Regulation of mucosal immunity in the female reproductive tract: the role of sex hormones in immune protection against sexually transmitted pathogens.

We have previously shown that polarized human uterine epithelial cells secrete antimicrobial molecules that kill or inhibit pathogenic bacteria, fungi, and viruses. 63 , 77 , 143 Apical secretion of these molecules is probably one reason why the uterine lumen has a relatively low number of microorganisms compared to the lower FRT despite flow of fluids throughout the entire tract. 39 Secretion of antimicrobials is both constitutive and induced by TLR agonists and microorganisms, and is affected by many factors such as age, menstrual status, health, and the presence of chemokines, cytokines and steroid hormones. As seen in Fig. 15 , uterine epithelial cell secretions inhibit Neisseria gonorrhoeae and Candida albicans (yeast and hyphal form), as well as reduce HIV-1 (R5) infection of target cells. Similar inhibition of pathogens is seen with secretions from Fallopian tube, cervical, and ectocervical epithelial cells (not shown). In contrast, none had an inhibitory effect on Lactobacillus crispatus , a common commensal in the vagina. Analysis of cytokines and chemokines in uterine secretions revealed several molecules that could account for pathogen inhibition. 144 These findings provide definitive evidence for the critical role of epithelial cells in protecting the FRT from infections, without comprising the beneficial presence of L. crispatus , which is part of the normal vaginal microflora of humans. The presence of sex hormones can dramatically alter anti-bacterial activity by epithelial cells. Secretions from human uterine epithelial cells treated with multiple concentrations of E 2 for at least 48 hr show greater inhibition of Staphylococcus aureus growth than those secretions from untreated cells ( Fig. 16 ), suggesting that E 2 induces the secretion of antimicrobial(s) in addition to that seen under control conditions. This reinforces the premise that E 2 can have a significant effect on antimicrobial secretion by human uterine polarized epithelial cells, and suggests that antibacterial, antifungal and antiviral activity in cell secretions changes during the menstrual cycle.

Sexually transmitted infections (STI) have reached epidemic proportions throughout the world with more than 20 pathogens transmitted through sexual intercourse. The World Health Organization estimates that more than 300 million new infections of Trichonomas vaginalis , Chlamydia trachomatis , or Neisseria gonorrhea occur annually ( http://www.who.int/mediacentre/factsheets/fs110/en/index.html ). Some STI can be transmitted vertically to the fetus, resulting in preterm deliveries and/or life-threatening systemic illness in newborn infants. Generally, adolescents and young adults are the demographic age groups most frequently affected with STI, and women are more likely than men to suffer the consequences of these serious infections ( http://www.who.int/hiv/data/2009_global_summary.png UGso-tAe2009). 1 Human Immunodeficiency Virus (HIV) is unique in its rapid spread and the depth of its impact. With 25 million deaths worldwide and an additional 33.2 million (of which 50% are women) infected, HIV is one of the world’s worst pandemics. 2 Since the 1980s, in parts of the world HIV has shifted from a disease spread predominantly through contaminated needles, blood products and male-male sexual contact, to a sexually transmitted disease in which women worldwide are more likely to be infected than men ( http://www.who.int/mediacentre/factsheets/fs360/en/index.html : HIV/AIDS2012). Presently, women and girls make up almost 57% of all people infected with HIV in Sub-Saharan Africa, where a striking 76% of young people (aged 15–24 years) living with HIV are female. 2 Within the female reproductive tract (FRT), the mucosal immune system functions as the first line of defense. 3 – 5 In response to the unique requirements of balancing immune protection with procreation, the immune system in the FRT, which consists of both innate and adaptive immune components, is responsive to and precisely regulated by the sex hormones estradiol (E 2 ) and progesterone, both of which are produced in a cyclic fashion by the ovary over the course of the menstrual cycle. In preparing the reproductive tract for fertilization and implantation, E 2 and progesterone simultaneously regulate the immune system in the Fallopian tubes, uterus, cervix, and vagina to complement the reproductive process (see ref. 6 for review). Key constituents of the mucosal immune system in the FRT are a dynamic population of immune cells which migrate into the uterus, cervix and vagina as well as resident epithelial cells and supportive stromal cells. 6 Sex hormones influence the migration of macrophages and dendritic cells (DC) as well as T and B cells by affecting the expression of adhesion molecules and chemotactic factors. 6 – 9 Epithelial cells, in addition to providing barrier protection, transport immunoglobulins (IgA and IgG) into FRT secretions, and produce antimicrobials that are both bactericidal and viricidal. 7 , 10 Through the production of cytokines and chemokines, these cells signal the recruitment and activation of other cells of the innate and adaptive immune systems. In this dynamic balance, epithelial cells, fibroblasts, and immune cells throughout the FRT respond directly to E 2 and progesterone, as well as indirectly to the cytokines and growth factors. What is clear is that this responsiveness is part of the bidirectional communication that occurs in which epithelial cells direct both reproductive as well as immune function to maintain an effective level of protection which distinguishes between pathogens, commensals, allogeneic sperm, and the developing fetus. The pleiotropic capacity of epithelial cells has led to their recognition as sentinels, the functions of which are only now being recognized. 6 , 11 , 12 This review focuses on our current knowledge regarding the role of epithelial cells, fibroblasts, and immune cells in the human FRT, with special emphasis on protection against STI. Our goal is to highlight some of the unique responses of these cells to E 2 and progesterone and to point out that, in addition to direct hormonal effects on particular cells, there are the equally important indirect actions of E 2 and progesterone mediated through growth factors, cytokines, and chemokines.

Fibroblasts are key structural components of reproductive tissues. However, recent findings have shown they are not passive bystanders, but active players in the immune response. Epithelial cell interactions with underlying stromal fibroblasts are essential in facilitating steroid hormone-induced growth and development in the EM. Cooke et al. 102 used FRT tissues from estrogen receptor knockout mice to demonstrate that underlying estrogen receptor positive stromal cells regulate the differentiation of adjacent epithelial cells. These studies indicated that E 2 acts on stromal cells to release one or more paracrine factors that then modulate E 2 effects on FRT epithelial cell growth and differentiation. Hepatocyte Growth Factor (HGF), a pleiotrophic agent initially shown to stimulate proliferation of hepatocytes in vivo, 103 is produced primarily by stromal fibroblasts. 104 , 105 HGF has been well characterized in terms of its normal physiological roles of increasing epithelial cell motility and proliferation, wound healing, 106 angiogenesis, 107 , 108 epithelial cell scattering, and embryogenesis 109 (see review ref. 110 ). Using primary human uterine epithelial cells, Sugawara et al. 111 found that HGF stimulates proliferation, migration, and morphological changes in human uterine epithelial cells and concluded that HGF may modulate the cyclic regeneration and development of the endometrial lining of the uterus. As shown in Fig. 8 , treating uterine stromal fibroblasts with estradiol significantly increased HGF secretion; E 2 -induced secretion of HGF increased with the duration of exposure to E 2 . 112 Cells treated with E 2 for 6 days secreted three times more HGF than cells exposed for 2 days. The long-lasting effect of E 2 on HGF secretion by uterine stromal fibroblasts could have implications for cancer or endometriosis where continued HGF secretion might lead to further proliferation of uterine epithelial cells. In related studies using a co-culture system, mouse uterine stromal fibroblasts cultured below uterine polarized epithelial cells in cell inserts produced HGF that significantly increased TER. 113 When epithelial cells and/or stromal cells were incubated with anti-HGF or anti-HGF receptor (HGFR) antibody prior to the addition of HGF, the effect of HGF was blocked. Addition of recombinant HGF to the basolateral compartment of polarized epithelial cells increased TER in a dose-dependent manner. These findings indicate that epithelial cells express the HGFR at their basolateral surfaces and that HGFR mediates the effects of HGF on TER. 114 In contrast, when cells were incubated with TGFβ, TER was markedly but reversibly suppressed. Based on these findings, we conclude that TGFβ and HGF may play regulatory roles in modulating epithelial cell tight junctions. Further studies are needed to determine the relative contributions of hormone balance ( Fig. 2 and Table I ), pathogen exposure, cytokine, and growth factor secretion ( Fig. 8 ), to innate and adaptive immune protection against pathogens in the FRT. The role of fibroblasts in mediating the immune response against viral pathogens including HIV in the FRT is not well-understood and few studies have addressed this. We have shown that confluent layers of purified fibroblasts from the distinct anatomical regions of the FRT secrete a panel of anti-HIV factors including CCL20/MIP3α ( Fig. 9 ), IL-8, RANTES, and SDF-1α (not shown). However, fibroblasts isolated from the EM and Fallopian tubes secreted substantially higher levels of CCL20 than those from the lower FRT. In preliminary studies, diluted conditioned media (CM) (1:4) from EM fibroblasts of two women showed anti-HIV activity, and reduced R5 (BaL) infection of TZM-bl cells by approximately 75% ( Fig. 10 ). As a part of these studies, we found that FRT fibroblasts demonstrated site-specific differences when treated with E 2 . Estradiol stimulated SDF-1 production in EM fibroblasts but had no effect on fibroblasts isolated from the cervix or Fallopian tubes (M. Patel, M. Rodriguez-Garcia, J.V. Fahey, C.R. Wira, Submitted). Monocytes circulate through the periphery, mediating immune recognition and pathogen clearance by phagocytosis and indirect stimulation of the immune system through production of key cytokines and chemokines. As shown in Fig. 11 (top), the TLR4 agonist LPS, the key antigenic component of many pathogenic bacteria, stimulates peripheral blood monocyte secretion of IL-1β, thereby causing a proinflammatory response. In the presence of E 2 , this response is further enhanced in a dose dependent manner ( Fig. 11 , bottom) 115 and may be detrimental to the host especially in the FRT. 91 , 92 In other studies, we found that increasing E 2 levels in the FRT can down-modulate immune associated HBD2 and IL-8 responses to IL-1β (see Fig. 5 ). We have further shown that this response is mediated through a down-regulation of interleukin-1 receptor type I (IL - 1RtI) protein expression ( Fig. 11 ). Overall, E 2 has the ability to enhance pathogen induced IL-1β expression by monocytes but at the same time decrease IL-1β receptor expression by epithelial cells. These studies demonstrate a close link between the endocrine and immune systems in the FRT that may be crucial for dampening proinflammatory responses during the time of ovulation or pregnancy. Even though sex hormones have been described to influence HIV infection in epidemiological studies and regulate different immune responses that may affect HIV infection, the direct role that female sex hormones play in altering the susceptibility of target cells to HIV-infection is largely unknown. To examine the direct effect of E 2 on HIV-infection in an in vitro infection assay, purified CD4 + T cells (>98% purity) were activated in vitro in the presence or absence of E 2 for 3 days and infected with R5 (HIV-1 BaL ) viral strain. 116 Secreted and intracellular p24 were measured 7 days after infection as an indication of viral replication. As shown in Fig. 12a , when CD4 + T cells were treated with E 2 prior to infection (pre) for 3 days, released p24 was significantly reduced 7 days after infection with HIV-1 BaL (56% reduction; P = 0.024). However, when E 2 was added for the entire length of incubation (pre/post) or 2 hr after infection (post) no differences were found compared to the control condition. This observation was confirmed by a significant reduction in the expression of intracellular p24 in CD4 + T cells pre-treated with E 2 before infection (not shown). 116 As a control to demonstrate that the shifts into the p24 positive population represent de novo infection and not residual virus, CD4 + T cells infected in the presence of azidothymidine (AZT) showed no infection (not shown). We investigated whether the reduction in HlV-1 susceptibility induced by E 2 could be due to differences in activation. CD4 + T cells treated or not with E 2 showed similar expression of CD25 and HLA-DR before HIV challenge (data not shown). Since these experiments were conducted with cells from both female and male donors, we analyzed if any differences could be found between them. Infection levels in the infected controls from female donors were lower than male donors, with median p24 values of 38,085 versus 71,541 respectively ( P = 0.03). Treatment with E 2 prior to viral challenge significantly reduced HIV-infection in women (54.9% reduction; P = 0.03), but the reduction (18.3%) was not significant in men. In other studies, recognizing that ethinyl estradiol (EE) is the main estrogenic component in most oral contraceptives and that contraceptives may be a risk factor for HIV infection, 117 , 118 we investigated if EE would have the same effect as E 2 in preventing HIV-infection of target cells. Following the same experimental design described above, CD4 + T cells were activated in the presence of EE and infected with an R5 viral strain. Irrespective of the time (pre) and length (pre/post/post) of hormone treatment, and unlike E 2 that inhibited HIV infection, EE had no effect on CD4 + T cell infection with HIV-1 BaL (not shown). Similar to the lack of inhibition with E 2 , we observed a pattern of reduced HIV-1 IIIb viral replication when EE was added 2 hr after infection, but no statistical significance. Recognizing that CD4 + T cells and macrophages are the most likely targets for HIV infection, we next focused on the effect of E 2 on macrophage infection. Monocyte-derived macrophages were differentiated in vitro in the presence of E 2 for 4 days and infected with HIV-1 BaL . As shown in Fig. 12b , similar to our results with CD4 + T cells, viral replication was significantly reduced in macrophages differentiated in the presence of E 2 ( P < 0.0003). This reduction in susceptibility to HIV-infection was induced by differentiating the macrophages in the presence of E 2 before infection (pre; 69% reduction) and maintained when E 2 was added back to the culture 2 hr after infection (pre/post; 73% reduction). In contrast, E 2 had no effect when added immediately after infection (post) ( Fig. 12a, b ). Intracellular p24 analysis confirmed a significant reduction in the percent of p24 positive cells when macrophages were differentiated in the presence of E 2 (not shown). Intracellular p24 values correspond to de novo infection, since macrophages infected in the presence of AZT as a control show intracellular p24 values equal to uninfected controls. Comparison between macrophages derived from female monocytes or male monocytes showed no statistical differences, and in both cases E 2 pre-treatment was able to significantly reduce HIV-infection (not shown). When Fig. 12a, b are compared, E 2 appears to be more effective in reducing susceptibility to HIV-infection in macrophages than CD4 + T cells. Furthermore, the suppressive effect was maintained when E 2 was present after infection. This maintenance of suppression represents a difference with respect to the effect observed in CD4 + T cells in that the inhibitory effect induced by pre-treatment with E 2 is lost when E 2 is present after infection ( Fig. 12a ). In contrast with T cells, EE was able to decrease macrophage susceptibility to HIV-infection, When macrophages were differentiated in the presence of EE, released p24 was significantly reduced 7 days after infection ( P < 0.005). When CD4 + T cells and macrophages from the same donors were analyzed in parallel to compare the effects of E 2 and EE, intracellular p24 staining confirmed the lack of effect on CD4 + T cells but decreased the% of infected macrophages for both E 2 and EE pretreatments. Pre-treatment with E 2 , however, was consistently 10–20% more effective than EE in suppressing viral replication (not shown). In other studies, E 2 suppresses HIV-infection in a dose-dependent manner with both T cells and macrophages. To investigate the possible mechanisms involved in E 2 suppression of viral replication, we first examined the expression of CCR5, the main coreceptor for R5 viral strains. CD4 + T cells were activated in the presence of E 2 and CCR5 expression was assayed by RNA and flow cytometry immediately prior to HIV-infection. No differences were observed in CD4 + T cell CCR5 gene expression after E 2 treatment, relative to controls. Expression levels of CD4 were also measured, but no differences were found between controls and E 2 treated cells (not shown). Macrophages differentiated in the presence of E 2 or EE were also analyzed for the expression of CCR5 at the time of infection and neither E 2 nor EE had any effect on CCR5 gene expression relative to controls. To better define the role of E 2 in affecting viral entry, we used a single cycle, VSV-G pseudotyped virus, which enters the cells by endocytocis, keeping the need for receptor/coreceptor attachment and fusion. 119 No differences in infection levels were found between the control conditions and cells pretreated with E 2 , strongly suggesting that E 2 inhibits infection at the steps of viral entry or fusion. Expressed by DC, DC-SIGN is a calcium-dependent carbohydrate-binding protein. 120 , 121 As an adhesion receptor DC-SIGN interacts with ICAM-2 on endothelial cells to both induce migration of DCs from blood to tissue 122 and mediate clustering of DCs with naive T cells through binding of ICAM-3. 120 Of equal importance in the FRT, DC-SIGN functions as a pattern recognition receptor that induces specific immune responses upon interaction with a number of pathogens. 123 This carbohydrate recognition pattern is the basis of its broad specificity for different pathogens and might also be responsible for its distinct signaling properties. While initial target cells for mucosal transmission of HIV-1 in the reproductive tract have not been clearly determined, studies conducted in non-human primate models indicate the potential role of DC as initial targets for mucosal infection and systemic dissemination of HIV-1. 124 – 126 DC can capture HIV-1 using surface expressed C-type Lectin Receptors (CLRs), 127 – 129 among which DC-SIGN is the best characterized. 122 HIV-1 captured by DC can be transferred to other target cells including CD4 + T lymphocytes via virological synapses, in a process referred to as trans infection. 20 , 130 Understanding the mechanisms by which DC-SIGN and other HIV-1 receptors are regulated on mucosal target cells remains critical to designing strategies to prevent sexual transmission. In previous studies aimed at elucidating the role of cell-cell interactions in the EM on the expression of HIV-1 receptors, we examined the effects of soluble factors produced by uterine epithelial cells on the expression of DC-SIGN by DC. 131 In these experiments, immature DC were differentiated from human monocytes in the presence of IL-4 and GM-CSF. To assess the CM effects on DC-SIGN expression, DC were differentiated in the presence or absence of CM consisting of basolateral secretions from polarized epithelial cells for 7 days prior to flow cytometry. As shown in Fig. 13a , we found that uterine epithelial cell secretions (i.e. CM) decreased DC-SIGN expression on immature DC via a TGFβ mechanism. These results suggested that DC generated in the presence of CM (i.e. CM-DC) might have reduced capacity for trans infection of HIV-1 relative to control DC. To directly test this hypothesis, HIV-1 trans infection assays were undertaken ( Fig. 13b ). When immature DC were pulsed with HIV-1, prior to virus infection of TZM-bl cells, CM inhibited DC-mediated trans infection of HIV-1. 131 These results provided direct evidence for uterine epithelial cell regulation of DC transmission of infection with reference and transmitted/founder HIV-1 variants. Interestingly, whereas recombinant TGFβ1 inhibited trans infection of prototypic reference HIV-1 by DC, TGFβ1 had a minimal effect on trans infection of transmitted/founder variants irrespective of the reporter system used to measure trans infection. These findings have immediate implications for designing strategies to prevent sexual transmission of HIV-1. We recently proposed a model for the existence of a window of vulnerability for HIV-1 infection during a woman’s menstrual cycle. 5 During this critical period, 7–10 days of the normal menstrual cycle, critical components of innate and adaptive immune responses are suppressed by E 2 and/or progesterone to facilitate reproductive processes. HIV-1 presumably exploits this time frame, during which antiviral factors are suppressed, to establish and propagate infection in the female genital tract mucosal. 5 The reported findings showing that E 2 increases DC-SIGN expression reveals yet another potential dimension by which sex steroid hormone may modulate susceptibility to HIV-1 infection in women. It is likely that in addition to suppressing antiviral effector mechanisms, the profile of sex steroid hormones present during the window of vulnerability in women may increase HIV-1 acquisition through enhanced expression of HIV-1 capture receptors by DC. More studies are needed to directly test the effects of sex steroid hormones on infection of DC with HIV-1. Natural killer (NK) cells are important in innate immunity, not only for their ability to kill certain tumour cells and viral-infected cells without prior immunization or MHC restriction, but also for their secretion of immunoregulatory cytokines that contribute to early host responses against viruses, bacteria and fungi. 132 NK cells account for a large percentage of leukocytes in the human EM and, since their numbers increase as the menstrual cycle progresses, 16 – 18 , 133 Sentman et al. 134 hypothesized that recruitment and/or expansion of uterine NK cells are regulated by sex hormones. As shown in Fig. 14a, b , they tested the effect of various doses of E 2 on human endometrial slices and found that 10 −9 M E 2 significantly enhanced mRNA expression of the chemokines CXCL10 and CXCL11. Since the estrogen receptor inhibitor ICI182780 blocked the E 2 -induced increase of these two chemokines, the effect was dependent upon E 2 receptor activity (not shown). Progesterone at 10 −8 M also increased the mRNA expression of CXCL10 (8.6-fold) and CXCL11 (12.2-fold) in the endometrial organ culture system. 134 Sentman et al. 134 have suggested that the sex hormones do not act directly on NK cells, but rather act on stromal and epithelial cells to produce chemokines for NK cell recruitment. Since E 2 and progesterone stimulate production of these cytokines in the EM, it is likely that these sex hormones contribute to the migration of NK cells needed for immune surveillance, pathogen response, pregnancy, and normal menstrual cycle function. One innate mechanism that regulates NK activity is TGFβ, which is regulated by sex hormones. TGFβ is typically produced in a pro-form that requires activation 135 ; the precise mechanisms that lead to activation of TGFβ in tissues are not well understood. It has previously been shown that uterine NK cell responses to monokines and blood cell NK cell responses to TLR agonists were inhibited by endogenous TGFβ. 136 , 137 Uterine NK cells express TLR and can respond to TLR agonists by producing cytokines. 138 Eriksson et al. 139 explored how endogenous TGFβ modulates the production of interferon-γ (IFN-γ) by human uterine NK cells. As shown in Fig. 14c, d , IL-12 and IL-15 in combination and poly (I:C) increase the percent of IFN-γ producing NK cells in the uterus, and antibody to TGFβ enhances the number and fold change of uterine NK cells induced by poly (I:C) that produce IFN-γ. Thus, endogenous TGFβ suppresses poly (I:C)-induced IFN-γ production by uterine NK cells. Since one way that E 2 regulates FRT immunity is by modulating the production and/or activation of TGFβ, it is likely that E 2 regulates NK activity via endogenous TGFβ. At mucosal surfaces, neutrophils are responsible for rapid elimination of potential pathogens through phagocytosis of microbes, release of antimicrobial compounds, and production of toxic oxygen and nitrogen species. 140 In studies to determine whether sex hormones modulate neutrophil phenotype, Smith et al. 19 , 141 examined blood neutrophils from women at days 7, 14, 21, and 28 of the menstrual cycle for expression of surface receptors, granule proteins, and intracellular cytokines. Blood neutrophil phenotype varied during the menstrual cycle with decreased expression of CD89 (IgA Fc receptor) and TNFα during the periovulatory period. In other studies, cytokines were analyzed for their ability to enhance the innate immune potential of neutrophils by altering receptor expression and cell function. 142 These studies indicated that GM-CSF, known to be produced by FRT epithelial cells, acted synergistically with the chemoattractant IL-8 to promote neutrophil chemotaxis. Use of antibody neutralization and CM from primary confluent cultures of epithelial cells led to the conclusion that FRT epithelial cells are a potent source of neutrophil chemoattractant activity.

The immune system in the FRT has evolved to be responsive to and precisely regulated by E 2 and progesterone, which are produced in a cyclic fashion by the ovary during the menstrual cycle. In preparing the reproductive tract for fertilization and implantation, E 2 and progesterone simultaneously regulate the immune system in the Fallopian tubes, uterus, cervix, and vagina to complement the reproductive process (see ref. 5 for review). Over the course of the menstrual cycle, immune cells are present in substantial numbers and non-uniformly distributed in both the stromal layer and the epithelium of the FRT. 13 – 16 Flow cytometry analyses of enzymatically and mechanically dispersed tissue fragments demonstrated that leukocytes are 6–20% of the total number of cells within the FRT. 17 When natural killer (NK) cells were included in this analysis, the population of immune cells doubled. 18 These studies indicated that T-lymphocytes (CD4 + and/or CD8 + ) are a major constituent of reproductive tract leukocytes from all tissues. The Fallopian tube contained granulocytes as a second major constituent, but were significantly less numerous in the other tissues. All tissues contained B-lymphocytes and monocytes as clearly detectable but minor components. Over the course of the menstrual cycle, subtle changes were observed in the migration of macrophages, B cells and neutrophils into the lower tract, 17 , 19 and in DC entering the squamous epithelium. 20 As seen in Fig. 1 , the distribution of immune cells in the FRT varies with the site examined. Exclusive to the uterus are lymphoid aggregates, which consist of a B cell core surrounded by T cells and an outer halo of macrophages ( Fig. 1a, f ). The B cell core (CD19 + ) was most often seen in large aggregates present in the late proliferative and secretory stages of the cycle. 21 Phenotypic analysis indicated that T cells are almost exclusively CD3 + , CD8 + , and CD4 − . Aggregates containing only cells of the CD3 + CD4 + phenotype were occasionally found as were individual CD4 + cells located outside the aggregates in the stroma. Monocytes/macrophages (CD14 + cells) were found as a mantle around the T cells. FRT aggregates are anatomically and functionally distinct from Peyer’s patches in the intestine. 21 The size of lymphoid aggregates was found to vary with the stage of the menstrual cycle, and was significantly larger during the secretory (3000–4000 cells) than the proliferative stage (300–400 cells). The distribution and frequency of CD8 + T cells in aggregates using expression of Vb2 or Vb8 as markers of clonality and Ki-67 as a marker of dividing cells led to the conclusion that lymphoid aggregates form largely by the trafficking of cells to nucleation sites within the endometrium (EM), rather than by division of precursor cells. 22 The lower reproductive tract, while lacking lymphoid aggregates, contains a full spectrum of immune cells located both within the submucosa and epithelial lining ( Fig. 1g–i ). Coincident with aggregate formation in the uterus, White et al. 23 found that CD8 + cytotoxic T lymphocyte (CTL) activity, measured in a redirected lysis assay, is suppressed in the uterus and Fallopian tubes during the secretory stage of the cycle. This suppression occurs without any drop in CD8 + T cell numbers. In the ectocervix and vagina, in contrast with the upper FRT, we found that CTL activity was measurable in tissues from women at the proliferative or secretory stages of the menstrual cycle. 24 Recognized as important molecules in implantation, 25 , 26 we measured the expression of chemokine receptors CXCR4 and CCR5 as well as CD4 on uterine epithelial cells and found that expression varies with the stage of the menstrual cycle ( Fig. 1a–d ). 27 All three were low during the proliferative stage of the cycle, peaked at the time of ovulation and then either plateaued (CXCR4, CD4) or declined (CCR5) during the secretory stage of the cycle. 27 Chemokine receptors on cultured endometrial epithelial cells showed an up-regulation and polarization of CXCR1, CXCR4, and CCR5 receptors when a human blastocyst was present. 25 The distribution and regulation of these receptors in the endometrial epithelium and the human blastocyst suggest that each is essential in the apposition and adhesion phases of human implantation. In addition to expression in the upper FRT, leukocytes and epithelial cells in the lower tract ( Fig. 1g–i ) express CCR5 and GalCer on both. 28 Yeaman et al. 28 showed that basal and parabasal epithelial cells of the ectocervix express CD4, CCR5, and GalCer, unlike the midzone and superficial cells lining the lumen. Although changes in protein expression were not as pronounced as those seen in the uterus, histological evidence supported the conclusion that CD4 and CCR5 expression was greater during the proliferative stage than during the secretory stage of the cycle. An unintended consequence of chemokine expression in the upper and lower tract is that HIV-1, as a sexually transmitted pathogen, most likely uses these coreceptors to infect cells in the FRT. The humoral immune system is hormonally controlled and varies with site analyzed and stage of the menstrual cycle (For review see ref. 10 ). In the uterus, levels of pIgR, the epithelial cell receptor responsible for transporting IgA from tissue to lumen, vary with the menstrual cycle. 29 When expressed as the percentage of total protein, luminal uterine secretory component (SC) levels were highest during the secretory phase, significantly reduced during the proliferative phase and lowest during menstruation. Total SC was also greatest during the secretory phase, averaging approximately twofold higher than SC in proliferative and menstrual samples. In other studies, IgG levels in secretions from the uterine mucosa were highest during the periovulatory phase, whereas levels in the Fallopian tube were lowest at that time. 30 This study reached the conclusion that each organ (Fallopian tubes, uterus, cervix, and vagina) and even different sites within each organ can respond independently from each other to changes in hormone levels, producing different types and amounts of secretory proteins. In the lower FRT, IgA, IgG, and lactoferrin levels in cervical mucus were depressed by 10–100-fold at midcycle relative to that seen early in the proliferative phase, only to rise toward the end of the menstrual cycle. 31 When women were placed on oral contraceptives, immunoglobulins and lactoferrin levels were suppressed for the duration of hormone exposure. In other studies, in which cervical mucus was evaluated from 5 days before to 3 days after ovulation, IgA and IgG had a biphasic pattern with a peak before ovulation followed by a small increase after ovulation. 32 Nardelli-Haefliger et al. 33 demonstrated that titers of anti-human papillomavirus 16 virus-like particle (VLP) IgG in cervical secretions dropped approximately ninefold at midcycle during ovulatory cycles suggesting increased vulnerability to pathogens at midcycle. Chemokines and cytokines are central to the progressive tissue growth and remodeling that occur during each menstrual cycle while antimicrobials are secreted proteins that protect against pathogens. 6 However, they are not mutually exclusive groupings, and many proteins are classified in both categories. For example, secretory leukocyte protease inhibitor (SLPI) is a potent inducer of neutrophil chemotaxis, alters Toll-like receptor (TLR) signaling, is bactericidal against Escherichia coli and Staphylococcus aureus amongst others, and possesses anti-HIV activity. While relatively little has been done in upper FRT during the menstrual cycle, several studies have demonstrated that ovarian stimulation markedly increases the profile of cytokines, chemokines, and growth factors in uterine secretions. 34 – 36 Analysis of cervical-vaginal lavages (CVL) demonstrate that chemokines and cytokines (IL-6 and IL-8) as well as endogenously produced antimicrobials (SLPI, HBD2, HNP1–3, and lactoferrin) dropped significantly at midcycle (day 13) and remained depressed for 7–10 days, returning to proliferative stage levels just before menstruation. 31 , 37 In contrast, total protein and TGFβ levels remained unchanged throughout the menstrual cycle. In other studies, human intestinal defensin-5 was highest in CVL during the secretory stage of the menstrual cycle. 38 As discussed elsewhere, 39 in contrast with CVL findings, studies using tampons for collection of vaginal fluid reported increased levels of HNP1–3, HBD2, and lysozyme while lactoferrin, HBD1, and SLPI decreased from proliferative to secretory stages of the menstrual cycle with no apparent mid-cycle decrease. 39 When Dacron swabs were used, SLPI peaked at mid-cycle compared to proliferative and secretory stages. 40 Further studies are needed to determine which recovery technique (CVL, tampon, or swab) most accurately reflects antimicrobial levels in the lower FRT. Overall, these studies support the hypothesis that hormonal changes during the menstrual cycle regulate the immune system throughout the FRT in a way that is synchronized with reproductive function which optimizes FRT conditions for successful sperm migration, fertilization, implantation, and pregnancy. The extent to which changes in the immune system are directly or indirectly mediated by E 2 is addressed in the following sections.

In conclusion, analysis of the FRT indicates that the innate and adaptive immune systems are present and functional throughout the menstrual cycle. Not widely appreciated is a growing body of data indicating that the immune system in the Fallopian tubes, uterus, cervix, and vagina function independently from each other to changes in hormone levels during the menstrual cycle. The net result is coordinated immune protection that complements the reproductive demands of each organ. As the first line of this protection, epithelial cells, macrophages, DC, NK cells, and neutrophils in the FRT function to meet the challenges of STI, while at the same time supporting an immunologically distinct fetal placental unit. The studies presented demonstrate overwhelmingly that E 2 secreted by the ovary during the menstrual cycle acts both directly and indirectly on epithelial cells and other immune cells in the FRT to regulate a spectrum of immune functions specific to each site in the FRT. At the center of immune protection, epithelial cells that line the FRT provide a level of protection not previously recognized. Acting as a physical barrier, these cells are sentinels that function as an integral part of the innate and adaptive immune systems by recognizing and protecting against bacterial and viral challenges as well as signaling the recruitment and activation of underlying immune protection when pathogenic challenge exceeds their protective capacity. Along with CD4 + T cell, macrophage, DC, epithelial cell, and fibroblast responses are uniquely programmed to respond to hormone changes so that immune coverage occurs in a way that confers continuous protection. As shown in Fig. 17 , this review indicates that epithelial, fibroblast and immune cell functions are complicated by their mutual interactions and the recognition that sex hormones alter cell-cell communication in the FRT to regulate innate and adaptive immune protection. As discussed previously, 5 aspects of the innate, humoral, and cell-mediated immune systems are suppressed by sex hormones to optimize conditions for procreation. Suppression occurs in the upper (Fallopian tubes, uterus, endocervix) and lower (ectocervix and vagina) FRT, and coincides with the recruitment of potentially infectable cells and upregulation of coreceptors involved in pathogen uptake. The studies presented suggest that by understanding the ways in which sex hormones regulate epithelial, fibroblast, and immune cell function in the FRT, new avenues may be identified both to protect against potential pathogens and to enhance the quality of women’s reproductive health.