Nature
Hepatitis B virus genotype, mutations, human leukocyte antigen polymorphisms and their interactions in hepatocellular carcinoma: a multi-centre case-control study
Received:
19 March 2015
Accepted:
14 October 2015
Published online:
16 November 2015
Abstract
Three genome-wide association studies (GWAS) have been conducted on the genetic susceptibility of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC), two of which consistently identified tagging single nucleotide polymorphisms (SNPs) around HLA-DQ/DR. In contrast, large multi-centre association studies between HBV genotype, mutations and the risk of HCC are relatively rare, and their interactions with host variants are even less. We performed a multi-centre study of 1,507 HBV-related HCC cases and 1,560 HBV persistent carriers as controls to evaluate the effects of HBV genotype, mutations, GWAS-identified HLA-DQ/DR SNPs (rs9272105 and rs9275319) and their interactions on HCC risk. We found HBV genotype C was more frequent in HBV-related HCC. And 11 HBV hotspot mutations were independently and significantly associated with HCC risk. We also detected significant interactions of rs9272105 with both the HBV genotype and mutations. Through stepwise regression analysis, HBV genotype, the 11 mutations, HLA-DQ/DR SNPs, and the interaction of rs9272105 with mutation A1752G were all entered into the HCC prediction model, and the area under the curve for the panel including the HLA-DQ/DR SNPs, HBV genotype and mutations was 0.840. The HBV genotype, the mutations and the HLA-DQ/DR SNPs may serve as biomarkers for the surveillance of HBV persistent carriers.
Introduction
Hepatocellular carcinoma (HCC) is the second leading cause of cancer-related deaths worldwide, with the incidence on the rise both in developed and developing countries1,2. HCC development is influenced by complex factors including viral infection, environmental factors3, and genetic makeup, with most studies having identified susceptibility loci at the human leukocyte antigen (HLA) class II region at 6p214,5,6.
Currently, three genome-wide association studies (GWAS), all from China, have been conducted on hepatitis B virus (HBV)-related HCC, two of which consistently identified HLA-DQ/DR as susceptibility loci5,6. Of the two independent GWAS, one study with 1,538 cases and 1,465 controls for GWA scan identified the single nucleotide polymorphism (SNP) rs9272105 (located between HLA-DQA1 and HLA-DRB1)5, while the other study with 1,161 HCC cases and 1,353 controls for GWA scan identified the SNP rs9275319 at HLA-DQ6. These findings highlight the importance of HLA-DQ/DR molecules in the development of HBV-related HCC.
In contrast, large multi-centre studies on the association between HBV genotype, mutations and HCC risk are relatively rare, especially regarding their interactions with host genetic variants. HBV has been classified into different genotypes according to a sequence divergence of >8% in the entire genome, and the genotypes are further separated into subgenotypes if the divergence in the nucleotide sequence is between 4 and 8%7,8. The HBV genotypes and subgenotypes have distinct geographic distributions and have been implicated to differ with regard to clinical liver diseases, disease outcomes, and responses to interferon treatment9,10,11,12. In East Asia, although HBV genotypes B and C are endemic, the HBV genotype frequency varies in these areas13. Because of the relatively small study sample sizes, the low success rates of HBV typing and the different study designs, the effect of HBV genotype on the outcomes of HBV persistent infection also varied greatly14,15,16,17,18,19. The basal core promoter, which is regulated by the enhancer II to a great extent, controls the transcription of precore mRNA20. The precore protein is processed to produce the secreted hepatitis B e antigen (HBeAg), which indicates active viral replication and is associated with an increased risk of HCC21,22,23. Three small-scale longitudinal studies with less than 50 cases demonstrated that some of the mutations in the enhancer II/basal corepromoter/precore (EnhII/BCP/PC) region would occur years before a diagnosis of HCC is made and gradually accumulate during the development of HCC24,25,26.
HBV mutations are most likely selected via virus-immune interactions in the inflammatory microenvironment. Therefore, we performed a large multi-centre study to evaluate the effects of HBV genotype, mutations in the EnhII/BCP/PC region, GWAS-identified HLA-DQ/DR SNPs (rs9272105 and rs9275319) and their interactions on HCC risk. Furthermore, we evaluated the risk prediction effects of these factors in HBV-related HCC. 作者: StephenW 时间: 2015-11-17 12:32
In this large multi-centre study, we found that HBV genotype C and subgenotype C2 were the risk factors for HCC, and 11 HBV mutations were found to be significantly associated with HCC risk. We also detected significant interactions of HLA-DQ/DR rs9272105 with both HBV genotype and mutations, which may imply a potential biological significance forrs9272105. Excitingly, the panel that combined the HLA-DQ/DR SNPs, HBV genotypes and mutations provided a high sensitivity and specificity to discriminate the HCC patients from the controls. The HBV carriers who infected with HBV genotype C and carrying the rs9272105 AA genotype, rs9275319 AA genotype and risky nucleotide of the 11 HBV mutations had a relatively high HCC risk, which was useful in screening of high-risk groups of HCC and needed to be validated in further prospective studies.
The development of HCC is a multistage process, and most HCCs arise from chronic hepatitis induced by HBV infection, particularly in China27. With the progression of chronic infection, HBV mutations gradually occur28,29. HBV reverse transcriptase lacks proofreading activity, resulting in an estimated mutation rate of 4.57 × 10−5 nucleotide (nt) substitutions per site per year30. Inflammatory factors could also promote HBV mutations, and the insufficient immune responses elicited by the HBV antigens select the disease-related HBV mutations during the long-term evolutionary process31,32. Only the HBV strains/variants best adapted to the host immune system will survive and thrive in liver33. The HLA system is the name of the locus of genes that encodes the major histocompatibility complex (MHC) in humans. This super-locus contains a large number of genes related to immune system function in humans. HLA class II molecules include three isotypes: HLA-DR, HLA-DQ, and HLA-DP, which have been reported on extensively for their association with HBV infection and hepatocarcinogenesis5,6,34,35,36. In this study, we detected significant interactions of HLA-DQA1/DRB1 rs9272105with HBV genotype and mutations on HCC risk. Thus, HLA- DQ/DR genetic polymorphisms might affect the outcomes of chronic HBV infection via regulating the immune selection of HBV mutations, thereby affecting the risk of HCC caused by the HBV mutations.
In this study, the HBV mutations A1846T and G1896A, which have been reported to be associated with an increased risk of HCC by several studies33,37,38, were not significantly associated with HCC. This may be due to different study areas, the sample sizes and the adjustment for other HBV mutations. There were also novel HBV mutations, including A1752G, G1915A/C and C1969T, which were found to be associated with HBV-related HCC. In addition, the effect of C1730G was reversed from a protective effect (adjusted OR = 0.18, 95% CI = 0.15–0.22) to a risk effect with borderline significance (adjusted OR = 2.07, 95% CI = 1.02–4.20) after being conditionally adjusted by the other mutations. Thus, the functional effects of these HBV mutations on HCC risk deserve further investigation.
We successfully validated the associations between HLA SNPs (rs9272105 and rs9275319) and HCC risk. SNP rs9272105 is located between HLA-DQA1 and HLA-DRB1, and rs9275319 is located between HLA-DQB1 and HLA-DQA2. HLA-DQ and -DR proteins make up the HLA class II complex, an α-β heterodimeric membrane glycoprotein that is expressed on the surface of antigen-presenting cells, such as B lymphocytes, macrophages and den-dritic cells. HLA class II glycoproteins present viral peptides to CD4+ T cells and influence the immune responses. Therefore, SNPs in HLA-DQ and -DR genes may have important roles in immune-mediated diseases, including liver diseases and HCC.
Recently, Cao et al. conducted a case-control study (1,108 HCC patients and 1,628 HBV-positive subjects without HCC) to evaluate the effects of HLA-DP polymorphisms (four SNPs reported by a GWAS of HBV persistent infection), HBV EnhII/BCP/PC region mutations and their interactions on HCC risk in subjects infected with HBV genotype B or genotype C. They sequenced the HBV EnhII/BCP/PC region successfully from 1,429 (52.2%) of the HBV-infected subjects and found that the interactions of rs9277535 AA with the T1674C/G or G1719T mutation in genotype C significantly decreased HCC risk33. The same group also evaluated the effect of the STAT3 SNPs and their interactions with HBV mutations on HCC risk with a sample size of 1,021 HCC patients and 990 HBV-positive subjects without HCC. In this study, they genotyped three SNPs of STAT3and sequenced the HBV EnhII/BCP/PC region successfully from 1,160 (57.7%) of the HBV-infected subjects. Finally, they found the interaction of rs1053004 with T1674C/G significantly increased the HCC risk37. Lately, the group performed a case-control study again with a larger sample size of 1,531 HCC patients and 2,489 HBV-positive subjects without HCC to evaluate the impacts of HLA-DQ SNPs (rs2856718 reported by a GWAS of chronic hepatitis B and rs9275319 reported by a HCC GWAS) and their interactions with HBV mutations on the risk of HCC. They sequenced the HBV EnhII/BCP/PC region successfully from 1,450 (36.11%) of all the HBV-infected subjects. And they found rs2856718 variant genotypes significantly decreased HCC risk and the variant genotypes of rs2856718 were significantly associated with an increased frequency of HBV A1726C mutation in genotype C. However, the association betweenrs9275319 and HCC risk was not observed (adjusted OR = 0.99, 95% CI = 0.83–1.18, P = 0.876), which failed to validate the results of the HCC GWAS (P = 2.72 × 10−17), and was different from our results (P = 1.008 × 10−6)6,39.The findings were exciting; however, the different study design (different sample size and matching criteria) and varying methodology (different models of nest multiplex PCR and primers, the detection rate and definition for hotspot mutations) could influence the reproducible of the results13,15,18,19,33,37,39,40.
To the best of our knowledge, this is the first large multi-centre study revealing that HBV genotype and mutations could affect HCC risk via interacting with HLA-DQ/DR genetic variants, and this is the first study constructing prediction models and estimating sensitivity and specificity. There are four advantages of our study. First, the multiple centres used included a large sample size matched by area, age and gender, and the rigid quality control provided sufficient statistical power and more convincing data. Second, because the HBV genotypes were varied among the controls, the controls from an ongoing large-scale, population-based cohort can be helpful for quality control. Third, the success rates of the determination of the HBV genotypes, subgenotypes and mutations were much higher than in previous studies, and the distribution of the HBV genotypes and subgenotypes between cases and controls was validated successfully in an additional independent sample set. Fourth, the diagnostic performance of the panel including the HLA SNPs, HBV genotype and mutations was relatively high (sensitivity = 81.3%, specificity = 74.8%), and the detection of these factors was non-invasive. However, this study had limitations as well. The sample size of the additional sample set was too small to validate the association of the HBV mutations or HLA-DQ/DR SNPs with HCC risk and the interaction of rs9272105 with the HBV genotype and mutations. Further studies with large sample size in diverse populations are warranted to validate and extend our findings. And HBV DNA levels, HBeAg status and ALT level were not detected in our case-control study because some participants had received antiviral treatments, especially for the HCC patients. In addition, since the controls were from a large-scale, population-based cohort, the sensitivity and specificity of the cirrhosis diagnosis may be relatively low, and the antiviral treatment scheme was not obtained. Therefore, cirrhosis status and antiviral therapy were also not shown. Further rigorous prospective studies were also needed to dynamically monitor the HBV persistent carriers. Since HBV EnhII/BCP/PC region is one of important regulatory regions for HBV replication and is less sensitive to antiviral treatments, we only amplified this region for HBV mutation analysis. The detection methods of HBV genotype and mutations were based on nested PCR and sequencing in this study, thus there might be a small number of HBV carriers who failed to detect HBV genotype and mutations for low viral loads. Moreover, we only focused on the HCC GWAS-identified HLA SNPs in terms of the host genetic polymorphism. Therefore, well-designed studies involving genome-wide genetic factors (at least immune-related genes), mutations of the entire HBV genome, and HBV DNA levels may improve the prediction accuracy.作者: StephenW 时间: 2015-11-17 12:33
讨论
在这个大的多中心研究，我们发现，HBV基因型C和亚型C2的危险因素为肝癌，并发现有11 HBV基因突变与肝癌的风险要显著相关。我们还检测到显著的HLA-DQ / DR rs9272105与两个HBV基因型和突变，这可能意味着一个潜在的生物学意义forrs9272105相互作用。令人兴奋的，该组合提供了一种高灵敏度和特异性，从控制判别HCC患者的HLA-DQ / DR的SNP，HBV基因型和突变的面板。谁感染HBV基因型C和携带rs9272105 AA基因型，rs9275319 AA基因型和11个HBV突变风险核苷酸的乙肝病毒携带者有一个比较高的肝癌的风险，这是非常有用的肝癌高危人群的筛查和需要验证在进一步的前瞻性研究。