Meta-Analysis of Associations of IL1 Receptor Antagonist and Estrogen Receptor Gene Polymorphisms with Systemic Lupus Erythematosus Susceptibility

Systemic lupus erythematosus (SLE) is an autoimmune disease that affects a number of different organs and tissues. Interleukin-1 (IL1) and estrogen are considered potential elements in the pathology of SLE. Recently, the variable number of tandem repeats (VNTR) polymorphism in the IL1 receptor antagonist gene (IL1-RN) and PvuII (rs2234693) and XbaI (rs9340799) polymorphisms in the estrogen receptor 1 gene (ESR1) have been associated with a predisposition to SLE. However, the evidence for these associations is inconclusive. We therefore conducted a meta-analysis to validate the roles of these polymorphisms in SLE susceptibility. We searched four databases and identified a total of 17 eligible articles comprising 24 studies. The Newcastle-Ottawa quality assessment scale was used to assess the qualities of the selected studies. We assessed the strengths of the associations using odds ratios (ORs) with 95% confidence intervals (95% CIs). Regarding the IL-1RN VNTR, the 2 allele significantly increased SLE susceptibility (2 vs. L: OR = 1.34, 95% CI = 1.03–1.73, P = 0.03). The ESR1 PvuII CC/CT genotype was also associated with SLE susceptibility (CC/CT vs. TT: OR = 1.25, 95% CI = 1.06–1.47, P = 0.01), and the difference was especially pronounced among Asians (CC/CT vs. TT: OR = 1.33, 95% CI = 1.04–1.69, P = 0.02). No significant association between the ESR1 XbaI polymorphism and SLE susceptibility was observed in the overall analysis. However, a marginally significant association between the GG/GA genotype was found in individuals of Asian descent (GG/GA vs. AA: OR = 1.30, 95% CI = 1.01–1.67, P = 0.04). These results indicate that the IL1-RN VNTR 2 allele, ESR1 PvuII CC/CT genotype and ESR1 XbaI GG/GA genotype may increase SLE susceptibility, especially in Asian individuals.


Introduction
Systemic lupus erythematosus (SLE) is an autoimmune disease that affects various organs and tissues, involving the production of a range of autoantibodies against serological, intracellular, nucleic acid and cell surface antigens [1]. Although the mechanisms underlying SLE are not fully understood, genetic, environmental and hormonal factors are all thought to impact on the development of the disease [2].
Cytokines are considered to be potential elements in the pathology of SLE. These include interleukin-1 (IL1), which plays a key regulatory role in initiating and modulating immunologic and inflammatory events [3,4]. Animals with experimental SLE produced increased levels of IL1 throughout the disease course [5]. The IL1 family consists of IL1a, IL1b and IL1 receptor antagonist (IL1-RA) [6]. IL1-RA is an important anti-inflammatory molecule that binds to IL1 receptors in competition with IL1a and IL1b, thus inhibiting their activities and modulating a variety of IL1-related immune and inflammatory activities [7]. The IL1-RA gene (IL1-RN) has a variable number of tandem repeats (VNTR) polymorphism of 86 base pairs (bp) in intron 2. Five alleles correspond to allele 1 (four repeats), allele 2 (two repeats), allele 3 (five repeats), allele 4 (three repeats) and allele 5 (six repeats), which can be further categorized into a long allele (L: 3-6 repeats) and a short allele (2:2 repeats). The genotypes are therefore classified as LL, 2L and 22 [8]. Blakemore et al. [9] first revealed that the frequency of the IL1-RN VNTR 2 allele was increased in SLE patients, since when mounting studies have explored the relationship between the IL1-RN VNTR polymor-phism and SLE susceptibility in different populations; however, the findings have been controversial [10][11][12].
Estrogen is another underlying element in the pathology of SLE. SLE typically presents in women of childbearing age [13] and its morbidity falls remarkably after the menopause, in line with the decline in endogenous estrogen [14]. In an SLE mouse model, female mice had poorer outcomes than male mice, and estrogens exacerbated, while androgens ameliorated, the disease [15]. One possible mechanism is that physiological concentrations of estrogen could affect the secretion of cytokines such as IL1 [16][17][18]. However, the roles of estrogen and IL1 in SLE remain unclear. Estrogen acts on target cells through binding to estrogen receptors (ERs). ERa, encoded by the ER 1 gene (ESR1), is the main form of ER. Two polymorphisms, ESR1 PvuII T/C (rs2234693) and ESR1 XbaI A/G (rs9340799), located in the first intron of the ESR1 gene, have been extensively studied, but the associations between these polymorphisms and SLE susceptibility remain controversial [19,20].

Search strategy
We searched the PubMed, Embase, Wanfang and Chinese National Knowledge Infrastructure databases using the search terms: 'systemic lupus erythematosus' or 'SLE', 'polymorphism' or 'allele' or 'genotype', 'interleukin-1 receptor antagonist' or 'IL1-RN' or 'estrogen receptor' or 'ER'. The literature search was updated on December 2013 and there was no date limit. The results were also supplemented with manual searches of references from the final published articles.

Study selection
The inclusion criteria were: (1) case-control design; (2) studies investigating the relationship between the IL1-RN VNTR, ESR1 PvuII or ESR1 XbaI polymorphisms and SLE susceptibility; (3) studies with sufficient data to provide odds ratios (ORs) and 95% confidence intervals (CIs); and (4) diagnosis of SLE patients performed according to the American College of Rheumatology criteria [32,33]. The exclusion criteria were: (1) studies with overlapping populations; and (2) studies with insufficient data.

Data extraction
The following information was sought from each publication: first author's surname, year of publication, participants' country, ethnicity, sex distribution, genotyping methods, the source of control groups (population-based or hospital-based controls) and matching numbers of genotyped cases and controls. The literature search, eligible study selection and data extraction were carefully conducted independently by two reviewers (Cai and Zhang) and consensuses were reached on all items.

Quality appraisal
Two reviewers (Cai and Zhang) independently rated the methodological quality of every included study according to the Newcastle-Ottawa quality assessment scale [34]. This scale contains nine items (1 point for each) in three parts: selection (four items), comparability (two items) and exposure (three items).

Statistical analysis
Statistical manipulations were conducted using Stata 10.0 (Stata Corporation, College Station, TX, USA). A x 2 test for goodness of fit was used to test for Hardy-Weinberg equilibrium (HWE) in the control group, with P,0.05 indicating a deviation from HWE. Crude ORs and 95% CIs were used to assess the strengths of the associations between the IL1-RN and ESR1 polymorphisms and SLE susceptibility. The statistical significance of the pooled ORs was determined by the Z test, with P,0.05 considered significant. Statistical heterogeneity among studies was assessed by the Q-test and the I 2 statistic was used to estimate heterogeneity quantitatively [35]. In the absence of heterogeneity (P.0.10 or I 2 ,50%), pooled ORs and 95% CIs were calculated by the fixed-effects model (Mantel-Haenszel method), otherwise the random-effects model (DerSimonian-Laird method) was used. Sensitivity analysis was used to evaluate the stability of the results of the meta-analysis by removing one study at a time to determine the influence of the individual data set on the pooled OR. Potential publication bias was examined by a funnel plot of log OR against its standard error using Begg's test, and the degree of asymmetry was assessed by Egger's unweighted regression asymmetry test. Publication bias may be present if P,0.05 [36].

Study characteristics
A detailed flow chart of the inclusion and exclusion processes is presented in Figure 1. Overall, 86 studies (44 in English and 42 in Chinese) were retrieved based on the search terms. Of these, 69 studies were excluded according to the inclusion and exclusion criteria and 17 eligible articles representing 24 studies were identified (one article was considered as several separate studies if it involved different populations or different target single nucleotide polymorphisms). Eleven studies including 1171 SLE patients and 1834 controls described IL-1RN VNTR genotypes, seven studies including 1012 SLE patients and 2442 controls described ESR1 PvuII genotypes and six studies including 816 SLE patients and 1478 controls described ESR1 XbaI genotypes. The first author's surname, publication year, ethnicity, quality score, sex distribution, frequencies of various genotypes in SLE patients and controls and HWE in controls for each study are listed in Tables 1, 2 and 3. The mean score of the quality appraisal was 6.54. In addition, most of the eligible studies were populationbased and polymerase chain reaction was performed in all studies. Genotype distributions in the control populations agreed with HWE in all except seven studies [12,19,22,23,25,29,30].  Table 4). However, a significant association was observed in an allelic contrast model (2 vs. L: OR = 1.34, 95% CI = 1.03-1.73, P = 0.03, Figure 2). After excluding the studies in which the genotype distributions in the control groups deviated from HWE, the IL1-RN VNTR polymorphism showed no significant association with SLE susceptibility in all genetic models (Table S1). We performed subgroup analyses in Asian and Caucasian populations and found no significant differences in either ethnic subgroup (Table S1).

Quantitative synthesis
Seven  Table 4). After excluding the study in which the genotype distribution in the control group deviated from the HWE, ESR1 PvuII polymorphism was significantly associated   Table S2). Six studies investigated the association between ESR1 XbaI polymorphism and SLE susceptibility. No significant relationships were identified for any of the genetic models in the whole study set (GG/GA vs. AA: OR = 1.19, 95% CI = 0.88-1.62, P = 0.27; GG vs. AA/GA: OR = 1.08, 95% CI = 0.77-1.51, P = 0.67; GG vs. AA: OR = 1.09, 95% CI = 0.77-1.54, P = 0.64; G vs. A: OR = 1.15, 95% CI = 0.89-1.49, P = 0.27, Table 4). Exclusion of the two studies in which the genotype distributions in the control groups deviated from the HWE had no significant effect on the results (Table S3). However, stratified analysis by ethnicity demonstrated a marginally significant association in the dominant model (GG/GA vs. AA: OR = 1.30, 95% CI = 1.01-1.67, P = 0.04, Figure 4, Table S3) in individuals of Asian descent.

Publication bias
No publication bias was detected among studies regarding the association between the IL1-RN VNTR polymorphism and SLE (P = 0.83 for 2 vs. L, Figure 5). Similarly, the results of Egger's and Begg's tests showed no publication bias for the ESR1 PvuII or ESR1 XbaI polymorphisms in all models.

Discussion
Studies of gene polymorphisms potentially related to SLE have recently attracted growing attention. In the present study, we performed a meta-analysis of the associations between IL1-RN VNTR, ESR1 PvuII, and ESR1 XbaI polymorphisms and SLE susceptibility. The analysis indicated an association between the 2 allele of the VNTR polymorphism in intron 2 of IL1-RN and increased SLE susceptibility (2 vs. L: OR = 1.34, 95% CI = 1.03-1.73, P = 0.03). There was also an association between ESR1 PvuII and SLE in the dominant model (CC/CT vs. TT: OR = 1.25, 95% CI = 1.06-1.47, P = 0.01), which was pronounced among Asian individuals (CC/CT vs. TT: OR = 1.33, 95% CI = 1.04-1.69, P = 0.02). There was no significant association between the ESR1 XbaI polymorphism and SLE suscepti- Table 3. Characteristics and ESR1 XbaI polymorphism genotype distributions in studies included in the meta-analysis. bility in the overall analysis, but the GG/GA genotype was associated with SLE susceptibility in Asians (GG/GA vs. AA: OR = 1.30, 95% CI = 1.01-1.67, P = 0.04). IL1 is a potent pro-inflammatory cytokine in acute and chronic inflammation in SLE [37]. IL1RA is a natural antagonist of IL1 and its anti-inflammatory activity is mediated through several different pathways [38] and investigations found decreased production of IL1RA in active SLE [39]. The IL1-RN VNTR 2 allele was associated with increased production of IL1b in vitro [40,41] and the concentration of IL1RA was shown to be    correlated with IL1b [42]. Also, this meta-analysis identified carriage of the 2 allele as a risk factor for SLE susceptibility (2 vs. L: OR = 1.34, 95% CI = 1.03-1.73, P = 0.03). In support of this, the IL1-RN VNTR contains three potential protein-binding sites: an acute phase response element, an interferon a and an interferon b silencer B [43]. The 2 allele of IL1-RN VNTR only has 2 repeats. This could affect mRNA length and subsequent protein processing and stability [44], which could in turn affect the production of IL1RA.
Our study also showed that the ESR1 PvuII CC/CT and ESR1 XbaI GG/GA genotypes could increase susceptibility to SLE. Estrogen can affect both innate and adaptive immune responses in mice [45] and SLE patients [46] through different pathways [47][48][49], and estrogen receptors are expressed in most immunocompetent cells [50]. Some researchers determined that IL1b levels were higher during the luteal period compared with the follicular period of the female reproductive cycle, which was consistent with the results of in vivo [51][52][53] and in vitro tests [18]. The polymorphisms PvuII and XbaI are located in intron 1 of ESR1 but are still able to affect the gene, and thus affect estrogen concentrations. The C allele of PvuII can produce a binding site for the B-myb transcription factor, which could enhance the ability to up-regulate downstream receptor structures compared with the T allele [54]. In the present study, SLE susceptibility was associated with the ESR1 PvuII C allele but not with XbaI in overall analysis. However, we could not rule out the possibility of an association between the ESR1 XbaI polymorphism and SLE susceptibility because PvuII and XbaI are tightly linked [55] and it is difficult to identify which one has a role to play.
Given its multifactorial nature, it is likely that the pathogenesis of SLE may be modulated by age, gender, ethnicity, environmental factors and other variables. We therefore carried out subgroup analysis based on ethnicity. Associations between SLE susceptibility and the ESR1 PvuII C allele and the ESR1 XbaI G allele were found in individuals of Asian descent. This may be attributable to genetic heterogeneity among different populations. Consistent with this, ethnicity contributed to the heterogeneity for ESR1 PvuII and ESR1 XbaI. Moreover, sensitivity analysis revealed that the heterogeneity was reduced by removing Johansson et al.'s [31] study, which was the only study in the meta-analysis of the association between ESR1 XbaI polymorphism and SLE susceptibility that was based on Caucasians. It is also possible that differences in lifestyle and environmental factors between different populations may interact with genes to affect the pathogenesis of SLE.
This meta-analysis had some inevitable limitations. First, three studies on the IL1-RN VNTR polymorphism did not provide genotype data, and the data used to analyze the various genetic models were thus not completely consistent. This may lead to misinterpretation of the association between the IL1-RN VNTR polymorphism and SLE susceptibility. Second, although all eligible studies were included in our study, the small sample size and low statistical power (Table S4, S5 and S6) associated with the low incidence of SLE means that there is a possibility of false negative results. We expect more participants being tested in the future to draw a more reliable conclusion. Third, deviation of genotype distributions from HWE in the control populations in some studies may reflect genotyping errors or control selection bias. The results relating to IL1-RN VNTR and ESR1 PvuII changed when these studies were excluded, suggesting that these results should be interpreted with caution. Fourth, as mentioned above, various factors affect the pathology of SLE; the lack of individual data meant that we only pooled the data based on unadjusted information. Finally, the quality scores were not high for some studies and these studies may have distorted the results.
In conclusion, this meta-analysis indicated that the IL1-RN VNTR 2 allele and the ESR1 PvuII CC/CT and ESR1 XbaI GG/GA genotypes may increase the susceptibility to SLE, especially in individuals of Asian descent. However, this conclusion should be interpreted cautiously because of the low statistical power and considerable heterogeneity. Also, further functional studies are needed to investigate the functions of these alleles. Well-designed, large studies in different ethnic groups and with more detailed information on age, sex, and age of onset of the disease are needed to validate our results. Studies of geneenvironment interactions in relation to IL1-RN-ESR1 should also be performed to confirm our preliminary findings.

Supporting Information
Table S1 Meta analysis of the association of IL1-RN VNTR polymorphism with SLE susceptibility. (XLS)

Author Contributions
Conceived and designed the experiments: X. Xia YZ. Performed the experiments: LC JWZ Jia-jia Wang. Analyzed the data: LC JWZ. Contributed reagents/materials/analysis tools: X. Xue ZGW Shao-wen Tang Jie Wang. Wrote the paper: LC JWZ Shai-di Tang.