Stronger Hepatitis C Virus-specific CD8+ T-cell Responses in HIV Coinfection

[Guest guest]

Stronger Hepatitis C Virus-specific CD8+ T-cell Responses in HIV Coinfection

Abstract

Hepatitis C virus (HCV) is a widespread chronic infection that shares routes of transmission with human immunodeficiency virus (HIV). Thus, coinfection with these viruses is a relatively common and growing problem. In general, liver disease develops over years with HIV coinfection, when compared to decades in HCV monoinfection. The role of the immune system in the accelerated pathogenesis of liver disease in HIV/HCV coinfection is not clear. In this study, we compared the frequency, magnitude, breadth and specificity of peripheral blood CD4+ and CD8+ T-cell responses between HCV-monoinfected and HCV/HIV-coinfected individuals and between HIV/HCV-coinfected subgroups distinguished by anti-HCV antibody and HCV RNA status. While HIV coinfection tended to reduce the frequency and breadth of anti-HCV CD8+ T-cell responses in general, responses that were present were substantially stronger than in monoinfection. In all

groups, HCV-specific CD4+ T-cell responses were rare and weak, independent of either nadir or concurrent CD4+ T-cell counts of HIV-infected individuals. Subgroup analysis demonstrated restricted breadth of CD8+ HCV-specific T-cell responses and lower B-cell counts in HIV/HCV-coinfected individuals without anti-HCV antibodies. The greatest difference between HIV/HCV-coinfected and HCV-monoinfected groups was substantially stronger HCV-specific CD8+ T-cell responses in the HIV-coinfected group, which may relate to accelerated liver disease in this setting.

Introduction

Hepatitis C virus (HCV) is a global problem, with over 170 million people infected. The insidious and chronic nature of HCV infection combine to create a large population of infected individuals unaware of their status. As HCV and human immunodeficiency virus (HIV) share similar routes of transmission, a substantial proportion of HIV-infected individuals are HCV coinfected. In this setting, the clinical course of HCV disease is accelerated and response to HCV therapy may be compromised.[1–5] With the increased life expectancy afforded by highly active antiretroviral therapy (HAART) in HIV infection, HCV-related disease has emerged as a significant cause of morbidity and mortality in HIV/HCV-coinfected individuals.[6]

The role of adaptive immunity in HCV clearance and disease pathogenesis remains poorly understood. Multiple studies in humans[7–10] and chimpanzees[11,12] associate viral clearance with strong anti-HCV CD4+ and CD8+ T-cell responses in acute infection, suggesting a central role for adaptive immunity. Vigorous and broadly directed HCV-specific CD4+ and CD8+ T-cell responses occur during spontaneous clearance, while these responses appear much less robust in chronic HCV infection.[8,13,14] Most found that HCV-specific CD8+ T cells demonstrate proliferative defects and HCV-specific

CD4+ T-cell responses are rare in persistent infection.[15–17] There is also evidence that CD4+ CD25+ regulatory T cells suppress anti-HCV CD8+ T-cell responses and promote chronic HCV infection.[18–20]

The impact of HIV coinfection on HCV-specific CD4+ and CD8+ T-cell responses has been investigated by several groups.[21–26] They report that T-cell responses are weak and narrow in the periphery, but stronger in the liver. Some studies report similar frequencies of CD4+ and CD8+ T-cell responses in monoinfection and HIV/HCV coinfection,[22] while others report rare[25,27] CD4+ T cell responses in the presence of weak but detectable CD8+ T-cell responses in HIV/HCV coinfection. The HCV-coinfected subgroup segregates into multiple groups: HCV-exposed, anti-HCV antibody positive, HCV

RNA-negative individuals (spontaneous clearers), chronic infections and chronically HCV-infected people without detectable anti-HCV antibodies (serosilent group). These subgroups have not been well studied in the context of HIV infection and may have distinct adaptive immune responses related to their status that are not apparent through global analyses.

In this study, we compared the frequency, magnitude, breadth and specificity of HCV-specific CD4+ and CD8+ T-cell responses between HIV/HCV-coinfected individuals and those with chronic HCV infection alone. We found the frequency and breadth of anti-HCV CD8+ T-cell responses was slightly decreased in HIV coinfection compared to chronic HCV monoinfection and HCV-specific CD4+ T-cell responses were rare in both groups, independent of CD4+ T-cell count for HIV-infected individuals. The clearest difference between groups was higher magnitude HCV-specific CD8+ T-cell responses in HIV-coinfected individuals. Not all coinfected groups were similar, with subgroup analysis indicating narrower anti-HCV CD8+ T-cell responses and lower B-cell counts in coinfected individuals lacking anti-HCV antibodies. Stronger HCV-specific CD8+ T-cell responses in subgroups of

HIV-coinfected individuals may contribute to rapid progression of immunologically mediated liver disease.

Study Participants

Individuals that were HIV-infected or HIV/HCV-coinfected were recruited from the St. 's General Hospital Infectious Disease Clinic, St. 's, NL, Canada. Individuals infected with HCV were recruited from the Capital Health Queen Hospital Hepatitis Clinic in Halifax, NS, Canada. Ethics approval for this project was obtained from the Human Investigation Committee at each institution, and all subjects provided informed written consent for blood collection and immunological studies.

Determination of HCV Infection Status

Exposure to HCV was ascertained by testing for serum anti-HCV antibodies using second or third generation EIA assays from Ortho Diagnostics. To demonstrate the presence of HCV RNA, total nucleic acids were extracted from plasma samples and reverse-transcribed with HCV-specific primers. Details are found in the supplementary material section. If individuals were serosilent, samples from multiple dates were tested for HCV RNA. Serosilent individuals were tested on multiple dates for antibodies.

Peripheral Blood Mononuclear Cell Isolation

Acid citrate dextrose-treated whole blood was obtained by venipuncture from each individual, and peripheral blood mononuclear cells (PBMC) were isolated by Ficoll-HyPaque Plus (GE Health Care, Baie d'Urfé, PQ, Canada) density gradient centrifugation. Cells were washed, counted and suspended at 1 × 106/mL in lymphocyte medium (RPMI supplemented with 10% FCS, 10 mm HEPES, 2 mm l-glutamine, 1% penicillin/streptomycin and 2 × 10−5 m 2-mercaptoethanol; all from Invitrogen, Burlington, ON, Canada) until use. In some cases, cells were frozen at −80 °C until use. Samples with <50% viability when thawed were discarded.

ELISPOT Assay for HCV-specific Interferon-γ Production

Eighteen (18)-mer peptides spanning the entire HCV genotype 1a polyprotein were obtained from the NIH AIDS Research and Reference Reagent Program. A total of 441 peptides overlapping by 11 amino acids were reconstituted in 80% sterile dimethylsulphoxide and then pooled in groups of 10 consecutive peptides at a concentration of 20 μg/mL peptide-1. Peptides were aliquoted and stored at −80 °C until use. Peptides were added to cells at a final individual concentration of 10 μg/mL. The number of HCV-specific IFN-γ-producing cells in PBMC was determined by ELISPOT (detailed methods are in the supplementary material section) and enumerated with the high-resolution Zeiss reader system and associated KS software (Carl Zeiss Canada, ON, Canada). Responses were converted to IFN-γ-producing cells per million PBMC by multiplying as appropriate. A peptide pool response was considered positive if the number of spots was greater than 10 and more than twice

background. With two coinfected samples, there was no difference in the number of peptide-induced IFN-γ+ spots with or without CD4+ cell depletion (data not shown). As each spot represents one cell, the area of the spot is related to IFN-γ production per cell and average spot area in arbitrary units serves as a relative measure of IFN-γ production per cell.[28,29]

Proliferation Assays

Proliferative cellular immune responses were measured by standard 5-day thymidine incorporation assay (details in the supplementary material).

Statistical Analyses

All statistical analyses were performed with SPSS version 9 (SPSS Inc., Chicago, IL, USA). Means were compared using either the Mann–Whitney U test or Student's t-test depending upon whether distribution of the data was normal. All graphs indicate average values, and error bars indicate the standard error of the mean. Pearson's correlation coefficients were used to assess relationships between independent continuous variables by linear regression.

Results

Demographic and Baseline Clinical Parameters of Study Participants

To investigate the effects of HIV coinfection on HCV-specific cellular immunity, we recruited 22 individuals with chronic HCV infection and 12 coinfected with HIV and HCV for this study. Baseline characteristics are reported in Table 1. Mean age (40 in the HIV/HCV groups and 47 in the HCV group) and gender distribution (75% and 77% men, respectively) did not differ significantly between the two groups. Those HCV genotypes determined were predominantly HCV subtype 1, as expected for North American infections. Intravenous drug use, blood product transfusion and organ transplant were the predominant identifiable risk factors in both groups. Estimated duration of HCV infection was similar in both groups with an average of 20 years since infection. Liver enzymes were considered elevated if plasma levels were greater than twice the upper limit of normal. All individuals in

the coinfected group and 68% in the chronic HCV group had normal plasma liver enzyme levels on sampling dates. Median ALT levels were similar in both the HIV/HCV-coinfected and HCV-monoinfected groups (57, IQR 45–77.5 vs 59, IQR 33–93.5, respectively). All coinfected individuals with the exception of HIV-087 were adherent to highly active antiretroviral regimens for greater than 6 months. There was no acute increase or decrease in CD4+ T-cell counts during the study period. Overall, the monoinfected and HIV/HCV-coinfected groups were similar in terms of their general demographics and HCV-related clinical parameters.

HIV/HCV-coinfected Subgroups

Table 2 illustrates three distinct groups of HIV-infected individuals exposed to HCV: those HCV seropositive and HCV RNA positive (chronic infection, n = 5); those HCV seropositive and HCV RNA negative (spontaneous clearers, n = 3); and those HCV seronegative and HCV RNA positive (serosilent HCV infection, n = 4). The HIV-associated parameters are described for the HIV/HCV-coinfected group in Table 2 and HCV-related parameters in Table 1 and Table 2. All subgroups were similar in age, gender distribution, duration of HIV infection and type of antiretroviral therapy at the time of testing. Only one

individual, HIV-087, was not receiving antiretroviral therapy at the blood collection time point. There were no statistically significant differences in either contemporaneous or nadir CD4+ T-cell counts between subgroups. Plasma HIV viral load was also not significantly different between groups, ranging from undetectable (<50 or <400 copies per mL) to 130 000 RNA copies per mL plasma. All liver function tests and plasma liver enzyme levels were normal on the date of testing. None of the HCV serosilent (lack of anti-HCV antibodies), self-limited HCV infection (undetectable plasma HCV RNA) or chronic infection groups were associated with a significant difference in any HIV-related parameter such as CD4+ T-cell count, CD4+ T-cell nadir or HIV plasma viral load. The HCV serosilent group lacked anti-HCV antibodies at all time points tested, suggesting that none of these individuals represented acute

seroconversions.

The Effect of HCV Infection on HIV-related Immunological Parameters

Parameters used to assess HIV disease progression were compared between the HIV/HCV-coinfected group and a control group of 75 HIV-infected individuals to assess the impact of HCV infection on HIV disease progression (Fig. 1a). Mean CD4+ T-cell counts were similar (370 ± 245 in the HIV group vs 380 ± 259 in the HIV/HCV group) as were CD4+ T-cell nadir (181 ± 175 and 255 ± 197 respectively), HIV viral loads (3.27 ± 1.08 and 3.80 ± 1.19 respectively) and CD8+ T-cell counts (886 ± 445 and 892 ± 582 respectively). B-cell counts (CD20+) were also similar in the HIV/HCV group and HIV-monoinfected group (248 ± 163 vs 194 ± 134, respectively); however, the 4 HCV-seronegative, HCV RNA-positive individuals showed a trend towards fewer B cells compared to the HCV-seropositive-coinfected group (mean 134 ± 65 vs 306 ± 187, Fig. 1b; t = 0.893, P = 0.398), although

this was not significantly different. Coinfection with HCV did not affect HIV-related clinical laboratory parameters in our study group, except for the trend towards lower B-cell counts in the group with plasma HCV RNA and no detectable anti-HCV antibodies.

(Enlarge Image)

Figure 1.

Comparison of immunological parameters in HIV-infected and HIV/HCV-coinfected study groups. (a) CD4+ and CD8+ T-cell counts, B-cell counts and HIV plasma virus loads (right hand axis) are shown for the sampling date. The CD4+ T-cell nadir was the lowest CD4+ T-cell count at any time in clinical records. Vertical bars represent the mean value for each parameter compared between study groups with error bars showing standard error of the mean. ( B-cell counts are shown for each HCV seropositive and seronegative HCV-exposed, HIV-coinfected individual with horizontal bars denoting the mean B-cell count for each group.

[ CLOSE WINDOW ]

Figure 1.

Comparison of immunological parameters in HIV-infected and HIV/HCV-coinfected study groups. (a) CD4+ and CD8+ T-cell counts, B-cell counts and HIV plasma virus loads (right hand axis) are shown for the sampling date. The CD4+ T-cell nadir was the lowest CD4+ T-cell count at any time in clinical records. Vertical bars represent the mean value for each parameter compared between study groups with error bars showing standard error of the mean. ( B-cell counts are shown for each HCV seropositive and seronegative HCV-exposed, HIV-coinfected individual with horizontal bars denoting the mean B-cell count for each group.

Frequency and Specificity of HCV-specific CD8+ T-cell Responses

To assess the frequency specificity, breadth and magnitude of anti-HCV CD8+ T-cell responses in PBMC from HIV/HCV-coinfected and HCV-monoinfected individuals, we carried out IFN-γ ELISPOT assays with an overlapping set of 441 18-mer peptides spanning the entire HCV polyprotein. Peptides were grouped in 44 sequential pools of 10 from the amino through carboxy terminus of the HCV polyprotein. Results were expressed both in terms of individual pools and HCV protein regions, with different numbers of peptide pools in relation to the protein size: 3 pools for core; 2 for E1; 6 for E2; 1 for p7; 3 for NS2; 9 for NS3; 2 for NS4A; 3 for NS4B; 7 for NS5A; 8 for NS5B.

Eighteen chronic HCV-infected individuals and 11 HIV/HCV-coinfected individuals were tested with 14/18 (78%) and 8/11 (72%), respectively, showing anti-HCV IFN-γ responses. Supplementary Figure S2 illustrates differences in the specificity and frequency of HCV-specific CD8+ T-cell responses between groups, depicted as the percentage of individuals from each group responding to each individual peptide pool. More than 40% of the coinfected group responded to NS3 region pool 18 and at least 30% responded to NS3 pool 16, NS3 pool 22 and NS5B pool 44. There were no responses to pools 2, 10–13, 15, 17, 31, 34, 36, 37 and 41. In the chronic HCV monoinfection group (supplementary Figure S1, panel 2), 40% of individuals responded to core pool 2 and NS3 pool 19. More than 30% responded to NS3 pool 16, NS3 pool 17, NS4A pool 25 and NS5A pool 30. There were no responses to pools 10, 36 and 38. Of note, no responses to pools 10 or 36 were detected

in either group. When results were expressed on a per region level (Fig. 2a), both groups had individuals with responses to all three structural proteins (core, E1 and E2). Responses to peptide pools from all nonstructural proteins occurred in each group except for p7 where no HIV/HCV-coinfected individuals responded. The most frequent region-specific responses for the coinfected group were NS3 (80%), NS5A and NS5B (both 40%). The HCV-monoinfected group also had the most frequent responses against NS3 (67%), NS4A (53%) and NS5A (50%). In general, the immunogenicity of different HCV proteins was similar in the HCV-infected and HIV-coinfected groups. While not statistically significant, there was a trend towards fewer NS2 and NS4A region responses in the HIV-coinfected group compared to the chronic infection group.

(Enlarge Image)

Figure 2.

Distribution of HCV-specific CD8+ T-cell responses by region in HCV and HIV/HCV-coinfected study groups. Peripheral blood mononuclear cells were incubated with pools of ten overlapping 18-mer HCV peptides for 18 hours and IFN-γ-producing cells enumerated by ELISPOT. (a) The percentage of HCV-monoinfected individuals responding to peptides from each HCV protein are shown in black bars with grey bars showing percentages adjusted for protein size (number of peptide pools per region). ( The percentage of HIV/HCV-coinfected individuals responding to peptides from each HCV protein are shown in black bars with grey bars showing percentages adjusted for protein size.

[ CLOSE WINDOW ]

Figure 2.

Distribution of HCV-specific CD8+ T-cell responses by region in HCV and HIV/HCV-coinfected study groups. Peripheral blood mononuclear cells were incubated with pools of ten overlapping 18-mer HCV peptides for 18 hours and IFN-γ-producing cells enumerated by ELISPOT. (a) The percentage of HCV-monoinfected individuals responding to peptides from each HCV protein are shown in black bars with grey bars showing percentages adjusted for protein size (number of peptide pools per region). ( The percentage of HIV/HCV-coinfected individuals responding to peptides from each HCV protein are shown in black bars with grey bars showing percentages adjusted for protein size.

Breadth of HCV-specific CD8+ T-cell Responses

The proportion of individuals responding to at least one HCV peptide pool was similar between groups as 8/11 HIV-coinfected (72%) and 14/18 HCV-infected individuals (78%) had anti-HCV IFN-γ responses. The median percentage of positive peptide pools in the HIV-coinfected group was 5% (IQR 0–14%) compared to 11.5% (IQR 1.5–25.5%) for the HCV-infected group, suggesting a somewhat restricted breadth to the HCV-specific CD8 ± T-cell responses in the HIV-coinfected group (Mann–Whitney 108.5, P = 0.342). When responses were compared on the basis of HCV protein (Fig. 2a,, there was little difference between HIV-coinfected and HCV-infected groups. In the HIV-coinfected group, we observed responses against 9/10 proteins; however, the proportion of individuals with responses to any one region was generally low (<50%). The exception was the NS3 region, where 82% of individuals responded. Responses were detected against 10/10 proteins in

the chronic HCV group. As in the coinfected group, the highest proportion of individuals responded to NS3 (67%), with greater than 50% of individuals also responding to NS4A and NS5A. As the size of the protein is one factor affecting the probability of generating immunogenic peptides, we adjusted for the effect of protein size. After this correction (grey bars of Fig. 2a,, it was clear that one of the reasons NS3 was so dominant was because it is the largest region in the genome. The response breadth did not change with size correction but differences in proportions between groups equalized. Responses against E1 became the most frequent on a per peptide basis in the HIV-coinfected group, and responses against NS4A became the most frequent in the chronic group. With or without correction for protein size, differences between the HCV-infected and HIV-coinfected group were relatively subtle with a trend towards reduced overall breadth of recognition and

reduced recognition of NS4a in the coinfected group.

HCV-specific CD8+ T-cell Responses in HIV-coinfected Subgroups

Supplementary Figure S2 illustrates differences in the specificity and frequency of anti-HCV IFN-γ responses between HIV-coinfected subgroups. In the chronic HCV group (n = 7), there were responses to all regions except p7 and NS2 (black bars) with responses against NS3 and NS5B the most frequent. In the HCV-serosilent group (n = 2), responses were limited to NS3 and NS5A. The two spontaneous clearers tested had responses against all proteins except p7, with both individuals responding against NS3. Responses to NS3 were most frequent, partly because of its size; however, the most commonly recognized peptide pool for all groups is also in the NS3 region, suggesting a true immunological hotspot therein.

Magnitude of HCV-specific CD8+ T-cell Responses

The average magnitude of anti-HCV responses per peptide pool as well as the cumulative number of HCV-specific SFU per 106 PBMC was compared between HIV-coinfected and HCV-infected groups after background subtraction. Many more IFN-γ-producing cells per positive pool were observed in the HIV-coinfected group (72 ± 22 SFC per 106 PBMC vs 21 ± 4.2 SFC per 106 PBMC, P = 0.02 Fig. 3a), and the cumulative magnitude of the HCV-specific response was also significantly higher in the HIV-coinfected group (1050 ± 369 SFC per 106 PBMC vs 162 ± 47 SFC per 106 PBMC P = 0.003 Fig. 3b). The greater magnitude of HCV-specific CD8+ T-cell responses in the HIV-coinfected group was the most substantial difference we observed between the two groups in terms of HCV-specific T-cell immunity.

(Enlarge Image)

Figure 3.

Comparison of pool-specific and cumulative magnitude of HCV-specific CD8+ T-cell responses in HIV/HCV-coinfected and HCV-infected individuals. Peripheral blood mononuclear cells were incubated with pools of ten overlapping 18-mer HCV peptides for 18 hours and IFN-γ-producing cells enumerated by ELISPOT. (a) Grey bars indicate the mean number of IFN-γ-producing cells responding to each positive pool in HIV/HCV-coinfected and HCV-infected groups with standard error of the mean shown. ( Grey bars indicate the mean cumulative number of IFN-γ-producing cells responding to the entire set of HCV peptides in HIV/HCV-coinfected and HCV-infected groups.

[ CLOSE WINDOW ]

Figure 3.

Comparison of pool-specific and cumulative magnitude of HCV-specific CD8+ T-cell responses in HIV/HCV-coinfected and HCV-infected individuals. Peripheral blood mononuclear cells were incubated with pools of ten overlapping 18-mer HCV peptides for 18 hours and IFN-γ-producing cells enumerated by ELISPOT. (a) Grey bars indicate the mean number of IFN-γ-producing cells responding to each positive pool in HIV/HCV-coinfected and HCV-infected groups with standard error of the mean shown. ( Grey bars indicate the mean cumulative number of IFN-γ-producing cells responding to the entire set of HCV peptides in HIV/HCV-coinfected and HCV-infected groups.

Proliferative T-cell Response Magnitude, Frequency and Specificity

Proliferative T-cell responses against mitogen, recall antigen and HCV viral antigens were assessed in all groups by standard five-day proliferation assay. An HIV-monoinfected group (n = 75) was included to compare the mitogen and recall antigen responses with those of the HIV/HCV-coinfected group. Magnitudes of the mitogen responses as SI are shown on the right hand axis of Fig. 4. The mean SI was not significantly different between HIV-infected and HIV/HCV-coinfected groups (160 ± 21.1 vs 129 ± 37.1). The mean SI of the HCV-monoinfected group (373 ± 242) was higher than either HIV-infected group but the difference was not significant because of high standard deviations. This would suggest that the overall CD4+ T-cell function in the HIV/HCV group is marginally reduced compared to the HCV-monoinfected group. The antigen-specific response magnitude measured against Candida antigen differed between groups in a pattern

similar to the PHA responses (Fig. 4). The HIV/HCV-coinfected group tended towards a lower mean SI in response to low-dose recall antigen (mean ± SEM; 10.9 ± 4.6) than the HCV-infected group (27.5 ± 9.4, P < 0.06), and this trend was maintained at the higher Candida dose (Fig. 4). In summary, the frequency and magnitude of CD4+ T-cell responses were somewhat decreased in HIV/HCV-coinfected individuals compared to HCV-monoinfected individuals, but no moreso than the HIV-monoinfected group. Anti-HCV-specific T-cell proliferative responses were rare in both groups and seen only against HCV core and NS3.

(Enlarge Image)

Figure 4.

HCV-specific CD4+ T-cell responses to mitogen, recall antigen, and HCV proteins. Peripheral blood mononuclear cells were incubated in triplicate for 5 days with phytohemagluttinin, Candida albicans antigen or HCV proteins and proliferation measured by 3H thymidine uptake. Responses were reported as stimulation indices calculated as the ratio of 3H counts in test wells divided by background 3H incorporation.

Discussion

Most HCV-exposed individuals develop chronic infection. In the setting of HIV infection, HCV establishes chronic infection even more frequently[4,30] and the rate of development of liver disease accelerates. The role of HCV-specific immunity in the pathogenesis of liver disease is not well understood, particularly in the context of HIV coinfection. We used recombinant HCV proteins and synthetic peptides spanning the entire HCV polyprotein to investigate how HIV infection affects the overall character (breadth, frequency, specificity and magnitude) of cell-mediated immunity against HCV. CD8+ T-cell response breadth was reduced in a subgroup of HCV-seronegative-coinfected individuals. Otherwise, the frequency, breadth and specificity of HCV-specific CD8+ T-cell responses in the two groups were similar. However, the HIV-coinfected

individuals had significantly more HCV-specific IFN-γ-producing CD8+ T cells in their peripheral blood. If paralleled in the liver, the heightened CD8+ T-cell response could play a role in the accelerated immunopathogenesis of liver disease.

Approximately 70% of individuals in both the HCV-infected and HIV/HCV-coinfected groups had detectable HCV-specific CD8+ T-cell responses. This is a higher frequency than reported in some previous studies. Dutoit et al. [22] found approximately 40% of individuals produced IFN-γ responses, and Kim et al. [23] described anti-HCV CD8+ T-cell responses in only 50% of HIV/HCV-coinfected individuals. An earlier study reported HCV-specific CD8+ T-cell responses in only 2/32 chronically infected individuals and 0/11 HIV/HCV-coinfected people.[26] All groups used similar experimental conditions to this

study; however, their definition of a positive response was a minimum of 20–50 spots per 106 PBMC and >3 times background. This absolute value would correspond to a frequency of 0.002%, which is 100-fold less than that found in other common viral infections (e.g. EBV with up to 0.2%). On the assumption that extrahepatic CD8+ anti-HCV T-cell responses are rare in chronic infection, we decided this high threshold might exclude responses relevant to overall outcome. Therefore, we chose a threshold of twice background with an absolute number of 10 spots as a cut-off. If we applied the criteria from other studies, responder frequency was unaffected in the HIV-coinfected group but fell from 70% to 50% in the HCV-infected group. We would have still observed more frequent responses than previously reported. The responses of the HCV-monoinfected infected individuals were broadly directed but low magnitude, and often twice, but not

three times background. The ELISPOT readout we used may underestimate total HCV-specific CD8+ T-cell responses by relying only on IFN-γ production. Functionally, cytotoxic anti-HCV CD8+ T cells can have a "stunned" phenotype and produce little IFN-γ in both acute[31] and chronic[32] infection. The estimate is further limited by the use of suboptimal 18-mer peptides derived from a consensus sequence that varies from the endogenous sequences found in individual patients.

In agreement with previous reports, we found the breadth and specificity of CD8+ T-cell responses were similar in coinfected and chronic monoinfected groups, with NS3 responses most common on a per protein basis.[23] One particular region of NS3 (pools 18 and 19) had a large number of responders in each group. These pools contain amino acids 1240–1390, where 12 human CTL epitopes have been defined (Los Alamos HCV database, http://hcv.lanl.gov), suggesting a strong immunological focus within that region.

In contrast to previous studies,[22,23,25] we found the magnitude of HCV-specific CD8+ T-cell responses elevated with HIV coinfection, compared to chronic HCV infection alone, independent of either current or nadir CD4+ T-cell count. This may significantly contribute to disease pathogenesis in the context of persistent infection if CD8+ T cell-mediated destruction of HCV-infected hepatocytes is an active mechanism. It appears that HIV-coinfected individuals experience only the negative aspects of elevated HCV-specific CD8+ T-cell responses as they are less likely to clear infection[33,34] yet develop liver disease faster. However, in our study group, none of the HIV-infected individuals showed evidence of

active liver disease in terms of elevated plasma levels of hepatic enzymes, decreased platelets or elevated INR. Several mechanisms have been proposed to explain the decrease in effective clearance of HCV with HIV coinfection, despite the presence, and perhaps even higher than usual levels, of HCV-specific CD8+ T cells. These include impaired effector function,[22] incomplete differentiation,[35] and decreased CD28-mediated costimulation.[36] Our study indicates the number of HCV-specific CD8+ T cells is higher in HIV coinfection than in chronic HCV infection, but the response is still low in absolute numbers compared to other

chronic viral infections, which may reflect hepatic localization of HCV-specific CD8+ T cells.

Responses to mitogen and recall antigen were decreased in the HIV-coinfected group, suggesting a generalized effect of HIV infection on CD4+ T-cell responses beyond HCV-specific attenuation. The frequency of HCV-specific CD4+ T cell-responses was low in both the coinfected and HCV chronic infection groups. A recent study of HCV-specific CD4+ T cells in HIV coinfection noted attenuated IFN-γ production independent of CD4+ T-cell count.[27] The proliferation assay reflects IL-2 production more than IFN-γ production; therefore, production of both cytokines by HCV-specific T cells may be limited. This study was cross sectional and only provided CD4+ T-cell counts and HCV-specific CD4+ T-cell reactivity at one time point in a disease process that spans decades. The status of

CD4+ T-cell responses changes over time from acute through chronic HCV infection depending on viral replication and stage of liver disease, as well as the state of the host immune system as affected by concurrent HIV infection. The lack of CD4+ T-cell responses in the periphery in the context of detectable CD8+ T-cell responses may reflect enhanced CD8+ T-cell responses favored by the high CD8+ T-cell counts characteristic of HIV infection, compartmentalization of CD4+ T cells to the liver, subset exhaustion with constant antigenic stimulation or a very low frequency of HCV-specific CD4+ T cells not detectable by proliferation assay.

Infection with HIV affects the magnitude of HCV-specific CD8+ T-cell responses and also affects HCV-specific humoral responses. There were three individuals in this study with plasma HCV RNA in the absence of detectable anti-HCV antibodies. These individuals had extremely narrow HCV-specific CD8+ T-cell responses and no detectable CD4+ T-cell responses. Thus, they have a different profile from the other HIV/HCV-coinfected individuals in terms of both humoral and cellular HCV-specific immunity and provide an opportunity for further study into HCV disease processes in a novel immunological setting. They may have a different course of clinical disease and need selective monitoring, while also providing a 'natural' experiment to assess viral and host factors impacting their impaired immune response and development of liver disease.

A limitation of our studies is that they were conducted on peripheral blood samples. The frequency of HCV-specific CD4+ and CD8+ T cells is many fold higher in the liver of HCV-infected individuals, and this also holds true in the context of HIV coinfection.[24,37,38] Qualitative aspects of HCV-specific T-cell responses in the periphery also appear to differ from those in the liver. Intrahepatic T cells produce less IFN-γ and have a phenotype consistent with T-cell exhaustion (PD-1high, CD127low) compared to the effector phenotype observed in periphery (PD-1+, CD127−).[39] This may reflect differences in HCV antigen levels or the different characteristics of viral quasispecies

within intra- and extra-hepatic compartments, although a recent study showed the same viral sequences predominate in the liver and peripherally in chronic HCV monoinfection.[40]

Cells outside the liver that have been associated with autonomous HCV replication and infection include PBMC[41] and neutrophils[42] and higher levels of lymphoid cell infection with HCV have been associated with HIV coinfection.[43,44] Together with generalized CD8+ T-cell activation, increased HCV replication within lymphoid cells in the context of HIV coinfection may promote stronger HCV-specific CD8+ T-cell responses in peripheral blood. Neither treatment-associated 'cure' nor spontaneous clearance of HCV infection appears to truly eradicate HCV from the extrahepatic compartment.[45] In these cases, peripheral T-cell responses may be an indicator of ongoing viral replication and also offer direction in developing vaccines to target extrahepatic quasispecies that become predominant post-treatment.

We have demonstrated an increase in the magnitude of HCV-specific CD8+ T-cell responses in HIV coinfection consistent with accelerated hepatic immunopathology mediated by CD8+ T-cell destruction of HCV-infected hepatocytes. This supports a rationale for early aggressive treatment of HCV infection in HIV-infected individuals and further highlights concerns that HIV-related immune dysregulation and immune reconstitution can provoke overly vigorous responses to some pathogens. While the benefits of HAART are clear in terms of controlling HIV replication, reducing HIV to undetectable levels may not totally prevent potentially negative interactions between generalized effects of HIV infection on the immune system and countercurrent infections. Residual immune activation or a propensity towards hyperactivation of the immune system in HIV infection may exacerbate the effects of chronic infection with HCV and other pathogens.

http://www.medscape.com/viewarticle/741379?src=rss