Testis-enriched ferlin, FER1L5, is required for Ca2+-activated acrosome reaction and male fertility

Spermatozoa need to undergo an exocytotic event called the acrosome reaction before fusing with eggs. Although calcium ion (Ca2+) is essential for the acrosome reaction, its molecular mechanism remains unknown. Ferlin is a single transmembrane protein with multiple Ca2+-binding C2 domains, and there are six ferlins, dysferlin (DYSF), otoferlin (OTOF), myoferlin (MYOF), fer-1–like 4 (FER1L4), FER1L5, and FER1L6, in mammals. Dysf, Otof, and Myof knockout mice have been generated, and each knockout mouse line exhibited membrane fusion disorders such as muscular dystrophy in Dysf, deafness in Otof, and abnormal myogenesis in Myof. Here, by generating mutant mice of Fer1l4, Fer1l5, and Fer1l6, we found that only Fer1l5 is required for male fertility. Fer1l5 mutant spermatozoa could migrate in the female reproductive tract and reach eggs, but no acrosome reaction took place. Even a Ca2+ ionophore cannot induce the acrosome reaction in Fer1l5 mutant spermatozoa. These results suggest that FER1L5 is the missing link between Ca2+ and the acrosome reaction.


INTRODUCTION
Fertilization is a union of two gametes, spermatozoa, and oocytes. To fertilize eggs, spermatozoa need to travel a long distance in the female reproductive tract using flagella. During the migration, the acrosome located anterior to the nucleus in sperm heads needs to undergo an exocytotic event called the acrosome reaction to fertilize eggs (1,2). During the acrosome reaction, the outer acrosomal membrane fuses with the sperm plasma membrane and hydrolytic enzymes such as a serine protease (acrosin) (3,4), phospholipase A2 (5), and hyaluronidases (6) in the acrosome are exposed to facilitate fertilization. For hyaluronidases, hyaluronoglucosaminidase 5 (HYAL5) is localized in both plasma and acrosomal membranes of acrosome-intact spermatozoa and released during the acrosome reaction, while sperm adhesion molecule 1 (SPAM1), also called PH-20, is localized in the plasma membrane of acrosome-intact spermatozoa and retained after the acrosome reaction (7,8). Furthermore, izumo sperm-egg fusion 1 (IZUMO1), a type I transmembrane protein localized in the acrosomal membrane of acrosome-intact spermatozoa, relocates onto the cell surface after the acrosome reaction, which is a prerequisite for sperm-egg fusion (9).
As Ca 2+ plays critical roles in the acrosome reaction, Ca 2+ ionophores such as A23187 are potent inducers of the acrosome reaction. Several sperm proteins involved in Ca 2+ influx were found, such as calcium channel, voltage-dependent, R type, alpha 1E subunit (Ca v 2.3) (10); phospholipase C, delta 4 (PLCD4) (11); and transient receptor potential cation channel, subfamily C, member 2 (TRPC2) (12). Knockout (KO) mice of Ca v 2.3 or Plcd4 result in impaired acrosome reaction in response to solubilized zona pellucida (ZP), the extracellular matrix surrounding the eggs (10,11). In contrast, A23187 could induce the acrosome reaction even in these mutant spermatozoa (10,11), indicating that these molecules play roles upstream of intracellular Ca 2+ increase. The Ca 2+activated acrosome reaction is likely mediated by soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins that are involved in fusing vesicles with the target membrane (13,14). Still, the molecule that links Ca 2+ increase and SNARE proteins in the acrosome reaction remains unknown.
Ferlin is a single transmembrane protein with multiple Ca 2+binding C2 domains and is involved in Ca 2+ -mediated membrane fusion. Ferlin was first found in Caenorhabditis elegans, in which mutations cause male infertility due to defective fusion of internal membranous organelles with the sperm plasma membrane (15). Misfire, a ferlin ortholog in Drosophila melanogaster, is also essential for male fertility, with mutations resulting in impaired sperm plasma membrane breakdown after fertilization (16). In mammals, there are six ferlin proteins: dysferlin (DYSF), otoferlin (OTOF), myoferlin (MYOF), fer-1-like 4 (FER1L4), FER1L5, and FER1L6. KO mice of Dysf, Otof, and Myof exhibit phenotypes related to impaired membrane fusion, such as muscular dystrophy in Dysf (17,18), deafness in Otof (19), and abnormal myogenesis in Myof (20). Disease-associated mutations in DYSF and OTOF were also found in humans (21)(22)(23). In contrast, it is unknown if ferlins play critical roles in male fertility in mammals.
This study generated Fer1l4, Fer1l5, and Fer1l6 mutant mice and found that only Fer1l5 is required for male fertility. Even A23187 cannot induce the acrosome reaction in Fer1l5 mutant spermatozoa, suggesting that FER1L5 is indispensable for the Ca 2+ -activated fusion of plasma membrane and acrosomal membrane.

Fer1l4, Fer1l5, and Fer1l6 are expressed in the mouse testis
Phylogenetic analyses of ferlin proteins in several model organisms are shown in Fig. 1A. There are six ferlin proteins, DYSF, OTOF,  MYOF, FER1L4, FER1L5, and FER1L6, in mice. While DYSF,  OTOF, MYOF, FER1L5, and FER1L6 are conserved, FER1L4 is missing in humans. We focused on Fer1l4, Fer1l5, and Fer1l6 because their functions in vivo remain unclear. Consistent with other ferlins, FER1L4, FER1L5, and FER1L6 have a single transmembrane domain at the C terminus and several C2 domains that are thought to bind to Ca 2+ and regulate Ca 2+ -dependent membrane targeting (Fig. 1B). We checked their expression in mouse tissues with reverse transcription polymerase chain reaction (RT-PCR) and found that Fer1l4, Fer1l5, and Fer1l6 are expressed in the testis with additional Fer1l4 expression in the uterus and Fer1l6 expression in the heart (Fig. 1C). We performed RT-PCR using postnatal testis and found that Fer1l4 and Fer1l6 are expressed at week one, while Fer1l5 starts to express at week three (Fig. 1D). These results suggest that Fer1l4, Fer1l5, and Fer1l6 may play roles in male reproduction.

Fer1l5 is required for male fertility
To analyze the function of Fer1l4, Fer1l5, and Fer1l6 in vivo, we generated mutant mice for each gene. Exon 5 and 6 of Fer1l4, Exon 2 and 3 of Fer1l5, and Exon 4 and 5 of Fer1l6 were deleted, which results in frameshift mutations (fig. S1, A to D). Each mutant mice were viable and had no gross abnormality in appearance or behavior. We then analyzed male fertility by mating each mutant male with wild-type females. While Fer1l4 and Fer1l6 mutant males were fertile ( fig. S2A), the fertility of Fer1l5 mutant males was severely impaired ( Fig. 2A). Because both Fer1l4 and Fer1l6 lack the DYSF domain ( Fig. 1B) and start expression from week one in the testis (Fig. 1D), the function of FER1L4 and FER1L6 may be similar in the testes and compensate each other. Therefore, we generated Fer1l4 and Fer1l6 double mutant mice by breeding, but the obtained males were still fertile ( fig. S2A). We then focused on the analysis of Fer1l5. In contrast to male fertility, Fer1l5 mutant females were fertile ( fig. S2B). To ensure that Fer1l5 is responsible for impaired male fertility, we generated transgenic (Tg) mice that expressed FLAG-tagged or PA-tagged Fer1l5 under testis-specific calmegin (Clgn) promoter ( fig. S2C). We detected FER1L5 in the testis of Tg mice with immunoblotting following immunoprecipitation ( fig. S2, D and E) and confirmed that both Fer1l5-FLAG and Fer1l5-PA transgenes rescued impaired male fertility of Fer1l5 mutant mice ( fig. S2F).

Fer1l5 mutant spermatozoa are defective in ZP penetration and sperm-egg fusion
To clarify the cause of impaired male fertility of Fer1l5 mutant mice, we analyzed testis and cauda epididymis sections, but no apparent abnormalities were found in Fer1l5 mutant mice ( Fig. 2B and fig.  S3A). Furthermore, no differences were observed in sperm morphology ( Fig. 2C) or motility parameters as analyzed by the computer-assisted sperm analysis system ( Fig. 2D and fig. S3, B and C) between the control and Fer1l5 mutant mice.
We then performed in vitro fertilization and found that Fer1l5 mutant spermatozoa could not fertilize cumulus-intact eggs (Fig. 3, A and B), consistent with in vivo fertility results. Removing cumulus cells could not rescue impaired fertilization rates (Fig. 3C). Furthermore, even without the ZP, Fer1l5 mutant spermatozoa could not fertilize eggs ( Fig. 3D and fig. S4A). Notably, when we used 10 times more spermatozoa for insemination, no fertilization occurred with cumulus-intact or cumulus-free eggs in Fer1l5 mutant mice ( fig. S4B), but a few ZP-free eggs were fertilized ( Fig. 3E and fig. S4A). These results indicate that Fer1l5 mutant spermatozoa cannot fuse with eggs. We also did not observe any spermatozoa in the perivitelline space of ZP-intact eggs, which can be seen in the mutant spermatozoa lacking any known oocyte-fusing factors (24,25), suggesting that Fer1l5 mutant spermatozoa have defects not only in fusing with eggs but also in penetrating the ZP.

Fer1l5 mutant spermatozoa cannot undergo the acrosome reaction in vitro
Because it is widely accepted that only acrosome-reacted spermatozoa can penetrate the ZP and fuse with eggs, we analyzed the acrosome reaction in vitro by observing IZUMO1 that relocates onto the equatorial segment after the acrosome reaction (9). IZUMO1 was originally localized in the apical region (acrosomal cap) in both control and Fer1l5 mutant spermatozoa (10 min after the collection from the cauda epididymis; Fig. 4A). After 1, 2, and 4 hours of incubation in capacitation medium, IZUMO1 relocates to the equatorial segment of the control spermatozoa; however, IZUMO1 stays in the apical region of almost all Fer1l5 mutant spermatozoa (Fig. 4, A and B). Furthermore, even after A23187 treatment, almost no spermatozoa undergo the acrosome reaction in Fer1l5 mutant mice (Fig. 4, A and B). Impairment of the acrosome reaction in Fer1l5 mutant spermatozoa was rescued by PA-tagged Fer1l5 transgene (Fig. 4, A and B), indicating that restored fertility in this line is due to the restored ability of spermatozoa to undergo the acrosome reaction.
Because immunostaining of IZUMO1 of fixed spermatozoa cannot tell the differences between live and dead spermatozoa, we also analyzed the acrosome reaction rates of live spermatozoa using Red Body Green Sperm (RBGS) Tg mice, whose spermatozoa lose the enhanced green fluorescent protein (EGFP) fluorescence signal after the acrosome reaction (26). Consistent with IZUMO1 staining results (Fig. 4, A and B), Fer1l5 mutant spermatozoa could not undergo the acrosome reaction not only after 4 hours of incubation in capacitation medium but also after A23187 treatment ( fig. S5A). These results indicate that FER1L5 is required for the acrosome reaction and functions downstream of Ca 2+ influx.

Tyrosine phosphorylation, hyperactivation, and acrosomal swelling occur in Fer1l5 mutant spermatozoa
Capacitation is the biochemical process that spermatozoa undergo after ejaculation into the female reproductive tract and is thought to be a prerequisite for the acrosome reaction (27). Therefore, we observed tyrosine phosphorylation, a hallmark of the capacitation. After 1 and 2 hours of incubation in capacitation medium, enhanced tyrosine phosphorylation signals were observed in both the control and Fer1l5 mutant spermatozoa (Fig. 5A), suggesting that the signal transduction pathways that induce tyrosine phosphorylation were not impaired.
Next, we checked the hyperactivation of sperm flagella, which is important for penetrating the ZP (28,29). Hyperactivation occurs during capacitation in which the spermatozoa exhibit a high amplitude and asymmetrical beating pattern of flagella (28,29). We examined the maximal bending angle of the midpiece in primary antihook curvature, which is used to quantitatively analyze hyperactivation (30). Consistent with the control spermatozoa, the maximum bending angles increased in Fer1l5 mutant spermatozoa after 2 and 4 hours of incubation in capacitation medium (Fig. 5B), suggesting that Fer1l5 mutant spermatozoa could undergo hyperactivation without the acrosome reaction.
Using transmission electron microscopy, we also observed the acrosome swelling that occurs during the capacitation process and is thought to be required for the fusion of the plasma membrane and the outer acrosomal membrane during the acrosome reaction (31). After 10 min of incubation in capacitation medium, no abnormalities were observed in the acrosomal morphology in Fer1l5 mutant spermatozoa (Fig. 5C). After 4 hours of incubation, acrosome swelling was observed in both control and Fer1l5 mutant spermatozoa; however, the acrosome reaction did not occur in the Fer1l5 mutant spermatozoa (Fig. 5C). These results indicate that Fer1l5 mutant spermatozoa can undergo capacitation-associated processes such as tyrosine phosphorylation, hyperactivation, and acrosomal swelling, but no fusion between the plasma membrane and outer acrosomal membrane takes place.

FER1L5 interacts with SNARE proteins
It is known that OTOF, a ferlin that is essential for hearing, interacts with SNARE proteins such as syntaxin 1 (STX1) and regulates Ca 2+dependent SNARE-mediated membrane fusion (19,32). Furthermore, SNARE proteins are thought to regulate membrane fusion during the acrosome reaction (13,14). Because these studies suggest an interaction between ferlins and syntaxins, we focused on STX2, a protein that was suggested to be involved in the acrosome reaction (13). It has been reported that Stx2 mutant male mice were infertile because of the absence of round spermatids and mature spermatozoa, confirming that STX2 plays a role in the testis (33). However, this phenotype makes it difficult to analyze the acrosome reaction in the Stx2 mutant mice because of the lack of spermatozoa. To uncover the relationship, we coexpressed FLAG-tagged FER1L5 and PA-tagged STX2 in human embryonic kidney 293T (HEK293T) cells and performed an immunoprecipitation analysis using an anti-FLAG antibody (Fig. 6A). After immunoblotting, we found that FER1L5 interacts with STX2 (Fig. 6A). In contrast, FER1L5 did not interact with negative control, a disintegrin and metallopeptidase domain 1b (ADAM1B), which is localized in the sperm plasma membrane (Fig. 6A) (34). These results suggest that FER1L5 may regulate SNARE-mediated membrane fusion during the acrosome reaction.
Fer1l5 mutant spermatozoa can reach the eggs but are not acrosome-reacted in vivo Last, we observed sperm migration in the female reproductive tract using RBGS Tg mice to analyze the acrosomal status in vivo. We observed spermatozoa in the female reproductive tract 2 and 8 hours after observing vaginal plugs [equivalent to 14 and 20 hours after human chorionic gonadotropin (hCG) injection]. Fer1l5 mutant spermatozoa could pass through the utero-tubal junction (UTJ) that connect the uterus and oviduct ( fig. S5B) and reach the ZP surrounding the eggs in the ampulla (Fig. 6B), indicating that the acrosome reaction is not essential for spermatozoa to migrate in the female reproductive tract and pass through the UTJ and the layer of cumulus cells. Although Fer1l5 mutant spermatozoa could reach the eggs, no spermatozoa were acrosome-reacted (Fig. 6, B and C). These results indicate that FER1L5 is required for the acrosome reaction and penetrating the ZP in vivo, consistent with the analyses in vitro.

DISCUSSION
In this study, we reveal that FER1L5 regulates the acrosome reaction in mice, similar to FER-1 regulating the membranous organelle fusion in C. elegans and MISFIRE regulating the sperm plasma membrane breakdown in D. melanogaster. These results indicate that ferlin function in spermatozoa is evolutionarily conserved, although vesicle fusion during fertilization takes place at different times. In C. elegans, vesicle fusion occurs during spermatid activation (15), while in D. melanogaster, it occurs during plasma membrane breakdown after sperm entry into the oocytes (16). In mice, fusion occurs in the acrosome reaction during sperm migration in the female reproductive tract (Fig. 6D) (1, 2). Although vesicle fusion takes place at different times, these processes may be regulated by similar molecular mechanisms through ferlins. Intriguingly, Fer1l5 was not found in zebrafish, in which spermatozoa lack the acrosome (Fig. 1A).
In C. elegans, it has been reported that FER-1 is localized in the membranous organelles (35). It is tempting to speculate that FER1L5 is localized in the acrosomal membrane like FER-1 in the membrane organelles, but we could not obtain an anti-FER1L5 antibody that worked for immunoblotting or immunofluorescence to analyze its localization. In Fer1l5-FLAG or Fer1l5-PA Tg mice, we detected FER1L5-FLAG or FER1L5-PA using large amounts of testicular proteins, which hamper the analysis of FER1L5 localization. Therefore, we cannot rule out the possibility that FER1L5 plays roles in the testes and impaired acrosome reaction in Fer1l5 mutant mice is a consequence of defective spermatogenesis; however, previous proteomics analysis showed that mature spermatozoa contained FER1L5 in several mammalian species such as humans (36), cattle (37), yaks (38), and rams (39). Furthermore, our mass spectrometry (MS) analyses detected FER1L5 peptides in mouse mature spermatozoa ( fig. S6, A to C), indicating that FER1L5 may play roles in mature spermatozoa.
Using heterologous expression system, we found that FER1L5 could interact with syntaxin consistent with OTOF (19,32), suggesting that FER1L5 may regulate Ca 2+ -dependent SNARE-mediated vesicle fusion during the acrosome reaction. This idea is supported by the result that even a Ca 2+ ionophore cannot induce the acrosome reaction in Fer1l5 mutant spermatozoa. Furthermore, recent in silico structural modeling predicted that the C2 domain of human FER1L5 could bind to Ca 2+ (40), suggesting that FER1L5 may functions as a calcium sensor that regulates the acrosome reaction. It has been reported that OTOF acts as a calcium sensor that regulates synaptic vesicle exocytosis in cochlear hair cells (32).
Fer1l5 mutant mice can be a good model to study the acrosome reaction. In this study, we found that Fer1l5 mutant spermatozoa could undergo hyperactivation (Fig. 5B), suggesting that the acrosome reaction is not required for hyperactivation. Furthermore, we confirmed that the acrosome-intact spermatozoa could reach the ZP in vivo (Fig. 6B). Because the acrosome reaction is severely impaired, Fer1l5 mutant spermatozoa could not fertilize cumulusintact eggs in vitro even with a higher concentration of spermatozoa ( fig. S4B). In contrast, a few pups were obtained from Fer1l5 mutant males in vivo ( Fig. 2A). These results suggest that the acrosome reaction is induced more efficiently in vivo than in vitro. Considering that A23187 rarely induced the acrosome reaction in Fer1l5 mutant mice, Ca 2+ -dependent signaling pathway and other pathways such as mechanical stimulus (41) may be involved in inducing the acrosome reaction in vivo.
In summary, our results indicate that FER1L5 is important for the acrosome reaction and male fertility in mice. Identification of FER1L5 as a key molecule of Ca 2+ -activated acrosome reaction will shed light on the molecular mechanism of the acrosome reaction. Furthermore, human FER1L5 exhibits testis-enriched expression (42), and its function may be conserved in humans as well. Revealing FER1L5-associated molecules and their functions may help us understand the etiology of human male infertility.

Animals
All mice used in this study were purchased from CLEA Japan (Tokyo, Japan) or Japan SLC (Shizuoka, Japan). Mice were maintained under specific pathogen-free conditions in a temperaturecontrolled environment, 12-hour light/12-hour dark cycle, and ad libitum feeding. All animal experiments were approved by the Animal Care and Use Committee of the Research Institute for Immunofluorescence staining for IZUMO1. IZUMO1 (green) relocates to the equatorial segment after the acrosome reaction (asterisks), which was used to analyze acrosome reaction rates. Fer1l5 mutant spermatozoa rarely exhibit IZUMO1 relocation. Impaired acrosome reaction was rescued by Fer1l5-PA transgene. Nuclei are stained blue. (B) The acrosome reaction rates were analyzed by IZUMO1 staining after 10 min, 1 hour, 2 hours, and 4 hours incubation in capacitation medium. After 4 hours of incubation, Ca 2+ ionophore A23187 was added to induce the acrosome reaction. n = 5 males for Fer1l5 +/− mice and n = 3 males for Fer1l5 −/− and Fer1l5 −/− PA-Tg mice. *P < 0.05, **P < 0.01, and ***P < 0.001 (Student's t test).
Microbial Diseases, Osaka University (#Biken-AP-H30-01) and were conducted in accordance with the guidelines established by the Research Institute for Microbial Diseases, Osaka University.

Generation of Fer1l4, Fer1l5, and Fer1l6 mutant mice
The targeting vectors were obtained from the International Mouse Phenotyping Consortium (Fer1l6) (www.mousephenotype.org/) (43) or generated using pNT1.1 (Fer1l4 and Fer1l5) (www.ncbi. nlm.nih.gov/nuccore/378747675) (44). The primers used to amplify long arms and short arms for pNT1.1 are listed in table S1. The vectors were electroporated into embryonic stem (ES) cells (EGR-G101 for generating Fer1l4 mutant mice and EGR-G01 for generating Fer1l5 and Fer1l6 mutant mice) (45) after linearization, and potentially targeted ES cell clones were selected with G418 (150 μg/ml; Thermo Fisher Scientific, Waltham, MA, USA). The homologously recombined ES cells were then injected into Institute of Cancer Research (ICR) eight-cell embryos. The embryos were cultivated in a potassium simplex optimized medium (KSOM) (46) until the next day and transferred into pseudo-pregnant females. The obtained chimera mice were mated with B6D2F1 females for germline transmission. For Fer1l6, floxed mice were further mated with B6D2 CAG-Cre Tg mice (47).

Generation of Fer1l5-FLAG and Fer1l5-PA Tg mice
Mouse Fer1l5 cDNA-FLAG tag or Fer1l5 cDNA-PA tag with a rabbit polyadenylate [poly(A)] signal under the mouse Clgn promoter ( fig. S2C) was prepared. The linearized DNA was injected into one of the pronuclei of fertilized eggs that were obtained by the mating of superovulated B6D2F1 females and B6D2F1 males, and the zygotes were cultivated in KSOM to two-cell stage and then transferred into pseudo-pregnant females.

Phylogenetic analysis
The phylogenetic trees of candidate genes were created using the neighbor-joining method with the GENETYX software (GENETYX, Tokyo, Japan). Amino acid sequences of each protein were obtained from the National Center for Biotechnology Information Entrez Protein database. The accession numbers of each amino acid sequence used were as follows: Drosophila

Reverse transcription polymerase chain reaction
Mouse RNAs were prepared from multiple adult tissues of C57BL/ 6N or testes from 1-to 5-week-old males using TRIzol (Thermo Fisher Scientific, Waltham, MA, USA). The obtained RNA was immediately reverse-transcribed to cDNA with SuperScript IV First-Strand Synthesis System (Thermo Fisher Scientific) using oligo(dT) as primers. PCR primers for each gene are shown in table S1.

Mating tests
Sexually matured male mice were individually caged with two 8week-old B6D2F1 female mice for 2 months, and the number of pups was counted. For Fer1l5 mutant mice, plugs were checked every morning.

Histological analysis of testes and cauda epididymides
Male mice (12 to 18 weeks old) were euthanized, and testes and cauda epididymides were dissected. Testes and cauda epididymides were fixed overnight at 4°C in Bouin's fluid (Polysciences Inc., Warrington, PA, USA), embedded in paraffin, and sectioned at a thickness of 5 μm. Paraffin sections were then rehydrated, treated with 1% periodic acid for 20 min at room temperature, and incubated with Schiff's reagent (FUJIFILM Wako Pure Chemical, Osaka, Japan) for 20 min at room temperature. The sections were stained with Mayer's hematoxylin solution (FUJIFILM Wako Pure Chemical) and observed using a BX-53 microscope (Olympus, Tokyo, Japan).

Morphological observation of spermatozoa
Spermatozoa were collected from the cauda epididymis and dispersed in phosphate-buffered saline (PBS) (Thermo Fisher Scientific). The spermatozoa were observed using a BX-53 microscope (Olympus).

Sperm motility analysis
Spermatozoa obtained from the cauda epididymis were incubated in a drop of Toyoda Yokoyama Hosi (TYH) medium (1) at 37°C under 5% CO 2 . Ten minutes, 2 hours, or 4 hours after incubation, spermatozoa were collected from the top of the drop. Sperm motility was analyzed using the CEROS sperm analysis system (Hamilton Thorne Biosciences, Beverly, MA, USA). Analysis settings were used as described previously (29,49). For analyzing maximum bending angles, spermatozoa were observed with an Olympus BX-53 microscope equipped with a high-speed camera (HAS-L1, Ditect, Tokyo, Japan).

In vitro fertilization
Spermatozoa collected from the cauda epididymis were incubated in a drop of TYH medium for 2 hours at 37°C under 5% CO 2 . For collecting eggs, Center for Animal Resources and Development (CARD) HyperOva (0.1 ml; Kyudo, Saga, Japan) was injected into the abdominal cavity of B6D2F1 female mice, followed by hCG (five units, ASKA Pharmaceutical, Tokyo, Japan) with 48-hour interval. Eggs were then collected from the superovulated females, treated with hyaluronidase (330 μg/ml; Sigma-Aldrich, St. Louis, MO, USA) for 10 min to remove the cumulus cells (cumulus-free eggs) or with collagenase (1 mg/ml; Sigma-Aldrich) for 10 min to remove the ZP (ZP-free eggs). The spermatozoa were added to the TYH drops that contain intact, cumulus-free, or ZP-free eggs at a final density of 2 × 10 5 or 2 × 10 6 spermatozoa/ml and incubated at 37°C under 5% CO 2 . The formation of pronuclei was observed 6 hours after insemination for ZP-free eggs. Two-cell embryos were counted the next day for intact or cumulus-free eggs.

Observation of sperm migration and analysis of the acrosome reaction rates in vivo
B6D2F1 female mice were superovulated as mentioned above and mated with male mice that contained the RBGS transgene. Two hours after observation of vaginal plugs or 20 hours after hCG injection, female reproductive tracts were dissected and spermatozoa inside the oviduct or ampulla were observed. To confirm that the spermatozoa were alive, motile spermatozoa were observed using a Nikon Eclipse Ti microscope connected to a Nikon C2 confocal module (Nikon, Tokyo, Japan). Spermatozoa with EGFP signals in the acrosome were counted as acrosome intact.

Collection of testis lysates for immunoprecipitation
Testes were lysed with a solution containing 1% Triton X-100, 50 mM NaCl, 20 mM tris-HCl (pH 7.4), and protease inhibitor mixture (Nacalai Tesque); incubated for 1 hour at 4°C; and centrifuged at 15,300g for 15 min at 4°C to collect the supernatants.
Generation of recombinant proteins for immunoprecipitation cDNAs encoding Fer1l5, Stx2, and Adam1b were amplified from mouse testis (C57BL/6N) and cloned into FLAG-tagged (C terminus), PA-tagged (C terminus), and untagged pCAG vectors, respectively, that contain the cytomegalovirus immediate enhancer/βactin (CAG) promoter and a rabbit globin poly(A) signal (51). The expression vectors were transfected transiently into HEK293T cells, and the cells were cultured for 24 hours before collection. The cells were then lysed with a solution containing 1% Triton X-100, 50 mM NaCl, 20 mM tris-HCl (pH 7.4), and protease inhibitor mixture (Nacalai Tesque); incubated for 30 min at 4°C; and centrifuged at 15,300g for 15 min at 4°C to collect supernatants.

Immunoprecipitation
The protein lysates were incubated for 60 min at 4°C with anti-FLAG antibody (M2, Sigma-Aldrich) or anti-PA antibody (FUJI-FILM Wako Pure Chemical)-conjugated Dynabeads (Thermo Fisher Scientific). The complexes were washed three times with a solution containing 40 mM tris-HCl, 150 mM NaCl, 0.1% Triton X-100, and 10% glycerol and eluted with a sample buffer. Immunoblotting with collected samples were performed as mentioned above.

Mass spectrometry analysis
Spermatozoa collected from the cauda epididymis were lysed with a solution containing 1% Triton X-100, 50 mM NaCl, 20 mM tris-HCl (pH 7.4), and protease inhibitor mixture (Nacalai Tesque); incubated for 2 hours at 4°C; and centrifuged at 15,300g for 15 min at 4°C to collect supernatants. The samples were subjected to MS analyses directly or after SDS-PAGE under reduced condition. To visualize proteins in gels, silver staining was performed following the manufacturer's protocol (#05773-11, Nacalai Tesque). Mass spectrometry analyses were performed as described previously with some modifications (52). Briefly, peptides were subjected to nanocapillary reversed phase liquid chromatography with tandem MS (LC-MS/MS) analysis using a C18 column on a NanoLC System (Bruker Daltoniks, Bremen, Germany) connected to a timsTOF Pro mass spectrometer (Bruker Daltoniks) and a modified nanoelectrospray ion source (CaptiveSpray, Bruker Daltoniks). The mobile phase consisted of water containing 0.1% formic acid (solvent A) and acetonitrile containing 0.1% formic acid (solvent B) at linear gradient elution. The ion spray voltage was set in the positive ion mode. Ions were collected in the trapped ion mobility spectrometry (TIMS) device. During the collection of MS/MS data, the TIMS cycle included parallel accumulation serial fragmentation-MS/MS scans (53), and nitrogen gas was used as collision gas. Proteins were identified using MASCOT (version 2.7.0, Matrix Science, London, UK) against the UniProt database. Scaffold (version Scaffold_5.0.1, Proteome Software Inc., Portland, OR, USA) was used to validate peptide and protein identifications. Peptide identification probability and protein identification probability were calculated using the PeptideProphet algorithm (54) and ProteinProphet algorithm (55), respectively.

Supplementary Materials
This PDF file includes: Figs. S1 to S6 Tables S1 to S3 View/request a protocol for this paper from Bio-protocol.