Seminal vesicle secretory protein 7, PATE4, is not required for sperm function but for copulatory plug formation to ensure fecundity

Abstract Seminal vesicle secretions (SVSs), together with spermatozoa, are ejaculated into the female reproductive tract. SVS7, also known as PATE4, is one of the major SVS proteins found in the seminal vesicle, copulatory plug, and uterine fluid after copulation. Here, we generated Pate4 knockout (−/−) mice and examined the detailed function of PATE4 on male fecundity. The morphology and weight of Pate4−/− seminal vesicles were comparable to the control. Although Pate4−/− cauda epididymal spermatozoa have no overt defects during in vitro fertilization, Pate4−/− males were subfertile. We found that the copulatory plugs were smaller in the vagina of females mated with Pate4−/− males, leading to semen leakage and a decreased sperm count in the uterus. When the females mated with Pate4−/− males were immediately re-caged with Pate4+/+ males, the females had subsequent productive matings. When the cauda epididymal spermatozoa were injected into the uterus and plugged artificially [artificial insemination (AI)], Pate4−/− spermatozoa could efficiently fertilize eggs as compared to wild-type spermatozoa. We finally examined the effect of SVSs on AI, and observed no difference in fertilization rates between Pate4+/+ and Pate4−/− SVSs. In conclusion, PATE4 is a novel factor in forming the copulatory plug that inhibits sequential matings and maintains spermatozoa in the uterus to ensure male fecundity.


Introduction
In mammals, spermatozoa generated in the testis cannot fertilize eggs until they acquire fertilizing abilities when they pass through the epididymis. Whereas spermatozoa collected from the cauda epididymis can efficiently fertilize eggs in vitro [1], male mice ejaculate spermatozoa into the female reproductive tract with accessory gland secretions (e.g. seminal vesicle, prostate, coagulating gland) in vivo. The majority of accessory gland secretions are produced from seminal vesicles in many species (e.g. human, bull) [2,3]. As mice without seminal vesicles show subfertility [4], seminal vesicle secretions (SVSs) are thought to play a beneficial role in fertilization in vivo.
As one of the physiological functions of the SVSs, copulatory plug formation is well known in several primates (such as chimpanzees) and rodents [5,6]. For copulatory plug formation, transglutaminase 4 (TGM4), a protein from the coagulating gland and prostate, catalyzes the formation of ε-(γ -glutamyl)lysine cross-bridges between some SVS proteins [6][7][8][9][10][11][12][13][14]. In fact, Tgm4-disrupted males show a plug formation defect [15]. Thus, the interaction between SVSs and proteins from the coagulating gland and prostate are required for copulatory plug formation. Mangels et al. showed that copulatory plugs prevent subsequent mating with other males [16], and their previous work suggested that the copulatory plug is involved with sperm release, sperm transition, and coital stimulation, but the critical role remained to be determined [17]. Further, other results showed that the attachment of SVSs to cauda epididymal spermatozoa protects spermatozoa from the immunological defense mechanism of the uterus and also prevents premature capacitation, leading to increased sperm viability and fertilizing ability in vivo [18][19][20]. Also, SVSs are reported to influence gene expression in the female reproductive tract and subsequent fetal development [21][22][23]. These results suggest that SVSs have several physiological functions, but the details remain to be elucidated.
The remaining SVS, SVS7, is encoded by the Pate4 (also known as Pate-B and Caltrin) gene on chromosome 9. Pate4 mRNA was reported to be expressed not only in the seminal vesicles but also in the testes, epididymides, and prostates by RT-PCR [28,29]. However, the later study showed that Pate4 mRNA is expressed not in testis and epididymis but in the prostate and seminal vesicle by northern blotting, and that the signal intensity of Pate4 mRNA in seminal vesicles is four times higher than in the prostates by qPCR [30]. Thus, the expression of mouse Pate4 in tissues remains unclear. PATE4 was localized to the head and flagella of mouse epididymal spermatozoa incubated in SVS7 solution [31]. It was found that Ca 2+dependent events relating to sperm fertilizing ability, such as in sperm capacitation, sperm motility, the acrosome reaction, and sperm-egg interaction, were regulated by PATE4 in vitro [26,[31][32][33]. Heckt et al. performed phenotypic analyses of Pate4−/− mice [30], but the detailed role of PATE4 in copulatory plug formation and spermfertilizing ability remains to be revealed. Here, we examined the physiological function of PATE4 in male fecundity using knockout mice.

Animals
All mice used in this study were purchased from Japan SLC or CLEA Japan. Mice were acclimated to a 12-h light/12-h dark cycle. All animal experiments were approved by the Animal Care and Use Committee of the Research Institute for Microbial Diseases, Osaka University, Japan (#Biken-AP-H25-02 and #Biken-AP-H30-01).

Sample collection
The testis, epididymis (caput, corpus, and cauda regions), coagulating gland (also known as anterior prostate), prostate (mixture of dorsal, lateral, and ventral regions), seminal vesicle, ovary, and uterus were collected from adult C57BL/6NCr and B6D2F1 mice. SVSs were obtained by squeezing the seminal vesicle immediately after euthanasia. The copulatory plug was collected from ICR males treated with vasoligation within 2 h of mating. These samples were fractured in TRIzol (Ambion) or in Tris-Buffered Saline with Triton-X {50 mM NaCl (Nacalai), 10 mM Tris-HCl (Nacalai), 1% (v/v) Triton-X 114 (Sigma), pH 7.5} containing 1% (v/v) protease inhibitor cocktail (Nacalai), and then used to detect PATE4 in various organs at mRNA and protein levels. Cauda epididymal spermatozoa were dispersed in TYH drops [34] for 5 min. The intrauterine fluid before and 1 h after mating with B6D2F1 males was collected by PBS perfusion (200 μL/each uterine horn). These samples were used for western blotting.
Surgical removal of the accessory glands B6D2F1 males (8-week-old) were used for the removal of accessory glands, as described previously [38]. Four weeks after surgeries, these males were used for the following experiments.

Count, morphology, motility, and in vitro fertilization of cauda epididymal spermatozoa
Protein extracts from the entire cauda epididymis were subjected to western blotting, and then the signal intensity of SLC2A3 was measured by ImageQuant TL software to estimate the sperm count within the cauda epididymis. Cauda epididymal spermatozoa were dispersed in PBS (for sperm morphology) or TYH drops {for sperm motility and in vitro fertilization (IVF)}. After an incubation period of 120 min, the sperm motility pattern was examined using the CEROS sperm analysis system (software version 12.3; Hamilton Thorne Biosciences) [39]. IVF was performed as described previously [40].

Collection of copulatory plugs and uterine sperm counts
Pregnant mare serum gonadotropin (PMSG) (5 units, ASKA Pharmaceutical) was injected into the abdominal cavity of B6D2F1 females, followed by human chorionic gonadotropin (hCG) (5 units, ASKA Pharmaceutical) 48 h after PMSG. Twelve hours after hCG, the hormone-treated females were caged with the Pate4 mutant, sham, CG (−), and SV (−) males (9-to 37-week-old) under observation. Immediately after mating, the existence of the copulatory plugs was checked, and then the leaked semen was collected with pipettes. Within 2 h of mating, the copulatory plug and the uterus were obtained after euthanasia. The spermatozoa in the uterus were collected by PBS perfusion (200 μL/each uterine horn), and then the total sperm count was examined with a hemocytometer.

Modified AI method
We modified the conventional AI method [41,42], by using a gel loading tip (cat#010-Q, BMBio) for sperm injection (Supplemental Figure S2A) and petroleum jelly for plugging (Nacalai) (Supplemental Figure S2B). AI was performed under Isoflurane anesthesia (Mylan Inc.). After inserting 10-15 mm of the tip (Supplemental Figure  S2B), about 50 μl of a sperm suspension was injected into the uterine body. Then, petroleum jelly at the end of the tip was injected to form the artificial plug around the uterine cervix using a plunger (Terumo) (Supplemental Figure S2C).

Sperm injection by AI
The SVSs of B6D2F1 and Pate4−/− males were suspended in 200 μL TYH or PBS, and subsequently centrifuged at 2000-5000 g for 5 min, as described previously [20]. Ten microliter of the supernatant was diluted in 90 μL TYH or PBS. Cauda epididymal spermatozoa of B6D2F1 and Pate4−/− males were dispersed in 100 μL TYH or PBS drops with and without SVSs for 10-40 min. The uterus of hormone-treated females was artificially injected with 0.08 to 8.8 × 10 6 cauda epididymal spermatozoa.

Sperm viability
Thirty minutes after AI, the uterine fluid was collected by TYH or PBS perfusion (200 μL/each uterine horn). The motility of spermatozoa in the middle layer of the drop was recorded using an Olympus BX-53 microscope equipped with a high-speed camera (HAS-L1, Ditect). Spermatozoa that showed flagellar beating were counted.

In vivo fertilization rates
Seven to fourteen hours after AI, the cumulus of collected eggs was removed with Hyaluronidase (300 μg/mL final conc., Sigma). The number of eggs with two pronuclei was counted.

Pregnancy rates and fetal development
After 19 days, offsprings were obtained by natural birth or Caesarean section.

Statistical analyses
All values are shown as the mean ± SD of at least three independent experiments. Statistical analyses were performed using the Student's t-test (for Figures 2E and F

Detection of PATE4
When we performed RT-PCR with cDNAs of male and female reproductive organs, Pate4 was strongly detected in the seminal vesicle ( Figure 1A). Further, we generated an anti-PATE4 antibody and performed western blotting analysis. While we could not detect any signals in the testis, cauda epididymis, coagulating gland, and prostate, a firm doublet signal was detected in seminal vesicles ( Figure 1B). To examine whether PATE4 is ejaculated into semen, we collected uterine fluid 1 h after copulation. Though we could not detect PATE4 in the uterus fluid before coitus, PATE4 was detected after coitus in uterine fluid which also contained semen and the copulatory plug ( Figure 1C). These results indicate that PATE4 is abundantly secreted from seminal vesicles and then ejaculated into the female reproductive tract.

Male fecundity of Pate4−/− mice
To reveal the function of PATE4 in vivo, we generated mice lacking Pate4 with a targeting vector from IMPC ( Figure 1D). Mice were genotyped by genomic PCR (Figure 1E Figure S1A) and weight (Supplemental Figure S1B) of seminal vesicles, PATE4 protein was not detectable from the seminal vesicles of Pate4−/− mice ( Figure  1F). SVS1 to SVS6 and TGM4 were detected in wild-type (WT) and Pate4−/− males at comparable levels (Figure 2A-C). As we expected from the western blot analysis, Pate4−/− cauda epididymal spermatozoa did not show any defects in sperm parameters in vitro such as morphology ( Figure 2D), motility parameters ( Figure 2E), and in vitro fertilizing ability ( Figure 2F).
To understand the effect of the decreased copulatory plug size, we observed the vaginal opening immediately after copulation. We found that the semen leaked out of females mated with Pate4−/− males ( Figure 4E and F). With the finding that fewer spermatozoa were found in females mated with Pate4−/− males ( Figure 3E), these results indicate that the copulatory plug has a role in keeping ejaculated spermatozoa in the uterus.

Discussion
The expression of mouse Pate4 in tissues remained to be clarified, but here we demonstrated PATE4 abundantly presents in the seminal vesicles at mRNA and protein levels ( Figure 1A and B). It is known that Pate4 is a member of the Pate family composed of 13 genes clustered on mouse chromosome 9 [28]. As some parts of the nucleotide sequence of Pate4 are similar to the other Pate family genes, nonspecific binding of primers might cause the discrepancy of RT-PCR results between this study and previous papers. As we had generated Pate4−/− mice before the report of Heckt et al. [30], we used our KO mice for this study. We used the same Pate4 targeting vector construct from KOMP as the previous report [30], but the genetic background used to maintain the mouse line was different between this study (B6D2 background) and the previous paper (B6J background). The previous paper indicated that fecundity of Pate4−/− males were comparable to the control males [30], but we revealed that Pate4−/− males were subfertile ( Figure 3A and C). When we used seven males of Pate4−/− mice for the mat-ing test, the male fecundity varied by individuals (see Supplemental  Table S2, no. of litters/female/caging of month, 0-0.96). Thus, the individual differences of Pate4−/− males may cause the difference in male fecundity between our study and the previous report. And, our result suggests that the subfertile phenotype may be easily masked by genetic background.
PATE4 was absent from the seminal vesicles of Pate4−/− mice ( Figure 1F), but the copulatory plug related proteins "SVS1 to SVS3" were detected in Pate4+/+ and Pate4−/− seminal vesicles at comparable level (Figure 2A). The sequence homology among PATE4, SVS1, SVS2, and SVS3 is low at the amino acid level (Supplemental Figure S3; PATE4 vs SVS1: 23.9%, PATE4 vs SVS2: 23.5%, PATE4 vs SVS3: 17.7%). These results suggest that the remaining factors for plug formation in Pate4−/− males cannot compensate for the lack of PATE4. To form the copulatory plug, it is known that TGM (e.g. TGM4) catalyzes the formation of ε-(γ -glutamyl)lysine which cross-bridges among SVS1 to SVS3 [6][7][8][9][10][11][12][13][14]. Lin et al. showed that the peptide sequence "QXK(S/T)" in SVS3 acts as the transglutaminase cross-linking sites by the reaction of guinea pig liver transglutaminase and recombinant polypeptides from SVS3 (Supplemental Figure  S3C) [10]. We also found this peptide sequence in SVS2, but the sequence is not conserved in PATE4 (Supplemental Figure S3B). SVS1 also does not contain the sequence "QXK(S/T)", but Tseng et al. showed that two glutamine residues in SVS1 were the major site for TGM4 crosslinking by mass spectrometry (Supplemental Figure   S3A) [8]. As PATE4 also has four glutamine and ten lysine residues (Supplemental Figure S3), our results suggest that PATE4 may have an unidentified target sequence for TGM4 or a function to promote plug formation independent from TGM4.
It is well known that proteins secreted from coagulating glands are required for plug formation [15,17]. The plug weight from CG (−) males was reduced to almost half of the sham-operated males ( Figure 4B and D), but CG (−) males were fertile ( Figure 3B), as previously reported [4]. Tgm4 mRNA was detected in the seminal vesicle, coagulating gland, and prostate ( Figure 2B), corresponding with the Unigene database and a previous study [44]. Further, we revealed that the TGM4 protein was detected in the prostate and coagulating gland ( Figure 2C). These results indicate that TGM4 localized in both the coagulating gland and the prostate contribute to copulatory plug formation.
Detailed functions of SVSs on sperm-fertilizing ability and fetal development remain to be known. Kawano et al. showed that SVSs improved sperm viability and increased the fertilization rate in vivo [20]. However, with sperm counts comparable to that observed in normal matings (1 × 10 6 or more), we could not find a significant difference in the rates of sperm motility between spermatozoa with and without SVSs ( Figure 5B). Also, 2 h after mating with SV (−) males, the ejaculated spermatozoa in the uterus survived (Supplemental Movie S5). Further, SVSs did not increase the pregnancy rates nor litter size ( Figure 5C and D). Given the data presented, SVSs do not appear to play an essential role in sperm-fertilizing ability, implantation, or fetal development in mice.
When we lowered the sperm count in AI, WT SVSs improved fertilization rates, implicating a positive function of WT SVSs on sperm function in vivo. Even in such a case, PATE4 does not contribute to increased sperm-fertilizing ability ( Figure 5E). Our results suggest that other components secreted from seminal vesicles support sperm-fertilizing ability in vivo but only when sperm count is low. Many researchers have analyzed the sperm-fertilizing ability using AI, but the fertilization rates varied among papers [20,41,42,45], indicating that an AI method has yet to be standardized. We used females after 12 h of hCG injection for AI, and then obtained the high fertilization rates, corresponding with the previous papers [42]. Thus, for AI with hormone-treated females, the timing after 12 h of hCG injection is considered as the best. Previous studies showed that sperm parameters, such as the morphology, sperm motility, and fertility rate, varied between genetic backgrounds of mouse strains [46,47]. Thus, the difference between the previous report and this study may be due to these problems.
Here, we revealed that PATE4 is an essential factor for copulatory plug formation. Through phenotypic analysis of Pate4−/− males, we found that the copulatory plug has a physiological function to keep the spermatozoa in the uterus, leading to an increase in fertilization rates. Further, females mated with Pate4−/− males became pregnant from subsequent matings with Pate4+/+ males due to the plug formation defect. Thus, the copulatory plug has dual functions not only to prevent subsequent matings but also to maintain proper sperm count for fertilization in the female reproductive tract, as a winner-take-all strategy to advance male reproduction.

Supplementary data
Supplementary data are available at BIOLRE online. Figure S1. Morphology and weight of Pate4−/− seminal vesicles, and the production of the male accessory gland-removed mice. (A) Morphology of seminal vesicles. Scale bars show 5 mm. (B) Seminal vesicle (SV) weight/body weight (BW). There was no difference between Pate4+/+ and Pate4−/− SVs (P = 0.78). N.S.: not significant. (C) Accessory glands were surgically removed, as described previously [38]. CG: coagulating gland. Supplemental Figure S2. Modification of AI and pups produced by AI. (A) Pipette tip for AI. Scale bar shows 5 mm. (B and C) Procedure of AI. Ten to fifteen millimeter of a tip containing the sperm suspension (50 μL) and petroleum jelly for plugging were inserted into the uterus of female mice (B). By inserting 4-5 mm of a plunger into the tip, the sperm suspension was injected into the uterus, and the petroleum jelly formed an artificial plug around the uterine cervix (C). (D) Pups produced from AI. There were no difference in the pregnancy rates and litter size between AI with and without SVSs (also see Figure 5C