Sperm calcineurin inhibition prevents mouse fertility with implications for male contraceptive

See allHide authors and affiliations

Science  23 Oct 2015:
Vol. 350, Issue 6259, pp. 442-445
DOI: 10.1126/science.aad0836

Mouse work may lead to male contraceptive

Unintended pregnancies are a major health issue worldwide. Although oral contraceptives were developed decades ago for use in women, there are no male oral contraceptives. Miyata et al. show that genetic deletion or drug inhibition of sperm-specific calcineurin enzymes in mice cause male sterility (see the Perspective by Castaneda and Matzuk). Although calcineurin inhibitors resulted in male infertility within 2 weeks, fertility recovered 1 week after halting drug administration. Because the sperm-specific calcineuin complex is also found in humans, its inhibition may be a strategy for developing reversible male contraceptives.

Science, this issue p. 442, see also p. 385


Calcineurin inhibitors, such as cyclosporine A and FK506, are used as immunosuppressant drugs, but their adverse effects on male reproductive function remain unclear. The testis expresses somatic calcineurin and a sperm-specific isoform that contains a catalytic subunit (PPP3CC) and a regulatory subunit (PPP3R2). We demonstrate herein that male mice lacking Ppp3cc or Ppp3r2 genes (knockout mice) are infertile, with reduced sperm motility owing to an inflexible midpiece. Treatment of mice with cyclosporine A or FK506 creates phenocopies of the sperm motility and morphological defects. These defects appear within 4 to 5 days of treatment, which indicates that sperm-specific calcineurin confers midpiece flexibility during epididymal transit. Male mouse fertility recovered a week after we discontinued treatment. Because human spermatozoa contain PPP3CC and PPP3R2 as a form of calcineurin, inhibition of this sperm-specific calcineurin may lead to the development of a reversible male contraceptive that would target spermatozoa in the epididymis.

Calcineurin is a Ca2+- and calmodulin-dependent serine-threonine phosphatase that plays a major role in calcium signaling (1, 2). In the immune system, calcineurin activates T cells by dephosphorylating the transcription factor NFAT (nuclear factor of activated T cells), and the dephosphorylated NFAT up-regulates the expression of interleukin-2 (1, 2). This process is suppressed by calcineurin inhibitors, such as cyclosporine A (CsA) and FK506, that are mainstays of immunosuppressive therapy after organ transplantation (1, 2). In the male reproductive system, animal experiments have revealed that CsA and FK506 have deleterious effects on spermatogenesis and epididymal sperm maturation (3, 4). Further, in vitro treatment of spermatozoa with these drugs impairs sperm motility and the acrosome reaction (5, 6). These data suggest important roles of calcineurin in male fertility; however, the existence of several isoforms expressed in the testis hampered the clarification of their functions and pharmacological processes.

Calcineurin exists as a heterodimer composed of a catalytic and a regulatory subunit. In mammals, three isoforms of the catalytic subunit (PPP3CA, PPP3CB, and PPP3CC) and two isoforms of the regulatory subunit (PPP3R1 and PPP3R2) have been identified. Ppp3ca, Ppp3cb, and Ppp3r1 are expressed ubiquitously, whereas Ppp3cc and Ppp3r2 are expressed strongly in the mouse testis (fig. S1A) (7). PPP3CC and PPP3R2 were not detected in the testis and epididymis of c-Kitw/wv mice, which lack differentiating germ cells (Fig. 1A). This indicates that Ppp3cc and Ppp3r2 are the only subunits expressed in spermatogenic cells. In mature spermatozoa, both PPP3CC and PPP3R2 are localized in the tail (Fig. 1B). Consistent with the existence of the calcineurin isoforms, Ca2+-dependent phosphatase activity is inhibited by CsA in mouse spermatozoa (fig. S1B). When mouse Ppp3cc and Ppp3r2 were expressed in human embryonic kidney 293T (HEK293T) cells, heterodimerizaion of PPP3CC and PPP3R2 stabilized the calcineurin complex (Fig. 1C). This is consistent with genetic deletion of Ppp3r1, which leads to loss of PPP3CA and/or PPP3CB (8). Protein extracts from transfected cells contain Ca2+-dependent phosphatase activity that was blocked by CsA (Fig. 1D). Thus, we conclude that the PPP3CC-PPP3R2 complex is the sperm-specific calcineurin (sperm calcineurin).

Fig. 1 Sperm calcineurin is a complex of PPP3CC and PPP3R2.

(A) Both PPP3CC and PPP3R2 were detected in differentiating germ cells. (B) Both PPP3CC and PPP3R2 are localized in the sperm tail. IZUMO1 localized in the head and BASIGIN localized in the tail were used as controls. (C) Mouse PPP3CC-FLAG (cc-FLAG) and/or mouse PPP3R2 (r2) were overexpressed in HEK293T cells. (D) HEK293T cells that overexpressed both PPP3CC-FLAG and PPP3R2 have Ca2+-dependent phosphatase activity that is blocked by 100 nM CsA. n = 3.

To elucidate the physiological functions of sperm calcineurin, we deleted (i.e., knocked out) the Ppp3cc gene in mice (fig. S2, A and B). We confirmed that PPP3CC was depleted in both testis and spermatozoa in Ppp3cc−/− males (fig. S2C). The PPP3R2 signal was less in Ppp3cc-null testis and spermatozoa, which further indicated that PPP3CC stabilizes PPP3R2 (fig. S2D). As expected, Ca2+-dependent phosphatase activity was reduced in Ppp3cc-null spermatozoa (fig. S2E). In Ppp3cc−/− mice, there were no overt abnormalities in spermatogenesis, epididymal sperm morphology, or sperm counts (fig. S3, A to F).

Although Ppp3cc−/− males copulated, they were infertile (Fig. 2A). When we investigated sperm migration in the female reproductive tract using transgenic spermatozoa that have DsRed2 in the mitochondria (9), fewer spermatozoa reached the ampulla compared with those of controls (fig. S4A). Although this could explain, in part, the male infertility, the presence of spermatozoa in the ampulla suggests that there are other factors that render the Ppp3cc−/− males infertile. To examine other possible factors, we performed in vitro fertilization (IVF) and found that Ppp3cc-null spermatozoa cannot fertilize cumulus-intact oocytes (Fig. 2B). Further analysis revealed that Ppp3cc-null spermatozoa could pass through the cumulus cell layers (fig. S4B) and could bind to the zona pellucida (ZP) (Fig. 2C) but failed to fertilize cumulus-free ZP intact oocytes (fig. S4C); therefore, the ZP was the site of the problem. Once the ZP was removed, Ppp3cc-null spermatozoa could fuse with oocytes (fig. S4D), which confirmed that Ppp3cc-null spermatozoa are defective in zona penetration. When IVF was performed in a medium containing glutathione as a reducing agent to destabilize the ZP (fig. S5A) (10, 11), oocytes were fertilized by Ppp3cc-null spermatozoa (Fig. 2D and fig. S5A). These fertilized eggs developed to term (Fig. 2E and fig. S5B), which indicates that sperm calcineurin is not required for sperm genomic integrity. Fertility of Ppp3cc−/− males was also rescued by expressing mCherry-tagged Ppp3cc transgene under a testis-specific Clgn promoter (fig. S6, A to C) (12).

Fig. 2 Ppp3cc−/− male is infertile because of impaired ZP penetration.

(A) Pregnancy rate (pregnancy over vaginal plug) is presented as a percentage. (B) IVF with cumulus-intact oocytes. (C) Sperm-ZP binding assay in vitro. (D) IVF with cumulus-intact oocytes in medium containing glutathione (GSH). (E) Pups were obtained from eggs fertilized with Ppp3cc-null spermatozoa in (D).

To understand why null spermatozoa could not penetrate the ZP, we investigated the acrosome reaction, as this is a prerequisite for spermatozoa to penetrate through the ZP (13). However, the acrosome reaction occurred normally in Ppp3cc-null spermatozoa (fig. S7, A and B). Next, we investigated sperm motility using computer-assisted sperm analysis. Although there were no differences in the percentage of motile spermatozoa (fig. S7C), velocity parameters of Ppp3cc-null spermatozoa (C57BL6/DBA2 background) were lower than those of control C57BL6/DBA2 spermatozoa (Fig. 3A). However, knockout (KO) sperm motility was still comparable to that of C57BL6 wild-type mice (WT) (table S1). Therefore, differences in velocities alone cannot explain the penetration defect.

Fig. 3 Sperm calcineurin is necessary for a flexible midpiece.

(A) Sperm motility at 10 min and 120 min after sperm suspension. VAP, average path velocity; VSL, straight-line velocity; and VCL, curvilinear velocity. (B) The percentage of hyperactivated spermatozoa. (C) Flagellar bending patterns. Single frames throughout one beating cycle are superimposed. The midpiece (black arrow) was inflexible in the Ppp3cc-null spermatozoa. (D) The percentage of the spermatozoa with a rigid midpiece. The number of spermatozoa with a rigid midpiece over the number of spermatozoa examined is presented in the parentheses above the column.

We then investigated the percentage of hyperactivated spermatozoa and discovered that it was significantly lower in Ppp3cc−/− mice than in control spermatozoa (Fig. 3B and fig. S7D), which suggests that Ppp3cc-null spermatozoa are defective in ZP penetration due to impaired hyperactivation. To elucidate why Ppp3cc-null spermatozoa fail to exhibit hyperactivation, we further analyzed sperm motility. Although beat frequencies of Ppp3cc-null spermatozoa were normal (fig. S7E), the midpiece of null spermatozoa bent slightly in a direction opposite from the hook of the acrosome (anti-hook) (99%, 110 out of 111 spermatozoa) (14) and was inflexible (Fig. 3, C and D, and movies S1 to S4). In contrast, defects were not observed in the principal piece of null spermatozoa (Fig. 3C). When WT spermatozoa are hyperactivated, the curvature of the midpiece bend increases (14); however, Ppp3cc-null spermatozoa do not exhibit this increase. Thus, PPP3CC is required to develop the flexible midpiece for hyperactivation and to penetrate the ZP. Identical phenotypes were observed in Ppp3r2 KO mice (fig. S8, A to F), which reconfirmed the importance of sperm calcineurin in midpiece flexibility and male fertility.

PPP3CC is localized close to CATSPER1, a Ca2+ channel required for hyperactivation (15, 16). To investigate whether PPP3CC is downstream of CATSPER1, we analyzed sperm motility of Catsper1−/− mice. However, the midpiece of Catsper1-null spermatozoa was not rigid (fig. S8G), which indicates that sperm calcineurin is activated by calcium influx from different calcium channels and/or internal calcium stores.

To understand why the Ppp3cc-null midpiece remains rigid, we investigated microtubule (MT) sliding using spermatozoa from which membranes had been removed. MTs extend from the neck region in both control and null spermatozoa with comparable frequencies, which suggests that axonemal dyneins can slide the MTs even in the KO midpiece (fig. S9, A to C, and movie S5). The midpiece that was bent in the direction of the hook of the acrosome (pro-hook) appeared in null spermatozoa (fig. S9A) (32% ± 8.0, no. of males = 3, no. of spermatozoa = 55). Because the mitochondrial sheath was removed in this assay (17), the mitochondrial sheath may contribute to the rigid midpiece bending in an anti-hook direction. Interaction between the mitochondrial sheath and the outer dense fibers may keep the midpiece rigid (18).

To further analyze the function of sperm calcineurin, we treated WT spermatozoa in vitro with CsA or FK506. Sperm motility and midpiece flexibility of these mature spermatozoa were not affected by calcineurin inhibitors (fig. S10, A and B). Furthermore, WT spermatozoa were able to fertilize normally in vitro, even in the presence of CsA or FK506 in the media (fig. S10C). Thus, sperm calcineurin activity is no longer necessary to confer midpiece flexibility in mature spermatozoa collected from the cauda epididymis.

To determine the effects of immunosuppressant drugs on immature spermatozoa, WT male mice were treated with CsA or FK506 for 2 weeks; these males became infertile (Fig. 4A), and spermatozoa from treated mice did not fertilize in vitro (Fig. 4B). When sperm motility was investigated, their midpiece was as inflexible as Ppp3cc- or Ppp3r2-null spermatozoa (Fig. 4C, movie S6 to S9). These results indicate that sperm calcineurin activity is important during the later stages of spermatogenesis or during sperm maturation in the epididymis. To elucidate where sperm calcineurin works, we administered CsA or FK506 for short periods. Spermatozoa with rigid midpieces appeared as early as 5 days after CsA administration and 4 days after FK506 administration (Fig. 4D). Because it takes ~10 days for spermatozoa to transit the epididymis (19), these results indicate that sperm calcineurin is essential to confer midpiece flexibility during epididymal transit. This is consistent with the fact that the sperm midpiece becomes flexible during epididymal transit (fig. S11A) (20). Fast transit of spermatozoa through the epididymis suggests that male fertility may recover quickly after stopping drug administration. As expected, fertility of male mice treated with CsA or FK506 recovered 1 week after halting drug administration (Fig. 4E). Sperm motility and midpiece flexibility also recovered after treatment (fig. S11, B to D). Considering these results in mice, we suggest that sperm calcineurin may be a target for reversible and rapidly acting human male contraceptives.

Fig. 4 Sperm calcineurin confers midpiece flexibility during epididymal transit.

(A and B) Male mice were treated with CsA or FK506 for 2 weeks, and we analyzed their in vivo fertility using superovulated females (A), and fertilization rate in vitro (with cumulus cells) (B). (C) The percentage of spermatozoa with a rigid midpiece. The number of spermatozoa with a rigid midpiece over the number of spermatozoa examined was presented in the parentheses above the column. (D) Male mice were treated with CsA or FK506 for a short period, and their midpiece was analyzed. (E) Male mice were treated with CsA or FK506 for 2 weeks, and their in vivo fertility was analyzed 1 week after stopping drug administration. Drug concentration for CsA, 80 mg/kg per day, and for FK506, 8 mg/kg per day.

In human spermatozoa, the catalytic subunit of calcineurin was detected using a pan-calcineurin antibody (21). With specific antibodies, we confirmed the existence of PPP3CC and PPP3R2 but not PPP3CA and PPP3CB in human spermatozoa (fig. S12A). The recombinant human PPP3CC-PPP3R2 exhibited Ca2+-dependent phosphatase activity that was blocked by CsA (fig. S12, B and C). Thus, human spermatozoa contain PPP3CC-PPP3R2 as a functional calcineurin. In support of this idea, it has been reported that human spermatozoa exhibit Ca2+-dependent phosphatase activity (21). In humans, sperm motility develops during epididymal transit (22). Although male patients treated with a therapeutic dosage of CsA did not show signs of infertility (23), there are studies showing that CsA treatment impairs the motility of spermatozoa (24). Further, the concentration of CsA in the blood correlates inversely with sperm motility (25), which suggests that the PPP3CC-PPP3R2 complex is also involved in sperm motility development in humans.

In the present study, we demonstrated that PPP3CC associates with PPP3R2 to form sperm calcineurin during spermatogenesis. Our KO mice reveal that sperm calcineurin is not essential for spermatogenesis. The impaired spermatogenesis reported with CsA and FK506 treatment could be attributed to the inhibition of somatic calcineurin and/or nonspecific inhibition of other molecules. We conclude that sperm calcineurin confers midpiece flexibility during epididymal transit, and this process is required to generate fertilization-competent spermatozoa. In the immune system, calcineurin dephosphorylates the transcription factor NFAT (1, 2); however, in spermatozoa, calcineurin plays roles in the epididymis after transcription has ceased. Therefore, sperm calcineurin does not confer midpiece flexibility by dephosphorylating transcription factors such as NFAT. Specific inhibition of sperm calcineurin or its interaction with substrates may lead to the development of reversible and rapidly acting male contraceptives that target spermatozoa in the epididymis but leave testicular function intact.

Supplementary Materials

Materials and Methods

Figs. S1 to S12

Table S1

References (2640)

Movies S1 to S9


  1. AKNOWLEDGMENTS: We thank Biotechnology Research and Development (Osaka, Japan) for generating mutant mice, D. E. Clapham (Howard Hughes Medical Institute, Boston, MA) for providing Catsper1 KO mice, and S. A. M. Young and F. Abbasi for review of the manuscript. We also thank M. Okabe for continuous encouragement. This work was partly supported by the Ministry of Education, Culture, Sports, Science and Technology and the Japan Society for the Promotion of Science (26830056 to H.M.; 15H05573 to Y.F.; 20670006, 25112007, and 25250014 to M.I.), JAMBIO, and Takeda Science Foundation. All of the gene-manipulated mouse lines were deposited into the Riken BioResource Center (with stock numbers as follows: Ppp3cc KO, RBRC05878; Ppp3r2 KO, RBRC09504; Ppp3cc-FLAG TG#1, RBRC09502; Ppp3cc-FLAG TG#2, RBRC09503; and Ppp3cc-mCherry TG, RBRC09501) and are available from M.I. under a material transfer agreement with Osaka University. The authors declare no conflict of interest.
View Abstract

Navigate This Article