The rapamycin analog Everolimus reversibly impairs male germ cell differentiation and fertility in the mouse

Abstract

Sirolimus, also known as rapamycin, and its closely related rapamycin analog (rapalog) Everolimus inhibit “mammalian target of rapamycin complex 1” (mTORC1), whose activity is required for spermatogenesis. Everolimus is Food and Drug Administration approved for treating human patients to slow growth of aggressive cancers and preventing organ transplant rejection. Here, we test the hypothesis that rapalog inhibition of mTORC1 activity has a negative, but reversible, impact upon spermatogenesis. Juvenile (P20) or adult (P>60) mice received daily injections of sirolimus or Everolimus for 30 days, and tissues were examined at completion of treatment or following a recovery period. Rapalog treatments reduced body and testis weights, testis weight/body weight ratios, cauda epididymal sperm counts, and seminal vesicle weights in animals of both ages. Following rapalog treatment, numbers of differentiating spermatogonia were reduced, with concomitant increases in the ratio of undifferentiated spermatogonia to total number of remaining germ cells. To determine if even low doses of Everolimus can inhibit spermatogenesis, an additional group of adult mice received a dose of Everolimus ∼6-fold lower than a human clinical dose used to treat cancer. In these animals, only testis weights, testis weight/body weight ratios, and tubule diameters were reduced. Return to control values following a recovery period was variable for each of the measured parameters and was duration and dose dependent. Together, these data indicate rapalogs exerted a dose-dependent restriction on overall growth of juvenile and adult mice and negative impact upon spermatogenesis that were largely reversed; following treatment cessation, males from all treatment groups were able to sire offspring.

Treatments of juvenile and adult mice with rapamycin and its analog (rapalog) Everolimus cause reproductive deficits, but males are able to sire offspring following cessation of treatment.

Introduction

Over the past 80 years in developed countries, semen quality and male fertility have been in decline [1]. The causes for this concerning decline are multifactorial, but coincide with the introduction of a myriad of novel dietary changes, toxicological exposures, and pharmaceutical usage. One recent example of the latter is the clinical application of sirolimus (rapamycin) and its water-soluble rapamycin analog (rapalog) Everolimus, both of which inhibit the “mammalian target of rapamycin complex 1” (mTORC1), a protein kinase complex that positively regulates growth, metabolism, autophagy, cell cycle progression and proliferation [2]. In human medicine, rapalogs have been approved by the Food and Drug Administration (FDA) for treatment of aggressive cancers such as renal cell carcinoma [3] as well as to reduce rejection of transplanted organs [4, 5]. A significant potential unintended side effect of long-term sirolimus treatment suggested in one case report was reversible male infertility [6]. However, to the best of our knowledge, the effects of Everolimus on mammalian reproduction have not been reported.

Millions of haploid sperm are produced daily throughout the lengthy male reproductive lifespan. The foundation of the high-throughput system of spermatogenesis is provided by spermatogonial stem cells (SSCs), which undergo asymmetric divisions to both replenish the SSC population and generate undifferentiated progenitor spermatogonia that proliferate and differentiate in response to retinoic acid (RA) [7–9]. During the week-long process of differentiation, spermatogonia undergo six mitotic divisions (in mice) prior to entering meiosis as preleptotene spermatocytes [10–12]. Our laboratory recently demonstrated that, in the developing testis, RA stimulates spermatogonial proliferation, induces maturation of cellular organelles, and directs efficient translation of messenger RNAs (mRNAs) encoding key regulators of differentiation such as KIT, SOHLH1, and SOHLH2 by activating the PI3K/AKT kinase signaling pathway [7, 13]. We tested the requirement for mTORC1 in spermatogonial differentiation by treating neonatal mice with sirolimus; inhibition of mTORC1 activity blocked spermatogonial differentiation and RA-stimulated protein synthesis [14–16]. Others have reported a requirement for mTORC1 signaling in normal proliferation and differentiation of undifferentiated progenitor spermatogonia in rodents [17–22]. Several case reports suggest similar effects in human male transplant patients [23–26].

Currently, the reproductive phenotypes resulting from rapalog treatments remain unclear. Here, we utilized juvenile and adult mice to study the reproductive consequences of Everolimus treatment to define the mode of action, which can be related to human treatment regimens. Importantly, we also assessed spermatogenesis and the ability to sire offspring following cessation of treatment. Overall, our data indicate Everolimus is a potent in vivo inhibitor of mTORC1 and spermatogenesis, with dose-dependent and largely reversible effects on spermatogenesis. Males from each of the rapalog treatment groups were able to successfully sire offspring after a recovery period. Of interest, this recovery of spermatogenesis and the ability to sire offspring following Everolimus treatment reveals targeting germ cell-specific aspects of the mTORC1 signaling pathway is a potential human male contraceptive strategy.

Materials and methods

Animal care

All procedures using animals adhered to guidelines outlined in the National Research Council Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee at East Carolina University (approval A3469-01). Outbred CD-1 mice (Charles River Laboratories) were used for all studies, and the day of birth was designated as postnatal day (P)0. Mice were humanely euthanized by asphyxiation in CO₂ followed by cervical dislocation.

Rapalog injections

Juvenile male mice (P20) received daily intraperitoneal (IP) injections of sirolimus (10 μg/g, LC Laboratories, R-5000) or Everolimus (5 μg/g, LC Laboratories, E-4040) using a 30-gauge needle for 30 days. Control mice received dimethyl sulfoxide (DMSO) as the vehicle-alone control. A subset of the mice was sacrificed in the morning following the last day of injection (age = P50), whereas remaining mice were allowed to recover for 70 days before tissue recovery (age = P120). Testes, seminal vesicles, cauda epididymides, and body weights were recorded, and cauda epididymal sperm numbers were determined using a hemocytometer [27]. Testes were immersion fixed in 4% paraformaldehyde (PFA, for cryosectioning) or Bouin solution (for histological analysis).

Adult male mice (P>60) received daily IP injections of sirolimus, Everolimus, or vehicle-alone (DMSO) for 30 days starting at P60 as described above. Additionally, a separate group of male mice received a low dose of Everolimus (0.17 μg/g). Since human clinical dose of Everolimus is 0.083 μg/g (5 mg dose, average human male = 60 kg, therefore 5/60 = 0.083 mg/kg, https://www.hcp.novartis.com/products/afinitor/sega/dosing-administration/), we used allometric scaling to account for the metabolic rate differences between humans and mice, and determined that an animal equivalent dose (AED) = 1.01 μg/g (0.083 × 12.1 K_m ratio) [28]. A dose of 0.17 μg/g of Everolimus, ∼6-fold lower than the AED, was then chosen to investigate the effects of the drug on spermatogenesis at a concentration significantly lower than the AED of human clinical dose. Juvenile and adult tissues were recovered as described above immediately after injection (at P50 and P90, respectively) and following 70 and 30 days of recovery (both juvenile and adult at P120). Additional adult tissues were collected from males used in the breeding trial at P160 (following 70 days of recovery).

Histology and indirect immunofluorescence

For histological analyses, whole testes were immersion fixed in Bouin solution for 24 h at 4°C, washed overnight in 1× phosphate-buffered saline (PBS), dehydrated through an ethanol series, processed using standard methods, and then embedded in paraffin. Sections (5 μm) were cut and stained with hematoxylin and eosin using standard methods, and images were taken on a Zeiss Axio Observer A1 microscope (Carl Zeiss Microscopy, LLC) outfitted with a Dage XL16 digital camera (Dage-MTI).

For indirect immunofluorescence, testes were immersion fixed for 24 h at 4°C in fresh 4% PFA, washed overnight in 1× PBS, and then incubated in 30% sucrose at 4°C for 24 h. Testes were frozen in Optimal Cutting Temperature Compound (OCT), and 5-μm cryosections were cut for immunostaining. Primary antibodies used were the following: anti-TRA98 (1:1000, #ab82527, Abcam, rat monoclonal), anti-GFRA1 (1:800, #AF560, R&D Systems, goat polyclonal), anti-KIT (1:1000, #AF1356, R&D Systems, goat polyclonal) anti-ZBTB16/PLZF (1:400, #AF2944, R&D Systems, goat polyclonal), anti-cleaved PARP1 (1:100, #94885, Cell Signaling Technology, rabbit monoclonal), anti-phospho (P)-RPS6 (1:800, #5364, Cell Signaling Technology, rabbit monoclonal), and anti-MTOR (1:400, #2983, Cell Signaling Technology, rabbit monoclonal). Primary antibody was omitted in negative controls. Following stringency washes, sections were incubated in secondary antibodies (1:500, Alexa Fluor donkey anti-rabbit-488, Alexa Fluor donkey anti-goat-488, Alexa Fluor donkey anti-goat-555, Alexa Fluor donkey anti-rat-555, ThermoFisher Scientific). Coverslips were mounted with Vectastain containing DAPI (Vector Laboratories), and images were obtained using a Fluoview FV1000 confocal laser-scanning microscope (Olympus America). Testes from at least three different mice were analyzed for each experiment, and immunostainings were repeated at least twice.

Analysis of seminiferous tubule diameter

Bouin-fixed and paraffin-embedded testis halves were cut into (5-μm) sections, which were stained with hematoxylin and eosin using standard methods. Digital images were taken of each section using a Zeiss Axio Observer A1 microscope (Carl Zeiss Microscopy, LLC) and a Dage XL16 digital camera (Dage-MTI) using a grid pattern to avoid overlap. Each tubule diameter was measured at is maximal (D1) and minimal (D2) distance and the average (D) used for statistical analyses. In order to ensure consistency in measurement of average tubule diameter, only round or nearly round tubules with D2/D1 ≥ 0.85 were included in the analyses [29].

Fertility trial

Following a 10-day recovery, two adult males (P100) from each treatment group (vehicle-alone, sirolimus, Everolimus, and Everolimus-low dose) were paired with untreated wild type (WT) CD-1 female mice for 60 days. Litter dates and sizes as times to first and subsequent litters were recorded from females for 90 days.

Statistics

One-way analyses of variance for tissue weights, sperm counts, and tubule diameters were performed using NCSS 2000 (NCSS Statistical Software). Statistical significance was set at P ≤ 0.05. Differences between individual means were determined using Newman–Keuls multiple-comparison test.

Results

Rapalog treatments significantly impaired reproductive parameters in adult mice

We employed an in vivo mouse model to study the effects of chronic rapalog treatment on the male reproductive system. In the first study, adult (P>60) mice were injected with vehicle-alone or the mTORC1 inhibitors sirolimus or Everolimus at high doses to define the mode of action (Figure 1A). As shown in Figure 1B, treatment with sirolimus and Everolimus negatively impacted body weights and all measured reproductive parameters. Testis weights, testis weights corrected for body weight, and seminal vesicle weights were all reduced in all treatments compared with those from vehicle-treated mice. Cauda epididymal sperm counts were significantly decreased in sirolimus- and Everolimus-treated mice (−78 and −77%, respectively, Figure 1B).

Figure 1.

Short-term treatment of adult mice with high-dose sirolimus and Everolimus significantly diminished reproductive parameters and disrupted testis histology. (A) Mice were treated daily from P60 to P90 and euthanized on P90. (B) Body weights, testis weights, cauda sperm counts, and seminal vesicle weights were all significantly reduced in treated mice. (C) Compared with vehicle-treated control tubules, diameters were reduced in rapalog-treated mice. (D–F) Testis sections from mice treated with vehicle-alone (D, D’), 10 μg/g sirolimus (E, E’), and 5 μg/g Everolimus (F, F’) were stained with hemotoxylin and eosin (H&E). D’–F’ are enlarged images of D–F. White arrows (E’) indicate remaining spermatogonia, black arrows (F’) indicate vacuoles. Scale bars: D–F = 200 μm; D’–F’ = 100 μm. Statistically significant differences (P < 0.05) are indicated by asterisks.

Decreased testis weights and reduced cauda epididymal sperm counts of rapalog-treated mice suggested disrupted spermatogenesis. To examine this, we compared the testis histology of vehicle- versus rapalog-treated animals. Seminiferous tubule diameters were reduced in both sirolimus- and Everolimus-treated mice (−36 and −30%, respectively, Figure 1C). Among animals from the same treatment groups, there was some variability in the extent of histological defects. Typical examples of testes from vehicle- and rapalog-treated testes are shown in Figure 1D–F. Seminiferous tubules from sirolimus-treated animals frequently had vacuoles, suggestive of sites of germ cell loss [30]. Spermatogonia were the only consistently present cell type, and more advanced germ cell types (e.g., spermatocytes, spermatids) were rarely seen (Figure 1E). Less-affected sirolimus-treated tubules contained numerous round and elongating spermatids, but low numbers of condensing spermatids (data not shown). The seminiferous epithelia of testes from Everolimus-treated mice were also vacuolated; there was substantial disorganization of some tubules, whereas others contained mature condensing spermatids (Figure 1F). There was also an increase in connective tissue in the interstitium surrounding the significantly shrunken tubules, and this effect was more pronounced in sirolimus-treated mice (Figure 1E and E’). Together, these data reveal treatments with sirolimus and Everolimus significantly adversely affected germ cell development in adult animals.

Rapalog treatments significantly impaired reproductive parameters in juvenile mice

Since rapalogs such as Everolimus are FDA approved for treatment of young human males [31], we next investigated their effects on reproductive health of male mice in which treatment was initiated in juveniles at P20, before the first round of spermatogenesis had completed (Figure 2A). In these young male mice, as in adults, there were significant decreases in body weight in response to both sirolimus and Everolimus apparent as early as 5 days after initiating treatment. Both treatment groups had decreased testis weights, which were statistically significant even after normalizing to reduced body weights (Figure 2B). To measure the output of spermatogenesis, we quantified cauda epididymal sperm numbers from each group. Compared with vehicle-treated controls, there were −84 and −65% decreases in sperm numbers in animals treated with sirolimus and Everolimus, respectively (Figure 2B). There were also significant decreases in the weights of seminal vesicles following rapalog treatment (Figure 2B). Similar to adult mice, seminiferous tubule diameters of mice receiving rapalogs as juveniles were reduced ~ 25% as compared with vehicle-treated controls (Figure 2C). Histological examination revealed that, in comparison to vehicle-treated controls (Figure 2D), seminiferous epithelia from sirolimus-treated mice contained frequent vacuoles and few elongating spermatids (Figure 2E). Testes from Everolimus-treated mice were also adversely affected with similar defects, but contained more elongating and condensing spermatids (Figure 2F).

Figure 2.

Short-term treatment of juvenile mice with high-dose sirolimus and Everolimus significantly diminished reproductive parameters and disrupted testis histology. (A) Mice were treated daily from P20 to P50 and euthanized on P50. (B) Body weights, testis weights, cauda sperm counts, and seminal vesicle weights were all significantly reduced in treated mice. (C) Compared with vehicle-treated control tubules, diameters were reduced in rapalog-treated mice. (D–F) Testis sections from mice treated with vehicle-alone (D, D’), 10 μg/g sirolimus (E, E’), and 5 μg/g Everolimus (F, F′) were stained with H&E. D’–F’ are enlarged images of D–F. Scale bars: D–F = 200 μm; D’–F’ = 100 μm. Statistically significant differences (P < 0.05) are indicated by asterisks.

Sirolimus and Everolimus inhibit mTOR signaling in juvenile and adult mouse testes

Rapalogs such as Everolimus were designed to inhibit mTORC1 kinase signaling, thus blocking its downstream activities including stimulation of cellular growth, proliferation, and differentiation. Since this is the first report of examining testes from Everolimus-treated animals, we verified mTORC1 inhibition in germ cells. Testes from juvenile and adult vehicle- and rapalog-treated mice were immunostained to detect total MTOR. As expected, both juvenile and adult rapalog-treated mice showed no apparent differences in total MTOR compared with vehicle-treated controls (Supplementary Figure S1A–H). We next assessed the phosphorylation status of a key downstream target, ribosomal protein S6 (P-RPS6), which we have previously shown was lost in differentiating spermatogonia when mTORC1 signaling was lost [14, 16, 32]. While P-RPS6 was readily detectable in germ cells of vehicle-treated controls, it was undetectable in those from mice of all ages treated with sirolimus or Everolimus (Supplementary Figure S1I–P).

Sirolimus and Everolimus inhibited spermatogonial differentiation in juvenile and adult mice

RA provides the requisite signal for spermatogonial differentiation in mice and rats [33–36] and activates mTORC1 [13]. The receptor tyrosine kinase KIT is essential for spermatogonial differentiation [37], and recent work from our laboratory showed that RA activated efficient translation of suppressed mRNAs encoding KIT (as well as the transcription factors SOHLH1 and SOHLH2) in an mTORC1-dependent manner [13, 14, 16]. Therefore, we posited the observed histological defects and reduced spermatogenesis output following rapalog treatment stemmed at least in part from impaired spermatogonial differentiation. To address this hypothesis, we evaluated spermatogonial fate by quantifying, in testes from juvenile and adult mice following rapalog treatment, numbers of undifferentiated (ZBTB16/PLZF+) and differentiating (KIT+) spermatogonia. Numbers of KIT+ spermatogonia were indeed reduced following treatment with either rapalog, regardless of the age of the mouse (Figure 3A and B). Conversely, the ratios of undifferentiated (ZBTB16+) spermatogonia to the total remaining germ cells were significantly increased in mice treated with sirolimus or Everolimus compared to vehicle-treated controls (Figure 3A and B).

Figure 3.

Juvenile and adult rapalog-treated testis exhibit impaired spermatogonial differentiation. (A) Immunostaining was done to detect KIT+ differentiating and ZBTB16+ undifferentiated spermatogonia in testes from juveniles (2 left-most columns) and adults (2 right-most columns). KIT and ZBTB16 are in green, and TRA98 is in red. (B) Numbers of KIT+ and ZBTB16+ germ cells (TRA98+) were quantified. Scale bars = 100 μm. Asterisks indicate statistically significant differences (P < 0.05).

We next examined whether the observed reduction in germ cells corresponded with changes in apoptosis. Testes from vehicle- and rapalog-treated animals were immunostained for established marker of apoptosis (cleaved or c-PARP1), as we have done before [13, 14, 27, 32, 38–40]. Testes from juvenile mice treated with sirolimus or Everolimus and sacrificed immediately after treatment had significantly increased numbers of c-PARP1+ apoptotic cells (Supplementary Figure S2B). Results were similar in adult mice, as both Everolimus- and sirolimus-treated animals had significant, albeit moderate increased numbers of c-PARP1+ apoptotic germ cells as compared with controls (Supplementary Figure S2A and B).

Lower dose of Everolimus exerted lesser impacts on reproductive parameters of adult mice

The rapalog doses used in Figures 1–3 were employed to define the mode of action, and were higher than those used clinically to treat human patients. Therefore, we next utilized a dose ~ 6-fold lower (0.17 μg/g) than the clinical dose used in humans to treat cancer (0.083 μg/g in humans; 1.02 μg/g in mice using allometric scaling). Importantly, this lower dose of Everolimus did result in loss of RPS6 phosphorylation, indicating inhibition of mTORC1 signaling (Supplementary Figure S1O and P). However, it did not significantly reduce body weights, although testis weights and testis weights corrected for body weights were significantly reduced, as compared with vehicle-treated controls (Figure 4B). Cauda epididymal sperm counts were not significantly reduced by these treatments, and histological analyses revealed smaller reductions in tubule diameters (−14%) when compared with high-dose Everolimus (Figure 4C). Compared with controls (Figure 4D), mice treated with low doses had observable histological defects (vacuolated tubules) within their seminiferous epithelia (Figure 4D and E). As expected, based on minor histological differences, testes of mice receiving this lower dose of Everolimus had no significant increases in apoptotic cell numbers (Supplementary Figure S2B).

Figure 4.

Short-term treatment of adult mice with low dose (0.17 μg/g) Everolimus decreases testis size. (A) Mice were treated daily from P60 to P90 and euthanized on P90. (B) Body weights, cauda sperm counts, and seminal vesicle weights were similar in control and Everolimus-treated mice, but testis weights were significantly reduced in treated mice. (C) Compared with vehicle-treated control tubules, diameters were reduced in Everolimus-treated mice. (D–E) Testis sections from mice treated with vehicle-alone (C, C’) and Everolimus (E, E’) were stained with H&E. D’–E’ are enlarged images of D–E. Arrows in E’ indicate vacuoles. Scale bar (in D) = 200 μm, (in D’) = 100 μm. Statistically significant differences (P < 0.05) are indicated by asterisks.

Reproductive parameters of adult and juvenile Everolimus-treated mice (all doses) significantly improved after cessation of treatment

Since rapalog treatments exerted their greatest apparent effects on differentiating spermatogonia and later germ cells (e.g., spermatocytes and spermatids), we predicted reproductive deficits in rapalog-treated mice would improve following treatment cessation and a recovery period. We tested this by examining treated mice after recovery periods: 1—adults, treated from P60 to P90 and allowed to recover for 30 days; 2—juvenile, treated from P20 to P50 and allowed to recover for 70 days.

In the first group, adult mice were allowed to recover for 30 days after cessation of treatment. As compared with vehicle-treated animals, adults treated with 10 μg/g sirolimus and 5 μg/g Everolimus retained significantly reduced body weights, testis weights, testis weights adjusted for body weight, cauda epididymal sperm counts, and seminal vesicle weights (Figure 5B). Seminiferous tubule diameters of these animals were increased (Figure 5C) compared with those immediately after treatment (Figure 1B); however, they remained significantly lower than vehicle-treated controls (Figure 5C and D). Histological analyses of seminiferous epithelia from these mice revealed improved cellular and morphological organization, although vacuoles and defects in tubular morphology were still regularly observed (Figure 5E and F). Testes from adult mice treated with the lower dose (0.17 μg/g) of Everolimus and allowed to recover for 30 days had normalized seminiferous tubule diameters and histology (Figure 5C, G and H). In these mice, cauda epididymal sperm numbers were similar to those of vehicle-treated controls, although testis weights and testis weights adjusted for body weight remained significantly reduced (Figure 5C). The histological abnormalities observed in the interstitium after treatment (Figure 1E and F) were not apparent after recovery (Figure 5E and F).

Figure 5.

Dose-dependent recovery of reproductive parameters in rapalog-treated adult mice. (A) Adult mice were treated from P60 to P90, and allowed to recover for 30 days. (B) Testis weights, testis/body weight ratios, cauda epididymal sperm counts, and seminal vesicle weights were measured from testes from each rapalog treatment group. (C) Seminiferous tubule diameters remained reduced only in mice treated with highest sirolimus (10 μg/g) and Everolimus (5 μg/g) doses. (D–G) Testis sections were stained with H&E from mice treated with vehicle-alone (D), sirolimus (E), or Everolimus (F–G), with doses indicated on each image. Asterisks indicate statistical differences at P < 0.05. Scale bar (in D) = 200 μm.

In the second group, sirolimus- and Everolimus-treated juvenile mice were allowed to recover for 70 days prior to euthanasia and tissue harvest. Testis weights, testis weights corrected for body weight, and cauda epididymal sperm counts of sirolimus-treated mice all remained lower than those of vehicle-treated controls (Figure 6B). In contrast, in Everolimus-treated mice, each of these parameters except for cauda epididymal sperm counts (which remained lower) returned to control levels (Figure 6B). From a histological perspective, there was a remarkable recovery of the seminiferous epithelium in both sirolimus- and Everolimus-treated mice (compare Figures 2E and F to 6E and F), such that they looked similar to vehicle-treated controls (Figure 6D). To determine whether mTORC1 signaling was restored in rapalog-treated animals, we immunostained testes from these treated mice to detect P-RPS6. All rapalog-treated testes contained spermatogonia that were robustly P-RPS6+ (Supplementary Figure S3A–H).

Figure 6.

Juvenile mice treated with high dose Everolimus (5 μg/g) recovered reproductive parameters. (A) Juvenile males were treated from P20 to P50, and then allowed to recover for 70 days. (B) Testis weights and testis/body weight ratios, cauda sperm counts, and seminal vesicle weights remained lower for sirolimus-treated mice. In those receiving 5 μg/g Everolimus, only sperm counts remained statistically reduced in comparison to vehicle-treated controls. (C) Tubule diameters of rapalog-treated mice were not different from vehicle-treated controls. (D–F) Testis sections from mice treated with vehicle-alone (D, D’), 10 μg/g sirolimus (E, E’), and 5 μg/g Everolimus (F, F′) were stained with H&E. D’–F′ are enlarged images of D–F. Scale bar (in D) = 200 μm, in (D’) = 100 μm.

Sirolimus- and Everolimus-treated mice are able to sire offspring following recovery

Taken together, results revealed markedly different dose-dependent extents of recovery of reproductive phenotypes (to control values) in sirolimus- and Everolimus-treated mice. Indeed, both juvenile and adult mice treated with higher doses of sirolimus (10 μg/g) and Everolimus (5 μg/g) still had impaired spermatogenesis after recovery, with reduced sperm counts (Figures 5B and 6B).

The ultimate readout of any treatment on spermatogenesis is the ability of treated mice to sire offspring. Therefore, we paired vehicle- and rapalog-treated males 10 days after cessation of treatment (n = 2 males treated with vehicle, 10 μg/g sirolimus, 5 μg/g Everolimus, or 0.17 μg/g Everolimus) with CD-1 WT females in a 60-day fertility trial (Figure 7A). Females were removed at the end of the 60-day period and observed for another 30 days, and any additional litters recorded. Results of this trial are summarized in Figure 7B. Vehicle-treated males had six litters (71 pups), whereas 10 μg/g sirolimus-treated males had three litters (30 pups) during that interval. Males from both Everolimus treatments also had litters; the 5 μg/g group had five (58 pups) and the 0.17 μg/g group had six litters (68 pups). All males were euthanized after the 60-day fertility trial, and testes from rapalog-treated males lacked apparent histological defects (Figure 7C–F and C’–F’).

Figure 7.

Rapalog-treated males were able to sire offspring and largely recovered normal reproductive parameters after 60-day breeding trial. (A–B) The results from a 60-day breeding trial (n = 2 males for each treatment) are depicted in table format. N/A = Not applicable, these litters were not produced. (C–F) Testis sections were stained with H&E from mice treated with vehicle-alone (C), sirolimus (D), or Everolimus (E–F), with doses indicated on each image. Asterisks indicate statistical differences at P < 0.05. Scale bars: C–F = 200 μm, C’–F’ = 100 μm.

Discussion

Summary

Here, we report the first comprehensive study of the effects of Everolimus treatment on mammalian spermatogenesis and male fertility. Doses were administered to define mode of action as well as to bracket human clinical doses. The ability to sire offspring was examined after cessation of treatment. Treatments caused reproductive deficits including decreased testis weights and reduced sperm counts, which were underlain by increased apoptosis and impaired spermatogonial differentiation. Importantly, Everolimus caused few negative impacts on spermatogenesis and male fertility, and all treated animals recovered spermatogenesis and were able to sire offspring after cessation of treatment.

We used mice as our experimental model to study rapalog effects on mammalian male reproduction. In addition to being relatively inexpensive and thus facilitating larger sample sizes, mice have high fecundity, highly organized stages of spermatogenesis, and a short time to reproductive maturity. We denoted mice as “juveniles” at P20, as this is a point when the first round, or wave of spermatogenesis is incomplete, and the most advanced germ cells are secondary spermatocytes and perhaps a small number of haploid round spermatids. This age was chosen to avoid the potential for rapalogs to disrupt significant developmental reproductive events such as Sertoli cell maturation and formation of the blood–testis barrier, which occur in mice around P15 [41]. We denoted “adult” mice as those older than P60; after this age, the first round of spermatogenesis is complete, and males are sexually mature and fertile. These ages were selected to model human patients receiving rapalog treatment who are juvenile (germ cells have not yet completed a round of spermatogenesis) and adult (germ cells are undergoing steady-state spermatogenesis).

In this study, mice were treated with rapalogs for relatively short intervals (1 month), as compared with the much longer (up to decades) treatment intervals expected in human patients. However, 1 month represents a significant portion of the laboratory mouse lifespan (∼24 months), and especially of the productive male reproductive lifespan (lasting from ∼2 to 9 months of age). The cumulative effects of multiple years of chronic rapalog treatment on human male fertility are currently unclear. The significant dose-dependent negative reproductive outcomes we observed and the minor variability in response between outbred CD-1 mice within the same treatment groups lead us to predict human male patients may have different threshold levels, which may need to be adjusted for each individual. Perhaps a compromise can be reached to choose the lowest effective dose for immunosuppression or chemotherapy balanced with the maximum output of spermatogenesis as measured by ejaculate sperm counts.

We recently reported, using pharmacologic and genetic approaches, the requirement for mTORC1 in mouse spermatogonial proliferation and differentiation in vivo during the first wave of spermatogenesis [14, 16, 32]. Our findings here using older (juvenile and adult) mice indicate spermatogonial differentiation was significantly impaired. Indeed, the percentage of spermatogonia that were differentiating (KIT+) was significantly reduced in rapalog-treated compared with vehicle controls. As expected, the ratios of ZBTB16+ undifferentiated spermatogonia to the total remaining TRA98+ germ cells were increased. In a recent study in mice [21], it was shown that sirolimus-impaired mTORC1 activation (assessed by decreased phosphorylation of RPS6KB1 or P70S6K) was restored by co-treatment with RA. This result suggests retinoid treatment may be worth considering as an option in rapalog-treated human patients to promote spermatogonial differentiation and increase sperm production.

We administered high rapalog doses (10 μg/g sirolimus, 5 μg/g Everolimus) for 1 month in order to define the in vivo mode of action. Results for sirolimus were similar to other recent rodent studies, which reported combinations of reduced testis and body weights, disrupted spermatogenesis, and reduced sperm counts [17, 19–22]. Adult and juvenile mice had significantly reduced seminal vesicle weights and testes with abnormal interstitial histological appearance. This appearance was typified by increased connective tissue surrounding the shrunken seminiferous tubules. Notably, the adult males treated for 30 days with 10 μg/g sirolimus would clearly be expected to be infertile, as they lacked advanced germ cells, including sperm. These effects were not observed in low dose (0.17 μg/g) Everolimus-treated mice. This suggests higher doses exerted a deleterious side effect on Leydig cell function and testosterone production, which we are currently investigating.

We were surprised to see the high amount of recovery of reproductive parameters in all groups in juvenile and adult mice, but especially in high-dose rapalog-treated adult mice. In total, 30 and 70 days after cessation of treatment, there were dose-dependent residual effects on testis weights and cauda epididymal sperm counts. Rapalog-treated juvenile mice were allowed to recover for an extended period of time (70 days). After this period, their body weights returned to control levels, although there were lingering effects on reproductive parameters. Sirolimus-treated mice retained smaller testes and reduced cauda epididymal sperm counts, whereas Everolimus-treated mice only retained reduced sperm counts. However, histology of treated animals was nearly indistinguishable from controls (compared with adult 70-day recovery when available).

Although Everolimus treatment did have the capacity, at the higher 5 μg/g dose, to reduce numbers of differentiating spermatogonia, it did not negatively impact numbers of undifferentiated spermatogonia. Since undifferentiated spermatogonia (particularly, SSCs) provide the foundation for steady-state spermatogenesis, they can restore spermatogenesis to normalcy during a recovery period after cessation of treatment, which occurred in all Everolimus-treated groups. These results suggest that the mTORC1 signaling pathway (or its downstream targets) represents therapeutic targets for male contraception, as rapalogs inhibited spermatogenesis with the following key outcomes: 1—safety (Everolimus is already FDA approved); 2—effectiveness (injected rapalogs clearly reached target spermatogenic cells to inhibit spermatogenesis); 3—reversibility (spermatogenesis was halted by 10 μg/g sirolimus as shown in Figure 1, but treated males were able to sire litters after drug treatment cessation, as shown in Figure 7).

In summary, we demonstrated several parameters of reproductive function were negatively impacted by administration, to both juvenile and adult mice, of sirolimus or Everolimus. Rapalog treatment reduced spermatogonial differentiation and increased germ cell apoptosis. While a significant reduction in cauda sperm numbers following administration of low-dose Everolimus was not observed, perhaps continued long-term treatment, as is the case for utilization of this drug in the clinical setting, may eventually lead to reduced sperm numbers. Results from this study provide a positive assessment of relatively low levels of detrimental effects of rapalogs on male germ cell development. However, it may be prudent for patients hoping to father children to have their sperm counts monitored during and after treatment, and to consider “biobanking” sperm (in the case of pubertal and adult patients) or testis tissue (in the case of prepubertal patients) prior to rapalog treatment.

Conflict of interest

The authors have declared that no conflict of interest exists.