A CCL5 Haplotype Is Associated with Low Seropositivity Rate of HCV Infection in People Who Inject Drugs

Objective The role of CC chemokine receptor 5 (CCR5) and its ligand CCL5 on the pathogenesis of HIV infection has been well studied but not for HCV infection. Here, we investigated whether CCL5 haplotypes influence HIV and HCV seropositivity among 373 Caucasian people who inject drugs (PWID) from Estonia. Methods Study included 373 PWID; 56% were HIV seropositive, 44% HCV seropositive and 47% co-infected. Four CCL5 haplotypes (A-D) were derived from three CCL5 polymorphisms (rs2107538/rs2280788/rs2280789) typed by Taqman allelic discrimination assays. The data of CCR5 haplotypes were used from our previous study. The association between CCL5 haplotypes with HIV and/or HCV seropositivity was determined using logistic regression analysis. Results Possessing CCL5 haplotype D (defined by rs2107538A/rs2280788G/rs2280789C) decreased the odds of HCV seropositivity compared to those not possessing it (OR = 0.19; 95% CI 0.09–0.40), which remained significant after adjustment to co-variates (OR = 0.08; 95% CI 0.02–0.29). An association of this haplotype with HIV seropositivity was not found. In step-wise logistic regression with backward elimination CCL5 haplotype D and CCR5 HHG*1 had reduced odds for HCV seropositivity (OR = 0.28 95% CI 0.09–0.92; OR = 0.23 95% CI 0.08–0.68, respectively) compared to those who did not possess these haplotypes, respectively. Conclusions Our results suggest that among PWID CCL5 haplotype D and CCR5 HHG*1 independently protects against HCV. Our findings highlight the importance of CCL5 genetic variability and CCL5-CCR5 axis on the susceptibility to HCV.


Introduction
A major role of chemokines is to promote leukocyte migration in order to participate in immune surveillance and innate/adaptive immune responses. Cysteine-cysteine chemokine ligand 5 (CCL5; formerly known as RANTES-regulated upon activation, normal T cell expressed and secreted) is a CC chemokine that binds to a variety of chemokine receptors including the major HIV co-receptor CCR5 [1]. On one hand, it poses antiviral activity by suppressing the entry of R5-tropic HIV strains by down-regulating CCR5 receptor expression through its occupancy of T cells and macrophages [2,3]. On the other hand, CCL5 exhibits viral replication inhibition in vitro [4].
CCL5 also plays a critical role in the recruitment of T cells into the liver and, through that, influences the progression of chronic hepatitis C infection [5]. Chemokines such as CCL3 and CCL4 can contribute to the provision of CD4 T cells to help optimal CD8 T cell priming [6]. The participation of CCL3, CCL4 and CCL5 in the early T cell response to infections including HIV and hepatitis C virus (HCV) is crucial [7][8][9].
The polymorphisms in the promoter and first intron region of CCL5 affect the expression of CCL5 and influence susceptibility as well as disease progression of HIV infection [10,11]. The single nucleotide polymorphisms (SNPs) of CCL5 and different haplotypes (depending on studied SNPs) have been associated with susceptibility to HIV and with the disease progression in HCV infection with various results (S1 Table). Most of these associations between CCL5 variability and acquisition have been found in the context of sexual or vertical transmission. We have previously evaluated the associations between the receptor of CCL5 (CCR5) and another ligand for CCR5 (CCL3L1) genetic variability and the susceptibility to HIV and HCV among PWID in terms of co-infection. We showed that a higher CCL3L1 copy number than population median decreases the odds of HIV positivity while CCR5 human haplotype (HH) G Ã 1 decreases the odds of HCV seropositivity [12,13].
In Estonia a rapid increase in the prevalence of hepatitis B (HBV) and HCV infection among people who inject drugs (PWID) was seen by the end of 1990s and was followed by the HIV epidemic in August 2000 (www.terviseamet.ee, [12]). As of now, approximately half of PWID are infected with HIV and more than 70% with HCV and HBV, providing an opportunity to look into the association between host factors and the abovementioned infections in the context of co-infections. Considering the potential role of CCL5 on HIV and HCV infection, here, we aimed to investigate associations between CCL5 haplotypes and susceptibility to HIV and/or HCV infection in Caucasian PWID.

Subjects and laboratory analyses
A total of 373 Caucasian PWID (300 male; 55 female and 18 gender unknown; median age of 26 years) were recruited in 2006 and 2007 from two syringe-exchange programs (n = 270) using a respondent-driven sampling [14,15], and from three Estonian prisons (n = 103). The duration of intravenous drug use (IVDU) was available for 67% subjects. All study subjects were Caucasians from Estonia and seven subjects (3.4%) reported that they had received or were receiving antiretroviral therapy. The data of HCV treatment was not available. Sample collection and the detection of HIV, HCV and HBV was performed as described previously [12].
The demographic characteristics (HCV serostatus, median age and gender) of subjects were similar both from the syringe-exchange programs and the prisons (p>0.1). It was not feasible to recruit HIV negative subjects from prisons. Prisoners were determined as PWID by their indication of previous drug use because, according to official sources, there is no drug use inside the prisons. The data on the duration of IVDU was not available for persons recruited from prisons but was recorded for the majority of PWID (92%) from the syringe-exchange programs [12].

Statistical analyses
The program R 2.8.1 [16] was used and a p value of <0.05 was considered statistically significant. Differences in the distribution of CCL5 haplotypes between study groups were compared by the Fisher exact tests as appropriate. Univariate and multivariate logistic regression models (including step-wise logistic regression analysis with backward elimination) were constructed to determine the associations of CCL5 haplotypes with HIV and HCV serostatus. Age, gender, HBV serostatus and the duration of IVDU were included as covariates. The study was approved by the Ethics Committees of Tallinn, University of Tartu and University of Texas Health Science Center at San Antonio; all participants signed the informed consent following the Helsinki declaration.

Study population and the distribution of CCL5 haplotypes
Altogether 56% (n = 208) and 76% (n = 284) of 373 subjects were infected with HIV or HCV, respectively, and approximately half (n = 177) had HIV/HCV co-infection. More detailed description of the study population is presented elsewhere [12].

CCL5 SNPs and HIV or HCV serostatus
The distribution of CCL5 SNPs was similar between HIV seropositive and seronegative subjects. Only the prevalence of rs2280788 GC genotype was higher in HCV negative compared to HCV seropositive subjects ( Table 1).

CCL5 haplotypes and HIV or HCV serostatus
The distribution of CCL5 haplotypes in HIV seronegative and HIV seropositive subjects was similar (data not shown). However, the frequency of CCL5 haplotype C was lower in HCV seronegatives compared to HCV seropositives (6.9% vs. 16.8%; p<0.05) and the frequency of CCL5 haplotype D was higher in HCV seronegatives than in the HCV seropositives (20.7% vs. 4.5%; p<0.05) (Fig 2). There were no differences between HCV seronegative and HCV seropositive subjects in terms of CCL5 haplotype A and B.

CCL5 haplotype C and D, and HCV serostatus
Due to the different distribution previously mentioned we developed univariate and multivariate logistic regression models to evaluate how the CCL5 haplotypes C and D influence the susceptibility to HCV infection. We observed that subjects possessing CCL5 haplotype C had almost three times increased odds of being HCV seropositive compared to those not possessing CCL5 haplotype C. At the same time, subjects with CCL5 haplotype D had five times decreased odds of HCV seropositivity compared to subjects without CCL5 haplotype D (

CCL5 haplotype D and HIV/HCV co-infection
The frequency of CCL5 haplotype D in HIV or HCV monoinfected PWID compared to HIV/ HCV co-infected PWID was similar. However, CCL5 haplotype D in HCV seronegative subjects was represented in significantly higher frequency than HCV seropositive subjects among HIV seronegative PWID but not among HIV seropositive PWID (Fig 3). In addition, HIV/ HCV co-infected subjects possessed CCL5 haplotype D at lower frequency than double negative subjects.F

and HCV serostatus
The CCL5 is the potent ligand of the HIV major coreceptor CCR5 [1] and in our previous study of the same population we showed that the presence of CCR5 HHG Ã 1 was associated with decreased odds of HCV seropositivity [12]. For that reason we tested whether both, CCL5 haplotype D and CCR5 HHG Ã 1, possess a protective effect against HCV infection. After using step-wise logistic regression analysis with backward elimination (including all co-variates), CCL5 haplotype D, CCR5 HHG Ã 1 and the duration of IVDU remained significantly associated with HCV serostatus. We saw that the first two decreased the odds of HCV seropositivity and longer duration of IVDU increased the odds of being HCV seropositive ( Table 3).

Discussion
To the best of our knowledge, this is the first study conducted in an IDU population to evaluate the effect of CCL5 haplotypes on the susceptibility of HCV and HIV infection. We demonstrate  Table 3. Associations between CCL5 haplotype D and CCR5 HHG*1 on HCV serostatus in people who inject drugs by step-wise logistic regression analysis with backward elimination. the involvement of CCL5 haplotype D (rs2107538A/rs2280788G/rs2280789C) in the susceptibility of HCV such that persons possessing CCL5 haplotype D have reduced odds of being HCV seropositive as compared to those not possessing this haplotype. Furthermore, analyzing CCL5 haplotype D and CCR5 HHG Ã 1 in the same model we saw that both of them had independent protective effect against HCV. This finding makes biological sense given the potent receptor ligand nature of CCR5 and CCL5. However, in contrast to some previous studies [17,18] we failed to find any associations between CCL5 haplotypes and the susceptibility to HIV infection in Caucasian PWID.
We are not aware of any studies evaluating the association between CCL5 SNPs or haplotypes and HCV susceptibility despite the fact that CCL5 participates in viral clearance during both acute and chronic HCV infection [19]. For example, patients with CCL5 rs2280789C had an improved response to interferon therapy and those with CCL5 rs2107538A showed milder portal inflammation than subjects without these polymorphisms [20,21]. In fact, CCL5 haplotype D is the combination of these two SNPs plus rs2280788G (Fig 1). While rs2107538A and rs2280788G have been associated with up-regulation of CCL5 transcriptional activity, rs2280789C down-regulates this [10,11]. However, the effect of CCL5 haplotypes on gene expression is unknown.
We have previously demonstrated the protective effect of CCR5 HHG Ã 1 in the acquisition of HCV infection [12]. In the present study we show that the possession of CCR5 HHG Ã 1 and CCL5 haplotype D independently protects from HCV infection. Our results suggest that despite CCR5 not being the receptor for HCV entry, the CCR5-CCL5 system is still essential in the susceptibility to HCV infection. Thus far the role of the Th1-CCL5-CCR5 system in the induction of immunity against HCV and the effect on the outcome of chronic HCV has been shown. The studies have hypothesized that CCR5 interacts with its ligands to elevate the recruitment of Th1 cells into the liver and, with this, mediates the clearance of HCV infected hepatocytes [22,23]. The data also indicate that the lack of Th1-type cytokines favors the establishment of chronic HCV infection [24]. In addition, HCV itself down-regulates CCR5 expression via direct interaction of the HCV E2 envelope protein with tetraspanin CD81 and through HCV core protein and NS5A increase the level of CCL5 [25,26] [27]. This leads to decreased CCR5 surface density [26] and possibly through that affects the Th1-CCL5-CCR5 system. Still, it remains unclear whether CCL5 and CCR5 directly participates in elimination of the virus or not, and what is the role of genetic variance of CCL5-CCR5 in this interaction.
We acknowledge that another host genetic factor, the polymorphism rs12979860 upstream of interleukin-28B (IL28B; also known as interferon-λ3) gene has been associated with HCV spontaneous clearance and the outcome of interferon based treatment [28][29][30]. One can speculate that IL28B genetic variability might also influence the susceptibility to HCV infection and therefore affect the results of the association between CCL5 and CCR5 haplotypes and HCV acquisition. However, we have shown that IL28B rs12979860 genotypes do not have an impact on HCV acquisition in Caucasian PWID population from Estonia [31]. Thus, we consider IL28B rs12979860 unlikely to interact with the results of current study.
Based on the previously described associations between CCL5 SNPs and susceptibility to HIV infection, we expected to observe differences in the distribution of CCL5 haplotypes between HIV negative and HIV positive subjects [10,17,18]. In contrast, our data seems to confirm the findings of Liu et al (1999) in haemophilia patients, which indicate that there is no association between CCL5 polymorphisms and susceptibility to HIV [11]. On the one hand, the potential effect may be shaded by the extent of viral exposure. Intravenously infected patients are exposed to a much greater amount of viruses than those infected through other routes [32]. On the other hand, the population studied in here is extremely hard to recruit and we acknowledge that associations in this group might be insignificant due to low power. Therefore, we cannot draw final conclusions in the context of HIV infection.
Some limitations of the study should be acknowledged. First, the duration of IVDU was only known for two thirds of the population (lacking from prison population). However, we assume that the duration of IVDU in the syringe exchange population reflects the IVDU in the prison population because these persons are similar in terms of demographic and risk behaviors as well as the short duration of the HIV epidemic in Estonia in general (Zettenberg et al., , Adojaan et al., 2005. Second, the lack of information on disease stages of HIV and HCV infection reflects the seroprevalent nature of the cohort and thus we were not able to analyze the effect of haplotypes on disease progression. In addition, previous studies have shown the association of HCV clearance with genetic diversity of hepatitis A virus cellular receptor 1 and HCV genotypes [33]. Unfortunately, there was no data available with regard to HCV RNA and HCV genotypes, so we could not analyze the associations between HCV clearance, HCV genotypes and CCL5 haplotypes. However, from the demographic background we know that in Estonia, the main HCV genotype is Ib (60% to 70%) followed by IIIa (around 30%) in HCV monoinfected as well as in HIV/HCV co-infected population [34,35]. Third, we were unable to draw conclusions with regard to the association between CCL5 haplotype D and HIV/HCV co-infection because of low proportion of HIV monoinfected subjects. Lastly, we acknowledge that we do not have access to CCL5 haplotype frequency in well-controlled normal population. However, 24% of our studied population is HCV seronegative and 16% of them are seronegative for both HCV and HIV. They had same demographic and similar risk behavior characteristics compared to the HCV seropositive population. In addition, we found that the CCL5 haplotype frequency in our studied population is comparable to other presumably HIV and HCV Caucasian populations (www.hapmap.org). Thus, the likelihood of our findings to be confounded due to the lack of control population is very low. Nevertheless, we believe that these limitations do not diminish the importance of our results.
In conclusion, we suggest that in PWID exposed to HIV and HCV, CCL5 haplotype D (-471A/-96G/In1.1C) and CCR5 HHG Ã 1 protect against HCV acquisition. Our findings highlight the importance of CCL5 variability and the CCL5-CCR5 axis on HCV acquisition.
Supporting Information S1 Table. Studies of associations between polymorphisms in CCL5 and acquisition of HIV-1 or HCV disease progression. (DOC)