Age-related decline in spermatogenic activity accompanied with endothelial cell senescence in male mice

Summary Male fertility decreases with aging, with spermatogenic decline being one of its causes. Altered testis environment is suggested as a cause of the phenotype; however, the associated mechanisms remain unclear. Herein, we investigated the age-related changes in testicular somatic cells on spermatogenic activity. The number and proliferation of spermatogonia significantly reduced with aging in mice. Interestingly, senescence-associated β-galactosidase-positive cells appeared in testicular endothelial cell (EC) populations, but not in germ cell populations, with aging. Transcriptome analysis of ECs indicated that senescence occurred in the ECs of aged mice. Furthermore, the support capacity of ECs for spermatogonial proliferation significantly decreased with aging; however, the senolytic-induced removal of senescent cells from aged ECs restored their supporting capacity to a comparable level as that of young ECs. Our results suggest that the accumulation of senescent ECs in the testis is a potential factor contributing to the age-related decline in spermatogenic activity.


INTRODUCTION
Age-related decline in fertility is widely observed in mammals, including humans.Currently, this adverse phenomenon has been demonstrated to be more pronounced in females.For example, mitochondrial dysfunction, 1 abnormal epigenetics, 2 and increased chromosomal missegregation 3 frequently occur in oocytes with aging, thereby reducing the developmental competency of oocytes and ultimately lowering the pregnancy or delivery rates. 4,5In contrast, studies have shown that fertility potential declines with aging in males (e.g., a decrease in semen volume 6 or sperm count 7 ).The most commonly used criterion to define the adverse paternal age of fertility in human males is an age of >40 years at conception. 8,9In the mouse model, spermatogenesis or fertility has also been shown to decrease with aging; in a previous study, male fertility declined after the age of approximately 12 months, with an increase in the proportion of seminiferous tubules without germ cells. 10urthermore, another study revealed that the number of spermatogonial stem cells (SSCs), which are considered the origin of all testicular germ cells, gradually decreases with aging. 11n the testis, spermatogenesis occurs inside the seminiferous tubules.Spermatogonia, including SSCs, are located in the outermost space of the seminiferous tubules and maintain their population through mitosis.Simultaneously, a group of spermatogonia is committed to differentiation by stimulating retinoic acid production by surrounding Sertoli cells. 12,13Germ cells that enter meiosis, also known as spermatocytes, undergo two rounds of meiotic divisions and differentiate into haploid spermatids.Then, spermatids undergo a period of maturation known as spermiogenesis and develop into flagellated, motile spermatozoa.
Although spermatogonia can proliferate in vitro (germline stem cells, GSC) for >5 years, which is longer than the average mouse life expectancy, a long-term culture of GSCs can lead to loss of their function as stem cells because long-term cultured GSCs cannot differentiate into spermatozoa after transplantation into the testis. 14However, when GSCs cultured in vitro for 2 years (the average lifespan of mice) are implanted in the testis of a young mouse, they could differentiate into spermatozoa and produce healthy offspring after fertilization. 15,16In addition, SSCs can proliferate for >3 years in vivo after being serially transplanted 9 times into the testes of young mice, where they can produce mature and competent sperm for producing healthy offspring. 10Thus, we hypothesized that age-related alteration in the testicular microenvironment is a cause of spermatogenic decline in aged males.However, the mechanisms underlying this phenomenon remain poorly understood.
In the present study, we showed that the proliferation of PZLF-positive spermatogonia declined in aged mice.Interestingly, our results clearly showed that senescent features were more pronounced in testicular endothelial cells (ECs), which are known to provide an environment essential for spermatogonial proliferation. 17,18Furthermore, we revealed that ECs from aged mice were less capable of supporting In this research, male mice aged 2 months or >2 years (designated as young or aged mice, respectively) were used.First, we conducted the mating test by housing male mice with young female mice to determine whether male reproductivity changes with aging.All four young male mice used for the mating test could consistently impregnate female mice.After 2 months of mating, the mean number of deliveries per head was 3.0 G 0.7, and the mean number of pups per delivery was 7.7 G 0.3.Conversely, only one of eight aged male mice could impregnate a female mouse, with the delivery occurring only once after 2 months of mating (Table 1).Next, the histological features of testes were compared between young and aged males.Although the body weight of aged males was significantly higher than that of young ones (Figure 1A, p < 0.001), the mean testis weight was comparable between the two groups (Figure 1B).Aged males had significantly lower sperm counts in the cauda epididymis (1.36 G 0.21 3 10 7 cells) than the young ones (2.89 G 0.28 3 10 7 cells) (Figure 1 C, p < 0.05).We also determined the correlation between body weight and sperm count and revealed no negative correlation (Figure S1).Regarding the histology of the testis cross section, some seminiferous tubules in aged males demonstrated a phenotype of spermatogenic decline, which is rarely observed in young males (Figure 1D, asterisk), and the proportion was significantly higher in aged males than in young males (Figure 1E, p < 0.001).Furthermore, germ cell layer thickness, which represents the activity level of spermatogenesis, [19][20][21] in normal-appearing seminiferous tubules was lower in aged mice than in young ones (Figure 1F, p < 0.05).Taken together, fertility in aged male mice declined markedly, with lower spermatogenic activity with aging.

Aged testes exhibited spermatogenesis dysregulation and less active spermatogonial proliferation
Considering that spermatogenic defect was observed in aged testes, we assessed spermatogenic progression in further detail using immunohistochemistry. MVH (a pan-germ cell marker with a more robust expression in the cytoplasm of spermatocytes but relatively weaker expression in spermatogonia or spermatids) and TRA54 (a marker of haploid spermatids) were used for immunohistochemical staining.Multiple layers of MVH-or TRA54-positive cells were noted in almost all seminiferous tubules in young testes.Conversely, insufficient layers of MVH-high spermatocyte or TRA54-positive haploid spermatids (Figure 2A, white and yellow arrows, respectively) were observed in some tubules in aged testes.These proportions were significantly higher in aged testes than in young ones (Figures 2B and 2C; p < 0.05).Furthermore, aged mice had significantly fewer PLZF-positive spermatogonia per seminiferous tubule than young mice (Figures 2D and 2E; p < 0.001).Considering that spermatogonial cell numbers decrease with aging, we hypothesized that the mitotic activity of spermatogonia is lower in aged mice.We confirmed the proliferative potential of spermatogonia by assessing the expression of Ki67, a cell proliferation marker (Figure 2F).Approximately 50% (52.2%G 6.7%) of PLZF-positive spermatogonia in young testes were positive for Ki67, whereas the proportion was reduced to 28.1% G 2.9% in aged testes (Figure 2G, p < 0.05).Therefore, the testes of aged mice showed spermatogenesis dysregulation and less active spermatogonial proliferation.

Testicular ECs in aged mice have a decreased capacity to support spermatogonial proliferation
6][27] SA-b-gal-positive cells were rare in young testes but were common in aged testes.Interestingly, almost all SA-b-gal-positive cells were located adjacent to the outer side of seminiferous tubules, whereas nearly no positive cells were noted inside the seminiferous tubules with germ cells (Figure 3A).In the testes, ECs closely surround the outer region of the seminiferous tubules and play essential roles in controlling the proliferation or self-renewal of spermatogonia. 17,189][30] Thus, owing to aging, testicular ECs in mice might undergo senescence.Testicular ECs were isolated according to the previously reported method. 18,31We confirmed that almost all isolated cells were positive for CD34, which is a marker of testicular ECs (Figure S2A).First, SA-b-gal staining was performed to evaluate cellular senescence.Most ECs that were recovered from young testes were SA-b-gal negative, whereas a significantly higher number of ECs found in aged testes were SA-b-gal-positive (Figures 3B and 3C; p < 0.001).Next, the protein expression level of p21 (a CDK inhibitor protein and one of the common markers of cellular senescence) in ECs of each age was quantified and compared via western blot analysis.The mean protein expression of p21 tended to be higher in the ECs of aged testes than in those of young ones (Figures 3D and 3E; p = 0.054).Furthermore, the in vitro proliferation rate of ECs was significantly lower in aged mice than in   young mice (Figures S3A and S3B; p < 0.05).Considering that testicular ECs are necessary for spermatogonial proliferation, 17,18 we next determined age-related changes in the capacity of ECs to support spermatogonial proliferation.We evaluated the supportive capacity of ECs for spermatogonial proliferation through coculturing of GSCs and ECs, as reported previously. 17,18GSCs, which were developed from 7-day-old postpartum neonates in this study, proliferate to form typical grape-like colonies, and their morphology is distinct from that of other testicular somatic cells (Figure 3F).When cocultured with ECs that were recovered from young testes, GSCs exhibited sustained proliferation.However, when cocultured with ECs recovered from aged mice, the proliferation of GSCs was significantly inhibited (Figures 3F and 3G).Thus, the senescent phenotype was more pronounced in testicular ECs of aged mice than in those of young mice, and the ability of ECs to support spermatogonial proliferation diminished with aging.
Transcriptome of aged ECs showed reduced proliferative features and increased expression of senescence-associated secretory phenotype (SASP)-related genes Age-related transcriptomic alterations in testicular ECs were examined via next-generation sequencing.ECs cultured for 9 days post-harvest (dph) were used for the assays.Initially, RNA sequencing (RNA-seq) data were validated in a randomly selected set of 10 genes using quantitative real-time PCR (qPCR).Of these 10 genes, 8 showed statistically significant trends in both qPCR and RNA-seq analyses.Although the differences in the expression of the two genes, Cdkn2a and Fst, in ECs of each age were not statistically significant in qPCR analysis, the average expression of both genes was upregulated in aged ECs, consistent with the RNA-seq data.The corresponding p values for the two genes in qPCR analysis were 0.0732 and 0.0722 (Figure S4), suggesting that our RNA-seq results were reliable.Next, clustering analysis was conducted to confirm individual differences and reproducibility of transcriptome variations in young and aged mice (N = 3 in each age group).Gene expression patterns in ECs obtained from mice of the same age were grouped into the same clusters (Figure 4A).Next, differentially expressed genes (DEGs) with a log2 fold change of R0.5 or % À0.5 and a p value of <0.05 were extracted (Figure 4B).Compared with young ECs, aged ECs had 540 upregulated and 547 downregulated genes.Interestingly, GDNF and FGF2 (essential factors for the proliferation of SSCs) were not included in the DEG list, although the p value for FGF2 expression was <0.05 (Figure S5).Further, the DEG list was subjected to Gene Ontology (GO) analysis to determine the types of biological processes or signal transduction pathways that are altered in testicular ECs with aging.In the DEG set with age-related downregulation, GO terms related to cell proliferation or cell cycle progression, such as cell cycle, DNA replication, or cell cycle phase transition, were listed (Figure 4C, red arrows).In the age-related upregulation set, GO terms involved in the inhibition of cell proliferation, such as negative regulation of cell population proliferation (Figure 4D, red arrow) or wound healing as regulation of response to wounding (Figure 4D, blue arrow), in which senescent cells are actively involved, 32 were included.In gene set enrichment analysis (GSEA), the number of several gene sets related to cell cycle and proliferation, such as G2M_checkpoint, Mitotic_spindle, and E2F_targets, decreased significantly in ECs obtained from aged mice compared with that in ECs obtained from young mice.Conversely, the P53_pathway gene set, which is related to mitotic arrest, was significantly accumulated in ECs obtained from aged mice, suggesting that the proliferative activity decreased in ECs obtained from aged mice (Figure 4E).Furthermore, to elucidate age-related changes in physiological and molecular pathways within ECs, we analyzed the DEG list using the Kyoto Encyclopedia of Genes and Genomes (KEGG).This analysis further confirmed the significant inhibition of pathways related to the cell cycle or DNA replication in aged ECs (Figure 4F).Next, we assessed the expression of SASP-related genes (one of the hallmarks of cellular senescence) based on the previously reported list. 33Of the 56 listed SASP-related genes, 19 showed a significant age-dependent difference in expression levels (Figure 4G, asterisk).Furthermore, compared with ECs obtained from young mice, 15 significant DEGs were upregulated in ECs obtained from aged mice (Figure 4H).Therefore, the ECs of mouse testes exhibited senescent characteristics with aging.

Treatment of aged ECs with senolytics recovered their potential for supporting spermatogonial proliferation
Removing senescent cells using senolytics reportedly mitigates age-related deleterious phenotypes such as diabetes, 34 cardiac dysfunction, 35,36 vascular hyporeactivity, 37 and renal dysfunction. 38Thus, removing senescent cells from aged ECs might recover ECs' capacity to support spermatogonial proliferation.Among the reported senolytics, the combination of the SRC/tyrosine kinase inhibitor dasatinib (D), which has been approved by the US Food and Drug Administration, and the natural flavonoid quercetin (Q) has been commonly applied to research as well as clinical trials (D + Q).Therefore, we determined whether D + Q treatment can remove senescent cells from ECs obtained from aged mouse testes.These ECs were first treated with D + Q for 6 days under in vitro culture; then, SA-b-gal staining was performed to compare the proportions of SA-b-gal-positive cells.We revealed that the ECs from aged testes that were treated with D + Q showed a significant decrease in the proportion of SA-b-gal-positive cells compared with untreated ECs (Figure 5A).Subsequently, the proportion of SA-b-gal-positive cells in ECs from aged testes was comparable to that in ECs from young testes (Figure 5B), suggesting that D + Q treatment successfully removed senescent ECs.Finally, we used the GSC and EC coculture system to determine whether ECs' capacity to support spermatogonial proliferation recovered after removing senescent cells by D + Q treatment.When GSCs were cocultured with D + Q-treated ECs obtained from aged testes, the proliferative rate of GSCs was similar to that of ECs obtained from young testes, with no significant difference between young ECs and aged ECs treated with D + Q (Figures 5C and 5D).Therefore, the decreased spermatogenic potential observed in the testes of aged mice might partly be due to the impaired spermatogonial proliferation caused by the accumulation of senescent ECs.

DISCUSSION
In this study, we analyzed the effects of aging on male fertility.Results showed that the mitotic activity of spermatogonia was weakened in aged mice, and sperm counts were significantly decreased, resulting in a marked decline in fertility.In addition, the accumulation of senescent ECs in the testis was observed with aging.Interestingly, the in vitro culture model indicated that removing senescent ECs by senolytics treatment could restore their supporting capacity for spermatogonial proliferation.These results indicate a possibility that an age-related increase in the proportion of senescent testicular ECs could be a cause of the spermatogenic decline.
The sperm count has been reported to decrease in aging male mammals, such as rats 39 and humans. 7A previous mouse study also showed that the number of spermatogonia decreases with aging. 11These results are consistent with our findings such as a significant decrease in the number of PLZF-positive spermatogonia in aged testes (Figures 2D and 2E) as well as a decrease in the proportion of Ki67-positive mitotic spermatogonia (Figures 2F and 2G).Although no stage-specific arrest of germ cell differentiation was observed, compared with young mice, the thickness of the germ cell layer reduced (Figure 1F), and the proportion of the seminiferous tubules with insufficient spermatocyte or spermatid layers increased in aged mice (Figures 2B and 2C).Thinning of the germ cell layer accompanied with reduced sperm counts resulting from decreased spermatogonial proliferation has been reported in various studies. 40,41Therefore, a reduction in the number of spermatogonia caused by their low proliferative activity could be a cause of the lower sperm counts observed in aged individuals.
Serial transplantation of SSCs into the testes of young mice allows the SSCs to continue spermatogenesis beyond the lifespan of mice. 10 When SSCs are transplanted into aged mice, their ability to form spermatozoa is suppressed. 10,11In addition, single-cell RNA-seq results using human testis samples revealed that the changes with aging are more pronounced in somatic cells but less evident in germ cells. 42Thus, the properties of somatic cells in the testes tend to alter with aging, leading to alterations in the testicular environment.Likewise, our study showed that SA-b-gal-positive germ cells were rarely detected within the seminiferous tubules even in aged mice.However, the proportion of somatic cells located outside the seminiferous tubules markedly increased with aging (Figure 3A).Histologically, ECs are localized in this region. 17,18,31We observed that the proportion of SA-b-gal-positive ECs predominantly increased; additionally, the protein expression of p21, which inhibits cell proliferation and is highly expressed in senescent cells, 43,44 increased in ECs obtained from aged mice.Furthermore, the expression of a set of genes related to proliferation or cell division decreased, whereas that of SASP-related genes, one of the characteristics of senescent cells, 33 increased in ECs derived from aged mice.Therefore, testicular ECs undergo senescence with aging; however, future in vivo studies are warranted to determine the senescence of other testicular somatic cells, such as Leydig and myoid cells.
Testicular ECs, which are located outside the seminiferous tubules, regulate the proliferation and self-renewal of spermatogonia by secreting GDNF 17 and FGF. 18ECs recovered from aged mice showed lower capacity to support GSC proliferation than those recovered from young mice.However, removal of senescent cells from ECs that were recovered from aged mice by D + Q treatment restored ECs' ability for supporting GSC proliferation.Therefore, the accumulation of senescent cells in ECs suppresses spermatogonial proliferation.Moreover, this study showed no significant difference in GDNF mRNA or a slight (1.46-fold) increase in FGF2 mRNA expression in aged ECs compared with that in young ones (Figure S5).The expression of factors other than GDNF or FGF, which accelerate the proliferation of germ cells, may decrease with aging.In fact, the in vitro culture of GSCs requires MEFs as feeder cells in addition to GDNF and FGF.In B6 and 129 background mice, GSCs develop poorly even after supplementation with GDNF and FGF. 15 Thus, other factors in ECs, in addition to GDNF and FGF, could play an important role in spermatogonial proliferation.Another possibility is that aged ECs produce factors that inhibit the proliferation of germ cells.ECs recovered from aged mice showed an elevated expression of proinflammatory cytokines related to SASP (Figures 4F and  4G).In mouse experimental autoimmune orchitis model, which exhibits a chronic physiological inflammation, the proliferation and differentiation of spermatogonia are reduced, resulting in spermatogenic defects. 45In addition, mitosis in GC-1 cells, a cell line originating from spermatogonia, can be arrested by adding the inflammatory cytokine TNF-a to the culture medium. 46Insights into the changes associated with the senescence of testicular ECs that affect the proliferative potential of spermatogonia should be analyzed in future research.

Limitations of the study
We also note the certain limitations of this study.We determined the age-related alteration of the transcriptome in the testicular ECs.However, cellular senescence is thought to be a complex phenomenon involving numerous changes in gene expression and molecular signalings due to genetic or epigenetic changes.For future projects, it would be interesting to conduct several epigenetic analyses, such as ChIP-seq for histone modifications or ATAC-seq for chromatin structure.In addition, it would be required to determine in detail what types of changes occur in ECs with age and how these changes work to suppress the proliferative ability of spermatogonia by using these genetic and epigenetic analyses.In addition, we have focused only on the testicular environment in this study.Since aging is an alteration of the multisystemic environments including hormone secretions or immune systems, future studies should analyze physiological changes of the whole body, such as endocrine and/or immunology.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:  (G) Heatmap showing age-related alterations in the expression levels of senescence-associated secretion phenotype (SASP)-related genes. 33Gene expression patterns in mice of the same age were grouped into the same cluster (n = 3 in each age, biological replicates).*p < 0.05, indicating that the gene showed a significant difference.
(H) The mean expression levels of SASP-related genes showing a significant difference in ECs between young and aged mice.Of the 19 genes with significant differences in expression, 15 were elevated in aged ECs.
A B C D  antibodies.After staining with ProLong Diamond Antifade Mountant with NucBlue Stain (Thermo Fisher Scientific), the sections were covered with a coverslip.We observed the stained sections under a microscope to detect fluorescein (Epi-fluorescent microscope, BZ-X700; Keyence, Osaka, Japan).The seminiferous tubules showing insufficient spermatocyte layers or incomplete meiosis (Figure 2A, indicated using white or yellow arrows, respectively) were counted, and their ratios with respect to the total number of seminiferous tubules were calculated.

Germline stem cell development and culture
We used wild-type mice (B6J background) for GSC development according to a previously reported method. 15Briefly, the testes of 7-day-old postpartum neonates were collected and enzymatically digested into single cells.Then, single cells were seeded into a GSC culture medium containing mitotically inactivated mouse embryonic fibroblasts (MEF; treated with 10 mg/mL mitomycin C [MMC] for 2 h before use); the medium was slightly modified. 40These cells were then cultured at 37 C under a humid atmosphere with 5% CO 2, wherein all testicular somatic cells inhibited proliferation within a short period, and stably self-renewing GSCs were developed after several passages.Subsequently, these GSCs were passaged every 7-10 days at a density of 1 3 10 5 cells/mL on MEF.

Testicular EC culture and immunofluorescence
Testicular ECs were recovered from young or aged mice using previously reported methods. 18,31Briefly, each testis was sliced using a pair of scissors and further digested at 37 C for 30 min with agitation using a buffer solution containing collagenase type IV (1 mg/mL; Worthington, Lakewood, NJ, USA), hyaluronidase (1 mg/mL; Sigma-Aldrich), and DNase I (10 mg/mL; Roche, Tokyo, Japan).Next, the digested tissues were washed once with Hanks' balanced salt solution.Then, the cells were seeded onto gelatin-coated dishes containing EC culture medium (50:50 mixture of alpha MEM [Nacalai, Kyoto, Japan] and StemPro-34 [Thermo Fisher Scientific]) supplemented with 20% (v/v) FBS and were cultured until specific time points for each assay.To confirm the purity of the obtained ECs, immunofluorescence was performed.In brief, cultured cells were fixed with ice-cold methanol (100%) for 5 min and then blocked with PBS containing Tween 20 (0.1%, v/v) and donkey serum (10%, v/v) for 30 min.Then, cells were exposed overnight at 4 C to an antibody against CD34 (Thermo Fisher, #14-0341-82), which is a marker of testicular ECs. 18,31Immunoreactivity was visualized using Alexa Fluor 555-conjugated anti-rat IgG secondary antibody.After staining with ProLong Diamond Antifade Mountant with NucBlue Stain (Thermo Fisher Scientific), the sections were covered with a coverslip.We observed the stained cells under an epi-fluorescent microscope (BZ-X700; Keyence).

Coculture of ECs and GSCs
Coculture of ECs and GSCs was performed as described previously. 17,18Briefly, ECs (1 3 10 5 cells seeded into a gelatin-coated 35-mm dish) at 9 dph were treated with MMC (10 mg/mL) for 3 h before being used for coculture with GSCs (2 3 10 5 GSC per dish).Coculture of ECs and GSCs was performed in GSC culture medium, and the number of GSCs at each passage was counted using the Luna-FL Dual Fluorescence Cell Counter (Logos Biosystems, Gyeonggi-do, South Korea).
Senescence-associated b-galactosidase staining (SA-b-gal) Senescence b-Galactosidase Staining Kit (Cell Signaling Technology) was used for SA-b-gal staining according to the manufacturer's instructions.For staining testis sections, testes without fixation were embedded into a 22-oxacalcitriol compound (Sakura Finetek, Tokyo, Japan), frozen in liquid nitrogen, and sliced to obtain 5-mm thick sections.The sectioned tissues were then fixed and used for SA-b-gal staining.After SA-b-gal staining, the nucleus of the section was counterstained with 4 0 ,6-diamidino-2-phenylindole (DAPI).These stained sections were covered with a coverslip and observed under a microscope (BZ-X700, Keyence).For in vitro EC staining, ECs were seeded into a gelatincoated culture dish containing EC medium until 9 dph, which were then fixed and used for SA-b-gal staining.Next, we stained the nucleus with DAPI and observed the ECs under a microscope (BZ-X700, Keyence).The number of SA-b-gal-positive cells was counted manually using Fiji cell counter plug-in (ImageJ, NIH).The number of DAPI-positive cells in the same field was counted using BZ-X Analyzer (Keyence).

Western blot analysis
We collected ECs that were cultured for 9 dph.Then, we extracted total protein via sonication using RIPA buffer (Nacalai, Kyoto, Japan) containing phenylmethylsulfonyl fluoride (Cell Signaling Technology, Danvers, MA, USA), protease inhibitor cocktail (Nacalai), and phosphatase inhibitor (Nacalai).The extracted protein lysate was analyzed via the standard western blotting protocol.An animal-specific secondary antibody conjugated to horseradish peroxidase was used to visualize each protein via Luminata Forte (Merck Millipore).The protein signal was detected using the ChemiDoc Imaging System (Bio-Rad, Hercules, CA, USA). Figure S6 presents all raw membrane images.

RNA-seq
We collected ECs that were cultured for 9 dph via trypsinization and used them for RNA-seq analysis.Total RNA concentration was calculated using Quant-IT RiboGreen (Invitrogen, #R11490).To assess the integrity of total RNA, we ran the samples on the TapeStation RNA screentape (Agilent, #5067-5576).Only high-quality RNA samples with an integrity number of >7.0 were used for RNA library construction.A library was independently constructed using 1 mg of total RNA for each sample via Illumina TruSeq Stranded mRNA Sample Prep Kit (Illumina, Inc., San Diego, CA, USA, #RS-122-2101).We first purified poly-A-containing mRNA molecules using poly-T-attached magnetic beads, fragmented the mRNA samples into small pieces using divalent cations at elevated temperatures, copied the cleaved RNA fragments into first-strand

Figure 1 .
Figure 1.Aged male mice show reduced fertility and lower sperm production (A) Body weight of young and aged mice.The dots in the bar graph represent biological replicates of individual males (n = 12, young group; n = 13, aged group).***p < 0.001, indicating a significant difference.(B) Testis weight of young and aged mice.The dots in the bar graph represent biological replicates of individual males (n = 12, young group; n = 13, aged group).(C) Sperm counts in the cauda epididymis (310 7 ).The dots in the bar graph represent biological replicates of individual males (n = 8, young group; n = 13, aged group).***p < 0.001, indicating a significant difference.(D) Representative periodic acid-Schiff (PAS) and hematoxylin staining of the testis section from young or aged mice (three individuals of each age).Highermagnification photos from an individual at each age are shown on the lower right panels.Seminiferous tubules with an asterisk (*) represent spermatogenic degeneration.The scale bar denotes 100 mm.(E) Proportion of seminiferous tubules showing spermatogenic degeneration.The dots in the bar graph represent biological replicates of individual males (n = 10, young group; n = 10, aged group).***p < 0.001, indicating a significant difference.(F) Thickness of the germ cell layer in the seminiferous tubules indicates normal spermatogenesis.The dots in the bar graph represent biological replicates of individual males (n = 8, young group; n = 8, aged group).*p < 0.05, indicating a significant difference.

Figure 2 .
Figure 2. Aged testes exhibit spermatogenesis dysregulation and less active spermatogonial proliferation (A) Immunohistochemistry of testes from young and aged mice.Sections were stained with anti-MVH (a marker of germ cells) and anti-Haploid sperm cell-specific antigen (clone: TRA54) (a marker of haploid spermatid) antibodies.Nuclei were stained with NucBlue.White and yellow arrows depict the seminiferous tubules showing insufficient spermatocyte layers and incomplete meiosis, respectively.The scale bar denotes 100 mm.(B and C) Proportion of seminiferous tubules showing insufficient spermatocyte layers (B) or incomplete meiosis (C).Fourteen to twenty-five sections per individual male were used for counting, and the average was designated as one biological replicate.The dots in the bar graph represent biological replicates of individual males (n = 4 per group).*p < 0.05, indicating a significant difference.(D) Immunohistochemical analysis of PLZF in young and aged testes.Nuclei were stained with NucBlue.The scale bar denotes 100 mm.

Figure 2 .Figure 3 .
Figure 2. Continued (E) Violin plots showing the number of PLZF-positive spermatogonia per seminiferous tubules in young or aged mice.Ninety-nine to a hundred and nine seminiferous tubules were observed for cell counting per individual male.Each violin plot shows data from individual mouse.Three males as biological replicates of each age were counted (n = 3, young group; n = 3, aged group).***p < 0.001, indicating a significant difference.(F) Immunohistochemical analysis of PLZF and Ki67 in young and aged testes.Nuclei were stained with NucBlue.Yellow arrowheads depict Ki67/PLZF doublepositive proliferating spermatogonia.The scale bar denotes 100 mm.(G) The ratio of the number of Ki67/PLZF double-positive cells to the total number of PLZF-positive cells.The dots in the bar graph represent biological replicates of individual males (n = 3, young group; n = 3, aged group).*p < 0.05, indicating a significant difference.

Figure 3 .Figure 4 .
Figure 3. Continued (F) Representative germline stem cell (GSC) colonies cocultured with ECs from young mice (left panel) or aged mice (right panel).The scale bar denotes 100 mm.(G) The mean proliferative rate of GSCs cocultured with ECs derived from young or aged mice.GSC numbers were counted every 7 days for 3 weeks.The dots in the bar graph represent biological replicates of individual males (n = 3, young group; n = 3, aged group).*p < 0.05, **p < 0.01, or ***p < 0.001, indicating a significant difference.

Figure 4 .
Figure 4. Continued (B) Volcano plot showing the distribution of each gene expression level (x axis) and their p value (y axis).In ECs, 547 genes were downregulated (dots in blue), whereas 540 genes were upregulated (dots in red) with aging based on the cutoff criterion of log2 fold change of R0.5 or % À0.5 and a p value of <0.05.(C and D) Gene Ontology (GO) analysis of the age-related downregulation (C) or upregulation (D) of genes in testicular ECs.Red arrows indicate GO terms related to cell cycle, cell division, or cell proliferation.(E) Gene set enrichment analysis (GSEA) of ECs derived from young and aged mouse testes.(F) KEGG enrichment analysis of upregulated and downregulated DEGs in ECs derived from aged and young testes.(G) Heatmap showing age-related alterations in the expression levels of senescence-associated secretion phenotype (SASP)-related genes.33Gene expression patterns in mice of the same age were grouped into the same cluster (n = 3 in each age, biological replicates).*p < 0.05, indicating that the gene showed a significant difference.(H) The mean expression levels of SASP-related genes showing a significant difference in ECs between young and aged mice.Of the 19 genes with significant differences in expression, 15 were elevated in aged ECs.

Figure 5 .
Figure 5. Treatment of aged ECs with senolytics recovers their potential for supporting spermatogonial proliferation (A) SA-b-gal staining of testicular ECs in vitro with or without dasatinib (50 nM) and quercetin (25 mM) (D + Q).The left and middle panels show ECs from young and aged mice, respectively, and the right panel shows D + Q-treated ECs from aged mice.The scale bar denotes 100 mm.(B) Proportion of SA-b-gal-positive ECs.The dots in the bar graph represent biological replicates of individual males (n = 4, young group; n = 4, aged group; n = 4, aged group treated with D + Q). *p < 0.001, indicating a significant difference.(C) Representative GSC colonies cocultured with ECs derived from young mice (left panel) or ECs derived from aged mice that were treated with (right panel) or without D + Q (middle panel).The scale bar denotes 100 mm.(D) The mean proliferative rate of GSCs cocultured with ECs derived from young or aged mice treated with or without D + Q. GSCs were counted every 7 days for 3 weeks.The dots in the bar graph represent biological replicates of individual males (n = 3, young group; n = 3, aged group; n = 3, aged group treated with D + Q). *p < 0.05, indicating a significant difference.

Table 1 .
Fertility of the young and aged male mice Age No. of males No. of fertile males (ratio) Total No. of delivery No. of delivery per head No. of pups per delivery

TABLE
d RESOURCE AVAILABILITY B Lead contact B Material availability B Data and code availability d EXPERIMENTAL MODEL AND MOUSE DETAILS B Animals and ethics