CD8⁺ T Cell Activation in Women Coinfected with Human Immunodeficiency Virus Type 1 and Hepatitis C Virus

Participants, materials, and methods

This is a substudy of the Women’s Interagency HIV Study (WIHS), a multicenter, prospective study of the natural history of HIV-1 infection and associated diseases in US women. During a single visit at 2 WIHS sites (one in Los Angeles and another in Chicago), we evaluated 169 women with and 51 women without HIV-1 infection.

We determined plasma HIV-1 RNA levels by means of the isothermal nucleic acid sequence– based amplification method (bioMérieux) in laboratories that participate in and are certified by the National Institute of Allergy and Infectious Diseases Virology Quality Assurance certification program.

At baseline, we determined HCV serostatus by means of the Abbott EIA 2.0 and 3.0 and HCV RNA level by means of the COBAS Amplicor Monitor 2.0 (detection range, 600 –500,000 IU/mL [Roche Diagnostics]) or the COBAS TaqMan (detection range, 10 –2 × 10 ^{IU/mL [Roche Diagnostics]); qualitative PCR (Amplicor 2.0; lower limit of detection, 50 IU/mL) was performed if HCV RNA was not detected, as previously reported [}6].

To determine real-time levels of CD4 ^{and CD8 T cell subsets, fresh whole blood specimens were collected in EDTA tubes and subjected to 3-color flow cytometry (FACSCalibur [Becton Dickinson]) [}7], in accordance with AIDS Clinical Trials Group consensus protocol. The analysis used fluorochrome-conjugated antibodies (anti-CD3, anti-CD4, anti-CD8, anti-CD19, anti–HLA-DR, anti-CD38, anti-RA, anti-62L, anti-CD28, anti-CD16, and anti-CD56 [Becton Dickinson]) and antibody combinations (anti–HLA-DR/anti-CD38 to measure T cell activation, anti-CD45RA/anti-CD62L to measure the levels of memory and naive cells for CD4 ^{and CD8subsets, and anti-CD8/anti-CD28 to measure T cell maturation, with HLA-DR as a marker of activation).}

χ ^{tests were used for comparison of proportions, and Kruskal-Wallis tests for comparison of median values of demographic, clinical, and immunological characteristics. Analyses of covariance, stratified by HIV-1 infection status, were used to investigate the effect of HCV status (HCV antibody positive [HCV] and HCV RNA positive [RNA], HCVand HCV RNA negative [RNA], and HCV antibody negative [HCV]) on the number and percentage of T cell subsets. Analyses were adjusted for age (<30, 30 –39, or ⩾40 years), race (black, white non-Hispanic, Hispanic, or other), injection drug use (yes or no), and HIV-1 treatment (any antiretroviral therapy or no antiretroviral therapy). Because of the nonnormal distribution of immunological markers, analyses of covariance used ranked data. Additional analyses of covariance also adjusted for CD4 T cell count (<200, 200 –500, or >500 cells/mm) and HIV-1 RNA level (not tested, <4000, 4000 –50,000, or >50,000 copies/mL) separately and together. Spearman rank correlation tests were used to evaluate relationships between the percentage of activated CD8 T cells and the HIV-1 RNA load among HIV-1–positive [HIV] women, by HCV status, and between the percentage of activated CD8 T cells and the HCV RNA load among HCV women, by HIV-1 infection status. A 2-tailed} P value of <.05 was considered to be statistically significant. SAS statistical software, version 9 (SAS Institute), was used to conduct these analyses.

Results

Of the 220 women enrolled in this study, 52 were HIV^{HCVRNA, 16 were HIVHCVRNA, 101 were HIVHCV, 11 were HIV-1 negative (HIV) and HCVRNA, 4 were HIVHCVRNA, and 34 were HIVHCV; 2 HIV women were excluded because of indeterminate results of HCV RNA tests. The majority of women were black (48.6%) or Hispanic (30.7%) and 30 –39 years of age (49.1%) (}table 1). More than 83% of women who were HIV^{HCVRNA or HIVHCVRNA reported a history of injection drug use.}

The course of HIV-1 disease in the majority of HIV ^{women was early: only 21.9% had a CD4 T cell count of <200 cells/mm, and 20.7% had a plasma HIV-1 RNA load of >50,000 copies/mL. Almost 40% had a CD4 T cell count of >500 cells/mm, and nearly 50% had an HIV-1 RNA level of <4000 copies/mL. Only 7.7% were receiving HAART, because HAART was not universally available at the time of the study. Of the HIVHCV women, 76.5% were RNA at baseline (median HCV RNA load, 1,985,000 IU/mL), and none were receiving HCV therapy.}

HIV^{HCVRNA women had a significantly increased percentage of activated CD8HLA-DR38 T cells, compared with HIVHCV women, after adjustment for age, race, injection drug use, and HIV-1 treatment (mean percentages, 30.7% for HIVHCVRNA women, 26.4% for HIVHCVRNA women, and 13.8% for HIVHCV women; HCVRNA effect, 16.9%;} P = .008) (table 2). Furthermore, we found that HIV^{HCVRNA women had a marginally higher number of activated CD8 T cells, compared with HIVHCV women (mean counts, 326 cells/mm for HIVHCVRNA women, 264 cells/mm for HIVHCVRNA women, and 167 cells/mm for HIVHCV women; HCVRNA effect, 159 cells/mm;} P = .07). These findings remained constant after separate adjustments for HIV-1 RNA load only, CD4 ^{T cell count only, and HIV-1 RNA load plus CD4 T cell count. In separate analyses of HIV women, the increased percentage of activated CD8 T cells for HIVHCVRNA women was similar to that for HIVHCV women, regardless of treatment status. However, because of sample size, differences were not statistically significant among untreated women (mean percentages, 24.1% for HIVHCVRNA women, 23.7% for HIVHCVRNA women, and 9.0% for HIVHCV women; HIVHCVRNA effect, 15.1%;} P = .22) but were statistically significant among treated women (mean percentages, 40.3% for HIV^{HCVRNA women, 31.0% for HIVHCVRNA women, and 18.4% for HIVHCV women; HCVRNAeffect, 21.9%;} P = .005). There were no differences in the percentage or number of activated CD4 ^{T cells between the study groups.}

The percentages of CD8^{HLA-DR38 T cells in relation to HIV-1 RNA load and HCV RNA load are presented in} figure 1. Spearman rank correlation tests demonstrated correlations of borderline significance between HIV-1 RNA level and immune activation among HIV^{HCVRNA women (Spearman rank-correlation coefficient [hereafter, “Spearman coefficient”], 0.26;} P = .08) and HIV^{HCVRNA women (Spearman coefficient, 0.47;} P = .06) but not among HIV^{HCV women (Spearman coefficient, 0.13;} P = .23) (figure 1). Among the HIV^{HCV women, there was no correlation between HCV RNA level and percentage of activated CD8 T cells (Spearman coefficient, 0.13;} P = .29); however, among HIV^{HCV women, the correlation between these values was statistically significant (Spearman co-efficient, 0.57;} P = .03) (figure 1). We also evaluated the number and percentage of naive and memory CD4^{T cells and CD8 T cells with respect to HIV-1 infection status, HCV antibody status, and HCV RNA status but found no significant differences between HIVHCVRNA women, HIVHCVRNA women, and HIVHCV women or between HIVHCVRNA women, HIVHCVRNA women, and HIVHCV women, after adjustment for age, race, injection drug use, and HIV-1 treatment (data not shown).}

Figure 1

A, Relationship between the plasma human immunodeficiency virus type 1 (HIV-1) load and the percentage of CD8^{DR38 cells, by hepatitis C virus (HCV) antibody and HCV RNA statuses, among women with HIV-1 infection (HIV). Lower limit of detection (LLD), ≤2.60 log}₁₀ copies/mL (≤4000 copies/mL). B, Relationship between the baseline plasma HCV RNA load and the percentage of CD8^{DR38 cells, by HIV-1 infection status, among HCV antibody–positive [HCV] women. LLD, ≤1 log}₁₀ IU/mL (≤10 IU/mL). HCV^{, HCV antibody negative; HIV, HIV negative; RNA, HCV RNA positive; RNA, HCV RNA negative.}

A subset of women was further evaluated for the expression of CD28, an important costimulatory receptor needed for activation, and the presence of HLA-DR, which is a marker of activation, on CD8 ^{T cells (}table 1). As reported earlier, in this study the count and percentage of CD8^{CD28 T cells in HIV women were less than those in HIV women. Furthermore, we found that HIVHCVRNA women had a significantly higher percentage of CD8CD28HLA-DR T cells, compared with HIVHCV women, after adjustment for age, race, injection drug use, and HIV-1 treatment (mean percentages, 19% for HIVHCVRNAwomen, 15.1% for HIVHCVRNA women, and 12.8% for HIVHCV negative women; HCVRNA effect, 6.2%;} P = .01).

Discussion

Our study demonstrates that the level of CD8 ^{T cell activation is higher in HIVHCV women than in HIVHCV women early during the course of HIV-1 disease and before HAART initiation. This was not the case for CD4T cell activation, which was not further impacted by HCV infection. Furthermore, the level of CD8 T cell activation was significantly higher among HIVHCV viremic women than among HIVHCV women, although HIVHCV nonviremic women also had higher activation levels. Finally, coinfected women also had an increased percentage of activated immature memory CD8 T cells. Because activation of CD8 T cells is related to T cell maturation and ultimately to HIV-1 disease progression, these data may have implications for the medical management of coinfected patients.}

CD8^{T cell immune responses are essential for the control of viral infection. Early after HIV-1 infection, there is a significant expansion of the CD8 T cell population. Activation of T cells occurs following antigen presentation, and the level of activation is related to the antigen load [}9 –11]. However, in chronic viral infections such as those due to HIV-1 and HCV, lack of clearance of HIV-1 and/or HCV results in continued activation of T cells and a chronic state of immune activation, ultimately resulting in T cell “exhaustion” characterized by anergy and senescence. Late in the course of HIV-1 disease, there is a decrease in the CD8 ^{T cell population. This is most likely due to apoptosis of cells that coexpress CD8 and CD38, as such cells have been shown to be more susceptible to apoptosis when there is ongoing viral replication. The combination of HIV-1 and HCV infection may further accelerate this process, because both viruses can increase the rate of apoptosis [}9 –11].

Our study also assessed for CD8 ^{T cells that coexpress CD28, an important costimulatory molecule that is necessary for functional responses and is expressed on naive T cells and early memory T cells. CD28 ligation with T cell receptor is essential for activation and expansion of CD8 T cells [}9 –12]. Cells that express CD28 are considered to be early differentiation T cells, whereas those that are CD28 ^{are late-differentiation memory cells. Recent studies have shown that viruses have the ability to impact maturation of CD8 T cells as a mechanism for immune evasion. In the case of HIV-1, there is a noted increase in HIV-1–specific CD8 T cells of the intermediate memory phenotype (i.e., CD8CD28CD27). On the other hand, HCV-specific CD8 T cells have an earlier memory phenotype (i.e., CD28CD27), whereas CMV-specific T cells have a late effector phenotype (i.e., CD2728) [}9]. Our current study confirms previously reported findings that HIV ^{persons have fewer CD8CD28 T cells than do HIV persons, whereas HCV persons have more CD8CD28 T cells than do HCV persons [}8, 12, 13]. Additionally, we found that HIV^{HCVRNA women have increased CD828HLA-DR T cells. We previously published the finding that coinfected women have increased primed activated T cells (i.e., CD8CD45ROCD2795), compared with HIVHCV women [}6], and postulated that this population of cells represented effector/central memory T cells. The population of activated CD8^{28cells in our current study may also represent early central memory T cells and further suggests that HCV may be playing a role in skewing CD8 maturation toward an earlier phenotype [}6, 9, 11–13].

In our previous study involving women who initiated HAART, we found no difference in percentages of activated CD4 ^{or CD8 T lymphocytes between HIVHCV and HIVHCV women, but we found decreased activation after HAART initiation. In our current study, HIV-1 disease in most women was at an early stage: almost 50% were untreated, with high CD4 T cell counts and low HIV-1 RNA levels, and only 7.7% were receiving HAART. Immune activation was marginally correlated with HIV-1 RNA level but not correlated with HCV RNA level among coinfected patients. Even after adjustment for HIV-1 RNA level, we found a marked increase in the percentage of CD8HLA-DR38 and CD828HLA-DR T cells in HIVHCV women. This is in agreement with a previous report that found increased activation in coinfected patients after HAART initiation [}14], but the present study involved a much larger number of patients and was limited to women. Furthermore, although treated women overall had greater levels of immune activation, most likely because treatment was selective, HIV^{HCV viremic treated women still had a significantly greater percentage of activated CD8 T cells. RNA women also had a greater percentage of activated CD8T cells. Activation of T cells may be occurring in the liver or at extrahepatic sites in some of these women, as previously reported by us [}15]. Although our study was large, one limitation is that we did not adjust for multiple comparisons; therefore, some of the significant findings might be due to chance.

In conclusion, our study suggests a potential mechanism for the observed increase in disease progression and immune dysregulation noted in some studies of patients coinfected with HIV-1 and HCV. Global activation of CD8 ^{T cells in the liver may result in CD8 dysfunction, including cytokine production. This in turn would lead to a defective response to invading pathogens, a decrease in T cells, and, ultimately, the development of opportunistic infections.}

Acknowledgments