Immune Responses to HIV in the Female Reproductive Tract, Immunologic Parallels with the Gastrointestinal Tract, and Research Implications

Abstract

The female reproductive tract is a major site of mucosa-associated lymphoid tissue and susceptibility to HIV infection, yet the tissue site(s) of infection and the impact of HIV infection on this important mucosal tissue remain poorly understood. CD4+ T-cells and other cell types expressing the major coreceptors for HIV, CCR5 and CXCR4, are abundant in both the lower reproductive tract (endocervix and vagina) and the upper tract (endocervix and uterus), and are highly susceptible to infection. Antiviral defenses in the female reproductive tract are mediated by a variety of soluble factors, and by mucosal effector cells that differ phenotypically from their counterparts in blood. The immunologic characteristics of the female reproductive tract parallel those of the gut, where major HIV-related immunologic injury occurs. The susceptibility of the female reproductive tract to HIV infection and immunopathogenesis suggests important new avenues for further research.

Introduction

Mucosal surfaces are the primary sites of most HIV transmission, and thus these tissues are a focus of attention for efforts to prevent HIV infection¹. Several mechanisms have been proposed to explain how HIV crosses the epithelial barriers of the female genital tract, including direct passage of virus and/or infected cells through epithelial breaches, transcytosis of free virus across polarized epithelial cells, and binding and/or uptake by mucosal dendritic cells^{2, 3}. Regardless of the mechanism(s) leading to infection via the reproductive mucosa, the rate of HIV transmission per high-risk sexual contact appears to be low⁴. However, once infection is initiated, virus amplification occurs rapidly, aided by local inflammation⁵. Systemic dissemination occurs prior to the development of the adaptive immune response⁶, leading to massive depletion of CD4+ T-cells, notably those lining the gastrointestinal mucosa^7–10. Like the gastrointestinal tract, tissues of the lower (ectocervix and vagina) and upper (endocervix and uterus) female reproductive tract contain partially activated, CD4+ memory cells expressing CCR5 and/or CXCR4, which can serve as targets for HIV. However, the effects of acute/early HIV infection on the reproductive tract have not been the focus of intensive study, while much attention has been devoted to the early effects of HIV infection on the gastrointestinal tract. This review summarizes literature on immune responses to HIV/SIV in the female reproductive tract, highlighting immunologic parallels between the reproductive and gastrointestinal mucosal tissues.

HIV infection of the female reproductive tract: local inflammation leads to rapid viral dissemination

Detailed studies of acute simian immunodeficiency virus (SIV) infection in rhesus macaques have revealed the mechanisms that lead to the rapid expansion of HIV/SIV in tissues, beginning with a small “founder” population of cells near the initial site of entry in the endocervix. Li and colleagues employed a combination of in situ hybridization and immunohistochemistry to elucidate the cells and soluble factors participating in the earliest steps of SIV dissemination following cervicovaginal inoculation of macaques⁵. At just one day post-infection, they found an accumulation in cervical mucosa of CD123+ plasmacytoid dendritic cells (pDC) expressing interferons α and β, and producing the chemokines MIP-1α and MIP-1β. These chemokines, in turn, attract CD4+ T-cells expressing the chemokine receptor CCR5, which serve as targets for SIV infection. This “inside-out” signaling, involving innate and inflammatory responses in mucosal tissues, leads to rapid amplification and expansion of the infection⁵.

If an inflammatory cascade leads to rapid expansion of HIV/SIV infection in mucosal tissues, might the inhibition of local inflammatory responses prevent viral dissemination following HIV/SIV mucosal exposure? To address this question, Li and colleagues selected an antimicrobial compound, glycerol monolaurate (GML), which inhibits the production of proinflammatory cytokines and chemokines by vaginal epithelial cell cultures in vitro. Strikingly, rhesus macaques treated with 5% GML prior to a high-dose intravaginal challenge with SIV were completely protected from systemic viral infection⁵. These findings demonstrate that innate immune defenses within mucosal tissues are, paradoxically, critical to the rapid dissemination of HIV/SIV within the host. They also suggest that microbicides designed to limit the host’s inflammatory response, rather than amplifying this response, may prove to be surprisingly effective at reducing HIV transmission. Furthermore, these findings imply that individuals who remain uninfected despite repeated mucosal exposure to HIV may respond to exposure by “immune quiescence” rather than an active inflammatory response. Caution is warranted in the study of topical microbicides, as some inflammatory modulators have an additional, unique role in the female reproductive tract by which they influence the outcome of embryonic implantation and early growth¹¹.

Innate immune defenses in the female reproductive tract

The female reproductive tract is equipped with a variety of physical and chemical defenses against viral infection [recently reviewed by Iwasaki¹²]. The vagina and ectocervix are lined with multilayered squamous epithelium. In contrast, the upper reproductive tract, including the endocervix, uterine endometrium and fallopian tubes, contains a less resilient monolayer of columnar epithelium. The endocervical canal is lined with mucus, which provides an additional defensive layer by trapping bacteria and viruses. Epithelial cells lining the reproductive tract produce a wide range of soluble factors with documented anti-HIV activity; these include secretory leukocyte protease inhibitor (SLPI)¹³, lactoferrin¹⁴, defensins^{15, 16}, cytokines and chemokines (e.g., RANTES and MIP-1β)¹⁷.

The serine protease inhibitor Trappin-2/Elafin has antibacterial activity against Gram-positive and -negative bacteria, as well as fungal pathogens¹⁸. Intriguingly, a recent proteomics study identified this protein as a potential biomarker of protection against HIV acquisition¹⁹. This study evaluated CVL samples from 315 women, including highly exposed, persistently seronegative (HEPS) commercial sex workers, HIV-positive women, and seronegative controls, using SELDI-TOF mass spectrometry. Trappin-2/Elafin was found to be significantly overexpressed in HEPS women as compared to both HIV-positive and seronegative groups.

A related study found that Trappin-2 was constitutively produced by primary cells from multiple reproductive tract sites, including uterus, Fallopian tube, cervix and ectocervix¹⁸. Levels of secreted Trappin-2/Elafin were higher in CVL samples from HIV-negative women than in samples from HIV-positive women, although these differences did not reach statistical significance. Recombinant Trappin-2/Elafin inhibited the infectivity of CXCR4 and CCR5-utilizing HIV strains in a TZM-bl assay¹⁸.

Other innate immune factors present in genital fluids include alpha and beta defensins and the antimicrobial peptide cathelicidin (LL-37). One recent study assessed levels of these factors in cervicovaginal secretions from HEPS commercial sex workers²⁰. The presence of multiple sexually transmitted infections (including C. trachomatis, N. gonorrheae, T. vaginalis and candidiasis) was associated with increased levels of human neutrophil peptide-1 (HNP-1), human β-defensin-2-3, and LL-37 in secretions. Levels of HNP-1 and LL-37 were strongly associated with in vitro HIV-neutralizing capacity of cervicovaginal secretions. However, in univariate analysis, the levels of these factors in secretions tended to be increased in high-risk women who eventually seroconverted, relative to those who remained uninfected²⁰.

HIV-specific IgA in cervical secretions

The presence of mucosal IgA at the sites of potential HIV exposure in the female reproductive tract could theoretically provide an important barrier against infection. IgG is also abundant in the reproductive tract, where it may be produced in the vagina or transported from serum by paracellular diffusion¹². Intriguingly, several studies have reported the detection of HIV-specific immunoglobulin A (IgA) in cervicovaginal secretions from highly-exposed, persistently seronegative (HEPS) women²¹. It has also been reported that such antibodies are capable of neutralizing primary strains of HIV in vitro²², and can block transcytosis of HIV across an intact epithelial monolayer²³. However, other studies of HEPS cohorts have failed to detect such antibodies, and this topic remains controversial in the HIV field^{24, 25}. Intriguingly, Tudor and colleagues recently described the construction of an antibody expression library derived from B-cells isolated from the cervical mucosa of HEPS women. Antibodies derived from this library included IgA antibodies specific for the membrane-proximal region of gp41²⁶.

HIV-specific T-cells, inflammation, and CD4 depletion in the lower female reproductive tract

Very few studies have characterized HIV-specific T-cell responses in the female reproductive tract of chronically infected individuals, likely due to the challenges associated with obtaining fresh tissues for analysis. Accordingly, it has been difficult to obtain a clear picture of the T-cell repertoire in the female reproductive tract, and the extent to which CD4+ and CD8+ T-cell populations mirror those found elsewhere in the body. Musey and colleagues compared the T cell receptor clonality of HIV-specific CD8+ T-cells from several mucosal sites, and found that the majority of clones were shared between mucosal tissues and PBMC²⁷. They also identified several clones with CTL function that were CD4+, MHC class II-restricted, and responded to peptides within the HIV Env protein²⁸.

Recently, progress has been made in utilizing minimally invasive clinical specimens, notably cervical cytobrushes and cervicovaginal lavage (CVL) for immunological assays^29–36. Gumbi and colleagues studied the relationship between mucosal inflammation and HIV-specific CD8+ T-cell responses in the cervix³². They found that increased levels of proinflammatory cytokines (i.e., TNF-α, IL-1β, IL-6 and IL-8) in cervicovaginal lavage fluid were significantly associated with HIV shedding in the genital tract. Plasma viral load was also positively associated with viral shedding from the cervix. However, there was no association between the magnitude of HIV Gag-specific CD8+ T-cell responses in the cervix and viral shedding. There was also no correlation between HIV-specific T-cell response magnitudes at the cervix and in blood. These findings suggest that (1) genital inflammation may promote HIV replication and viral shedding, and (2) HIV-specific T-cell responses in blood do not necessarily predict those in cervical mucosa³².

In a related study, Bebell and colleagues collected cervicovaginal lavage and plasma samples from women in a longitudinal cohort in Durban, South Africa, including 44 women with acute HIV infection²⁹. Levels of IL-6, IL-10 and IL-12p70 in CVL were significantly greater in women with acute HIV infection than in pre-seroconversion samples from the same women, or in CVL from HIV-negative controls. Importantly, cytokine levels in the genital tract were similar in women with acute HIV infection regardless of concurrent STI or bacterial vaginosis. Furthermore, elevated levels of IL-1β, IL-6 and IL-8 correlated with lower systemic CD4+ T-cell counts during acute HIV infection. Unfortunately, mucosal CD4+ T-cell depletion was not measured; nevertheless, these findings reveal that genital inflammation is associated with acute HIV infection and is significantly associated with systemic CD4+ T-cell depletion²⁹.

Nkwanyana and colleagues evaluated the composition and yield of cytobrush-derived T-cells from HIV infected women and uninfected controls³⁴. In the HIV-positive women, yields of CD3+ T-cells as well as CD19+ B-cells were significantly greater than in uninfected controls, although CD4+ T-cells were depleted relative to CD8+ T-cells. The majority of cytobrush T-cells had an effector memory (CD45RA-CCR7-CD27-) phenotype. There was a significant positive correlation between cytobrush T-cell counts and cervical concentrations of IL-1β, TNFα and IL-12.

The low number of viable lymphocytes isolated from cytobrush specimens limits the range of experiments that may be performed to characterize HIV-specific T-cell responses at the cervical mucosa. To address this issue, Bere and colleagues compared several methods for in vitro expansion of cytobrush-derived T-cells^{30, 31}. Expansion with anti-CD3/28 beads and IL-2, IL-7 and IL-15 gave rise to the greatest expansion after 7 days, with an accumulation of central memory T-cells³¹. The same authors used polyclonally expanded T-cell lines from 16 HIV-positive women to assess the relationship between HIV Gag-specific responses in blood and cervix in Elispot assays³⁰. Both the total response magnitude and the peptide pools targeted were similar in both sites, and cervical responses to HIV Gag peptides were detected only in women with systemic Gag-specific responses >1000 spot-forming units/10⁶ cells³⁰. This finding underscores the limitations inherent in cytobrush sampling, which typically yields between 0.01–1 × 10⁶ viable cells per cytobrush^{30, 32}.

Although highly active antiretroviral therapy (HAART) is associated with control of viremia and variable restoration of CD4+ T-cell counts in the periphery, little is known of the response to HAART in genital mucosa. In a recent cross-sectional study of 62 HIV-positive women, including 35 on HAART, Mkhize and colleagues demonstrated that HAART is associated with higher CD4+ T-cell percentages at the cervical mucosa³³. Genital tract HIV shedding was significantly associated with the amount of HIV RNA circulating in blood, and was fully suppressed in HAART-compliant women. As previously reported, HAART was also associated with decreased HIV-specific CD8+ T-cell responses in blood. Surprisingly, however, Gag-specific T-cell responses in cervical mucosa were preserved in women on HAART. These findings may suggest that low levels of virus continue to replicate in cervical mucosa, even when virus reaches very low to undetectable levels in plasma and cervical secretions. Further studies will be required to address this issue.

Sexually transmitted infections alter the mucosal microenvironment

Other sexually transmitted infections are frequently present when HIV exposure occurs, and these infections may alter the mucosal immune environment in ways that facilitate HIV transmission [reviewed by Kaul et al³⁷]. For example, infection with Chlamydia trachomatis, a pathogen of the columnar epithelium, results in increases in the number of CD4+ and CD8+ T cells, neutrophils, CD14+ monocytes, and CD83+ dendritic cells^{38, 39}. Similarly, increased concentrations of a wide range of cytokines are found in genital tract tissues and secretions during infection with C. trachomatis, including INFγ, TNFα, IL-1α, IL-6, IL-8, and IL-10, contributing to an escalating inflammatory response. In contrast, human papillomavirus infections do not elicit an intense inflammatory response, but intraepithelial lesions produced by these viruses may impair barrier function and increase HIV access to susceptible cell types⁴⁰. Less invasive infections, such as bacterial vaginosis, also produce immunologic responses that may facilitate HIV transmission. St. John and colleagues compared the effects of vaginal fluid from women with bacterial vaginosis or healthy controls on monocyte-derived dendritic cells. Fluids from women with bacterial vaginosis, but not controls, induced production of IL-12, IL-23, increased expression of dendritic cell activation markers (HLA-DR, CD40 and CD83) and stimulated proliferation of T-cells in allogeneic mixed lymphocyte reactions⁴¹. Chancroid is one of the STDs most closely associated with HIV transmission, perhaps due to extensive disruption of the epithelial barrier, bleeding and recruitment of susceptible cells⁴². In addition, Humphreys reported that experimental infection with Haemophilus ducreyi resulted in increased expression of HIV coreceptors by macrophages and CD4+ T-cells cells in chancroid lesions⁴³. Genital infections with herpes simplex viruses (HSV) lead to increases in cervical dendritic cells and CCR5-expressing CD4+ T-cells⁴⁴. Additionally, some HSV proteins can transactivate HIV gene expression, while others can serve as immune modulators⁴⁵.

Exposure to semen induces inflammatory responses in the female genital tract

While it is vital to elucidate the immunologic characteristics of the intact female genital tract, HIV exposure often occurs in the context of semen, thus the effects of semen on these immunologic functions must also be understood. In addition to perturbations related to genital tract infections, semen and seminal plasma also alter the immunologic milieu of the female genital tract in a manner that facilitates conception and pregnancy¹¹. The proinflammatory effects of insemination are well documented and referred to as the “post-mating inflammatory response” that is characterized by an increase in the synthesis of cytokines including GM-CSF, MCP-1 and IL-6 and the migration of macrophages, dendritic cells and granulocytes into the endometrial stroma⁴⁶. Leukocytes also migrate into the cervical epithelium, an occurrence that requires direct contact between seminal plasma and female genital tissues, and is related to increases in GM-CSF, IL-6 and IL-8⁴⁶. Seminal fluid taken from healthy men, and free of leucocytes, significantly increased TGF-β1 and IL-1β levels in the supernatants of endometrial epithelial cell cultures⁴⁷. In mouse experiments, TGFβ was the principal elicitor of uterine inflammation following mating⁴⁸. The TGFβ1 content of semen is comparable to that of colostrum, the most potent biological source of this cytokine⁴⁶. Semen from healthy men also contains significant levels of IL-7, SDF1α, MCP-1 and IL-8⁴⁹. After insemination, TGFβ1 tends to limit Th1-type responses, including those to paternal alloantigens which could undermine fertility⁵⁰. Robertson reported that exposure to semen results in increases in CD4+CD25+Foxp3+ regulatory T-cells in mice, and tolerance to paternal antigens⁵¹. Thus semen is a potent immune modulator in the cervix and uterus, and its effects on susceptibility to HIV infection are largely undefined.

The upper reproductive tract and HIV infection

The common occurrence of sexual transmission of HIV to women via vaginal intercourse was debated early in the AIDS epidemic; and anal intercourse was implicated as a possible mode of male-to-female transmission in the few incidents that were recognized⁵². Two key reports^{53, 54} indicated that male-to-female transmission occurred more regularly, but in these studies infection was detected in women who had undergone hysterectomy, and though these cross-sectional studies could not determine that hysterectomy preceded HIV transmission, the studies may have been interpreted as indicating that the upper genital tract was not a common site of HIV transmission. Most studies of HIV transmission to women, including studies of vaccine responses and microbicides, have focused on the lower genital tract.

The assumption that sexual HIV transmission to women occurs in the lower genital tract is not surprising since the upper tract is generally thought to be a sterile environment that is not easily accessed. However, MRI studies demonstrate that intravaginal gels reach the endocervix and uterus within minutes of application.^{55, 56} Once in the uterus, nonoxynol gels produce significant disruption of the endometrium including focal absence of the uterine epithelium proximal to the cervix⁵⁵, hemorrhage and areas of necrosis⁵⁵, as well as increased densities of CD8+ and Foxp3+ cells.⁵⁷

The uterus contains a variety of cell types that are susceptible to HIV infection, including CD4+ T-cells expressing CXCR4 and/or CCR5 as well as monocyte/macrophages. A polarized human endometrial epithelial cell line, HEC-1, has been shown to efficiently transcytose both X4 and R5 tropic strains of HIV^{58, 59}, allowing transfer of virus to susceptible target cells situated near the basolateral surface. Primary uterine epithelial cells can also transfer virus to CD4+ T-cells in vitro⁶⁰. In one recent study, primary tissue explants derived from ectocervix and endometrium of HIV-negative premenopausal women were infected ex vivo with R5-tropic HIV, and viral transcription levels in the two tissue types were compared using real-time polymerase chain reaction. Significantly higher levels of HIV transcription were detected in ectocervical tissues as compared to endometrial explants, suggesting that ectocervix may be more permissive than endometrium for HIV replication⁶¹. This may be related to a relatively higher density of T cells and dendritic cells in the cervix, particularly the cervical transformation zone, as compared to the upper reproductive tract^{61, 62}.

Few studies have addressed HIV-specific T-cell responses in the upper female reproductive tract. White and colleagues measured HIV-specific cytotoxic T-cell (CTL) activity in PBMC and a variety of upper and lower reproductive tissues from three HIV-positive women, including one long-term nonprogressor (LTNP) and two women on highly active antiretroviral therapy (HAART)⁶³. HIV-specific T-cell responses were found in ovaries, endometrium and endocervix of the LTNP, but such responses were not detected in the two women on HAART. Although limited in scope, these findings demonstrate that HIV-specific T-cell responses are present in the upper reproductive tract under some conditions, and provide a basis for further study.

Changes in the upper reproductive tract during the menstrual cycle

The immune microenvironment of the uterine endometrium fluctuates during the menstrual cycle, and significant changes in cell populations, growth factors, cytokines and adhesion molecules may be required to promote embryo implantation. There is a short window during the mid-secretory phase, limited to about 48h, during which conditions are favorable for implantation (for a review, see Aghajanova et al., 2008⁶⁴). This implantation window typically begins 6–10 days after the surge in luteinizing hormone (LH), roughly days 20–24 of an idealized menstrual cycle. Gene expression changes include upregulation of endometrial growth factors, cytokines and adhesion molecules, coupled with downregulation of genes associated with cell-mediated immunity and the classical complement pathway.

The hormonal fluctuations associated with the menstrual cycle regulate a variety of features of the reproductive tract that may directly impact susceptibility to HIV infection and persistence. Several years ago, White and colleagues demonstrated that cytolytic activity by CD8+ T-cells in the upper reproductive tract, as measured by redirected lysis assays, was present during the proliferative phase of the menstrual cycle (prior to ovulation), yet absent during the secretory phase⁶⁵. Notably, in contrast to the upper reproductive tract, CTL activity in the cervix and vagina persisted throughout the menstrual cycle⁶⁶. Postmenopausal women also retained strong cytotoxic T-cell populations in the upper tract. Taken together, these observations suggest tight post-ovulatory regulation of cell-mediated immunity in order to facilitate the implantation of the semi-allogeneic embryo. It is proposed that these fluctuations might contribute to a cycle-dependent “window of vulnerability” to HIV infection in the upper reproductive tract⁶⁷. Expression of CD4, CCR5 and CXCR4 by uterine epithelial cells and other cell types may also be temporally regulated during the menstrual cycle, with consequences for HIV infection and dissemination^{68, 69}.

Finally, the uterine endometrium also contains hormonally regulated, organized cellular aggregates, consisting of a B-cell core surrounded by CD8+ T-cells and macrophages, which are located in the stratum basalis layer⁷⁰. Aggregate size is smallest during the proliferative phase of the menstrual cycle, largest during the secretory phase, and absent from postmenopausal women. Whether these structures play an active role in antiviral defense is currently unknown.

Uterine natural killer (NK) cells: A novel cell type with an unusual dual role

Natural killer (NK) cells are distributed throughout the body and play a major role in immunosurveillance of virally infected and neoplastic cells. NK cells utilize a variety of effector mechanisms, including the release of cytolytic granules containing perforin and granzymes, the production and release of cytokines, and antibody-dependent cellular cytotoxicity (ADCC)⁷¹. Uterine NK cells (uNK) differ in cell surface phenotype from their counterparts in peripheral blood, typically expressing CD56 but not CD16 or CD57. Their density in the uterine endometrium increases after ovulation. Uterine NK cells appear to secrete a variety of cytokines and growth factors that help prepare the endometrium for embryo implantation. However, high uterine NK activity is also associated with recurrent miscarriage^{72, 73}. The extent to which these unique cells may participate in antiviral defense in the reproductive tract is currently unknown.

Immunological parallels between the reproductive and gastrointestinal tracts

The gut, of course, differs considerably from endometrium in terms of location, morphology and basic function (intake of nutrients and elimination of wastes). The endometrium’s function is to support implantation of the embryo and sustenance until the placenta is fully developed. However, the similarities between the tissues are numerous. Both sites have a dual role in maintaining a tolerogenic microenvironment while retaining rapid responsiveness to pathogens (Table 1)^{12, 74}. The ectocervix and vagina, like the lower anorectal canal, are characterized as Type II mucosal tissues, lined with stratified squamous epithelia and lacking organized mucosa-associated lymphoid tissue (MALT) structures. The endocervix and uterus, as well as the rectum and most of the large and small intestine, are characterized as Type I mucosal tissues, lined with simple columnar epithelium. These tissues produce secretory immunoglobulin, type A (sIgA), and can have organized MALT structures, including hormonally regulated uterine cellular aggregates, lymphoid follicles in the rectum, and intestinal Peyer’s patches.

Table 1.

Immunologic Properties Shared by the Female Reproductive Tract and Gastrointestinal Tract

Property	Female Reproductive Tract	Gastrointestinal Tract
Major portal of entry for HIV	Activated CD4+ cells expressing CCR5, CXCR4	Activated CD4+ cells expressing CCR5, CXCR4
Presence of Type I Mucosa Simple columnar epithelium Secretory IgA Presence of MALT	Endocervix, Uterus	Rectum, Large and Small Intestine
Presence of Type II Mucosa Stratified squamous epithelium Major antibody: IgG Absence of MALT	Ectocervix, Vagina	Lower Anorectal Canal
Mucus contributes to barrier function	Produced by vaginal epithelial cells, cervical crypts	Produced by goblet cells
Epithelial cells play multiple roles: Barrier function, antigen presentation, cytokine secretion, transcytosis of antigens	Yes	Yes
Beneficial endogenous flora	Lactobacilli prevent overgrowth of harmful bacteria Secretion of lactic acid maintains low pH	Role in metabolism, vitamin synthesis Stimulate production of iNOS and IgA TLR engagement promotes epithelial homeostasis
Immunologic tolerance	Directed towards preservation of conceptus Relies on complex mechanisms involving cytokines and Treg	Directed towards flora and food antigens Relies on complex mechanisms involving cytokines and Treg
Novel effector cell populations	Uterine NK cells with unique phenotype TCR gamma delta cells Intraepithelial CD8+ T-cells Th17 cells	Gut NK cells with unique phenotype TCR gamma delta cells Intraepithelial CD8+ T-cells Th17 cells
Innate effector molecules	SLPI, lactoferrin, defensins PRR (TLR, NLR) very important in defense	lysozyme, cathelicidins, defensins PRR (TLR, NLR) very important in defense
Sex steroid responsiveness	Significant modulation of tissue architecture, development of inducible MALT	Estrogen receptor ligands affect crypt/villus architecture and may play a role in inflammatory bowel diseases

Both the reproductive and gastrointestinal tracts are protected by mucus layers, and by epithelial cells that play multiple roles in barrier function, antigen presentation, secretion of cytokines, chemokines and growth factors, and transcytosis of macromolecules. Both tissue systems are extensively colonized by endogenous flora. These flora both prevent the overgrowth of pathogenic bacteria and provide useful “housekeeping” functions; in the case of intestinal flora, these include the synthesis of vitamins and hormones, stimulation of iNOS⁷⁵ and IgA^{76, 77} production, and the maintenance of epithelial homeostasis through signaling of Toll-like receptor (TLR) pathways⁷⁸.

Both the reproductive and gastrointestinal tracts engage in immunological tolerance. In the reproductive tract, tolerance is related to the implantation and preservation of a semi-allogeneic embryo; in the GI tract, tolerance is maintained with respect to food antigens and commensal bacteria. At both sites, the mechanisms regulating tolerance are complex and appear to involve Foxp3+ regulatory T-cells and immunosuppressive cytokines such as TGF-β.

Th17/CD4+ cells and IL-17 play important roles at each site. Depletion of Th17+ cells and lack of IL-17 is associated with impaired intestinal barrier function, as illustrated by increased dissemination of S. typhimurium in simian immunodeficiency virus infected macaques^{79, 80}. In humans, IL-17 and related cytokines are involved in a number of autoimmune and inflammatory disorders, including inflammatory bowel disease^{81, 82}, conditions that, like HIV infection, are characterized by altered gut barrier function⁸³. In the female genital tract, Th17 cells and IL-17 are associated with increased IL-8 and endometriosis,⁸⁴ and have been implicated in spontaneous abortion⁸⁵.

From the perspective of HIV infection, both the reproductive and gastrointestinal tracts serve as important portals of entry. Both tissues are rich in partially activated CD4+ memory T-cells, many of which express CCR5 and/or CXCR4. Both harbor significant populations of HIV-specific memory CD8+ T-cells. Our laboratories are currently studying HIV-specific CD8+ T-cell responses in blood, reproductive mucosa and gastrointestinal mucosa of HIV-infected women^{74, 86}. Although these studies are still preliminary, early results suggest that there are significant differences in antigen responsiveness between CD8+ T-cells isolated from endometrial biopsy as compared to those from PBMC or the gastrointestinal tract of the same individuals (Figure 1). Additional studies will be required to extend these results.

Figure 1.

CD8+ T-cell responses to HIV Gag peptides (left) and a mixture of immunodominant peptides derived from cytomegalovirus, Epstein-Barr virus and Influenza A (C-E-F) (right). Freshly isolated lymphocytes from PBMC, endometrial biopsy, and sigmoid colon were tested in a cytokine flow cytometry assay for production of IFNγ, and the cytolytic granule marker CD107a in response to peptide stimulation. Numbers in individual quadrants indicate the number of cells producing either IFNγ (upper left quadrant), CD107 (lower right quadrant), or both IFNγ and CD107 (upper right quadrant) in response to peptide stimulation. Assays were performed as described elsewhere⁸⁸, ⁸⁹. The samples used in this experiment were obtained from a premenopausal HIV-positive woman on long-term HAART (9 years), with plasma viral load <75 copies/mL, blood CD4+ T-cell count 1180, from whom blood and tissue sampled were obtained on day 24 of the menstrual cycle.

Implications for studies on prevention of HIV infection and HIV immunopathogenesis

Women are increasingly affected by the HIV epidemic worldwide, and now account for nearly half the 33 million individuals living with HIV⁸⁷. In sub-Saharan Africa, women account for approximately 60% of HIV infections⁸⁷. As the HIV/AIDS epidemic enters its second quarter century, many questions remain unanswered concerning the disease process, and many more studies are needed to advance our understanding of HIV pathogenesis in women.

Due to the susceptibility of endocervical and endometrial cell populations, these tissues are potential sites of HIV infection. Current data do not indicate the relative frequency of any female genital site as the location of initial HIV infection, and it is possible that tissue specific susceptibility to infection depends on the existence of factors such as sexually transmitted infections, trauma, ovulatory cycle, seminal plasma effects or hormonal modulators of genital immune responses. Thus the upper tract tissues should be considered in studies of HIV vaccines, microbicides and other prevention interventions. Several studies point to a potential “window of vulnerability” to HIV infection at 7 to 10 days following ovulation⁶⁷. Addressing this question in the laboratory will require detailed studies of mucosal immunity in women across the menstrual cycle, and a better understanding of the influence of reproductive hormones on adaptive and innate mucosal immunity. A second, and related area of interest is the potential influence of hormonal contraception on HIV transmission.

Since semen and seminal fluids influence female genital immunology, studies of the tissue effects of prevention modalities must also include evaluations of semen exposure. Studies of the post-menopausal genital tract, and HIV susceptibility and shedding are also needed given the increased survival of HIV patients, and the large number of women who are living with this infection worldwide.

The limited existing evidence indicates that HIV exerts parallel effects on mucosal lymphoid tissues in the gut and endometrium. Although the rapid CD4+ T-cell decline that occurs during acute HIV infection of the GI tract has been extensively studied, much less is known about the effects of acute HIV infection on reproductive tissues. Similarly, the role of immune activation and inflammation within the reproductive tract on HIV disease has not been thoroughly studied, nor have the effects of HAART on CD4+ T-cell repopulation in the reproductive mucosa. These questions should prompt further study since endometrial tissues are readily obtained and no enteric preparation is required. Since tissue sampling in human studies of infrequent HIV phenotypes, such as in acute infection, are often problematic, the use of endometrial sampling may provide an opportunity to obtain MALT tissues more readily.