Canagliflozin treatment prevents follicular exhaustion and attenuates hallmarks of ovarian aging in genetically heterogenous mice

Abstract

Ovarian aging is characterized by declines in follicular reserve and the emergence of mitochondrial dysfunction, reactive oxygen species production, inflammation, and fibrosis, which eventually results in menopause. Menopause is associated with increased systemic aging and the development of numerous comorbidities; therefore, the attenuation of ovarian aging could also delay systemic aging processes in women. Recent work has established that the anti-diabetic drug Canagliflozin (Cana), a sodium-glucose transporter 2 inhibitor, elicits benefits on aging-related outcomes, likely through the modulation of nutrient-sensing pathways and metabolic homeostasis. Given that nutrient-sensing pathways play a critical role in controlling primordial follicle activation, we sought to determine if chronic Cana administration would delay ovarian aging and curtail the emergence of pathological hallmarks associated with reproductive senescence. We found that mice receiving Cana maintained their ovarian reserve through 12 months of age, which was associated with declines in primordial follicles FoxO3a phosphorylation, a marker of activation, when compared to the age-matched controls. Furthermore, Cana treatment led to decreased collagen, lipofuscin, and T cell accumulation at 12 months of age. Whole ovary transcriptomic and proteomic analyses revealed subtle improvements, predominantly in mitochondrial function and the regulation of cellular proliferation. Pathway analyses of the transcriptomic data revealed a downregulation in cell proliferation and mitochondrial dysfunction signatures, with an upregulation of oxidative phosphorylation. Pathway analyses of the proteomic data revealed declines in signatures associated with PI3K/AKT activity and lymphocyte accumulation. Collectively, we demonstrate that Cana treatment can delay ovarian aging in mice and could potentially have efficacy for delaying ovarian aging in women.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11357-024-01465-w.

Introduction

Ovarian aging has garnered significant interest in recent years due to the recognition that menopause is associated with accelerated systemic aging [1], greater chronic disease burden [2–4], and increased all-cause mortality risk [5]. Menopause is characterized by ovarian follicular exhaustion and myriad of aging hallmarks including mitochondrial dysfunction [6], reactive oxygen species production [7, 8], inflammation [9–11], and fibrosis [12, 13] within the ovarian microenvironment. Several recent reports also indicate that mice and humans accumulate a variety of different immune cells in the aging ovary [9–11, 14, 15], which are believed to contribute to the observed rise in inflammation, fibrosis, and accumulation of multinucleated giant cells (MNGCs). The cell-type-specific mechanisms and time-course of events underlying these age-related ovarian hallmarks are still being elucidated, but a few interventional strategies have proven effective in attenuating the aforementioned phenotypes.

Both dietary [16–19] and pharmacological interventions [18, 20] have been reported to elicit benefits on ovarian aging. Interventions that target nutrient-sensing pathways such as phosphoinositide 3-kinase (PI3K)/protein kinase B (AKT) and mammalian target of rapamycin (mTOR) have proven the most efficacious in preclinical models. This likely occurs because the synergistic actions of PI3K/AKT and mTOR play a pivotal role in regulating primordial follicle activation from a quiescent state, which is the dominant determinant of follicular exhaustion, and thus, reproductive lifespan [21]. PI3K/AKT is activated in response to insulin and insulin-like growth factor 1 (IGF-1) signaling, which promotes primordial follicle activation by inactivating tuberous sclerosis complex 1 (TSC1) in granulosa cells and forkhead box protein O 3a (FoxO3a) in oocytes [21–23]. Mouse models of growth hormone deficiency or impaired signaling represent an example of this phenotype because they possess low circulating levels of insulin and IGF-1, a suppression of systemic mTOR activity, and a preservation of ovarian reserve [24, 25]. Circulating glucose also plays a direct role in modulating primordial follicle activation. Increased glycolysis in granulosa cells due to excess glucose availability decreases AMP-activated protein kinase (AMPK) activity, which promotes mTOR activation. mTOR activation in granulosa cells subsequently induces Kit ligand (KITL) expression and secretion, which bind to KIT receptors on the oocyte, thereby activating PI3K-AKT and inactivating FOXO3a via phosphorylation, resulting in the activation of quiescent primordial follicles [23, 26]. As such, maintaining lower glucose levels is associated with lower levels of primordial follicle activation in mice, ewes, and humans [23, 27, 28]. Given that treatment with sodium-glucose transporter 2 inhibitors (SGLT2i) is known to reduce circulating glucose and insulin [29, 30], we speculate that SGLT2i may have therapeutic potential for attenuating ovarian aging.

Canagliflozin (Cana), an FDA-approved SGLT2i, is primarily prescribed to treat type 2 diabetes [31]. Cana antagonizes SGLT2 in the proximal tubules of the kidneys, thereby inhibiting glucose reabsorption and curtailing hyperglycemia in both sexes [29, 30, 32]. Cana has also proven beneficial when prescribed “off-label” for treating patients with polycystic ovarian syndrome (PCOS) [33, 34], although it remains unclear if these benefits occur in response to direct actions of Cana in the ovary. Interestingly, the National Institute on Aging Interventions Testing Program has reported that Cana significantly extends lifespan and attenuates the incidence of pathology in male, but not female, mice when treatment began at 7 months of age [35, 36]. In contrast to changes in lifespan, Cana improves markers of glucose homeostasis in both sexes [35], suggesting that lifespan and disease pathology may be linked to metabolism in a sex-specific manner. Subsequent studies found that Cana also elicits neuroprotective effects in a sex-specific fashion [37, 38]. For example, age-related hippocampal microgliosis and astrogliosis were reduced in male, but not female, mice receiving Cana [37]. However, Cana treatment did attenuate age-related gliosis and microglia activation in the hypothalamus of both sexes [38], which could conceivably affect the hypothalamic-pituitary–gonadal (HPG) axis in a beneficial way since local proinflammatory stress has been linked to impairments in reproductive function and gonadal hormone production [39].

In the current study, we sought to determine if chronic Cana administration from 6 to 12 months of age would delay ovarian aging and pathological hallmarks associated with reproductive senescence. We chose to evaluate Cana during this period because it represents the timeframe when declines in reproductive function deteriorate the most. We hypothesized that Cana treatment would delay ovarian follicular exhaustion and that this would correspond to declines in age-related ovarian mitochondrial dysfunction, inflammation, fibrosis, and MNGC accumulation.

Materials and methods

Control and experimental diets

TestDiet, a division of Purina Mills (Richmond, IN), prepared all the diets for the studies performed herein. LabDiet 5LG6 (62.2% CHO, 21.7% PRO, 12.1% FAT) was provided as the control (Con) diet and was supplemented with Cana (180 ppm) for the treatment diet.

Animals

UM-HET3 female mice were purchased from The Jackson Laboratory (strain #: 036603; N = 44) at 5 months of age and were acclimated to their new environment for a month. Mice were group-housed at 22 ± 0.5 °C on a 12:12-h light–dark cycle. Unless otherwise noted, mice had ad libitum access to food and water throughout the experimental timeframe. At 6 months of age, mice were randomly assigned to one of three groups: [1] 6-month-old control mice (n = 6), [2] 12-month-old control mice (n = 18), or [3] 12-month-old mice that were treated with Cana from 6 to 12 months of age (n = 20). The 6-month-old control mice were immediately anesthetized in the fasted state (5–6 h) with isoflurane and euthanized by exsanguination due to cardiac puncture prior to tissue collection, and the remaining mice were aged to 12 months under their respective treatment conditions. At 12 months of age, the remaining mice were anesthetized in the fasted state (5–6 h) with isoflurane and euthanized by exsanguination due to cardiac puncture prior to tissue collection. Blood was collected into EDTA-lined tubes, and plasma was collected and frozen. One ovary from each animal was excised, flash frozen, and stored at − 80 °C. The other ovary from each animal was collected into 4% PFA, processed, and serially sectioned. Ovaries from six animals per group were used for histological assays, while the opposing ovary from these mice were used for bulk RNA sequencing. Frozen ovaries from the remaining 12-month-old control (n = 12) and 12-month-old Cana-treated groups (n = 14) were pooled together (2 ovaries/sample) to provide enough tissue for proteomic and NADH oxidase activity assays. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee at Wayne State University.

Plasma analyses

Fasting glucose was measured in whole blood at the time of euthanasia using Accu-Chek Aviva Plus glucometers (Roche, Basel, CH). Fasting insulin and IGF-1 levels were measured in plasma collected during the terminal harvests using kits from Alpco using the ELISA kits from Alpco (cat #: 80-INSMSU-E01) and R&D Systems (Cat #: MG100), respectively, as previously described [40, 41].

Histological analyses

Ovarian reserve was determined by three blinded staff members counting follicles in H&E-stained slides as previously described [14]. Picrosirius Red staining for collagen deposition and Sudan Black staining for lipofuscin accumulation were performed on one randomly assigned section from the middle of each ovary and analyzed as previously described [14]. Immunofluorescence was performed as previously described [14], with minor adaptations. In brief, slides were incubated with primary rabbit anti-phospho-FoxO3a (pFoxO3a; Ser318/321) antibody (cat #: 9465; Cell Signaling; 1:200) and chicken anti-DEAD-Box Helicase 4 (DDX4) antibody (cat #: ab314236; Abcam; 1:200) or rabbit anti-CD3 antibody (cat #: 17,617–1; ProteinTech; 1:200). Primary antibodies were incubated overnight at 4 °C, followed by secondary goat anti-rabbit IgG Alexa Fluor 488 antibody (cat #: 111–545-003; Jackson ImmunoResearch Laboratories; 1:1000) and Rhodamine Red donkey anti-chicken (cat #: 703–295-155; Jackson ImmunoResearch Laboratories; 1:1000) for 1 h at room temperature and DAPI for 5 min. Sudan Black staining was performed during the immunofluorescence protocol in order to quench the autofluorescence from the ovaries. Images were captured on a confocal microscope (Zeiss LSM 880 w/ Airyscan) at 100 × magnification in the green (pFoxO3a), red (DDX4), and blue (DAPI) channels and at 10 × magnification in the green (CD3) and blue (DAPI) channels. The mean fluorescent intensity of pFoxO3a from 10 to 15 primordial follicles were quantified using the Image J, and averaged. CD3 fluorescent intensity was measured in the whole ovarian section.

RNA sequencing

Whole ovary RNA was extracted from one of the ovaries from each mouse from Con-12mo and Cana-12mo groups by the Trizol method [42]. cDNA Libraries were constructed from 40 ng total RNA using the NEB Ultra II Directional Library preparation kit (#NEBE7760L, New England Biolabs) with the NEBNext Poly(A) mRNA Magnetic Isolation Module (#NEBE7490L, New England Biolabs), as previously described [43, 44]. Library sizing was performed with HSD1000 ScreenTape (#5067–5584, Agilent Technologies) and quantified by Qubit dsDNA HS Assay Kit (#Q32851, ThermoFisher Scientific) on a Qubit 4 Fluorometer (#Q33226, ThermoFisher Scientific). The sequence depth obtained was ~ 20,000 reads. The libraries for each sample were pooled at 4 nM concentration and sequenced using an Illumina NovaSeq 6000 system. RNA-Seq analysis was conducted in Partek Flow software (Version 11.0, Illumina). Briefly, the “Illumina_TruSeq_Stranded_RNA-Seq” pipeline was used to align FASTQ files to the mm39 genome using the STAR 2.5.3A aligner. Partek E/M was used to quantify reads, and raw counts were normalized to the median ratio. DESeq2 was used to call differentially expressed genes (DEGs) by treatment (Cana vs. Control) (false discovery rate [FDR] < 0.05). Ingenuity pathway analysis (IPA) was then run on DEGs to identify upstream regulators and canonical pathways differentially regulated by treatment. Heatmaps of differentially expressed genes and pathways were made in Morpheus (https://software.broadinstitute.org/morpheus), and PCA plots were generated in R using ggplot2.

Mitoplast isolation and mitochondrial ETC activity assessment

Total homogenates and mitoplasts were prepared in isolation buffer containing 210 mM mannitol, 70 mM sucrose, 5.0 mM MOPS, and 1.0 mM EDTA, pH 7.4, from the same ovarian samples. As addressed above, each sample consisted of two ovaries pooled from two mice. During the process of mitoplasts isolation, a portion of total homogenates was separated for proteomic analyses. Mitoplasts were then isolated from the remaining volume for assessment of electron transport chain (ETC) activity. To determine the overall ETC activity through complexes I-III-IV, NADH oxidase activity was evaluated as previously described [45]. In brief, isolated ovarian mitoplasts were diluted to 0.05 mg/mL in 25 mM MOPS buffer (pH 7.4). NADH oxidase activity was measured spectrophotometrically (Agilent 8453) as the rate of NADH oxidation (340 nm, ε = 6200 M⁻¹ cm⁻¹) following addition of 150 µM NADH in the presence of 10 mM KCl. Activity was sensitive to inhibition by rotenone (2.5 µM), indicating the necessity of complex I function for NADH utilization. The protein concentration of each sample was determined by the Bradford method (Thermo Scientific) with a BSA as the standard.

Proteomic analyses

Samples were digested using previously described methods [46, 47]. In brief, 20 μg of each sample were immobilized in a short run 12.5% SDS-PAGE gel (Criterion, Bio-rad) followed by fixation and staining with Coomassie blue (Pierce). Each gel lane was cut out as a single sample, cut into smaller pieces, and washed/de-stained. After the proteins were reduced, alkylated, and digested with trypsin, they were subjected to extraction using 70% methanol/5% acetic acid in water, evaporation to dryness, and reconstitution with 1% acetic acid. Samples were then analyzed by both selected reaction monitoring (SRM) and data independent acquisition (DIA) [46–48]. A TSQ Quantiva instrument was used in the SRM mode with Q1 and Q3 resolution = 0.7 FWHM [45]. Each SRM method was constructed with the program Skyline in a series of steps that selects and validates the best responding peptides for each protein target (the best flyers) [49–51]. We typically use 2–3 peptides for each protein. Data were analyzed using the program Skyline [52]. This program finds and integrates the peptide used for each protein. These results are exported into excel worksheets for processing and sorting. A QEx plus instrument was used in the DIA mode with a 20 m/z window working from m/z 350 to 950. The orbitrap was operated at a resolution of 17,500. A full scan spectrum at a resolution of 70,000 was acquired each cycle. These conditions give 7–8 data points across our typical 30 s chromatographic peaks. Data were analyzed using the program DIANN. This program uses neural network methods to analyze the DIA datasets versus the human Uniprot proteome database from Embl. These results are exported into excel worksheets for processing and sorting. For data and pathway analyses, proteins with statistical differences across experimental groups were identified and subjected to ingenuity pathway analysis (IPA). All raw proteomics data are available upon request. The full list of differentially expressed proteins is available in Supplemental File 1.

Statistical analyses

Results are presented as mean ± SEM with p values less than 0.05 considered to be significantly different. Analyses of differences between groups in plasma, histological assays, and NADH oxidase activity were performed by Student’s t-tests or one-way ANOVA with Tukey post hoc analyses where appropriate using the GraphPad Prism 9.0 software. For RNA sequencing and proteomics, differentially expressed gene (DEGs) lists were imported into IPA software and filtered on FDR < 0.1 and logFC > 0.25 for pathway analyses. Pathways with p values less than 0.05 were considered statistically significant, and the activation z-scores are reported by heatmap or bar charts. Corrections for multiple comparisons were made, where appropriate, using the Benjamini, Krieger, and Yekutieli correction for multiple comparisons. Significant differences were defined at p < 0.05 or FDR < 0.05 (for multiple comparisons).

Results

Cana treatment reduced fasting glucose but not insulin or IGF-1

To confirm that Cana elicited the anticipated effects on metabolic parameters, we evaluated fasting glucose, insulin, and IGF-1 in 12-month-old control and Cana-treated mice. As expected, we found that Cana treatment reduced circulating glucose by approximately 25% (Fig. 1A), which is similar to prior reports [35]. Cana treatment did not elicit changes in circulating insulin (Fig. 1B) or IGF-1 (Fig. 1C). The lack of effects of Cana treatment on insulin and IGF-1 was not unexpected because the mice being evaluated were unchallenged and metabolically healthy.

Fig. 1.

Cana treatment reduces fasting glucose in breeding age female mice. A Fasting glucose (n = 6/group), B fasting insulin (n = 5–6/group), and C Fasting IGF-1 (n = 5/group) in 12-month-old control and Cana-treated mice. All data are presented as mean ± SEM and were analyzed by Student’s t-tests. *p < 0.05

Cana treatment attenuated primordial follicle activation, thereby maintaining follicular reserve

To determine if reductions in circulating glucose levels following Cana treatment were associated with declines in primordial follicle activation, we evaluated pFoxO3a in oocytes from primordial follicles via immunofluorescence. We found that lower circulating glucose concentrations were indeed associated with a suppression of FoxO3a phosphorylation by approximately 50% in oocytes from primordial follicles (Fig. 2A–B). The suppression of FoxO3a phosphorylation in oocytes translated to significantly greater numbers of primordial follicles in Cana-treated mice, which nearly doubled the number seen in age-matched control mice. Surprisingly, 12-month-old Cana-treated mice displayed similar follicular reserve to 6-month-old control mice (Fig. 2C), which was mirrored by the primordial:primary follicle ratio (Fig. 2D). The higher the primordial:primary follicle ratio, the greater number of quiescent primordial follicles remain in the ovary. Collectively, these results indicate that Cana treatment preserves ovarian follicular reserve in breeding age female mice.

Fig. 2.

Cana treatment attenuates primordial follicle activation and preserves follicular reserve in breeding age female mice. A Representative immunofluorescence images of pFoxO3a, DDX4, and merged pFoxO3a:DDX4 (magnification = 100 × ; scale bar = 20 μm), B quantified immunofluorescent intensity of merged pFoxO3a:DDX4 (n = 5/group), C estimated number of follicles in each stage (n = 6/group), and D primordial-to-primary follicle ratio (n = 6/group) in 6-month-old control, 12-month-old control, and 12-month-old Cana-treated mice. All data are presented as mean ± SEM and were analyzed by one-way ANOVA with Tukey post hoc analyses *p < 0.05; **p < 0.01

Cana treatment curtailed the emergence of ovarian hallmarks of aging

We next sought to determine if the Cana-mediated preservation of ovarian reserve was associated with changes in age-related ovarian hallmarks. We evaluated ovarian fibrosis, MNGC formation, and T cell accumulation. We found that Cana treatment attenuated age-related ovarian fibrosis by approximately 50% (Fig. 3A–B). Cana treatment also significantly reduced age-related lipofuscin accumulation in the ovary (Fig. 3C–D), which we recently reported to be an excellent marker of ovarian MNGCs [14]. Lastly, we also found that Cana treatment reduced the age-related increase in ovarian T cells (CD3), which are known to accumulate in the aging ovary [11, 14, 15] (Fig. 3E–F).

Fig. 3.

Cana treatment attenuates age-related ovarian fibrosis and MNGC accumulation in breeding age female mice. A Representative Picrosirius Red stained images (magnification = 4 × ; scale bar = 500 μm), B Quantified Picrosirius Red positivity (n = 6/group), C representative lipofuscin stained images (magnification = 4 × ; scale bar = 500 μm), and D quantified lipofuscin positivity; E representative CD3 (green) and DAPI (blue) immunostained images (magnification = 10 × ; scale bar = 200 μm); and F quantified immunofluorescent intensity of CD3 (n = 5–6/group) in 6-month-old control, 12-month-old control, and 12-month-old Cana-treated mice. All data are presented as mean ± SEM and were analyzed by one-way ANOVA with Tukey post hoc analyses. *p < 0.05; **p < 0.01; ***p < 0.001

Cana treatment elicits transcriptomic changes associated with ovarian mitochondrial function

To better understand how Cana treatment may elicit benefits in the aging ovary, we performed bulk RNA sequencing in ovaries from 12-month-old control and Cana-treated mice. We found that Cana treatment only elicited significant changes in the expression of approximately 140 genes, as shown in volcano plot format (Fig. 4A). Only 42 genes were downregulated by Cana, whereas 95 genes were upregulated. Overall transcriptomic differences between control and Cana-treated mice were mild as evidenced by only a subtle separation of the groups in the principal components analysis (PCA) (Fig. 4B). We subsequently evaluated the DEGs using IPA and found that changes in mitochondrial activity were modulated the most by Cana treatment. For instance, the mitochondrial dysfunction pathway was the most downregulated by Cana treatment, whereas the oxidative phosphorylation pathway was the most upregulated (Fig. 4C). The analysis of upstream regulators revealed that pathways associated with mitochondrial function including AMPK, PPARGC1B, and DAP3 were upregulated in Cana-treated ovaries (Fig. 4D). While AMPK helps restore energy balance by promoting energy-generating processes and inhibiting energy-consuming processes [53], PPARGC1B (a downstream target of AMPK) drives the expression of genes involved in mitochondrial biogenesis and function [54]. Additionally, the activation of DAP3 plays a crucial role in the assembly of respiratory chain complexes [55]. These observations suggest that Cana treatment is associated with improvements in ovarian mitochondrial function, which is known to decline significantly during aging [56–58]. Interestingly, the ovarian Sirtuin signaling pathways were also modulated by Cana treatment, with SIRT3 being suppressed and SIRT6 being upregulated. We suspect that declines in SIRT3 signaling may be due to declines in oxidative stress [59], whereas increased SIRT6 signaling may be related to the inhibition of NF-κB activity [60]. Several regulators of cell proliferation were also modulated by Cana treatment including declines in RICTOR [61], FLCN [62], and EIF6 [63] and increased activity of LIF [64], FBXW7 [65], and STK11 [66], which we surmise may be related to greater control of follicular activation observed in the Cana group. Other notable pathways and/or regulators that were modified by Cana treatment included estrogen receptor signaling, the granzyme A pathway, and the neutrophil extracellular trap pathway which are suggestive of increased ovarian function [14], reduced T cell accumulation [67], and declines in chronic inflammation [14, 68], respectively. The full list of pathways and upstream regulators from the transcriptomic analysis is available in Supplemental File 1.

Fig. 4.

Cana treatment attenuates transcriptional signatures of age-related mitochondrial dysfunction in ovaries from breeding age female mice. A Volcano plot showing upregulated and downregulated transcripts, B principal component analysis (PCA) of the transcriptomic data, C IPA canonical pathway analysis, and D IPA upstream regulator analysis showing inhibition or activation of pathways associated with mitochondrial function and cell proliferation from bulk ovarian RNAseq from 12-month-old control (n = 5) and Cana-treated (n = 5) mice. Statistical analysis was performed by t-test. p-values < 0.05 and log2 change > 1 and <  − 1 were considered significant

Cana treatment upregulates pathways associated to decreased PI3K/Akt signaling in the ovary

In an effort to determine if the aforementioned changes in transcriptional outcomes were mirrored by changes in translation, we performed proteomics in ovaries from 12-month-old control and Cana-treated mice. Similar to changes in transcriptional outcomes, we found that Cana treatment only elicited significant changes in the expression of approximately 80 proteins (Fig. 5A). Only 40 proteins were downregulated by Cana, while only 37 proteins were upregulated. In alignment with transcriptional differences, proteomic differences between control and Cana-treated mice were also mild (Fig. 5B). Differentially expressed proteins were then evaluated using IPA. Similar to the transcriptomic results, proteomic data also shows upregulation of the estrogen signaling pathway (Fig. 5C). Interestingly, IL12 signaling pathway was downregulated in the Cana-treated ovaries which is aligned with a prior report that showed IL12 signaling is upregulated in aged ovaries [14]. Additionally, pathways associated with the inhibition of the Pi3K/AKT signaling pathway, including PTEN [69] and cachexia signaling [70], were upregulated by Cana. PTEN as an upstream regulator was also downregulated by Cana (Fig. 5D). Similar to the RNAseq observations, proteomics also revealed a downregulation of upstream regulators of cell proliferation, such as RICTOR [61] and SMAD3 [71] and an upregulation of the cell proliferation inhibitor, KLF15 [72]. Insulin as an upstream regulator was also downregulated along with other regulators associated to glucose metabolism (e.g., HNF4A [73]). Antioxidant response regulators FOXA2 [74] and NFE2L2 [75] were also increased by Cana treatment. Endoplasmic reticulum (ER) stress also appeared to be reduced by Cana treatment as evidenced by downregulation of regulators ATF4 [76] and XBP1 [77]. As shown in our other data, Cana also appeared to beneficially alter proteomic readouts related to inflammation and immune responses. In particular, upstream regulators associated with lymphocyte accumulation (e.g., CD38 [78], NFAT5 [79], STAT4 [80]) were downregulated by Cana. These results corroborate the decrease in CD3 staining observed in the Cana-treated ovaries. The full list of pathways and upstream regulators from the proteomic analysis is available in Supplemental File 1.

Fig. 5.

Cana treatment attenuates proteomic signatures of age-related proinflammatory and ER stress in ovaries from breeding age female mice. A Volcano plot showing upregulated and downregulated proteins, B principal component analysis (PCA) of proteomic data, C IPA canonical pathway analysis, and D IPA upstream regulator analysis showing inhibition or activation of pathways associated with cell proliferation, metabolism, endoplasmic reticulum (ER) stress, and inflammation/immune responses from ovarian proteomic analyses from 12-month-old control (n = 7) and Cana-treated (n = 6) mice. Statistical analysis was performed by t-test. p-values < 0.05 and log2 change > 0.25 and <  − 0.25 were considered significant

Cana exerts no effects on NADH oxidase activity and proteins associated with oxidation and mitochondrial function

To further investigate how Cana treatment may alter mitochondrial function, we performed a NADH oxidase activity assay [45], an overall measure of mitochondrial electron transport chain activity, and a targeted proteomic analysis. Our proteins of interest were from the beta oxidation, glycolysis, mitochondrial function, antioxidants, and stress response pathways. Surprisingly, we observed no differences in NADH oxidase activities between groups (Fig. 6A). Similarly, no differences were observed for most of the proteins selected for the panels except for acadvl (Very long-chain specific acyl-CoA dehydrogenase), ubb/c (Bbiquitin C), hsp90b1 (heat shock protein 90 beta family member 1), and hspa5 (heat shock protein family A member 5), which showed statistical significance (p < 0.05) (Fig. 6B–E). Nevertheless, the fold-changes observed in these proteins were low, which implies that the statistical differences may not represent robust biological effects.

Fig. 6.

Cana treatment only elicits mild changes in metabolic and antioxidant response pathway in ovaries from breeding age female mice. A Ovarian NADH oxidase activity in 12-month-old control (n = 7) and 12-month-old Cana-treated mice (n = 6). Heatmaps showing minor to no changes in protein panels associated with B beta oxidation, C glycolysis, D mitochondrial markers, and E antioxidants and stress responses from ovarian proteomic analyses in 12-month-old control (n = 7) and 12-month-old Cana-treated mice (n = 6). Data is presented as mean ± SEM. * < 0.05 by t-test

Discussion

In the current study, we evaluated the effects of Cana on ovarian aging in mice. Cana inhibits glucose reabsorption in the kidneys by antagonizing SGLT2, thereby reducing circulating glucose [29, 30, 32]. Since glucose and insulin signaling have direct roles in activating quiescent primordial follicles, we hypothesized that Cana may be efficacious for delaying ovarian aging. Cana has also been proven effective for treating metabolic disturbances in patients with PCOS [33, 34]. To test our hypothesis, we provided UM-HET3 female mice Cana from 6 to 12 months of age and evaluated several hallmarks of ovarian aging. We found that Cana treatment curtails the age-related loss of primordial follicles, which was mirrored by an attenuation of fibrosis, proinflammatory signatures, T cell accumulation, and MNGC accumulation. Whole ovary transcriptomic and proteomic analyses revealed that the most significant changes were associated with declines in mitochondrial dysfunction and proinflammatory processes with generalized improvements in metabolic plasticity and antioxidant responsiveness. Our findings, which will be discussed in detail below, demonstrate that Cana may hold therapeutic potential for delaying ovarian aging in women.

Our data suggests that Cana may maintain follicular reserve by reducing follicle activation rate [21, 26], which is supported by our observation that FoxO3a phosphorylation is reduced in primordial follicles following Cana treatment. Since the phosphorylation of FoxO3a is downstream of Pi3K/AKT [21, 22], we surmise that reductions in glucose and/or insulin signaling in the ovary may be causally implicated in Pi3K/AKT activation. Although Pi3K/AKT was not differentially expressed in our study, we did find that upstream regulation by PTEN, an inhibitor of Pi3K/AKT activity [69], was significantly upregulated by Cana treatment. This is aligned with a prior report showing that PTEN ablation in oocytes results in the overactivation of these follicles [81], thereby demonstrating that PTEN is a crucial regulator of primordial follicle activation. We also found that ovarian mTOR signaling was downregulated by Cana treatment. mTOR activity in granulosa cells results in the activation of quiescent primordial follicles through KITL-KIT receptor signaling on oocytes, which signals through the Pi3K-AKT pathway [23, 26]. Prior work has also established that hyperactivation of mTOR in granulosa and oocytes accelerates follicular exhaustion [26, 82–84]. We have observed similar results in mice that possess mTOR hyperactivation in theca cells [unpublished observation].

In addition to maintaining ovarian reserve, Cana also attenuated several key hallmarks commonly seen in aging ovaries, which is suggestive that Cana treatment directly, or indirectly, elicits benefits within the ovarian microenvironment. One of the most common hallmarks of aging ovaries is the induction of fibrogenesis [10, 12–14]. Cana treatment reduced ovarian collagen content by half, which was similar to levels observed in 6-month-old control mice. The mechanism underlying this outcome remains unclear because our transcriptomic and proteomic data do not show significant changes in transforming growth factor-β expression, the primary driver of fibrogenesis [85]. Conversely, we did find that Cana treatment reduced several markers indicative of a proinflammatory environment, which is known to be strongly linked to fibrogenesis [10, 12–14, 86]. In particular, we saw declines in the granzyme A and IL-12 signaling pathways, both of which modulate proinflammatory responses and T cell activity [67, 87]. Interestingly, these results were mirrored by differences in T cell positivity across groups as measured by CD3 immunofluorescence. Cana treatment reduced age-related ovarian T cell accumulation by more than half. We recently reported that aging ovaries accumulate non-conventional double-negative (CD4⁻ and CD8⁻) T cells [14, 88], but the mechanisms underlying this observation, and the role of these T cells in ovarian aging, remain unclear and are currently being explored. We speculate that these non-conventional T cells play a role in ovarian fibrogenesis and MNGC formation due to their proinflammatory effector systems [89–91]. Although we did not evaluate T cell effector systems in this report, our observed declines in ovarian T cells following Cana treatment was associated with a reduction in lipofuscin accumulation, which we recently demonstrated to be an excellent marker of ovarian MNGCs [14].

Cana treatment also appeared to mildly affect mitochondrial function, although these effects were only seen in transcriptomic analyses. Moreover, no differences in NADH oxidase activity were observed between groups. Similarly, markers of ER stress were also reduced by Cana treatment, but these effects were only seen in proteomic analyses, which makes these findings difficult to interpret. We suspect that Cana-mediated changes in mitochondrial and ER functionality are occurring in a cell-type-specific manner, and thus, the bulk tissue approach employed in this study has limited our ability to detect these changes with sufficient granularity. Our suspicion is supported by prior reports indicating that changes to ovarian mitochondrial and ER functional parameters are dominant to follicles, particularly oocytes and granulosa cells [92, 93]. Future studies that employ single-cell approaches and/or the evaluation of isolated follicles are needed to definitely determine how Cana affects ovarian mitochondrial and ER functionality.

There are several notable limitations to the current study that should be acknowledged. First, there is little to no expression of SGLT2 in the ovary [94]; thus, Cana-mediated benefits on ovarian outcomes likely occur in response to changes in systemic metabolic homeostasis, or through the modulation of other pathways as shown in heart [95]. Future Cana treatment studies that employ SGLT2 knockout mice will provide insight into this unresolved question. Secondly, although we show consistent reductions in hallmarks of ovarian aging, we have not yet established that these benefits confer reproductive benefits because we did not perform fertility assessments in the current study. Future Cana treatment studies that evaluate ovulation, oocyte quality, and reproductive fitness in the aforementioned SGLT2 knockout mice will answer this question. Third, we only evaluated the effects of Cana on ovarian health at a single timepoint so it remains unclear which hallmark of ovarian aging is affected first and/or most robustly. Future studies should evaluate the effects of Cana treatment longitudinally, which will help disentangling the time course of cell-type-specific events that promote ovarian aging and how Cana modulates these phenotypes. Despite these limitations, our report serves as the first proof-of-concept study that clearly demonstrates the ability of Cana treatment to ameliorate several age-related phenotypes within the ovary.

In summary, the data presented herein are the first to show that Cana treatment reduces pathological hallmarks associated with reproductive senescence. We found that mice receiving Cana from 6 to 12 months of age had greater follicular reserve than age-match controls, and significantly less fibrosis, proinflammatory signatures, T cell accumulation, and MNGC accumulation. Cana treatment also appeared to elicit mild benefits on mitochondrial and ER-related outcomes, although additional work is needed to confirm these results. Future studies will also be needed to determine the mechanism(s) by which Cana directly, or indirectly, modulates ovarian aging. Our study demonstrates that Cana may hold therapeutic potential for the treatment of ovarian aging in women.

Supplementary information

Below is the link to the electronic supplementary material.

Author contributions

J. V. V. I., H. J., M. S., and M. B. S. conceived the project and designed the experiments. J. V. V. I., H. J., and J. D. H. performed the experiments with contributions from S. B., S. M., and M. T. K. J. V. V. I., C. R. H., K. M. H., M. T. K., and S. R. O. analyzed the data and created figures with contributions from S. B. J. V. V. I., S. B., and M. B. S. wrote the manuscript, and all authors edited and approved the final version.

Funding

This work was supported by grants from the National Institutes of Health (R01 AG069742 to M. B. S.; RF1 AG078170 to M. S.; R01 ES033171 to M. S.; R56 ES034765 to M. S.; R24 GM137786 to M. T. K.; P30 AG050911 to M. T. K.; P20 GM103447 to M. T. K.; P20 GM139763 to K. M. H.) and the Global Consortium for Reproductive Longevity and Equality (GCRLE-4501 to M. B. S.; GCRLE-0523 to S. R. O.).

Data availability

The datasets generated through this work are available upon reasonable request from the corresponding author.

Declarations

Disclaimer

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflict of interest

The authors declare no competing interests.