Summary
Behçet disease is a multi‐system disease associated with human leukocyte antigen (HLA) class I polymorphism. High‐resolution next‐generation sequencing (NGS) with haplotype analysis has not been performed previously for this disease. Sixty Egyptian patients diagnosed according to the International Study Group (ISG) criteria for Behçet disease and 160 healthy geographic and ethnic‐matched controls were genotyped for HLA class I loci (HLA‐A, B, C). For HLA class II loci (DRB1, DRB3/4/5, DQA1, DQB1, DPA1, DPB1), 40 control samples were genotyped. High‐resolution HLA genotyping was performed using NGS and the results were analyzed. Clinical manifestations were oral ulcers (100%), genital ulcers (100%), eye (55%) and neurological (28%) and vascular involvement (35%). HLA‐B*51:08 [odds ratio (OR) = 19·75, 95% confidence interval (CI) = 6·5–79; P < 0·0001], HLA‐B*15:03 (OR = 12·15, 95% CI = 3·7–50·7; P < 0·0001), HLA‐C*16:02 (OR = 6·53, 95% CI = 3–14; P < 0·0001), HLA‐A*68:02 (OR = 3·14, 95% CI = 1·1–8·9; P < 0·01) were found to be associated with Behçet disease, as were HLA‐DRB1*13:01 and HLA‐DQB1*06:03 (OR = 3·39, 95% CI = 0·9–18·9; P = 0·04 for both). By contrast, HLA‐A*03:01 (OR = 0·13, 95% CI = 0–0·8; P = 0·01) and HLA‐DPB1*17:01 were found to be protective (OR = 0·27, 95% CI = 0·06–1·03; P = 0·02). We identified strong linkage disequilibrium between HLA‐B*51:08 and C*16:02 and A*02:01 in a haplotype associated with Behçet disease. HLA‐B*51:08 was significantly associated with legal blindness (OR = 2·98, 95% CI = 1·06–8·3; P = 0·01). In Egyptian Behçet patients, HLA‐B*51:08 is the most common susceptibility allele and holds poor prognosis for eye involvement.
Keywords: Behçet disease, epidemiology, genetics, next‐generation sequencing, susceptibility
This is the first study to use Next Generation sequencing technology for high resolution HLA typing for patients with behcet disease. HLA B*5108 is associated with Egyptian behcet patients and a poor prognostic factor in regards to eye disease
Introduction
Behçet disease (BD) is a rare vasculitis having differing clinical presentations, with oral and genital ulcers the most common manifestations of the disease [1, 2, 3]. BD is more common among the population of the old Silk Road of the Middle East and Asia, with clustering through Turkey, Iran, China, Korea and Japan [4, 5]. While the exact etiology of the disease remains unknown, the underlying immune response shows evidence both of innate and adaptive immune system dysregulation [4, 6]. Previous reports have shown association with nucleotide‐binding oligomerization domain containing 2 (NOD2) variants and activated neutrophils. Importantly, activation of T cells and the presence of autoantibodies, especially retinal autoantibodies, gives clear evidence for adaptive immune system involvement [7, 8].
A cornerstone in understanding the disease pathogenesis is its genetic predisposition, determined by specific human leukocyte antigen (HLA) alleles [9, 10]. HLA‐B51 is consistently identified among distinct populations, and meta‐analysis has confirmed the increased susceptibility for developing BD that is associated with this allotype [11, 12]. Although extensive studies in different populations have been reported for HLA class I, very few have studied class II. Given the implication of the adaptive immune system in the pathogenesis of BD, further in‐depth studies of HLA class II become compelling.
The original association between HLA‐B51 and BD has been further dissected with the help of high‐resolution typing of HLA. Interestingly, Ombrello et al. reported susceptibility of HLA‐B*51 carriers, whereas HLA‐B*52 was found to be protective [13]. Advances in genetic typing have been introduced with the advent of next‐generation sequencing (NGS) methods that are targeted specifically to HLA [14]. NGS technology has also helped in identifying associations with the innate immune system, as illustrated in the report by Burillo‐Sanz [8]. However, the use of NGS for typing of HLA has not yet been applied for understanding the role of HLA in BD.
The aim of this study was to utilize NGS to help to understand the genetic background of BD by investigating HLA class I and class II loci, together with haplotype analysis of these genes. A further aim was to identify associations between HLA polymorphism and the distinct disease presentations.
Methods
In our pilot study, 60 Egyptian BD patients (age > 18 years) diagnosed according to the International Study Group (ISG) criteria for BD, together with 160 healthy geographically and ethnic‐matched controls, were genotyped for HLA class I loci (HLA‐A, B, C). For HLA class II loci (DRB1, DRB3/4/5, DQA1, DQB1, DPA1, DPB1) only the patients and 40 control samples were typed. Patients and the control group were HLA genotyped at high resolution for all 11 loci using the NGSgo® Workflow (GenDx, Utrecht, the Netherlands) on an Illumina MiSeq® platform by paired‐end sequencing [2 × 150 base pairs (bp)], and HLA alleles were called using the NGSengine® software (GenDx).
Clinical characteristics were available for 56 patients for data analysis and comparison. Briefly, these were demographic and clinical data including clinical presentation and patterns of organ involvement as well as the severity of eye involvement. Vascular or neurological involvement was also documented, and results of the imaging studies were included whenever imaging studies were deemed necessary at the time by the managing physician. To the best of our knowledge, patients involved in the study were not related.
All data analysis, including allele counts and frequency estimations, were performed in the R environment for statistical computing and visualization [15]. The haplotype analysis was carried out using the R ‘haplo.stats’ package. Numerical data were expressed as means with standard deviations. Categorical data were summarized into percentages. Comparisons between categorical data were conducted using χ2 test or Fisher’s exact test when appropriate.
The work was part of participation in the 17th International Histocompatibility and Immunogenetics Workshop (IHIWS) in the disease association component. The study was conducted in compliance with the declaration of Helsinki and was approved by the Research Committee at Al‐Azhar University.
Results
In this study, the clinical characteristics for 56 Egyptian BD patients were recorded. The majority (49 patients, 87·5%) were male. The mean age of the patients was 35·28 ± 45·1 years with mean age of onset at 25·73 ± 32·47. All the included patients experienced oral and genital ulcers, and nearly half the patients (55·3%) had ocular involvement, with (30·3%) fulfilling the diagnosis of legal blindness. Erythema nodosum was noted in 26 patients (46·4%) while vascular and neurological involvement were documented in 20 (35·7%) and 16 (28·5%) patients, respectively (Table 1).
Table 1.
Demographic data in the studied Egyptian patients with Behçet disease and prevalence of the clinical manifestations in the patient group
| Demographic data | Value |
|---|---|
| Age, mean ± SD in years | 35·29 (9·82) |
| Age at onset, mean ± SD in years | 25·73 (6·74) |
| Disease duration, mean ± SD in years | 9·55 (7·5) |
| Duration of follow up, mean ± SD in years | 3·54 (5·23) |
| Clinical presentation | Number of patients (percentage) |
| Male sex | 49 (87·5%) |
| Genital ulcers | 56 (100%) |
| Erythema nodosum | 26 (46·4%) |
| Ocular involvement | 31 (55·3%) |
| Legal blindness | 17 (30·3%) |
| Vascular involvement | 20 (35·7%) |
| Arterial involvement | 8 (14·2%) |
| Venous involvement | 15 (26·7%) |
| Neurological involvement | 16 (28·5%) |
| Intestinal involvement | 3 (5·3 %) |
SD = standard deviation.
HLA typing for patients and controls
HLA class I typing revealed distinct allele frequencies characterizing the patient and the control groups. In the patient group, the most frequent HLA‐A alleles were HLA‐A*02:01 (21·6%), HLA‐A*01:01 (10%) and HLA‐A*24:02 (8·3%), the most frequent HLA‐B alleles were HLA‐B*51:08 (20%), HLA‐B*15:03 (13·3%) and HLA‐B*51:01 (5·8%) and the most frequent HLA‐C alleles were HLA‐C*16:02 (21·6%), HLA‐C*07:01 (16·6%) and HLA‐C*17:01 (11·6%). By contrast, in the control group the three most frequent alleles at each locus were HLA‐A*02:01 (15·9%), HLA‐A*01:01 (12·1%) and HLA‐A*03:01 (6·25%), HLA‐B*41:01 (7·5%), HLA‐B*52:01 (7·1%) and HLA‐B*49:01 (5·6%) and HLA‐C*07:01 (11·2%), HLA‐C*12:03 (10·9%) and HLA‐C*04:01 (9·6%). Full details of the HLA class I allele frequencies are given in Table 2.
Table 2.
Allele frequency for Class I HLA in Egyptian Behçet disease patients and controls
| Locus | Allele | Controls (%) | Patients (%) | Locus | Allele | Controls (%) | Patients (%) | Locus | Allele | Controls (%) | Patients (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| A | 01:01 | 12·188 | 10·0 | B | 07:02 | 3·438 | 0 | C | 01:02 | 0·938 | 0·833 |
| A | 01:03 | 0·938 | 0·833 | B | 07:05 | 0·938 | 0 | C | 02:02 | 2·812 | 0·833 |
| A | 01:43 | 0·312 | 0 | B | 07:10 | 0·312 | 0 | C | 02:10 | 0·625 | 4·167 |
| A | 02:01 | 15·937 | 21·667 | B | 07:96 | 0·312 | 0 | C | 03:02 | 1·562 | 0 |
| A | 02:02 | 3·438 | 5·0 | B | 08:01 | 3·125 | 1·667 | C | 03:03 | 0 | 0·833 |
| A | 02:05 | 3·438 | 3·333 | B | 13:02 | 3·125 | 1·667 | C | 03:04 | 2·188 | 0·833 |
| A | 02:14 | 0·312 | 0 | B | 14:01 | 0·625 | 0·833 | C | 04:01 | 9·688 | 7·5 |
| A | 02:17 | 0·312 | 0·833 | B | 14:02 | 5·0 | 0·833 | C | 05:01 | 1·25 | 1·667 |
| A | 03:01 | 6·25 | 0·833 | B | 14:03 | 0 | 0·833 | C | 05:37 | 0·312 | 0 |
| A | 03:02 | 2·5 | 0·833 | B | 15:01 | 0 | 0·833 | C | 06:02 | 9·062 | 7·5 |
| A | 03:05 | 0·312 | 0 | B | 15:02 | 0·312 | 0 | C | 07:01 | 11·25 | 16·667 |
| A | 11:01 | 3·438 | 2·5 | B | 15:03 | 1·25 | 13·333 | C | 07:02 | 5·312 | 0·833 |
| A | 23:01 | 4·062 | 3·333 | B | 15:10 | 0·625 | 0 | C | 07:04 | 0·625 | 0·833 |
| A | 23:17 | 0·312 | 0 | B | 15:16 | 0·625 | 0 | C | 07:05 | 0·312 | 0 |
| A | 24:02 | 5·0 | 8·333 | B | 15:17 | 0·625 | 0·833 | C | 07:06 | 0·312 | 0 |
| A | 24:03 | 1·875 | 2·5 | B | 15:18 | 0 | 0·833 | C | 07:18 | 1·25 | 0 |
| A | 24:17 | 0·312 | 0 | B | 15:22 | 2·812 | 0 | C | 08:01 | 0·312 | 0 |
| A | 25:01 | 0·312 | 0 | B | 18:01 | 5·0 | 5·0 | C | 08:02 | 5·938 | 3·333 |
| A | 26:01 | 2·812 | 1·667 | B | 27:02 | 0·625 | 0·833 | C | 12:02 | 7·812 | 3·333 |
| A | 29:01 | 1·562 | 1·667 | B | 27:03 | 1·875 | 0 | C | 12:03 | 10·938 | 9·167 |
| A | 29:02 | 1·562 | 2·5 | B | 27:07 | 1·562 | 1·667 | C | 14:02 | 1·562 | 2·5 |
| A | 29:10 | 0·312 | 0 | B | 35:01 | 3·438 | 4·167 | C | 15:02 | 3·438 | 1·667 |
| A | 30:01 | 3·438 | 4·167 | B | 35:02 | 1·25 | 0·833 | C | 15:05 | 1·875 | 0 |
| A | 30:02 | 3·438 | 1·667 | B | 35:03 | 1·562 | 0 | C | 16:01 | 4·062 | 1·667 |
| A | 30:04 | 4·375 | 8·333 | B | 35:08 | 2·188 | 0 | C | 16:02 | 4·062 | 21·667 |
| A | 30:10 | 0 | 0·833 | B | 37:01 | 0·312 | 1·667 | C | 16:04 | 1·25 | 0·833 |
| A | 30:99 | 0·312 | 0 | B | 38:01 | 5·312 | 2·5 | C | 17:01 | 8·125 | 11·667 |
| A | 31:01 | 1·25 | 1·667 | B | 39:01 | 0·312 | 0 | C | 17:02 | 0·312 | 0 |
| A | 31:04 | 0·312 | 0 | B | 39:10 | 0·625 | 0 | C | 17:03 | 0·312 | 0 |
| A | 32:01 | 5·0 | 3·333 | B | 40:01 | 1·875 | 0·833 | C | 18:02 | 0·312 | 0 |
| A | 33:01 | 2·5 | 1·667 | B | 40:06 | 0·312 | 0 | ||||
| A | 33:03 | 0·938 | 0·833 | B | 41:01 | 7·5 | 4·167 | ||||
| A | 34:02 | 1·25 | 0 | B | 41:02 | 0·625 | 0·833 | ||||
| A | 66:01 | 0·938 | 0·833 | B | 42:01 | 0·938 | 5·0 | ||||
| A | 68:01 | 1·562 | 0 | B | 42:02 | 0·312 | 0·833 | ||||
| A | 68:02 | 2·812 | 8·333 | B | 44:02 | 2·5 | 2·5 | ||||
| A | 69:01 | 1·875 | 1·667 | B | 44:03 | 3·125 | 1·667 | ||||
| A | 74:01 | 0·625 | 0·833 | B | 44:252 | 0·312 | 0 | ||||
| B | 45:01 | 2·812 | 0·833 | ||||||||
| B | 49:01 | 5·625 | 5·0 | ||||||||
| B | 50:01 | 3·75 | 3·333 | ||||||||
| B | 51:01 | 3·438 | 5·833 | ||||||||
| B | 51:05 | 0·312 | 0 | ||||||||
| B | 51:07 | 0·625 | 0 | ||||||||
| B | 51:08 | 1·25 | 20·0 | ||||||||
| B | 52:01 | 7·187 | 3·333 | ||||||||
| B | 52:19 | 0·312 | 0 | ||||||||
| B | 53:01 | 1·25 | 1·667 | ||||||||
| B | 53:05 | 0·312 | 0 | ||||||||
| B | 55:01 | 0·625 | 0·833 | ||||||||
| B | 56:01 | 0·312 | 0 | ||||||||
| B | 57:01 | 0·625 | 0·833 | ||||||||
| B | 57:03 | 1·25 | 1·667 | ||||||||
| B | 58:01 | 2·188 | 2·5 | ||||||||
| B | 58:02 | 0·625 | 0 | ||||||||
| B | 73:01 | 0·938 | 0 | ||||||||
| B | 78:01 | 0·625 | 0 | ||||||||
| B | 82:02 | 0·625 | 0 |
HLA = human leukocyte antigen.
Bold values indicate statistically significant result.
Class II HLA typing for the patients showed the allele frequencies of HLA‐DPA1*01:03 and HLA‐DPB1*02:01 to be 70 and 30·8%, respectively, HLA‐DQA*03:01 and HLA‐DQB1*03:02 as 25·8 and 18·3%, respectively, and HLA‐DRB1*04:02 was the most frequent HLA‐DRB1 allele at 12·5% in the BD patients. In the control group the most frequent HLA class II alleles were HLA‐DPA1*01:03 and HLA‐DPB1*04:01 (72·5 and 3%, respectively), HLA‐DQA*05:01 and HLA‐DQB1*02:01 (both 26·2%) and HLA‐DRB1*03:01 at 13·75%. Full details of the HLA class II allele frequencies are given in Table 3.
Table 3.
Allele frequency for class II HLA in Egyptian Behçet disease patients and controls
| Locus | Allele | Controls (%) | Patients (%) | Locus | Allele | Controls (%) | Patients (%) | Locus | Allele | Controls (%) | Patients (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| DRB1 | 01:02 | 5·0 | 4·167 | DQA | 01:01 | 13·75 | 10·833 | DPB1 | 01:01 | 3·75 | 10·0 |
| DRB1 | 03:01 | 13·75 | 8·333 | DQA | 01:02 | 12·5 | 10·833 | DPB1 | 02:01 | 28·75 | 30·833 |
| DRB1 | 03:02 | 1·25 | 8·333 | DQA | 01:03 | 10·0 | 16·667 | DPB1 | 03:01 | 6·25 | 8·333 |
| DRB1 | 04:01 | 0 | 0·833 | DQA | 02:01 | 12·5 | 5·833 | DPB1 | 04:01 | 30·0 | 20·0 |
| DRB1 | 04:02 | 6·25 | 12·5 | DQA | 03:01 | 18·75 | 25·833 | DPB1 | 04:02 | 7·5 | 10·833 |
| DRB1 | 04:03 | 6·25 | 5·0 | DQA | 04:01 | 2·5 | 9·167 | DPB1 | 05:01 | 1·25 | 0 |
| DRB1 | 04:04 | 3·75 | 4·167 | DQA | 05:01 | 26·25 | 19·167 | DPB1 | 06:01 | 1·25 | 0 |
| DRB1 | 04:05 | 2·5 | 0·833 | DQA | 06:01 | 1·25 | 0 | DPB1 | 09:01 | 1·25 | 0 |
| DRB1 | 04:06 | 0 | 0·833 | DQB1 | 02:01 | 26·25 | 16·667 | DPB1 | 10:01 | 0 | 0·833 |
| DRB1 | 07:01 | 12·5 | 5·833 | DQB1 | 03:01 | 16·25 | 12·5 | DPB1 | 11:01 | 0 | 1·667 |
| DRB1 | 08:01 | 1·25 | 0 | DQB1 | 03:02 | 13·75 | 18·333 | DPB1 | 13:01 | 5·0 | 7·5 |
| DRB1 | 08:04 | 2·5 | 0·833 | DQB1 | 03:03 | 2·5 | 0 | DPB1 | 14:01 | 2·5 | 0·833 |
| DRB1 | 09:01 | 0 | 1·667 | DQB1 | 03:05 | 0 | 1·667 | DPB1 | 15:01 | 0 | 2·5 |
| DRB1 | 10:01 | 7·5 | 1·667 | DQB1 | 04:02 | 5·0 | 11·667 | DPB1 | 165:01 | 0 | 1·667 |
| DRB1 | 11:01 | 6·25 | 5·833 | DQB1 | 05:01 | 12·5 | 7·5 | DPB1 | 17:01 | 11·25 | 3·333 |
| DRB1 | 11:02 | 0 | 0·833 | DQB1 | 05:02 | 1·25 | 2·5 | DPB1 | 23:01 | 1·25 | 0·833 |
| DRB1 | 11:04 | 3·75 | 3·333 | DQB1 | 05:03 | 1·25 | 2·5 | DPB1 | 29:01 | 0 | 0·833 |
| DRB1 | 12:01 | 0 | 1·667 | DQB1 | 06:01 | 6·25 | 5·0 | ||||
| DRB1 | 12:02 | 1·25 | 0 | DQB1 | 06:02 | 1·25 | 1·667 | ||||
| DRB1 | 13:01 | 3·75 | 11·667 | DQB1 | 06:03 | 3·75 | 11·667 | ||||
| DRB1 | 13:02 | 10·0 | 8·333 | DQB1 | 06:04 | 8·75 | 5·833 | ||||
| DRB1 | 13:03 | 2·5 | 1·667 | DQB1 | 06:09 | 1·25 | 2·5 | ||||
| DRB1 | 13:05 | 1·25 | 1·667 | DPA1 | 01:03 | 72·5 | 70·0 | ||||
| DRB1 | 14:01 | 1·25 | 4·167 | DPA1 | 01:04 | 0 | 2·5 | ||||
| DRB1 | 15:01 | 1·25 | 0 | DPA1 | 02:01 | 21·25 | 15·0 | ||||
| DRB1 | 15:02 | 6·25 | 5·0 | DPA1 | 02:02 | 6·25 | 6·667 | ||||
| DRB1 | 16:01 | 0 | 0·833 | DPA1 | 03:01 | 0 | 0·833 |
HLA = human leukocyte antigen.
Allele susceptibility
Several HLA alleles were found to be associated with susceptibility to Behçet disease and their odds ratios (OR) and 95% confidence intervals (CI) are given below. Among HLA class I alleles, HLA‐B*51:08 (OR = 19·75, 95% CI = 6·5–79·5; P < 0·0001), HLA‐B*15:03 (OR = 12·15, 95% CI = 3·7–50·7; P < 0·0001), HLA‐C*16:02 (OR = 6·53, 95% CI = 3–14; P < 0·0001) and HLA‐A*68:02 (OR = 3·14, 95% CI = 1·1–8·9; P < 0·01) were all found to be associated with BD. By contrast, HLA‐A*03:01 was found to be associated with protection from the development of BD, with OR = 0·13 (95% CI = 0–0·8; P = 0·01). In analyzing HLA class II, HLA‐DRB1*13:01 and HLA‐DQB1*06:03 were associated with susceptibility to BD with OR = 3·39 for both (95% CI = 0·9–18·9; P = 0·04), while HLA‐DPB1*17:01 was found to be protective, with OR = 0·27, 95% CI = 0·06–1·03; P = 0·02).
Haplotype analysis
To further delineate the true association of the different HLA alleles, pairwise haplotype analysis was performed. HLA‐B*51:08, HLA‐A*02:01 and HLA‐C*16:02 were found to be in linkage disequilibrium (P < 0·0002 for A–C correlation and P < 0·0001 for B–C correlation).
Association between HLA alleles and clinical manifestations
We next analyzed HLA allele diversity in relation to the distinct disease manifestations (Table 4). In analyzing blinding eye disease, we identified HLA‐B*51:08 to be associated with development of legal blindness (OR = 2·98, 95% CI = 1·06–8·3; P = 0·01), while HLA‐DQA1*03:01 was found to be protective (OR = 0·34, 95% CI = 0·09–1·03; P = 0·03). Conversely, HLA‐DQB1*03:01 was found to be associated with susceptibility to eye involvement in general (OR = 3·19, 95% CI = 0·9–11·8; P = 0·03), whereas both HLA‐B*15:03 and HLA‐DPB1*04:01 were found to be protective of eye involvement in general (OR = 0·31 and P < 0·03 for both, and 95% CI = 08–1·08 and 0·07–1·05, respectively). Interestingly, HLA‐C*07:01 was found to be protective from neurological involvement (OR = 0·12, 95% CI = 0–0·8; P = 0·01) and HLA‐DPB1*04:01 was found to be protective from venous thrombosis (OR = 0·09, 95% CI = 0·07–1·05; P = 0·006).
Table 4.
Allele association with clinical manifestations in Egyptian patients with Behçet disease
| Clinical manifestation | HLA allele | Association | Odds ratio | 95% CI | P‐value |
|---|---|---|---|---|---|
| Eye involvement | B*15:03 | Protective | 0·31 | 0·08–1·08 | 0·03 |
| Eye involvement | DPB1*04:01 | Protective | 0·31 | 0·07–1·05 | 0·03 |
| Eye involvement | DQB1*03:01 | Susceptibility | 3·19 | 0·9–11·8 | 0·03 |
| Blinding eye disease | B*51:08 | Susceptibility | 2·98 | 1·06–8·3 | 0·01 |
| Blinding eye disease | DQA*03:01 | Protective | 0·34 | 0·09–1·03 | 0·03 |
| Neurological involvement | C*07:01 | Protective | 0·12 | 0–0·8 | 0·01 |
| Venous thrombosis | DPB1*04:01 | Protective | 0·09 | 0·07–1·05 | 0·006 |
CI = confidence interval.
Discussion
Behçet disease is a multi‐system immune‐mediated disease affecting mainly the blood vessels, with major organ involvement including the eyes, brain, gastrointestinal trace and the skin [2]. In this study of Egyptian patients and controls, Behçet disease was found to be associated with HLA‐A*68:02, B*51:08 and C*16:02, as well as HLA‐DRB1*13:01 and DQB1*06:03. Conversely, HLA‐A*03:01 and HLA‐DPB1*17:01 were found to be protective. Furthermore, HLA‐B*51:08 was found to be associated with blinding eye disease while HLA‐B*15:03 was found to be protective of eye involvement. Similarly, HLA‐C*07:01 was found to be protective of neurological involvement. Using pairwise haplotype analysis, HLA‐B*51:08, HLA‐A*02:01 and HLA‐C*16:02 were found to be in linkage disequilibrium, giving a probable explanation for the identified association of HLA‐C*16:02 with development of Behçet disease.
The relation between HLA class I and susceptibility to Behçet disease has been the focus of multiple studies in various populations [16, 17, 18]. Most of these studies were included in the meta‐analysis by de Menthon, that showed increased susceptibility in HLA‐B51 carriers compared to the general population [11]. However, the meta‐analysis did not find a significant association with disease presentation. More results have been reported recently, which may help in understanding the underlying association of HLA polymorphism with Behçet disease [19, 20, 21, 22].
In our current report, HLA‐B*51:08 was found to associated with Behçet disease and more importantly the blinding eye disease, which is in line with the reported susceptibility of HLA‐B*51:08 with Behçet disease in the Italian and Spanish populations [23, 24]. However, this finding was not confirmed in the Turkish population [25]. Conversely, HLA‐A*68:02 that we identified here was also found to be associated with Behçet disease in Moroccan females [26], while HLA‐C*16:02 was reported as a susceptibility allele in the Spanish and Portuguese patients [27, 28], but not in the Turkish or the Japanese patients [25, 29].
Few studies have addressed the association of HLA class II with Behçet disease, and these include Turkish, Saudi and Italian populations [9, 23, 25, 30]. None of the prior studies identified susceptibility of HLA‐DRB1*13:01 and HLA‐DQB1*06:03 that is reported in the current study. Interestingly, HLA‐DRB1*04 and DRB1*07 were associated with susceptibility to Behçet disease in the Turkish population [25], while no significant susceptibility alleles were reported in the Saudi and Italian studies [23, 30]. A possible explanation for the difference in the susceptibility alleles, as the authors of the above‐cited articles have also proposed, is the genetic variation between these populations; an additional effect that could be considered is the differences in the sample size, methodology and HLA typing techniques used in these studies.
Previous work has shown an association of HLA‐A2 with Behçet disease in Egyptian patients [21, 31]. This association of HLA‐A as well as HLA‐C reported in the current study triggered a more thorough examination of these associations. After haplotype, analysis of both alleles was found to be in linkage disequilibrium with HLA‐B51, making this allele the probable driver of the significant associations at HLA‐A and ‐C. Furthermore, we performed a deeper look into the amino acid sequences in the HLA molecules, and this identified amino acid residue 116 in the HLA‐A molecule and amino acid residue 97 in the HLA‐B molecule. This finding agrees with the association reported by Ombrello et al. [13], who postulated that these polymorphic sites, which occur in the antigen binding groove, are an explanation for the susceptibility to the disease.
Historically, Behçet disease has been associated with the geographical distribution of the ancient Silk Road [4]. Using the data from this report and with the help of the reported HLA allele frequencies, our map of the distribution of HLA‐B*51:08 also shows correspondence with the ancient Silk Road (Fig. 1). This analysis also agrees with the recent report by Mizuki, who mapped HLA‐A*26 and found a similar distribution [22]. Further studies are needed to understand the biological context of this geographic distribution of Behçet disease.
Fig. 1.
World map showing the HLA‐B*51:08 allele frequency among several populations. Maps were prepared using data extracted from Allelefrequencies.net and 2015 update: new features for human leukocyte antigen (HLA) epitopes, killer cell immunoglobulin‐like receptors (KIR) and disease and HLA adverse drug reaction associations [34].
One strength of our study is the use of high‐resolution typing by NGS technology which is – to the best of our knowledge – first used here for the study of Behçet disease susceptibility. This approach allows more detailed information regarding the allele susceptibility, and is encouraging for further studies of this enigmatic disease. Similarly, our haplotype analysis has identified that the underlying association of HLA‐A and HLA‐C with BD is probably driven by their linkage disequilibrium with HLA‐B*51. Lastly, the study identified associations between HLA alleles and specific disease manifestations.
A weakness of our study is the small sample size, requiring further confirmation of the results. Similarly, although there is a paucity of patients with gastrointestinal manifestations in our cohort, the prevalence of this manifestation in the Egyptian population is reportedly low compared to other populations [3, 31, 32, 33]. Given the pilot nature of the study and the novel technology, it would be of great interest to determine if our results were replicated or contrasted in other ethnic groups or populations.
In conclusion, Behçet disease is associated with HLA‐A*68:02, B*51:08, C*16:02, DRB1*13:01 and DQB1*06:03 in Egyptian patients, while HLA‐A*03:01 and HLA‐DPB1*17:01 protect from development of the disease. HLA‐B*51:08 is the susceptibility allele in the Egyptian population, particularly with the blinding eye manifestation, and is in strong linkage disequilibrium with HLA‐A*02:01 and HLA‐C*16:02. HLA‐B*51:08 can be considered as a poor prognostic marker in Egyptian patients with Behçet disease; however, more studies are needed to confirm the association. Mapping the prevalence of HLA‐B*51:08 yielded a distribution resembling the Silk Road, giving a possible biological origin of the prevalence of the disease.
Disclosures
M. E., G. M. , D. G. A., G. M.‐M., P. J. N., M. F.‐V., A. A., J. H., K. Z. and S. E. have nothing relevant to disclose. H. de B., L. V. de P. and E. R. are employees of GenDx and E. R. is a stockholder in GenDx. All the authors included in the article meet the journal’s criteria for authorship. This manuscript has not been submitted or is not being submitted simultaneously elsewhere and is not, at the time of submission, under consideration by another journal or other publication and no portion of the data have been or will be published elsewhere while the manuscript is under review by the journal.
Acknowledgements
This work was made possible through collaboration during the 17th International Histocompatibility and Immunogenetics Workshop (IHIWS) in the disease association component, with support from GenDx Inc, Utrecht, the Netherlands.
Data Availability Statement
Data sharing of the manuscript is not applicable.
References
- 1. Jennette JC. Overview of the 2012 Revised International Chapel Hill Consensus Conference Nomenclature of Vasculitides. Clin Exp Nephrol 2013; 17:603–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Yazici H, Seyahi E, Yurdakul S. Behçet’s syndrome is not so rare: why do we need to know? Arthritis Rheum 2008; 58:3640–3. [DOI] [PubMed] [Google Scholar]
- 3. Yazici H, Seyahi E, Hatemi G, Yazici Y. Behçet syndrome: a contemporary view. Nat Rev Rheumatol 2018; 14:107–19. [DOI] [PubMed] [Google Scholar]
- 4. Greco A, De Virgilio A, Ralli M et al. Behçet’s disease: new insights into pathophysiology, clinical features and treatment options. Autoimmun Rev 2018; 17:567–75. [DOI] [PubMed] [Google Scholar]
- 5. Jun J‐B, Kim HJ, Kazmi SZ et al. Significant decline in the Incidence of Behcet’s disease in South Korea: a nationwide population‐based study (2004–2017). Arthritis Care Res [internet] [cited 2020 Aug 30]. Available at: https://onlinelibrary.wiley.com/doi/abs/10.1002/acr.24408. [DOI] [PubMed] [Google Scholar]
- 6. Takeuchi M, Kastner DL, Remmers EF. The immunogenetics of Behçet’s disease: a comprehensive review. J Autoimmun 2015; 64:137–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Wildner G, Thurau SR. Cross‐reactivity between an HLA‐B27‐derived peptide and a retinal autoantigen peptide: a clue to major histocompatibility complex association with autoimmune disease. Eur J Immunol 1994; 24:2579–85. [DOI] [PubMed] [Google Scholar]
- 8. Burillo‐Sanz S, Montes‐Cano M‐A, García‐Lozano J‐R et al. Mutational profile of rare variants in inflammasome‐related genes in Behçet disease: a next generation sequencing approach. Sci Rep 2017; 7:8453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Piga M, Mathieu A. Genetic susceptibility to Behcet’s disease: role of genes belonging to the MHC region. Rheumatol Oxf Engl 2011; 50:299–310. [DOI] [PubMed] [Google Scholar]
- 10. Giza M, Koftori D, Chen L, Bowness P. Is Behçet’s disease a ‘class 1‐opathy’? The role of HLA‐B*51 in the pathogenesis of Behçet’s disease. Clin Exp Immunol 2018; 191:11–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. de Menthon M, Lavalley MP, Maldini C, Guillevin L, Mahr A. HLA‐B51/B5 and the risk of Behçet’s disease: a systematic review and meta‐analysis of case–control genetic association studies. Arthritis Rheum 2009; 61:1287–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Maldini C, LaValley MP, Cheminant M, de Menthon M, Mahr A. Relationships of HLA‐B51 or B5 genotype with Behçet’s disease clinical characteristics: systematic review and meta‐analyses of observational studies. Rheumatology 2012; 51:887–900. [DOI] [PubMed] [Google Scholar]
- 13. Ombrello MJ, Kirino Y, de Bakker PIW, Gül A, Kastner DL, Remmers EF. Behçet disease‐associated MHC class I residues implicate antigen binding and regulation of cell‐mediated cytotoxicity. Proc Natl Acad Sci USA 2014; 111:8867–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. De Santis D, Dinauer D, Duke J et al. 16(th) IHIW: review of HLA typing by NGS. Int J Immunogenet 2013; 40:72–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. R Core Team . R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2013. http://www.R‐project.org/. (accessed 11 December 2018) [Google Scholar]
- 16. Choukri F, Chakib A, Himmich H, Hüe S, Caillat‐Zucman S. HLA‐B*51 and B*15 alleles confer predisposition to Behçet’s disease in Moroccan patients. Hum Immunol 2001; 62:180–5. [DOI] [PubMed] [Google Scholar]
- 17. Itoh Y, Inoko H, Kulski JK et al. Four‐digit allele genotyping of the HLA‐A and HLA‐B genes in Japanese patients with Behcet’s disease by a PCR‐SSOP‐Luminex method. Tissue Antigens 2006; 67:390–4. [DOI] [PubMed] [Google Scholar]
- 18. Shahram F, Davatchi F, Nadji A et al. HLA‐B51 frequency in Iranian patients with Behçet’s disease. Adv Exp Med Biol 2003; 528:229–30. [DOI] [PubMed] [Google Scholar]
- 19. Hamzaoui A, Houman MH, Massouadia M et al. Contribution of Hla‐B51 in the susceptibility and specific clinical features of Behcet’s disease in Tunisian patients. Eur J Intern Med 2012; 23:347–9. [DOI] [PubMed] [Google Scholar]
- 20. Lennikov A, Alekberova Z, Goloeva R et al. Single center study on ethnic and clinical features of Behcet’s disease in Moscow, Russia. Clin Rheumatol 2015; 34:321–7. [DOI] [PubMed] [Google Scholar]
- 21. Elfishawi MM, Elgengehy F, Mossallam G et al. HLA class I in Egyptian patients with Behçet’s disease: new association with susceptibility, protection, presentation and severity of manifestations. Immunol Invest 2019; 48:121–9. [DOI] [PubMed] [Google Scholar]
- 22. Nakamura J, Meguro A, Ishii G et al. The association analysis between HLA‐A*26 and Behçet’s disease. Sci Rep 2019; 9:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kera J, Mizuki N, Ota M et al. Significant associations of HLA‐B*5101 and B*5108, and lack of association of class II alleles with Behçet’s disease in Italian patients. Tissue Antigens 1999; 54:565–71. [DOI] [PubMed] [Google Scholar]
- 24. González‐Escribano MF, Rodríguez MR, Walter K, Sanchez‐Roman J, García‐Lozano JR, Núñez‐Roldán A. Association of HLA‐B51 subtypes and Behçet’s disease in Spain. Tissue Antigens 1998; 52:78–80. [DOI] [PubMed] [Google Scholar]
- 25. Pirim I, Atasoy M, Ikbal M, Erdem T, Aliagaoglu C. HLA class I and class II genotyping in patients with Behcet’s disease: a regional study of eastern part of Turkey. Tissue Antigens 2004; 64:293–7. [DOI] [PubMed] [Google Scholar]
- 26. Bennani N, Atouf O, Benseffaj N, Brick C, Essakalli M. HLA polymorphism and Behçet’s disease in Moroccan population. Pathol Biol 2009; 57:403–9. [DOI] [PubMed] [Google Scholar]
- 27. Sanz L, González‐Escribano F, de Pablo R, Núñez‐Roldán A, Kreisler M, Vilches C. HLA‐Cw*1602: a new susceptibility marker of Behçet’s disease in southern Spain. Tissue Antigens 1998; 51:111–4. [DOI] [PubMed] [Google Scholar]
- 28. Bettencourt A, Pereira C, Carvalho L et al. New insights of HLA class I association to Behçet’s disease in Portuguese patients. Tissue Antigens 2008; 72:379–82. [DOI] [PubMed] [Google Scholar]
- 29. Mizuki N, Ohno S, Ando H et al. HLA‐C genotyping of patients with Behçet’s disease in the Japanese Population. Hum Immunol 1996; 50:47–53. [DOI] [PubMed] [Google Scholar]
- 30. Yabuki K, Ohno S, Mizuki N et al. HLA class I and II typing of the patients with Behçet’s disease in Saudi Arabia. Tissue Antigens 1999; 54:273–77. [DOI] [PubMed] [Google Scholar]
- 31. Assaad‐Khalil SH. Clinical, genetic, immunological, and biochemical features of 180 Egyptian patients with Behçet’s disease. In: O’Duffy JD, Kokmen E, eds. Behçet’s disease: basic and clinical aspects. New York: Marcel Dekker, 1991:269–277. [Google Scholar]
- 32. El Menyawi MM, Raslan HM, Edrees A. Clinical features of Behcet’s disease in Egypt. Rheumatol Int 2009; 29:641–6. [DOI] [PubMed] [Google Scholar]
- 33. Sibley C, Yazici Y, Tascilar K et al. Behçet syndrome manifestations and activity in the United States versus Turkey – a cross‐sectional cohort comparison. J Rheumatol 2014; 41:1379–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. González‐Galarza FF, Takeshita LYC, Santos EJM et al. Allele frequency net 2015 update: new features for HLA epitopes, KIR and disease and HLA adverse drug reaction associations. Nucleic Acids Res 2015; 43:D784–788. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data sharing of the manuscript is not applicable.