Polymorphisms in and Genes Are Associated with Rheumatoid Arthritis in a Chinese Han Population

Background

Rheumatoid arthritis (RA) is a kind of multisystem inflammatory disease that mainly affects synovium and surrounding tissues [1,2]. The prevalence rate of RA is 0.8–1.0% in Europeans and Americans, and around 0.5% in Chinese [3]. The incidence is highest in adults aged 20–50 years, and the incidence rate in women is 2–3 times higher than that in men [4]. The etiology of RA is unknown and is thought to be related to genetic factors and environmental factors such as infection, smoking, and pregnancy [5–7].

Recent studies have confirmed that the interferon family and its immunomodulatory pathways, especially type I interferon (IFNs) pathway, play an important role in the pathogenesis of RA [8]. IFNs are cytokines produced, among others, by plasmacytoid dendritic cells during infection [9]. Studies have found that IFNs gene expression is abnormal in whole blood of patients with autoimmune diseases, especially systemic lupus erythematosus (SLE) [10] and RA [11]. Genetic association analysis has confirmed that interferon regulator 5 (IRF5) and tyrosine kinase 2 (TYK2) in the type I interferon pathway are risk genes for SLE [12]. RA and SLE have similar autoimmune abnormalities, and the polymorphisms of IRF5 and TYK2 genes may also affect susceptibility to RA.

IRF-5 is a member of the IRF family of transcription factors and plays an important role in inflammation and autoimmune response [13]. The gene encoding IRF5 maps to chromosome 7q32 [14]. Overexpression of IRF5 gene leads to increased levels of IFN and IL6 proteins and promotes the occurrence and development of inflammation in autoimmune diseases [15]. rs2004640 is located in the cleavage region of exon 1 at the 5′ end of the IRF5 gene, and different bases can form a variety of IRF5 mRNA splices, affecting the stability of the IRF5 gene [16]. In addition, rs10954213 was previously reported to change the polyadenylate site of IRF5 and is associated with IRF5 mRNA overexpression [17]. Multiple case-control association studies have reported the genetic association of IRF5 gene polymorphisms with RA in white and Asian populations [18–20]. However, the susceptibility loci of the IRF5 gene were inconsistent in different races, and there was racial heterogeneity. Moreover, the genetic association between the IRF5 gene polymorphisms and RA risk in Chinese Han populations were controversial.

TYK2 is a member of the non-receptor tyrosine kinase-linked Janus kinase (JAK) family and is part of the Janus kinase/signal transduction and transcriptional activator 4(JAK-STAT) pathway [21]. Several studies have confirmed the important role of TYK2 in the type I interferon signaling pathway [22,23], and in SLE [12], systemic sclerosis (SS) [24], multiple sclerosis [25], and other immunological diseases. Zheng et al reported that expression of the TYK2 gene is closely related to RA disease progression and may be involved in the development of RA by regulating the levels of IL-2, IL-17, and IL-21 [26]. The association between variants in the TYK2 gene and increased expression of type 1 IFN gene has been identified in a previous study [27]. Graham et al indicated that TYK2 rs280500 was a susceptibility locus for SLE [28]. Subsequently, Tang et al revealed that TYK2 rs280500, rs8108236, and rs280519 polymorphisms were significantly associated with SLE risk [12]. However, no study has investigated the polymorphisms of the TYK2 gene and RA risk in a Chinese Han population.

Therefore, the present study sought to investigate the genetic association between IRF5 and TYK2 gene polymorphisms and RA susceptibility in a Chinese Han population.

Material and Methods

SAMPLES:

The Ethics Committee of Changshu Hospital Affiliated to Soochow University (CSHEC-190922) approved the study protocol. Written informed consent for genetics analysis was obtained from all subjects. The case group was composed of 578 unrelated patients (51 men and 527 women) from Jiangsu province who fulfilled the American College of Rheumatology 1982 criteria for RA [29]. The control group was composed of 578 healthy controls (55 men and 523 women) matched for sex, ethnicity, and age. Clinical features of RA patients and controls are shown in Table 1. All the individuals were of Chinese Han ethnicity.

SINGLE-NUCLEOTIDE POLYMORPHISMS (SNPS) SELECTION AND GENOTYPING:

Genomic DNA was extracted from peripheral leukocytes using the standard phenol-chloroform method. The SNP selection of the IRF5 gene was performed using the methods described by Tang et al [12]. Four SNPs in the IRF5 gene were selected: exon 6 (de/in), rs2004640 (G/T), rs2070197 (C/T), and rs10954213 (G/A). The SNP selection of the TYK2 gene was performed using Haploview Software with minor allele frequency (MAF) higher than 0.05 and r2 ≥0.8 based on the HapMap database (CHB, Chinese Han population) (http://www.hapmap.org/index.-html.ja). After screening, 5 tag SNPs (rs280500, rs280519, rs280521, rs8108236, and rs12720253) were selected. All the SNPs were genotyped by direct sequencing with the ABI 3730XL DNA Sequencer.

STATISTICAL ANALYSIS:

SHEsis software was used in statistical analysis (http://analysis.bio-x.cn/myAnalysis.php). The χ² test was used to analyze the significance of Hardy-Weinberg equilibrium (HWE), genotype, and allele frequency of single-nucleotide polymorphisms. Stringent Bonferroni correction was applied to correct the P values obtained by logistic regression, in multiple comparisons, for associations with RA (P=0.05/N). Haploview 4.2 software was used for linkage disequilibrium (LD) analysis and haplotype analysis. P<0.05 was considered statistically significant. Calculation power was obtained at the 0.05 level of significance, assuming an OR of 1.5 (small effect size) by using G*Power software (www.gpower.hhu.de).

Results

ASSOCIATION OF IRF5 POLYMORPHISMS AND RA:

The frequency of rs2004640T was significantly higher in cases than in controls (P=0.0003) after stringent Bonferroni correction (0.05/4) (Table 3). Genotype analysis ascertained that rs2004640 was associated with RA according to the dominant and recessive models (dominant model: P=0.001; recessive model: P=0.01) after stringent Bonferroni correction (0.05/4) (Table 3). No significant difference in the other 3 variants – exon 6 (de/in), rs2070197, and rs10954213 – were detected between the cases and healthy controls (P>0.05) (Tables 2, 3).

As shown in Table 4, 8 major haplotypes were identified by 3 SNPs and the exon 6 insertion/deletion with the lowest frequency threshold (LFT)>0.01). Significant associations were observed between the haplotype (IH1) involving exon 6 (de), rs2070197T, rs10954213A, and rs2004640T and RA (P=0.0009, OR (95%CI): 1.54 [1.191~1.976]) after stringent Bonferroni correction (0.05/8) (Table 2). Additionally, a protective haplotype (IH4) carrying the exon 6 deletion, rs2070197T, rs10954213G, and rs2004640G, was identified (P=4.45e-005, OR (95%CI): 0.48 [0.337~0.688]) after stringent Bonferroni correction (0.05/8) (Table 2).

ASSOCIATION OF TYK2 POLYMORPHISMS AND RA:

Significant association between rs280500A, rs280521A, and rs8108236A and RA were observed after the stringent Bonferroni correction (0.05/5) (rs280500: P=0.0001; rs280521: P=1.06e-008; rs8108236: P=0.0002) (Table 3). Similar results were obtained for the association between the dominant model of rs280500, rs280521, and rs8108236 and RA (rs280500: P=0.0001; rs280521: P=1.35e-008; rs8108236: P=0.0003) after stringent Bonferroni correction (0.05/5) (Table 2). No correlation was detected between the TYK2 rs280519 and rs12720253 polymorphisms and RA risk (P>0.05) (Tables 2, 3).

As shown in Table 4, 15 major haplotypes containing rs280500, rs280519, rs280521, rs8108236, and rs12720253 alleles with the lowest frequency threshold (LFT)>0.01) were identified. Haplotypes containing rs280500A- rs280519A- rs280521A- rs8108236G- rs12720253T(TH1) (P=9.88e-005, OR (95% CI): 3.18 [1.723~5.878]), rs280500A- rs280519G- rs280521G- rs8108236A- rs12720253T(TH6) (P=9.82e-005, OR (95% CI): 2.35 [1.513~3.667]), and rs280500G- rs280519A- rs280521A- rs8108236A- rs12720253T (TH8) (P=0.001, OR (95% CI): 2.52 [1.427~4.460]) were shown to be significantly associated with RA after stringent Bonferroni correction (0.05/15). A protective haplotype – rs280500G- rs280519A-rs280521G- rs8108236G- rs12720253T (TH11) – was also identified in our study (P=2.60e-005, OR (95% CI): 0.65 [0.538~0.799]) after stringent Bonferroni correction (0.05/15).

Discussion

Our results confirmed significant associations between rs2004640 and RA in a Chinese Han population according to allele, dominant, and recessive models. The frequencies of rs2004640 T allele and TT genotype were lower in controls than in the RA cases, which is consistent with previous studies conducted in Chinese Han populations with RA [30,31]. Kozyrev et al found that IRF5 gene rs2004640 had significant ethnic heterogeneity [32]. Maalej et al reported a close association between rs2004640 and RA in Tunisians, but not in Spanish, Swiss, or Argentine populations [19]. In the present study, we verified that rs2004640 was associated with RA in a Han population in Jiangsu, China. Although the frequency of rs2004640 T allele in Chinese Han RA patients in this study was much lower than that in white RA patients, the correlation of rs2004640T allele with RA risk remained. Therefore, the IRF5 rs2004640T allele is a risk factor of RA in both the white and Asian populations.

No0, no association was found between the IRF 5 exon 6 (de/in), rs2070197 (C/T), and rs10954213 (G/A) and RA susceptibility in the Chinese Han population. No association between rs10954213 and RA was identified in the present study, which agrees with the results reported by Li et al [30]. Moreover, rs2070197 was not found to be polymorphic in our sample. This polymorphism has been shown to be strongly correlated with SLE in Argentina, Spain, and Germany [32], although no genetic association was identified for IRF 5 exon 6 (de/in), rs2070197, and rs10954213 and RA risk in this study. The difference between white and Chinese cohorts may explain this inconsistency. In addition, there could be other variants in linkage disequilibrium in this gene conferring RA susceptibility.

Significant associations were detected between the rs280500A, rs280521A, and rs8108236A alleles in TYK2, as well as the dominant model of TYK2 rs280500, rs280521, and rs8108236 and RA risk. As for the variants in non-coding regions of TYK2, most of the variants were reported in SLE. The TYK2 polymorphisms were reported to be significantly associated with SLE risk in Scandinavian and Finnish populations [33]. The rs280500 polymorphism, located upstream of the 5′ UTR of the TYK2 gene, has been investigated in patients with SLE in white and Asian populations. Only 1 study has reported a significant association between TYK2 rs280500 polymorphism and SLE risk in a Chinese Han population [12]. The TYK2 rs280500 polymorphism was not related to SLE risk in other populations in Chinese Han [34] and whites [33]. Although we found that rs280500 polymorphism was significantly associated with RA risk in a Chinese Han population, the function of this polymorphism in the pathogenesis of RA needs further investigation.

Few studies have been conducted on the association between TYK2 rs280521 and rs8108236 polymorphisms and autoimmune diseases susceptibility. Our study has shown a significant association between TYK2 rs280521 and RA risk for the first time, which was different from the results reported by Sigurdsson et al [33] in patients with SLE. We also found that TYK2 rs8108236 polymorphism was significantly associated with RA risk, which was different from that reported by Tang et al [12] and Li et al [30] in patients with SLE in Chinese Han populations. These conflicting results suggest there is a difference in the pathogenesis of SLE and RA. Although the function of TYK2 rs280521 and rs8108236 polymorphisms in RA is unclear, we may hypothesize that these variants are in moderate to high LD with other functional variants. Thus, further investigation with a larger sample size and multiple ethnicities is necessary.

Furthermore, research in a white population confirmed a significant association between TYK2 rs280519 and SLE [35]. However, no association was found between TYK2 rs280519 and SLE in a Chinese Han population [12] or in a Japanese population [36]. In the present study, we found that TYK2 rs280519 was not associated with RA risk in a Chinese Han population in Jiangsu province, which may indicate TYK2 rs280519 is a susceptible factor for autoimmune diseases including SLE and RA in whites but not in Asians. To confirm this result, larger studies with multiple ethnicities are necessary in the future.

Haplotype analysis has shown that the haplotype (IH1) involving exon 6 (de), rs2070197T, rs10954213A, and rs2004640T and haplotype (IH4) carrying the exon 6 deletion, rs2070197T, rs10954213G, and rs2004640G were associated with the RA susceptibility. This is the first time that haplotypes defined by exon 6, rs2070197, rs10954213, and rs2004640 polymorphisms were found to be associated with RA in a Chinese Han population. However, comparison of IRF5 haplotypes among Japanese [37], whites [28], and Han Chinese [12] indicated differences between Whites and Han Chinese. The white risk haplotype was not present in Japanese and Han Chinese for the absence of rs2070197C allele. Instead, a risk haplotype [(exon 6 (de)- rs2070197T- rs10954213A- rs2004640T)] was identified in Han Chinese. Furthermore, neither of these 2 susceptible haplotypes [exon 6 (in)- rs10954213A- rs2004640T and exon 6 (de)- rs10954213A- rs2004640G] that were identified in patients with SLE [12] have been found in patients with RA in a Chinese Han population, which may indicate the different pathogenesis RA and SLE. Regarding the haplotypes of TYK2 gene polymorphisms in RA, we found 2 risk haplotypes and 1 protective haplotype defined by rs280500, rs280519, rs280521, rs8108236, and rs12720253 polymorphisms in patients with RA in a Chinese Han population for the first time. Notably, the SNPs included in the haplotypes in our research were a little different from those in Tang et al [12]. Therefore, to confirm these results, larger case-control studies are necessary in the future.

There are limitations in the present study. First, although the calculation power indicated that the sample size is large enough to detect an association at an odds ratio of 1.5, the sample size included in the study was relatively small. Our results need to be confirmed in studies with larger sample sizes. Second, both genetic and environmental factors were determined to play a role in the development of RA, but we did not assess the influence of environmental factors and RA risk due to insufficient data.

Conclusions

Our results suggest that IRF5 rs2004640, TYK2 rs280500, rs280521, and rs8108236 are risk factors for RA in a Chinese Han population, and haplotypes of IRF5 DTAT and DTGG, and TYK2 AAAGT, AGGAT, GAAAT, and GAGGT may be risk factors for RA susceptibility in a Chinese Han population.