Genetic analysis of innate immunity in Behcet’s disease identifies an association with IL-37 and IL-18RAP

Interleukin-1 (IL-1) and the IL-1 receptor (IL-1R) family play an important role in the pathogenesis of inflammatory diseases. This study aimed to investigate the association between single nucleotide polymorphisms (SNP) of IL-1 and IL-1R family genes with Vogt-Koyanagi-Harada (VKH) and Behcet’s disease (BD) in Han Chinese. The case-control study was divided into two stages and included 419 VKH cases, 1063 BD cases and 1872 healthy controls. The MassARRAY platform (Sequenom), iPLEX Gold Assay and TaqMan SNP assays were used to score genotypes of 24 SNPs. The expression of IL-37 and IL-18Rap was measured by ELISA and real-time PCR in genotyped healthy individuals. A significantly lower frequency of the AG genotype, and a higher frequency of the GG genotype and G allele of IL-37/rs3811047 were observed in BD as compared to controls. AA genotype and A allele frequency of IL-18RAP/rs2058660 was significantly decreased in BD as compared to controls. Functional studies performed in healthy controls showed that rs3811047 AG genotype carriers had a higher IL-37 gene expression in peripheral blood mononuclear cells (PBMCs) than GG carriers. GG carriers showed a higher cytokine expression as compared to AG carriers. No association was detected between the tested SNPs and VKH.

Genotype and allele frequencies of the examined SNPs in cases and controls in the first-phase study. In the first phase study, we tested the association between 24 SNPs of 10 IL-1 and IL-1R family genes with BD (n = 416) or VKH (n = 419) as compared to healthy controls (n = 627). The results demonstrated significant differences between BD cases and healthy controls for only two SNPs (rs2058660 and rs3811047) in two genes (IL-18Rap and IL-37) ( Table 2). No association was observed for VKH for any of the 24 SNPs tested (Supplemental Tables 1 and 2).
Compared to controls, the frequency of the IL-37/rs3811047 AG genotype was significantly lower in BD (P c = 4.32 × 10 -4 , OR = 0.522). A significant increase in the frequency of the GG genotype and G allele was observed (P c = 1.66 × 10 −3 , OR = 1.800 and Pc = 2.19 × 10 −2 , OR = 1.529, respectively). The frequency of the IL-18RAP /rs2058660 AA genotype and A allele was significantly decreased in BD as compared to healthy controls (Pc = 2.08 × 10 −2 , OR = 0.603 and Pc = 1.17 × 10 −3 , OR = 0.685, respectively). The frequencies of the IL-37/rs3811047 AA genotypes and the frequencies of the IL-18RAP/rs2058660 AG and GG genotypes were not significantly different between BD cases and healthy controls ( Table 2).
Genotype and allele frequency of the examined SNPs in cases and healthy controls in the second phase and combined study. Since IL-37/rs3811047 and IL-18RAP/rs2058660 displayed a significant association with BD in the first phase exploratory study, a confirmatory second phase study was performed for these two SNPs in a separate set of BD patients (n = 647) and healthy controls (n = 1245). The data again showed that the frequency of the IL-37/rs3811047 GG genotype and G allele was significantly higher in BD as compared to controls (Pc = 3.60 × 10 −4 , OR = 1.657; P c = 2.45 × 10 −5 , OR = 1.639). The AG genotype was significantly lower in BD (P c = 3.25 × 10 −2 , OR = 0.672). The frequency of the AA genotype and A allele of IL-18RAP/rs2058660 was significantly decreased in BD cases as compared to controls (Pc = 3.74 × 10 −3 , OR = 0.621 and Pc = 2.51 × 10 −2 , OR = 0.788, respectively) ( Table 2). Combination of the data from the first and second phase studies further strengthened the association between IL-37/rs3811047 and BD (GG genotype: Pc = 7.27 × 10 −8 , OR = 1.693; AG genotype: Pc = 3.67 × 10 −6 , OR = 0.616; G allele: Pc = 7.68 × 10 −8 , OR = 1.584) and the observed association between IL-18RAP/rs2058660 and BD (AA genotype: Pc = 1.27 × 10 −5 , OR = 0.620; A allele: Pc = 5.94 × 10 −6 , OR = 0.750) ( Table 2).
The Influence of IL/37rs3811047 and IL-18RAP/rs2058660 genotypes on mRNA and cytokine expression. To study the effect of genotype on mRNA or cytokine expression, we performed experiments with peripheral blood mononuclear cells (PBMCs) taken from healthy genotyped individuals instead of using patients. This approach was chosen in view of the ongoing inflammation in these patients and the fact that they often were receiving immunosuppressive drug therapy. Real-time PCR analysis was applied to determine whether rs3811047 and rs2058660 polymorphisms altered the mRNA expression of the IL-37 or IL-18RAP gene, respectively. Functional studies using PBMCs from healthy  Table 2. Genotype and allele frequencies of IL-18Rap and IL-37 polymorphisms in BD and healthy controls. SNP = single-nucleotide polymorphism; BD = 5% confidence interval; Pc = Bonferroni corrected p value.
genotyped donors indicated that rs3811047/AG carriers had a higher expression of the IL-37 gene as compared to GG carriers (Fig. 1). Polymorphisms of rs2058660 did not significantly affect the IL-18RAP expression in PBMCs from healthy individuals (Fig. 2).
The effect of IL-37/rs3811047 genotype on cytokine production was tested in lipopolysaccharide (LPS) stimulated PBMCs taken from genotyped healthy controls. An enzyme-linked immunosorbent assay (ELISA) was applied to examine the level of tumor necrosis factor (TNF-α ), IL-1β and IL-6 in 72h cell culture supernatants. GG carriers showed a two-fold higher level of IL-1β and TNF-α as compared to that detected in AG carriers, whereas IL-6 production was only modestly increased (Fig. 3). We did not test AA carriers since the frequency of this genotype is very low (3.5%; see data of controls in Table 2).

Discussion
This study shows that IL-37 and IL-18RAP gene polymorphisms are associated with BD but not with VKH uveitis in Han Chinese. No association could be found for gene polymorphisms in other members of the IL-1 and IL-1R family with either BD or VKH. Functional studies showed that carriers of the GG risk genotype of IL-37/rs3811047 had a higher mRNA expression of the gene and that PBMCs from such carriers showed a higher pro-inflammatory cytokine response compared to carriers of the AG genotype. An explanation for the novel association of BD with the IL-18RAP/rs2058660 gene polymorphism could not yet be supported by functional assays with PBMCs, although a recent report did show that this polymorphism was linked to a low expression of the β -chain of this factor, causing a lower response to IL-18, as detected both on the RNA and protein level of CCL3, CCL20, TNF-α and CXCL8 in granulocytes 29 . Another study that used human monocyte-derived macrophages showed that polymorphisms in this region had a marked effect on the expression of cell-surface IL-18RAP protein, leading to an altered cytokine secretion following stimulation of a wide range of pattern-recognition receptors 30 . The IL-18RAP/rs2058660 tag SNP belongs to a linkage block on chromosome 2.q12 that has a strong association with leprosy and Crohn's disease(CD) 31 .
Our study confirms earlier reports on the association of IL-37/rs3811047 with immune mediated diseases such as PsA 14 , RA 31 and Graves' disease (GD) 32 . To our knowledge this is the first report addressing the association of IL-37 gene polymorphisms with uveitis. The fact that only an association was found with ocular BD and not with VKH uveitis suggests that our finding cannot be generalized for all uveitis entities and that further research is necessary to study the association of IL-37 with other uveitis types. Why only an association was found with BD and not with VKH may be caused by differences in disease causing mechanisms. BD is considered an autoinflammatory disorder caused by an aberrant response against microbial antigens, whereas VKH is mediated by a loss of immunological tolerance against melanin associated antigens. An association with factors regulating innate immunity is therefore more plausible for BD than for VKH. Earlier studies from our group however showed that IL-37 expression in PBMCs was decreased in both VKH and BD patients and that it was related to disease activity 33,34 . Further study in VKH patients showed that treatment with corticosteroids and cyclosporine A was associated with an increased expression of IL-37 34 . A decreased IL-37 expression in BD was confirmed recently 35 , although other groups reported that IL-37 expression was increased in immune-mediated inflammatory conditions such as RA, PsA, AS, GD and SLE [36][37][38][39][40][41] . The reasons for these discrepancies concerning IL-37 expression between uveitis and other autoimmune diseases is not clear and deserves further study.
IL-37 is coded by a gene that belongs to the so called IL-1 gene cluster that is located in a 360 kb region on human chromosome 2q13 42,43 . Other proteins coded by this region include IL-1α , IL-1β , IL-1 receptor antagonist, IL-36α , IL-36β and IL-36γ , IL-36 receptor antagonist and IL-38. We were not able to find an association between BD with other genes than IL-37 in this cluster, although others have reported an association between BD and rs16944 in the IL-1β locus 43 . The rs16944 locus is in strong linkage disequilibrium (LD) with the rs1143627 locus used in our study. The discrepancy may be due to a different BD patient subgroup, since only 54.6% of these Tunisian patients showed ocular involvement, whereas all our patients had intraocular inflammation.
There is ample evidence that IL-37 suppresses both innate as well as adaptive immune responses 44 . It is expressed in PBMCs and can be induced by Toll-like receptor agonists. IL-37 interacts with a heterologous receptor combination comprised of IL-18Ra and IL-1R8 which leads to an inhibition of the activation of inflammatory pathways in the cell 21 .
An earlier Genome-wide analysis (GWAS) in both BD and VKH did not reveal an association with IL-37 or IL-18RAP, but showed MHC class I, IL10, IL23R/IL12RB2, HLA-B*51 and ERAP1 associations with BD 45,46 . This may be due to the strong p value threshold used in GWAS, sample size or the use of insensitive tag SNPs. It should be noted that the observed associations were present in common variants of these genes and that the odds ratios (OR) were modest. The data are nevertheless supportive of an inflammatory pathway that involves triggering of toll like receptors (TLRs) 47 followed by the activation of the IL-1 family. A pro-inflammatory genotype involving a myriad of cytokines and their receptors may render an individual at risk for a disease like BD, despite the fact that the OR for each individual cytokine might be low 48 .
The current study has several limitations. Although great effort was made to exactly match patients and controls for gender this was not completely achieved due to the large male preponderance in our BD group. On the other hand, we did not detect differences in genotype frequencies after stratification for gender (data not shown). Gender differences have been noted for certain geographical regions and males often suffer from the more severe forms of BD uveitis 49 . Validation of our findings is therefore necessary in other ethnic samples. All our BD patients had uveitis and it is not known whether our findings are restricted to patients with ocular BD. Further studies are therefore needed in BD patients recruited via other disciplines than ophthalmology, such as dermatology or rheumatology. A further restriction of our findings concerns the functional assays we performed. These were restricted to the testing of PBMCs, whereas recent studies suggest that addressing the response in neutrophils or macrophages might be more suitable 50,51 .
In conclusion, this study reports a novel association between IL-37/rs3811047 and IL-18RAP/ rs2058660 polymorphisms with BD in Han Chinese, which supports the important role of the IL-1 pathway in the pathogenesis of this disease and may provide a future target for its treatment.

Materials and Methods
Study Population. The study was divided into two phases: an exploratory and a confirmatory phase. A total of 3354 unrelated out-patients including 419 VKH cases, 1063 BD cases and 1872 healthy controls seen at the First Affiliated Hospital of Chongqing Medical University between April 2008 and August 2015 were included. For the first phase, 419 VKH patients and 416 BD patients as well as 627 healthy controls were tested. For the second phase, a different set of 647 BD patients and 1245 healthy controls were investigated. All subjects, either patients or controls, were Han Chinese. The controls were randomly selected among normal unrelated individuals with suitable age, ethnicity and geographic origin. Diagnostic criteria for BD and VKH disease strictly followed the International Study Group for BD and the modified diagnostic standards for VKH disease [51][52][53] . The prominent features of the BD and VKH cases in this study are shown in Table 1.
Ethical considerations. Before the collection of blood, all the investigated individuals had signed the informed consent. The investigation protocols obtained the approval of the Clinical Research Ethics Committee of the First Affiliated Hospital of Chongqing Medical University. All experiments were conducted in accordance with the approved guidelines and regulations and the study was conducted according to the tenets of the Declaration of Helsinki.
All blood samples were stored in 3.2% sodium citrate-treated tubes at − 80 °C, from which the DNA of BD and VKH patients and controls was extracted by utilizing the QIA amp DNA Blood Mini Kit (QIAGEN, CA). Subsequently, the collected DNA specimens were kept at − 20 °C. The polymerase chain reaction (PCR) was performed on a 9700 thermocycler of the ABI Gene Amp PCR System according to the manufacturer's instructions (Applied Biosystems, Foster City, CA). The MassARRAY platform (Sequenom, San Diego, CA) and iPLEX Gold Assay were used to score genotypes. The MassArray Designer of Sequenom was utilized to design the PCR as well as extended primers of the respective SNPs. Experimental data were analyzed through TYPER software Cells isolation and culture, RNA preparation and real-time quantitative PCR. Peripheral blood mononuclear cells (PBMCs) were isolated from blood obtained from healthy controls using Ficoll-Hypaque density-gradient centrifugation. TRIzol reagent (Invitrogen, Carlsbad, CA)) was used to extract RNA from PBMCs. The PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa, Japan) was used to synthesize cDNA according to the manufacturer's instructions. The Power SYBR Green PCR Master MIX (Biosystems, UK) was applied to test IL-37/IL-18RAP expression with the ABI 7500 Real-Time PCR System. The following primers of IL-37/IL-18RAP and ß-actin were used: IL-37 (forward: 5′ -AGTCATCCATCCCTTCAGC-3′ and reverse: 5′ -CCCAACAGGCTCATTACAA-3′ ), IL-18RAP (forward: 5′ -CAGATATTCTGGATCCTGTCGAG-3′ and reverse: 5′ -TGCTTTGCAGCTAATAGTTAAAGG-3′ ), β -actin (forward: 5′ -CGAGAAGATGACCCAGATCATG-3′ and reverse: 5′ -CAGAGGCGTACAGGGATAGCA-3′ ). Expressions of IL-37/IL-18RAP were calculated and compared with the expression of β -actin through the 2-Δ Δ CT approach. Measurement of cytokines by ELISA. Human Duoset ELISA kits (R&D Systems) were used to analyze the level of IL-6, IL-1β and TNF-α in the culture supernatants of PBMCs following stimulation with LPS (100ng/ml; Sigma-Aldrich, USA) for 72 hours.
Statistical analysis. The Chi-square (χ 2 ) test was adopted to examine the Hardy-Weinberg equilibrium(HWE) in healthy samples while the genotype frequency was estimated by direct counting. The χ 2 test was used to examine the differences between patients and healthy controls with regard to allele and genotype frequencies with SPSS17.0 statistical software package (version17.0, SPSS, Chicago, IL). Correction for multiple comparisons was performed using the Bonferroni method whereby the p value was multiplied with the number of comparisons (P corrected (P c )). A P c < 0.05, was taken as statistically significant. The Mann-Whitney test, one-way analysis of variance (ANOVA), and Kruskal-Wallis H test and Bonferroni correction were adopted by SPSS Statistics 17.0 software, where p values smaller than 0.05 were regarded as significant.