miR-23a, miR-146a and miR-301a confer predisposition to Vogt-Koyanagi-Harada syndrome but not to Behcet’s disease

Ninety-eight miRNAs are involved in the immune response. However, the genetic roles of these miRNAs remain unclear in Behcet’s disease (BD) and Vogt-Koyanagi-Harada (VKH) syndrome. This study aimed to explore the association and functional roles of copy number variants (CNV) in several miRNAs with BD and VKH syndrome. Genotyping of CNVs was examined by TaqMan PCR. The expression of miR-23a, transfection efficiency and cytokine production were measured by real-time PCR, flow cytometry or ELISA. First, replication and combined studies for miR-23a, miR-146a and miR-301a demonstrated a similar association with VKH syndrome (Combined: P = 5.53 × 10−8; P = 8.43 × 10−31; P = 9.23 × 10−8, respectively). No association of CNVs of the above mentioned miRNAs was observed in BD patients. mRNA expression of miR-23a showed a positive association with its copy numbers. Additionally, individuals with high copy number of miR-23a show an increased production of interleukin-6 (IL-6), but not IL-8 and monocyte chemoattractant protein-1 (MCP-1) by stimulated PBMCs. miR-23a transfected ARPE-19 cells modulated the production of IL-6 and IL-8, but not MCP-1. Our results suggest that CNVs of miR-146a, miR-23a and miR-301a confer susceptibility to VKH syndrome, but not to BD. The contribution of miR-23a to VKH syndrome may be mediated by increasing the production of IL-6.

Copy number variations (CNVs) belong to structural variations of DNA, and capture more genetic information than SNPs (5%-10% vs < 1%). CNVs of various genes are linked to gene expression 37 and are associated with uveitis 14,38,39 . However, to our knowledge no reports have appeared concerning miRNAs CNVs and uveitis. We therefore performed a study to examine whether the CNVs in miRNAs were possibly implicated in the pathogenesis of uveitis. Ninety-eight miRNAs are implicated in immune response (Supplementary Table 1), however, only eight immune related miRNAs including miR-146a, miR-23a, miR-22, miR-143, miR-205, miR-9, miR-132 and miR-301a known to have CNVs in the human genome variants database were addressed in the present study. Our results showed that copy number variants of miR-146a, miR-23a and miR-301a are significantly associated with VKH syndrome, but not with BD. Functional studies showed that miR-23a may be implicated in the development of VKH syndrome by the upregulated production of inflammatory cytokines such as IL-6.

Results
Clinical feature of patients with BD and VKH syndrome. The clinical characteristics of the uveitis patients were assessed at the time of diagnosis and summarized in Table 1.

First stage study for copy numbers of investigated miRNAs between patients and controls.
A total of 383 VKH syndrome patients, 377 BD patients and 660 healthy controls were enrolled in the first stage study. The most common counts of miRNA gene copy numbers are 2 copies, so comparison between patients and controls was performed after a division into three groups: less than 2, equal to 2 or more than 2 gene copy numbers. The frequencies of copy numbers of the investigated miRNAs (miR-9-3, miR-22, miR-23a, miR-143, miR-146a, miR-205, miR-301a) was not different in BD patients and healthy controls ( Table 2). Increased frequencies of more than 2 copies for miR-23a and miR-146a were found in VKH syndrome patients as compared with normal controls (miR-23a > 2, P = 2.00 × 10 −4 , OR = 2.9; miR-146a > 2, P = 1.21 × 10 −11 , OR = 51.98, respectively) ( Table 2). An increased frequency of less than 2 copies for miR-301a was found in VKH patients (P = 4.71 × 10 −4 , OR = 3.8) ( Table 2).There was no significant difference for the remaining miRNAs including miR-9-3, miR-22, miR-143 and miR-205 copy numbers between VKH patients and healthy controls ( Table 2). Two different probes were used to test miR-132 CNVs, however, the result was unreliable and data for miR-132 were excluded from the study.

The influence of miR-23a copy number variants on cytokine production in peripheral blood mononuclear cells (PBMCs).
Since miR-23a CNVs correlate with the expression of miR-23a at the transcriptional level, further experiments were performed to investigate whether CNVs of miR-23a affected cytokine production. Seventy-seven normal controls, which are different from those used for the expression study, were enrolled in the examination of cytokine production. As no individual with less than 2 copies was found in the enrolled samples in this study, only groups with more than 2 or equal to 2 gene copy numbers were examined. Previous studies demonstrated that the expression of IL-6, IL-8 and MCP-1 were significantly increased in the vitreous fluid, aqueous humor or serum of uveitis patients [40][41][42] . Additionally, the treatment with anti-IL-6 antibody could inhibit the severity of a mouse model of uveitis 43 . These three cytokines were therefore examined in this study. An increased production of IL-6 by stimulated PBMCs was found in individuals with a high copy number of miR-23a (P = 4.68 × 10 −9 ) (Fig. 2). There was no significant association between the production of IL-8, MCP-1 by stimulated PBMCs and CNV of miR-23a (Fig. 2).

Optimal concentration of miR-23a mimics and inhibitors in human retinal pigment epithelial cell line (ARPE-19).
Results of flow cytometry showed a successful transfection efficiency of 5Cy3-labeled miRNA mimic control in human RPE cells, which was up to 98.1% (Fig. 3). According to the manufacturer's reagent supplies manual, we chose concentrations of 50 nM and 100 nM for miR-23a mimics and 100 nM and 200 nM for miR-23a inhibitors. To obtain the optimal concentration for transfection, expression of miR-23a was detected by quantitative RT-PCR (qRT-PCR) and compared between different groups. ARPE-19 cells transfected with miRNA mimics at a final concentration of 100 nM showed an optimal overexpression of miR-23a (Fig. 4). For miR-23a inhibitors, concentration of 200 nM had a stronger inhibitory effect than 100 nM. Therefore, retinal pigment epithelium (RPE) cells were transfected with miRNA mimics and inhibitors at a final concentration of 100 nM and 200 nM respectively (Fig. 4).

miR-23a promotes the expression of IL-6 and IL-8 in human ARPE-19 cells.
To elucidate the role of miR-23a in regulating the cytokine production of IL-6, IL-8 and MCP-1 in human RPE cells, cells were transfected with miR-23a mimics as well as inhibitors to affect the endogenous miR-23a level. Results of ELISA showed an increased IL-6 and IL-8 production, but not for MCP-1 in supernatants of RPE cells transfected with miR-23a mimics (P < 0.001) (Fig. 5). Additionally, transfected RPE cells with miR-23a inhibitors showed downregulated expression of IL-6 and IL-8, but not MCP-1 (Fig. 5).

Discussion
The present study aimed to explore the association of CNVs of eight miRNAs with VKH syndrome and Behcet's disease. The results showed that three miRNAs including miR-23a, miR-146a and miR-301a are associated with VKH syndrome but not with BD. No association was found between the remaining CNVs in four miRNAs including miR-9-3, miR-22, miR-143, miR-205 and VKH syndrome or BD. The functional study showed that mRNA expression of miR-23a was positively related with its copy numbers. Additionally, increased production of IL-6, but not IL-8 and MCP-1 by stimulated PBMCs was found in the group with a high copy number of miR-23a.  and MCP-1 (c) by stimulated PBMCs from healthy controls carrying different gene copies of miR-23a (miR-23a < 2: n = 72, miR-23a > 2: n = 5). Significance was analyzed by Student's t test or two independent samples Nonparametric test.  Consistently with the result in PBMCs, miR-23a mimics significantly increased the production of IL-6 and IL-8, but not MCP-1 in RPE cells, whereas miR-23a inhibitors suppressed cytokines such as IL-6 and IL-8. Many miRNAs have been shown to be involved in the development of immune diseases such as multiple sclerosis (MS), rheumatoid arthritis and uveitis 22,35,44,45 . However, the genetic role of these miRNAs in uveitis has to our knowledge not yet been addressed and was therefore the purpose of the study described here. The tested candidate miRNAs were principally chosen due to their role in modulating the immune response. Many immune-related miRNAs have been described. However, only eight miRNAs including miR-146a, miR-23a, miR-22, miR-143, miR-205, miR-9, miR-132 and miR-301ashow gene copy number variation and probes are available to test this.
To the best of our knowledge, this is the first study to identify an association of miRNAs CNVs with human disease. The results showed that high copy numbers of miR-146a, miR-23a and low copy number of miR-301a confer susceptibility to VKH syndrome. Inconsistent with the results in VKH syndrome, no association of these three miRNAs CNVs were found for BD. The discrepancy in the association between VKH syndrome and BD may be partially due to the fact that these three miRNAs including miR-146a, miR-23a and miR-301a may be involved in the development of autoimmune disease rather than inflammatory disease. VKH syndrome is generally considered as autoimmune disease directed against melanocytes, whereas BD is now defined as an autoinflammatory disease [46][47][48] . miR-23a plays a suppressive role on CD8 + T cell effector functions 49 , suggesting the involvement of this miRNA in immune disease. The level of miR-23a was decreased in serum of MS patients 44 , and its rs3745453 variant was associated with susceptibility for MS 50 . Inconsistent with the aforementioned result, this miRNA was found to be upregulated in gastric cancer 51 , suggesting a variable disease involvement for miR-23a. miR-146a acts as a negative regulator of immune cell activation 52 by repressing its target genes including TRAF6 and IRAK1. A promoter SNP of miR-146a confers risk for SLE by modulating its expression 53 . Zhou et al. 8 also showed the association of a miR-146a SNP with Behcet's disease, but not with VKH syndrome. However, there are no reports on the association of miR-146a CNV with human disease. Our present study showed that miR-146a CNV may contribute to the risk for human diseases such as VKH syndrome but not BD. These results suggest that certain genetic variants of miR-146a such as gene polymorphisms and copy number variants can both be involved in the development of uveitis.
As the frequency of less than 2 copies of miR-301a is as low as 1.7% in controls, only one sample with less than 2 copies of miR-301a was found in our population of healthy controls. The relationship between the expression of miR-301a and its copy numbers was therefore not examined in this study. Additionally, our group tested the functional role of a genetic variant of miR-146a in uveitis in a previous study 8 . We therefore focused on the functional role between miR-23a CNV and its expression and the effect of miR-23a CNV on the production of inflammatory cytokines. Since there is a marked heterogeneity in the clinical presentation of VKH syndrome and in view of the general treatment regimen of these patients that includes immunosuppressive drugs, we performed functional studies using PBMCs from genotyped normal controls. A positive association was found between miR-23a CNV and its expression in PBMCs. Individuals with a high copy number of miR-23a showed an increased production of IL-6 in PBMCs. Additionally, since the RPE plays an important role in the pathogenesis of uveitis, we also examined the functional role of miRNA in these cells. Similar to the results obtained in PBMCs, we found enhanced productions of IL-6 and IL-8 in transfected RPE cells with miR-23a mimics, whereas IL-6 and IL-8 production was suppressed in RPE cells that were transfected with miR-23a inhibitors. Our results are in agreement with a recent study that showed that miR-23a increased the expression of IL-6 in LPS-stimulated macrophages 54 . Contrary to the aforementioned findings, previous studies showed no significant effect of miR-23a on IL-6 levels in trophoblast cells 55 and a suppressive role of miR-23a on IL-6 production by purified CD8 + T cells 56 . These studies suggest that the role of miR-23a in disease may affect different pathways which may depend on the cell type investigated such as CD8 + T cells, PBMCs, RPE cells or macrophages. A previous study also showed that miR-23a has a suppressive role on the production of interferon-γ (IFN-γ ) 49 . We tested the expression of IFN-γ by RPE cells but the expression level was below the detection limit of our assay.
It is worthwhile to point out that our experiments concerning miR-23a CNV on cytokine production are unbalanced for the different CNV groups, which may affect the interpretation of these functional studies. Our cDNA bank of healthy controls is however much smaller than our DNA bank and since the frequency of the miR-23a > 2 individuals is relatively small. We were only able to enroll 5 controls for this group in our expression study. Further studies using larger samples are need to further examine the functional role of miR-23a.
In conclusion, we found that high copy numbers of miR-146a, miR-23a and a low copy number of miR-301a confer risk for VKH syndrome, but not for BD. miR-23a may be involved in the development of VKH syndrome by increasing the production of inflammatory cytokines such as IL-6.

Materials and Methods
Study population. Individuals selected for this study included a total of 377 BD patients, 1,243 VKH patients and 2,291 healthy controls. All individuals were Han Chinese and were recruited via the First Affiliated Hospital of Chongqing Medical University (Chongqing, China) or the Zhongshan Ophthalmic Center, Sun Yat-sen University (Guangzhou, China). The revised diagnostic criteria 2001 for VKH syndrome and the International Study Group for BD were used to make the diagnosis of VKH syndrome and BD, respectively 57,58 . Patients with a doubtful diagnosis were excluded from the study. The healthy controls were age-, and ethnicity-matched with the patients as well as free of any intraocular or systemic inflammatory condition. Written informed consent was obtained from every participant. The present study was approved by the Ethics Committee of the First Affiliated DNA extraction and genotyping. Genomic DNA was extracted from peripheral blood by using the QIAamp DNA Blood Mini Kit (QIAGEN,Valencia, CA). Copy numbers of examined genes were determined by the TaqMan-based qPCR which was performed in 96-well optical plates on a 7500 real time PCR system (Applied Biosystems). Nine TaqMan probes (labeled with FAM) specific for our studied genes were as follows: miR-9, Hs05330221 -cn; miR-22, Hs01095539 -cn; miR-23a, Hs04021842 -cn; miR-143, Hs03578207 -cn; miR-146a, Hs06722002 -cn; miR-205, Hs07483345 -cn; miR-301a, CXN1ENO; miR-132, Hs03965136_cn or CCN1FDA. RNaseP (labeled with VIC) was used as the control assay for normalization. Both TaqMan probes and RNaseP were obtained from Applied Biosystems (ABI, FosterCity, CA). Standard thermal cycler conditions were set as follows: 95 °C for10 min, 40 cycles of 95 °C for 15s and 60 °C for 1 min. Measurements of CNVs for each sample were performed in three replicates.

Transfection of miRNA mimics and inhibitors in RPE cells. ARPE-19 cells were obtained from the
American Type Culture Collection (ATCC, Manassas, VA), and cells at passages 21 to 25 were used for the present experiment. ARPE-19 cells were cultured in DMEM/F12 medium (Invitrogen, Carlsbad, CA) supplemented with 10% FBS, 100 U/ml penicillin and 100 ug/ml streptomycin in an incubator at 37 °C in 5% CO 2 . ARPE-19 cells were seeded in 12-well plates at 5 × 10 5 cells/well and cultured for 24 h to reach 50-70% confluence and then transfected with miR-23a mimics or inhibitors (RIBOBIO, Guangzhou, China) using Lipofectamine RNAiMAX (Invitrogen) and OPTI-MEM I reduced serum medium (Invitrogen) according to the manufacturer's protocol for 24 h. The cells were harvested for RNA analyses. Negative control mimics or inhibitors were used as matched controls.
To detect the effect of miR-23a on cytokine production by ARPE-19 cells, the cells were transfected with miR-23a mimics or inhibitors for 6 h and then stimulated with LPS (1 ug/ml, Sigma-Aldrich) for 18 h. The culture supernatants were collected for detection of IL-6, IL-8 and MCP-1 by ELISA.
Flow cytometry. To evaluate the transfection efficiency of miRNAs in human RPE cells, cells were gently washed with PBS after transfection with 5Cy3 labeled mimics for 24 h and directly analyzed by flow cytometry. Flow cytometry was performed on a FACS Aria cytometer (BD Bioscience, San Diego, CA) and the data were processed using FlowJo software (Treestar, Inc., San Carlos, CA). Statistical analysis. The differences in copy numbers of miRNAs between patients and controls were analyzed by chi-square test using SPSS (v. 17.0; SPSS Inc., Chicago, IL). Bonferroni correction was applied to correct for multiple comparisons. The expression of miRNAs and various cytokines was analyzed by independent samples t test or two independent samples Nonparametric test using SPSS 17.0 software. Corrected P values less than 0.05 were considered significant.