Ovariectomized mice and postmenopausal women exhibit analogous loss of genital epithelial integrity.

ABSTRACT

Roughly half of all postmenopausal women are affected by the genitourinary syndrome of menopause (GSM). Symptoms of GSM, including vaginal irritation and dyspareunia, occur as reduced estrogen (E) production elicits loss of elasticity and other changes in genital tract tissue. While the use of the injectable contraceptive depot-medroxyprogesterone acetate (DMPA) likewise lowers serum E concentrations in reproductive age women and is associated with decreased genital levels of the cell-cell adhesion molecules desmoglein-1 (DSG1) and desmocollin-1 (DSC1) and impaired genital epithelial barrier function, the relevance of these findings to women in menopause is uncertain. Exploring the impact of menopause on genital epithelial integrity herein, we detected significantly lower levels of DSG1 and DSC1 in ectocervical tissue from menopausal and postmenopausal vs premenopausal women. Using ovariectomized (OVX) mice as a menopause model, we comparably saw significantly lower vaginal tissue levels of DSG1 and DSC1 in OVX mice vs. mice in estrus. Compared to estrus-stage mice and E-treated OVX mice, DMPA-treated ovary-intact mice and OVX mice also exhibited significantly reduced genital epithelial barrier function, greater susceptibility to genital herpes simplex virus type 2 infection, and delayed clearance of genital Chlamydia trachomatis infection. Current studies thus identify analogous loss of genital epithelial integrity in OVX mice and menopausal and postmenopausal women. By showing that loss of genital epithelial integrity is associated with increased mouse susceptibility to bacterial and viral pathogens, our findings also prioritize the need to resolve if reduced genital epithelial integrity in postmenopausal women is a significant risk factor for genital infection.

Introduction

The injectable progestin depot-medroxyprogesterone acetate (DMPA) provides contraception for reproductive-age women primarily via a disruption of the hypothalamic-pituitary-ovarian (HPO) axis that inhibits ovarian follicle maturation and synthesis and secretion of ovarian estradiol.^1,2 As evidence of DMPA-mediated HPO disruption, median estradiol levels in serum collected weekly from women after a single 150 mg injection of DMPA was approximately 15 pg/mL, values that approach those measured in postmenopausal women.^3,4 Previous reports also show the hypoestrogenemia induced by DMPA in reproductive age women is associated with decreased genital levels of the cell-cell adhesion molecules desmoglein-1 (DSG1) and desmocollin-1 (DSC1) and impaired genital epithelial barrier function.⁵ Likewise, compared to estrus-stage mice, DMPA-treated mice displayed lower genital DSG1 levels and greater genital mucosal tissue permeability to low molecular mass molecules and activated leukocytes.^5,6 Consistent with this weakening of genital epithelial barrier function, DMPA-treated wild type mice and humanized mice were uniformly susceptible to genital infection with herpes simplex virus type 2 (HSV-2) and human immunodeficiency virus type 1 (HIV-1), respectively, whereas estrus-stage mice and ovary-intact mice administered DMPA and a conjugated estrogens (E) cream were recalcitrant to these infections.^5–7 While the above results indicated that DMPA disrupts the HPO axis and diminishes E-regulated promotion of genital epithelial integrity and barrier function,⁸ the relevance of these data to the reproductive health of postmenopausal women is relatively unexplored.

Greater clarity on this subject is essential however as approximately 1.2 billion women worldwide will be menopausal or postmenopausal by the year 2030.⁹ Half of these women will likely also be affected by a constellation of genital (e.g., dryness and burning), sexual (e.g., dyspareunia and lack of lubrication), and urinary (e.g., dysuria, urgency, and recurrent urinary tract infections) signs and symptoms termed the genitourinary syndrome of menopause (GSM).¹⁰ These signs and symptoms result from lower levels of serum E that produce epithelial thinning, loss of collagen, hyaluronic acid, and elastin, increased connective tissue density, and other genital tract tissue changes.^11,12 These changes diminish vaginal lubrication, reduce tissue elasticity, narrow the vagina, and promote vaginal bleeding, tears, and fissures.¹³ The risk of urinary tract infection (UTI) and recurrent UTI are also increased by menopause, and well-established risk factors for UTI in older women include urinary incontinence and urinary retention.^14,15 However, as prior research identified DMPA induces hypoestrogenemia that reduces genital epithelial integrity and barrier function, it is also possible the hypoestrogenic state associated with menopause is an important risk factor for urinary or genital infection in postmenopausal women. In the current study, we explored this possibility by comparing ectocervical levels of the desmosomal cadherins DSG1 and DSC1 in premenopausal vs. menopausal and postmenopausal women. Using ovariectomized (OVX) mice to model the hypoestrogenic state associated with menopause, we also explored the impact that lost ovarian function has on genital epithelial integrity and barrier function and susceptibility to viral and bacterial genital pathogens.

Materials and methods

Mice and in vivo procedures in mice

All murine studies were approved by the Stanford University Administrative Panel on Animal Care prior to study initiation. For these studies, 8–10-week-old C57BL/6 J female mice were acquired from the Jackson Laboratories (Bar Harbor, ME, USA), and, where indicated, sedated with 1.8 mg ketamine hydrochloride (JHP Pharmaceuticals, Rochester, MI) and 0.18 mg xylazine (Lloyd Laboratories, Shenandoah, IA). Also, as indicated, ovary-intact mice (hereafter termed intact mice) were subcutaneously (sc) injected with 0.3 mg of DMPA or diagnosed in the estrus stage of the estrous cycle by light microscope examination of crystal violet-stained cells obtained by vaginal lavage.¹⁶ Where specified, mice ovariectomized 14 days earlier were sedated daily for 3 days to intravaginally (ivag) administer placebo or the conjugated estrogens cream Premarin® (Wyeth Pharmaceuticals, Philadelphia, PA).

Assessment of genital permeability

To evaluate genital mucosal permeability to low molecular mass (LMM) molecules, sedated mice were ivag administered a 10 µL PBS solution containing 90 µM of dextran Texas–Red (DR) (70 KDa) and 10 mM of Lucifer Yellow CH, lithium salt (LY) (457 Da) (Life Technologies, Carlsbad CA). At indicated time points after treatment, mice were euthanized and vaginal tissue fixed for 24 h at 4°C in PBS containing 4% methanol-free formaldehyde. For confocal microscopy analysis of fluorescent molecule penetration of the vaginal mucosa, tissue was agarose-embedded, sliced into 200 µm sections, counter-stained with 4,6-diamidino-2-phenylindole (DAPI) (Sigma–Aldrich, St. Louis, MO), and embedded in VECTASHIELD® mounting medium (Vector Laboratories, Burlingame, CA). Data were obtained using a Nikon A1 confocal laser microscope and images were acquired by sequential scanning. LY penetration into epithelial and stromal tissue was quantified using Image J software.^17,18

RNA isolation and quantitative real-time PCR (qPCR)

As indicated, vaginas from euthanized mice were immediately immersed in 1 mL of RNAlater® (Qiagen, Hilden, Germany) and stored for 24 h at 4ºC prior to longer-term storage at −80ºC. RNA was isolated with RNeasy Lipid Tissue kits (Qiagen) and re-suspended in nuclease-free water. A SpectraMax® Abs Plus absorbance microplate reader (Molecular Devices, San Jose, CA) quantified RNA concentrations (all samples had 260/280 and 260/230 ratios >1.80). To quantify gene expression, cDNA was generated from 500 ng of RNA using SuperScript ™ IV VILO™ master mix with ezDNase™enzyme (ThermoFisher Scientific, Rockford, IL) and a MiniAmp Thermal Cycler (Applied Biosystems, Foster City, CA). qPCR was performed using TaqMan™ Universal PCR Master Mix and QuantStudio® 3 Real-Time PCR Systems (Applied Biosystems). Relative desmocollin-1 (Mm00809994_s1) and desmoglein-1a (Mm00809994_s1) gene expression was determined using ∆C_T methodology with pyruvate carboxylase (Mm00500992_m1) as the housekeeping reference gene (all primers Life Technologies).

Histology

Formaldehyde-fixed paraffin-embedded (FFPE) ectocervical tissue blocks from premenopausal (n = 12) (mean 35 years; range 23–38 years) and postmenopausal women (n = 11) (mean 71 years; range 54–87 years) archived in the Stanford University School of Medicine Department of Pathology were obtained with Institutional Review Board approval and 5 µm tissue sections were stained with hematoxylin and eosin (H&E). Images from all sections were acquired using the NanoZoomer 2.0-RS slide scanner (Hamamatsu Photonics KK Hamamatsu City, Japan), and genital epithelial thickness defined using NDP.view2 software (Hamamatsu Photonics KK). Murine FFPE blocks were prepared by immersing vaginal tissue from intact and OVX mice in 4%-buffered formaldehyde. Tissue was paraffin-embedded and 5 µm tissue sections stained with H&E.

Immunofluorescence staining

5 µm sections from murine FFPE vaginal tissue blocks‎ were deparaffinized by sequential immersion in xylene, ethanol, and distilled water. Antigen retrieval was performed using 10 mM citrate buffer at pH 6.0 with 0.05% Tween®20 for 20 min at 95ºC. Sections were blocked for 2 h at 4ºC with 5% normal goat serum (Abcam, Cambridge MA), and incubated at room temperature (RT) for 1 h with rabbit anti-DSG1 monoclonal antibody (mAb) (clone EPR6766(B), Abcam) or for 2 h with rat anti-DSC1 mAb (clone 772906, R&D Systems, Minneapolis, MN) (both used at 1/200 dilutions). Sections were incubated for 30 min at RT with Alexa Fluor® 488-labeled goat anti-rabbit IgG polyclonal secondary antibody or Alexa Fluor® 594-labeled goat anti-rat IgG polyclonal secondary antibody (both Abcam and both used at 1/500 dilutions), DAPI counterstained, and evaluated by Nikon A1 confocal laser microscope. Image J software was used to determine the relative expression of DSC1 or DSG1 protein.^17,18

Immunohistochemical staining

Similar to the methods used for murine tissue immunofluorescence studies, 5 µm sections from human FFPE ectocervical tissue blocks were deparaffinized for antigen retrieval. Sections were incubated with 3% hydrogen peroxide for 10 min and blocked overnight with 5% normal goat serum (for anti-DSG1 studies) or 5% normal rabbit serum (for anti-DSC1 studies). Sections were incubated 24 h at 4ºC with rabbit anti-DSG1 mAb (clone EPR6766(B), Abcam) (1:200 dilution in PBS) or 2 h at 4ºC with rat anti-DSC1 mAb (clone 772906) (R&D Systems) (1:400 dilution in SignalStain® Antibody Diluent) (Cell Signaling Technology, Beverly, MA). After primary mAb staining, sections were washed and incubated with goat anti-rabbit IgG polyclonal secondary antibody conjugated with horseradish peroxidase (HRP) (for anti-DSG1 studies) or rabbit anti-rat IgG polyclonal secondary antibody conjugated with HRP (for anti-DSC1 studies). Sections were washed and using the manufacturer’s instructions, serially incubated with SignalStain® Boost Detection and SignalStain® DAB Chromogen (both Cell Signaling Technology). Once the optimal signal was obtained, sections were counterstained with hematoxylin and coverslipped with SignalStain® mounting medium (Cell Signaling Technology). For quantifying DSC1 or DSG1 protein, images were captured using the NanoZoomer 2.0-RS slide scanner and DAB Chromogen optical density calculated using ImageJ software.

Genital infection

Sedated mice were ivag infected with 10³ plaque-forming units (PFU) of WT HSV-2 333 (provided by Dr. Robert Hendricks, University of Pittsburgh) in 10 μL of RPMI and evaluated daily post-infection. Genital pathology was assessed using a previously defined 5-point scale: 0, no pathology; 1, mild vulvar erythema; 2, moderate vulvar erythema; 3, severe vulvar erythema and perineal fur loss; 4, perineal ulceration; 5, an extension of perineal ulceration and fur loss to surrounding tissue.¹⁹ Mice were euthanatized when pathology scores were ≥3 or encephalopathic signs detected. Other sedated mice were ivag infected with 10⁴ inclusion-forming units (IFU) of Chlamydia trachomatis serovar D (ATCC® VR-885D™) in 10 µL of sucrose-phosphate-glutamate (SPG) buffer. This bacterium had been propagated on McCoy cells (ATCC® CRL-1696) and isolated elementary bodies (EB) stored at −80°C in SPG prior to genital inoculation.²⁰ Cervicovaginal lavages (CVL) were obtained at 3, 5, 7, 14, 21, 28, and 35 days post-infection (dpi) to assess Chlamydia clearance via quantification of its 16S rRNA gene levels, as previously described.²¹

Statistical considerations

All statistical analyses were performed using Prism 8 software (GraphPad, La Jolla, CA). Normal distribution was tested by D’Agostino and Pearson omnibus test or evaluation of the residuals (when group numbers were <8). To assess HSV-2-induced genital pathology and Chlamydia clearance, the area under the curves (AUC) was calculated for individual mice and AUC values were compared. Two-group samples were compared using the unpaired Student’s t-tests. For studies with more than two groups, one-way ANOVA with Dunnett’s or Tukey’s post hoc test were used for parametric and nonparametric data, respectively. For comparison of two groups in experiments that contained more than two groups, two-way ANOVA with Bonferroni’s multiple comparison test was used. P values less than 0.05 were deemed statistically significant.

Results

Ectocervical epithelial thickness and levels of DSG1 and DSC1 protein were significantly reduced in postmenopausal vs. premenopausal women

As various reports found that genital tract changes associated with the onset of menopause include genital mucosal thinning,^22–24 we began the current study by comparing mucosal thickness in ectocervical tissue from premenopausal vs. menopausal or postmenopausal women. Congruent with prior reports, we found ectocervical mucosa was significantly thinner in menopausal and postmenopausal women (Figure 1(a,b)). Also, compared to premenopausal women, menopausal and postmenopausal women displayed significantly lower levels of the desmosomal cadherins DSG1 and DSC1 (Figure 1(a,c)), cell-cell adhesion molecules known to promote integrity and diminish mechanical insult in other epithelial tissues.²⁵ These initial studies were therefore consistent with earlier reports that identified genital epithelial tissue in menopausal and postmenopausal women as more fragile and susceptible to fissures and tears.²⁶

Figure 1.

Ectocervical tissue from menopausal and postmenopausal women displayed reduced epithelial thickness and lower DSG1 and DSC1 protein levels. (a) Representative images depicting epithelial thickness and DSG1 and DSC1 protein levels in ectocervical biopsy tissue from pre- and postmenopausal women; scale bar denotes 100 µm. (b) Panel displays significantly thinner ectocervical epithelium in menopausal and postmenopausal vs. premenopausal women. (c) DSG1 and DSC1 protein levels were also significantly lower in ectocervical epithelium from menopausal and postmenopausal vs. premenopausal women. In panels (b,c), values represent results from individual woman and bars indicate means and two-tailed unpaired Student’s t-test was used for statistical comparisons. PRE, premenopausal women; POST, menopausal or postmenopausal women; DSC1, desmocollin-1; DSG1, desmoglein-1; H&E, hematoxylin and eosin; DAB, 3,3ʹ-diaminobenzidine

DMPA-treated and OVX mice display significantly reduced vaginal DSG1 and DSC1 protein levels

As a follow-up to the human ectocervical tissue studies, OVX mice were used to model the reduced levels of circulating estrogen (E) associated with the cessation of ovarian function in postmenopausal women. Comparing genital tracts from intact mice in the estrus stage of the estrous cycle, DMPA-treated intact mice, OVX mice, and OVX mice administered the conjugated estrogen cream Premarin®, these studies identified significant thinning of the vaginal epithelium in DMPA-treated and OVX mice vs. estrus-stage mice (upper panel of Figure 2(a,b)). Conversely, vaginal epithelium thickness in OVX mice was restored to levels measured in mice in the estrus stage of the estrous cycle by Premarin® treatment (upper panel of Figure 2(a,b)). There was also significantly lower vaginal epithelial Dsc1 and Dsg1a gene expression in DMPA-treated vs. estrus-stage mice (Figure 2(c)). Whereas reduction in vaginal expression of these cell-cell adhesion genes was more severe in OVX vs. DMPA-treated mice, treatment of OVX mice with the conjugated estrogen cream restored Dsc1 and Dsg1a gene expression levels to those found in estrus-stage mice (Figure 2(c)). Congruent with gene expression results, we saw significantly lower protein levels of DSG1 and DSC1 in vaginal epithelium of DMPA-treated intact mice and OXV mice vs. estrus-stage mice, and Premarin® treatment of OVX mice restored levels of these proteins to those measured in estrus-stage mice (middle and lower panels of Figure 2(a,d)). Together with our human ectocervical tissue data, these findings revealed there is analogous genital mucosal thinning and compromise of genital epithelial integrity in menopausal and postmenopausal women and OVX mice and strongly indicated that mouse ovariectomization generates significant compromise to genital epithelial barrier function.

Figure 2.

DSG1 and DSC1 protein levels in mouse vaginal epithelium were significantly reduced by DMPA administration or ovariectomization. (a) Representative images that characterize epithelial thickness and DSG1 and DSC1 protein levels in vaginal tissue from estrus-stage and DMPA-treated mice with intact ovaries and untreated and E-treated OVX mice; all images magnified at 200X; DAPI (blue); DSG1 (green); DSC1 (red); scale bars denote 100 µm. (b) Thickness of the vaginal epithelium was significantly reduced in DMPA-treated mice and OVX mice compared to estrus-stage mice and OVX mice administered an E-containing cream. (c) Reduced Dsg1a and Dsc1 gene expression in vaginal tissue from DMPA-treated and OVX mice was restored in E-treated OVX-mice to values measured in estrus-stage mice. (d) Levels of DSG1 and DSC1 protein were significantly higher in estrus-stage and E-treated OVX-mice vs. DMPA-treated intact mice and untreated OVX mice. In displayed results, 2 independent experiments with 4–6 animals per group were performed and 1-way ANOVA and Dunn’s post hoc tests used for comparisons. Bars indicate mean ± SD. DAPI, 4,6-diamidino-2-phenylindole; DMPA, depot-medroxyprogesterone acetate; (e); conjugated estrogens vaginal cream; DSC1 or Dsc1, desmocollin-1; Dsg1a, desmoglein-1α; DSG1, desmoglein-1; OVX, ovariectomized

Compromise of genital epithelial barrier function was more severe in OVX mice vs. DMPA-treated intact mice

Based on between-group differences in vaginal DSG1 and DSC1 protein levels observed in the previous set of experiments (Figure 2), we posited that genital epithelial barrier function is weaker in DMPA-treated intact mice and OVX mice vs. mice in the estrus stage of the estrous cycle or Premarin®-treated OVX mice. To test this hypothesis, LMM fluorescent molecules were ivag administered to these four groups of mice to compare the penetration of these molecules into vaginal tissue. As posited, LMM molecule entry into vaginal tissue 45 min after ivag administration was comparably increased in DMPA-treated and OVX mice vs. estrus-stage mice but restored to values measured in estrus-stage mice by Premarin® treatment of OVX mice (Figure 3(a)). As levels of vaginal DSG1 and DSC1 protein were lower in OVX vs. DMPA-treated mice (Figure 2), we next hypothesized that OVX mice display greater loss of genital epithelial barrier function than DMPA-treated intact mice. To test this hypothesis, DMPA-treated and OVX mice were used to compare fluorescent LMM molecule penetration of vaginal tissue 15, 30, or 45 min after ivag administration of these molecules. While we again observed comparable levels of fluorescence in submucosal vaginal tissue 45 min after LMM molecule administration, submucosal penetration was significantly greater in OVX vs. DMPA-treated mice when tissues were examined 15 or 30 min after ivag administration of these molecules (Figure 3(b)). Combined, these experiments revealed DMPA treatment and mouse ovariectomization impairs genital epithelial integrity and barrier function and that these effects appear more severe in OVX vs. DMPA-treated animals.

Figure 3.

Compromise of genital epithelial barrier function was greater in OVX mice vs. DMPA-treated mice. (a) Left side of panel shows representative images of vaginal tissue from estrus-stage and DMPA-treated intact mice and untreated and E-treated OVX mice obtained 45 min after mice were genitally administered 10 µL of a PBS solution containing lucifer yellow (LY, green) and 70 KDa dextran-Texas Red® (DTR, red). Tissue was processed as defined in Methods and counterstained with DAPI (blue) for confocal microscopy studies assessing submucosal LY molecule penetration. Bar graph on panel’s right side shows similar tissue permeability in DMPA-treated and OVX mice and that epithelial barrier function was restored in OVX mice treated with E to levels found in estrus-stage mice. One-way ANOVA and Dunn’s post hoc test was used for statistical comparison. (b) Though comparable results were seen in DMPA-treated and OVX mice when vaginal tissue was collected 45 min after ivag fluorescent molecule administration, identically performed studies saw LY entry into vaginal submucosa significantly enhanced in OVX vs. DMPA-treated mice if tissues were obtained 15 or 30 min after these molecules were administered. Left side of panel shows typical results. Bar graph on the right side of panel identifies more rapid LY molecule penetration in OVX vs. DMPA-treated intact mice. Two-way ANOVA with Bonferroni’s multiple comparisons test provided statistical comparison. In (a) and (b), results shown are representative of 2 independent experiments with 4 animals per group. For images, scale bar denotes 100 µm and white line drawn to delineate vaginal epithelium from adjacent stroma. DAPI, 4,6-diamidino-2-phenylindole; DMPA, depot-medroxyprogesterone acetate; (e), conjugated estrogens vaginal cream; OVX, ovariectomized

Susceptibility to genital viral and bacterial pathogens was significantly increased in OVX mice

Because OVX mice displayed a significant loss of genital epithelial barrier function, we hypothesized such mice have greater genital pathogen susceptibility. To test this hypothesis, estrus-stage mice, DMPA-treated intact mice, and untreated and Premarin®-treated OVX mice were observed daily after genital infection with 10³ PFU of HSV-2. These studies showed estrus-stage mice and Premarin®-treated OVX mice were resistant to genital HSV-2 infection, while all DMPA-treated and OVX mice developed genital pathology and succumbed to infection (Figure 4(a–c)). Interestingly, whereas DMPA-treated and OVX mice were uniformly susceptible to HSV-2 infection, the OVX mice showed earlier morbidity and mortality and significantly higher viral burden in CVLs collected at 2 dpi (Figure 4(a–d)). To extend our results to bacterial pathogen susceptibility, we also measured genital clearance of C. trachomatis after estrus-stage mice, DMPA-treated intact mice, and untreated and Premarin®-treated OVX mice were genitally inoculated with 10⁴ IFU of this pathogen. Compared to acutely infected estrus-stage mice and Premarin®-treated OVX mice, Chlamydia load was larger in the genital tract of DMPA-treated intact mice and OVX mice (Figure 4(e,f)). Together, these studies revealed there is comparably increased susceptibility to viral and bacterial genital pathogens in OVX mice and intact mice treated with DMPA, the synthetic progestin used to achieve uniform infectivity in various murine models of genital infection.^27–29

Figure 4.

Genital pathogen susceptibility was significantly increased in DMPA-treated and OVX mice. (a) After estrus-stage and DMPA-treated intact mice and untreated and E-treated OVX mice were genitally inoculated with 10³ PFU of HSV-2, genital pathology developed in all DMPA-treated intact mice and untreated OVX mice, but no E-treated mice or mice infected in estrus. Conversely, onset of genital pathology occurred 2 days earlier in OVX vs. DMPA-treated mice. (b) AUC analyses of HSV-2 induced genital pathology through 4 dpi (i.e., the last day that each group contained equal number of animals) identified significantly greater pathology in DMPA-treated and OVX mice vs. estrus stage and E-treated OVX mice (unshown P value <.0001) and OVX mice vs. DMPA-treated mice (P < .0001); differences assessed with one-way ANOVA with Dunnett’s post hoc test. (c) Consistent with genital pathology data, while all DMPA-treated mice and OVX mice succumbed to infection, mortality occurring significantly earlier in OVX mice; log-rank test assessed between-group differences. (d) Panel identifies that genital HSV-2 burden was significantly larger in OVX vs. DMPA-treated mice at 2 dpi; unpaired Student t-test used for comparison. (e) In separate studies, estrus-stage and DMPA-treated intact mice and untreated and E-treated OVX mice were genitally infected with 10⁴ IFU of C. trachomatis. CVL collected at various time points after infection were used in qPCR assays that showed greater bacterial burden in DMPA-treated and OVX-mice vs. estrus-stage mice and E-treated OVX mice. (f) AUC analyzes using one-way ANOVA with Dunnett’s post hoc test identified greater Chlamydia burden in DMPA-treated and OVX-mice vs. estrus-stage mice and E-treated OVX mice. AUC; area under the curve; CVL; cervicovaginal lavage; dpi, days post-infection; DMPA, depot-medroxyprogesterone acetate; E; conjugated estrogens vaginal cream; IFU, inclusion-forming unit; PFU; plaque-forming unit; qPCR, quantitative real-time PCR

Discussion

Cessation of ovarian function and loss of circulating E are associated with menopause and the GSM, a syndrome whose symptoms include vaginal dryness and irritation, lack of lubrication, dyspareunia, urinary urgency, and dysuria. In addition, the incidence of UTI rises sharply after menopause, and clinical studies suggest that 20% of the women 65–70 years of age have bacteriuria.^30,31 While less explored, there are additional data to suggest that susceptibility to HIV and other sexually transmitted infections (STI) is enhanced after menopause. One prospective study of 181 women at low risk for STI identified genital C. trachomatis infection in 0.8% of the women 30–49 years of age vs. 23% of the women 50–65 years of age.³² Whereas the author of this publication concluded that acquisition of Chlamydia infection in the older cohort probably occurred at a younger age (with the bacterium persisting in the genital tract for decades), the paucity of infection in the younger cohort makes postmenopausal genital tract changes promoting Chlamydia susceptibility the more plausible explanation for the higher number of cases in women 50–65 years of age. As other examples, a European study of HIV discordant heterosexual couples found women greater than 45 years of age were nearly fourfold more likely to acquire HIV than younger women and an African study found this age group eightfold more likely than women less than 45 years of age to acquire HIV-2.^33,34 While no mechanisms were explored in these clinical studies, elevated UTI incidence, more frequent bacteriuria, greater susceptibility to C. trachomatis infection, and increased transmission of HIV among older women all suggest that age-related changes in genital tract anatomy or function impair host immune responses to urogenital pathogens and represent important risk factors for urogenital infection.

The possibility menopausal and postmenopausal women are at greater risk for HIV, and other STI is also suggested by current findings that show compared to reproductive-age women, ectocervical epithelial tissue from older women has significantly lower levels of the desmosomal cadherins DSC1 and DSG1. These intercellular junctional molecules impart mechanical strength to epithelial tissue by anchoring a network of flexible intermediate filaments to the plasma membrane,³⁵ and their key role in epithelial integrity and barrier function is supported by the profound disruption of these functions in humans and mice deficient in DSG1.^36–39 The lower levels of ectocervical DSC1 and DSG1 protein observed in menopausal and postmenopausal women in the current study are thus indicative of reduced epithelial integrity and are consistent with the enhanced vulnerability of genital tissue of postmenopausal women to physical irritation and trauma.⁴⁰ While the loss of genital epithelial integrity and increased risk for injury to genital tract tissue in menopausal and postmenopausal women may also facilitate the sexual transmission of bacterial and viral pathogens, further research is needed to connect weakened genital epithelial integrity with reduced genital epithelial barrier function or greater STI susceptibility and resolve if there is a greater compromise of genital epithelial barrier integrity in women affected by GSM or specific GSM symptoms.

Analogous to the lower genital levels of DSC1 and DSG1 protein in menopausal and postmenopausal women, current findings reveal that levels of Dsc1 and Dsg1a gene expression and DSG1 and DSC1 protein amount are significantly lower in vaginal tissue from DMPA-treated intact mice and OVX mice vs. mice in the estrus stage of the estrous cycle. These results confirm and extend findings from our prior work that showed DMPA administration to mice significantly reduced vaginal expression of Dsc1 and Dsg1a.⁵ Because vaginal tissue from estrus-stage and DMPA-treated mice conversely displayed comparable gene expression levels of tight junction protein 1, occludin, and claudin-1,⁵ together our findings imply the hypoestrogenic states induced by DMPA and menopause are less likely to impact transcriptional regulation of genital epithelial tight junctions (zonula occludens) vs. the desmosome (macula adherens). Current findings also suggest that compromise of genital epithelial barrier function and genital pathogen susceptibility is greater in OVX vs. DMPA-treated mice. Whereas genital epithelial barrier function was impaired in both DMPA-treated and OVX mice, we saw ivag administered LMM molecules more rapidly penetrate the vaginal submucosa of OVX mice. Likewise, although DMPA-treated and OVX mice were uniformly susceptible to genital HSV-2 infection, the latter displayed a significantly higher burden of the virus at 2 dpi and earlier onset of morbidity and mortality. It, therefore, seems possible that differences in genital epithelial barrier function and pathogen susceptibility are coupled to the lower concentrations of estradiol that circulate in OVX vs. DMPA-treated mice.^6,41–43

Whereas DMPA-treated intact mice and OVX mice were uniformly susceptible to genital HSV-2-infection, ivag administration of Premarin® made OVX mice wholly resistant to this genital infection. Compared to untreated OVX mice at early time points after infection, Premarin®-treated OVX mice also had a lower genital tract burden of Chlamydia. These findings are consistent with prior work that found humanized mice administered DMPA and Premarin® significantly less susceptible to genital infection with cell-associated HIV than humanized mice treated with DMPA alone and OVX nonhuman primates treated with E significantly more resistant to genital simian immunodeficiency virus infection than untreated controls.^7,44–46 Current results also indicate that OVX mice model important elements of the impact of ovarian senescence on human female genital tissue and suggest this experimental model can help elucidate cellular mechanisms by which endogenous and exogenous sex steroids regulate genital epithelial integrity and barrier function. Along with the important effects exerted by progestogenic and estrogenic compounds on genital DSG1 and DSC1 protein levels, prior studies demonstrated that the hypoestrogenemia generated in DMPA-treated mice significantly enhanced genital expression of several kallikrein-related peptidases,²⁸ serine proteases known to promote DSG1 cleavage.⁴⁷ It is therefore possible that the hypoestrogenic states induced by mouse ovariectomization and ovarian senescence similarly impact the genital expression of kallikrein-related peptidases, and our research group is exploring this possibility. In addition to supporting the continued use of animal models to define mechanisms by which reduced levels of E promote loss of genital epithelial integrity, current findings seem to newly prioritize clinical research that can address: how host immune defense systems are impacted by menopause; if the loss of genital tissue integrity in menopausal women is a significant risk factor for STI acquisition; and the intriguing possibility that reduced STI susceptibility is an underrecognized benefit among women using E-containing compounds to relieve GSM symptoms.

Funding Statement

Support for this work was provided by the Eunice Kennedy Shriver National Institute of Child Health and Human Development [R01HD094634].

Disclosure of potential conflicts of interest

The authors report no conflict of interest.