Methodological strategies for mapping HLA susceptibility loci in vitiligo: A focused minireview

Abstract

Some genes, such as HLA class I and class II genes, play important roles in the pathogenesis of vitiligo and its associated autoimmune diseases. Attempts to identify these genes have involved several approaches, including analysis of HLA expression patterns, genetic association studies, and genome-wide association studies. These strategies employ a variety of laboratory techniques, each with distinct advantages and limitations. This review provides a comprehensive overview of current strategies for identifying HLA genes linked to vitiligo susceptibility, with a critical comparison of study designs and methodological approaches. Enhancing these methods is essential to improve accuracy and reliability, ultimately facilitating the development of therapeutic targets and clinically relevant diagnostic biomarkers for patients with vitiligo.

Key Words: Vitiligo; Human leukocyte antigen; HLA typing; Case-control studies; Family studies; Genome-wide association studies

Core Tip: Sequencing-based HLA typing offers higher resolution and can detect novel HLA alleles in vitiligo that may be missed by serological or conventional molecular methods. While gene expression and association studies are susceptible to false positives, rigorous genome-wide association studies incorporating quality control, population stratification adjustment, and replication enhance reliability. Computational modeling of the HLA-peptide-TCR complex holds promise for guiding the development of targeted precision immunotherapies.

INTRODUCTION

Vitiligo is the most common pigmentary disorder, affecting approximately 0.40% of the global population[1], and causes a significant clinical and psychosocial burden due to its visible and often stigmatizing skin depigmentation[2,3]. It is a heterogeneous disorder characterized by various phenotypes based on clinical features, age of onset, family history, triggers, comorbidities, the extension of the patches, and underlying immunopathological mechanisms[4]. The etiology of vitiligo is multifactorial, involving both immunogenetic[5-7] and environmental factors[8].

Among immunogenetic factors, variations in HLA genes are particularly important because they play a central role in T cell–mediated elimination of melanocytes[9-11] and in the development of autoimmunity[12-14]. Environmental and psychological determinants, such as chemical exposures[8], viral infections[15], and stress[16], also contribute to disease onset and progression.

Recent advances in genomic approaches and deoxyribonucleic acid (DNA)-based technologies have facilitated efforts to map and identify specific HLA alleles and haplotypes associated with vitiligo susceptibility and pathogenesis across different ethnic populations[7,17]. Each approach offers distinct advantages and limitations, and their findings can be complementary.

This minireview summarizes current approaches for identifying HLA genes associated with vitiligo susceptibility and compares study designs, including gene expression analyses, genetic association studies, and genome-wide association studies (GWAS).

OVERVIEW OF THE HLA SYSTEM

The HLA region, located on chromosome 6 (6p21.3), represents one of the most polymorphic regions of the human genome. These loci are categorized into three main classes with distinct features and functions.

HLA class I (HLA-I) region includes three classical genes (HLA-A, HLA-B, and HLA-C) coding for the alpha (α)-chain of the HLA-I molecules. These glycoproteins acquire short peptides, typically 8-11 residues in length, generated through proteasomal degradation of intracellular proteins. The complex of transporter associated with protein processing (TAP) subsequently facilitates HLA-I/peptide complex translocation into the endoplasmic reticulum. After glycosylation within the Golgi apparatus, these complexes are migrated to the cell membrane, enabling CD8+ T cells to recognize displayed peptides. HLA-I molecules are expressed on the surface of almost all nucleated cells.

HLA class II (HLA-II) comprises three classical loci (DR, DP, and DQ), which encode both α and β chains. These molecules acquire long peptides, typically 13-20 residues in length, from exogenous pathways and present them on the surface of antigen-presenting cells (APCs), particularly myeloid dendritic cells (DCs). HLA-II/peptide complexes are displayed on the cell surface and recognized by naïve CD4⁺ T lymphocytes, triggering their differentiation into effector cells, including follicular Th, T helper (Th) 1, Th2, Th17, and regulatory T cells. HLA-II molecules are primarily expressed on APCs, activated T lymphocytes, and activated endothelial cells under inflammatory conditions.

HLA class III region encodes a variety of molecules involved in immune regulation, including inflammatory mediators such as tumor necrosis factors and components of the complement system.

HLA-peptide binding groove contains multiple distinct pockets or cavities (generally six or more) that are preferentially located in the β-pleated sheet forming the floor of the groove, interacting with specific amino acid side chains from peptides. Because these pockets are lined by amino acid residues, which differ from one HLA molecule to another, they appear to be the most important factor in the specificity of peptide binding.

HLA TYPING TECHNIQUES

Studies investigating HLA genes and vitiligo across different ethnic populations differ in the reporting of allele designations and the extent of HLA information reported. HLA genes are complex and extremely polymorphic, with over 35000 alleles documented to date[18]. In vitiligo research, three primary approaches have been used for HLA typing: Serological assays, molecular techniques, and sequencing-based methods, each differing in their strategy for interrogating HLA antigens and alleles[19] (Figure 1).

Figure 1 Overview of HLA typing techniques. GWAS: Genome-wide association studies; PCR: Polymerase chain reaction; SSP: Sequence-specific primer; SSO: Sequence-Specific oligonucleotide; RFLP: Restriction fragment length polymorphism; SBT: Sequence-based typing; NGS: Next-generation DNA sequencing.

SEROLOGICAL ASSAYS

In early vitiligo studies, HLA polymorphisms were identified using serological methods[20]. These assays involve HLA antigen, specific anti-HLA antibodies, and complement using the microlymphocytotoxicity technique[21]. In some population case-control studies, data were reported only at the antigen level for HLA-I and HLA-II, such as HLA-A13[20], HLA-B13[22], HLA-A30, HLA-Cw6, HLA-DQw3[23], HLA-B21, HLA-Cw6, and HLA-DR53[24]. Such techniques assess the expression of HLA-I and HLA-II molecules, a functional feature that molecular and sequencing techniques cannot always verify.

MOLECULAR TECHNIQUES

Molecular testing for HLA genes has now completely replaced the microlymphocytotoxicity assay in research and clinical laboratories. Amplification methods involve extracting the DNA and amplifying the exons that encode for HLA-I and HLA-II molecules using polymerase chain reaction (PCR)[19]. Sequence-specific primer (SSP)[13,25-29], sequence-specific oligonucleotide (SSO)[30-33], and real-time PCR (qPCR)[34] are common approaches used to detect the highly variable sequence motifs within the HLA susceptibility loci in vitiligo. Moreover, these methods offer faster results at a lower cost. However, their limited haplotype- or allele-specific resolution frequently leads to ambiguities in HLA typing. Additionally, the oligonucleotide probes used in SSO and SSP methods are static and thus unable to accommodate novel HLA alleles.

Another approach to HLA typing involves restriction fragment length polymorphism PCR (PCR-RFLP), which is used to detect variations within small amplified regions of human DNA, including those within the HLA locus in vitiligo[35]. Variations of the PCR-RFLP led to the development of microarray methods, where multiple targets or multiple patient samples could be investigated simultaneously. Such techniques are typically used for the HLA typing of individuals with generalized vitiligo in different ethnic populations[12,36].

SEQUENCING-BASED METHODS

Sequencing methods such as sequence-based typing (SBT)[37], which relies on conventional Sanger sequencing, and next-generation DNA sequencing (NGS)[38] provide high-resolution HLA results. These approaches are particularly effective in resolving phase ambiguities and identifying single-nucleotide polymorphisms (SNPs) as well as novel alleles that may be missed by amplification, RFLP, or hybridization-based methods. Furthermore, NGS has proven to be more cost-effective and easier to implement than SBT.

ASSESSMENT OF T CELL RESPONSES TO HLA-RESTRICTED PEPTIDES

The enzyme-linked immunospot (ELISpot) assay is a widely used immunological technique for assessing T cell responses to HLA-restricted peptides in the context of vitiligo. ELISpot assays conducted by Cui et al[11] demonstrated that among HLA-A*02: 01-positive individuals, peptides P28 and P119 derived from the melanocyte differentiation antigens gp100, tyrosinase, and MelanA/MART-1 may elicit dominant Th1 responses in 15 vitiligo patients. These epitopes were further shown to drive robust cytotoxic T cell activity against melanocytes in vivo. Consistently, ELISpot analyses from an independent study revealed that patients carrying the HLA-A*02 allele responded to gp100-derived and modified peptides, including gp100 209-217, 210M, and gp100 280-288, 288V[39].

QUANTIFICATION ANALYSIS OF HLA GENE EXPRESSION

Quantification of HLA gene expression has significantly advanced our understanding of vitiligo pathogenesis. Initially, HLA expression in unrelated healthy individuals and vitiligo patients was assessed using antibody-based techniques such as flow cytometry and immunohistochemistry to measure protein levels at the cell surface[12,40], and by qPCR for mRNA transcription levels[41]. However, these methods require careful antibody and primer design to detect diverse HLA alleles, which can lead to specificity and sensitivity limitations and potentially ambiguous results[42,43]. More recently, RNA sequencing (RNA-seq) has become the method of choice for assessing HLA allele-specific expression in large whole-transcriptome datasets from vitiligo patients across different populations[7,12]. This approach typically begins with reverse transcription of RNA into complementary DNA, followed by different library preparation strategies, including whole transcriptome sequencing[44,45], PCR amplification with universal gene-specific primers[44,46], or target enrichment using biotinylated oligonucleotide probes[47]. To facilitate these analyses, bioinformatics tools like Seq2HLA have been developed to quantify HLA expression from RNA-seq data. Seq2HLA, a Python- and R-based pipeline first applied in vitiligo studies[12], processes standard RNA-seq FASTQ reads and uses Bowtie to align sequences against exons 2 and 3 of HLA-DRB1, HLA-DQA1, and HLA-DQB1, enabling both HLA-II typing and expression quantification.

STUDY DESIGNS AND STATISTICAL CONSIDERATIONS FOR MAPPING HLA SUSCEPTIBILITY LOCI IN VITILIGO

Three main study designs have been employed to investigate HLA loci associated with vitiligo susceptibility: Gene expression studies, genetic association studies (including both case-control and family-based designs), and GWAS. An overview of the statistical considerations relevant to these approaches includes factors such as population stratification, multiple testing correction, linkage disequilibrium (LD), haplotype estimation, transmission disequilibrium test, and meta-analysis, all of which influence the validity and reproducibility of findings. In Table 1, the main study designs in vitiligo are summarized.

Table 1 Main study designs in vitiligo.

Type of study	Unit of study	Common techniques
HLA expression	HLA protein levels	Flow cytometry, immunohistochemistry
	HLA mRNA levels	qPCR, RNA-seq, microarray
Case-control	Unrelated case-control, exposure and non-exposure group	Microlymphocytotoxicity, SSP, SSO, qPCR, PCR-RFLP, microarray, SBT, and NGS
Family-based association	Related-case control	SSP, SSO
GWAS	Unrelated case-control, exposure and non-exposure group	RNA-seq, microarray

GWAS: Genome-wide association studies; PCR: Polymerase chain reaction; SSP: Sequence-specific primer; SSO: Sequence-specific oligonucleotide; RFLP: Restriction fragment length polymorphism; SBT: Sequence-based typing; NGS: Next-generation DNA sequencing; RNA-seq: RNA sequencing.

HLA EXPRESSION STUDIES

Quantitative differences in HLA-I and II mRNA levels and surface protein expression may influence immune responses and contribute to vitiligo risk. Utilizing the new capture RNA-Seq method, Jin et al[12] found that HLA-DQB1 mRNA and HLA-DQ protein expression in peripheral blood monocytes and DC from vitiligo patients are specifically associated with rs145954018del-rs9271597A haplotype located within lymphoid-specific enhancers. Moreover, this haplotype was primarily associated with early onset and more severe clinical phenotype of vitiligo[12]. Similarly, Hayashi et al[41] reported that elevated expression of HLA-A*02: 01, driven by an SNP haplotype located 20 kb downstream of the gene and associated with an open chromatin configuration, contributes to primary autoimmune vitiligo susceptibility. In contrast, decreased expression of certain HLA molecules may be protective. Jalel et al[40] observed significantly higher HLA-G expression in skin biopsies from healthy individuals than from vitiligo patients, suggesting a potential immunoregulatory role. While these studies are instrumental in identifying candidate HLA-I and II alleles associated with vitiligo, they cannot definitively separate genes with primary pathogenic effects from those showing secondary dysregulation or natural expression variability in genetically diverse populations.

GENETIC ASSOCIATION STUDIES

Several genetic association studies have suggested a potential link between vitiligo and HLA genes across different ethnic populations. This approach typically involves testing specific HLA loci through case-control and family-based association analyses to evaluate genetic associations and explore potential causal relationships.

Case-control genetic association studies

The earliest HLA genetic investigations of vitiligo employed case–control association study designs. These studies genotyped various HLA alleles and haplotypes in individuals exhibiting diverse clinical phenotypes of vitiligo and compared them to healthy controls across multiple populations[26-30,33,37,48-51]. However, heterogeneity in HLA distribution across these studies limits the generalizability of associations with protective or susceptible HLA alleles at a global scale. Factors contributing to this heterogeneity include genetic diversity across populations, vitiligo disease heterogeneity, differences in LD patterns, and variability in the statistical methods employed. These methodological and biological differences may introduce variability both across and within individual studies.

First, variability in HLA distribution across multiple populations may be attributed to the clinical heterogeneity of vitiligo disease, including differences in family history, disease severity, age of onset, clinical subtypes, and the presence of associated autoimmune diseases[12,30,52-55].

Second, positive associations arising from LD may yield conflicting results depending on the population studied and the methods used for haplotype reconstruction. Only a few studies have investigated the association between HLA haplotypes and vitiligo in different ethnic populations[26,31,33,56-59] (Table 2). Familial generalized vitiligo is characterized by an increase in DRB1*04-DQB1*03: 01 in European-American Caucasian families[43], whereas a significant association exists between DRB1*07–DQB1*02 and familial vitiligo in the Moroccan population[26]. This haplotype inference is commonly performed using algorithms based on expectation-maximization, which typically rely on the assumption that the studied populations are in hardy–weinberg equilibrium (HWE)[26,31,58].

Table 2 HLA haplotypes associated with vitiligo and the study population.

HLA haplotypes	Controls	Vitiligo	Haplotype estimation	P value	Risk	Study population
DRB1*07–DQB1*02	300	100	EM algorithm	Pc = 0.024	Predisposing	Moroccan[27]
DRB1*03–DQB1*02				Pc = 0.012	Protective
A*02-B*51-C*15-DRB1*07-DQB1*02	243	116	EM algorithm	P = 0.0113	Predisposing	Brazilian[34]
A*02-B*15-C*07-DRB1*13-DQB1*06				P = 0.0340	Predisposing
A*29-B*44-C*16-DRB1*07-DQB1*02				P = 0.0340	Predisposing
A*33:01-DRB1*07:01	29	251	EM algorithm	P = 6.97 × 10-26	Predisposing	North India[32]
DRB1*04– DQB1*0301	3798	76	Family-based studies	Pc < 0.003	Predisposing	White European and American[59]
DRB1*15–DQB1*0602				Pc < 0.003	Protective
A25-B13-Cw*0602	273	187	Direct counting	P = 0.00166	Predisposing	Chinese[58]
A25-B27-Cw*0602				P = 0.00032	Predisposing
DQA1*0302-DQB1*0303-Cw*0602				P = 0.02300	Predisposing
B13-DQB1*0303-Cw*0602				P = 0.00380	Predisposing

Pc: Corrected P value; HLA: Human leukocyte antigen; EM: Expectation–maximization.

Third, variation in the significance of HWE estimates may occur by chance, leading to false positives and contributing to inconsistencies in HLA association results across studies[27,28,57,59]. Furthermore, statistical tests for HWE generally have limited power, particularly in small sample sizes, and are primarily sensitive to large deviations. To mitigate the risk of false positives, some case-control studies on vitiligo apply corrections for multiple testing, commonly using Bonferroni adjustment (α’ = α/n) or the Sidak correction [α’ = 1 - (1 - α)1/n], where alpha (α) denotes the original significance level, n is the number of tests, and alpha prime (α’) represents the corrected significance threshold[26,33,52,55].

Fourth, a meta-analysis of genetic association studies has demonstrated a significant association between vitiligo and the HLA-A2 allele[60]. Specifically, the predominant HLA-A*02 subtype conferring generalized vitiligo risk is HLA-A*02: 01 in both Japanese populations[59] and individuals of European ancestry[38]. Another meta-analysis identified associations of vitiligo with HLA-A*02, HLA-A*33, and HLA-Aw*31, while HLA-A*09 and HLA-Aw*19 were associated with a decreased risk[61]. However, such studies may be prone to false-positive findings, often exacerbated by reporting and publication biases. Additionally, true variability in association signals may result from factors such as population stratification, epistatic interactions, environmental influences, and heterogeneity in LD structure across populations.

Finally, a replication study conducted in Gujarat confirmed a positive HLA association (e.g., HLA-A*33: 01, HLA-B*44: 03, and HLA-DRB1*07: 01) previously identified in an initial study of North Indian individuals[31]. In this replication study, the authors employed the exact strategy in which only the specific HLA loci that showed significant associations in the original study were genotyped in the replication cohort. This approach offers a practical balance between statistical power and genotyping efficiency[62].

Family-based association studies

Family-based association tests, which build upon the transmission disequilibrium test, are conducted in families ascertained through an affected offspring (proband). These studies are robust against population stratification and other confounding factors common in population-based designs, though their implementation is generally more complex and less frequent. Notably, Zamani et al[32] identified associations between DRB4*01: 01 and DQB1*03: 03 alleles and generalized vitiligo in Dutch families using family-based approaches. In a separate study, Fain et al[57] reported a significant association of HLA-II haplotype DRB1*04-(DQA1*03: 02)-DQB1*03: 01 with both increased susceptibility and earlier onset of vitiligo in multiplex Caucasian families. Casp et al[63] also described an association between vitiligo and loci within the TAP1–PSMB8 region of the HLA in a cohort of 35 families; however, subsequent investigations suggest that these findings may reflect long-range LD with the classical HLA-II region[64].

GWAS

GWAS are a powerful approach for identifying associations of genotypes with phenotypes by comparing allele frequency differences of genetic variants among individuals with similar ancestry but differing phenotypes. In contrast to candidate gene association analysis, GWAS are less susceptible to a priori bias introduced by preselected candidate loci, and they allow for systematic correction of population stratification and multiple testing, thereby enhancing the reliability and generalizability of detected associations. Accordingly, GWAS have identified reproducible association signals that likely represent genuine susceptibility loci within the HLA region for vitiligo. Genotyping of individuals with generalized vitiligo is typically performed using microarrays to identify common variants[12,36,65]. Following data acquisition, rigorous quality control procedures are applied to both case and control samples, as well as to SNPs within the HLA region, using tools such as PLINK[36]. After sample and variant quality control of the GWAS array data, genotypes are typically pre-phased to generate haplotypes using a sequenced haplotype reference panel[66], and imputation is then performed using efficient algorithms such as the positional burrows–wheeler transform[67]. The associations identified for vitiligo within the HLA region were successfully replicated in independent cohorts, thereby supporting the initial GWAS findings (Figure 2). However, the cumulative proportion of disease risk explained by these loci remains incompletely defined.

Figure 2 Overview of the key steps involved in identifying HLA susceptibility loci in vitiligo using genome-wide association studies. The process includes population selection, genotyping, quality control, population stratification adjustment, association testing, fine-mapping of significant loci, particularly within the HLA class I and class II regions, and replication study. DNA: Deoxyribonucleic acid; mRNA: Messenger ribonucleic acid; QC: Quality control; SNPs: Single nucleotide polymorphisms; HLA: Human leukocyte antigen.

GWAS conducted in both Chinese and Caucasian populations have revealed significant association signals within the HLA region on chromosome 6p21.3, although the specific alleles and haplotypes implicated differ across populations. In individuals of European ancestry, GWAS have identified distinct and independent major susceptibility signals within both the HLA-I and HLA-II regions. The HLA-I signal is primarily represented by the HLA-A*02 allele, while the HLA-II association maps predominantly between the HLA-DRB1 and HLA-DQA1 Loci, in LD with HLA-DRB1*04[68]. These findings are therefore consistent with earlier studies implicating both HLA-A*02[60] and HLA-DRB1*04[57] in susceptibility to generalized vitiligo. Indeed, the strong LD between the risk allele at rs12206499 and HLA-A*02 supports the hypothesis that the HLA-I A*02 allele may represent the causal variant at this locus, contributing to vitiligo susceptibility. Recent analysis in the early-onset subgroup identified a novel association with the HLA-II indel rs145954018, along with an independent association at the established HLA-II GWAS locus represented by rs9271597[12]. The strongest association was observed for the rs145954018del–rs9271597A haplotype, which resides on the extended HLA haplotype rs145954018del-DRB1*13: 01-DRB3*01: 01-rs9271597A-DQA1*01: 03 DQB1*06: 03[12]. In contrast, a GWAS in the Chinese Han population, involving 1117 cases of generalized vitiligo and 1429 healthy controls, identified the primary HLA association signal within the HLA-III region, with additional evidence of independent correlation with the HLA-II region[65]. Further analyses indicated that the significant association observed at rs11966200 could correspond to previously documented associations with HLA-I alleles HLA-A*30: 01, HLA-B*13: 02, and HLA-C*06: 02, as well as the class II HLA-DRB1*07: 01 allele, while the signal at rs9468925 may represent a novel HLA risk allele. These population-specific differences highlight the importance of LD and haplotype variation across ancestral groups.

CONCLUSION

Methodological challenges and opportunities coexist in identifying susceptibility HLA alleles and haplotypes in vitiligo research. Key challenges include the risk of false-positive results driven by reporting and publication biases, as well as genuine biological variability both within and between populations. Ambiguous HLA types and typing strings remain a limitation of conventional HLA typing methods. Moreover, data standard hackathons for NGS can impede collaboration when clinical HLA laboratories engage in research partnerships and must exchange data with other laboratories.

Conversely, opportunities arise from increased statistical power through rigorous multiple testing corrections, appropriate selection of statistical models, and careful evaluation of HWE. A comprehensive understanding of these factors is essential before undertaking HLA studies in this domain. Employing standardized, high-resolution HLA genotyping methods, adjusting for confounders, and consistently addressing genetic subgroups enables more robust analyses.

Computational approaches that precisely determine how melanocyte differentiation antigens bind to specific HLA alleles can deepen our understanding of the HLA–peptide–TCR complex. This insight can guide strategies to modulate cytotoxic CD8⁺ T cell responses against melanocytes in vitiligo, inform personalized immunotherapies, refine risk prediction models, and support early intervention strategies for at-risk individuals. Additionally, integrating multi-omics approaches, including single-cell transcriptomics, epigenomics, and proteomics, can provide a comprehensive understanding of vitiligo across biological levels. Such integration could identify vitiligo-specific HLA-presented peptides as potential drug targets for vitiligo immunotherapy and as clinically relevant diagnostic biomarkers.

Finally, collaborative efforts across research groups enhance the capacity for prospective genotyping of additional polymorphisms, further advancing the field.