Association between IL28B Polymorphisms and Spontaneous Clearance of Hepatitis B Virus Infection

Background/Aims Single-nucleotide polymorphisms (SNPs) near the interleukin 28B gene (IL28B; interferon [IFN]-λ-3) are associated with outcomes of chronic hepatitis C virus (HCV) and hepatitis B virus (HBV) infection treated with peginterferon (PEG-IFN) alpha-based antiviral therapy. In this study, we investigated the influence of IL28B polymorphisms on spontaneous clearance of HBV infection in a large Korean cohort. Methods Between January 2007 and June 2010, a total of 208 patients with chronic HBV infection and newly diagnosed HBV-related hepatocellular carcinoma were recruited as the CC group [HBsAg(+) for >6 months, anti-HBc(+), and anti-HBs(-)]. In addition, 351 organ donors were stratified into the UE group [n = 106; HBsAg(-), anti-HBc(-), and anti-HBs(-)] or the SC group [n = 245; HBsAg(-), anti-HBc(+), and anti-HBs(+)]. The SNaPshot ddNTP Primer Extension Kit (Applied Biosystems, Foster City, CA) was used for SNP detection. Direct full sequencing of the IL28B coding region was attempted. Results Regardless of group, rs12979860 CC was most frequently identified (85.0% in UE, 85.9% in SC, and 93.5% in CC, respectively), whereas rs12979860 TT was not identified in any group. Similarly, rs12980275 AA and rs8099917 TT were most frequently identified (≥85%) regardless of group, whereas rs12980275 GG was identified in only one subject in the SC group. In addition, rs8099917 GG was not identified. The prevalences of CC in rs12979860, AA in rs12980275, and TT in rs8099917 were significantly higher in the CC group when compared with the UE and SC group (all P<0.05). Among 19 novel SNPs in the IL28B coding region, the proportions of 6 SNPs were significantly different among the UE, SC, and CC groups (all P<0.05). Conclusions The SNP upstream of IL28B that has the strongest genetic association with HCV recovery has an inverse influence on HBV recovery. Additional studies are needed to understand the mechanisms of this SNP in HBV infection.


Introduction
Recently, several independent genome-wide association studies have shown that genetic polymorphisms at or near the interleukin 28B gene (IL28B; also known as interferon [IFN]-l-3), including rs12979860, rs12980275, and rs8099917, are associated with higher rates of sustained virologic response (SVR) in patients with chronic hepatitis C (CHC) treated with peginterferon (PEG-IFN) alpha and ribavirin [1][2][3][4]. Although the mechanism by which IL28B influences the response to PEG-IFN therapy has remained elusive, it is likely that the relationship is not specific to hepatitis C virus (HCV) infection. Indeed, the type of IFN coded for by IL28B, IFN-l, has been previously shown to be active against several other viruses including hepatitis B virus (HBV) [5,6]. Recently, a human study also demonstrated that IL28B polymorphisms were independently related to serological responses to PEG-IFN in a large global cohort of patients with hepatitis B e antigen positive chronic hepatitis B (CHB), similar to the results from chronic HCV infection [7,8].
In addition to this association between IL28B polymorphisms and treatment responses to PEG-IFN, the relationship between IL28B polymorphism and spontaneous clearance of HCV infection has also been reported in several previous studies [1,4,[9][10][11]. However, only one study, as a brief report, investigated the association between IL28B polymorphism and spontaneous clearance of HBV infection. The study concluded that the single-nucleotide polymorphism (SNP) 4 kb upstream of IL28B (rs12979860), which has a strong genetic association with HCV recovery, had no association with spontaneous recovery from HBV infection [13]. Although it is unknown how rs1297860 affects the activity of IL28B, the authors proposed that it presumably alters the immune function to HCV, but not HBV [12].
Therefore, the IL28B polymorphism seems to have similar influences on both the response to antiviral treatment and spontaneous clearance in HCV infection, whereas it exhibits different influences in HBV infection between treatment outcomes using PEG-IFN and spontaneous clearance [7,12]. However, data on this issue are extremely rare in terms of explaining the different influences of IL28B polymorphism on treatment outcomes using PEG-IFN and spontaneous clearance in HBV infection [7,12]. Furthermore, previous studies have only attempted to characterize the IL28B polymorphism upstream of the IL28B coding region. Thus, we first investigated the influence of the IL28B polymorphism on HBV infection in a large Korean cohort and evaluated whether there is any novel polymorphism in the IL28B genome, especially in the coding region.

Patients
Between January 2007 and June 2010, a total of 208 patients with CHB infection and newly diagnosed CHB-related hepatocellular carcinoma (HCC) were prospectively recruited at the Severance Hospital, Yonsei University College of Medicine, Seoul, Korea and defined as the CC group [HBsAg(+) for .6 months, anti-HBc(+), and anti-HBs(-)]. During the study period, a total of 351 subjects who were admitted as living liver or kidney donors were also prospectively recruited and stratified into the UE group (n = 106) or SC group (n = 245) according to their HBV serological markers. Subjects were stratified into the UE group when their HBV serological markers were HBsAg(-), anti-HBc(-), and anti-HBs(-), whereas subjects who were HBsAg(-), anti-HBc(+), and anti-HBs(+) were stratified into the SC group. All subjects were native Korean and provided written informed consent for participation and for use of their genetic material for this study. The study protocol conformed to the ethical guidelines of the 1975 Helsinki Declaration and was approved by the Institutional Review Board of Severance Hospital.
The exclusion criteria were as follows: (1) no available HBV serological markers, (2) vaccinated subjects who were positive for anti-HBs alone, but negative for anti-HBc, (3) HCV or human

Laboratory Tests
All serum specimens were stored at 270uC before testing. Human immunodeficiency virus antibody was determined by enzyme-linked immunoassay (ELISA) with Western blot confirmation. HCV antibody and HCV RNA were assayed using commercially available kits. HBV serological markers for HBsAg, anti-HBc, and anti-HBs were conducted with ELISAs (Dade Behring, Marburg, Germany).

IL28B Genotyping
Genomic DNA was extracted using a Qiagen DNA Blood Mini Kit (Qiagen, Mildren, Germany) according to the manufacturer's instructions. Three primer sets were used to amplify three SNPs (rs12979860, rs12980275, and rs8099917) in the IL28B gene ( Table 1). Reactions were screened by gel electrophoresis with ethidium bromide staining and the polymerase chain reaction (PCR) product was purified by a PCR purification kit (Qiagen, Mildren, Germany).
For detection of the three SNPs (rs12979860, rs12980275, and rs8099917) near the IL28B gene on chromosome 19, which was identified in previous studies, primer extension reactions were performed using SNaPshot ddNTP Primer Extension Kit (Applied Biosystem, Foster City, CA). These SNPs were chosen because they were previously reported in three independent studies including mostly white and Asian patients with chronic HCV infection [2,3,11]. To clean up the primer extension reaction, one unit of shrimp alkaline phosphatase was added to the reaction mixture, and the mixture was incubated at 37uC for 1 hour followed by 15 min at 72uC for enzyme inactivation. The DNA sample, containing extension products and the Genescan 120 Liz size standard solution, was added to Hi-Di formamide (Applied Biosystems, Foster City, CA) according to the manufacturer's recommendations. The mixture was incubated at 95uC for 5 min followed by 5 min on ice. The results were analyzed using the ABI Prism GeneScan and Genotyper program (Applied Biosystems, Foster City, CA).

Full Sequencing of the IL28B Coding Region
Sequencing of the IL28B coding region was performed using the primers listed in Table 1. Nine primer sets were used to amplify the IL28B coding region. PCR products were identified by electrophoresis, and PCR products were then purified with a PCR purification kit and direct sequencing by an ABI 310 automated sequencer (Applied Biosystems, Foster City, CA).

Statistical Analysis
Data are expressed as the median (range), n (%), or n, as appropriate. Observed numbers of each genotype were compared with the expected values in order to test whether the sample was in Hardy-Weinberg equilibrium using the chi-Square test with one degree of freedom. Haplotypic association analysis was performed using PLINK version 1.06 (http://pngu.mgh.harvard.edu/ purcell/plink/). The odds ratio (OR) was calculated to indicate the associated risk and presented with 95% confidence intervals (CI). A P-value ,0.05 on a two-tailed test was considered statistically significant. Statistical analyses were performed with SPSS version 11.0 (SPSS, Inc., Chicago, IL).

Baseline Non-genetic Data
The mean age of the subjects in the UE (53 men and 53 women), SC (135 men and 110 women), and CC groups (153 men and 55 women) were 38 (median, 38; range, 18-59), 42 (median, 43; range, 18-69), and 44 (median, 45; range, 18-68) years, respectively. Subjects in the UE group were significantly younger than those in the SC (P = 0.004) and CC groups (P,0.001). The proportion of male subjects in UE group (50.0%) was significantly lower than that in the SC (55.5%, P,0.001) and CC groups (73.6%, P,0.001). Genotypic analysis of HBV showed that all subjects in this study had genotype C. IL28B SNP had no association to age (P.0.05) and was not prominent in different age group (P.0.05 by one way ANOVA).
Prevalence of rs12979860, rs12980275, and rs8099917 in the UE, SC, and CC Groups Excluding cases with failed genotypic analyses (0,26.0%), genotypes were successfully called for rs12979860 in 100 of 106 (94.3%) subjects in the UE group, 220 of 245 (89.8%) in the SC group, and 154 of 208 (74.0%) in the CC group, respectively; for rs12980275 in 106 of 106 (100%) subjects in the UE group, 243 of 245 (99.2%) in the SC group, and 203 of 208 (97.6%) in the CC group, respectively; and for rs8099917 in 106 of 106 (100%) subjects in the UE group, in 241 of 245 (98.4%) in the SC group, and in 204 of 208 (98.1%) in the CC group, respectively ( Table 2).
Regardless of group, rs12979860 CC was most frequently identified in more than 85% of patients in each group (85.0% in the UE, 85.9% in the SC, and 93.5% in the CC group, respectively), whereas rs12979860 TT were not identified in any group (0% in all groups) ( Table 2). Similarly, rs12980275 AA and rs8099917 TT were most frequently identified ($85%) regardless of subject group, whereas rs12980275 GG was identified in only one subject in the SC group and rs8099917 GG was not identified ( Table 2). The prevalence of CC in rs12979860, AA in rs12980275, and TT in rs8099917 was significantly higher in the CC group than in the UE and SC groups (all P,0.05) ( Table 2). In addition, allelic frequency of three IL28B SNPs is also described in Table 2. All genotype frequencies of these SNPs were in Hardy-Weinberg equilibrium (P = 0.172, 0.382, and 0.227, respectively).

Full Sequencing of the IL28B Coding Region
Fifty samples from each group (150 samples total) were randomly selected for full sequencing of the IL28B coding region. The primer sequences that were used for full sequencing are described in Table 1. Among 19 novel SNPs that were identified in our study (Figure 1), the proportions of 6 SNPs (SNP 2, 5, 7, 8, 17 and 19) were significantly different among the UE, SC, and CC groups (all P,0.05).

Coding Region
In addition to the 150 samples that were used for full sequencing of the IL28B gene, 137 samples from the SC group and 109 samples from the CC group that remained after analyzing the 3 SNPs (rs12979860, rs12980275, and rs8099917) were additionally used for investigation of the 6 novel SNPs in the IL28B coding region.  Table 3). In contrast, only one case each of SNP2 CC, SNP7 AA, and SNP8 CC was identified in the CC group, whereas only one case of SNP17 GG was noted in the SC group ( Table 3). Only the prevalence of SNP8 GG was significantly higher in the CC group (P = 0.026; odds ratio, 0.219, 95% CI, 0.055-0.870), where the prevalence of SNP19 GG trended higher in the CC group (P = 0.051). In addition, allelic frequency of six novel SNPs is also described in Table 3. All genotype frequencies of these SNPs were in Hardy-Weinberg equilibrium (P = 0.865, 0.715, 0.871, 0.799, 0.180, and 0.259, respectively).

Discussion
IL28B (IFN l3) triggers a cascade through the JAK-STAT pathway that upregulates the IFN-stimulated genes (ISGs) [13,14]. The effects of IL28B are similar to those of type I IFNs such as IFN-a and -b; however, IL28B binds to a distinct receptor that may modulate a different set of ISGs [15]. Recently, this cytokine was identified as a key factor of the immune response to HCV.
Because IFN-a and ISGs are thought to be important in the immune response to HBV, and PEG-IFN-a is used to treat chronic HBV infection similar to HCV infection, we can assume that IL28B may also be important in treatment and recovery from HBV infection. Currently, two reports investigated the clinical role of the IL28B polymorphism in HBV infection are currently available [7,12]. One study investigated whether the IL28B polymorphism is associated with responses to PEG-IFN in patients with hepatitis B e antigen positive CHB and proposed that polymorphisms near IL28B are independently associated with serological responses to PEG-IFN in CHB [7]. The other study investigated whether the polymorphism marking the haplotype (rs12979860) also affects other INF-a responsive chronic viral illness, namely HBV, and concluded that the rs12979860 CC genotype was not associated with spontaneous HBV recovery [12]. Therefore, in contrast to HCV infection, the IL28B polymorphism seems to act via different mechanism in terms of treatment outcomes using INF-a and spontaneous clearance in HBV infection.
In our study, to determine the potential influence of the IL28B polymorphism on HBV infection outcome in a natural history setting, we genotyped this polymorphism in HBV cohorts comprising subjects who spontaneously cleared HBV or had persistent infection and those who were never exposed to HBV infection. In contrast to a previous study [12], our study demonstrated that IL28B polymorphisms (rs12979860, rs12980275, and rs8099917) are significantly associated with the outcomes of HBV infection ( Table 2). However, the prevalence of the major alleles rs12979860 CC, rs12980275 AA, and rs8099917 TT (.85%), which are favorable predictors in terms of spontaneous clearance of HCV infection, were unexpectedly higher in subjects who did not clear HBV than in those who experienced spontaneous clearance of HBV or those without exposure to HBV infection. This high prevalence of the major IL28B polymorphism alleles has been reported in several previous Korean studies [16] and is consistently higher than the reported prevalences from Western studies [9]. Based on this finding, higher SVR rates in Korean patients with CHC have recently been reported [16]. However, because IFN-based treatment outcomes have not been investigated in Korean patients with CHB showing a high prevalence of major IL28B polymorphism alleles, further studies are needed to investigate whether the IFN-based treatment outcomes for CHB are similar to those of Western data adjusting for the influence of HBV genotype (nearly 100% of genotype C in Korea). Considering that prior studies have shown that IL28B polymorphism is associated with spontaneous and treatmentinduced clearance of HCV and has one of the strongest known genetic associations with any chronic viral infection to date [1][2][3][4][5], we can assume that IL28B polymorphisms have a distinct effect on the immune response to HCV in spite of the same signal from both type I IFNs and IL28B through the JAK-STAT pathway. Because ISGs were shown to be a major mechanism of non-cytolytic inhibition of HBV replication in a transgenic mouse model [17] and IFN-l2 (IL28A) inhibits HBV replication through upregulated ISGs in HCC cell lines [18], it is difficult to explain the negative correlation between major alleles of IL28B polymorphism and HBV spontaneous clearance given that IL28B stimulates ISGs, which play an important role in the immune response to HBV infection. However, differences between HBV and HCV infection in terms of their replication strategies, pathways to viral persistence and clearance, and host responses, including the production of type I IFNs, are well known [19]. In addition, several unique characteristics of IL28B may explain this unexpected finding. In addition to inducing ISG expression, IL28B may activate alternate antiviral pathways, such as the adaptive immune response, which may be more important in HCV [20]. IL28B may also lead to a different antiviral state than type I IFNs, in terms of the upregulation of different ISGs and different speeds of phosphorylation of STAT1 and STAT2 [15]. It is also possible that the IL28B pathway may respond differently to HCV and HBV (more dominant in HCV), which can lead to a different effect on HCV and HBV infection. Some previous studies support this hypothesis by proposing lower antiviral activities of IFN-l2 (IL28A) and IFN-l1 (IL29) [16,18], although extrapolation of these findings to IL28B should be determined. Finally, it is also plausible that the IL28B and IFN-a pathways are synergistic because of common downstream signaling pathways and that this synergism results in different immune responses to HCV and HBV. However, all of these assumptions, including a lack of influence of the major IL28B polymorphism alleles on HBV clearance [10] or the negative influence observed in our study, should be validated in future work. Interestingly, we found one novel SNP in the IL28B gene (SNP8) that was located immediately ahead of the IL28B exon 1 (Figure 1). Its major allele (GG) prevalence was significantly higher in subjects with persistent HBV infection ( Table 3), although we failed to obtain SNP data in a number of cases using the SNaPshot assay, possibly because of high homology in the nucleotide arrangement on chromosome 19. Thus, we could assume that the genomic site of the SNP8 polymorphism might serve as a binding area for certain transcriptional factors or polymerases and may therefore control gene expression levels by regulating binding capacity. However, functional studies such as reporter assays are needed to validate this assumption. In addition, four haplotypic combinations were significantly different between SC and CC groups. However, further studies are required to reveal the clinical relevance of these haplotypic combinations.
Our study has several strengths. First, this is the first Korean study to focus on the influence of the IL28B polymorphism on the natural clearance of HBV infection and to investigate IL28B polymorphisms in a healthy Korean population. Thus, our results can help establish the worldwide database of IL28B polymorphisms not only for subjects with persistent HBV infection, but also for apparently healthy populations. Second, although the IL28B polymorphism has been known to have a significant correlation with IFN-based treatment of HCV and spontaneous clearance of HCV infection, how this polymorphism influences gene expression levels relatively far from the actual IL28B coding region has been unclear. Thus, in our study, we attempted to sequence the full IL28B coding region using highly qualified PCR techniques with multiple primers to investigate whether there is any novel polymorphism in the actual IL28B coding region that might be related to the clearance of HBV infection. However, because only native Korean subjects were recruited for this study, our results should be interpreted with cautions and external validation in other ethnic groups should be followed. Furthermore, anti-HBs information was not available in our study, hence the correlation between IL28B SNP and anti-HBs could not be analyzed.
In conclusion, the SNP upstream of IL28B that has the strongest genetic association with HCV recovery to date inversely influences on HBV recovery. Thus, the effects of this SNP cannot be generalized to chronic viral infections in which IFN-a treatment is important. Additional studies are needed to understand the mechanisms underlying the effects of this SNP in HBV infection.