Genetic Polymorphisms of rs3077 and rs9277535 in HLA-DP associated with Systemic lupus erythematosus in a Chinese population

Although the SLE risk gene loci of HLA-DR and HLA-DQ within the major histocompatibility complex (MHC) region has been gradually revealed by recent Genome-Wide Association studies (GWAS), the association of HLA-DP polymorphisms with SLE was minimally reported. Considering that the variants in rs3077 and rs9277535 in the HLA-DP region could influence the immune response by affecting antigen presentation of HLA class II molecules to CD4+ T cells, the present study aimed to explore the role of HLA-DP polymorphisms in SLE. In total, samples from 335 SLE patients and 635 healthy controls were collected and genotyped by a polymerase chain reaction-high resolution melting (PCR-HRM) assay. A significant positive correlation was observed between the SNP rs3077, rs9277535 of HLA-DP and SLE susceptibility (rs3077, OR = 0.74, 95%CI = 0.60–0.91, P = 0.004; rs9277535, OR = 0.72, 95%CI = 0.59–0.88, P = 0.001). Rs3077 polymorphism was corelated to IL-17, INF-γ and cutaneous vasculitis (P = 0.037, P = 0.020 and P = 0.006, respectively). Additionally, rs3077 AA genotype carriers showed lower concentration of inflammatory cytokines and lower cutaneous vasculitis incidence than did the other two genotype. No significant association was observed between rs9277535 and cytokines or any clinical features. In conclusion, HLA-DP polymorphisms (rs3077 and rs9277535) were associated with SLE susceptibility and the levels of some inflammatory cytokines in SLE patients.

Scientific RepoRts | 7:39757 | DOI: 10.1038/srep39757 HLA-DPB1 were also shown associated with SLE 16 . Besides, HLA-DPB1*05:01 was reported relating to the presence of autoantibodies in Japanese SLE patients 17 . However, these studies neither have found the role of allele variants on HLA-DPA1 nor have explored the further association between the reported SNPs and the process of SLE.
Additionally, considering that a number of studies have revealed that HLA-DP genes were associated with susceptibility to many autoimmune diseases, including Wegener's granulomatosis, systemic sclerosis, multiple sclerosis and others [18][19][20] , HLA-DP polymorphisms are likely to play an important role in SLE. In addition, recent Genome-Wide Association studies (GWAS) validated that rs3077 and rs9277535 genetic variants in the HLA-DP locus were significantly associated with HBV infection in Asian population 21 . Due to the fact that HLA-DP genes encode antigen-binding sites and exon 2 of HLA-DPs is highly polymorphic, the variants of rs3077 and rs9277535 may regulate the immune responses to infection by means of affecting the function of antigen presentation of HLA class II molecules in immune cells.
Meanwhile, antigen presentation is related to CD4 + T cells directly. And active CD4 + T cells, which could differentiate into disparate T cells subsets, play a critical role in development of SLE by secreting all kinds of cytokines, such as inflammatory cytokines (IL-1β , IL-6, INF-γ , TNF-α and IL-17), or the inhibitory cytokine IL-10. Serum INF-γ was primarily derived from Th1 cells, IL-6 and IL-10 were mainly from Th2 cells and IL-17 were from Th17 cells. Some studies have shown that these cytokines influenced SLE 22 .
Therefore, in the present study, we aim to explore the association of the HLA-DP polymorphisms rs3077 and rs9277535 with SLE susceptibility and further investigate the influence of HLA-DP polymorphisms rs3077 and rs9277535 on critical serum cytokine levels and clinical features in SLE in Chinese Han nationality.

Results
Characteristics of the included subjects. The primary demographic and clinical characteristics were illustrated in Table 1. A total of 335 SLE patients and 635 health people participated in the study. The age and sex of the participants from two groups were matched (P = 0.795 and 0.480, respectively). The average age of SLE patients and health controls were 35.77 and 36.28 years old, respectively. In SLE patients, the average disease duration was 36.00 (12.00-96.00) months and the average SLEDAI score was 11.76 ± 5.53. Among the SLE group, ANA, anti-dsDNA and anti-Sm antibody positive frequencies were 91.94%, 35.82% and 28.66%, respectively; while the average serum C3 concentration was 0.67 ± 0.30 g/L, and the average C4 was 0.15 ± 0.08 g/L. Additionally, the prevalences of cutaneous vasculitis, arthritis, albuminuria, rash, pleuritis, pericarditis and neurological disorder were 9.55%, 39.40%, 74.63%, 29.85%, 1.49%, 1.49% and 9.85%, respectively.
Genotyping and LD evaluation. All participants were genotyped via the PCR-HRM methods for SNPs rs3077 and rs9277535 in the HLA-DP region. The veracity of the results was confirmed by direct sequencing of PCR products from randomly selected samples. The direct sequencing results were consistent with all of the corresponding genotyping results. All genotypes were distributed in concordance with the Hardy-Weinberg equilibrium (HWE), as determined at the 0.05 significance level.
Association analysis of HLA-DP polymorphisms with serum cytokine levels. Analysis of the influence of the two HLA-DP polymorphisms on serum cytokine characteristics in SLE patients is illustrated in Table 3.   clinical symptoms, a significant difference could be observed between rs3077 and cutaneous vasculitis (CV) (χ 2 = 10.132, P = 0.006) ( Table 4). Moreover, as shown in Table 5, a decreased frequency of the minor allele A in CV patients (10.94%) compared with non-CV patients (28.22%) was also observed (χ 2 = 8.860, P = 0.003), which suggests the relationship between the minor A allele of rs3077 and decreased susceptibility to CV in SLE patients (P = 0.003, OR = 0.312, 95%CI = 0.140-0.699). In addition, under the dominant model, the minor allele A carriers (GA + AA) exhibit lower risk for CV than those with the GG genotype (GA + AA versus GG: P = 0.001, OR = 0.245, 95%CI = 0.098-0.612).

Discussion
Systemic lupus erythematosus (SLE) is a chronic complex autoimmune disease characterized by autoantibody deposits and involvement of multiple systems 23 . The high heritability, high sibling relapse risk ratio and higher concordance rate in monozygotic twins confirm the genetic effect on the etiology of SLE 24,25 . Therefore, the exact genetic pathogenesis of SLE has long been one of the most difficult and intriguing issues.
In the present study, we found two SNPs, rs3077 and rs9277535 of HLA-DP, did associate with the SLE susceptibility and the A allele of the two SNPs are potential protective allele in SLE in the Chinese Han population. To our knowledge, HLA-DP rs3077 and rs9277535 were not reported to be related to autoimmune diseases (e.g., SLE, RA).  It was our study that first investigated the association between the variants of HLA-DP and SLE susceptibility in Chinese Han population residing in Southwest China. As previous studies reported, there were some other SNPs of HLA-DPB1 significantly associated with autoimmune diseases such as rheumatoid arthritis, Wegener's granulomatosis, systemic sclerosis and infectious diseases like chronic hepatitis B infection [18][19][20][21]26,27 . Our study suggested that HLA-DP region may contain some genes that have potential protective effects on SLE, which implied HLA-DP rs3077 and rs9277535 were likely to represent a shared autoimmune or infectious immune locus.  28,29 , it is most likely that the SNP rs3077 was not only associated with SLE susceptibility but also correlated with the levels of some inflammatory cytokines.
HLA-DP is mainly expressed on the surface of antigen-presenting cells like macrophages, dendritic cells and B cells, and belongs to HLA α molecules, which can bind and present antigen epitopes to CD4 + T helper cells [30][31][32] . Rs3077 and rs9277535 are located in the 3′ untranslated regions of HLA class II genes HLA-DPA1 and HLA-DPB1, correspondingly 33,34 . Therefore, rs3077 and rs9277535 are most likely to affect the disease through influencing the binging and presenting antigen epitopes to CD4 + T helper cell. Though without exact evidence, this can be partly supported by the research of O'Brien and his colleagues. In O'Brien's study, genotype and gene expression data collected and integrated from the normal liver samples obtained from 651 Europeans. They found that rs3077 allele G and rs9277535 allele G were correlated with the down-regulated level of HLA-DP mRNA, and allele A was associated with the increased level of HLA-DP mRNA in normal human liver tissue 35 . Therefore, considering that allele G of HLA-DP rs3077 and rs9277535 have been confirmed predisposed to chronic hepatitis B, carrying allele G of these two SNPs could down-regulate the expression of HLA-DP, which turns out to influence the antigen presentation. Compared to HBV infection, SLE is another kind of disease. However, in our present study, allele G was also found predisposed to SLE. These two consistent results suggest that though in different diseases, the role of variants of HLA-DP might be the same. Apart from that we have revealed the significant association between the two SNPs of HLA-DP and SLE, in the present study, we also found the association between rs3077 and the lower concentration of inflammatory cytokines, as well as the lower risk of cutaneous vasculitis, which might explain partly the function of rs3077 in SLE. Obviously, studies on the structure-function of the two SNPs are still needed. Therefore, the details of the mechanism, including how exactly the gene variants influence the gene expression like HLA-DP mRNA regulation, and the gene function like antigen presentation to CD4 + T cells, and so on, would be explored in our future studies. To our knowledge, because these two variants were first reported in SLE, the studies on linkage disequilibrium of HLA-DP in SLE or in autoimmune diseases are inadequate. However, we could take a sight into this issue from the reported linkage disequilibrium of the two variants in other diseases. The allele variant rs3077 in HLA-DPA1 and rs9277535 in HLA-DPB1 were thought that could be independently considered to be possible causative SNPs for HBV in a study which the linkage disequilibrium of multiple-single nucleotide polymorphism genotype data was researched by the textile plot 36 . But in order to clarify the role of these two variants in the numerous alleles, complex structure and the tight linkage disequilibrium of the HLA region, the relationship between these two SNPs and other reported SNPs in HLA would also be discussed in further studies.
In summary, our study firstly found that the two variants (rs3077 and rs9277535) of HLA-DP were associated with decreased SLE susceptibility, and the SNP rs3077 was correlated with the lower IL17, IL-1β and INF-γ concentrations in peripheral blood and reduced susceptibility to CV in SLE patients. This study could provide further evidence to improve our understanding of the exact function of HLA-DP in the pathogenesis of autoimmune diseases. It is worthwhile to mention that there are several limitations in our present study. The sample of patients in our study is relatively small and only Han Chinese individuals in western China were included. Therefore, further studies using large sample sizes and other ethnic populations are needed to confirm the results observed in this study. Serological testing. Laboratory assays of SLE were analyzed by the following methods: ANA, anti-dsDNA and anti-Sm were detected with ANA, anti-dsDNA and ENA reagent kits (Euroimmun company, Germany). Complement 3(C3) and Complement 4(C4) were detected by rate nephelometry on Beckman Coulter IMMAGE 800 immunoassay (Beckman Coulter, Inc, CA, USA). All the tests were conducted according to manufacturers' instruction.
HLA-DP polymorphism genotyping. Genomic DNA was isolated from the peripheral blood by using Genomic DNA kit (Biotake Corporation, Beijing, China), with the concentration of DNA tested on Nanodrop 2000c spectrophotometer (Thermo Scientific, DE). The polymerase chain reaction-high resolution melting (PCR-HRM) was applied to genotype rs3077 and rs9277535 and the analysis was performed on the Light Cycler 480 (Roche Diagnostics, Penzberg, Bavaria, Germany). SNP genotyping was performed in a 20 μ L reaction system which contains 10 μ L Roche Master Mix (Roche Applied Science) (comprising FastStart Taq DNA Polymerase and the High Resolution Melting Dye in a reaction buffer), 2.4 μ L 25 mM MgCl 2 , 0.2 μ L 10 mmol/L Forward Primer, 0.2 μ L 10 mmol/L Reverse Primer, 6.2 μ L deionized water and 1 μ L DNA sample as recommended by the manufacturer. The whole genotyping process was performed under the following conditions: 95 °C for 10 min in the initial denaturation step followed by 50 cycles of 95 °C for 15 s, touchdown cycling (decreasing 1 °C/cycle), 65-55 °C for 10 s, and 72 °C for 10 s in the annealing step and then denatured at 95 °C for 1 min and cooled to 40 °C for 1 min after the amplification phase. After HRM analyses, melting process and cooling, the results were analyzed by the corresponding Gene Scanning Software v1.2 (Roche Diagnostic). Statistical analysis. Statistical power was calculated by a software "PS: Power and Sample size Calculation" (http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize). The Hardy-Weinberg equilibrium (HWE) was evaluated for each polymorphism independently. Mean ± SD, median and interquartile was used to describe the continuous variables with normal and skewed distribution, respectively. Student's t test or Mann-Whitney U test were applied to compare demographic and clinical data between groups as appropriate. Allele case-control comparisons were analyzed by pearson's chi-square test or Fisher's exact test. Association of SNPs with susceptibility of SLE was assessed by figuring out the odds ratio (OR) and 95% Confidence Interval (CI). The following analytic methods were used when it came to compare the subjects from two groups: allelic frequency distribution of the two groups (allele A versus allele B, A as the major allele, B as the minor allele, this applied to the following methods); dominant model (AB + BB versus AA); recessive model (AA + AB versus BB). All statistical analyses were executed applying the Statistical Package for the Social Sciences (SPSS, SPSS Inc., Chicago, IL, USA), version 19.0. A two-sided P value < 0.05 was considered to be statistically significant.