Introduction

Systemic lupus erythematosus (SLE) is a complex autoimmune disease and is characterized by production of pathogenic autoantibodies and deposits of immune complexes. The cause of SLE is unknown, but strong genetic and environmental components are involved [1]. SLE is estimated to affect about one in 2000 people in some populations, and the clinical presentation is diverse and can sometimes be fatal [2]. Despite the fact that SLE has been intensively studied, the underlying cause of this autoimmune disease remains elusive [3]. Epidemiological evidence, together with recent linkage and association studies, suggested that susceptibility to SLE in humans is strongly influenced by genetic factors [4]. The major histocompatibility complex (MHC) and the human leukocyte antigen (HLA) system have been confirmed as important genetic risk factors associated with SLE. However, the HLA system constitutes 40 % of the overall estimated genetic risk for SLE, and it has been suggested that a substantial proportion of SLE genetic susceptibility is encoded by non-HLA genes [5]. The rs763361 single-nucleotide polymorphism (SNP) within the CD226 gene has recently been reported as a novel susceptibility locus for multiple autoimmune diseases [6]. It has also been reported that CD226 expression deficiency causes high sensitivity to apoptosis in NK T cells obtained from patients with SLE, providing supporting evidence for the role of CD226 in autoimmune diseases [7].

CD40 is a member of the tumor necrosis superfamily, and its gene is located on chromosome 20q12-13.2 [8]. It has nine exons of between 29 and 412 bp in length [9]. CD40 is expressed on B cells, monocytes, dendritic cells, and endothelial cells [10]. CD40 plays an important role in T cell-dependent B cell humoral responses, regulation of cytokines, growth factor production in monocytes, and modulation of adhesion molecule density on the endothelial cell surface [11]. CD40 may be involved in the pathogenesis and exacerbation of different human autoimmune diseases, including SLE [12]. Therefore, the single-nucleotide polymorphisms (SNPs) modulating CD40 gene expression may increase the risk of autoimmune disease. Also several studies have focused on the association between CD40 and CD226 gene polymorphism and SLE risk in different populations; however, no study focused on the impact of gene–gene interaction between CD40 and CD226 gene polymorphism on SLE risk was conducted in Chinese populations. So, the aim of this study was to investigate the impact of CD40 and CD226 gene SNP and additional gene–gene interaction on SLE risk in Chinese Han populations.

Materials and methods

Subjects

This was a case–control study. Participants were consecutively recruited between July 2008 and June 2014 from the Affiliated Hospital to Changchun University of Chinese Medicine. Consecutive, unrelated patients diagnosed with SLE according to classification criteria by the American College of Rheumatology (ACR) [13, 14] were included from the in- and outpatient clinics of a tertiary referral rheumatology center with special interest in SLE. Prevalent in- or outpatients aged more than 18 years were eligible for inclusion. The control group had neither family history nor symptoms related to SLE and was randomly selected from the same regions with nearly 1:1 matched to cases on the basis of age (±3 years) and sex. Blood samples were collected from each participant. At last, a total of 646 subjects (125 males, 521 females), with a mean age of 43.1 ± 11.1 years old, were selected, including 326 SLE patients and 320 normal subjects. Blood samples were collected for genotyping of polymorphisms from each participant. Written informed consent was obtained from each individual prior to participation in the study. This study was approved by the ethics committee of Changchun University of Chinese Medicine.

Data on demographic information, diet style, and family history of SLE for all participants were obtained using a questionnaire administered by trained staffs. Cigarette smokers were those who self-reported smoking cigarettes at least once a day for 1 year or more. Alcohol consumption was expressed as the sum of milliliters of alcohol per week from wine, beer, and spirits. Body weight, height, and waist circumference (WC) were measured. Body mass index (BMI) was calculated as weight in kilograms divided by the square of the height in meters. WC was measured two times at 1 cm above the umbilicus at minimal respiration by trained observers; the mean of the two WC measurements was utilized in the analysis. Blood samples were collected in the morning after at least 8 h of fasting.

Genomic DNA extraction and genotyping

We selected SNPs within the CD40 and CD226 gene, which has been reported that it associated with SLE and minor allele frequency (MAF) >4 %. Two SNPs of CD40 gene and one SNP of CD226 gene were selected for genotyping in the study: rs4810485, rs763361, and rs3765456. All SNPs were detected by TaqMan fluorescence probe. ABI Prism 7000 software and allelic discrimination procedure were used for genotyping of fore-mentioned three SNPs. Probe sequences of all SNPs are shown in Table 1. Genomic DNA from participants was extracted from EDTA-treated whole blood, using the DNA Blood Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. A 25-μl reaction mixture including 1.25 μl SNP Genotyping Assays (20×), 12.5 μl Genotyping Master Mix (2×), 20 ng DNA, and the conditions was as follows: initial denaturation for 10 min and 95 °C, denaturation for 15 s and 92 °C, annealing and extension for 90 s and 60 °C, 50 cycles.

Table 1 Description and probe sequence for three SNPs used for TaqMan fluorescence probe analysis
Full size table

Statistical analysis

The mean and SD were calculated for normally distributed continuous variables, and percentages were calculated for categorical variables. The categorical data were analyzed using Chi-squared test. Further, continuous variables were analyzed using Student’s t test. Hardy–Weinberg equilibrium (HWE) was performed by using SNPStats (available online at http://bioinfo.iconcologia.net/SNPstats). Logistic regression was performed to investigate association between SNP and SLE. Generalized multifactor dimensionality reduction (GMDR) [15] was used to analyze the interaction among three SNPs, and cross-validation consistency, the testing balanced accuracy, and the sign test, to assess each selected interaction, were calculated.

Results

A total of 646 subjects (125 males, 521 females), with a mean age of 43.1 ± 11.1 years old, were selected, including 326 SLE patients and 320 normal subjects. Participants’ characteristics stratified by cases and controls are shown in Table 2. There is no significant difference between case group and control group in age, gender, smoking and drinking status, high-fat and low-fiber diet, BMI, WC, and marriage.

Table 2 General characteristics of study participants in case and control group
Full size table

All genotypes were distributed according to Hardy–Weinberg equilibrium in controls (p > 0.05). The frequencies for T allele of rs4810485 in CD40 gene were significantly higher in SLE cases (28.8 vs 19.5 %) than that in controls, and T allele of rs763361 in CD226 gene was also significantly higher in SLE cases (30.1 vs 20.3 %) than that in controls (Table 3). Logistic regression analysis showed that SLE risk was significantly higher in carriers of T allele of rs4810485 in CD40 gene than those with GG genotype (GT+ TT vs GG), adjusted OR (95 % CI) 1.84 (1.40–2.29). In addition, we also found SLE risk was also significantly higher in carriers of T allele of rs763361 in CD226 gene than those with CC genotype (CT+ TT versus CC), adjusted OR (95 % CI) 1.89 (1.38–2.13). However, we did not find any significant association between rs3765456 and SLE risk after covariates adjustment.

Table 3 Genotype and allele frequencies of three SNPs between case and control group
Full size table

We employed the GMDR analysis to investigate the impact of the interaction among three SNPs after adjustment for covariates. Table 4 summarizes the results obtained from GMDR analysis, and we found that there was a significant two-locus model (p = 0.0100) involving rs4810485 and rs763361, indicating a potential gene–gene interaction between rs4810485 and rs763361. Overall, the cross-validation consistency of this two-locus model was 10/10, and the testing accuracy was 62.17 %. In order to obtain the odds ratios and 95 % CI for the joint effects of the two SNPs on SLE risk, we conducted interaction analysis between two SNPs by using logistic regression. We found that subjects with GT or TT of rs4810485 and CT or TT of rs763361 genotype have the highest SLE risk, compared with subjects with GG of rs4810485 and CC of rs763361 genotype, and OR (95 % CI) was 2.14 (1.67–3.08), after covariates adjustment (Table 5).

Table 4 Best gene–gene interaction models in GMDR model
Full size table
Table 5 Interaction analysis for rs4810485 and rs763361 by using logistic regression
Full size table

Discussion

In current study, we investigated the impact of CD40 and CD226 gene polymorphism on SLE risk. We found that SLE risk was significantly higher in carriers of rs4810485 T allele in CD40 gene than those with GG and was also significantly higher in carriers of rs763361 T allele in CD226 gene than those with CC. However, we did not find any significant association between rs3765456 and SLE risk after covariates adjustment. Till now, just two studies have focused on the association between CD40 SNP and SLE risk, and the results on association between CD40 gene and SLE were inconsistent [16, 17]. Piotrowski et al. [16] indicated that the CD40 T variant might be negatively associated with some clinical disease manifestations in patients with SLE. CD40 has been identified as a new susceptibility locus with SLE in Greek and Turkish populations. However, Vazgiourakis et al. [17] indicated that the rs4810485 minor allele T is underrepresented in SLE and correlates with reduced CD40 expression in peripheral blood monocytes and B cells, with potential implications for the regulation of aberrant immune responses in the disease. This result was inconsistent with that in current study. The rs3765456 was another SNP in CD40 which was reported in the previous study. Joo et al. [18] conducted a study in Korean population and indicated that CD40 gene polymorphisms are possible risk factors for SLE development, especially rs3765456 in the dominant model. CD40 polymorphisms are also associated with SLE clinical manifestation, mainly nephritis and arthritis. However, in this study, we found that the rs3765456 was not associated with SLE risk, although the frequencies for the A allele of rs3765456 in CD40 gene were higher in SLE cases than that in control; however, there was no significance.

CD226 (also known as the DNAX accessory molecule-1; DNAM-1) plays a role as a costimulatory molecule and is expressed in cells of haematopoietic origin, including T lymphocytes, natural. Several SNPs in CD226 have been reported, and investigation on association between CD226 SNPs and SLE and other autoimmune diseases was conducted in different populations. In the SNPs within CD226, rs763361 was a SNP which was most studied previously. Song [19] and Qiu et al. [20] conducted a meta-study and indicated that the CD226 rs763361 polymorphism confers susceptibility to autoimmune disease in Europeans, South Americans, and Asians, and in particular, shows that the CD226 rs763361 polymorphism is associated with SLE and other autoimmune diseases. These results support the existence of an association between the CD226 gene and a subgroup of autoimmune diseases. Du et al. [21] conducted a study in Chinese population and suggested that polymorphism of Gly307Ser (rs763361) in exon 7 of the CD226 gene may be associated with the development of SLE. Maiti et al. [22] also demonstrate that the coding variant rs763361 in CD226 gene is associated with multiple ADs in non-European populations.

SLE risk was influenced by many genetic factors and both CD40 and CD226 were associated with increased SLE risk, so it was necessary to investigate the impact of gene–gene interaction between CD40 and CD226 gene on SLE risk. In current study, we found a significant gene–gene interaction between rs4810485 and rs763361, and subjects with GT or TT of rs4810485 and CT or TT of rs763361 genotype have the highest SLE risk, compared with subjects with GG of rs4810485 and CC of rs763361 genotype. To our knowledge, this was the first study for investigating interaction between CD40 and CD226 gene on SLE risk in Chinese population. To date, just one study [23] focused on the gene–gene interaction on another autoimmune diseases risk (rheumatoid arthritis), which was conducted for Colombian population and provided evidence for gene–gene interaction between SNP in MMEL1 (rs3890745) and C80rf13-BLK gene.

Several limitations of this study should be considered. Firstly, limited number of SNP in CD40 and CD226 gene was chosen in this study. More SNPs, not only in CD40 and CD226, but also in others SLE-related gene, should be included in the further studies. Secondly, environmental factors should be included in the study, and gene–environment interaction should be investigated.

In conclusion, the results of current study indicated an important association of rs4810485 in CD40 gene and rs763361 in CD226 gene polymorphism, combined effect of rs4810485 and rs763361 with increased risk of SLE.