Mouse estrous cycle regulation of vaginal versus uterine cytokines, chemokines, α-/β-defensins and TLRs

Abstract

This study investigates the cyclic changes in innate immunity in the female reproductive tract (FRT) of mice during the estrous cycle. By examining uterine and vaginal tissues and secretions we show that innate immunity varies with the stage of the estrous cycle and site in the FRT. Secretions from the uterine lumen contained cytokines and chemokines that were significantly higher at proestrus and estrus relative to that measured at diestrus. In contrast, analysis of vaginal secretions indicated that only IL-1β and CXCL1/mouse KC changed during the cycle, with highest levels measured at diestrus and estrus. In contrast, vaginal α-defensin 2 and β-defensins 1–4 mRNA levels peaked at proestrus and estrus and are expressed 1–4 logs greater than that seen in the uterus. These studies further indicate that TLR5 and TLR12 in the uterus, and TLR1, TLR2, TLR5 and TLR13 in the vagina varies with stage of the estrous cycle, with some peaking at proestrus/estrus and others at diestrus. Overall, these studies indicate that innate immune parameters in the uterus and vagina are separate and discrete, and regulated precisely during the estrous cycle.

Introduction

The innate immune system has evolved to protect the host against a wide range of pathogenic organisms. The relatively non-specific mechanisms of innate immunity in the female reproductive tract (FRT) include a physical mucus lining and epithelial barrier, in addition to the secretion of cytokine and antimicrobial peptides (see review Hickey et al.¹). The production of cytokines and defensins provides a dual function in coordinating the recruitment and activation of immune cells, such as dendritic cells (DCs), but many also possess direct antimicrobial properties that mediate the killing or neutralization of pathogens, including bacteria and viruses in the FRT through direct or indirect mechanisms (see reviews by Ganz et al., Oppenheim et al. and Ouellette et al.^2–4). The activation of the innate immune system involves the stimulation of pattern recognition receptors (PRRs), such as membrane bound TLRs and cytoplasmic NOD-like receptors (NLRs) by conserved pathogen-associated molecular patterns (PAMPs) present on many microorganisms (see review by Akira et al.⁵). Previously, we showed that isolated human and mouse uterine epithelial cells, when directly stimulated by TLR ligands in vitro, secrete a wide variety of antimicrobial peptides, including chemokines and defensins.^6–9

The FRT is unique in that changes in uterine and vaginal physiology are cyclically controlled by the sex steroid hormones estradiol and progesterone. In addition to mediating physiological changes, we have demonstrated in vitro that estradiol treatment directly alters the constitutive and TLR-mediated secretion of cytokines and antimicrobials from polarized human and mouse uterine epithelial cells in a molecule-specific manner.^10–14 For example, estradiol treatment of uterine epithelial cells enhances production of human β-defensin 2 (HBD2) and secretory leukocyte protease inhibitor (SLPI), but inhibits the secretion of CCL20/MIP3α both constitutively and following TLR stimulation.

Evidence that innate immunity varies with the stage of the reproductive cycle has been reported previously. For example, in humans, Keller et al. identified a mid-cycle drop in the concentration of cytokines, chemokines, antimicrobial peptides and Abs in cervicovaginal lavages.¹⁵ Antimicrobial peptides measured that declined included SLPI, HBD2 and human neutrophil peptides 1–3. The relationship between sex hormone and antimicrobial peptide secretion may have direct implications for a cyclic susceptibility to pathogenic organisms.¹⁶ Understanding changes in innate immunity during the reproductive cycle is important in defining the factors that prevent entry of sexually-transmitted pathogens that cause pathology and disease. Recently, Kumar et al. reported that SLPI expression differs in human vaginal, cervical and endometrial tissue, with highest expression in the endocervix as observed by immunohistochemisty and quantitative real time PCR mRNA analysis. Furthermore, SLPI production is regulated differentially by hormonal treatment at each site. Postmenopausal women receiving hormonal treatment, i.e. combined oral estrogen/ progesterone or topical intravaginal estrogen, observed attenuated SPLI protein localized in the vagina and ectocervix, but increased in the endocervix compared with women not receiving hormonal treatment.¹⁷ The coordination of multiple innate immune parameters in the upper and lower FRT, and how these change throughout the reproductive cycle in vivo is unknown.

Using BALB/c female mice, we examined the expression of a wide variety of innate immune molecules simultaneously, including cytokines, chemokines, defen-sins and TLRs. Similar to the human, the mouse FRT undergoes structural changes during the estrous cycle. The estrogen-dominant proestrus and estrus stages are comparable to the human follicular/proliferative phase, while the progesterone-dominant metestrus and diestrus stages are analogous to the secretory stage seen in humans. The present study demonstrates for the first time the normal cyclic changes in multiple innate immune paramenters in the upper and lower FRT during the estrous cycle. Furthermore, we show that the regulation of innate immunity in the vagina and uterus are regulated independently. To the best of our knowledge, our in vivo findings are the first to show that innate immunity in the murine FRT is compartmentalized and hormonally controlled during the estrous cycle.

Materials and methods

Mice

Sexually-mature virgin BALB/c female mice were obtained from the National Cancer Institute colony at Charles River Laboratories. Mice were maintained in a constant-temperature room with Exed 12 h light/dark intervals and allowed food and water ad libitum. All procedures involving animals were conducted after approval of the Dartmouth College Institutional Animal Care and Use Committee.

Determining the stage of estrous cycle

Daily vaginal smears were collected through gentle lavage with 50 μl sterile PBS using a sterile 200 μl pipette tip inserted 3 mm into the vagina. Stage of estrous cycle was determined by examining the proportion and morphology of leukocytes and epithelial cells present under 40 × objective light microscope as previously described.¹⁷ Briefly, proestrus was characterized by the primary presence of nucleated epithelial cells, estrous was characterized by the primary presence of stratified squamous epithelial cells (non-nucleated) and diestrus was characterized but the primary presence of leukocytes. Animals were followed for a minimum of three cycles to establish that each animal had normal 4–5 day cycles. Animals were subsequently sacrificed during proestrus, estrus or diestrus (day 1 or 2 diestrus). Animals who showed prolonged diestrus were not used in this study; at least 10 mice per estrous stage were used for analysis. Owing to the short transient nature of metestrus, this phase was not examined in this study.

Luminex cytokine and chemokine analysis

Vaginal and uterine lavages were collected in 50 μl of sterile PBS. Briefly, vaginal lavages were collected as described above for smears. For uterine fluid collection, upon sacrifice, the entire upper FRT was removed, rinsed repeatedly to remove blood after which uterine horns were cut at the level of the uterine corpus. The ovary and oviducts were removed, after which each uterine lumen was perfused with 50 μl of PBS. Samples were centrifuged at 10,000 g for 5 min and supernates frozen at −80°C. Luminex analysis was used to determine the concentration of cytokines (IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-9, IL-10, IL-12 (p40), IL-12 (p70), IL-13, IL-17, GM-CSF, G-CSF, TNF-α and IFN-γ) and chemokines (CCL2/MCP-1, CCL3/MIP-1α, CCL4/MIP-1β, CCL5/Rantes, CCL11/Eotaxin and CXCL1/ Mouse KC) in reproductive secretions. Washes from two mice were combined and a total of three individual wells for each stage and site (vaginal and uterine) were assayed individually. Cytokine/chemokine concentrations (pg/ml) were measured using a Bio-Rad Mouse 23-plex kit (Bio-Rad, Hercules, CA, USA), as it is ideally suited to measure multiple cytokines from one sample. All assays were carried out in a 96-well filtration plate (Millipore, Billerica, MA, USA) at room temperature (20–22°C) and protected from light as per the manufacturer's instructions. Calibration curves from recombinant cytokine/chemokine standards were prepared for the 8-point standard dilution set with 4-fold dilution steps in sterile PBS. The plate was analyzed using a Bio-Rad Bio-Plex 200 Array Reader. For data analysis and to calculate cytokine concentrations in samples, Bio-Plex Manager software's five-parameter logistic curve fitting method was used. The amount collected in 50 μl wash was calculated and results are expressed as amount in pg per animal. Furthermore, we determined the percentage (%) change at each stage compared with the mean concentration at proestrus.

Real-time PCR analysis of mRNA

The expression of TLRs and α-/β-defensin mRNA in whole uterine and vaginal tissues was determined by real-time PCR. Tissues from individual mice were homogenized using the PowerGen 125 tissue homogenizer (Fisher ScientiEc Inc., Pittsburgh, PA, USA) and RNA isolated using RNeasy® mini kit as per the manufacturer's instructions (Qiagen, Valencia, CA, USA). Following reverse transcription of 400 ng of RNA, the relative mRNA expression compared to β-actin house keeping gene was determined for TLRs (TLR1, TLR2, TLR3, TLR4, TLR5, TLR6, TLR7, TLR8, TLR9, TLR11, TLR12 and TLR13) and defensins (adefensin 1, α-defensin 2, α-defensin 5, β-defensin 1, bdefensin 2, β-defensin 3, β-defensin 4) (see Table 1 for Applied Biosystems primer information). Gene expression was measured using the 5’ Fexpression was measured in real-time quantitative PCR and Taqman chemistry on the ABI 7700 Prism real-time PCR instrument (ABI, Foster City, CA, USA). The cycle parameters were as follows: 95°C 12 min for 1 cycle (95°C 20 s, 60°C 1 min) for 40 cycles. Analysis was carried out using the sequence detection software supplied with the ABI 7700. The software calculates the threshold cycle (Ct) for each reaction and this was used to quantify the amount of starting template in the reaction. β-Actin RNA was used as an internal standard to control for variability in amplification due to differences in initial RNA concentrations. The level of the cytokine mRNA, relative to β-actin mRNA, was calculated using the following formula: relative mRNA expression = 2^{−(Ct of cytokine − Ct of βactin)} × 10¹⁰, where Ct is on the threshold cycle value.¹⁸

Table 1.

Real time PCR primers were obtained from Applied Biosciences using TaqMan® expression assay pre-designed gene primer and probe sets. The table shows gene target and Applied Biosystems order identification. The relative expression of each gene is normalized to Beta Actin Control.

Antimicrobials gene	Applied Biosystems #	Toll-like receptors gene	Applied Biosystems #
α-Defensin 1	Mm00655851_gH	TLR1	Mm00446095_m1
α-Defensin 2	Mm00651548_g1	TLR2	Mm00440346_m1
α-Defensin 5	Mm00432803_m1	TLR3	Mm00446577_g1
β-Defensin 1	Mm00657074_m1	TLR4	Mm00445274_m1
β-Defensin 2	Mm0161469_m1	TLR5	Mm00546288_s1
β Defensin 3	Mm00731768_m1	TLR6	Mm01208943_s1
β-Defensin 4	Mm00441527_m1	TLR7	Mm00446590_m1
Beta Actin (Control)	Mm00607939_s1	TLR8	Mm01157262_m1
		TLR9	Mm00446193_m1
		TLR11	Mm01701924_s1
		TLR12	Mm01180208_s1
		TLR13	Mm0123318_m1

Statistical analysis

For tissue masses and mRNA analysis, samples were collected from individual animals at different stages of the estrous cycle; a total 10 mice per stage was determined. For Luminex analysis, we determined cytokine/ chemokine concentration (pg) from three separate wells for each stage of the estrous cycle. Each well contained combined vaginal or uterine washes from wo individual mice. Data are presented as the mean SEM. ANOVA followed by Bonferroni's post-test was used to examine the differences between each stage of the estrous cycle. When directly comparing vaginal and uterine tissue results, an unpaired t-test was used. The significance level was set at P < 0.05 for all tests. Statistical analysis was performed using GraphPad Prism version 4.00 (GraphPad Software, San Diego, CA, USA).

Results

Cytokine and chemokine concentrations in vaginal and uterine lavages change throughout the estrous cycle

Cytokines are signaling molecules secreted by immune and epithelial cells to mediate tissue remodeling, cellular migration and immunity in the FRT. As such, changes in secreted cytokines play significant roles in innate immune protection within FRT secretions. Figure 1 shows the percentage change in vaginal cytokines secreted at each stage of the estrous cycle compared with the mean values measured at proestrus. Changes in cytokines in vaginal secretions were observed in a molecule-, stage-specific manner (Figure 1A–C). For example, significant increases in IL-1β, IL-12p40 (Figure 1A), GM-CSF (Figure 1B) and CCL4/MIP-1β (Figure 1C) were observed at dies-trus, whilst CXCL1/Mouse KC increased at estrus (Figure 1C) compared with proestrus. In contrast, changes in cytokines present in uterine lavages were synchronized over estrous cycle compared with that seen in the vagina. As seen in Figure 1D–F, significant decreases by 50% in interleukins and greater than 75% in other cytokines and chemokines were observed at diestrus relative to that seen at proestrus. One exception was IL-12p40, which decreased by 30% at diestrus and 50% at estrus (Figure 1D). For specific cytokine and chemokine concentrations (pg per secretion) at each stage and statistical analysis, please refer to supplementary Figure 1.

Figure 1.

Changes in cytokines present in vaginal and uterine secretions. Data represent percent change (%) in cytokine/chemokine concentrations determined by Luminex analysis in vaginal (panels A, B & C) and uterine (panels D, E & F) secretions at each stage of the reproductive cycle compared to proestrus. The percent change in cytokine and chemokine proteins are secreted in the vaginal tract in a protein specific/cycle dependent manner. Uterine lavages show a 25–50% reduction in all proteins analyzed at diestrus compared to proestrus. Data represents combined results from 3 separate wells containing washes from two individual animals (n = 3) and shown as mean ± std err. For intra-estrous cycle statistical analysis refer to supplementary table.

Comparison of interleukin cytokine levels in vaginal and uterine secretions

To compare the distribution of interleukins in the mouse vagina and uterus, we measured and directly compared the concentrations of vaginal and uterine cytokines and chemokines at each stage of the estrous cycle. As seen in Figure 2, levels of 6 out of the 11 interleukins analyzed were significantly higher in uterine lavages relative to that seen in vaginal lavages (IL-2, IL-3, IL-6, IL12-p40, IL-13 and IL-17). Of these, with the exception of IL12-p40, which was highest at proestrus and diestrus, interleukins were measured to be highest at proestrus and estrus. In contrast, in the vagina only 2 out of the 11 interleukins were significantly greater in vaginal lavages compared with uterine lavages (IL-1β and IL-12p70). Interestingly, vaginal IL-1β levels rose by approximately two logs at diestrus, when compared with that measured in the vagina at proestrus and estrus. In contrast, vaginal IL-12p70 levels at estrus and diestrus, when compared with that seen in the uterus, were 2- and 10-fold higher respectively. Overall, these findings indicate that in the intact mouse, the uterus and vagina function as two discrete endocrine-responsive immune compartments, each with a unique profile of secreted interleukin cytokines.

Figure 2.

Comparison of vaginal (black bars) and uterine (grey bars) levels of interleukin cytokines (pg/animal) at various stages of the estrous cycle (proestrus, estrus and diestrus). Data represents combined results from 3 separates wells containing washes from two individual animals in each well (total wells 3), and shown as mean ± std err. *P < 0.05, **P < 0.01, ***P < 0.001 significant difference between vaginal and uterine tissue at each stage of the estrous cycle using t-test.

Comparison of cytokine and chemokine levels in vaginal and uterine secretions

Figure 3 indicates that, similar to interleukin levels in FRT secretions, the cytokines G-CSF, GM-CSF, TNF-α and IFN-γ, and most chemokines, are significantly higher in uterine secretions when compared with vaginal lavages. Eight out of 10 cytokines/chemokines examined were higher in the uterus compared with those measured in the vagina (CCL2, CCL4, CCL5, CCL11, GM-CSF, G-CSF, TNF-α and IFN-γ). Of those chemokines assayed, only CCL3 was enhanced in the vagina at diestrus, relative to that seen in the uterus. Interestingly, the cytokine GM-CSF had a unique profile in that uterine levels at proestrus and estrus were higher than that measured in the vagina but then switched with levels higher in the vagina at diestrus. Overall, cytokines and chemokines were present at higher concentrations in the uterus compared with the vagina.

Figure 3.

Comparison of vaginal (black bars) and uterine (grey bars) levels of cytokines (GM-CSF, G-CSF, TNF-α & IFN-γ) and chemokines (pg/animal) at various stages of the estrous cycle (proestrus, estrus and diestrus). Data represent combined results from 3 separate wells containing washes from two individual animals in each well (total wells 3), and shown as mean ± std err. *P < 0.05, **P < 0.01, ***P < 0.001 significant difference between vaginal and uterine tissue at each stage of the estrous cycle using t-test.

Cyclic changes in α- and β-defensin tissue mRNA expression

Defensins are broad-spectrum antimicrobial peptides present in tissues and secretions of the FRT. Given the limited availability of mouse ELISA assays for measuring secreted proteins, we undertook studies to measure the levels of defensin mRNA of FRT tissues. Using real-time PCR analysis, the relative expression of α-defensins 1, 2 and 5, and β-defensins 1, 2, 3 and 4 mRNA in vaginal (Figure 4A) and uterine (Figure 4B) tissue was determined. As seen in the left panels of Figures 4A and B, expression levels for adefensins 1, 2 and 5 mRNA in the vagina were the same as those measured in the uterus. In contrast, β-defensins (right panels) were expressed at higher levels in vaginal tissues with enhanced β-defensin 1 (2-logs), β-defensin 3 (3-logs) and β-defensin 4 (2-logs) measured when compared with uterine tissues. Within each tissue significant changes in α- and β-defensin mRNA expression were observed during the estrous cycle. In the vagina, for example, α-defensin 2, and β-defensins 1, 2, 3 and 4 were significantly elevated at proestrus and/or estrus (Figure 4A) relative to diestrus. In uterine tissues, α-defensin 1 was enhanced at diestrus compared with proestrus, whereas β-defensins 2 and 3 were significantly inhibited at diestrus compared with proestrus and estrus respectively (Figure 4B).

Figure 4.

Changes in vaginal (panel A) and uterine (panel B) tissue mRNA expression of α- and β-defensins normalized to β-actin at proestrus (black bars), estrus (white bars) and diestrus (grey bars). Data represent combined results from 10 individual animal for each stage of the estrous cycle and shown as mean ± std err. *P < 0.05, **P < 0.01, ***P < 0.001 1-way ANOVA, Bonferroni post hoc.

Cyclic changes in vaginal and uterine tissue TLR mRNA expression

The production of cytokines, chemokines and defensins are, in part, mediated through the activation of TLRs by wide range of PAMPs, including endotoxin, bacterial DNA and viral RNA. Recognizing that murine reproductive tract tissues express mRNA for mouse TLRs 1–9,¹⁹ studies were undertaken to measure the relative expression of TLRs 1–9 and mouse TLRs 11–13 mRNA in vaginal and uterine tissue during the estrous cycle. Figure 5 shows that while all mouse TLRs are present in uterine and vaginal tissues. TLR 2, TLR4 and TLR5 mRNA were expressed predominately in both the upper and lower FRT compared with the other TLRs (Figure 5A and B). Interestingly, with the exception of TLR2 and TLR5, which were expressed at greater than twofold increase in the vagina compared with the uterus, similar levels of TLR expression were found in both tissues for all other TLR measured. In contrast, TLR13, a murine-specific receptor, was higher in vaginal tissue than in uterine tissue, whereas TLR12 was predominately expressed in the uterus, relative to that seen in the vagina. As a part of these studies we found that stage of the estrous cycle influenced TLR expression for some but not all TLRs. As seen in Figure 5A, vaginal TLR1 and TLR5 levels were highest at estrus relative to proestrus, whilst TLR2 and TLR13 peaked at diestrus relative to proestrus and estrus. In contrast, as shown in Figure 5B, uterine TLR5 was lowest at diestrus compared with proestrus, whilst TLR12 was highest at diestrus compared with proestrus.

Figure 5.

Changes in TLR expression 1–9 and 11–13 at proestrus (black bars), estrus (white bars) and diestrus (grey bars) in vaginal (panel A) and uterine (panel B) tissue. Data represent combined results from 10 individual animal for each stage of the estrous cycle and shown as mean ± std err. * p < 0.05, **p < 0.01 1-way ANOVA, Bonferroni post hoc.

Discussion

To the best of our knowledge, this is the first study to simultaneously characterize multiple innate immune parameters across the three main stages of the mouse estrous cycle. This work quantitatively compares the lower and upper FRT of individual mice, and thereby demonstrates, for the first time, the strict compartmentalization of innate immunity. In both the vagina and the uterus, cytokines and defensins are produced in an estrous stage- and molecule-specific manner. The luminal secretion of uterine cytokines and β-defensins are significantly higher at proestrus and estrus compared with diestrus. Of those cytokines analyzed in vaginal secretions, only IL-1b and CXCL1/mouse KC changed significantly during the estrous cycle, with the highest levels measured at diestrus and estrus respectively. In contrast, vaginal α-defensin 2, and β-defensins 1–4 peaked at proestrus and/or estrus. We found that levels of β-defensins in the vagina are 1–4 logs greater than that seen in the uterus. These studies further indicate that TLR5 and TLR12 in uterine tissue and TLR1, TLR2, TLR5 and TLR13 in vaginal tissue vary during the stage of the estrous cycle, with TLR5 being greatest during the estrogen dominant phase in the uterus and vagina.

The striking observation of this study is the expression of uterine cytokines, β-defensins and TLRs during proestrus and estrus, followed by a post-ovulatory decline at diestrus. This coordinated cycle-dependent change in innate immunity facilitates endometrial receptivity, fertilization and implantation. For example, a number of cytokines mediate leukocyte recruitment, angiogenesis, proliferation and differentiation (see review by Kayisli et al.²⁰). These include TNF-α, IL-9, IL-12p40, IL-13, IL-6, CCL2/MCP-1, CXCL1/mouse KC, and CXCL11/Eotaxin, which we show are expressed predominately during the proestrus/estrus stages, but not the diestrus stage of the cycle. Exceeding all other cytokines was the production of G-CSF during proestrus, which is secreted locally in the upper FRT and not derived from serum.²¹ G-CSF is an important cytokine-promoting endometrial thickening cytokine prior to ovulation and is integral to reproduction.^22,23 The dominance of cytokines, specifically G-CSF, at the beginning of endometrial proliferation supports a cyclic role for uterine cytokines in preparing the uterine environment for reproduction. Uterine DCs have been shown to be essential for successful embryo implantation and decidualization, but are not integral in endometrial proliferation.^24,25 This functional difference correlates with an altered secretory pattern of the potent DC chemoattractant chemokine CCL20. Unlike the chemokines examined in this study, CCL20 is present in uterine fluid in increasing concentrations over the estrous cycle peaking at diestrus (Hickey et al. unpublished data). This suggests that, whereas some uterine cytokines are suppressed at diestrus, other specific cytokines may be increased.

Coordinating with cytokine preparation of the endometrial lining is the production of potent antimicrobial peptides, including defensins and chemokines. The presence of antimicrobials may play a role in cleansing the upper FRT of potential pathogens that might hinder implantation and pregnancy. We believe that the overall mid-cycle suppression of innate immunity in accordance with ovulation seen here is essential to facilitate implantation of an immunologically-distinct fetus. Freshly isolated mouse and human uterine epithelial cells secrete a variety of cytokines and antimicrobial molecules in vitro both constitutively and in response to hormone treatment. Furthermore, freshly isolated uterine epithelial cells collected at diestrus secrete a lower concentration of cytokines in vitro than cells isolated at proestrus and estrus suggesting long-lasting cyclic priming.²⁶ In addition, progesterone-mediated physical remodeling/absorption of uterine epithelium and stroma during diestrus may also play a direct role in dampening immunity in vivo.²⁷

In direct contrast with uterine responses and with the exception of CXCL1 and IL-1β, the secretion of cytokines in the vaginal tract is not regulated by the estrous cycle. However, the fact that CXCL1 and IL-1β are cyclically regulated is not surprising. CXCL1 is a potent chemokine that mediates the migration of neutrophils and in mice acts as a functional substitute for human IL-8 along with CXCL2 (MIP-2) and CXCL3 (GROγ).^28–30 In our work, the peak expression of CXCL1 at estrus corresponds with neutrophil accumulation in vaginal tissue at this time which immediately precedes neutrophil migration into the vaginal lumen at diestrus. The peak secretion of CXCL1 during estrus was confirmed using ELISA analysis (unpublished observation). In other studies CXCL2 has shown to be also secreted prior to diestrus and directly controls 50% of neutrophil migration in the mouse vaginal tract.³¹ This suggests that neutrophil migration within vaginal tissue may be a coordinated event with other CXC chemokines. The direct role of CXCL1 on mediating vaginal neutrophil migration has not been studied. The influx of neutrophils may further mediate the concentration of IL-1β in vaginal secretions. IL-1β is an important molecule mediating many innate and adaptive immune responses; however, its specific role in the vaginal tract is unclear. In the mouse, we show that IL-1β is present at diestrus, corresponding with the accumulation of neutrophils in the vaginal lumen which mimics human cervicovaginal lavages where a direct correlation between neutro-phil numbers and IL-1β concentrations is observed.³² Currently, the direct association between neutrophils and IL-1β remains unclear. Our findings show that estrous cycle regulation of cytokines in the FRT is individualized, potentially to facilitate a specific physiological function.

Unlike cytokines, vaginal defensins (α- and β-) peak during the estradiol-dominant phases of the cycle (protestrus and estrus). In laboratory rodents, commensal bacteria, including Staphylococcus aureus, Enterococcus species, Lactobacillus species, enteric bacilli anaerobic species and coagulase-negative staphylococci, are elevated significantly at proestrus and estrus compared with metestrus and diestrus.^33,34 Cyclic variations in the vaginal microbiome are also observed in humans with increased lactobacilli numbers and adherence occurring when estradiol is high.^35,36 We show that the production of vaginal β-defensins exceeds that observed in the uterus by several logs at all stages of the estrous cycle. As commensals are confined predominately to the lower FRT correlating with higher vaginal β-defensins compared with the uterus, our findings suggest that antimicrobial peptides are produced locally in relation to changes in vaginal commensals. Whether this is a direct response to commensal growth, which in turn limits bacterial growth, remains to be determined. What is interesting is that along with increases in vaginal β-defensins we measured increases in expression of vaginal TLRs, namely TLR2 and TLR5, compared with the uterus. This finding suggests that commensal growth may lead to TLR responses that in turn up-regulate antimicrobial production in the vaginal tract. The influence of commensals on regulating innate immunity, including cytokine, antimicrobials and TLR expression, in the FRT requires further study.

The cyclic change in innate immunity observed in this study may be an important mechanism for limiting infection by sexually-acquired pathogens. In laboratory rodent infection models, animals must either be challenged at diestrus or following progesterone treatment to sustain genital infection with Chlamydia trachomatis and human papilloma virus. Conversely, estradiol treatment is protective against the development of these infections.^37,38 The cyclic change in the production of antimicrobial peptides in the upper and lower FRT reported in the present study correlates with the known susceptibility of animal models to sexually transmitted infections. Antimicrobials, such as defensins and chemokines, directly inhibit potential genital tract pathogens. Human α-defensins, β-defensins and antimicrobial chemokines inhibit HIV, human papilloma virus, and Chlamydia infection and/or replication.³⁹ We hypothesize that the beneficial role of estradiol in protecting against STIs in vivo may, in part, be mediated through cyclic changes in antimicrobials—mainly defen-sins—that peak in both the upper and lower tracts during the estradiol-dominant phases of the estrous cycle. Although in mice, estradiol and the estrogen-dominant phases of the estrous cycle confer a level of protection against many STIs, some infections, such as Neisseria gonorrhoeae, in fact require estrogen dominance to establish infection.⁴⁰ Furthermore, the influence of the reproductive cycle and hormones to mediate protection is species-specific. For example, infections of female guinea pigs with Chlamydia infections are enhanced with estradiol treatment leading to greater bacterial load, longer infection times and ascension of infection to the upper FRT.⁴¹ In humans, women have been shown to be more resistant to STIs during the progesterone-dominant phase of cycle suggesting that, like swine, estradiol, but not progesterone, facilitates infection. We do acknowledge species differences and therefore cannot extrapolate the finding shown here in mice directly to humans. However, this work does suggests that changes in innate immunity influenced by cyclic changes in sex hormones may mediate susceptibility to STIs throughout the reproductive cycle.

Overall, these studies provide an insight into the complexity of the innate immune system in the murine FRT and demonstrate that sex hormone-mediated changes during the estrous cycle selectively regulate components of innate immune protection in both the uterus and vagina. Our findings show that uterine innate immune parameters are up-regulated prior to ovulation primarily to facilitate reproduction, whereas vaginal responses are regulated to provide continuous broad-spectrum protection against microorganisms with enhancement at the time of ovulation, when mating is most likely to occur. Our work demonstrates for the first time that the levels of uterine and vaginal cytokines and chemokines, in addition to defensins and TLRs, are regulated precisely by the stage of the estrous cycle in vivo, with each site showing independent compartmentalized expression of innate components. Further studies are needed to unravel the complex inter-relationships between sex hormones, commensals and the innate immune system in the rodent and human FRT.