Accumulation of senescent cells in the stroma of aged mouse ovary

Abstract

Senescent cells play a detrimental role in age-associated pathogenesis by producing factors involved in senescence-associated secretory phenotype (SASP). The present study was conducted to examine the possibility that senescent cells are present in aged ovaries and, if so, to determine the tissue region where senescent cells accumulate using a mouse model. Female mice at 2–4 and 8–10 months were used as reproductively young and aged models, respectively; the latter included mice with and without reproductive experience. Cells positive for senescence-associated β-galactosidase (SA-β-Gal) staining, one of the markers of cellular senescence, were detected in the stromal region of aged, but not young, ovaries regardless of reproductive experience. Likewise, the localization of cells expressing CDKN2A (cyclin dependent kinase inhibitor 2A), another senescence marker, in the stromal region of aged ovaries was detected with immunohistochemistry. CDKN2A expression detected by western blotting was significantly higher in the ovaries of aged mice with reproductive experience than in those without the experience. Moreover, cells positive for both γH2AX (a senescence marker) and fluorescent SA-β-Gal staining were present in those isolated from aged ovaries. In addition, the transcript levels of several SASP factors were significantly increased in aged ovaries. These results suggest that senescent cells accumulate in the ovarian stroma and may affect ovarian function in aged mice. Additionally, reproductive experience may promote accumulation.

In mammals, the female reproductive system is one of the first organs to exhibit overt symptoms of aging. Reproductive aging in females is associated with ovarian aging, which is characterized by a reduction in the number of ovarian follicles and the quality of oocytes [1,2,3]. Although ovarian aging is one of the major issues in successful infertility treatment in assisted reproductive technology (ART) and efficient animal production, the underlying mechanism is not fully understood.

Cellular senescence is a state of permanent cell-cycle arrest elicited in response to various stresses. This phenomenon was first reported in vitro by Hayflick and Moorhead who demonstrated the limited replicative potential of human diploid fibroblasts in culture, known as the “Hayflick limit” [4]. Senescent cells can be also observed in vivo and are known to be involved in many physiological processes [5, 6]. For instance, programmed senescence has been shown to play a beneficial role during mammalian embryogenesis [7, 8]; likewise, during the wound healing process, senescent cells transiently appear and contribute to the optimal repair of damaged tissues [9, 10]. Cellular senescence is also a protective stress response with a potent anticancer mechanism [11]. In fact, senescent tumor cells are observed in the pre-malignant stages of tumorigenesis [12,13,14,15] and are cleared by immune cells, such as macrophages, to ensure tumor regression [16, 17]. Therefore, when senescent cells transiently appear in tissues, they mainly perform their beneficial functions [5, 6].

In contrast to the beneficial role of the transient appearance of senescent cells, their accumulation in tissues appears to negatively affect the restoration of tissue homeostasis. Senescent cells accumulate in the tissues of aged mammals, including rodents [18,19,20] and humans [21,22,23], and play a detrimental role in age-associated pathogenesis. Senescent cells have been implicated in many age-related disorders, including atherosclerosis [24,25,26], bone disease [27,28,29], Alzheimer’s and Parkinson’s disease [30], and type 2 diabetes mellitus [31]. Therefore, cellular senescence appears to be a critical determinant of tissue/organ aging; however, whether senescent cells are present in aged ovaries has not yet been examined.

The present study was conducted to assess whether senescent cells are present in aged ovaries, and if so, to determine the region in which these cells accumulate. Senescent cells are characterized by elevated activity of β-galactosidase due to increased lysosomal contents [32, 33] and constitutive CDKN2A (cyclin dependent kinase inhibitor 2A, also known as p16^INK4a) expression which contributes to the maintenance of cell-cycle arrest [18, 20, 34]. The phosphorylation of histone H2A variant, H2AX, at the site of DNA damage (γH2AX), is another well-known feature of senescent cells [35]. In addition, senescent cells acquire the ability to secrete growth factors, cytokines, chemokines, and matrix-remodeling proteins, a phenomenon known as the senescence-associated secretory phenotype (SASP) [36]. Therefore, high activity of senescence-associated β-galactosidase (SA-β-Gal), constitutive CDKN2A expression, and γH2AX accumulation, as well as elevated expression of SASP factors, are commonly accepted biomarkers of senescent cells [37] and were monitored as such in the present study.

Materials & Methods

Mice

Mice were purchased from Sankyo Lab Service (Tokyo, Japan), housed under a 12-h light/12-h dark schedule, and provided with food and water ad libitum. All animal experiments were conducted in accordance with institutional guidelines and approved by the Institutional Animal Care and Use Committee at the University of Tokyo.

Fertility assessment

To assess reproductive performance, 8-week-old C57BL/6N female mice were mated with age-matched males, and the number of pups per litter was recorded until the mice reached 10 months of age. The offspring were weaned at 3 weeks old. The number of pups for the periods of 2–4 and 8–10 months for the same mating pairs is presented.

Experimental models

In the present study, C57BL/6N female mice at 2–4 and 8–10 months of age were used as reproductively young and aged models, respectively. For the young group, 2-month-old female mice were purchased from Sankyo Lab Service (Tokyo, Japan) and habituated for at least two weeks before use. Two models were used for the aged group. The “aged virgin” models were mice that had been purchased at 2 months of age and kept in the authors’ laboratory for at least 6 months, while the “aged breeder” models were mice that were purchased at about 6 months of age as retired breeder mice, then kept in the laboratory for an additional 2 months or more before use. Since aged breeder mice are more easily obtainable, these mice were mainly used in the present study.

Histological analysis

For histological assessment, ovaries were fixed in Bouin’s solution, embedded in paraffin, sectioned at a thickness of 6 µm, and stained with hematoxylin and eosin using standard procedures. At least 50 sections of the ovaries around the largest cross section were assessed for each mouse. To assess autofluorescence, frozen ovarian sections (7 µm) were fixed with acetone and mounted using the Vectashield mounting medium (Vector Laboratories, Burlingame, CA, USA). Autofluorescence of ovarian tissue was observed under a fluorescence microscope ECLIPSE Ts2R-FL (Nikon, Tokyo, Japan) using a 470-nm filter. The numbers of mice examined in each experiment are described in the figure legends.

SA-β-Gal staining using whole ovaries or frozen sections

SA-β-Gal staining was conducted as previously reported [38] using whole ovaries or frozen sections (7 µm). After staining, whole ovaries were fixed in Bouin’s solution, embedded in paraffin, sectioned at a thickness of 6 µm, and counterstained with eosin using standard procedures. At least 50 paraffin sections or 10 frozen sections of the ovaries around the largest cross-section, respectively, were assessed for each mouse. The numbers of mice examined in each experiment are described in the figure legends.

Western blotting analysis

Western blotting was conducted as reported previously [39] using whole ovaries. Anti-CDKN2A antibody (ab211542, Abcam, Cambridge, UK) and anti-ACTB antibody (GTX109639, GeneTex, Irvine, CA, USA) were used as primary antibodies, and peroxidase-conjugated goat anti-rabbit IgG (AP132P, Merck Millipore, Burlington, MA, USA) was used as the secondary antibody. The ImmunoStar LD western blotting detection kit (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) and a C-DiGit Blot Scanner and Image Studio for the C-DiGit system (LI-COR, NE, USA) were used to visualize the signals. Protein expression was quantified using ImageJ software (NIH, Bethesda, MD, USA). One ovary from each mouse was used for the western blotting. The numbers of mice examined independently in each experiment are shown in the figure legends.

Immunohistochemical detection of CDKN2A

For the immunohistochemistry, ovaries were fixed in 4% paraformaldehyde for 20 h at 4°C, embedded in paraffin, sectioned at a thickness of 5 µm. The slides were deparaffinized, rehydrated, and heated in 10 mM sodium citrate buffer (pH 6.0). The slides were blocked with 2.5% horse serum (S-2012, Vector Laboratories) for 30 min, and incubated with anti-CDKN2A antibody (ab211542, Abcam) overnight at 4°C. The next day, after washing with phosphate-buffered saline (PBS), the slides were incubated with the secondary antibody from the ImmPRESS-HRP detection kit (MP-7401, Vector Laboratories) for 30 min and washed with PBS at room temperature. Immunoreactive signals were detected using a DAB substrate kit (SK-4105, Vector Laboratories). The slides were counterstained with hematoxylin, dehydrated, and mounted using Permount (Fisher Scientific, Waltham, MA, USA). Negative controls were performed in the absence of the primary antibody. The numbers of mice examined independently in each experiment are shown in the figure legends.

Fluorescent SA-β-gal staining and immunofluorescent detection of γH2AX in cultured ovarian cells

Ovaries of aged breeder mice were cut into small pieces and incubated in 0.25% trypsin-EDTA at 37°C while pipetting several times. After passing through a cell strainer (Falcon, Corning, NY, USA), the cells were centrifuged at 300 × g for 5 min and washed twice with PBS containing 5% fetal bovine serum (FBS). The cells were then cultured overnight in an 8-well chamber slide (Thermo Fisher Scientific) pre-coated with the ECL cell attachment matrix (Merck Millipore). Culture medium used was minimum essential medium alpha (MEMα, Fisher Scientific) with 75 µg/mL penicillin G (Meiji Seika Pharma, Tokyo, Japan), 50 µg/ml streptomycin sulfate (Meiji Seika Pharma), and 5% FBS. Fluorescent SA-β-gal staining and immunofluorescent detection of γH2AX was performed using a Cellular Senescence Detection Kit-SPiDER-βGal (Dojindo Laboratories, Kumamoto, Japan) in accordance with the manufacturer’s protocol. Anti-phospho-histone H2A.X (Ser139) antibody (05-636, Merck Millipore) and rhodamine-conjugated goat anti-mouse IgG (AP181R, Merck Millipore) were used as primary and secondary antibodies, respectively.

Reverse-transcription quantitative-PCR (RT-qPCR)

RT-qPCR was conducted using whole ovaries as previously reported [40]. Briefly, total RNA was isolated using a ReliaPrep RNA Cell Miniprep system (PROMEGA, Tokyo, Japan), and reverse transcription was performed using the ReverTra Ace qPCR RT Master Mix with a gDNA Remover kit (TOYOBO, Osaka, Japan). qPCR was conducted in duplicate using the THUNDER BIRD qPCR Mix (TOYOBO) and an ABI StepOne Real-time PCR System (Applied Biosystems, Foster City, CA, USA). To confirm the specificity of the PCR products, dissociation curve analysis was performed at the end of amplification. In addition, the PCR products were analyzed using agarose gel electrophoresis to confirm the product size for each set of primers. The levels of transcripts of interest were standardized to the levels of a housekeeping gene, ribosomal protein L19 (Rpl19), by the 2^-ΔΔCt method [41]. The PCR primer sets used are listed in Table 1. The numbers of mice examined independently in each experiment are shown in the figure legends.

Table 1. PCR primer sets used in this study.

Symbol	Accession No.	Forward	Reverse	Product size (bp)
Il1a	NM_010554	GAAGCTCGTCAGGCAGAAGT	TCCCGACGAGTAGGCATACA	145
Il1b	NM_008361	GTGCTGTCGGACCCATATGA	TGGGTGTGCCGTCTTTCATT	181
Il6	NM_001314054	CCACTCCCAACAGACCTGTC	GCCACTCCTTCTGTGACTCC	380
Il15	NM_001254747	GGGATCCTGCTGTGTTTGGA	AGCAAGGACCATGAAGAGGC	123
Igfbp2	NM_008342	GCCAAACACCTCAGTCTGGA	GCCATGCTTGTCACAGTTGG	177
Igfbp3	NM_008343	CCTAAGCACCTACCTCCCCT	TGGCATGGAGTGGATGGAAC	143
Igfbp4	NM_010517	CTGCGTACATTGATGCACGG	ACCTGTGATCATGGGCACTG	145
Igfbp6	NM_008344	GGGTCTACAGCCCTAAGTGC	CCTTGGGGTTTGCTCTCCTT	151
Igfbp7	NM_001159518	CCCCCAAGGACATCTGGAAC	GTTCTGTCCGCTGAACTCCA	129
Ctsb	NM_007798	GAGGGTGCCTTCACTGTGTT	CGTGGCCACCCATCATATCA	88
Mmp12	NM_008605	TCCATATGGCCAAGCATCCC	CACAGATGCAGAGAAGCCCA	165
Mmp14	NM_008608	GGCGACAGTACACCCTTTGA	TGCACAGCCACCAAGAAGAT'	158
Timp1	NM_0011593	TTCTTGGTTCCCTGGCGTAC	TCTGGTAGTCCTCAGAGCCC	178
Timp2	NM_011594	GCTGGACGTTGGAGGAAAGA	TCCCAGGGCACAATGAAGTC	99
Serpine1	NM_008871	GGCTATGCTGCAGATGACCA	TGGCATCCGCAGTACTGATC	121
Rpl19	NM_009078	GCCGGCTTCTCAGGAGATAC	TCCATGAGGATGCGCTTGTT	112

Statistical analyses

All experiments were repeated at least three times. The Student’s t-test was used for pairwise comparisons using Microsoft Excel (Microsoft). Statistical significance was set at P < 0.05. Values were shown as mean ± SEM.

Results

Histological changes of ovaries associated with aging

In the present study, female mice at 2–4 and 8–10 months of age were used as “reproductively” young and aged models, respectively. The aged model used in the present study (i.e., 8–10-month-old) was the model at the early stage of reproductive aging, as it showed a significant decrease in the average number of pups per litter and the number of litters per two-month period, resulting in a significant decrease in the total number of pups delivered during the test period compared to young (2–4-month-old) mice (Fig. 1A).

Fig. 1.

Changes in reproductive performance and ovarian histology associated with aging. (A) Reproductive performance of female mice at 2–4 and 8–10 months of age. The average number of pups per litter (Pups/litter, left panel), the number of litters per two-month period (Litter/period, middle panel), and the total number of pups delivered during the test period (2–4 and 8–10 months) (Total number of pups, right panel) are shown (n = 9). * P < 0.05. (B) Representative hematoxylin-eosin-stained ovarian sections from young and aged breeder models (n = 9 and 5, respectively). Lower- and higher-magnification photographs are shown in the upper and lower panels, respectively. The regions shown in the lower panels are indicated with dashed line squares in the upper panels. Arrowhead indicates multinucleated and enlarged macrophages located in the stroma. AF, antral follicle; CL, corpus luteum. (C) The autofluorescence of stromal macrophages was observed (n = 3). Left panel, bright field; right panel, autofluorescence observed using a 470-nm filter.

First, the histological changes of ovaries associated with aging were assessed using young and aged mice with prior reproductive experience (aged breeder mice) (see Materials & Methods for details). Although many pre-antral and antral follicles and corpora lutea were observed in the ovaries of young mice (2–4 months), only a few follicles and corpora lutea, and more stromal regions, were observed in aged ovaries (8–10 months) (Fig. 1B). Many large cells were observed in the stromal region of the aged ovaries (Fig. 1B, arrowheads). These large cells contained multiple nuclei, the cytoplasm was filled with frothy vacuoles (Fig. 1B), and exhibited strong autofluorescence (Fig. 1C). Therefore, these cells were enlarged macrophages, which have been well-characterized in previous studies [42,43,44].

Identification of senescent cells in ovaries of aged mice

To test the possibility that senescent cells are present in aged ovaries, SA-β-Gal staining was conducted using whole ovaries of young and aged breeder mice, and the ovarian sections were examined. In young ovaries, positively stained cells were detected in the atretic follicles, which were characterized by the presence of degenerating oocytes (Fig. 2A, arrows). In contrast, many enlarged macrophages, observed as yellowish colors, were observed in aged ovaries (Fig. 2A, arrowheads), and positive signals were detected as clusters situated near the macrophages in the stromal region. Stromal-positive signals were not detected in young ovaries. Moreover, similar SA-β-Gal staining patterns were observed when SA-β-Gal staining was conducted on cryostat sections (Fig. 2B). The brown-yellow granules observed in aged ovaries were lipofuscin, a non-degradable intra-lysosomal polymeric substance that commonly accumulates in aged tissues (Fig. 2B, arrow) [45].

Fig. 2.

Presence of senescent cells in the stromal region of the ovaries of aged breeder mice. (A) Whole ovaries of young and aged breeder mice (n = 5) were stained with SA-β-Gal staining (blue color), and ovarian sections counterstained with eosin were observed. Lower- and higher-magnification photographs are shown in the upper and lower panels, respectively. Arrow indicates degenerative oocytes within atretic follicles. Multinucleated and enlarged stromal macrophages exhibit a yellowish color (arrowheads). (B) Representative photographs of cryostat sections of the ovaries of young and aged breeder mice (n = 3) after SA-β-Gal staining. Weak SA-β-Gal staining was observed in some follicles, likely atretic follicles, in young mice, whereas cells strongly positive for SA-β-Gal staining were observed in aged ovaries. The arrow indicates lipofuscin, and the arrowhead indicates stromal macrophages. (C) Detection of CDKN2A expression in the ovaries of young and aged breeder mice with immunohistochemistry (n = 3). Negative controls were conducted without the first antibody. Arrowheads indicate stromal macrophages. (D) Detection of CDKN2A expression in the ovaries of young and aged breeder mice with western blotting (n = 3). Band intensities were quantified and are shown in a lower panel. * P < 0.05. (E) Representative photographs of SPiDER-βGal staining and fluorescent immunostaining of γH2AX on cells isolated from aged breeder ovaries (n = 3). Fluorescent SA-β-Gal staining (SPiDER-βGal, green), γH2AX immunofluorescence (γH2AX, red), DNA (DAPI, blue), and merged image are shown. Arrowheads indicate the cell is positive for both fluorescent SA-β-Gal staining and γH2AX.

The localization of CDKN2A, another well-known marker of cellular senescence, was analyzed by immunohistochemistry in ovarian sections of young and aged breeder mice (Fig. 2C). While no signals were observed in young ovaries or in the negative controls, CDKN2A-positive cells were detected around enlarged macrophages in the stromal region of aged ovaries (Fig. 2C). Western blotting analysis showed that young ovaries were negative for CDKN2A expression, whereas strong expression was detected in aged ovaries. (Fig. 2D).

To further examine the presence of senescent cells in aged ovaries, we attempted to colocalize multiple senescent markers in aged ovarian cells. To achieve this, ovaries from aged breeders were dissociated, cultured as a monolayer, and subjected to immunofluorescence staining of γH2AX along with a fluorescent SA-β-Gal (SPiDER-βGal) staining. The results showed that there were cells positive for both senescent markers in aged ovaries (Fig. 2E, arrowhead).

Collectively, these results strongly suggest that senescent cells are present in the stromal regions of aged ovaries.

Expression of SASP factors in aged ovaries

To gain insight into the possible effects of senescent cells on ovarian function, the expression of SASP factors was examined in the ovaries of young and aged breeder mice. The SASP factors examined were interleukins (ILs), insulin-like growth factor-binding proteins (IGFBPs), and proteases/regulators [36, 37]. As shown in Fig. 3, the expression levels of transcripts encoding Ils, i.e., Il1a, Il1b, Il6, and Il15, were significantly higher in the ovaries of aged mice than in those of young mice (Fig. 3A). Transcript levels of Igfbp3, but not Igfbp2, 4, 6 or 7, were significantly higher in aged ovaries than in young ovaries (Fig. 3B). The levels of transcripts encoding proteases/regulators, i.e., Ctsb (cathepsin B), Mmp12 (matrix metallopeptidase 12), Timp2 (tissue inhibitor of metalloproteinase 2), and Serpine1 (serine peptidase inhibitor, clade E, member 1, also known as plasminogen activator inhibitor-1) were significantly elevated in aged ovaries (Fig. 3C).

Fig. 3.

Comparison of SASP factor expressions in the ovaries of young and aged breeder mice (n = 3). Transcript levels of (A) interleukins (ILs), (B) insulin-like growth factor-binding proteins (IGFBPs), and (C) proteases/regulators were examined with RT-qPCR. * P < 0.05.

Identification of senescent cells in ovaries of aged virgin mice

To test whether reproductive experience affects the accumulation of senescent cells in the ovaries, the presence of senescent cells was assessed in the aged virgin mouse model (see Materials & Methods for details). Similar to the ovaries of aged breeder mice, the ovaries of aged virgin mice contained fewer follicles and more stromal regions than those of young mice (compare Fig. 1B and Fig. 4A). Additionally, large yellowish cells, likely enlarged macrophages [42,43,44], were observed in the stromal region (Fig. 4A, arrowheads). Cells positively stained with SA-β-Gal staining were detected in the stromal region of ovaries of aged virgin mice (Fig. 4B). Moreover, strong CDKN2A expression was detected in the ovaries of aged virgin mice (Fig. 4C); however, the expression levels were significantly lower in the ovaries of aged virgin mice than in those of aged breeder mice (Fig. 4D). These results suggest that senescent cells accumulate in the stromal region of aged ovaries irrespective of their reproductive experience. Moreover, the reproductive experience appears to promote the accumulation of senescent cells in the ovaries.

Fig. 4.

Presence of senescent cells in the stromal region of the ovaries of aged virgin mice. (A) Representative hematoxylin-eosin-stained ovarian sections from aged virgin models (n = 4). Lower- and higher-magnification photographs are shown in the upper and lower panels, respectively. AF, antral follicle; CL, corpus luteum. (B) Whole ovaries of aged virgin mice were subjected to SA-β-Gal staining, and ovarian sections counterstained with eosin were observed (n = 3). Lower- and higher-magnification photographs are shown in the upper and lower panels, respectively. Arrowheads indicate stromal macrophages. (C) Detection of CDKN2A expression in the ovaries of young and aged virgin mice with western blotting (n = 3). (D) Comparison of CDKN2A expression in the ovaries of aged breeder and aged virgin mice with western blotting (n = 3). Band intensities were quantified and are shown in the lower panels. * P < 0.05.

Discussion

Although ovarian aging is one of the major issues in human ART and animal production, the underlying mechanisms have not been fully elucidated. The present results showed that cells positive for SA-β-Gal staining and those expressing CDKN2A were both present in the ovarian stroma, and CDKN2A expression was significantly elevated in aged ovaries. Moreover, cells positive for both γH2AX and SA-β-Gal staining were observed in those isolated from aged ovaries. In addition, transcripts encoding well-known SASP factors were significantly upregulated in the ovaries of aged mice. Therefore, ovarian aging appears to be accompanied by cellular senescence, and the accumulated senescent cells may play a detrimental role in age-associated ovarian pathogenesis through the production of SASP factors. These findings provide important insights into the mechanisms underlying ovarian aging.

Although our results suggest that senescent cells accumulate in the stromal region of aged mouse ovaries, the cellular origin of these senescent cells remains unknown. In contrast to the well-studied process of folliculogenesis, the developmental or age-associated dynamics of the ovarian stroma and its components have not been extensively studied. Generally, stromal tissues comprise components such as immune cells, blood vessels, nerves, fibroblasts, and lymphatic vessels. In addition to these components common to most stromal tissues, the ovarian stroma contains specific components such as the ovarian surface epithelium, tunica albuginea, intraovarian rete ovarii, hilar cells, stem cells, and many other uncharacterized stromal cells [46]. The origin of the ovarian senescent cells, that is, whether the known components of the ovarian stroma become cellular senescence with aging or whether senescent cells migrate from outside the ovaries, needs further investigation.

In addition to its cellular origin, the mechanism by which senescence is induced in the ovarian stroma needs to be determined. Cellular senescence is induced by a wide range of cellular stressors, including genomic damage, oncogene activation, and oxidative stress [47]. Once the cells become senescent, they acquire the ability to produce SASP factors that enforce senescence in an autocrine manner [48]. For example, SASP factors, such as IL-1 and IL-6, produced by senescent human fibroblasts, act in a positive feedback loop to reinforce senescence [49, 50]. Interestingly, some SASP factors, including IL-1 and IL-6, are known to induce a senescent phenotype in healthy neighboring cells in a paracrine manner, thus spreading the senescent phenotype [50, 51]. The present results identified that transcripts encoding IL-1A and IL-6 were upregulated in aged ovaries. It is an intriguing possibility that these SASP factors act in a paracrine manner to promote ovarian senescence, making them attractive targets for therapeutic intervention. The fact that senescent cells were not uniformly distributed in the stromal region of aged mouse ovaries, but rather clustered in some parts, supports this possibility. Further studies testing these possibilities and determining the cellular origin and original inducer of cellular senescence in the ovarian stroma are warranted.

What are the consequences of the accumulation of senescent cells in the ovarian stroma with increasing age? Ovarian aging is often associated with an expanded stromal cell compartment and increased fibrosis in humans [52, 53] and mice [43, 54]. Moreover, polycystic ovary syndrome (PCOS), one of the most common causes of infertility among women of reproductive age, is characterized by thickening of the ovarian capsule and stromal fibrosis [55]. Abrogating fibrosis using hormonal treatment or mechanical stimuli has been shown to restore follicular development and fertility in a mutant mouse model of precocious ovarian aging [56, 57]. In addition, antifibrotic drug treatment eliminates fibrotic collagen and restores ovulation in reproductively aged mice [58]. Therefore, stromal fibrosis is considered one of the principal reasons for diminished ovarian function [59]; however, the mechanism underlying fibrosis formation in the ovarian stroma is not well understood. Interestingly, the involvement of cellular senescence in fibrogenesis has been reported in several mammalian tissues. For example, the activation of Wnt/β-catenin signaling triggers the transition of tubular epithelial cells to the senescent phenotype and promotes renal fibrosis in chronic kidney disease [60]. Senescent cell-induced fibrosis is also implicated in idiopathic pulmonary fibrosis, as locally induced senescence in the lung causes lung fibrosis and defective physical function in a mouse model, whereas clearing senescent cells improves pulmonary function in the same model [61]. In contrast, the premature senescence of myofibroblasts plays an antifibrotic role in myocardial fibrosis [62]. Therefore, whether senescent cells exhibit fibrotic or antifibrotic effects appear to depend on the cellular/tissue context; however, SASP factors secreted by these cells seem to play a significant role as determinants of these effects. In fact, some of the SASP factors found to be elevated in aged ovaries in the present study have been reported to be involved in fibrosis. For example, IGFBP3 inhibits Wnt/β-catenin signaling together with DKK1/3 (dickkopf WNT signaling pathway inhibitor 1/3) and suppresses fibrosis in livers [63], whereas its expression has been reported to promote cardiac fibrosis [64]. Similarly, SERPINE1 induces cellular senescence in alveolar progenitor cells and contributes to fibrotic lung diseases [65], and inhibition of CTSB suppresses fibrosis in the liver [66]. Moreover, MMPs exhibit inhibitory or promotive effects on fibrosis depending on the tissue, and their actions are regulated by TIMPs, which are inhibitors of MMPs [67]. As senescent cells produce SASP factors that include both promoters and inhibitors of certain processes, including fibrogenesis, the ultimate effect of senescent cells appears to be determined by a balance of these factors [68]. Whether ovarian senescent cells and SASP factors secreted by these cells are involved in ovarian fibrogenesis, and if so, whether they play fibrotic or anti-fibrotic roles in the ovaries, requires further investigation.

Whether the SA-β-Gal staining positive cells observed in atretic follicles in young ovaries are senescent cells requires further investigation. As mentioned in the introduction, the transient appearance of senescent cells plays a beneficial role in maintaining regular tissue function [5, 6]. Therefore, it is possible that senescent cells transiently appear in atretic follicles and play some roles in follicular atresia or clearing residual tissues.

In summary, our results suggest that senescent cells accumulate in the stromal regions of the ovaries of aged mice. In addition to testing whether similar phenomena are observed in other mammals, identification of the cellular origin and inducers of senescence will be an important next step toward understanding the mechanisms governing ovarian aging and seeking therapeutic targets for human ART and animal production. Moreover, it is important to investigate whether the removal of senescent cells using senolytic drugs can improve the reproductive performance of aged female mice.

Conflict of interests

Natsumi Maruyama, Isuzu Fukunaga, Tomoaki Kogo, Tsutomu Endo, Wataru Fujii, Masami Kanai-Azuma, Kunihiko Naito, and Koji Sugiura declare no conflicts of interest.