Polymorphisms in the Th17 cell-related RORC gene are associated with spontaneous clearance of HCV in Chinese women

In the present study, 190 HBsAg-negative, anti-HCV-positive, Han Chinese females were identified by screening a total of 1252 residents (80% of the local population) in Wangying Village, Shangcai County, Henan province, in August 2009. More than 90% of patients were former plasma donors, and the remaining individuals composed of their parents, spouses, or children. Samples were tested for anti-HCV antibody using the Architech anti-HCV system (Abbott Diagnostics, USA), and those with signal/cut-off ratios between 1.0 and 5.0 were confirmed by RIBA assay (HCV BLOT 3.0, MP diagnostics, USA). Plasma HCV viral load was determined using the Abbott RealTime HCV Amplification Kit (Abbott Molecular Inc., USA) according to the manufacturer’s instructions. Spontaneous resolvers were defined as subjects who were positive for HCV antibody and negative for plasma HCV RNA with no history of HCV-specific treatment. HCV RNA-negative status was confirmed in a second sample collected in 2012 and/or 2013. Chronic HCV carriers were defined as those patients positive for both HCV antibody and plasma HCV viral load and also confirmed by the follow-up detection. No participants had received any type of HCV-specific antiviral therapy. All subjects were screened for HIV-1 infection status. Anti-HIV-1 antibody status was initially tested by ELISA assay (GBI biotech Co., Ltd., Beijing, China) and confirmed by HIV Blot 2.2 WB assay (Genelabs Diagnostics, Singapore). All HIV-positive patients had received first-line antiretroviral therapy for 6~ 8 years, regularly or intermittently. None of the participants received any types of anti-HCV treatment.

Of 190 female anti-HCV-positive individuals, HIV^pos women with chronic HCV infection (n = 53) were excluded from the study, as HIV infection could affect the ability of some individuals to spontaneously resolve their HCV infection. The final case number of primary cohort was 137, which consisted of a “Chronic” group (HIV^neg HCV carriers, n = 64) and a “Resolved” group (n = 73, including 45 HIV^neg HCV resolvers and 28 HIV^pos HCV resolvers). Overall, the cohort exhibited a similar high frequency of IFNL3 favorable genotypes to Han Chinese nationality (Additional file 1: Table S1), indicating that the study was based on a representative cohort of the Chinese population [12, 16]. To eliminate the influence of IFNL3 SNPs on viral clearance, a sub-cohort carrying favorable IFNL3 genotypes (rs12979860CC, rs8099917TT, rs12980275 AA) was constructed and analyzed. The IFNL3 favorable sub-cohort included 54 and 67 participants in the “Chronic” and “Resolved” (HIV^neg, n = 43; HIV^pos, n = 24) groups, respectively. A flow diagram for cohort construction was provided as Fig 1. The clinical and biochemical characteristics of the primary cohort were presented in Table 1.

A total of 111 SNPs in 23 genes encoding HCV co-receptors (CD81, SCARB1, CLDN1, OCLN, NPC1L1, APOE, LDLR), transcription factors(IRF3, RORC, TBX21, FOXP3, BCL6), Toll-like receptors (TLR3, TLR7, TLR9), co-stimulating molecules (ICOS, CXCR5, CD40LG), and cytokines(IL4, IFNG, IL21, IFNL3, CXCL13) were selected for SNP analysis (Additional file 1: Table S2). The selected SNPs met at least one of the following criteria:1) a reported minor allele frequency > 10% in the Han Chinese population according to SNP browser software 4.0 (Applied Biosystems), with reference to the NCBI SNP database (http://​www.​ncbi.​nlm.​nih.​gov/​projects/​SNP/​); 2) published evidence that the SNP was associated with disease.

Genomic DNA extracted from whole blood samples using a DNeasy Blood & Tissue kit (Qiagen, USA) was dissolved in sterile double distilled water and stored at − 20 °C until use. DNA purity was checked and DNA samples with 260/280 ratios< 1.7 were re-purified. SNP genotyping was performed using the iPLEX Sequenom MassARRAY system (Sequenom Inc., USA).

Statistical and graphical analyses were performed using GraphPad Prism 5.0, Microsoft Excel 2007, or SPSS 20.0. The allele frequency and genotype distributions of each SNP were descriptively summarized as numbers of cases and frequencies. Chi-square (χ²) and Fisher’s exact tests were used to examine differences in frequencies of individual SNPs between HCV carriers and spontaneous resolvers. SNP-specific deviation from Hardy–Weinberg Equilibrium (HWE) in the whole study population was tested using a χ² test in SHEsis software, and SNPs with HWE violation (P < 0.01) were excluded. Pairwise D′ and r² measures of linkage disequilibrium for RORC SNPs were calculated using SHEsis (http://​analysis.​bio-x.​cn/​myAnalysis.​php). Specific parameters were set as previously described [27]. P-values, odds ratios (ORs), and 95% confidence intervals (95% CIs) were used for association analysis. All P-values were two-tailed, and were considered significant at < 0.05.

Genotype distributions and allele frequencies were calculated for 111 candidate SNPs in both the “Chronic” and “Resolved” groups of the primary cohort (Additional file 1: Table S2). The frequencies of favorable IFNL3 genotypes were higher in “Resolved” group than “Chronic” group (resolved vs. chronic: rs12979860, CC 96% vs. 86%; rs8099917, TT 95% vs. 91%; and rs12980275, AA 96% vs. 86%), although only the difference in the frequency of the rs12979860 C allele was statistically significant (P = 0.043) (Table 2).

Of the 22 remaining genes tested (excluding IFNL3), only SNPs rs9826 and rs1521177, located in non-coding regions of RORC, exhibited frequency differences between the “Resolved” and “Chronic” groups (rs9826: P = 0.024 for CC/TT/CT, P = 0.015 for C allele/T allele, OR = 1.97, 95% CI: 1.14-3.41; rs1521177: P = 0.017 for GG/TT/GT, P = 0.006 for G allele/T allele, OR = 2.18, 95% CI: 1.24-3.82)(Table 2). The rs9826 TT and rs1521177 TT genotypes were associated with improved viral clearance. Overall, the primary cohort exhibited a similar distribution of RORC genotypes to Han Chinese nationality (Additional file 1: Table S3). The Hardy–Weinberg Equilibrium (HWE) test of three IFNL3 SNPs and two RORC SNPs in the primary cohort were tested, and no SNPs with HWE violation were presented in the study (Additional file 1: Table S4).

In the present study, a minority of “Resolved” individuals were co-infected with HIV (HIV^pos resolved, n = 28) and very similar IFNL3 (rs12979860, rs8099917, and rs12980275) and RORC (rs9826 and rs1521177) SNP genotype distributions were identified among HIV^neg, HIV^pos, and total resolved individuals (Additional file 1: Figure S1). No association was observed between RORC SNPs (rs9826 and rs1521177) and HCV RNA levels in HIV-uninfected “Chronic” individuals (Additional file 1: Figure S2). In addition, we did not find any differences in genotypes and allele frequency distributions of RORC SNPs (rs9826 and rs1521177) between HIV^pos HCV carriers and HIV^pos HCV resolvers (Additional file 1: Table S5), and no associations between RORC SNPs and type of HCV genotypes (HCV 2a vs. HCV 1b) was found in HIV^neg HCV carriers (Additional file 1: Table S6).

In the IFNL3 favorable sub-cohort, cases with unfavorable IFNL3 genotypes in the “Chronic” (n = 10) and “Resolved” (n = 6) groups were excluded. Analysis of the sub-cohort indicated that significant differences between the two groups remained for rs1521177 (P = 0.040 for GG/TT/GT, P = 0.021 for G allele/T allele, OR = 2.00, 95% CI: 1.11-3.55), although no significant differences were identified for rs9826 (P = 0.092 for CC/TT/CT, P = 0.062 for C allele/T allele, OR = 1.74, 95% CI: 0.97-3.19) (Table 3). These data suggested that SNP alleles in RORC had an independent effect on HCV viral clearance, which couldnot be explained by the distribution of IFNL3 polymorphisms.

Linkage disequilibrium tests for RORC and IFNL3 SNPs in the primary cohort were shown in Additional file 1: Figure S3. The RORC SNPs, rs9826 and rs1521177, were determined to be in linkage disequilibrium (Primary cohort: D’ = 0.831, r² = 0.639; IFNL3 favorable sub-cohort: D’ = 0.812, r² = 0.645) (Additional file 1: Table S7). Two main haplotypes were present in the primary cohort (rs9826-T/rs1521177-T: 62% vs. 79%, chronic vs. resolved; rs9826-C/rs1521177-G: 27% vs. 16%, chronic vs. resolved) and the IFNL3 favorable sub-cohort (rs9826-T/rs1521177-T: 64% vs. 78%, chronic vs. resolved; rs9826-C/rs1521177-G: 26% vs. 17%, chronic vs. resolved). Overall, the presence of the rs9826-T/rs1521177-T haplotype was associated with a higher likelihood of spontaneous resolution in both the primary cohort (P = 0.0027, OR = 0.445, 95% CI: 0.26–0.76) and the IFNL3 favorable sub-cohort (P = 0.0117, OR = 0.485, 95% CI:0.28–0.86). In addition, the frequency of the rs9826-C/rs1521177-G haplotype was significantly different between the chronic and resolved groups in the primary cohort (P = 0.0315, OR = 1.893, 95% CI:1.05–3.40), suggesting that individuals carrying this haplotype was difficult in spontaneous resolution of HCV infection (Table 4).