Development of a Vaginal Extracellular Matrix Hydrogel for Combating Genitourinary Syndrome of Menopause

Abstract

Genitourinary syndrome of menopause (GSM) is initiated by the hypoestrogenic state consequent to cessation of ovarian function and involves changes in all vaginal layers. These alterations include decreased epithelial proliferation and differentiation, impaired extracellular matrix (ECM) maintenance, pro‐inflammatory immune milieu, and smooth muscle atrophy. Patient satisfaction with existing treatments for GSM is low, due in part to their lack of therapeutic effect throughout all vaginal layers and reliance on hormonal treatments. Decellularized ECM therapeutics are known to facilitate soft tissue repair in a variety of applications. Thus, a porcine vaginal tissue‐derived decellularized ECM (vECM) hydrogel suitable for noninvasive topical intravaginal administration is developed specifically for vaginal repair. When applied in a rat model of surgical menopause, vECM significantly improves vaginal epithelial thickness and epithelial stem cell phenotype. Despite topical intravaginal application, vECM is also observed in the lamina propria and fibromuscularis, exerting a regenerative effect on the vaginal smooth muscle layer. Furthermore, vECM modulates macrophage density and promotes pro‐regenerative M2‐like phenotype within the vaginal tissue. Overall, the work herein demonstrates a new nonhormonal biomaterial treatment that counteracts pathological vaginal alterations – a hallmark of GSM – in an established preclinical model of menopause.

In this study, a vaginal specific ECM hydrogel (vECM) is developed, optimized for intravaginal topical administration, and evaluated for mitigatation of vaginal atrophy associated with menopause. When applied in a rat model of menopause, vECM improves vaginal epithelial thickness and stem cell phenotype, and despite topical application, vECM also exerts a regenerative effect on the deeper smooth muscle layer.

1. Introduction

Genitourinary syndrome of menopause (GSM) refers to the pathological changes in the vulvovaginal and lower urinary tract tissues as a result of the hypoestrogenemia in perimenopausal and menopausal women.^{[ 1 ]} Up to 85% of women over 40 years of age are affected by some form of GSM, with 30–60% of women with GSM suffering from vaginal pruritis and irritation.^{[ 1 ]} Not surprisingly, women with GSM report that this condition negatively affects their sexual activities, interferes with their enjoyment of life and sleep, and hinders traveling, work, and athletic and social activities.^{[ 2 ]} Given that an estimated 47 million women become menopausal every year^{[ 3 ]} and that the expected lifespan is continuously increasing, this population is rapidly expanding.^{[ 4 ]} As such, GSM represents a major public health issue, making it imperative to design accessible and effective treatments for this morbid condition that interferes with healthy aging.

GSM is initiated by the hypoestrogenic state consequent to the cessation of ovarian function, defined as menopause.^{[ 1 ]} Estrogen deprivation results in decreased proliferation and differentiation of vaginal epithelial cells, leading to the disappearance of vaginal rugae and leaving a thin layer of basal epithelium.^{[ 5} , ^{6 ]} Further pathological changes include diminished vascularization, reduced smooth muscle content and contractility within the fibromuscularis, and biochemical and biomechanical alterations of the extracellular matrix (ECM).^{[ 7} , ^{8 ]} Lastly, it is known that immune cells of the female reproductive tract are responsive to estrogen; in the hypoestrogenic state, increased expression of pro‐inflammatory cytokines and a reduction in important immune populations, including T‐regulatory cells, are hallmarks of the GSM phenotype.^{[ 9 ]} The immune system is intricately involved with ECM remodeling^{[ 10 ]} and both immune activation and ECM remodeling within vaginal tissues are impacted by estrogen deprivation in menopause.^{[ 11 ]}

Considering the complex cascade of events consequent to the hypoestrogenic state, none of the existing treatments mitigate pathological alterations in every vaginal layer when used alone.^{[ 2 ]} Low‐dose vaginal estrogen therapy is the gold‐standard GSM treatment, and it is highly effective at restoring epithelium through estrogen‐sensitive immunomodulation pathways; however, it fails to improve the fibromuscular phenotype.^{[ 12 ]} Furthermore, despite the benefits and established safety of low‐dose vaginal estrogen, a number of patients and even some clinicians still reject this treatment because of the concerns of increased risk of hormone‐sensitive cancers, especially with the cream formulation of vaginal estrogen.^{[ 13 ]} Importantly, menopause can alter the expression of estrogen receptors in genitourinary tissues, further compromising the effectiveness of estrogen‐based therapeutics for GSM.^{[ 14 ]} Alternatively, laser therapy has shown some efficacy for repairing the fibromuscularis, but side effects of epithelial burns, scarring, and chronic pain incited an FDA warning against this treatment.^{[ 15 ]} Importantly, only 35% of women suffering from GSM are satisfied with their treatment, whether prescribed or over‐the‐counter.^{[ 2 ]}

Some biomaterials have been used for the treatment of GSM or are under investigation. For example, hyaluronic acid gels have been used to alleviate vaginal dryness and discomfort, though these are inferior to vaginal estrogen and do not exert a regenerative effect on vaginal tissues.^{[ 16 ]} Additionally, one group explored intravaginal administration of a recombinant collagen‐derived adhesion protein and demonstrated increased epithelial proliferation and differentiation following treatment in a rat model of menopause.^{[ 17 ]} Our lab previously demonstrated that collagen‐based biomaterials are inferior in inducing regeneration compared to tissue‐specific ECM hydrogels.^{[ 18 ]} Decellularized ECM hydrogels have been extensively investigated as pro‐regenerative scaffolds in various pre‐clinical tissue engineering applications^{[ 19 ]} and have demonstrated safety and feasibility in a Phase I Clinical trial (clinicaltrial.gov NCT02305602) for the treatment of myocardial infarction.^{[ 20 ]} Decellularized ECM hydrogels have demonstrated tissue regenerative capabilities in numerous injury and disease models through mechanisms of cellular recruitment, increased cellular proliferation and differentiation, immune modulation, neovascularization, and ECM remodeling.^{[ 19 ]} However, these applications have mainly utilized needle‐based injections into tissue rather than topical applications in hollow organs, and none have been applied intravaginally. In this study, we set out to develop a vaginal specific ECM hydrogel (vECM), optimize it for intravaginal topical administration, and evaluate whether it could mitigate vaginal atrophy and other alterations in vaginal phenotypic properties associated with menopause.

2. Results

2.1. Fabrication and Characterization of vECM Hydrogel

Given the lack of existing published protocols for a decellularized vaginal tissue‐derived ECM hydrogel, we started with extensive optimization of a new decellularization protocol, informed by our successful workflow used to develop ECM hydrogels from cardiac and skeletal muscles.^{[ 21 ]} A summary of the decellularization conditions tested in iterative optimization batches is shown in Table 1 . Notably, the decellularization conditions previously utilized for cardiac and skeletal muscle hydrogels (Table 1, Row A) did not result in successful gelation of the material and had high residual sodium dodecyl sulfate (SDS). In the production of decellularized ECM hydrogels, over‐decellularization (too high detergent concentration or too long decellularization) is a main contributor to failed gelation due to the stripping of critical ECM proteins necessary for the formation of a hydrogel.^{[ 22 ]} Therefore, in subsequent iterative decellularization optimization batches, SDS concentration, decellularization duration, and water rinsing steps were varied, as shown in Table 1, with the goal of minimizing residual SDS and achieving successful decellularization and gelation. A reduced duration of decellularization using 1% SDS also failed gelation (Table 1, Row B), after which lower concentrations of SDS were tested (Table 1, Row C, D). Of these conditions, only the 3‐day decellularization using 0.1% SDS passed the gelation test. We next modified the tissue processing step to use a meat grinder instead of hand‐chopping to improve decellularization throughput and consistency; this process also resulted in successful gelation (Table 1, Row E). In addition to gelation testing, all batches were evaluated for residual DNA, residual SDS, and sulfated glycosaminoglycan (sGAG) content. Residual DNA and sGAG content were at acceptable levels in all batches, indicating thorough decellularization without stripping important biochemical components of the ECM^{[ 21 ]} while residual SDS was acceptable in all conditions except the first batch (Table 1, Row A).^{[ 23 ]}

Table 1.

Decellularization conditions tested in optimization batches.

	Tissue Processing	Decell. Conditions	Water Rinsing	Gelation Success	dsDNA (ng mg−1)	% SDS content	sGAGs (µg mg−1)
A	Minced with a knife	5 days 1% SDS	1 day	No	1.6 ± 0.4	0.018 ± 0.001	14.5 ± 0.6
B	Minced with a knife	3 days 1% SDS	3 days	No	1.5 ± 0.2	0.0016 ± 0.0001	9.6 ± 0.7
C	Minced with a knife	3 days 0.5% SDS	3 days	No	1.5 ± 0.2	0.0017 ± 0.0001	13 ± 3
D	Minced with a knife	3 days 0.1% SDS	3 days	Yes	1.6 ± 0.1	0.0043 ± 0.0009	10.8 ± 0.8
E	Meat grinder	3 days 0.1% SDS	3 days	Yes	0.4 ± 0.4	0.0081 ± 0.0005	10.6 ± 0.2

Ultimately, the vaginal decellularization protocol necessitated a much lower concentration of SDS for a shorter duration – 0.1% for 3 days – with a longer water rinsing period of 3 days to remove residual SDS, as compared to established cardiac and skeletal muscle decellularization protocols. The final decellularization protocol is depicted in Figure 1 . To further evaluate this final protocol, we performed hematoxylin and eosin staining on tissue cross‐sections taken throughout the decellularization process, and found that the ECM structure was maintained while nuclear content was sufficiently removed from the tissue (Figure 2A–D).

Figure 1.

Production of a porcine vaginal tissue‐derived ECM hydrogel (vECM). A) Whole porcine vaginas were cleaned and frozen B) before being minced with a commercial meat grinder, C) then spun in SDS to decellularize the tissue. D) Decellularized tissue was lyophilized and E) milled to a fine and unform powder. F) Decellularized vECM powder was partially enzymatically digested with pepsin, then pH and salt balanced before incubating at 37 °C to form G) a stable hydrogel.

Figure 2.

Characterization of vECM structure, composition, and material properties. Vaginal samples were taken from A) fresh tissue or tissue after either B) 1, C) 2, or D) 3 days of decellularization. Tissue samples were cryosectioned and stained with hematoxylin and eosin to demonstrate removal effective decellularization. E) Porcine vaginal tissue and decellularized vECM were analyzed via ECM targeted proteomics using QConCAT (n = 3/group). F) 6 and 8 mg mL⁻¹ vECM preparations (n = 3/group) underwent viscosity testing with a parallel plate rheometer, demonstrating shear thinning properties for both groups, and an increased viscosity in the 8 mg mL⁻¹ group at lower shear rates. G) vECM was prepared at 6 and 8 mg mL⁻¹ concentrations, and n = 3 wells per concentration underwent a turbidity assay, in which optical density (OD) was read at 405 nm for two hours or until equilibrium. H) The time to 50% equilibrium OD was determined for 6 and 8 mg mL⁻¹, demonstrating a significantly lower time for the 8 mg mL⁻¹ group.

We next analyzed vECM and native porcine vaginal tissue with ECM targeted proteomics using quantitative concatemers (QConCAT) as previously described.^{[ 24 ]} Proteins were identified and categorized according to ECM function, and relative intensities of each category were found for porcine vaginal tissue and decellularized vECM (Figure 2E). vECM shared overall similar distribution in proteomic content, with a larger proportion of fibrillar collagen and a smaller proportion of crosslinking enzyme, which is expected from the decellularization process.

vECM was then prepared at two concentrations, 6 and 8 mg mL⁻¹, which we have shown in previous studies to be viable concentrations for cellular interactions.^{[ 25 ]} While 6 mg mL⁻¹ has been determined to be the optimal concentration for intramuscular injections in terms of injectability, spread, and degradation^{[ 25 ]}, the intravaginal application is unique in that it is applied topically in the lumen, and thus we would not anticipate gelation into a solid hydrogel given the constant motion. As such, the 8 mg mL⁻¹ formulation was tested in anticipation that increased viscosity may improve retention and efficacy in this application. We tested the viscosity (Figure 2F) of 6 and 8 mg mL⁻¹ vECM, as well as turbidity to evaluate self‐assembly at 37 °C (Figure 2G). The 8 mg mL⁻¹ demonstrated higher viscosity than 6 mg mL⁻¹ vECM, although at a level still amenable to delivery via a syringe applicator, and both materials were shear thinning. As expected, 8 mg mL⁻¹ vECM demonstrated faster self‐assembly (see time to 50% OD, Figure 2H), and a higher equilibrium OD than 6 mg mL⁻¹.

2.2. vECM Improves Epithelial Thickness and Morphology in Ovariectomized Rats

We next conducted a study to assess the efficacy of intravaginally administered vECM for the restoration of vaginal epithelium in ovariectomized (OVX) rats. The study design (Figure 3A) utilized rats that were two weeks post‐OVX, which is a well‐established model of surgically induced menopause.^{[ 26 ]} Uterine horns were excised from all specimens along with vaginal tissues to visually confirm menopausal state for animals that underwent OVX procedure in comparison to healthy (unperturbed) controls (Figure S1, Supporting Information). As epithelial atrophy is a hallmark of GSM,^{[ 26 ]} we also demonstrated that the epithelial thickness of OVX animals was significantly less than healthy controls (Figure S2, Supporting Information), further validating this model for the present study.

Figure 3.

Evaluation of vECM efficacy in restoring vaginal epithelium in menopausal rats. A) Experimental timeline to assess the therapeutic efficacy of intravaginal administration of vECM, alongside saline and collagen, in a rat model of surgical menopause via OVX. Vaginal tissue sections, isolated from healthy controls and OVX animals after 14 days of treatment, were stained with hematoxylin and eosin to assess vaginal epithelial thickness. Representative images are shown for B) healthy (unperturbed) controls and OVX rats treated with C) saline, D) collagen, E) 6 mg mL⁻¹ vECM, or F) 8 mg mL⁻¹ vECM. G) Data for epithelial thickness were averaged per animal and compared among groups with a one‐way ANOVA and Tukey's pairwise comparisons. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; n = 3–9/group.

Animals were randomly assigned to receive daily topical intravaginal administration of 500 µL of vECM at two concentrations (6 or 8 mg mL⁻¹), collagen, or saline. Collagen was chosen as a control, given that it is the predominant component of vECM. The concentration of collagen was chosen to match the low‐shear‐rate viscosity of 8 mg mL⁻¹ vECM (Figure S3, Supporting Information) given the liquid intravaginal delivery in this application.

Representative tissue sections for healthy controls and OVX animals treated with either saline, collagen, 6 mg mL⁻¹ vECM, or 8 mg mL⁻¹ vECM are shown in Figure 3B‐F. Treatment with the higher concentration 8 mg mL⁻¹ vECM resulted in significantly greater epithelial thickness compared to saline, collagen, and 6 mg mL⁻¹ vECM (Figure 3G), indicating that vECM has a dose‐dependent effect on epithelium in menopausal rats.

We next assessed epithelial stem cells in vaginal tissue sections from healthy controls and OVX animals treated with either saline, collagen, or 8 mg mL⁻¹ vECM. Previous studies have demonstrated that hypoestrogenism leads to decreased proliferation and differentiation of vaginal epithelial stem cells and a consequent decrease in vaginal epithelial stratification.^{[ 6 ]} We showed that in premenopausal controls, the p63⁺ epithelial stem cells constitute a relatively small portion of the total epithelial cells (Figure 4A). Conversely, in OVX saline‐treated animals, the p63⁺ stem cells occupy almost the entirety of the vaginal epithelium, with very little differentiated epithelium present (Figure 4B). Importantly, vECM [8 mg mL⁻¹] treatment led to restoration of a healthy‐appearing vaginal epithelial phenotype with respect to nuclei density (number of total nuclei normalized by epithelial area, Figure 4E), p63⁺ nuclei density (number of epithelial stem cell nuclei normalized by epithelial area, Figure 4F), and proportion of p63⁺ nuclei (number of epithelial stem cell nuclei normalized by total number of nuclei, Figure 4G). These analyses demonstrate that vECM treated animals display morphologically similar vaginal epithelium to premenopausal animals and suggest that vECM enables epithelial stem cells to proliferate and differentiate as they would in healthy animals.

Figure 4.

Characterization of vaginal epithelial stem cells in vECM treated animals. Vaginal tissue sections, isolated from healthy controls and OVX animals after 14 days of treatment, were stained against nuclei (DAPI, blue), an epithelial stem cell transcription factor (p63, magenta), and an epithelial membrane marker (filaggrin, green) to identify epithelial stem cells. Representative immunofluorescent images are shown for A) healthy (unperturbed) controls and OVX rats treated with B) saline, C) collagen, or D) 8 mg mL⁻¹ vECM. Vaginal epithelium was assessed for E) nuclei density, F) p63⁺ vaginal epithelial stem cell nuclei density, and G) percentage of nuclei within the epithelium that were p63⁺ vaginal epithelial stem cells. Data were compared among groups with a one‐way ANOVA and Tukey's pairwise comparisons **P < 0.1; ***P < 0.001; ****P < 0.0001, n = 3–9/group.

2.3. vECM Modulates Smooth Muscle Thickness and Proliferation

In addition to epithelial atrophy, smooth muscle atrophy is a notable consequence of the hypoestrogenic state.^{[ 27 ]} Thus, we conducted immunohistochemical staining against alpha smooth muscle actin (ɑSMA) to assess the smooth muscle layer thickness in all groups. Representative immunohistochemical staining for healthy controls, and saline, collagen, 6 mg mL⁻¹ vECM, and 8 mg mL⁻¹vECM treated groups are shown in Figure 5A‐E, and quantitative comparison of these groups is shown in Figure 5F. Topical 8 mg mL⁻¹ vECM treatment resulted in significantly greater smooth muscle thickness compared to saline and collagen, with no significant differences relative to the healthy controls or 6 mg mL⁻¹ vECM treated OVX animals.

Figure 5.

Evaluation of vECM efficacy in restoring vaginal smooth muscle in menopausal rats. Vaginal tissue sections, isolated from healthy controls and OVX animals after 14 days of treatment, were stained against nuclei (DAPI, blue) and smooth muscle (ɑSMA, yellow). Representative immunofluorescent images are shown for A) healthy (unperturbed) controls and OVX rats treated with B) saline, C) collagen, D) 6 mg mL⁻¹ vECM, or E) 8 mg mL⁻¹ vECM. F) Data for smooth muscle thickness were averaged per animal and compared among groups with a one‐way ANOVA and Tukey's pairwise comparisons. **P < 0.01, n = 3–9/group.

Tissue sections were additionally stained against the nuclear proliferation marker Ki67 alongside ɑSMA to assess the amount of proliferating smooth muscle cells. Representative immunohistochemical staining for healthy and saline, collagen, and 8 mg mL⁻¹ vECM treated groups are shown in Figure 6A–D, and quantitative comparison is shown in Figure 6E,F. vECM treatment significantly improved both the density of nuclei within the smooth muscle layer (Figure 6E) and the percentage of proliferating nuclei within the smooth muscle (Figure 6F) compared to saline and collagen treatment, and was also greater than the baseline level in healthy controls. These data suggest that topical vECM treatment impacted smooth muscle regeneration and phenotype.

Figure 6.

Assessment of vaginal smooth muscle cell proliferation following treatment. Vaginal tissue sections, isolated from healthy controls and OVX animals after 14 days of treatment, were stained against nuclei (DAPI, blue), smooth muscle actin (ɑSMA, green), and a nuclear marker of proliferating cells (Ki67, cyan). Representative immunofluorescent images are shown for A) healthy (unperturbed) controls and OVX rats treated with B) saline, C) collagen, or D) 8 mg mL⁻¹ vECM. Groups were assessed for E) Ki67+ nuclei density within the smooth muscle layer and F) proportion of Ki67+ nuclei over total nuclei within the smooth muscle layer. Data were averaged per animal and compared among groups with a one‐way ANOVA and Tukey's pairwise comparisons (F). *P < 0.05; ****P < 0.0001, n = 3–9/group.

2.4. Material Retention, Distribution, and Macrophage Interactions

Given the dose‐dependent response of vaginal epithelial and smooth muscle regeneration with daily administration of topical biomaterial in this model, we conducted a short‐timepoint material retention study. A single 500 µL dose of saline or fluorescently pre‐labeled collagen, 6 mg mL⁻¹ vECM, or 8 mg mL⁻¹ vECM was delivered topically via intravaginal administration. Vaginal specimens were isolated at 1, 2, or 3 days post‐administration to assess material spread and retention throughout the vaginal tissues. Fluorescently pre‐labeled material was observed in all material groups (Figure 7A–C, cyan); interestingly, material was not only observed in the vaginal lumen, where it was delivered, but also in the fibromuscular layer. Material retention throughout the vagina was quantified by averaging the area of material across all vaginal sections per animal (Figure 7D), but no significant differences in material retention were observed between groups at any time point. Given the observation of material in the fibromuscularis, the frequency of observation was quantified by the percentage of sections that contained material in the fibromuscularis (Figure 7E). This phenomenon was observed in all material groups, but only 8 mg mL⁻¹ vECM had significant increases over collagen over the 3 days. Overall, all three biomaterial groups demonstrated similar retention and localization in the vaginal lumen, with diminished retention at the 3 day timepoint, with some differences in fibromuscular localization between collagen and 8 mg mL⁻¹ vECM groups.

Figure 7.

vECM at two doses and collagen control demonstrate similar retention in the vaginal lumen and fibromuscularis. Two weeks following OVX, rats were treated with a single 500 µL intravaginal topical administration of fluorescently pre‐labeled A) collagen, B) 6 mg mL⁻¹ vECM, or C) 8 mg mL⁻¹ vECM. Tissue sections were harvested at 1, 2, or 3 days post‐administration (n = 2/group/timepoint). Representative tissue sections are shown for each group, in which fluorescently tagged biomaterial is shown in cyan, and tissues were stained against cell nuclei (DAPI, blue). D) Tissue sections were assessed for material retention by assessing the average area of material throughout the tissue. E) Additionally, material from all groups demonstrated material in the vaginal fibromuscularis; the frequency of this observation was reported as a percentage of sections. F) Sections were stained against macrophage marker (CD68, red), which was observed colocalized with fluorescently pre‐labeled materials in both the vaginal lumen and fibromuscularis in all groups.

As both collagen and vECM materials are too large to passively bypass the vaginal epithelial layer, we hypothesized that cellular trafficking may be aiding the transfer of materials from the vaginal lumen to fibromuscularis. Indeed, in all events where vECM or collagen was observed in subepithelial layers, it was co‐localized with nuclei. To further investigate, tissue sections from all groups were stained against macrophages, as they are a cornerstone of the immune response to biomaterials. In particular, macrophages demonstrate a unique response to acellular ECM scaffolds^{[ 28 ]} and facilitate the pro‐regenerative activity of ECM biomaterials by activating endogenous tissue repair mechanisms.^{[ 28} , ^{29 ]} Representative tissue sections from each group are shown in Figure 7F, in which fluorescently pre‐labeled collagen and vECM are shown in cyan, nuclei are shown in blue (DAPI), and macrophage membrane markers are shown in red (CD68, pan‐macrophage marker). In all groups, macrophages were seen in both the vaginal lumen and fibromuscularis, colocalized with the administered materials.

Given that the data suggested the interaction of vECM and collagen with macrophages, macrophage density in vaginal tissues was assessed at the end of treatment, utilizing tissues from the efficacy study previously discussed. Tissues were stained against CD68 (pan‐macrophage marker) and CD163 (M2 macrophage marker). Representative immunofluorescent images for healthy controls (Figure 8A) and OVX animals treated with saline (Figure 8B), collagen (Figure 8C), or 8 mg mL⁻¹ vECM (Figure 8D) are shown. The 6 mg mL⁻¹ vECM group was not assessed for macrophages as this group did not demonstrate improvement in either epithelial or smooth muscle phenotype. There was no significant difference among groups in nuclei density (Figure 8E), but the vECM‐treated group demonstrated significantly greater proportion of macrophages overall (Figure 8F) and specifically M2‐like macrophages (Figure 8G) compared to saline and collagen‐treated groups. The vECM‐treated group was not significantly different from healthy controls in either metric, although healthy controls demonstrated higher macrophage and M2‐like macrophage densities compared to saline‐treated animals. These data indicate that vECM modulates the immune response in vaginal tissues to a healthier phenotype.

Figure 8.

Assessment of macrophage infiltration in vaginal tissues after various treatments. Vaginal tissue sections, isolated from healthy controls and OVX animals after 14 days of treatment, were stained against nuclei (DAPI, blue), a pan‐macrophage membrane marker (CD68, red), and an M2 macrophage membrane marker (CD163, green). Representative immunofluorescent images are shown for A) healthy (unperturbed) controls and OVX rats treated with B) saline, C) collagen, or D) vECM 8 mg mL⁻¹. Groups were assessed for E) nuclei density within the lamina propria and muscularis, F) proportion of CD68+ nuclei, indicating macrophages, and G) proportion of CD163+ nuclei, indicating anti‐inflammatory M2‐like macrophages. Data were averaged per animal and compared among groups with a one‐way ANOVA and Tukey's pairwise comparisons. *P < 0.05; ***P < 0.001; ****P < 0.0001, n = 3–9/group.

3. Discussion

More than half of menopausal women experience vaginal pruritis, dryness, pain, and discomfort associated with GSM. Unfortunately, the only effective treatment, vaginal low‐dose estrogen, is not effective for restoring fibromuscular phenotype in GSM, and some reject it out of fear of hormone‐associated cancer risks despite existing evidence to the contrary.^{[ 13 ]} Thus, we aimed to develop a nonhormonal, acellular, and easily manufactured pro‐regenerative biomaterial as an alternative treatment for vaginal changes associated with GSM.

While other groups have investigated acellular vaginal scaffolds for vaginal reconstruction^{[ 30 ]} or 3D printed vaginal tissue analogues incorporating gelatin and sodium alginate,^{[ 31 ]} to the best of our knowledge, no group has published on an acellular vaginal tissue‐derived ECM hydrogel, the physical properties of which make this material well suited for topical intravaginal administration. Additionally, ECM hydrogels have not been previously examined for this application, which is uniquely challenging as the intravaginal environment, owing to constant motion in an open cavity, does not allow for true gelation of the material. In the context of other ECM biomaterials, our work demonstrated that optimal decellularization conditions for vaginal tissue differed considerably from those of striated muscle^{[ 21 ]} and small intestinal submucosa^{[ 32 ]} while sharing similarities with some protocols for skin decellularization.^{[ 33 ]} As the vagina has a relatively low cellular content and is largely composed of ECM, we found that vaginal tissue required gentler decellularization and a more thorough water rinse to preserve ECM structure and sufficiently remove residual detergent. Considering the differences in ECM structure and composition between tissue sources,^{[ 34 ]} the differing therapeutic effect of ECM hydrogels depending on tissue source,^{[ 35 ]} as well as evidence for the superiority of vaginal derived materials over SIS in vaginal reconstruction,^{[ 36 ]} we chose to test a tissue‐specific biomaterial for this novel application of an ECM hydrogel for intravaginal topical delivery in a model of menopause.

In the present study to assess the therapeutic potential of vECM, we mirrored the treatment regimen of locally delivered vaginal estrogen creams, which are applied once daily for a period of two weeks, after which 2–3 applications per week are used for maintenance.^{[ 37 ]} Locally delivered estrogen, the current clinical gold standard of treatment for vaginal atrophy associated with GSM, activates estrogen‐sensitive pathways in the vaginal tissues to promote epithelial growth and maturation index^{[ 7 ]} though neither local nor systemic estrogen treatment effectively improves the fibromuscular layer.^{[ 27} , ^{38 ]} In the Sprague‐Dawley rat OVX model of menopause, estrogen typically returns epithelial thickness to healthy values, which range from 50–70 µm, compared to untreated or sham‐treated OVX controls, in which epithelial thickness ranges from 10–30 µm.^{[ 39 ]} We found that vECM treatment, at the higher dose, not only significantly improved epithelial thickness and returned epithelial phenotype to a healthier state, but additionally impacted vaginal smooth muscle thickness and proliferation. For both of these metrics, we observed a dose‐dependent response in regeneration; the 8 mg mL⁻¹ formulation of vECM performed better than the 6 mg mL⁻¹ formulation although the 8 mg mL⁻¹ was not able to completely restore epithelial thickness to healthy values; it is possible that even higher concentrations may deliver improved results.

Decellularized ECM biomaterials in various formulations, including hydrogels, particulates and sheets, promote cellular proliferation, differentiation, and viability^{[ 35} , ^{40 ]} as well as ECM remodeling^{[ 41 ]} immunomodulation,^{[ 42} , ⁴³ , ^{44 ]} and vascularization.^{[ 35} , ^{42 ]} Our data demonstrate increased macrophage density, specifically M2‐like macrophages, within the vaginal fibromuscularis, which supports the extensive literature describing the macrophage‐mediated pro‐repair response to ECM biomaterials.^{[ 41} , ^{44 ]} Additionally, it is well‐established that ECM components are critical for proper epithelial growth and stratification^{[ 45 ]} and ECM biomaterials can support epithelial regeneration in wound healing,^{[ 46} , ^{47 ]} lung radiation,^{[ 48 ]} and intestinal perforation.^{[ 49 ]} The data from the present study support these findings, as we have demonstrated that vECM promotes vaginal epithelial proliferation and stratified differentiated epithelium.

The impact of vECM on the subepithelial tissues are likely due to the observed vECM localization in lamina propria and fibromuscularis, where the material colocalized with cell nuclei in all observations, oftentimes with co‐staining of macrophage membrane markers. vECM and collagen biomaterial components are too large to passively bypass the vaginal epithelium at the acute timepoints; as such, we hypothesize that cells may be trafficking vECM and collagen through the epithelium and into subepithelial tissues. One study demonstrated that macrophages can uptake synthetic biomaterials, which then facilitate intracellular modulation of macrophage activity.^{[ 50 ]} Overall, cellular interactions and localization were consistent between all biomaterial groups, suggesting that the superior therapeutic efficacy of the 8 mg mL⁻¹ vECM may be due to the increased concentration of bioactive ECM components. Decellularized ECM components have been shown to polarize macrophages to an M2‐like phenotype and stimulate endogenous tissue repair mechanisms in a variety of applications, such as spinal cord injury,^{[ 51 ]} wound healing,^{[ 46 ]} and osteochondral defects.^{[ 52 ]} Furthermore, various studies have explored the pivotal role of macrophages in epithelial remodeling and regeneration in nasal^{[ 53 ]} colonic,^{[ 54 ]} and lung^{[ 55 ]} epithelia as well as smooth muscle regeneration.^{[ 56 ]} In the present study, after 2 weeks of treatment, the 8 mg mL⁻¹ vECM group demonstrated significant recruitment of macrophages and polarization to M2‐like phenotype in vaginal tissues; this supports our hypothesis that the tissue regeneration observed in the epithelium and fibromuscularis may result from pro‐regenerative immune modulation. Overall, the literature and current study encourage further development of decellularized ECM formulations beyond injection to mediate immune modulation and regeneration of unique soft tissue environments, such as the vagina.

4. Conclusion

We developed and characterized a vaginal tissue‐derived decellularized ECM hydrogel and found that when delivered intravaginally as a topical application in a rat model of surgical menopause, it improved epithelial thickness and phenotype, smooth muscle thickness and proliferation, and recruited macrophages, which were polarized to a pro‐regenerative phenotype. Furthermore, we found that biomaterial formulations are well retained in the vagina and are even trafficked to subepithelial tissues following intravaginal topical administration. In conclusion, we have demonstrated the development and application of a novel acellular biomaterial therapeutic for vaginal atrophy associated with GSM.

5. Experimental Section

Fabrication and Characterization of vECM Hydrogel

Porcine vaginal tissues were acquired from young female pigs sourced from Midwest Research Swine (Collagen Solutions). Several rounds of decellularization optimization were conducted, in which the following conditions were tested: tissue processing method (hand‐chopped or meat grinder), SDS concentration (0.1, 0.5, or 1%), duration of decellularization (3 or 5 days), and duration of water rinse (1 or 3 days). For all decellularization conditions, decellularized ECM was lyophilized, milled, and passed through a #60 filter. Milled and filtered vECM underwent pepsin digestion at a 10 mg ECM per mL pepsin solution in 0.1 m HCl for 48 h. Digested vECM was pH and salt balanced, then incubated at concentrations of 4, 6, or 8 mg mL⁻¹ at 37 °C for 24 h to test gelation. vECM was screened for residual nuclear material, using PicoGreen (Thermo Fisher) fluorescent assay; residual SDS, using methylene blue assay; and sulfated glycosaminoglycans, using 1,9‐dimethylmethylene blue assay.

ECM‐specific protein makeup of fresh tissue and decellularized powder was assessed via ECM targeted proteomics (n = 3/group) using quantitative concatemers as previously published.⁷⁰ Sample preparation and LC‐MS/MS analysis were completed utilizing previously developed and published methods.^{[ 57 ]} Data were then searched using MSFragger v3.8 via FragPipe v20.0 against UniProt as previously described.^{[ 58 ]} Results were filtered to 1% FDR at the peptide and protein levels.

To evaluate vECM mechanical properties, vECM was prepared at 6 and 8 mg mL⁻¹ concentration. Optical density was used to evaluate the turbidimetric gel kinetics of vECM. 100 µL of vECM was pipetted into a 96‐well plate (n = 3 wells/concentration group). The plate was loaded into a preheated (37 °C) spectrophotometer, and absorbance was read at 405 nm every minute for two hours or until equilibrium was reached. To measure viscosity, 200 µL of vECM (n = 3 runs/concentration group) was dispensed on a parallel plate ARG2 rheometer stage (TA Instruments) at 25 °C. Gap height was set to 500 µm, and a flow procedure was executed with shear rate ranging from 0.1 – 1000 Hz.

Preparation of vECM and Collagen for Intravaginal Administration

Two different concentrations of vECM (6 mg mL⁻¹ or 8 mg mL⁻¹) were prepared for intravaginal administration by modulating water and salt (phosphate buffered saline solution) addition during the resuspension stage. As a non‐tissue specific control, collagen 1 derived from rat tail (Corning) was pH and salt balanced and used at a 6.5 mg mL⁻¹ concentration in order to match the rheological properties (specifically low‐shear rate viscosity) of vECM at the 8 mg mL⁻¹ concentration.

Intravaginal Administration of Saline, Collagen, and vECM

All procedures were approved by the Institutional Animal Care and Use Committee at the University of California, San Diego (AAALAC Accreditation #000503). A well‐established bilateral OVX rat model of surgical menopause (Charles River) was utilized. Two weeks after OVX, a duration previously established to be sufficient to induce changes driven by estrogen deprivation in this model,^{[ 26 ]} 3‐month‐old Sprague Dawley rats were anesthetized with 2.5% isoflurane in oxygen for the duration of the procedure and placed in a supine position. A 1 mL syringe without a needle was loaded with 500 µL of saline, collagen, or vECM at 6 mg mL⁻¹ or 8 mg mL⁻¹ concentration (n = 4‐9/group). The animals were left to rest for 10 min, after which they were returned to housing and observed until mobile. These treatments were carried out daily for a duration of two weeks. The day after the end of treatment, animals were euthanized, and full‐thickness vaginas were harvested, flash frozen, and stored in ‐80C for subsequent cryosectioning, staining, and analyses. Age‐matched non‐OVX animals were used as healthy controls (n = 3).

To assess retention and localization of vECM in the vagina, a separate group of OVX rats was used. For all biomaterial groups (vECM 6 mg mL⁻¹, vECM 8 mg mL⁻¹, and collagen) materials were mixed with an NHS Succinimidyl Ester fluorescently tagged with Alexa Fluor 647 and incubated at room temperature for 1 h to ensure complete binding of the dye to free amines within the vECM. Fluorescently prelabeled biomaterials (500 µL) were administered intravaginally as described above. Animals were euthanized 1, 2, or 3 days later (n = 2/time point), and vaginas were harvested and prepared for imaging and analyses.

Histological Assessment of Epithelial Thickness

Five tissue sections spanning the proximal and mid regions of the vagina were stained with hematoxylin and eosin and imaged with an Olympus VS200 slide scanning microscope at 20X magnification. Using QuPath v5 software, uniform gridlines were superimposed on the posterior portion of each tissue cross‐section, and 20–25 measurements of epithelial thickness were made where the epithelium intersected the grid. The measurements were averaged for each section and again across all sections for each biological replicate.

Histological Assessment of Epithelial Stem Cells

Tissue sections were incubated with DAPI nuclear stain, and antibodies against epithelial stem cell nuclear marker p63 (OriGene) and against a vaginal epithelium membrane marker filaggrin (Biorbyt), conjugated with an Alexa Fluor 647 or 488 secondary antibody (Invitrogen), respectively. Tissue sections were imaged with an Olympus VS200 slide scanning microscope at 40X magnification. Using QuPath v5 software, uniform gridlines were superimposed on 5 sections per specimen, sampling throughout the proximal and mid vagina, with five 250 µm x 250 µm grids of epithelium located in the posterior vagina analyzed per each tissue section. The region of epithelium was traced using positive filaggrin signal, and the QuPath cell counter was used to mark all nuclei within the grid to enable quantification of p63⁺ cells within the epithelial region. The overall nucleated cell density (number of nuclei/epithelium region area); density of epithelial stem cells within the epithelium (number of p63⁺ nuclei/epithelium region area), and proportion of epithelial stem cells (number of p63⁺ nuclei per total number of nuclei within the epithelium region area). For all metrics, data were averaged for each section and across all sections per specimen.

Histological Assessment of Vaginal Smooth Muscle Layer

Tissue sections were incubated with DAPI nuclear stain, and antibodies against alpha smooth muscle actin (Dako) and against a nuclear marker for proliferating cells Ki67 (Abcam), conjugated with an Alexa Fluor 568 or 647 secondary antibody (Invitrogen), respectively. Tissue sections were imaged with an Olympus VS200 slide scanning microscope at 40X magnification. Using QuPath v5 software, uniform gridlines were superimposed on 5 sections per specimen, sampling throughout the proximal and mid vagina, with 20–25 measurements analyzed per each tissue section in the posterior vagina to determine smooth muscle thickness. To quantitatively assess smooth muscle proliferation, five 250 µm x 250 µm grids of fibromuscularis located in the posterior vagina were analyzed per each tissue section. The region of smooth muscle within the grid was traced, and QuPath cell counter was used to count the total number of nuclei and the number of Ki67⁺ nuclei within the smooth muscle region. Data were averaged as described above.

Histological Assessment of Macrophage Infiltrate and Phenotype Induced by the Study Treatments

Tissue sections were incubated with DAPI nuclear stain, and antibodies against a pan‐macrophage membrane marker CD68 (Abcam) and against a marker of M2 macrophages CD163 (BioRad), conjugated with an Alexa Fluor 568 or 448 secondary antibody (Invitrogen), respectively. Tissue sections were imaged with an Olympus VS200 slide scanning microscope at 40X magnification. Using QuPath v5 software, uniform gridlines were superimposed on 5 sections per specimen, sampling throughout the proximal and mid vagina. To assess macrophage density, five 250 µm x 250 µm grids of fibromuscularis located in the posterior vagina were analyzed per each tissue section. QuPath cell counter was used to count the total number of nuclei, and the number of CD68⁺ and CD163⁺ nuclei within each grid. Data were averaged as described above.

For the short‐timepoint retention study, tissue sections were stained against DAPI, CD68, and CD163 as described and imaged with an Olympus VS200 slide scanning microscope. All tissue sections were inspected for the presence of cells in the lumen, fluorescently prelabeled material in the lumen, and fluorescently prelabeled material in the fibromuscularis. For these three metrics, the percentage of sections with positive identifications was compared among the groups analyzed. Finally, for all sections with fluorescently prelabeled material in the lumen or fibromuscularis, the tissue section was exported to ImageJ, after which a uniform threshold was applied to all sections. The positive area of fluorescent prelabeled material signal was then measured for each section and averaged across all sections per animal.

Statistical Analysis

For all analyses, a one‐way ANOVA followed by post‐hoc pairwise comparisons with Tukey's test were used to compare data between all groups. Significance was set to 0.05 with two‐sided testing. The quantitative analyses were conducted by the investigators, blinded to the group identity. Data, presented as mean±standard error of the mean, were analyzed using GraphPad Prism v10.2, San Diego, CA.

Conflict of Interest

Dr. Christman is a co‐founder, consultant, and board member for, and holds equity interest and receives income from Ventrix Bio, Inc.

Supporting information

Supporting Information

Zelus E. I., Grime J., Saviola A., et al. “Development of a Vaginal Extracellular Matrix Hydrogel for Combating Genitourinary Syndrome of Menopause.” Adv. Mater. 37, no. 39 (2025): 2419977. 10.1002/adma.202419977

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.