Plasma viral load predicts genital tract human immunodeficiency virus (HIV) shedding in HIV-infected women. We investigated whether local mucosal T-cell activation (HLA-DR, CD38, CCR5, and Ki67) contributed to HIV shedding in the genital tracts of HIV-infected women. We showed that cervical cytobrush-derived T cells expressed higher frequencies of T-cell activation markers (CD38+ and HLA-DR+) than blood-derived T cells. Expression was significantly higher in HIV-infected women than in uninfected women. We found that the frequency of activated proliferating cervical T cells (Ki67+; Ki67+CCR5+) broadly predicted HIV shedding in the genital tract in HIV-infected women, independently of plasma viral loads. Furthermore, activated cervical T cells (HLA-DR+CD38+ and HLA-DR+CCR5+) and local HIV shedding were independently associated with CD4 depletion in the genital tract. These data suggest that the presence of high frequencies of activated T cells in the female genital mucosa during HIV infection facilitates both local HIV shedding and CD4 T-cell depletion.
The micro-environment in the female genital tract has a substantial influence on both the acquisition and transmission of human immunodeficiency virus type 1(HIV-1) during penetrative vaginal sex. Although several studies have shown that plasma viral loads are a strong predictor of HIV transmission risks , local factors in the genital tract, such as either concurrent sexually transmitted infections (STIs) or genital inflammation [2, 3], have also been shown to increase both shedding of HIV in the genital tract and susceptibility to infection via the genital route . Genital tract concentrations of interleukin 6 (IL-6) and interleukin 8 (IL-8) are significantly associated with STIs [5, 6]. It is assumed that this increase in pro-inflammatory cytokines drives HIV replication, although the mechanisms remain unresolved.
Although immune activation in response to invading pathogens is crucial to protective immunity, such responses may ironically also contribute to HIV pathogenesis in the female genital tract by providing the virus with a steady supply of susceptible target cells . HIV preferentially infects activated CD4+CCR5+ T cells in the female genital tract . In macaques, activated CD4+ T cells generally contain 5 times more simian immunodeficiency virus (SIV) RNA than do resting CD4+ T cells . Although rapid death of HIV-infected CD4 cells contributes to CD4+ T-cell depletion in HIV-infected individuals, it is clear that generalized immune activation is primarily responsible for HIV-associated disease progression [10, 11]. Similarly, nonpathogenic SIV infection in nonhuman primates is characterized by low levels of immune activation despite high viral load and severe CD4 T-cell depletion .
An improved understanding of how immune activation in the female genital tract contributes to HIV pathogenicity and infectivity could aid the design of interventions aimed at effectively disrupting transmission of the virus. In this study, we investigated whether local mucosal immune activation is associated with HIV shedding and CD4 T-cell depletion in the female genital tract.
MATERIALS AND METHODS
Description of Study Participants
Thirty-three chronically HIV-1–infected women and 11 healthy uninfected women attending the Emphilisweni Clinic in Cape Town, South Africa, were enrolled in this study. Women who were menstruating at the time of sampling, were postmenopausal, or had undergone a hysterectomy were excluded from the study. Women who had symptoms or signs of STI were not sampled. The study was approved by the Research Ethics Committee of the University of Cape Town, South Africa, and informed written consent was obtained from all women before initiation of the study. All women with CD4 cell counts <200 cells/mm3 were referred to the nearest HIV treatment center. Of the 19 women for whom detailed clinical histories were available, 8 (42%) were receiving combination antiretroviral therapy.
Cervical Mucosal Sample Collection and Processing
Cervical mucosal mononuclear cells (MMCs) were collected using a Digene cervical sampler, as described elsewhere . Cervical samples that had visible red blood cell contamination were discarded. Cervical cells were isolated within 4 hours after collection by flushing the cytobrush bristles to dislodge all cervical-derived cells. The cell suspension was pelleted, and the supernatant fraction was stored at −80°C until analysis for HIV shedding. Pelleted cells were resuspended in 2 mL of 10% fetal calf serum (FCS) RPMI. Cervical MMCs were used immediately for flow cytometry.
Peripheral Blood Mononuclear Cell Isolation
Blood samples were collected using venipunture into sterile vacutainer tubes (Becton Dickenson). Peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll-Hypaque density gradient centrifugation with Leucosep tubes. All experiments were performed using fresh PBMCs.
Evaluation of Immune Activation by Flow Cytometry
PBMCs and cervical MMCs were incubated with LIVE/DEAD Fixable Violet Dead Cell Stain for 20 minutes at room temperature, then washed twice with 1% FCS phosphate-buffered saline (PBS). Staining was done in 96-well plates. Extracellular staining was performed for 20 minutes at room temperature with phenotypic markers: peridinin-chlorophyll protein-cyanine 5.5 (PerCP-Cy5.5)–labeled anti-CD4 (Becton-Dickinson), QDot605-labelled anti-CD8 (Invitrogen), allophycocyanin (APC)–labeled CCR5 (BD), phycoerythrin cyanine 7 (PE-Cy7)–labeled anti-CD38 (E-Biosciences), PE-labeled anti–HLA-DR (BD), and dump markers Pacific Blue–labeled anti-CD14 (BD) and -CD19 (Caltag). Cells were washed twice with 1% FCS PBS, centrifuged, fixed, and permeabilized using BD CytoFix/CytoPerm for 20 minutes at room temperature and washed twice with BD Permawash. Cells were then stained with Alexa Flour700–labeled anti-CD3 (BD) and fluorescein isothiocyanate–labeled anti-Ki67 (BD). Cells were finally washed with BD Permawash, centrifuged, and fixed with BD Cell Fix. Cell fluorescence was measured using an LSRII (BD Immunocytometry Systems), and compensation and analysis were performed using FlowJo (Tree Star). Fluorescence minus one was used to distinguish continuous populations. Gating strategy is shown in Supplemental Figure 1.
Determination of Viral Load in Cervical Supernatant and Plasma
Viral loads in cervical supernatants and plasma samples were determined using Nuclisens Easyq HIV 1, version 1.2. The detection limit of this assay was 50 copies/mL.
Measurement of Cervical Supernatant Cytokines
Cervical supernatant samples were centrifuged and stored at −80°C. Before cytokine measurements, cervical supernatants were filtered by centrifugation using 0.2 μm cellulose acetate filters (Sigma). Total protein concentrations were measured in all samples with use of the Bio-Rad Protein Assay. Interleukin 6 (IL-6) and interleukin 8 (IL-8) concentrations in cervical samples were determined using Milliplex Human Cytokine kits (Millipore). The sensitivity of these kits range from 0.2 to 0.7 pg/mL for the 2 cytokines measured. Data were collected using a Bio-Plex Suspension Array Reader (Bio-Rad Laboratories), and a 5 PL regression formula was used to calculate sample concentrations from the standard curves. Data were analyzed using BIO-plex manager software (version 4). Cytokine levels that were below the lower limit of detection of the assay were reported as the midpoint between the lowest concentrations measured for each cytokine and zero.
Mann–Whitney U tests were performed for independent sample comparisons, and Spearman’s rank tests were applied for correlations. Adjustment for multiple comparisons was performed using a step-down approach . All tests were 2-tailed with P < .05. The association between viral load and the activation markers were examined using a zero-inflated negative binomial regression to take into account the large numbers of zero viral loads and, therefore, skewed data for individual blood and cervical sample measurements. Plasma viral load was included as a predictor variable. All analyses were performed using Stata, version 10.0 (StataCorp).
A total of 44 women were included in this study, of whom 33 were HIV-infected and 11 uninfected. The median age of the women was 34 years (interquartile range [IQR], 30–42 years), and there was no statistically significant difference in the ages of HIV-infected and uninfected women (37 vs 32 years; P = .48). The median CD4 cell count in the HIV-infected women was 334 cells/mm3 (IQR, 238–488 cells/mm3), and the median log peripheral viral load in the HIV-infected women was 3.3 HIV RNA copies/ml (IQR, 1.7–4.5). Fewer than half of the women were receiving antiretroviral therapy.
Comparison Between Cervical and Blood T-Cell Activation
We compared the frequencies of T cells expressing markers of activation (CCR5, CD38, and HLA-DR) and proliferation (Ki67) in matched cervical and blood samples from HIV-infected and uninfected women. We assessed dual expression of T-cell activation markers together with the coreceptor CCR5, because activated CD4+ T cells expressing CCR5 would be target cells for HIV infection. In addition, we assessed dual expression of CD38 and HLA-DR, because these are considered to be highly activated cells. Irrespective of infection status, we found that a higher proportion of cervical CD4+ and CD8+ T cells expressed CCR5, compared with blood (Figure 1). After adjustment for multiple comparisons, this difference was statistically significant for CD4 T cells in both the HIV-infected (P < .05) and uninfected (P < .001) women. In uninfected women, significantly higher frequencies of cervical CD4+ T cells dually expressed CCR5/HLA-DR alone HLA-DR alone, compared with CD4+ T cells in blood samples (P < .05 for both after adjusting for multiple comparisons). Of interest, the extent of immune activation in blood significantly predicted activation by CD4 and CD8 T cells in the cervical compartment (HLA-DR and CD38 expression by CD4 and CD8 T cells; P < .001 for both markers and both T-cell subsets after adjusting for multiple comparisons).
Inflammatory Cytokines and Immune Activation
We measured the concentrations of IL-6 and IL-8, 2 inflammatory cytokines commonly associated with STIs, in genital secretions from both HIV-infected and uninfected women. HIV-infected women had a median IL-6 concentration of 11.4 pg/mL in cervical secretions, whereas uninfected women had a median concentration of 22.9 pg/mL (P = .7). HIV-infected women had a median IL-8 concentration of 183.7 pg/mL in genital secretions, whereas uninfected women had 120.2 pg/mL (P = .6).
We found no statistically significant relationship between the concentration of either of these cytokines and T-cell activation marker expression in the female genital tract, suggesting that T-cell activation in the genital tract was independent of local mucosal inflammation, but rather dependent on the level of activation in blood.
Impact of HIV Infection and Viremia on Immune Activation at the Cervix and in Blood
We compared the frequency of activation marker expression on T cells isolated from the cervix and blood samples from uninfected and HIV-infected women (Figure 1A–D). T cells from both compartments of HIV-infected women were generally more activated than those obtained from uninfected women (Figure 1). Of interest, CCR5+ receptor expression by CD4 T cells was similar in HIV-infected and uninfected women irrespective of compartment.
Because 42% of HIV-infected women in this study were receiving combination antiretroviral therapy and were aviremic, we compared the extent of T-cell activation in the blood and genital tract of HIV-infected women with and without detectable plasma viral loads (Figure 1A). Of the 28 HIV-infected women with plasma samples available for viral load testing, 18 (64%) had detectable HIV RNA in their plasma sample, with viral loads ranging from 2000 to 63 000 RNA copies/mL. There were no statistically significant differences in blood CD4 T-cell immune activation between viremic and aviremic women, suggesting that HIV viremia was not an important determinant of activation in these women. At the cervix, HIV-infected women with detectable plasma viral loads had significantly higher cervical CD4+ T cells expressing HLA-DR/CD38+ than aviremic HIV-infected women (Figure 1C), although no other differences in genital activation were observed for the other markers. In contrast to CD4 T cells, CD8 T cells in blood samples and at the cervix were significantly more activated in viremic than in aviremic women (HLA-DR, CD38, HLA-DR/CD38, HLA-DR/CCR5, and CD38/CCR5) (Figure 1B and D).
HIV-1 Shedding in the Female Genital Tract During HIV Infection
Viral loads were determined in cervical samples from 26 of these chronically HIV-infected women, 13 (50%) of whom had detectable HIV RNA in their cervical samples and were therefore classified as shedding virus. In these 13 women, the median cervical HIV load was 790 RNA copies/mL (range, 160–6300 RNA copies/mL). There was a significant correlation between plasma and cervical viral loads (Rho = 0.59; P = .002; data not shown), and women who were shedding HIV from the cervical mucosa (ie, those with detectable cervical HIV RNA) had significantly higher plasma viral loads than did women who were not detectably shedding virus from the cervix (P < .001; data not shown).
Immune Activation and Genital Tract HIV Shedding
We investigated the relationship between T-cell activation at the cervix and in blood samples from the HIV-infected women with HIV loads in each compartment (Figure 2). T-cell activation in blood was not associated with plasma viral loads (data not shown). At the cervix, however, we found a significant correlation between the expression of certain activation markers by CD4+ and CD8+ T cells (Ki67 and CCR5/Ki67) with HIV load in genital secretions, which remained significant even after adjusting for plasma viral loads and multiple comparisons (Figure 2). A β-coefficient of 0.052 for CD4+Ki67+ represents a 12.7% increase in cervical viral load for every unit increase in Ki67 expression by CD4+ T cells. Similarly, a β-coefficient of 0.089 for CD8+Ki67+ represents a 22.7% increase in cervical viral load for every unit increase in Ki67 expression by CD8+ T cells. In addition, expression of CD38 by cervical CD4 T cells was inversely related to cervical viral load, although this association did not exist after adjustment for multiple comparisons.
Immune Activation, Genital HIV Shedding, and CD4 T-Cell Depletion
We next investigated the impact of HIV infection and immune activation on genital tract CD4 T-cell depletion (Figure 3). We found that the proportion of CD3 T cells that were CD4+ was reduced in HIV-infected women, compared with uninfected women, in both the genital tract and blood samples (both P < .001) (Figure 3A) and that the extent of CD4 T-cell depletion in blood significantly predicted CD4 T-cell proportions at the cervix (Rho = 0.71; P < .001) (Figure 3B). Furthermore, the percentage of CD4 T cells at the cervix was inversely associated with the extent of cervical CD4 T-cell activation (HLA-DR, HLA-DR/CCR5, and HLA-DR/CD38) (Figure 3C), even after adjusting for cervical and peripheral viral loads. In contrast, we observed no correlation between CD4 T-cell depletion and CD4 T-cell activation at the cervix of uninfected women.
We calculated the association between genital viral load and CD4 T-cell depletion in the female genital tract. After adjusting for plasma viral load and T-cell activation marker expression, genital HIV shedding was independently predictive of CD4 T-cell depletion at that site. Together, these data suggest that CD4 T-cell activation in the genital tract during HIV infection is associated with both CD4 T-cell depletion in the genital mucosa and HIV shedding.
Systemic immune activation during HIV infection has been identified as an important factor driving elevated plasma viral load, accelerated CD4+ T-cell death in blood, and more rapid disease progression [7, 10]. In gut-associated mucosal tissue, a number of key studies have demonstrated that there is severe depletion in HIV-infected individuals of certain CD4 T-cell subsets, and this depletion is associated with immune activation and microbial translocation . We show that, in the female genital tract, T-cell activation is related to cervical CD4 T-cell depletion and elevated HIV shedding in genital secretions.
Few studies have investigated T-cell activation in the female genital tract during HIV infection. Elevated genital T-cell CCR5 expression, compared with blood, has been previously described in HIV-uninfected women . High frequencies of CD69 expression have been described in the genital tract of HIV-uninfected African women . In this study, we investigated the impact of HIV infection on immune activation in the genital tract and the relationship between genital T-cell activation, HIV shedding, and local CD4 T-cell depletion. In both HIV-infected and uninfected women, we found that certain activation markers were expressed at significantly higher frequencies on cervical T cells, compared with blood-derived T cells. In particular, we found significantly higher proportions of activated CD4+ T cells (HIV target cells) in the cervix than in blood samples from uninfected women. Although we found that, as in other studies, plasma viral loads were associated with HIV shedding at the genital tract, we showed that immune activation at the cervix independently predicted genital HIV shedding.
Although women with symptomatic STIs were not included in this study, various asymptomatic and clinically subtle infections may have been present. African women are known to have a high prevalence of bacterial vaginosis, candida infection , and human papillomavirus infection . Candidal or trichomonal infections are associated with elevated genital tract inflammation and cellular recruitment and may provide HIV with a source of activated targets [19, 20]. Chlamydial and gonorrheal infections in women are frequently asymptomatic. Many of these infections, however, are associated with high levels of inflammatory cytokines. including IL-6 and IL-8 [5,6]. In addition, it has also been shown that even STI-free HIV-uninfected Kenyan women had higher levels of cellular activation than their North American counterparts . Mitchell et al  found that cervical IL-8 concentrations predicted HIV shedding from the female genital tract independent of other genital infections.
Whatever the mechanism driving T-cell activation in this study, these findings suggest that highly activated T cells, which are both targets and reservoirs of HIV infection, are present at high frequencies at the cervix of both HIV-infected and uninfected women. In HIV-infected women, the presence of these highly activated T cells predicted HIV shedding from the genital tract. It is possible that cervical immune activation drives local viral replication in HIV-infected women and not vice versa, as is commonly believed. Of importance, although both local viral replication and immune activation were found to predict CD4 T-cell depletion in the genital tract, viral replication is a necessary cofactor for local CD4 T-cell loss, because CD4 T-cell depletion was not evident in the genital mucosa of HIV-uninfected women.
Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.
We thank all the women who kindly participated in the study and Sister Ntombizonke Makhonza for collecting the specimens.
This work was supported by the Wellcome Trust and South African HIV/AIDS Research Platform, Wellcome Trust Intermediate Fellowship in Infectious Diseases (to J.-A. P.); and National Institutes of Health Training Grant (5T32HD007233-28 to H. J.).
Potential conflicts of interest.
All authors: No reported conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.