CCR5 and Outcome of HCV Infection
This report contains excerpts from study article in current Hepatology journal and an Editorial.
Association of genetic variants of the chemokine receptor CCR5 and its ligands, RANTES and MCP-2, with outcome of HCV infection
Simon Hellier and colleagues. Hepatology, December 2003, Volume 38, Number 6
ABSTRACT. The effect of host genetic variation on the outcome of hepatitis C virus (HCV) infection and its treatment is poorly understood. The chemokine receptors CCR5, CCR2, and CCR3 and their ligands, RANTES, MCP-1, MCP-2, and MIP-1a, are involved in the immune responses and the selective recruitment of lymphocytes to the liver in HCV infection. We studied 20 polymorphisms within these genes and investigated their association with persistent carriage of HCV, severity of liver disease, hepatic inflammation, and response to treatment in a large European cohort. Significant associations were found between CCR5-D32 and reduced portal inflammation (P = .011, odds ratio [OR]: 2.3, 95% confidence interval [CI]: 1.09-4.84) and milder fibrosis (P = .015, OR: 1.97, 95% CI: 1.13-3.42). A promoter polymorphism at position –403 in the RANTES gene was associated with less severe portal inflammation (P = .004). An amino acid change in MCP2, Q46K, was associated with severity of fibrosis (P = .018, OR: 2.29, 95% CI: 1.14-4.58). In conclusion, our study suggests a possible role of the polymorphisms CCR5-D32, RANTES –403, and MCP-2 Q46K in the outcome of HCV infection.
Hepatitis C virus (HCV) is a major global health problem with 1% to 2% of the world's population chronically infected. The prevalence of infection in the United Kingdom is approximately 0.5%, but it is likely that this figure will continue to rise as more cases come to light. Persistent HCV infection has a very variable outcome, with 25% of those infected developing cirrhosis within 20 years, of whom 3% to 5% per year will proceed to hepatocellular carcinoma. A further 5% per year will develop liver failure, and 2% per year will die of a liver-related death. There is also a very variable response to treatment because only 40% of those treated with interferon alfa and ribavirin successfully clear the virus. There are a number of epidemiologic and viral factors that influence susceptibility to persistent HCV infection, progression of HCV-related liver disease, and response to treatment, but host genetic factors are also influential. This study has examined 20 polymorphisms in the genes for the CC chemokine receptors (CCR) CCR5, CCR2, and CCR3 and their ligands, RANTES (regulated and normal T-cell expressed and secreted), MCP-1 (monocyte chemotactic protein), MCP-2, and MIP-1 (macrophage inflammatory protein), to try and elucidate their role in the immunopathologic outcome of HCV infection.
Chemokines constitute a large family of small (8-10 kD) cytokines that are up-regulated in inflammation and whose effects are mediated by members of a family of 7 transmembrane domain, G protein-coupled receptors. Their major role is in leukocyte migration and dependent processes such as immune surveillance and innate and adaptive immune responses. Chemokines have been linked with several disease states including psoriasis, atherosclerosis, arthritis, and multiple sclerosis.
It is now known that chemokine receptors also have a role in infectious disease either because of over expression of receptors or by facilitating entry of pathogens into permissive cells. In 1996, the CCR5 was shown to act as a cofactor for entry of macrophage-tropic strains of HIV-1. Shortly after this, a series of reports described the defective CCR5-D32 allele, which established the role of CCR5 in HIV pathogenesis. CCR5-D32 is a 32-bp deletion in the CCR5 gene resulting in a nonfunctional protein, explaining the almost complete protection against HIV-1 infection in individuals homozygous for the defective allele and delayed progression to AIDS in heterozygotes.
CCR5 is a strong candidate gene for the outcome of HCV infection and the course of HCV-related liver disease. The immune response in persistent hepatitis C is compartmentalized, with a predominant Th2 or Th0 response in the periphery and a Th1 response in the liver associated with progressive liver injury. Differences in chemokine receptor expression between Th1 and Th2 cells may influence their selective recruitment to tissues. In vitro, Th1 cells that express CCR5 migrate to their respective chemokines, RANTES and MIP-1, which are largely confined to the portal regions. Portal inflammation is the predominant pattern of inflammation seen in persistent viral infection, including HCV, and is associated with a less aggressive course of disease. Antigen-presenting dendritic cells are located in the portal area and infiltrating T cells are exposed to infected hepatocytes in the periportal area. Thus, these areas could be sites of the initial immune response to hepatitis C viruses. Apart from T-cell migration, CCR5 and other chemokine receptors mediate cell activation, costimulation, and differentiation of T cells and monocytes during innate and adaptive immunity. These are all processes relevant to HCV clearance or persistence, which are likely to be modified by the loss of a functional CCR5 receptor. In addition to CCR5-D32, 7 CCR5 promoter single nucleotide polymorphisms (SNP) were studied here.
CCR2 also serves as an HIV-1 coreceptor, and a role in modulating the immune response as well as recruiting monocytes/macrophages to sites of inflammation has been suggested. Only 1 variant has been reported in CCR2, which leads to a ValineIsoleucine substitution at amino acid position 64. This is within the first transmembrane region and has been shown to be associated with a delay in progression to AIDS; this protection is genetically independent of that conferred by CCR5-D32.
Five polymorphisms have been identified in the CCR3 gene: 2 silent mutations (T51C, C240T) and 3 that encode amino acid changes, an arginine to glutamine at amino acid position 275 (G824A), a leucine to proline at amino acid position 351 (T1052C), and a cysteine to serine substitution at amino acid position 218 (T652A).
Two polymorphisms have been described in the RANTES promoter region. It has been suggested that the RANTES –28G mutation increased RANTES expression in HIV-1-infected individuals and thus caused a delay in the progression of HIV-1 disease. The second polymorphism is a GÆA substitution at position –403. There is evidence that the mutant A allele leads to increased transcription of RANTES and that the A allele is associated with an increased susceptibility to atopy, asthma, and HIV.
MIP-1 has both phagocyte stimulating and proinflammatory properties and is a key ligand for CCR5. A biallelic dinucleotide microsatellite repeat has been identified within the MIP-1 promoter region. MCP-1 and 2 are active on multiple leukocyte populations showing chemotactic activity at low concentration in vitro. Hepatic stellate cells (HSC) have been shown to regulate leukocyte trafficking by secreting MCP-1 and MIP-1a, and it has been suggested that MCP-1 may have a direct profibrogenic action via HSC chemotaxis. Two polymorphisms have been identified in the distal regulatory region of the MCP-1 gene. The polymorphism at –2076 does not appear to affect MCP-1 transcription. However, in vitro cells from individuals who are heterozygous or homozygous for G at –2518 appear to produce more MCP-1 than cells from individuals homozygous for A at –2518. Little is known about the role of MCP-2 in pathology, although it may act as an effector molecule in the inflammatory events occurring in multiple sclerosis. One SNP has been identified in the MCP-2 gene to date. No significant difference in biologic activity has been observed between the 2 isoforms.
RESULTS. A significant association was found between severe fibrosis and carriage of CCR5 D32 (P = .015; OR: 1.97, 95% CI: 1.13-3.47). When looking at the subgroup comparison for portal inflammation, although the number of patients was markedly reduced to 124, there was a significant association between –D32/–D32 homozygotes and mild portal inflammation (P = .011, OR: 2.3, 95% CI: 1.09-4.84). The promoter polymorphism at position –2132 was found to be significantly associated with susceptibility to persistent HCV infection, with presence of the C allele increasing risk of persistent carriage (P = .048). Carriage of the C allele at –2132 was also associated with an initial response to interferon, i.e., sustained responders and relapsers versus nonresponders (P = .023, OR: 3.15, 95% CI: 1.14-8.72).
The genotype frequency of the RANTES promoter polymorphism at position –403 was significantly associated with portal inflammation (P = .015). AA homozygotes are associated with milder portal inflammation than those carrying the G allele (P = .004).
For MCP-2 Q46K, a significant association was found between severity of liver fibrosis and genotype (P = .027). Carriage of the K variant is associated with more severe fibrosis (P = .018, OR: 2.29, 95% CI: 1.14-4.58).
When logistic regression was applied for the variables sex, ethnicity, and age at infection, the associations became nonsignificant because of the reduction in sample numbers and reduced statistical power. However, if each variable was assessed individually, there was no significant change in the OR, suggesting that these variables are not significant confounders.
There was no association between the polymorphisms analyzed in CCR2, CCR3, MIP-1a, or MCP-1 and any of the phenotypes. Three of the SNP described in CCR3 (C240T, T652A, and T1052C) were not polymorphic in this population.
Discussion by Authors
This is a large case control study, looking at 20 polymorphisms in a group of 7 CC chemokines and their receptors. Interest in these genes was initially raised by the presence of the functional polymorphism CCR5-D32, which gives rise to a truncated protein that, in the context of HIV-1, does not act as a competent receptor. With the finding that CCR5 is partly responsible for the recruitment of T cells to the portal region, it was expected that an incompetent receptor would result in the presence of less inflammation in this region. The finding that there is an association between CCR5-D32 and lower levels of portal inflammation is in agreement with this hypothesis. The same finding was not found when CCR5-D32 was assessed in the context of interface hepatitis, or with overall necroinflammatory score. How this observation can be put in context with the other finding that carriage of CCR5-D32 is associated with more severe fibrosis is a matter of conjecture. It is likely that the recruitment of T cells to the liver is not dependent on the locally expressed chemokine receptors but that they are responsible for differential recruitment once the cells enter the liver. It is possible, therefore, that, if fewer cells are recruited to the portal regions, which are associated with a more benign fibrosis outcome, then more cells will be available for recruitment to areas of interface or lobular hepatitis, which are associated with a more severe outcome in liver fibrosis. However, in this study, an association between CCR5-D32 and increased interface hepatitis was not found, although data were only available on a relatively small number of patients. Other groups have suggested that the presence of CCR5-D32 may be associated with a less favorable outcome in HCV infection. A study in Germany compared the frequency of this mutation among patients with HCV infection, HIV infection, and those with HIV/HCV coinfection. They found an increased frequency of D32/D32 genotypes among patients with HCV infection and also increased viral loads, suggesting unfavorable effects of this mutation on the course of HCV infection. Data on viral load were not available to make this comparison in our study. However, the frequency of D32/D32 genotypes in those chronically infected with HCV was 1.6% close to the expected frequency in Caucasian subjects and much less than the 7.8% reported by Woitas et al. This has also been shown in another recent study, and the high prevalence of CCR5-D32 homozygosity in the earlier study may reflect resistance to HIV in hemophiliacs rather than a susceptibility to HCV infection.
Of the 7 promoter polymorphisms studied in CCR5, only –2132 was found to have significant associations with HCV phenotypes, although it does not appear to have an effect on CCR5 expression. In this study, the C allele was weakly associated with the persistent carriage of HCV, although the numbers in the group of individuals who cleared the virus spontaneously are very small, so this could be a chance finding. There also appeared to be an association between this allele and interferon response. In this instance, individuals who carried the C allele were more likely to have an initial response to interferon (i.e., SR + RL) than individuals who were homozygous for the T allele.
The RANTES promoter polymorphism at position –28 showed no significant association with any of the phenotypes studied. However, this is a rare polymorphism in Caucasian subjects with a mutant allele frequency of only 4%, and a study of this size may miss an association. This is also the case for the CCR3 polymorphisms, CCR2 V64I, and the closely linked CCR5 promoter polymorphisms at position –1835.
The polymorphism at position –403 appears to be functional, increasing RANTES expression, and has been associated with increased susceptibility to asthma, atopy, and HIV. RANTES may have a role in the selective recruitment of T cells to portal and, in particular, periportal regions. It might be expected, therefore, that the mutant allele would be associated with increased portal inflammation, interface hepatitis, and severity of fibrosis. A significant association was found with portal inflammation, suggesting that homozygotes for the less frequent allele (A) at position –403 had less severe portal inflammation than homozygotes for the wild-type allele (G). This is counterintuitive because the A allele has been associated with increased transcription of RANTES in some studies and might, therefore, have been expected to lead to more severe portal inflammation. RANTES is not, however, a dedicated CCR5 ligand because it also binds CCR1 and CCR3; it is also expressed at sites of interface hepatitis. It is possible, therefore, that, if RANTES expression is increased as a result of this polymorphism, T cells are not recruited specifically to the portal regions. No associations were found with the degree of interface hepatitis in this study, although only small numbers could be included in this subset analysis.
Although there is no published evidence of a role for MCP2 in the immunopathology of HCV, it is a reasonable candidate gene because it could potentially be involved in susceptibility to persistent carriage, inflammatory response, or fibrosis. The results from this study showed no significant association with susceptibility to persistent HCV carriage. They did, however, show a significant association between carriage of the K allele and severe fibrosis (P = .018, OR: 2.29, 95% CI: 1.14-4.58) but, interestingly, not with inflammation. In HCV, fibrosis is largely driven by the inflammatory response, and it would be expected that a chemokine involved in T-cell migration would lead to fibrosis through increased inflammation. Functional studies revealed no difference in chemotactic and calcium-mobilizing abilities of the 2 isoforms of MCP-2. Critically, these functional assays did not assess the abilities of the 2 isoforms to block HIV-1 binding to CCR5, but it is unlikely that the polymorphism is directly responsible for associations seen in either HIV or HCV and may be acting as a marker for a functional polymorphism in MCP2 or a nearby gene. Given that MCP2 is on chromosome 17 within the b-chemokine gene cluster, this is a strong possibility, and it is, therefore, likely that the associations with HIV and HCV do not necessarily imply that the same gene is involved. Therefore, to elucidate the role of MCP2 in HCV, particularly that related to fibrosis, further functional studies will need to be performed.
A recent publication has looked for associations of these HCV phenotypes with polymorphisms in CCR5, CCR2, and RANTES. An association was identified for RANTES –403 homozygotes and inflammation but with the overall necroinflammatory score rather than the portal inflammation subgroup. An association between CCR5-D32 and fibrosis was not observed and possible association with portal inflammation not assessed. A marginal association between CCR5 promoter polymorphism at position –2459 (position 59029 from U95626) was noted, which we cannot confirm. As in this study, no associations were found with CCR2 –64I or RANTES –28.
In this study, 20 polymorphisms in 7 genes have been studied, raising the issue of multiple comparisons. A Bonferroni correction is too conservative in the domain of human complex trait genetic association, in which many genes are expected to be associated with the phenotype to some extent. Taking the limitations of our data as outlined above into account, we therefore believe that we have carried out the most appropriate evaluation of our results but acknowledge that our subgroup analysis is based on smaller sample numbers. Thus, the associations in this study should be reassessed in other studies and populations, but all of the positive results are with mutations that are known to be (CCR5-D32, RANTES –403) or may well be (MCP-2) functional.
In conclusion, this study has attempted to elucidate the role of CC chemokines and their receptors in the immunopathologic outcome of HCV. This is a complex area because of the large number of chemokines in this group and the likelihood that there is an overlap in their roles and, therefore, a degree of redundancy. This may influence the observed effect of polymorphisms on individual chemokines or receptors. Importantly, the results add further weight to the role of CCR5 and its ligands in the recruitment of T cells to the portal regions of the liver and suggest that CCR5-D32, RANTES, and MCP-2 may have an effect on the outcome of HCV-related liver disease.
Chemokine systems and hepatitis C virus infection: Is truth in the genes of the beholders?
December 2003, Volume 38, Number 6
Kittichai Promrat, M.D.
Gastroenterology Division, Brown University, Providence, RI
T. Jake Liang, M.D.
Liver Diseases Section, NIDDK, National Institutes of Health, Bethesda, MD
Hepatitis C virus (HCV) infection leads to a wide spectrum of liver injury. While viral factors probably play a role in the outcome of infection, host factors including genetic susceptibility may contribute to the variable manifestations of the disease and treatment response. The concept of disease susceptibility genes, with effects ranging from simple Mendelian inheritance to complex polygenic predisposition, has been widely demonstrated for decades and proven in a variety of human diseases.1 The number of association studies of candidate genes in infectious disease has increased rapidly as more polymorphisms are being identified in genes considered to have important roles in disease acquisition and outcome. Continuing identification of immunoregulators including cytokines, chemokines, growth factors, and their receptors has provided a rich source of new candidate genes to study disease susceptibility or resistance. Various candidate genes, most of them related to host immune response in microbial infection, have defined genetic polymorphisms that are associated with variable manifestations of infections including malaria, tuberculosis, leprosy, acquired immunodeficiency syndrome (AIDS), and viral hepatitis.
The initial immunogenetic studies in chronic viral hepatitis had focused on association with certain major histocompatibility complex antigen (human leukocyte antigen [HLA]) polymorphisms. Specific major histocompatibility complex class II alleles (HLA-DQB1*0301 and HLA-DRB1*1101) have been associated with spontaneous clearance of HCV infection. Several other class II alleles have been linked with disease progression, but results are inconsistent, possibly due to differences in designs and subjects. HLA class I has been less well studied. HLA-Cw*04 was found to be associated with persistent HCV infection in a large case-control study. Several other candidate genes that are related to antigen presentation, type 1 and type 2 cell-mediated immunity, and inflammation have been investigated.
Chemokine and chemokine receptor genes are strong candidate genes for outcome of HCV. Chemokines are 8- to 10-kd proteins with 20% to 70% homology in amino acid sequences. There are approximately 40 chemokines identified to date, which are classified according to the configuration of cysteine residues near the N-terminus into 4 families: CC-, CXC-, C-, and CX3C-. They play important roles in leukocyte trafficking to sites of infection and in regulating T helper cell polarization. They also have crucial roles in linking innate and adaptive immunity. Chemokine receptors are G-protein coupled, 7-transmembrane receptors, which are categorized based on the chemokine class they bind. Chemokine-chemokine receptor interactions are likely to be important in chronic hepatitis C, where T cells are recruited to the liver parenchyma to mediate clearance of HCV-infected hepatocytes. Several allelic variants of chemokine receptors have been shown to be important in the pathogenesis of viral infection. The most widely known, a 32-base-pair deletion in the open reading frame of CCR5 (CCR5D32) is associated with protection against human immunodeficiency virus 1 (HIV-1) infection and delayed progression to AIDS in white populations. The importance of the chemokine system in the pathogenesis of HCV has been the main focus of several recent epidemiologic studies. Outcomes of HCV infection that have been examined to date in most immunogenetic studies can be categorized into 4 areas: susceptibility to infection, clearance or persistence, histology (inflammation and fibrosis), and treatment response.
SUSCEPTIBILITY OF INFECTION
The frequency of CCR5D32 homozygosity was recently reported to be dramatically increased in patients with hepatitis C. In this study, 7.8% of patients with HCV exhibited the CCR5D32/D32 genotype, which was 3-fold higher than expected in the general population at Hardy-Weinberg equilibrium. Based on this finding, it was suggested that individuals with CCR5D32/D32 genotype had increased susceptibility to HCV infection. This finding, however, appeared to be specific to a subgroup of patients with hemophilia who were infected with hepatitis C and remained uninfected with HIV, and was not found in other groups of patients with hepatitis C. Several subsequent reports, including the report by Hellier et al. in this issue of HEPATOLOGY have confirmed similar frequency of CCR5D32 alleles among patients with hepatitis C and healthy reference controls. One possible explanation for this discrepancy is that the high frequency of CCR5D32 homozygosity observed does not represent an increase in susceptibility to HCV infection but rather a resistance to HIV infection in individuals who were heavily exposed to both viruses through multiple contaminated blood product transfusion. These studies underscore the difficulty in experimental design to examine the influence of host genetics on susceptibility to HCV infection. Several studies were performed by comparing anti-HCV-positive individuals with either the general population or anti-HCV-negative individuals. It is, however, difficult to ascertain whether the control groups have the same level of exposure to HCV. Additionally, certain individuals may lose anti-HCV and may be left with no markers of HCV infection after spontaneous clearance of HCV infection. Without an appropriate control group, this may lead to a misinterpretation of results or a failure to detect a true susceptibility gene.
Clearance or Persistence
To date, the associations of chemokine polymorphisms on clearance or persistence of HCV infection have been largely negative. However, most studies included a relatively small number of patients with spontaneous clearance of HCV.
Histology (inflammation and fibrosis)
In this issue of HEPATOLOGY, Hellier et al. further expand our knowledge of the influence of chemokine polymorphisms on the outcome of HCV infection. In this large case-control study, participants with HCV infection were recruited from several hepatology clinics across Europe. Twenty polymorphisms of chemokines and chemokine receptors were genotyped. A promoter polymorphism at position –403 in the RANTES gene (P = .004) and CCR532 allele (P = .011) were found to be associated with reduced hepatic inflammation in the portal areas. RANTES serves as a key ligand for CCR5 and plays an important role in attracting T cells to the portal area of hepatitis C-infected liver.
After HCV infection, the initial host defense mechanism involving innate immune response is activated. Important mediators participating in this phase include cytokines, macrophages, dendritic cells, NK cells, and NK T cells. The chemokine system plays an important role in linking innate and adaptive immune responses against HCV infection. Interactions between CCR5 and its ligands RANTES, MIP-1a, and MIP-1b are crucial in leukocyte recruitment, transendothelial migration, and retention at sites of infection. RANTES and CCR5 are also involved in (1) HSC migration and proliferation and (2) cross-talk between HSCs and leukocytes during fibrogenesis.
Studies have shown that two chemokine receptors, CCR5 and CXCR3, are found on a high proportion of T lymphocytes isolated from livers of patients with chronic hepatitis C. Moreover, these receptors appear to be functional in that their ligands, including CCR5 ligands RANTES, MIP-1a, and MIP-1b, are locally up-regulated in the portal areas where most necroinflammatory activity is located. Importantly, the polymorphism at position –403 of the RANTES gene promoter appears to have functional difference with the –403A allele leading to increased transcription of RANTES. The association of RANTES–403A and reduced portal inflammation in this study is in agreement with a recent study that showed milder overall necroinflammatory scores among HCV patients with RANTES–403A. Additionally, Hellier et al. also reported that CCR5D32 allele is associated with reduced portal inflammation. Taken together, these findings identify chemokine-chemokine receptor interaction as an important mediator of hepatic inflammation in hepatitis C. Another observation reported by Hellier et al. is the association between CCR5D32 allele with more advanced fibrosis. This was, however, not observed in another study, which used a fairly similar approach. In vitro experiments have shown that CCR5 and its ligand RANTES play significant roles in hepatic stellate cell (HSC) migration and proliferation and communication between HSCs and leukocytes during fibrogenesis. It is therefore difficult to explain this association from the biological point of view. Moreover, this association does not fit well with their own observation that CCR5D32 allele is associated with reduced portal inflammation.
Interleukin 10, CTLA-4, and certain HLA class II haplotypes have been associated with response to treatment with interferon. Dorak et al. performed a comprehensive genotyping of CCR5 and CCR2 genes and showed that a CCR5 haplotype (genotype E/E) was associated with sustained virologic response to interferon therapy. In another study, 59029A allele, which is an allele of the CCR5 haplotypes E, was also found to be associated with sustained response to interferon therapy in multivariate analyses. Hellier et al. were not able to replicate this association but instead found a positive association with another CCR5 promoter variant at –2132. However, this finding needs to be confirmed in a future study since it was based on only a small subgroup of patients (n = 98).
Many studies, including the one by Hellier et al., have weaknesses that are worth mentioning. First, the sample size is often inadequate to detect differences (power <80%) in the effects of several genetic variants, in particular those with allele frequency less than 15%. Moreover, incomplete clinical and genotyping data (in some cases less than 50% of samples) further reduces the statistical power to detect significant associations. Another analytical challenge that is inherent to most immunogenetic studies is multiple testing. Many genetic markers in a candidate gene or many candidate genes in a pathway are usually being tested at the same time. Bonferroni correction is typically used to reduce type I error in multiple testing, but at the same time will reduce the statistical power. Since many of the tested markers or genes are not independent, this type of correction may be too conservative. One solution is to select genetic markers using a hierarchical scheme based on the strength of the association. Another approach is to divide a cohort for a 2-stage analysis or confirm positive results in an independent sample.
Over a relatively short time span, we have seen a surge of interest in the area of immunogenetic study of chemokine system and HCV. It is encouraging to see that some of the associations have been confirmed (in part) in independent cohorts, thus minimizing chance findings. Thus far, most immunogenetic studies of chemokine systems have been performed in cohorts composed of predominantly white populations. Further research in different ethnic populations and a combination of genetic markers is needed to convincingly identify genetic variants affecting the outcome of a disease that is polygenic in nature. In addition, these polymorphic markers need to be defined in terms of their functional significance in the pathogenesis of the disease. By identifying relevant host factors genetically and investigating their molecular interactions with HCV, we may gain additional insights into HCV pathogenesis and uncover new potential targets for vaccine development and treatment intervention.