Heterosexual transmission of HIV-1 accounts for the most new HIV infections worldwide.1 The cervicovaginal mucosa is the first site of encounter with the virus in most women, and HIV shed from this site may infect sexual partners or neonates. The endocervix is a major reservoir of T lymphocytes in the female genital tract and is also the primary site of antibody production.2-5 Hence, local immune dynamics at the level of the endocervical mucosa would be expected to play an important role in the control of local viral replication, and thus in the establishment of infection. Cytotoxic cell activity and antibody-dependent cell-mediated cytotoxicity have been detected in the cervix of HIV-seropositive women6-8 and in highly exposed uninfected sex workers in Africa.9,10 Whether the presence or the type and quality of HIV-specific immunity at these sites confer protection from transmitting or acquiring infection is not clearly established, however.
Primate and human studies have indicated that T-cell populations residing in the gastrointestinal (GI) mucosa are depleted more severely and rapidly than populations in the periphery; a recent study indicated a similar pattern in the simian genital mucosa.11-15 Although a limited number of studies have indicated that human cervical T cells exhibit some phenotypic markers characteristic of cells in the GI mucosa, the repertoire and function of this genital population, and its alterations in different stages of HIV infection and in response to treatment, are not well characterized. Further, little research has been done on the effect of HIV infection on genital tract humoral immunity at this site. Antibody profiles in peripheral blood and mucosal secretions are often abnormal in HIV-infected persons, presumably because of insufficient T-cell help.16,17 Limited studies have suggested dysregulation of local humoral immunity in the female lower genital tract during HIV infection.18 The extent and nature of perturbations in this mucosal immune repertoire could clearly have substantial ramifications for local viral control, spread of virus, susceptibility to opportunistic infections or sexually transmitted pathogens, and the extent to which highly active antiretroviral therapy (HAART) could reconstitute immunity at this site.
To evaluate the impact of HIV infection on the T-cell and antibody repertoire in the endocervix, we compared T-cell immunophenotypes and antibodies in the cervices of HIV-seronegative women, HIV-infected women with a low plasma viral load not taking antiretroviral therapy (well-controlled viremia), and HIV-infected women with a high plasma viral load (uncontrolled viremia) who had to be started on antiretroviral therapy. T cells were sampled using the cytobrush technique,4 and antibodies were quantified from cervical mucus collected by absorbent ophthalmic sponges.19 Both of these collection techniques have been described previously by our laboratories as being appropriate and noninvasive procedures for assessing local cervical immunity.4,19
MATERIALS AND METHODS
Study participants were HIV-seropositive and HIV-seronegative women who met the following enrollment criteria. The control group included 12 seronegative women at low risk for HIV infection (mean age = 43 years, range: 27-55 years); 10 of 12 women had regular menstrual periods, whereas the other 2 were perimenopausal/menopausal. The slow-progressing HIV-seropositive (HIV+ SP) group included 10 women with documented HIV infection for at least 10 years on no antiretroviral therapy who had CD4+ cell counts >400 cells/mm3 or CD4% >36% and a peripheral viral load <10,000 copies/mL (mean age = 40 years, range: 30-61 years); 8 of 10 had regular menstrual periods, whereas the other 2 were perimenopausal/menopausal. The HIV-seropositive failing (HIV+ F) group included 9 HIV-infected women with a CD4+ T-cell count <200 cells/mm3 and a peripheral viral load >3000 copies/mL in the past year, currently taking antiretroviral therapy (mean age = 38 years, range: 29-46 years); 7 of 9 had regular menstrual periods, whereas 1 was anovulatory and 1 was on Depo-Provera. The HIV+ SP group had a median CD4+ T-lymphocyte count of 640 cells/mm3 (range: 429-1094 cells/mm3) and a median viral load of 2033 copies/mL (range: 63-8111 copies/mL); the HIV+ F group had a median CD4+ T-lymphocyte count of 95 cells/mm3 (range: 46-201 cells/mm3) and a median viral load of 172,336 copies/mL (range: 3000-500,000 copies/mL). Institutional Review Board approval was obtained at the Brigham and Women's Hospital (Boston, MA) and Miriam Hospital (Providence, RI) and at the Centers for Disease Control and Prevention (CDC); all participants consented to the study.
Specimen Collection and Processing
Cervical secretions were collected by absorption onto 2 preweighed ophthalmic sponges; sponges were reweighed and sealed in 2-mL cryovials. Sponges were stored at −80°C until elution and analysis. After sponge sampling of secretions, a cytobrush was inserted into the cervical os and gently rotated a single 360° turn. The cytobrushes were cut from their wire and placed in RPMI medium supplemented with gentamicin and amphotericin. Blood was collected in ethylenediaminetetraacetic acid (EDTA) collection tubes. Cytobrush and blood samples were processed within 6 hours of collection.
Isolation and Flow Cytometric Analysis of T-Cell Populations
Cytobrushes were processed as described previously,4 and the isolated cell population was enumerated for red blood cells (RBCs), polymorphonuclear cells (PMNs), and total and viable leukocytes. Cells were stained for flow cytometric analysis with the use of a direct immunofluorescence technique in which an anti-CD3 antibody was always included. Antibodies used were in the following combinations: CD3-Cy/CD4-FITC/CD8PE, CD3-Cy/CD4-FITC/CD45RA-PE, CD3-Cy/CD4-FITC/HLA-DR-PE, and CD3-Cy/CD103-FITC (BD Pharmingen, San Jose, CA). T cells were identified and gated based on their CD3 expression and low side scatter characteristics.
T-Cell Receptor Repertoire Analysis
Cytobrush and blood samples from 2 women in the study (1 seronegative and 1 failing) were used for T-cell receptor (TCR) repertoire analyses. Cervical cells and peripheral blood mononuclear cells (PBMCs) were isolated as described previously, and CD4+ T cells were positively selected using specific-antibody-treated immunomagnetic beads, as described previously (Dynabeads; Invitrogen, Carlsbad, CA).20 Equal numbers of PBMCs and cervical CD4+ T cells were used for RNA extraction, depending on the total number of cells retrieved from the cytobrush. A second tube of 1 × 106 CD4+ T cells from the seronegative woman provided a control for sampling small numbers of T cells. The immunoscope technique, a polymerase chain reaction (PCR)-based method that determines variable, diversity, and joining (VDJ) junction size patterns in 24 human TCR Vβ subfamilies, was then used to analyze these paired blood and cervical cell populations.21,22 TCR β transcripts were reverse transcribed and amplified using Vβ- and Cβ-specific primers, and a fluorophore-labeled Cβ-specific primer was used to visualize amplified products; primer sequences and methodology were as described elsewhere by Puisieux et al.22
Measurement of Total Immunoglobulin
Total serum and cervical mucus immunoglobulin concentrations were determined by quantitative enzyme-linked immunosorbent assay (ELISA), as described previously.19 In brief, microtiter plates were coated overnight at 4°C with affinity-purified F(ab')2 fragments of goat antibodies specific for human IgG and IgA (Jackson Immunoresearch, West Grove, PA) diluted in phosphate-buffered saline (PBS). Total immunoglobulin assays were standardized using an immunoglobulin calibrator (The Binding Site, San Diego, CA) of known IgA and IgG concentrations. The extracted samples and the calibrator were diluted in PBS containing 1% fetal calf serum and added to the plates in duplicate. After an overnight incubation at 4°C, the plates were washed with PBS with 0.05% Tween 20 (PBS-T); isotype-specific biotinylated affinity-purified F(ab')2 fragments of goat antihuman antibodies (Biosource International, Camarillo, CA) were then added and incubated for 2 hours at room temperature. The plates were then washed with PBS-T and incubated for 30 minutes with horseradish-peroxidase-conjugated avidin (Sigma, St. Louis, MO) diluted to 0.5 μg/mL in 0.87% saline containing 0.05% Tween 20. The plates were washed and exposed to substrate consisting of 2,2'-azino-bis (3-ethylbenzthiazoline-6-sulfonic acid [ABTS]; Sigma) at a concentration of 0.25% (wt/vol) in 0.1 M of citrate buffer, pH 4.2, containing 0.0075% H2O2 (ABTS Substrate). Substrate turnover was monitored with spectrophotometry at 415 nm using an automated reader (BioTek, Winooski, VT). A standard curve was produced using Delta Soft ELISA analysis software (Biotek Instruments, Winooski, VT).
Measurement of HIV-Specific Immunoglobulin
Serum and cervical mucus HIV-1-specific IgG concentrations were measured by quantitative ELISA using Nunc Maxisorp plates (Fisher Scientific, Pittsburgh, PA) coated with 100 μL of HIV-1IIIB lysate (Advanced Biotechnologies, Columbia, MD). HIV-specific IgG was quantitated using an HIV-specific IgG standard prepared from breast milk from an HIV-1-infected mother and used between 1 and 1000 ng/mL. Serum and cervical mucus were added to the plates at serial dilutions starting at 1:1000 for serum and 1:5 for cervical mucus. The plates were then developed with biotinylated goat polyclonal antibodies to human IgG, followed by avidin-labeled peroxidase and ABTS. To account for the large variation in total IgG levels among individual people, results were expressed as specific activity, which was calculated by dividing the anti-HIV antibody concentration by the total antibody concentration (μg/mL). A positive sample was defined as an HIV-1 antibody concentration >2 SDs above the mean concentration measured from the group of HIV-1-seronegative women.
None of the variables satisfied the conditions of normal distribution or homogeneity of variance. Therefore, nonparametric statistics were used for the data analysis. Specifically, for continuous variables, Mann-Whitney U tests were used for comparisons of 2 independent groups and Wilcoxon signed-rank tests were used for 2 dependent comparisons. Kruskal-Wallis nonparametric analysis of variance with the Dunn multiple comparison test was used for comparisons of more than 2 independent groups. The Spearman rank correlation coefficient was used for correlation of continuous variables. Finally, the Fisher exact probability test and Freeman-Halton test were used for analysis of categoric data from independent samples. In all cases, statistical significance was assumed when P < 0.05. Data analysis was conducted using StatView (version 5.0.1; SAS Institute, Cary, NC), Prism (version 2.0b, GraphPad Software, San Diego, CA), and StatXact (version 4; Cytel Software Corporation, Cambridge, MA) statistical software.
Peripheral Blood and Cervical T-Cell Populations
Samples for the cytobrush study were available from 11 seronegative women, 5 HIV+ SP patients, and 6 HIV+ F patients. (Not all women participated in all analyses because of logistical considerations, such as delays in specimen transport that rendered several cytobrush specimens unusable.) In a previous study, we examined the issue of blood contamination (natural or collection induced) in healthy seronegative women and found a minimal impact of blood on cervical immune parameters.4 By using markers seen predominantly in blood (CD45RA) or mucosa (CD103) to assess the impact of peripheral blood contamination of samples collected on cytobrushes for this study, we found no indication of significant blood contamination (data not shown). In fact, in the HIV-infected group, a significant negative correlation between RBCs and PMNs was observed (r = −0.62, P = 0.05), indicating a minimal impact of blood, at least in this population.
A smaller number of leukocytes (total and viable) were retrieved from the cytobrushes of HIV-positive women, particularly the patients with uncontrolled viremia (Table 1). This decrease was statistically significant in the women whose treatment was failing as compared with the control women (P < 0.01). Phenotypic differences between paired peripheral blood and cervical cytobrush T cells in control, HIV+ SP, and HIV+ F women were seen (see Table 1). In HIV-seronegative women, the proportion of CD4+ to CD8+ T cells was slightly higher in the cervix compared with the blood; this is in contrast to other mucosal sites, where CD8+ T cells usually predominate. However, HIV-infected women, especially those women with more advanced disease (failing treatment), had a low frequency of CD4+ T cells in peripheral blood and in the cervix. They had corresponding increases in the proportion of CD8+ T cells.
Highly elevated proportions of activated (positive for human leukocyte antigen-DR [HLA-DR+]) T cells were found in the cervix of the group of seronegative women as compared with their blood (P < 0.01), and a significant difference was also found in the group of slow-progressing HIV-seropositive women (P = 0.04; see Table 1). In contrast, activated CD4+ T cells seemed to be more common in the blood than in the cervix of patients with uncontrolled viremia, although the differences were not significant. Similar to other tissue sites, the cervix contained rare naive cells (CD45RA+) as compared with the peripheral blood in all groups of women (see Table 1). Approximately 20% to 30% of cervical CD3 T cells expressed CD103; as expected, this marker was not expressed on peripheral blood T cells (see Table 1). No significant differences were found among groups of patients in the proportion of CD3 cells expressing CD103.
T-Cell Receptor Use in Peripheral Blood and Endocervix
We next investigated the pattern of use of TCRs in the endocervical CD4+ T-cell population. Highly restricted utilization of Vβ families and limited clonality within these families have been reported at other mucosal sites, but the pattern of use has never been investigated in the genital tract. We used the immunoscope technique21,22 on the CD4+ T-cell population isolated from the blood and the cervix using specific antibody-treated immunogenetic beads (1.2 × 104 for the seronegative woman and 0.8 × 104 viable CD+ T cells for the failing woman). We analyzed samples from 1 of the seronegative women and found utilization of all Vβ families in the peripheral blood and no evidence of oligoclonal expansions; this finding was in contrast to the extremely limited utilization of Vβ families in the cervix, with only 4 families represented (Table 2). Furthermore, the restricted peak numbers seen in these families utilized in the cervix suggested a further restriction of available TCRs (data not shown). We next examined paired blood and endocervical T cells from 1 of the failing (HIV+ F) women. Although all TCR Vβ families were utilized in the blood, we again found highly restricted utilization of Vβ families in the cervix, with only 3 families represented. Similar to the seronegative woman, all families in the cervix were highly oligoclonal.
Peripheral Blood and Endocervical Antibody Profiles
Before endocervical immunoglobulin analyses were performed, cervical mucus specimens were analyzed for heme content to determine whether there were differences between the HIV-negative and HIV-positive women and whether blood, present naturally or induced by sampling, influenced cervical antibody measurements. No significant differences were observed in heme content among the 3 groups (P = 0.88). Further, an analysis of heme and IgG antibody ratios in cervical mucus specimens showed no relation (P = 0.80). If blood contamination of cervical specimens had been an issue, a direct relation between heme and IgG would have been observed.
As in other studies, we found that IgG predominated over IgA in the normal endocervix (Table 3), which is a phenomenon that is unusual for a mucosal site. As others have reported, we also found that total serum IgG levels were elevated in HIV-positive women, a difference that was highly significant between the HIV-positive women whose viremia was uncontrolled and the seronegative group (P < 0.01). Similar to serum, the endocervix contained elevated IgG levels in HIV-infected women with uncontrolled viremia as compared with uninfected women (P < 0.01). No significant differences were found among groups with respect to total IgA concentrations in serum or the endocervix.
Analysis of HIV-specific IgG in the serum and cervix (Table 4) revealed HIV-specific IgG in the serum of all the HIV-infected participants, regardless of clinical status. In this analysis, a positive value for HIV-specific IgG was defined by a cutoff value of 2 SDs greater than the mean of the seronegative group. Seven of 10 slow progressors and 6 of 8 women with uncontrolled viremia demonstrated HIV-specific IgG in their endocervical mucus. Serum HIV-specific IgG concentrations were higher in slow progressors than in patients with uncontrolled viremia, although this difference did not reach statistical significance (P = 0.06). When adjusting for the total amount of immunoglobulin collected at each site, we found that the slow-progressing group of HIV-positive women had greater IgG HIV-specific activity in the serum than did the women with uncontrolled viremia (P = 0.01). This group difference was not observed in the endocervix. A significant positive correlation was found between the peripheral CD4+ T-cell count and serum IgG HIV-specific activity (r = 0.53, P = 0.03); however, there was no significant correlation between the peripheral CD4 cell count and cervical IgG HIV-specific activity. Hence, the peripheral CD4 cell count may most accurately predict IgG HIV-specific activity in the serum, although not necessarily at a local site. We did not find significant correlations between cervical CD4 cell counts and cervical IgG HIV-specific activity (data not shown).
We observed a severe depletion of leukocytes and CD4+ T lymphocytes in the endocervix of HIV-infected women as compared with uninfected women, and the severity of this depletion correlated with disease stage. CD4+ T cells expressing HLA-DR were found in significantly higher numbers in the endocervix of the seronegative and slow-progressing women compared with their blood. This was not observed among women failing antiretroviral therapy, where no significant difference was seen. This may be attributable to generalized immune activation in the periphery because of a high viral load, or preferential killing of activated cells in the cervical mucosal environment. Interestingly, CD4+ T cells with a naive phenotype were a rarity in the cervix of all groups of women, suggesting that HIV infection at any stage did not have a significant impact on the trafficking of this cell subset. CD103 expression was high in the cervix, consistent with its mucosal nature. We did not find changes in CD103 expression in HIV-infected patients, unlike Schneider et al,23 who found decreased CD103 expression on intestinal CD8 T cells, although not on CD4 cells, of HIV-infected patients. The fact that our data were based on all CD3 cells may explain this difference. Vβ family utilization in the cervical CD4+ T cells of the seronegative and HIV+ F patients we examined seemed to be highly restricted. This finding is similar to results in the GI tract.24 Oligoclonality was evident within the represented Vβ families, suggesting further restriction of available TCRs and, possibly, clonal expansion of specific populations. Ideally, these studies should be expanded to look at greater numbers of patients; it would also be of particular interest in HIV-infected patients who are controlling virus and in longitudinally studied patients who are placed on HAART to determine if the cervical repertoire does have the capacity to expand in response to infection, or to reconstitute, respectively.
Our findings confirm previous studies that IgG predominates over IgA in the cervix.5 Even though HIV-infected women had higher levels of cervical and blood IgG when compared with HIV-negative women, no such differences were observed for IgA; this finding concurs with the report by Artenstein et al.25 HIV-specific IgG was detected in most of the cervical specimens of HIV-infected women.
Mucosal T lymphocytes constitute the largest population of T cells in the body. T cells derived from the GI tract are particularly well characterized and express unique phenotypic markers, a restricted TCR repertoire, and a functional capacity that reflects their location.24 Studies in HIV-infected persons11 and simian immunodeficiency virus (SIV)-infected macaques14,15 have indicated a profound loss of CD4+ cells in the GI tract early in infection; this loss was more pronounced and occurred earlier than the respective loss in peripheral blood. This selective depletion also continued to occur at all stages of HIV disease.13 The loss of CD4+ cells seems to occur because HIV replicates most efficiently in activated CD4+ CCR5+ cells, a population that is more dominant in the gut than in the blood or lymph nodes; this population is killed massively by HIV early in infection.14,15 Much less is known about the effects of HIV on the human genital tract, although a recent study in macaques indicated that genital T cells have a similar profile to intestinal T cells and are also rapidly depleted early in SIV infection.12 Our preliminary studies indicate depletion of CD4+ T cells at all stages of infection, quite probably because of the high proportion of activated memory cells.
It has been reported that T cells in the human female genital mucosa express a mucosal-like profile,26 but the effector function of this population is not well characterized. Further, little research has been done on the effects of HIV infection on humoral mucosal immunity at this site. An understanding of local immunity in the female genital tract is extremely important, because vaccines must be designed to induce appropriate protective immunity at this most common site of initial infection. In addition, it is important to know whether HAART could reconstitute mucosal tissues with CD4+ helper T cells, because this could possibly lead to normalization of antibody concentrations and specificity profiles in mucosal secretions.
In conclusion, despite the small number of patients, our study demonstrates a distinct pattern of T-cell subpopulations and antibody concentrations in the endocervix of HIV-infected women, which is characterized by CD4+ T-cell depletion and increased levels of cervical IgG. Further studies conducted on a larger number of women focusing on the function and antigen specificities of the lymphocyte populations and antibodies of the genital mucosa, including studies of immune reconstitution in women with control of viral replication on HAART, should shed more light on the role of this site in sexual acquisition or transmission of HIV.
Support for this subproject (CSA-99-242 to AJQ) was provided by the CONRAD Program, Eastern Virginia Medical School, under a Co-operative Agreement with the United States Agency for International Development (USAID) (HRN-A-00-98-00020-00), which in turn receives funds for AIDS research from an interagency agreement with the Division of Reproductive Health, Centers for Disease Control and Prevention (CDC). The views expressed by the authors do not necessarily reflect the views of USAID, CDC, or CONRAD.
1. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report
. Atlanta, GA; Centers for Disease Control and Prevention; 2003.
2. Pudney J, Quayle AJ, Anderson DJ. Immunological microenvironments in the human vagina and cervix: mediators of cellular immunity are concentrated in the cervical transformation zone. Biol Reprod
3. Cohen MS, Anderson DJ. Genitourinary mucosal defenses. In: Holmes KK, Sparling PF, March PA, et al, eds. McGraw-Hill Professional, New York, NY. Sexually Transmitted Diseases
4. Quayle AJ, Shah M, Cu-Uvin S, et al. Implications of blood contamination for assessment of local cellular immunity in the endocervix
. AIDS Res Hum Retroviruses
5. Kutteh WH, Mestecky J. Secretory immunity in the female reproductive tract. AM J Reprod Immunol
6. Musey L, Hu Y, Eckert L, et al. HIV
-1 induces cytotoxic T lymphocytes in the cervix of infected women. J Exp Med
7. Nag P, Kim J, Sapiega V, et al. Women with cervicovaginal antibody
-dependent cell-mediated cytotoxicity have lower genital HIV
-1 RNA loads. J Infect Dis
8. Shacklett BL, Cu-Uvin S, Beadle TJ, et al. Quantification of HIV
-1-specific T-cell responses at the mucosal cervicovaginal surface. AIDS
9. Kaul R, Plummer FA, Kimani J, et al. HIV
-1-specific mucosal CD8+ lymphocyte responses in the cervix of HIV
-1-resistant prostitutes in Nairobi. J Immunol
10. Belec L, Ghys PD, Hocini H, et al. Cervicovaginal secretory antibodies to human immunodeficiency virus type 1 (HIV
-1) that block viral transcytosis through tight epithelial barriers in highly exposed HIV
-1-seronegative African women. J Infect Dis
11. Lim SG, Condez A, Lee CA, et al. Loss of mucosal CD4 lymphocytes is an early feature of HIV
infection. Clin Exp Immunol
12. Veazey RS, Marx PA, Lackner AA. Vaginal CD4+ T cells express high levels of CCR5 and are rapidly depleted in simian immunodeficiency virus infection. J Infect Dis
13. Brenchley JM, Schacker TW, Ruff LE, et al. CD4+ T cell depletion during all stages of HIV
disease occurs predominantly in the gastrointestinal tract. J Exp Med
14. Veazey RS, DeMaria M, Chalifoux LV, et al. Gastrointestinal tract as a major site of CD4+ T cell depletion and viral replication in SIV infection. Science
15. Veazey RS, Lackner AA. HIV
swiftly guts the immune system. Nat Med
16. Shirai A, Cosentino M, Leitman-Klinman S, et al. HIV
infection induces both polyclonal and virus-specific B cell activation. J Clin Invest
17. Muller F, Froland SS, Hvatum M, et al. Both IgA subclasses are reduced in parotid saliva from patients with AIDS. Clin Exp Immunol
18. Ahmed SM, Al Doujaily H, Johnson MA, et al. Immunity in the female lower genital tract and the impact of HIV
infection. Scand J Immunol
19. Crowley-Nowick PA, Bell MC, Brockwell R, et al. Rectal immunization for induction of specific antibody
in the genital tract of women. J Clin Immunol
20. Quayle AJ, Xu C, Mayer KH, et al. T lymphocytes and macrophages, but not motile spermatozoa, are a significant source of human immunodeficiency virus in semen. J Infect Dis
21. Ria F, van den Elzen P, Madakamutil LT, et al. Molecular characterization of the T cell repertoire
using immunoscope analysis and its possible implementation in clinical practice. Curr Mol Med
22. Puisieux I, Even J, Pannetier C, et al. Oligoclonality of tumor-infiltrating lymphocytes from human melanomas. J Immunol
23. Schneider T, Ullrich R, Bergs C, et al. Abnormalities in subset distribution, activation, and differentiation of T cells isolated from large intestine biopsies in HIV
infection. The Berlin Diarrhoea/Wasting Syndrome Study Group. Clin Exp Immunol
24. Blumberg RS, Yockey CE, Gross GG, et al. Human intestinal intraepithelial lymphocytes are derived from a limited number of T cell clones that utilize multiple V beta T cell receptor genes. J Immunol
25. Artenstein AW, VanCott TC, Sitz KV, et al. Mucosal immune responses in four distinct compartments of women infected with human immunodeficiency virus type 1: a comparison by site and correlation with clinical information. J Infect Dis
26. Prakash M, Patterson S, Gotch F, et al. Recruitment of CD4 T lymphocytes and macrophages into the cervical epithelium of women after coitus. Am J Obstet Gynecol