 2008). Since the application of genome-wide association study (GWAS), more and more vitiligo susceptibility loci have been identified, including MHC region (Jin et al., 2010; Quan et al., 2010; Jin et al., 2011; Birlea et al., 2013). In 2010, Jin et al. identified two major MHC signals associated with generalized vitiligo in the European population, represented by rs12206499 (HLA-A ~ HCG9) and rs532098 (HLA-DRB1 ~ HLA-DQA1). Moreover, another two association signals within the MHC region were reported in the Han Chinese population, including rs11966200 (might reflect HLA-A ∗ 30:01, HLA-B ∗ 13:02, HLA-C ∗ 06:02 and HLA-DRB1 ∗ 07:01) and rs9468925.

Although we have established the association between HLA genes and generalized vitiligo in the Han Chinese subjects by both linkage study and GWAS (Liang et al., 2007; Quan et al., 2010), due to the complexity of the MHC region, it is difficult to fully reveal the genetic architecture of the MHC region for vitiligo. Recently, the newly developed computational strategy SNP2HLA (Jia et al., 2013) enables researchers to perform fine-mapping studies within the MHC region. This has resulted in many achievements in the research of autoimmune diseases of the European population, including idiopathic achalasia (Gockel et al., 2014), psoriasis (Okada et al., 2014), ankylosing spondylitis (Cortes et al., 2015), inflammatory bowel diseases (Goyette et al., 2015), and type 1 diabetes (Hu et al., 2015). However, the susceptibility on HLA genes for diseases such as vitiligo may differ among different populations and continents. To accelerate the fine-mapping strategy in the Han Chinese population, our group constructed the Han-MHC reference panel, which has shown better performance for the imputation of HLA genes (Zhou et al., 2016). Recent studies also implied that the autoimmunity risk conferred by HLA genes was more likely determined by the variants of their amino acid residues (Raychaudhuri et al., 2012; Foo et al., 2013; Kim et al., 2014). However, a fine-mapping study of the MHC region in vitiligo to assess the contribution of amino acid variants of HLA genes has not been implemented. To this end, we conducted a fine-mapping study on 2818 subjects to systematically investigate the coding variants of HLA genes within the MHC region for generalized vitiligo in the Han Chinese population based on a HLA imputation method with our newly built Han-MHC reference panel.