All work in this study was performed in Turkey.
Association of human leukocyte antigen class II alleles with pemphigus vulgaris in a Turkish population
Article first published online: 26 FEB 2010
© 2010 Japanese Dermatological Association
The Journal of Dermatology
Special Issue: Blistering Diseases (pages 193-239)
Volume 37, Issue 3, pages 246–250, March 2010
How to Cite
TUNCA, M., MUSABAK, U., SAGKAN, R. I., KOC, E. and AKAR, A. (2010), Association of human leukocyte antigen class II alleles with pemphigus vulgaris in a Turkish population. The Journal of Dermatology, 37: 246–250. doi: 10.1111/j.1346-8138.2009.00743.x
- Issue published online: 26 FEB 2010
- Article first published online: 26 FEB 2010
- Received 20 June 2009; accepted 18 September 2009.
- class II allele;
- human leukocyte antigen;
- pemphigus vulgaris;
- Turkish patients
Pemphigus vulgaris (PV) is a severe autoimmune blistering skin disorder that is strongly associated with major histocompatibility complex class II alleles. Human leukocyte antigen (HLA) subtypes vary with racial/ethnic backgrounds. The purpose of this study was to determine the association of HLA class II alleles and haplotypes with PV in Turkish patients. Twenty-five patients with PV and 113 healthy transplant donors were genotyped for HLA class II alleles. HLA DNA typing was performed by the polymerase chain reaction/sequence specific primer method. The frequency of HLA DRB1*04 allele was 68.00% in patients compared to 30.97% in controls (P = 0.0012) and the frequency of HLA DRB1*14 allele was 32.00% in the patient group compared to 8.85% in the control group (P = 0.0054). Also, the frequency of HLA DRB1*04/DQB1*03 and HLA DRB1*14/DQB1*05 haplotypes in PV patients was significantly higher than controls (32.0% vs 6.2%, χ2 = 28.142, P < 0.001; and 16% vs 2.7%, χ2 = 15.143, P = 0.001, respectively). A preventive allele or haplotype for the manifestation of PV has not been identified in this study. Our findings suggest that HLA DRB1*04 and DRB1*14 alleles, and HLA DRB1*04/DQB1*03 and HLA DRB1*14/DQB1*05 haplotypes are genetic markers for general susceptibility to PV in the Turkish population.
The human leukocyte antigen (HLA) super-locus is a genomic region in the chromosomal position 6p21 that encodes the six classical transplantation HLA genes and at least 132 protein coding genes that have important roles in the regulation of the immune system as well as some other fundamental molecular and cellular processes. This small segment of the human genome has been associated with more than 100 different diseases, including various autoimmune disorders. HLA molecules perform a crucial function in the regulation of the immune response.1
Pemphigus vulgaris (PV) is an autoimmune blistering disease of the skin of unknown etiology. While various environmental factors have been implicated as triggering agents, HLA association has been suggested to be the most important probable predisposing factor.2 Studies consistently showed that the major histocompatibility complex (MHC) locus, in particular HLA class II alleles, is associated with PV.3 In distinct ethnic/racial populations, disease and associated HLA antigens distribution is variable For example, HLA DRB1*04, DRB1*1401, DRB4, DQA1*0104, DQA1*03011, DQB1*0302 and DQB1*0502 alleles have been found to be significantly increased in Iranian patients.4 The gene frequencies of HLA DRB1*14, DRB1*12 and DQB1*0503 alleles in PV patients were reported to be significantly higher than a control group in Han subjects of Northeast China.5 In Korean PV patients, the frequency of the DRB1*01 allele was found to be significantly higher.6 The aim of this study was to investigate HLA class II alleles in our population with PV and identify susceptible and preventive HLA antigens.
Patients and controls
Twenty-five unrelated Turkish patients with PV (14 men and 11 women) aged 20–69 years were included in this study. Only patients who had a diagnosis of PV based on clinical and histopathological findings, and confirmed with immunofluorescent microscopy were included in the study. One hundred and thirteen healthy Turkish transplant donors (65 men and 48 women) were used as controls for HLA class II DNA typing. The study was approved by a local ethics committee and all patients and controls gave informed consent. Demographic data were gathered from all patients.
HLA class II DNA typing
DNA extraction and polymerase chain reaction (PCR) amplification was performed as previously described.7,8 Briefly, DNA extraction was carried out by classical phenol-chloroform method from fresh whole blood samples which were anti-coagulated by ethylene diamine tetra acetate. The PCR sequence specific primer (PCR/SSP) method was used to determine HLA DNA typing. A dried primer stock solution consisting of a HLA specific primer mix, (i.e. allele- and group-specific primers and internal positive control primer pairs) was aliquoted in 0.2-mL PCR tubes. To check PCR conditions, the internal positive control primer pairs that amplify the human growth hormone gene segments were used in each specimen. The PCR master mix contained: 0.4 U/μL Taq Polymerase, 200 μmol/L deoxyribonucleotide triphosphate, 50 mmol/L KCl, 1.5 mmol/L MgCl2, 10 mmol/L Tris-HCl, pH 8.3, 0.001% w/v gelatin, 5% glycerol and 100 μg/mL cresol red at final concentration. Amplification reaction contained: 2 μL template DNA (20–100 ng/mL), 3 μL PCR master mix, 5 μL nuclease-free dH2O for one SSP reaction.
Polymerase chain reaction cycling parameters were: denaturation at 94°C for 2 min, one cycle; denaturation at 94°C for 10 s, annealing and extension at 65°C for 60 s, 10 cycles; and denaturation at 94°C for 10 s, annealing at 61°C for 50 s, extension at 72°C for 30 s, 20 cycles, as per manufacturer instructions (HLA DR & DQ Combi SSP; Olerup SSP AB, Stockholm, Sweden). PCR products were electrophoresed on 2% agarose gel. Presence and relative lengths of the specific PCR products were interpreted with the Helmberg-Score interpretation software.
SPSS for Windows ver. 11.5 was used for statistical analysis. The frequency of antigens and haplotypes from patients and controls were compared using χ2-test analysis with Yates correction, and strength of associations was estimated by odds ratio (OR) of χ2-test determination. If five or fewer persons were present per group, Fisher’s exact two-tailed test was used. Global-likelihood ratio was calculated for haplotype frequency estimation. Additionally, relative risk (RR) values for the alleles and haplotypes which were found to be significantly associated with PV were assessed by a calculator created by D. J. R. Hutchon (http://www.hutchon.net/ConfidRR.htm).
The frequencies of HLA alleles are shown in Table 1. The allele HLA DRB1*04 was found in 68.00% of patients compared to 30.97% in controls (P = 0.00124). Also, the frequency of HLA DRB1*14 allele was increased in the patient group compared to the control group (32.00% vs 8.85%, P = 0.0054). DRB1*16 allele was 22.12% in the control group, whereas in the patient group it was 0.00%. There was a statistically significant difference between the two groups (P = 0.0075). But after correction, this difference did not reach the level of significance. There were no significant differences for the other DQ and DR alleles tested between patients and controls.
|Alleles||Controls n = 113 % (n)||Patients n = 25 % (n)||P-values||χ2||OR|
|DRB1*01||14.16% (16)||12.00% (3||NS|
|DRB1*03||23.01% (26)||8.00% (2)||NS|
|DRB1*04||30.97% (35)||68.00% (17)||0.00124||10.43||4.74|
|DRB1*05||7.08% (8)||0.00% (0)||NS|
|DRB1*06||0.88% (1)||0.00% (0)||NS|
|DRB1*07||12.39% (14)||12.00% (3)||NS|
|DRB1*08||3.54% (4)||4.00% (1)||NS|
|DRB1*09||3.54% (4)||0.00% (0)||NS|
|DRB1*10||4.42% (5)||0.00% (0)||NS|
|DRB1*11||36.28% (41)||24.00% (6)||NS|
|DRB1*12||2.65% (3)||0.00% (0)||NS|
|DRB1*13||26.55% (30)||28.00% (7)||NS|
|DRB1*14||8.85% (10)||32.00% (8)||0.00540||7.74||4.85|
|DRB1*15||17.70% (20)||8.00% (2)||NS|
|DRB1*16||22.12% (25)||0.00% (0)||NS||5.35||0.00|
|DQB1*02||32.74% (37)||12.00% (3)||NS|
|DQB1*03||62.83% (71)||84.00% (21)||NS|
|DQB1*04||6.19% (7)||8.00% (2)||NS|
|DQB1*05||33.63% (38)||40.00% (10)||NS|
|DQB1*06||35.40% (40)||16.00% (4)||NS|
In patients with PV, HLA DRB1*04/DQB1*03 was the most common haplotype (32.0%) and the frequency of this haplotype was significantly higher than that of control subjects (32.0% vs 6.2%, χ2 = 28.142, P < 0.001). In control subjects, HLA DRB1*11/DQB1*03 was the most common haplotype (7.5%), but the frequency of this haplotype was not significantly different from that of PV patients (7.5% vs 12%, χ2 = 1.075, P = 0.393). Other frequent haplotypes were HLA DRB1*14/DQB1*05 (16%), and HLA DRB1*11/DQB1*03 (12%) in patients with PV, and HLA DRB1*04/DQB1*03 (6.2%) and HLA DRB1*13/DQB1*03 (4.9%) in controls. Among these, the frequency of HLA DRB1*14/DQB1*05 haplotype in PV patients was significantly higher than controls (16% vs 2.7%, χ2 = 15.143, P = 0.001). No significant differences were found between patients and healthy controls regarding the comparisons of other frequent haplotypes (data not shown).
For PV, RR of HLA DRB1*04 allele was 2.19 (95% confidence interval [CI] = 1.49 < RR < 3.22), and RR of DRB1*14 allele was 3.61 (95% CI = 1.58 < RR < 8.23), RR of DRB1*04/HLA DQB1*03 haplotype was 8.60 (95% CI = 4.68 < RR < 15.81), and RR of DRB1*14/HLA DQB1*05 haplotype was 6.02 (95% CI = 2.18 < RR < 16.60).
To the best of our knowledge, this is the first study investigating HLA class II alleles among Turkish patients with PV using a biomolecular method. Our results suggest that DRB1*04 and DRB1*14 alleles are markers for general susceptibility to PV in the Turkish population. PV was consistently found to be associated with DR4 and DR14 in various previous studies performed in different ethnic populations.3 Our results are consistent with these findings.
There are few studies investigating the relationship between HLA antigens and PV in a Turkish population.9,10 Birol et al.9 reported that among class II HLA antigens, DR4, DR14, DQ8 and DQ4 antigens were significantly higher in PV patients whereas HLA DR11, DQ7 and DQ2 antigens were higher among the healthy controls. The frequencies of DR4 and DR14 antigens were present in 75.8% and 30.3% of the patients as compared to 29% and 15% of controls, respectively, in their study. In our study, the frequencies of HLA DRB1*04 and DRB1*14 alleles were 68.00% and 32.00% in patients compared to 30.97% and 8.85% in controls, respectively. In our study, higher frequencies of HLA DRB1*04 and DRB1*14 alleles among patients was consistent with their findings. However, despite them reporting a positive association between PV and DQ8 and DQ4 antigens, we did not note any positive association between the corresponding alleles and PV. In another study from Turkey, Kavala et al.10 reported that HLA class II antigen DR14 was significantly higher in PV patients whereas HLA DR16 was higher among controls. They also studied healthy relatives of their patients and reported that HLA class II antigens DR4, DR14 and DQ1 were significantly higher while HLA DR3, DR11, DQ5 and DQ6 were significantly lower among healthy relatives of these patients. A higher incidence of HLA DR14 in both the patient and healthy relative groups implies an association of PV in a Turkish population with this antigen. Consistent with this finding, in our study there was also a positive association with the corresponding allele and pemphigus. Also, they found a positive relationship between healthy relatives and HLA DR4 antigens. This may also imply a relationship between genetic susceptibility to this disease and the corresponding HLA allele as shown in our study. Birol et al.9 reported a negative association with HLA DR11, DQ7 and DQ2 antigens and PV whereas Kavala et al.10 reported negative association with HLA DR16 and Turkish PV patients. However, these negative associations between various antigens and PV in Turkish populations in these studies do not confirm each other. Our results do not support these negative associations.
In a recent study which was carried out in the USA among both Jewish and non-Jewish patients, Lee et al.11 reported that the frequencies of three alleles at the DRB1 locus, DRB1*0402, 1401 and 1404, were significantly overrepresented in non-Jewish US patients with PV. They found that 50% of these non-Jewish patients carried the DRB1*0402 allele, 36% carried 1401, and 11% carried 1404. Our findings are consistent with their positive associations at the DRB1 locus in non-Jewish patients with PV. However, they also reported positive associations with DQB1*05 and 03 alleles, whereas in our study there was no such association. In the same study, among 32 Ashkenazi Jewish patients with PV, DRB1*0402 was the only significantly overrepresented allele at the DRB1 locus with 31 of 32 patients carrying it. Three other DRB1 alleles (07, 1104 and 0301) were underrepresented in patients versus controls. When our results are compared with the Jewish patients with PV in this study, HLA DRB1*04, which was the only significantly overrepresented allele in the DRB1 locus in Jewish patients, was also more prevalent in our patient group than controls.
Other studies in different countries and different populations also reported associations between HLA groups and PV. Shams et al.4 found that HLA DRB1*04, DRB1*1401, DRB4, DQA1*0104, DQA1*03011, DQB1*0302 and DQB1*0502 alleles were significantly increased in Iranian patients with PV. In contrast, HLA DRB1*15, DRB1*0301, DRB1*07, DRB1*11, DRB5, DQA1*0101, DQA1*0103, DQA1*201, DQA1*05, DQB1*0201, DQB1*0301, DQB1*06011 and DQB1*0602 were decreased significantly in their patient group. They concluded that HLA DRB1*04 and DRB1*1401 alleles might be among major PV susceptibility factors in their population study.
In a recent paper, Brick et al.12 reported that in a Moroccan population, they observed no significant association with any of the HLA A or B antigens with PV. However, they detected a significant increase of DRB1*04, DRB1*14 and DQB1*03 allele frequencies and significant decrease of DRB1*15 and DQB1*06 allele frequencies. They also reported that the HLA DRB1*15–DQB1*06 haplotype seemed to confer a protective effect in their population while DRB1*04–DQB1*03 and DRB1*14–DQB1*05 haplotypes induced susceptibility to PV. However, we could not detect any haplotype being protective, and there was also a positive association with the DRB1*04–DQB1*03 and DRB1*14–DQB1*05 haplotypes with PV in our study.
Miyagawa et al.13 genotyped HLA DRB1, DQA1, DQB1 and DPB1 alleles in Japanese patients with PV by PCR restriction fragment length polymorphism. They detected that all nine patients with PV in their study carried one or two alleles of HLA DRB1*04 and HLA DRB1*14. They concluded that their findings, together with previous HLA studies on PV patients of different ethnic groups, suggest that HLA DRB1*04 and DRB1*14 alleles are commonly associated with PV across racial barriers.
Sáenz-Cantele et al.14 reported that statistically significant HLA DR frequency differences between patients with PV and controls were observed for DRB1*0402 and DRB1*1401 in Venezuela. They concluded that their results support the hypothesis that DRB1*0402 without DQB1*0302 is the most relevant HLA DRB1 allele responsible for the pathogenesis of pemphigus in Venezuelan patients with PV, and discard the DQB1*0302 influence observed in other populations.
Lombardi et al.15 reported that in Italian patients, PV and pemphigus foliaceus shared DRB1*1401 and DQB1*0503 as susceptible HLA alleles, whereas DRB1*0402 was only found in association with PV.
González-Escribano et al.16 found that among Spanish patients with PV, HLA DR4 allele was high and HLA DR14 frequency was also increased. Interestingly, HLA DR13, a frequent HLA DR specificity in the Spanish general population, was absent among the PV patients. They concluded that in the Spanish population, PV is preferentially and strongly associated with HLA DRB1*0402, whereas DRB1*13 seems to confer a protective effect.
These epidemiological studies in different ethnic populations indicate that genetic factors are involved in the occurrence of PV. Various studies from different countries and populations have shown that different HLA antigens or alleles confer susceptibility to PV in different ethnic populations, and some other alleles have been shown to be protective for PV. Especially HLA class II alleles, particularly DRB1*04 and DRB1*14, have been shown to be more prevalent in PV patients. The results of our study indicate that these two alleles are associated with the susceptibility to PV in a Turkish population, and there are no significant protective alleles in our population. According to the results of our study, we also suggest that HLA DRB1*04/DQB1*03 and HLA DQB1*05/DRB1*14 haplotypes are related to PV in a Turkish population.
Because our results show that RR of the HLA DRB1*04/DQB1*03 haplotype for the disease is greater than that of the HLA DRB1*04 allele (8.60 vs 2.19), and RR of the HLA DRB1*14/DQB1*03 haplotype for the disease is greater than that of the HLA DRB1*14 allele (6.02 vs 3.61), these haplotypes may be more important than these alleles in conferring susceptibility to PV. However, studies with larger patient series are needed to clear the roles of these HLA alleles and haplotypes in this disease.
- 5Association between pemphigus vulgaris and human leukocyte antigen in Han nation of northeast China. Chin Med Sci J 2005; 20: 166–170., , , , , .
- 10HLA haplotypes in Turkish patients with pemphigus and their healthy relatives. Turkderm 2007; 41: 86–89., , , , , .