A Celiac Gene HLA-DQB1∗02:01 Is Associated with Linear IgA Bullous Dermatosis in the Chinese Population

Linear IgA bullous dermatosis (LABD) is a rare autoimmune blistering skin disease characterized by linear deposits of IgA autoantibodies against the basal membrane zone by direct immunofluorescence. Clinical features include polycyclic groupings of bullae with central crusting (Genovese et al., 2019Genovese G. Venegoni L. Fanoni D. Muratori S. Berti E. Marzano A.V. Linear IgA bullous dermatosis in adults and children: a clinical and immunopathological study of 38 patients.Orphanet J Rare Dis. 2019; 14: 115Crossref PubMed Scopus (38) Google Scholar). Serological markers are LAD-1 and LABD97, but clinically used ELISA kits are unavailable. Our current understanding of various autoimmune blistering diseases has been greatly increased by genetic investigations mainly focusing on HLA, such as the understanding of bullous pemphigoid (Sun et al., 2018Sun Y. Liu H. Wang Z. Fu X. Wang C. Mi Z. et al.The HLA-DQB1∗03:01 is associated with bullous pemphigoid in the Han Chinese population.J Invest Dermatol. 2018; 138: 1874-1877Abstract Full Text Full Text PDF PubMed Scopus (17) Google Scholar), pemphigus (Sun et al., 2019Sun Y. Liu H. Yang B. Wang C. Foo J.N. Bao F. et al.Investigation of the predisposing factor of pemphigus and its clinical subtype through a genome-wide association and next generation sequence analysis.J Eur Acad Dermatol Venereol. 2019; 33: 410-415Crossref PubMed Scopus (10) Google Scholar), and dermatitis herpetiformis (Sun et al., 2016Sun Y. Lin Y. Yang B. Wang C. Fu X. Bao F. et al.The HLA alleles B∗0801 and DRB1∗0301 are associated with dermatitis herpetiformis in a Chinese population.J Invest Dermatol. 2016; 136: 530-532Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar). However, the majority of LABD studies mainly summarized the clinical or histological features. Only a few studies reported sparse results of low-resolution HLA serotype associations with LABD in limited sample sizes, including HLA-B8, -DR3, -DQ2, and TNF2 (Collier et al., 1999Collier P.M. Wojnarowska F. Welsh K. McGuire W. Black M.M. Adult linear IgA disease and chronic bullous disease of childhood: the association with human lymphocyte antigens Cw7, B8, DR3 and tumour necrosis factor influences disease expression.Br J Dermatol. 1999; 141: 867-875Crossref PubMed Scopus (60) Google Scholar; Wojnarowska et al., 1988Wojnarowska F. Marsden R.A. Bhogal B. Black M.M. Chronic bullous disease of childhood, childhood cicatricial pemphigoid, and linear IgA disease of adults. A comparative study demonstrating clinical and immunopathologic overlap.J Am Acad Dermatol. 1988; 19: 792-805Abstract Full Text PDF PubMed Scopus (321) Google Scholar), and none reported on the Chinese population. In this study, we aimed to establish HLA associations for LABD by employing a next-generation sequence–based HLA-typing analysis in 84 patients with LABD and 976 healthy controls of the Chinese population. All LABD cases were recruited from our institute from 2004 to the present. The HLA typing results of 976 healthy controls were reported previously (Sun et al., 2018Sun Y. Liu H. Wang Z. Fu X. Wang C. Mi Z. et al.The HLA-DQB1∗03:01 is associated with bullous pemphigoid in the Han Chinese population.J Invest Dermatol. 2018; 138: 1874-1877Abstract Full Text Full Text PDF PubMed Scopus (17) Google Scholar). The study was approved by the Institutional Review Board of Shandong Provincial Institute of Dermatology and Venereology (W2020060136), and all the subjects or their parents gave written informed consent. The diagnosis of LABD was based on clinical manifestation, skin histology, and a predominant linear deposition of IgA along the basal membrane zone by direct immunofluorescence. To discover the genetic risk factors within the major histocompatibility complex region, high-coverage next-generation sequence–based HLA-typing analysis was conducted on five classical HLA alleles: HLA-A, -B, -C, -DRB1, and -DQB1. In detail, the whole-genomic region of class I major histocompatibility complex alleles and the polymorphic region (exons 2–4) of class II major histocompatibility complex alleles were amplified and sequenced using the Illumina Miseq platform (Illumina, San Diego, CA). High-resolution classical HLA alleles were subsequently defined with default settings by the software NGSengine (GenDX, Utrecht, The Netherlands). In addition, the HLA-DQA1 genotype of all patients with LABD was typed by sequence-specific oligonucleotide typing using the LIFECODES SSO Typing Kit (Immucor, Stamford, CT) according to the manufacturer’s instructions. The basic characteristics of 84 patients with LABD are shown in Table 1. A total of 40 male and 44 female patients were included. Disease subtypes were classified by the onset age, with 16 pediatric (aged ≤12 years) and 68 adult (aged ≥18 years) LABD group. Strong sex bias was observed in the pediatric group between males (11; 68.75%) and females (5; 31.25%). A total of 51.19% of patients showed neutrophil-dominant dermal infiltration, whereas 38.10% of patients showed mixed infiltration of both neutrophil and eosinophils. Solely, linear deposits of IgA at the basal membrane zone of direct immunofluorescence were present in 60 patients (71.43%).Table 1Clinical, Histological, GI Symptom Characteristics, Celiac Antibody Tests and HLA AssociationsCategoryPatients with LABD (n = 84)Healthy ControlsSubtype and gendern (%)Mean age (range)n (%)Mean age (range)Pediatric LABD16 (19.2%)6.2 (1–12)——Adult LABD68 (80.8%)46.3 (18–88)976 (100%)61.7 (22–91)AllMale 40 (47.6%)/female 44 (52.4%))38.7 (1–88)Male 596 (61.1%)/female 380 (38.9%)61.7 (22–91)Dermal infiltraten (%)Neutrophilic43(51.2%)Eosinophil and neutrophil32(38.1%)Eosinophil3 (3.6%)Lymphocyte6 (7.1%)DIFn (%)IgA predominant at BMZ60 (71.4%)IgA with C3 at BMZ18(21.4%)IgA with IgG or IgM at BMZ6 (7.2%)IIF (47 available)n (%)IgA positive25 (53.2%)GI symptom (51 available)n (%)Persistent diarrhea unrelated to gluten2 (3.9%)Gluten intolerance1 (2.0%)Occasional stomachache, diarrhea, or constipation6 (11.8%)Diagnosed ulcerative colitis2(3.9%)Serum test for CeDn (%)Any celiac antibody positive3 (5.08%)2% (Yuan et al., 2017Yuan J. Zhou C. Gao J. Li J. Yu F. Lu J. et al.Prevalence of celiac disease autoimmunity among adolescents and young adults in China.Clin Gastroenterol Hepatol. 2017; 15: 1572-1579.e1Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar)Significant HLA associationsPP/PA/AAPP/PA/AAHLA-DQB1∗02:011/22/611/80/895∗P = 9.4 × 10−8, OR = 4.01 (95% CI = 2.41–6.68)F_A = 14.29%F_U = 4.20%HLA-DB1∗03:011/22/611/80/895∗P = 9.4 × 10−8, OR = 4.01 (95% CI = 2.41–6.68)F_A = 14.29%F_U = 4.20%Abbreviations: BMZ, base membrane zone; CeD, celiac disease; CI, confidence interval; DIF, direct immunofluorescence; GI, gastrointestinal; IIF, indirect immunofluorescence; LABD, linear IgA bullous dermatosis.Celiac antibodies for anti-tTG IgG/IgA and anti-gliadin IgG/IgA are shown. Genotypes represent PP/PA/AA for homozygote/heterozygote/absent; F_A represents allele frequency in cases, and F_U represents allele frequency in controls. P-value, corresponding OR per present allele, and 95% CI are shown. Open table in a new tab Abbreviations: BMZ, base membrane zone; CeD, celiac disease; CI, confidence interval; DIF, direct immunofluorescence; GI, gastrointestinal; IIF, indirect immunofluorescence; LABD, linear IgA bullous dermatosis. Celiac antibodies for anti-tTG IgG/IgA and anti-gliadin IgG/IgA are shown. Genotypes represent PP/PA/AA for homozygote/heterozygote/absent; F_A represents allele frequency in cases, and F_U represents allele frequency in controls. P-value, corresponding OR per present allele, and 95% CI are shown. We applied the logistic regression framework to test for associations. The significant threshold was considered as P < 2.30 × 10−4 using Bonferroni correction (217 four-digit HLA alleles were tested totally, and thus significant threshold was 0.05/217), and all tested alleles with frequency and statistical results are shown in Supplementary Table S1. Two HLA alleles were identified as significantly associated with LABD. The significant risk association was demonstrated to be HLA-DQB1∗02:01 (P = 9.4 × 10−8; OR = 4.01; 95% confidence interval = 2.41–76.68) (Table 1 and Figure 1a). The allele was present in 23 of 84 patients with LABD (27.38%) with one homozygote, whereas it was only present in 81 of 976 healthy controls (8.30%). HLA-DRB1∗03:01, which showed perfect linkage disequilibrium (R2 = 1, D’ = 1) with DQB1∗02:01, showed the same effect as expected (Table 1 and Figure 1a). All the patients with LABD with DQB1∗02:01/DRB1∗03:01 also presented with HLA-DQA1∗05:01 (Supplementary Table S2). The pediatric group (6 of 16; 42,86%) presented with higher DQB1∗02:01 frequency than the adult group (16 of 68, 27.12%) without a statistical significance (P > 0.05). Dapsone is the primary treatment for LABD but can cause dapsone hypersensitivity syndrome, which has been reported to be associated with HLA-B∗13:01. We found that four cases carried the risk HLA allele, which meant that these patients were at high risk if they took dapsone as treatment. HLA-DQB1∗02:01 has been well-documented as a celiac gene because nearly 90–95% of patients with celiac disease (CeD) express HLA-DQ2 (Espino and Núñez, 2021Espino L. Núñez C. The HLA complex and coeliac disease.Int Rev Cell Mol Biol. 2021; 358: 47-83Crossref PubMed Scopus (9) Google Scholar). The haplotype DR3-DQ2 also played an important role in CeD (Delgado et al., 2014Delgado J.F. Amengual M.J. Veraguas A. Rodríguez E. de Los Santos M.M. Guallarte M.P. Paediatric celiac patients carrying the HLA-DR7-DQ2 and HLA-DR3-DQ2 haplotypes display small clinical differences.Acta Paediatr. 2014; 103: e238-e242Crossref PubMed Scopus (15) Google Scholar). Several cases have been reported on the relationship between LABD and CeD (Daoud et al., 2006Daoud W. El E.D. Mokni M. Cherif F. Ben T.N. Azaiz M.-I. et al.[Linear IgA bullous dermatosis associated with celiac disease].Ann Dermatol Venereol. 2006; 133: 588-589Crossref Google Scholar; Egan et al., 2001Egan C.A. Smith E.P. Taylor T.B. Meyer L.J. Samowitz W.S. Zone J.J. Linear IgA bullous dermatosis responsive to a gluten-free diet.Am J Gastroenterol. 2001; 96: 1927-1929Crossref PubMed Scopus (37) Google Scholar). To investigate the complicated CeD and gastrointestinal burden on patients with LABD, we conducted follow-up and serum tests for available patients with LABD. A total of 51 patients (60.71%) were successfully followed using telephone interviews with questionnaires. We identified two cases diagnosed with ulcerative colitis; two cases with persistent diarrhea unrelated to gluten; one patient with gluten intolerance; and six patients with occasional stomachache, diarrhea, or constipation. In total, 11 patients with LABD (21.57%) were identified to have various gastrointestinal symptoms (Table 1). Four CeD antibodies, including anti-gliadin (GAF-3X) IgA/IgG antibody and anti-tTG IgA/IgG antibody, were tested in 59 patients using ELISA kit (Euroimmun AG, Lüebeck, Germany). Three patients (5.08%) showed positive results but none in the controls (details are presented in Supplementary Table S3). We might speculate that higher gastrointestinal and celiac burdens were seen in patients with LABD than the published data that 2% of 19,778 adolescents or young adults in China tested positive for serum markers for CeD (Yuan et al., 2017Yuan J. Zhou C. Gao J. Li J. Yu F. Lu J. et al.Prevalence of celiac disease autoimmunity among adolescents and young adults in China.Clin Gastroenterol Hepatol. 2017; 15: 1572-1579.e1Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). In summary, this study represents the largest genetic study of LABD in the Chinese population. We identified HLA-DQB1∗02:01 and DRB1∗03:01 as having significant risk associations with LABD. Consistent with its role in CeD, we found a significantly high gastrointestinal and celiac burden in patients with LABD. Although the precise mechanisms of HLA alleles are unclear, identification of the HLA allele of LABD offers molecular insight into the pathophysiological mechanisms and benefits the diagnostics of LABD. All data generated or used during the study appear in the submitted article. Lichen Li: http://orcid.org/0009-0008-7623-5689 Lele Sun: http://orcid.org/0000-0001-7363-7297 Gongqi Yu: http://orcid.org/0000-0001-7654-5232 Qianqian Xia: http://orcid.org/0000-0001-6315-8073 Tingting Liu: http://orcid.org/0000-0002-8168-6117 Qing Zhao: http://orcid.org/0000-0002-5511-1352 Zhenzhen Wang: https://orcid.org/0000-0001-5927-2471 Chuan Wang: http://orcid.org/0000-0003-3079-8431 Baoqi Yang: http://orcid.org/0000-0003-3521-7130 Yongxia Liu: http://orcid.org/0000-0002-6181-4150 Xuechao Chen: http://orcid.org/0000-0002-3773-7001 Shengli Chen: http://orcid.org/0009-0004-7705-0484 Guizhi Zhou: http://orcid.org/0009-0006-9990-1942 Yonghu Sun: http://orcid.org/0000-0003-4553-2871 Furen Zhang: http://orcid.org/0000-0002-3383-1973 The authors state no conflict of interest. We thank all the participants involved in this study. This work was supported by the Taishan Scholars Program of Shandong Province (tsqn201909141); the Outstanding Youth Grant of Shandong Natural Science Foundation (ZR2020YQ56); the National Natural Science Foundation of China (82073441 and 81874244); the Clinical Innovation Project of Jinan, Shandong Provincial Key research and development program (2019RKC03002); and the Academic promotion program of Shandong First Medical University (2020RC001). Conceptualization: FZ, YS; Data Curation: YS, ZW, LL; Formal Analysis: YS, ZW, LL; Funding Acquisition: FZ, YS; Investigation: LL; Project Administration: YS; LL, FZ; Resources: GY, YS, QX, BY, YL, XC, SC, GZ; Supervision: YS; LL, FZ; Validation: LS, TL, QZ, CW; Visualization: LL, YS; Writing - Original Draft Preparation: LL; Writing - Review and Editing: YS Download .pdf (.15 MB) Help with pdf files Supplementary Table S1 to S3