Mechanosensitive Vaginal Epithelial ATP Release and Pannexin 1 channels in Healthy, in Type 1 Diabetic and in Surgically Castrated Female Mice

Abstract

Background:

Distension of hollow organs is known to release ATP from the lining epithelium, which triggers local responses and activates sensory nerves to convey information to the CNS. However, little is known regarding participation of ATP and mediators of ATP release, such as Pannexin 1 (Panx1) channels, in mechanisms of vaginal mechanosensory-transduction and of changes imposed by diabetes and menopause, conditions associated with vaginal dysfunction and risk for impaired genital arousal.

Aim:

Investigate if intravaginal mechanical stimulation triggers vaginal ATP release and if (a) this response involves Panx1 channels and (b) is altered in animal models of diabetes and menopause.

Methods:

Diabetic Akita female mice were used as a Type 1 diabetes (T1D) model and surgical castration (ovariectomy; OVX) as a menopause model. Panx1-null mice were used to evaluate Panx1 participation in mechanosensitive vaginal ATP release. Vaginal washes were collected from anesthetized mice at baseline (non-stimulated) and at 5 minutes after intravaginal stimulation. For the OVX and Sham groups, samples were collected before surgery and at 4, 12, 22, 24 and 28 weeks post-surgery. ATP levels in vaginal washes were measured using the Luciferin-luciferase assay. Panx1 mRNA levels in vaginal epithelium were quantified by qPCR.

Main Outcome Measure:

Quantification of mechanosensitive vaginal ATP release and evaluation of impact of Panx1 deletion, OVX and T1D on this response.

Results:

Intravaginal mechanical stimulation induced vaginal ATP release that was 84% lower in Panx1-null (P<0.001) and 76% lower in diabetic (P<0.0001) mice compared to controls, and was reduced in a progressive and significant manner in OVX mice when compared to Sham. Panx1 mRNA expression in vaginal epithelium was 44% lower in diabetics when compared to controls (P<0.05), and 40% lower in OVX when compared to Sham (P<0.05).

Clinical Translation:

Panx1 downregulation and consequent attenuation of mechanosensitive vaginal responses may be implicated in mechanisms of female genital arousal disorder (FGAD), thereby providing potential targets for novel therapies to manage this condition.

Strengths & Limitations:

Using animal models, we demonstrated Panx1 involvement in mechanosensitive vaginal ATP release, and effects of T1D and menopause on this response and on Panx1 expression. A limitation is that sex steroid hormone levels were not measured, precluding correlations and insights into mechanisms that may regulated Panx1 expression in the vaginal epithelium.

Conclusions:

Panx1 channel is a component of the vaginal epithelial mechanosensory-transduction system that is essential for proper vaginal response to mechanical stimulation and is targeted in T1D and menopause.

Introduction

The mechanosensitivity of the female genital organs and its importance for proper perception and response to penetrative sexual stimulation are well recognized. However, our current knowledge of the molecular mediators and mechanisms implicated in vaginal mechanosensation and mechanotransduction is still limited.

In hollow organs such as the vagina, ATP and its purinoceptors have been proposed to be involved in mechanisms of mechanosensory transduction.¹ Distension leads to ATP release from the epithelial lining of these organs, and through both autocrine and paracrine signaling, ATP modulates local responses to mechanical stimulation and relays information to the central nervous system. In the urogenital system, this role for ATP in mechanosensitive responses has been reported for the ureters and extensively for the bladder. ^2–5

In the bladder, several channels, receptors and transporters participate in mechanisms of urothelial mechanosensation and mechanosensitive ATP release.⁶ Pannexin 1 channels are among those and are unique, as they function not only as mechanosensors but also as mechanotransducers by providing a direct conduit for controlled cellular ATP efflux.⁷ Absence of Panx1 or pharmacological inhibition of these channels blunt urothelial ATP release and signaling, and impair proper responses to bladder distension, as normally occurs during the micturition cycle.^{7, 8}

The vagina, similar to the bladder, is equipped to respond to ATP signaling. The vaginal epithelium expresses purinoceptors of both P2Y and P2X subtypes.^{9, 10} Activation of P2Y₂ receptors with selective agonists has been shown to stimulate vaginal moisture in ovariectomized rabbits, and both P2X5 and P2X7 receptors have been proposed to regulate proliferation and differentiation of the continuously renewing vaginal epithelium.^11–14 P2 receptors are also present in the vaginal wall and have been shown to induce relaxation that does not involve an indirect action on intrinsic nerves and release of secondary transmitters.¹⁵ Subepithelial sensory fibers expressing P2X3 and/or P2X3/2 receptors may also be present in the vaginal wall, as described for the rat cervix, urinary bladder and ureters, and would similarly participate in the transmission of mechanosensory information to the CNS.^16–18

Purinergic signaling in the vaginal wall is likely mediated by ATP released from both neuronal and non-neuronal sources. However, signaling evoked in response to direct stimulation of the vaginal canal, as imposed during penetrative intercourse, is expected to originate in the vaginal epithelium. It is also conceivable that the vaginal epithelium, similar to the bladder urothelium, not only provides the primary source for the ATP that triggers the vaginal mechanosensitive responses, but also serves as a mechanosensor. Such putative roles of the vaginal epithelium in mechanisms of vaginal mechanosensation and mechanotransduction would significantly expand its functional repertoire. They could also implicate the vaginal epithelium in mechanisms that can lead to genital arousal problems, and further add to the complexity of biological factors that can contribute to the etiology of female genital arousal disorder (FGAD).¹⁹

In this study, we specifically address this potential role of the vaginal epithelium in mechanisms of vaginal mechanosensation and purinergic mechanosensory transduction. We first used wildtype and Panx1-null female mice to determine whether intravaginal mechanical stimulation, as occurs during penile intromission, triggers vaginal ATP release, and define the extent to which this response involves the activation of Panx1 channels. We next investigated whether vaginal epithelial Panx1 expression and mechanosensitive ATP release are altered in animal models of menopause and Type 1 diabetes. These conditions are known to dysregulate vaginal function and to impair the female genital arousal response. Use of these animal models are thus expected to provide interesting insights into the possible involvement of the vaginal epithelial mechanosensory and mechanotransduction system in mechanisms that can lead to genital arousal problems in menopausal and Type 1 diabetic women.

Materials and Methods

Animals

In this study we used 3 month old female C57BL/6 mice purchased from Charles Rivers Laboratories (Wilmington, MA) and global Panx1 deficient female mice (Panx1-null) on the C57BL/6 background, generated in our animal facility by breeding heterozygous Panx1^{tm1a(KOMP)Wtsi} purchased from the Knockout Mouse Project (KOMP; www.komp.org) at UCDavis.^{7, 20, 21} Age-matched wildtype (WT) littermates bred in house from our colonies or C57BL/6 female mice purchased from Charles Rivers were used as controls for studies conducted with the Panx1-null mice. Akita heterozygote female mice (C57BL/6 ^Ins2Akita/J) purchased from Jackson laboratory (Bar Harbor, ME) and bred in our animal facility were used as a Type 1 diabetes (T1D) model. Age-matched littermate wildtype (WT) Akita females were used as controls. Experiments were conducted with T1D Akita and healthy WT females at 5 months of age. The diabetic status was confirmed by measuring both urine glucose levels (Diastix® reagent strips for urinalysis; Bayer, Mishawaka, IN) prior to starting the experiments and blood glucose levels (OneTouch Ultra 2, Blood Glucose Monitoring System; LifeScan, Inc., Milpitas, CA) at conclusion of the experiments before tissue harvesting. Surgical castration was used as the experimental model of menopause. Procedures to generate this model were as follows. The C57BL/6 female mice were anesthetized via intraperitoneal injections of a ketamine (KETATHESIA™, Henry Schein, Melville, NY) and xylazine (AnaSed®, Akorn Inc., Lake Forest, IL) mixture (150:10 mg/kg). The ventral abdominal wall and perineum were shaved with an electrical shaver and cleansed with 10% povidone-iodine solution (Betadine®, Purdue Pharma, Stamford, CT). A low midline abdominal incision was made and the ovaries were identified. In the ovariectomized group (OVX), the ovaries were clamped, transected with cauterization and then excised. In the Sham group, the ovaries were only identified. The abdominal incision was then sutured and the animals allowed 4 weeks of recovery before proceeding with experimental procedures. All animals were housed in AAALAC-approved animal facilities and studies conducted in accordance with the Guide for the Care and Use of Laboratory Animals.

Determination of Estrous Cycle Stage

Vaginal lavages were collected to assess the stage of the estrous cycle. The vaginal canal was flushed with 20 μl of Dulbecco’s Phosphate Buffered Saline (DPBS; Corning Cellgro, Mediatech, VA), the vaginal lavage placed on a glass slide and vaginal smear viewed at 40x magnification. The stage of the estrous cycle was then determined based on cell type and abundance, as previously described.²² Estrous cycle stage was determined for all animals before the experimental procedures.

Collection of Vaginal Washes and Protocol for Intravaginal Mechanical Stimulation

Mice were anesthetized using a ketamine and xylazine mixture (150:10 mg/kg) and placed dorsal supine on a heating pad. The stage of the estrous cycle was determined as described above. A custom made stabilizing platform was then placed over the pelvic region of the anesthetized mice for controlled mechanical stimulation and collection of the vaginal washes. Experiments were conducted 30 min after assessment of the estrous cycle, and mechanical stimulation of the inner vaginal canal wall was performed using a custom made intromission device. This device consisted of a plastic micropipette tip with a ceramic bead (2.07 ± 0.03 mm in diameter; n=6) inserted ~1 mm up into its lower end (stimulation probe) and attached to a micropipette (Gilson Pipetman P100, Middleton, WI) to allow collection of vaginal washes. The size of the ceramic beads used in this device is comparable to that of the polished brass “penis” used by Diamond in studies where mechanical stimulation was also employed to mimic normal vaginal stimuli during copulation in mice.²³ Samples for baseline measurement of luminal vaginal ATP levels were collected before stimulation. For this, the intromission probe was gently inserted into the vaginal orifice and the vaginal canal instilled with 30 μL of warm DPBS. After 30 min, the instilled DPBS (herein referred to as vaginal wash) was collect and fresh 30 μL DPBS was instilled again in the vagina. After a 30 min rest, intravaginal mechanical stimulation was applied by gently thrusting the probe along the length of the vaginal canal up to the cervix. The vaginal wash was then collected 5 min post-stimulation. Studies of copulatory behavior in male mice have shown that depending on the mouse strain, the time of penile intromission lasts for 8 to 25 sec with number of thrusts ranging from 5 to 20 per intromission, resulting in stimulus frequencies ranging from 0.63 to 1.25 Hz.^24–28 The mechanical stimulus or loading chosen for this study was based on these reported parameters. The vaginal canal was stimulated with three consecutive deep thrusts at 1 Hz, applied with average speed of 8.7 ± 0.3 mm/s (n=6) and force of 103.2 ± 5.3 nN (n=6), imposing approximately 7% radial tissue strain. After collection, all vaginal wash samples were immediately snap-frozen in liquid nitrogen and stored at −80°C. Baseline and post-stimulation luminal vaginal ATP levels were then quantified using the Luciferin-luciferase assay as described below.

Measurement of ATP Levels in Vaginal Washes

ATP levels in vaginal washes were measured using the Luciferin-luciferase method as previously described (ATP Determination Kit, Molecular Probes®, Thermo Fisher Scientific, Life Technologies, Eugene, OR).⁷ Briefly, 5 μl of the collected vaginal wash and DPBS (for background correction) were individually placed in duplicates in white walled 96-well plates and 45 μl of a standard reaction solution (0.5 mM D-luciferin, 1.25 μg/mL firefly luciferase, 25 mM Tricine buffer, pH 7.8, 5 mM MgSO₄, 100 μM EDTA and 1 mM DTT) was added to each well. Immediately after, the plate was transferred to a FLUOStar plate reader (BMG Labtech, Ortenberg, Germany) and luminescence was measured using a 5 sec integration time. The ATP concentration in the vaginal wash samples was calculated from standard curves constructed using ATP from 1 nM to 1000 nM.

Collection of Vaginal Tissue

The mice were euthanized in a CO₂ chamber. The vaginas were harvested, transferred to cold phosphate buffer saline (PBS; Corning Cellgro, Mediatech, VA) and cleaned of connective tissue. A vaginal ring (2 mm) was cut from the middle section of the vaginal canal and embedded in O.C.T. (Optical Cutting Temperature) compound (Thermo Fisher Scientific, Waltham, MA) for cryosectioning. The other two vaginal rings were cut open and incubated with Dispase (1 U/mL in DMEM/F-12; STEMCELL Technologies, Vancouver, Canada) at 36°C for 1 hour to facilitate separation of the vaginal epithelium from underlying muscle layer. The vaginal epithelium was then carefully scraped from the underlying muscle using a scalpel blade under a dissecting microscope. The vaginal epithelium segments were then cut in small pieces and kept at −80°C until processing for real-time quantitative PCR analysis.

Vaginal Tissue Histology and Immunohistochemistry

The vaginal ring OCT blocks were serially cryosectioned at 10 μm thickness and mounted in Superfrost Plus microscope slides (Thermo Fisher Scientific). Tissue sections were then processed for standard Hematoxylin and Eosin (H&E) staining and for Panx1 immunostaining, as previously described.⁷ Panx1 immunodetection was performed using the rabbit polyclonal Pannexin 1 NT antibody (N-terminus, 1:100, Invitrogen, Carlsbad, CA) followed by Alexa Fluor 488 donkey anti-rabbit IgG (H+L) (Invitrogen), and DAPI was used for nuclei staining. Tissue sections were examined and photographed using an Olympus microscope coupled to a DP72 cooled digital color camera.

Molecular Analysis

Real-time quantitative PCR (qPCR) was used to quantify and compare mRNA expression levels of Panx1 in the vaginal epithelial tissue from experimental and control mice. Total RNA was extracted from homogenized vaginal epithelial tissue using the RNeasy Plus mini kit (Qiagen, Valencia, CA) and 1 μg of total RNA was reverse transcribed into cDNA using Superscript VILO (Invitrogen). RT reactions were diluted in RNase-free water in a 1:100 ratio and 3 μL of this dilution was used in amplification reactions carried out for 40 cycles with annealing temperature of 60 °C in a 25 μL of final volume. Finally, a dissociation profile of the PCR product(s) was obtained by a temperature gradient running from 60 to 95 °C. Two technical replicates were performed per tissue sample. Relative gene expression levels of the mRNA analyzed were calculated using the delta-delta CT method, where values obtained for the gene of interest are first normalized to those of the reference gene (18S rRNA gene) and subsequently to those of their respective age-matched controls. Primers used in this study were as follows: Panx1, Fwd: TATTGCCGTGGGTCTACCTC and Rev: TGTCGCCAGGAGAAAGAACT, and 18S ribosomal RNA, Fwd: CACGGCCGGTACAGTGAAAC and Rev: AGAGGAGCGAGCGACCAAA.

Statistical Analysis

The GraphPad Prism 7 software was used for data and statistical analyses. Data are represented as mean±SEM. Statistical differences between non-stimulated (baseline) and mechanically-induced vaginal ATP released amounts were determined by paired Student’s t-test. One-way ANOVA followed by Tukey’s multiple comparison test or Multiple t tests with statistical significance determined using the Holm-Sidak method were used for comparisons between groups. Statistical differences in vaginal epithelial Panx1 mRNA levels in OVX and in diabetic mice relative to their controls were determined by unpaired Student’s t-test. Statistical significance was set at P<0.05 for all tests.

Results

Effect of intravaginal mechanical stimulation on ATP released in the vaginal lumen of healthy female mice.

ATP levels in the vaginal lumen of female mice were measured under non-stimulated (baseline, control) condition and after gentle mechanical stimulation of the vaginal canal. Intravaginal mechanical stimulation induced significant ATP release in the vaginal lumen, which was 3 times higher than measured under non-stimulated conditions (Baseline: 122.6 ± 24.7 nM vs. Stimulated: 389.3 ± 25.3 nM; n=10 per group; P<0.0001).

Effect of Pannexin 1 absence on responses to vaginal mechanical stimulation.

Examination of immunostained vaginal wall cross-sections of female mice demonstrated that Panx1 is abundantly expressed in the vaginal epithelium and is also present in the vaginal smooth muscle layer (Figure 1A). To investigate the potential involvement of these channels in mechanisms of vaginal mechanosensation and mechanosensory transduction, we determine the impact of Panx1 absence on the mechanosensitive vaginal ATP release response. Figure 1B compares the responses induced by intravaginal mechanical stimulation of WT and Panx1-null female mice. Both Panx1-null and WT female mice responded to vaginal mechanical stimulation with release of ATP that was significantly higher than that measured at non-stimulated, baseline conditions. However, when these responses are compared, the vaginal ATP release in mechanically stimulated Panx1-null female mice was 6 times lower than the response observed in mechanically stimulated WT mice (WT: 323.8 ± 5.36 nM, n=11 vs Panx1-null: 52.7 ± 20.6 nM, n=5; P<0.001). Gross examination and comparison of the H&E stained Panx1-null and WT vaginal wall cross-sections indicate that there are no apparent differences in the vaginal epithelium of these mice (Figure 1C). The observed differences in mechanosensitive ATP release can thus be primarily attributed to the lack of Panx1 expression in the female Panx1-null mice, and supports the participation of these channels in mechanisms of vaginal mechanosensation and mechanotransduction.

Figure 1: Pannexin 1 (Panx1) channels are expressed in the mouse vaginal epithelium and mediate the ATP release response induced by mechanical stimulation of the vaginal canal.

(A) Representative immunostaining for Panx1 (green) in vaginal wall sections from a C57BL/6 female mice. (Blue = nuclear DAPI staining; scale bar = 20 μm). (B) In vivo mechanical stimulation of the vaginal canal induces release of significant amounts of ATP from the vaginal epithelium of wildtype (WT) female mice that is blunted in Panx1-null female mice. Data correspond to mean±SEM, n=11 WT and n=5 Panx1-null females. Statistical differences were determined by paired t-test and by ANOVA followed by Tukey’s multiple comparison test: **P<0.01, ***P<0.001 and ****P<0.0001. (C) Photomicrographs of hematoxylin and eosin (H&E) staining of vaginal wall cross-sections from a 3 months old (a) WT female mouse and a (b) Panx1-null female mouse (scale bar: 40 μm). Note that vaginal structure is not altered in the absence of Panx1 expression.

Effect of surgical castration on the mouse vaginal ATP release.

ATP levels in the vaginal lumen were measured before (baseline, non-stimulated) and after intravaginal mechanical stimulation both prior to surgical intervention and then longitudinally at 4, 12, 22, 24 and 28 weeks after ovariectomy (OVX) or sham laparotomy (Sham) surgery. Before surgery, there were no differences between the animals randomly selected for the Sham and OVX groups with regard to both baseline and simulated vaginal ATP release (Figure 2A). However, post-surgical assessment of vaginal responses to mechanical stimulation of OVX mice relative to Sham mice indicated a progressive and a significant decrease in ATP released amounts along the course of the study, which plateau at 24-28 weeks after OVX at levels that were 10 times lower than those measured from age-matched Sham mice (Figure 2B).

Figure 2: Surgical castration impairs the ATP release response induced by mechanical stimulation of the vaginal canal.

ATP released into the vaginal lumen under non-stimulated (baseline) and after intravaginal mechanical stimulation was quantified before operation (Pre-op) and then at 4, 12, 22, 24 and 28 weeks after surgical castration (ovariectomy, OVX) or sham laparotomy. Note that before surgery (A) the baseline and the mechanically-induced vaginal ATP release were not different between the animal cohorts. In contrast, after surgery (B), the relative response of OVX mice to intravaginal stimulation were significantly lower than those of Sham mice. Note lower vaginal ATP release at 4 weeks post-OVX that progressively decreased in time and plateau at 24-28 weeks Post-Op. In (B), ATP released in response to intravaginal mechanical stimulation is expressed relative to average Sham ATP released amounts. Data correspond to mean±SEM, n=5 Sham and n=5 OVX mice. Statistical differences were determined by paired t-test and by ANOVA followed by Tukey’s multiple comparison test in (A), and by Multiple t tests in (B): *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001. (C) Photomicrographs of hematoxylin and eosin (H&E) staining of vaginal wall sections from a Sham-operated female mouse (a) age-matched to an OVX female mouse (b) at 28 weeks after ovariectomy (scale bar: 100 μm). Note the thinner vaginal epithelium in the OVX compared to the Sham female mouse, which is one of the characteristic signs of the vaginal atrophy imposed by surgical castration.

Surgical castration was validated at the completion of the experiment by measuring and comparing both body and uterus weight of OVX and Sham mice, and also by examining the H&E staining of the vaginal wall cross-sections. At 28 weeks after surgical castration, the body weights of OVX mice (45.0 ± 3.2 g; n=5) were significantly higher than those of Sham mice (36.1 ± 1.8 g; n=5; P<0.05). In addition, surgical castration induced a significant reduction in uterus size. When compared by weight, the uterus of OVX mice (43.7 ± 11.8 mg) was more than 2-fold lighter than that of sham mice (99.3 ± 12.2 mg; P<0.01). Gross examination of H&E stained vaginal wall sections indicates that the vagina of Sham-operated C57BL/6 female mice has a thicker wall and well differentiated histological features in comparison to the vagina of age-matched female mice at 28 weeks after ovariectomy, which appears atrophic with only 2-3 cell layers in thickness (Figure 2C). This finding is in line with previous reports demonstrating that surgical castration, mimicking a menopausal low-estrogen state, imposes marked structural and histological changes on the vagina. ²⁹

Effect of Type 1 diabetes on the mouse vaginal ATP release.

ATP levels in the vaginal lumen of diabetic Akita and healthy control WT mice were measured before (baseline, non-stimulated) and after intravaginal mechanical stimulation. Average baseline levels of vaginal luminal ATP did not differ between diabetic and healthy WT female mice (Control: 35.1 ± 4.2, n = 9 vs Akita: 27.1 ± 9.1, n=12; P=0.99; Figure 3A). However, diabetes significantly reduced the response to intravaginal mechanical stimulation. As shown in Figure 3A, diabetic females responded to intravaginal stimulation with release of ATP but this response was 4 times lower than that observed in mechanically stimulated healthy control mice (Control: 338.0 ± 63.7 nM, n=9 vs Akita: 79.8 ± 27.1 nM, n=12; P<0.0001).

Figure 3: Type 1 diabetes impairs the ATP release response induced by mechanical stimulation of the vaginal canal.

(A) In vivo mechanical stimulation of the vaginal canal of healthy control female mice resulted in significant release of ATP from the vaginal epithelium when compared to baseline levels. Intravaginal mechanical stimulation of diabetic female mice also induced ATP release but this response was blunted when compared to that of healthy females. Data correspond to mean±SEM, n=9 control and n=12 diabetic Akita females. Statistical differences were determined by paired t-test and by ANOVA followed by Tukey’s multiple comparison test: *P<0.05 and ****P<0.0001. (B) Photomicrographs of hematoxylin and eosin (H&E) staining of vaginal wall sections from (a) a 5-month old healthy control female mouse and (b) an age-matched diabetic Akita female mouse (scale bar: 40 μm). Note the absence of apparent structural differences between the healthy control and the T1D mice at the age employed in this study.

Gross examination of H&E stained vaginal wall cross-sections of diabetic Akita and healthy age-matched WT control mice (Figure 3B) indicated that there were no apparent structural differences between the healthy WT and the T1D mice at the age of 5 month employed in this study. This finding indicates that observed effects of T1D on vaginal responses to mechanical stimulation are related to physiological rather than structural changes in the vaginal epithelium.

Effects of surgical castration and Type 1 diabetes on Pannexin 1 expression in the mouse vaginal epithelium.

Panx1 mRNA levels in the vaginal epithelium were determine by real-time quantitative PCR and compared between OVX and Sham mice at 28 weeks post-surgery, and between 5 month old diabetic Akita and healthy age-matched WT female mice. As shown in Figure 4A, surgical castration resulted in a significant decrease in Panx1 mRNA levels in the vaginal epithelium of OVX mice, which were in average 39.8 ± 3.4 % (n = 5) lower than those measured from Sham operated mice (P<0.05). T1D also induced a significant downregulation of Panx1 expression in the vaginal epithelium. As shown in Figure 4B, Panx1 mRNA levels in the vaginal epithelium of diabetic Akita female mice were 44.4 ± 12.8 % (n = 5) lower than those measured from their age-matched control counterparts (P<0.05). These findings indicate that downregulation of Panx1 expression in the vaginal epithelium is likely the main factor driving the deleterious effects of T1D on the vaginal mechanosensitive responses of the 5 month old Akita female mice. The observed effects of surgical castration on both the vaginal epithelial structure and Panx1 expression levels indicate that not only structural changes but also dysregulation of Panx1 expression contribute to the decrease in vaginal mechanosensitive responses observed in the OVX female mice.

Figure 4: Surgical castration and Type 1 diabetes alters Pannexin 1 expression in the vaginal epithelium.

(A) Panx1 mRNA levels in the vaginal epithelium of ovariectomized (OVX) and age-matched Sham-operated controls (Sham) at 28 weeks post-surgery. Note significantly lower Panx1 expression in OVX compared to control healthy tissues. (B) Panx1 mRNA levels in the vaginal epithelium of 5 month old diabetic Akita (Diabetic) and age-matched healthy WT controls (Control). Note significantly lower Panx1 expression in diabetic compared to healthy tissues. Data is represented as mean±SEM, n=5 per group, normalized by 18S and then by average control values. Statistical differences were determined by unpaired t-test: *P<0.05.

Discussion

ATP is an important regulator of vaginal function. Local activation of purinergic receptors has been shown to induce relaxation of the vaginal smooth muscle, to stimulate vaginal moisture and regulate proliferation and differentiation of the vaginal epithelium. ^11–15 Data obtained from this study support additional roles for ATP and expand the functional repertoire of the vaginal epithelium. We have shown that the vaginal epithelium is mechanosensitive and responds to mechanical stimulation with release of ATP, a response that is mainly mediated by Panx1 channels. These findings are in line with the proposed involvement of purinergic signaling in mechanisms of mechanosensory transduction in hollow organs and suggest that the vaginal epithelium may exert roles in vaginal mechanosensation and mechanotransduction that can parallel those described for the urothelium in the urinary bladder.⁶

The urothelial mechanosensory-transduction machinery is comprised of various receptors and channels, among those the Panx1 channels, that directly or indirectly participate in mechanisms of urothelial ATP release.⁶ It remains to be determined whether aside of Panx1, other channels and receptors are involved in mechanosensitive vaginal epithelial ATP release. Nonetheless, our findings demonstrated that in the absence of Panx1 the responses to intravaginal mechanical stimulation are virtually abolished, which indicates that the Panx1 channel is a major component of the vaginal epithelial mechanosensory-transduction system. Moreover, because of the unique properties of these channels, we can expect that the role of Panx1 in the vaginal epithelium would go beyond that of serving exclusively as a mechanosensor and direct conduit for controlled cellular ATP release. Besides being mechanosensitive, Panx1 channels respond to other stimuli, including activation of ATP receptors, particularly the P2X7R subtype.^30–32 In previous studies, we demonstrated that this Panx1 sensitivity to P2X7R activation generates a feed-forward mechanism of ATP-induced-ATP release that can significantly amplify purinergic signaling.⁷ Considering that P2X7R are expressed in the vaginal epithelium, such mechanism may also operate in the vagina and thereby contribute to markedly amplify ATP release and purinergic signaling within the vaginal epithelium and from the epithelium to the underlying vaginal smooth muscle and sensory fibers.^{11, 12, 14} Propagation of the ATP signal through the vaginal wall has the potential to influence vaginal function. For example, it can activate the vaginal epithelial P2Y₂R-mediated mechanism of vaginal lubrication, independently of blood flow.¹³ It can directly modulate vaginal tonus through activation of P2 receptors shown to induce vaginal smooth muscle relaxation.¹⁵ Finally, it can also indirectly contribute to activation of the neuro-vascular components of the female genital arousal response, when it activates the vaginal afferent fibers to convey sensory information pertaining the mechanical stimulation of the vaginal canal, as occurs during sexual penetration. Panx1 channels, alongside P2 receptors, can thus be regarded as important molecular mediators in mechanisms that regulate vaginal function.

Panx1 channels are also gaining increasing prominence for their emerging roles in pathological processes. For example, in the urogenital system, changes in expression and function of Panx1 channels have been implicated in the development of neurogenic bladder in an animal model of Multiple Sclerosis, and these channels have also been proposed to participate in mechanisms of bladder overactivity.^{20, 33} In line with the role of Panx1 channels in mechanosensitive responses, we have also shown in previous studies that absence of these channels impairs the response of bone cells to fluid shear stress and blunts the load-induced responses of long bones.^{34, 35} Data that we have now obtained in this study provide evidence that Panx1 channels are also likely involved in mechanisms of vaginal dysfunction observed in the context of T1D and menopause.

Postmenopausal and T1D women experience reduced vaginal lubrication, diminished vaginal sensation and genital arousal, and vaginal atrophy.^36–38 In postmenopausal women, changes in vaginal function and structure are primarily attributed to changes in circulating levels of sex steroid hormones, with marked decline in estrogen levels and also changes in androgens, in particular of dehydroepiandrosterone.^{37, 39–41} Oxidative stress, damage to the vascular, autonomic and peripheral nervous systems, as well as lower levels of estrogen are viewed as the main factors leading to vaginal dysfunction in T1D.^{38, 42, 43} Little is known, however, of the molecular mechanisms responsible for the effects of menopause and T1D on vaginal function. Studies with animal models and with human genital tissues have shown that these conditions are accompanied by reduced levels of vaginal endothelial and neuronal nitric oxide synthase, arginase I and downregulation of sex steroid hormone receptors.^{42, 44–48} These findings are in line with the proposed disruption of sex steroid hormone signaling in vaginal T1D and postmenopausal tissues, and involvement of the NO/cGMP pathway in decreased genital arousal responses, given its recognized role in regulating vaginal blood flow and engorgement during sexual arousal. ^{40, 49, 50} In T1D women, a marked decline in vaginal responses to vibration has been reported and proposed to result, at least in part, from the degeneration of the vaginal myelinated sensory fibers.^{51, 52} Our findings that vaginal epithelial Panx1 expression is markedly downregulated not only in the diabetic Akita mice but also in the surgically castrated mice, indicate that these mechanosensitive channels may provide a molecular substrate for the observed decrease in vaginal sensation in T1D and postmenopausal women. Moreover, because downregulation of Panx1 channels would also decrease purinergic signaling within the vaginal wall, it is conceivable that these channels are also indirectly involved in mechanism that can lead to reduced vaginal moisture in T1D and menopause by attenuating the P2Y₂R-mediated vaginal lubrication.

This study focused primarily on investigating whether mechanical stimulation of the vaginal canal, as occur during sexual penetration, triggers ATP release and determining if this response involved activation of mechanosensitive Panx1 channels and was altered in conditions known to be associated with vaginal dysfunction and risk for impaired genital arousal. Data obtained from wildtype and Panx1-null mice demonstrated that Panx1 channels and ATP participate in mechanisms of vaginal epithelial mechanosensation and mechanotransduction. In addition, we demonstrated that mechanosensitive ATP release was reduced in T1D Akita and in OVX mice, and that this response is accompanied by marked downregulation of vaginal epithelial Panx1 expression. Previous studies reporting on the involvement of P2 receptors in mechanisms that regulate vaginal moisture and vaginal smooth muscle tonus allowed us to make inferences as to the potential impact of Panx1 expression and the role of mechanosensitive ATP signaling on vaginal function and dysfunction. The data presented in this study are limited in their support for this proposed role of Panx1 channels and vaginal epithelial mechanosensitive ATP signaling in mechanisms of vaginal function and dysfunction, including risk for FGAD in T1D and in menopause. Nonetheless, this study brings a new focus into this area of investigation and provides the basis for future studies that should address these points and further our understanding of the role of the vaginal mechanosensory and mechanotransduction system in peripheral mechanisms that increase the risk and ultimately contribute to emergence of FGAD.

We also acknowledge an important limitation of this study. We did not measure the plasma levels of sex steroid hormones to confirm the hypo-estrogenic state of the ovariectomized mice. Instead, we relied on the structural and histological changes that occur in the uterus and vagina as a readout to validated the decline in hormonal levels following surgical castration. Changes in sex steroid hormonal levels have also been reported in animal models of diabetes. In Streptozotocin-induced T1D rats, a significant decrease in estrogen and increase in testosterone levels was observed, and histological examination of vaginal tissue cross-sections was indicative of vaginal atrophy.⁴² We did not observed any apparent differences in the vaginal tissue structure of T1D Akita when compared to control female mice. It is likely that the decline in mechanosensitive vaginal epithelial responses reported here for the T1D Akita mice occur early and precede the late effects of estrogen deprivation on vaginal structure. However, quantification and comparison of circulating levels of not only estrogen but also androgens in OVX and T1D mice would be interesting as would allow correlation of changes in sex steroid hormone levels with those observed in mechanosensitive vaginal epithelial responses and Panx1 expression. Such comparisons could provide insights into a possible common role of sex steroid hormone signaling in mechanisms that regulate vaginal epithelial Panx1 expression in T1D and in menopause. In previous studies with bone cell lines, we shown that exposure to high extracellular glucose levels is one of the factors regulating Panx1 expression in the T1D bone.³⁴ Our ongoing studies also indicate that exposure to high extracellular glucose is also one of the factors regulating Panx1 expression in the T1D bladder urothelium. Future studies examining the effects of sex steroid hormone and insulin therapy in rescuing vaginal epithelial Panx1 expression and mechanosensitive ATP release are currently planned and should clarify whether common or independent mechanisms regulate Panx1 expression in the context of T1D and menopause.

Conclusions

This study provides evidence for the participation of ATP and Panx1 channels in mechanisms of vaginal mechanosensation and mechanotransduction. It advances our understanding of vaginal intrinsic regulatory mechanisms and identifies the Panx1 channel as one of the components of the vaginal epithelial mechanosensory and mechanotransduction system. This study also indicates that Panx1 channels have the potential of becoming a new molecular target for development of alternative therapies to manage vaginal dysfunction and improve distressing symptoms of FGAD in T1D and in postmenopausal women.