Detection of extracellular vesicles in the mouse vaginal fluid: Their delivery of sperm proteins that stimulate capacitation and modulate fertility

Abstract

Extracellular vesicles (EVs) were isolated by ultracentrifugation of vaginal luminal fluid (VLF) from superovulated mice and identified for the first time using transmission electron microscopy. Characterized by size and biochemical markers (CD9 and HSC70), EVs were shown to be both microvesicular and exosomal and were dubbed as “Vaginosomes” (VGS). Vaginal cross‐sections were analyzed to visualize EVs in situ: EVs were present in the lumen and also embedded between squamous epithelial and keratinized cells, consistent with their endogenous origin. Western blots detected Plasma membrane Ca²⁺‐ATPase 1 (PMCA1) and tyrosine‐phosphorylated proteins in the VGS cargo and also in uterosomes. Flow cytometry revealed that following coincubation of caudal sperm and VLF for 30 min, the frequencies of cells with the highest Sperm adhesion molecule 1 (SPAM1), PMCA1/4, and PMCA1 levels increased 16.4‐, 8.2‐, and 27‐fold, respectively; compared with control coincubated in phosphate buffered saline (PBS). Under identical conditions, sperm tyrosine‐phosphorylated proteins were elevated ~3.3‐fold, after VLF coincubation. Progesterone‐induced acrosome reaction (AR) rates were significantly (p < 0.001) elevated in sperm coincubated with VGS for 10–30 min, compared with PBS. Sperm artificially deposited in the vaginas of superovulated females for these periods also showed significant (p < 0.01) increases in AR rates, compared with PBS. Thus in vitro and in vivo, sperm acquire from the vaginal environment factors that induce capacitation, explaining recent findings for their acrosomal status in the isthmus. Overall, VGS appear to deliver higher levels of proteins involved in preventing premature capacitation and AR than those promoting them. Our findings which have implications for humans open the possibility of new approaches to infertility treatment with exosome therapeutics.

Extracellular vesicles (EVs) were identified in the murine vaginal fluid and shown to carry, among fertility modulating proteins, tyrosine‐phosphorylated proteins that induce capacitation and the acrosome reaction. The EVs are able to deliver these proteins to sperm after in vitro and in vivo exposure. Our findings provide an explanation for the induction of the acrosome reaction of mouse sperm in the isthmus.

1. INTRODUCTION

Extracellular vesicles (EVs) are spherical membranous organelles released by epithelial cells into body fluids where they play a role in cell‐to‐cell communication by conveying molecules from one cell or tissue to another (Thery, Zitvogel, & Amigorena, 2002; Valadi et al., 2007). These nano‐sized organelles carry a variety of regulatory molecules such as proteins, microRNAs, RNAs, and lipids (Thery et al., 2002; Valadi et al., 2007), and have been identified in both male and female reproductive biofluids (Martin‐DeLeon, 2016). In the male, EVs in the epididymal luminal fluid, dubbed epididymosomes, and those in the seminal plasma, prostasomes, have been extensively studied (Saez & Sullivan, 2016; Sullivan, 2016). It has been well‐established that epididymosomes are responsible for the transfer of a variety of proteins to sperm to facilitate epididymal maturation and the remodeling of the sperm membrane (Caballero, Frenette, & Sullivan, 2010; Sullivan & Saez, 2013). On the other hand, prostasomes have been known to transfer proteins that regulate sperm function, such as motility, and those that have a stimulatory effect on capacitation, as well as those with an inhibitory effect on this process (Arienti, Carlini, De Cosmo, Di Profio, & Palmerini, 1998a; Siciliano, Marciano, & Carpino, 2008).

Compared with the well‐characterized EVs in male secretions, EVs in the female reproductive fluids have received very little attention as the studies have only just been emerging. In the female tract EVs have been identified in the uterine luminal fluid (ULF) of mice (Griffiths, Galileo, Reese, & Martin‐DeLeon, 2008; Griffiths, Miller, Galileo, & Martin‐DeLeon, 2008), sheep (Burns et al., 2014), and humans (Ng et al., 2013) where they have been dubbed uterosomes (UTS); and in the oviductal luminal fluid (OLF) in mice (Al‐Dossary, Strehler, & Martin‐DeLeon, 2013), bovine (Alminana et al., 2017; Lopera‐Vasquez et al., 2016), and humans (Bathala et al., 2018), where they are referred to as oviductosomes (OVS). During the murine estrus, UTS and OVS were shown to carry both glycosyl phosphatidylinositol‐ (GPI‐) linked and transmembrane proteins that can be transferred to sperm in vitro (Al‐Dossary et al., 2013; Griffiths, Miller et al., 2008). Notable among the GPI‐linked proteins that UTS carry and deliver to sperm, is the highly conserved Sperm adhesion molecule 1 (SPAM1) which is known to play multiple roles in sperm function (Martin‐DeLeon, 2011). Importantly, it was shown that following uterosomal transfer of SPAM1 to murine sperm during their coincubation with ULF, there was increased hyaluronic acid binding capacity and cumulus penetration efficiency of sperm (Griffiths, Miller et al., 2008).

An essential transmembrane protein that is present in the uterine and oviductal epithelia and in the cargoes of UTS and OVS is Plasma membrane Ca²⁺‐ATPase 4 (PMCA4; Al‐Dossary et al., 2013). It is the major calcium efflux pump in murine sperm (Wennemuth, Babcock, & Hille, 2003), and its absence results in male infertility (Okunade et al., 2004). Interestingly, in the mouse PMCA4 is most highly expressed in estrus where it is not only present in the ULF and the OLF via UTS and OVS, but is also found in the vaginal luminal epithelium and the vaginal luminal fluid (VLF; Al‐Dossary et al., 2013). However, in the vaginal epithelium and VLF, it exists in a significantly lower abundance than in the uterine and oviductal counterparts (Al‐Dossary et al., 2013). This is in contrast to SPAM1 which was not studied in the VLF, but was detected in the vaginal luminal epithelium (which releases proteins into the luminal fluid) in significantly greater abundance than in oviductal tissues (Zhang & Martin‐DeLeon, 2003). Thus, the abundance of a protein that is released in the luminal fluids appears to be dependent on the particular protein rather than the region of the reproductive tract from which it is released during estrus. Following coincubation of murine sperm with the combined luminal fluids (OLF, ULF, and VLF), PMCA4 could be delivered to sperm in vitro (Al‐Dossary et al., 2013), suggesting that this occurs in vivo.

OVS have been shown to transfer PMCA4 to murine sperm, using super‐resolution structured illumination microscopy, by a fusogenic mechanism (Al‐Dossary, Bathala, Caplan, & Martin‐DeLeon, 2015). This mechanism which was also reported for protein transfer to sperm from bovine epididymosomes (Schwarz et al., 2013) appears to be conserved, because uterosome‐like vesicles secreted by human endometrial epithelial cells in vitro were shown to fuse with sperm following a 15‐min coincubation period (Franchi, Cubilla, Guidobaldi, Bravo, & Giojalas, 2016). This fusion of the uterosome‐like vesicles stimulated capacitation by significantly increasing sperm protein tyrosine‐phosphorylation, a hallmark of capacitation (Baker, Hetherington, & Aitken, 2006; Kwon, Rahman, & Pang, 2014; Visconti et al., 1995).

Because membrane fusion that mediates the transfer of cargo constituents from EVs to sperm occurs in as short a time period as 15 min (Franchi et al., 2016), we were prompted to investigate the presence of EVs in the murine VLF during estrus and their ability to transfer proteins to sperm. Herein we show the presence of exosomes and microvesicles in murine VLF, and their ability to transfer fertility‐modulating proteins in their cargo to sperm, following their coincubation for 30 min. Importantly, we show the physiological consequence of protein transfer from vaginal EVs (vaginosomes [VGS]) to sperm in vitro and in vivo, with respect to a significant increase in the rate of acrosome reaction (AR) after coincubation for 10 min.

2. MATERIAL AND METHODS

2.1. Mice and reagents

Sexually mature 10–12 week‐old males and 4–12 week‐old female mice (FVB/N and C57BL/6J strains; Harlan, Indianapolis, IN) were used for this study. In addition to these WT mice, Pmca4 ^−/− males and females on the FVB/N background were used to obtain testis and vaginal tissues, luminal fluid, and EVs for western blotting analysis. These mutant mice, generated in the laboratory of Dr. Gary Shull (University of Cincinnati), were a generous gift. They were bred and genotyped as described previously (Okunade et al., 2004). The studies were approved by the Institutional Animal Care and Use Committee at the University of Delaware and were in agreement with the Guide for the Care and Use of Laboratory Animals published by the National Research Council of the National Academies, 8th ed., Washington, D.C. (publication 85–23, revised 2011). All enzymes and chemicals were purchased from Fisher Scientific Co. (Malvern, PA), Sigma (St Louis, MO) or Invitrogen (Carlsbad, CA), unless otherwise specified.

2.2. Antibodies

Rabbit monoclonal anti‐PMCA1 antibody (ab190355) and the isotype control (ab172730) were purchased from Abcam (Cambridge, MA) and used for western blots, and flow cytometric studies. Mouse monoclonal anti‐HSC70 antibody (sc‐7298), mouse monoclonal anti‐panPMCA1/4 antibody (sc‐20028), goat polyclonal pan‐PMCA4 antibody (Y20, sc‐22080), rat monoclonal anti‐CD9 antibody (sc‐18869), and anti‐phosphotyrosine antibody (PY99) which is a mouse monoclonal (sc‐7020) were purchased from Santa Cruz Biotechnology (Dallas, TX). Also, mouse monoclonal anti‐phosphotyrosine antibody (clone 4G10) was purchased from Millipore (Temecula, CA). A rabbit anti‐mouse SPAM1 antiserum, a polyclonal antipeptide ([C]NEKGNASRRKESSD in the C‐terminus [#381‐395] custom‐made by Zymed (South San Francisco, CA) and previously reported to be specific for SPAM1 via peptide inhibition (Deng, He, & Martin‐DeLeon, 2000; Zhang & Martin‐DeLeon, 2003) was used. Secondary antibodies were purchased from Santa Cruz Biotech, Inc., Life Technologies or Molecular Probes Inc. (Eugene, OR). Fluoro‐Gel II with diamidinophenylindole (DAPI) (17985‐50), used for immunofluorescence, was obtained from Electron Microscope Sciences (Hatfield, PA).

2.3. Staging of females and the hormonal induction of estrus (superovulation)

The stages of the estrus cycle of 8–12 week‐old FVB/N virgins were classified based on the proportion of different cell types identified in the vaginal secretion (Byers, Wiles, Dunn, & Taft, 2012). To hormonally induce estrus, 4–6 week‐old FVB/N and C57BL/6J females were sequentially injected intraperitoneally with pregnant mare serum gonadotropin (7.5 i.u.) and human chorionic gonadotropin (7.5 i.u.), 44–48 hr apart. After 13.5–14 hr from the last injection, females were killed and their reproductive tissues collected for processing and analysis, as described below.

2.4. Collection of VLF and isolation of EVs

Immediately after sacrificing females hormonally‐induced into estrus, vaginas were removed. The luminal fluid was collected in phosphate buffered saline (PBS) with protease inhibitors after mincing the vaginas as previously described for the oviduct (Al‐Dossary et al., 2013). After mincing the tissues in PBS, the tissues were allowed to gravity settle and the fluid was clarified by centrifugation at 16,000g for 20 min to exclude cells, cells debris, and tissue fragments. The supernatant is referred to as the clarified VLF and was then frozen immediately at −80°C for further processing. Fresh or thawed frozen clarified VLF was subjected to ultracentrifugation at 120,000g at 4°C for 2 hr using Beckman Optima 2–70 k ultracentrifuge and a Ti60 rotor, as described (Griffiths, Galileo et al., 2008). Resulting pellets (EVs) were resuspended in homogenization buffer (62.5 mM Tris‐HCl, 10% glycerol, 1% SDS, pH 6.8) with protease inhibitor (1 mM phenylmethylsulfonyl fluoride (PMSF)) for western blot analysis, or in PBS for transmission electron microscopy (TEM) or further processing.

2.5. Negative staining for TEM

Aliquots of the VLF pellet suspension from females in estrus were negatively stained, using a methodology similar to that previously described (Al‐Dossary et al., 2013; Griffiths, Galileo et al., 2008). Briefly, nickel TEM grids (Electron Microscopy Sciences), 400 mesh with formvar/carbon film were floated on a drop of the pellet suspension. Several drops of water were then used to wash the grids which were then stained with 1% uranyl acetate, a phospholipid stain, before being subjected to TEM (Zeiss LIBRA 120, Germany) analysis and image capture of the membrane vesicles.

2.6. Perfusion of mice for vaginal sectioning

The lumen of the vaginas of sexually mature FVB/N females were flushed with a glutaraldehyde/formaldehyde perfusate (2% each in PBS) immediately after sacrifice, and the vaginas fixed in 2% glutaraldehyde and 2% formaldehyde in 0.1 M sodium cacodylate buffer pH 7.4 containing 2 mM calcium chloride. The tissue was cut into 1–2 mm³ pieces. Samples were sectioned on a Reichert‐Jung Ultracut E ultramicrotome, and sections collected onto 200 mesh formvar‐carbon coated copper grids and post‐stained with 2% uranyl acetate in 50% methanol and Reynolds lead citrate (Reynolds, 1963).

2.7. Sodium dodecyl sulphate‐polyacrylamide gel electrophoresis (SDS‐PAGE) and western blot analysis

Samples of homogenized EVs from VLF and of testicular tissues (used as positive control) for electrophoresis were diluted in 5× Laemelli sample buffer and heated at 95°C for 5 min. Twenty to 40 µg of proteins (or as specified) from tissues, fluids, or EVs were loaded in each lane of 10% polyacrylamide gels and transferred onto nitrocellulose membrane (Amersham Biosciences, UK). Blots were blocked for 1 hr at RT, incubated in anti‐PMCA1 (1:2,000) primary antibody or anti‐CD‐9 (1:1,000), and anti‐phosphotyrosine (1:250), overnight at 4°C, and processed as previously described (Al‐Dossary et al., 2013). Membranes were reprobed with HSC70 antibody (1:1,000), which served as an internal loading control for normalization and as an EV biomarker. The normalization approach was as described (http://bitesizebio.com/23411/the‐4‐important‐steps‐for‐western‐blot‐quantification/). To quantify the intensity of the bands, images from shorter film exposures were selected and Image Lab 6 software (Bio‐Rad, Hercules, CA) was used to subtract the background from the tyrosine‐phosphorylated bands and that of the HSC70 (n = 3).

2.8. In vitro sperm uptake of PMCA1, PMCA1/4, SPAM1, and tyrosine‐phosphorylated proteins from VLF EVs

Aliquots of caudal sperm from C57BL/6J males were coincubated with VLF EVs (total protein being 0.55 µg/µl) suspended in PBS or PBS (vehicle control) for 30 min or 3 hr at 37°C and assayed for protein uptake by flow cytometry as previously described (Al‐Dossary et al., 2013; Andrews, Galileo, & Martin‐DeLeon, 2015). After coincubation, sperm were washed with PBS and fixed in 1.5% paraformaldehyde for 1 hr at RT. After washing they were permeabilized with 0.1% triton X‐100 for 10 min at RT. They were then washed with PBS and blocked with 2% bovine serum albumin (BSA) and treated with the primary antibodies diluted 1:200 for anti‐PMCA1, 1:400 for SPAM1, and 1:50 for anti‐phosphotyrosine, anti‐PMCA1/4, and anti‐PMCA4, followed by the appropriate secondary antibodies conjugated to Alexa Fluor 488 and subsequent washing. Uptake was quantified in 50,000 sperm, using the appropriate IgG control and a FACSCalibur (Becton Dickinson, San Diego, CA) flow cytometer equipped with an argon laser with excitation at 488 nm.

To examine immunolabeled sperm with confocal microscopy, aliquots of the sperm suspension used for flow cytometry were smeared on slides which were then counterstained with DAPI in Prolong Gold.

2.9. In vivo exposure of caudal sperm to the vaginal environment of superovulated females

C57BL/6J females were superovulated with hormonal injections as described above. Thirteen and one‐half to 14 hr after the last injection freshly collected caudal sperm were re‐suspended in PBS and with the use of a pipette with an Eppendorf tip aliquots were deposited in the vagina of each female which was immobilized in a CO₂ chamber with a flow rate of 4–5 L/min. Females were immediately removed from the chamber after depositing sperm and, after 30 min, were killed to retrieve sperm from the vagina. In other experiments, 36 × 10⁵ sperm were deposited in the vagina for 10, 15, 20, or 30 min, before females were killed for recovery of sperm from the vagina. Sperm was collected after mincing the vaginal tissue in PBS. The tissue in the suspension was allowed to gravity settle and the supernatant was spun at 500g to collect the sperm. The control sperm were kept in PBS at 37°C and processed with the sperm collected from the vaginas for the frequency of acrosome‐reacted cells.

2.10. Hyaluronic acid‐enhanced progesterone‐induced AR

The AR was induced under physiological conditions using a modification of the method previously described (Morales et al., 2004; Smith et al., 2014). Briefly, caudal sperm samples that had been incubated with PBS or VGS reconstituted in PBS, or that had been deposited in the vagina were collected by centrifugation and washed in PBS. In the absence of incubation in a capacitating medium, they were then treated with 100 μg ml⁻¹ hyaluronic acid for 30 min followed by progesterone (3.18 μmol L⁻¹ for 5 min). Sperm were then pelleted (500g, 15 min) and resuspended in 4% paraformaldehyde and stored overnight at 4°C. The sperm were then resuspended in PBS to prepare a 1:5 dilution of the samples and ~200 μl of the suspension was dragged across a clean glass slide and allowed to air‐dry. Slides were then stained with 0.44% Coomassie Brilliant Blue G‐250 in 60% methanol‐acetic acid for 15 min. Alternatively, the acrosomal matrix of sperm on slides was stained with fluorescein isothiocyanate‐conjugated peanut agglutinin (PNA) lectin (L7381), 10 µg/ml, for 30 min in the dark. Slides were then washed in PBS three times and counterstained in DAPI with Prolong Gold. They were then examined under the confocal microscope (blindly) to determine the presence/absence of the acrosomal cap (in the case of Coomassie blue staining) or the green fluorescence signal for the acrosomal matrix, in a minimum of 200 sperm/sample from replicates of the assay.

2.11. Statistical analysis

Data were processed using Chi‐squared (χ²) analysis with Yates correction for contingency tables. The Students' t test was used to compare means of band intensities for western blot analysis and the means of the percentages of cells with the highest fluorescence intensities in the flow cytometric analysis.

3. RESULTS

Ultracentrifugation of VLF followed by negative staining of the resulting pellets which were analyzed by TEM revealed the presence of EVs of varying sizes. In Figure 1a, we show that EVs of both exosomal (<100 nm) and microvesicular (100nm‐1µm) sizes are present and have dubbed them as VGS. To show that these EVs have an endogenous vaginal origin and are not transported to the vaginal fluid from the uterine fluid, we investigated their presence in situ by examining vaginal sections after perfusion of the tissue. TEM analysis of the sections shows that the majority of EVs are entrapped in the vaginal tissue between keratinized and squamous epithelial cells, whereas others are present in the lumen (Figure 1b).

Figure 1.

Isolation of extracellular vesicles from VLF and their detection in situ in vaginal sections. (a) Negative staining of TEM of the VLF pellet reveals the presence of extracellular vesicles of varying sizes, *exosomes of size <100 nm and **microvesicles ranging in size from 100 nm to 1 µm diameter. Both are referred to as VGS. (b) TEM image of a vaginal cross section showing the membrane vesicles that are present in between the layers of vaginal tissue, as indicated by yellow arrows. An EV is seen in the lumen (arrowed). EV: extracellular vesicle; KC: keratinized cells; SEC: squamous epithelial cells; TEM: transmission electron microscopy; VGS: vaginosomes; VLF: vaginal luminal fluid [Color figure can be viewed at wileyonlinelibrary.com]

Western blot analysis was used for biochemical characterization of the EVs. Using OVS as a positive control, we show the presence of the 24 kDa CD9 tetraspanin band in the VLF EVs, VGS (Figure 2a). A second biochemical marker, HSC70, was detected in VGS via western blot analysis. Western blots also showed the presence of a calcium efflux pump, PMCA1, before the membrane was stripped and reprobed for HSC70 (Figure 2b).

Figure 2.

VGS are characterized biochemically as EVs by the presence of CD9 and HSC70, and contain PMCA1 in their cargo. (a) Western blot showing the presence of CD9 (24 kDa) on proestrus VGS (Pro‐VGS). Positive controls were proestrus and metestrus oviductosomes (Pro‐OVS and Met‐OVS). (b) Western blot showing VGS collected from WT and Pmca4 KO VLF carrying PMCA1 (139 kDa). WT and Pmca4 KO testis protein lysates were used as positive controls, whereas HSC70 was used as a loading control and extracellular vesicular marker. EV: extracellular vesicle; Pmca4: plasma membrane Ca²⁺‐ATPase 4; VGS: vaginosomes; VLF: vaginal luminal fluid; WT: wild type

To determine if VGS can transfer PMCAs and other proteins in their cargo to sperm, coincubation assays were performed. After washing and processing the sperm for immunofluorescence, VGS were shown to be capable of delivering SPAM1, PMCA1/4, and PMCA1, to sperm (Figure 3a). On the basis of the frequencies of sperm with the highest fluorescence intensity, flow cytometric analysis revealed an increase of 16.4‐fold for SPAM1, 8.2‐fold for PMCA1/4, and 27‐fold for PMCA1 when sperm was coincubated with VGS for 30 min, compared with the PBS control (Figure 3a‐i,’ ii,’ iii’). A t test for the mean number of sperm with the highest fluorescence intensities shows significant differences (p < 0.01) between sperm coincubated in VGS and PBS for the three proteins (Figure 3b). When the immunofluorescence was visualized by confocal microscopy, uptake of SPAM1 appeared to have occurred mostly on the flagellum whereas for PMCA1/4 both the flagellum and the head were involved (Figure 3c). When coincubation was extended for 3 hr, there was only an ~3‐fold increase in the proportion of sperm with the highest levels of PMCA4 on sperm for VGS/VLF, compared with PBS (Supporting Information S1).

Figure 3.

VGS are capable of delivering PMCAs and SPAM1 to caudal sperm following coincubation. (a) Flow cytometry reveals the specificity of the antibodies: (i) anti‐SPAM1, (ii) anti‐PMCA1/4, and (iii) anti‐PMCA1 show sperm with a right peak shift of fluorescence intensity, compared with the control treated with the IgG (rabbit, PMCA1 and SPAM1; mouse; PMCA1/4). Sperm aliquots in PBS or VGS reconstituted in PBS were incubated for 30 min, and subsequently washed and processed for immunofluorescence. (a‐iʹ, iiʹ, iiiʹ) Fluorescence was quantified in 50,000 cells for each sample, and the region with the highest intensity marked. The percentage of cells in this region reveals increases in frequencies with 16.4‐fold, 8.2‐fold, and 27.2‐fold for SPAM1, PMCA1/4, and PMCA1 for sperm in VGS, compared with those in PBS. (b) When the means of the percentage of PBS‐ and VGS‐coincubated cells in the marked region were analyzed by a t test, the differences were significant (p < 0.01) for each of the proteins (n = 3). (c) Confocal images of sperm immunolabeled for SPAM1, PMCA1/4, and PMCA1 are seen following coincubation in PBS (iʹ, iiʹ, iiiʹ) and VGS (iʺ, iiʺ, iiiʺ). IgG control of each antibody is seen in (i, ii, iii). The signals (green) are most intense in iʺ, iiʺ, and iiiʺ (sperm in VGS). The sperm heads are stained blue with DAPI in the middle panel followed by the merged image (original magnification = ×630). DAPI: 4',6‐diamidino‐2‐phenylindole; PBS: phosphate buffered saline; PMCAs: Plasma membrane Ca²⁺‐ATPase; SPAM1: Sperm adhesion molecule 1; VGS: vaginosomes [Color figure can be viewed at wileyonlinelibrary.com]

On the basis of the finding that tyrosine‐phosphorylated proteins are delivered to human sperm by uterosome‐like EVs (Franchi et al., 2016) and to mouse sperm by OVS (Bathala et al., 2018), we performed western blot analysis to investigate their presence in VGS as well as UTS. Using OVS as a positive control, 6 tyrosine‐phosphorylated proteins (including those with molecular weight (MWs) of ~55, 95, 130, and ~180 kDa) seen in OVS are present in the cargoes of both VGS and UTS (Figure 4a). Although the bands for these proteins in OVS tend to be more intense than in those in VGS and UTS, when they were normalized as a group with HSC70 there were no significant differences in the mean intensities between pairs of the three EVs (p = 0.18 for OVS and UTS; p = 0.38 for OVS and VGS). However, when the intensities of the 95 kDa band which is associated with the AR (Kwon et al., 2014) were compared independently among the three EVs, those for the OVS and the UTS were found to be significantly (p < 0.001) higher than that in VGS (Figure 4a‐ii). Importantly, VGS were able to transfer these proteins to caudal sperm. Flow cytometric analysis showed a 3.3‐fold increase in the frequency of cells with the highest fluorescence intensity/protein level when sperm was co‐incubated for 30 min in reconstituted VGS, compared with PBS (Figure 4b‐ii). This difference in frequency is significant (p < 0.001), as shown in Figure 4b‐iii. Confocal microscopy showed that uptake of tyrosine‐phosphorylated proteins occurred on both the sperm head and the flagellum (Figure 4c).

Figure 4.

VGS carry tyrosine‐phosphorylated proteins which they can deliver to sperm following coincubation. (a) Western blot shows 6 tyrosine‐phosphorylated proteins (PY; ranging from MW 55 to >180 kDa) in the cargoes of VGS and UTS, using OVS as a positive control. The membrane was stripped and reprobed for HSC70, as a loading control (40 µg protein/lane) to assess the relative intensities of the protein bands. (b) For the 95 kDa band, the intensities for OVS and UTS were significantly (p < 0.001) higher than that in VGS (n = 3). (b‐i) Flow cytometric analysis was performed on 50,000 cells for each sample. The specificity of the anti‐phosphotyrosine (PY) antibodies used is seen by the right peak shift of fluorescence intensity, compared with the IgG control. (b‐ii) The graph shows a right peak shift in fluorescence intensity after 30 min coincubation, with the highest fluorescence intensity (the gated or marked region) displayed by 72.2% of sperm in VGS compared with 21.8% of those in PBS. (b‐iii) The difference in the frequencies of the cells in the marked region is significant (p < 0.001), (n = 3). (c) Indirect immunofluorescence shows the localization of PY on the sperm head and flagellum, following coincubation in PBS and VGS (c‐ii,iii), with more intense staining in the latter after PY uptake and none seen in the IgG control (c‐i). (Original magnification = ×630). OVS: oviductosomes; PBS: phosphate‐buffered saline; VGS: vaginosomes; UTS: uterosomes [Color figure can be viewed at wileyonlinelibrary.com]

Finally, to investigate the physiological consequence of uptake of tyrosine‐phosphorylated proteins, among other fertility‐modulating proteins, we assessed the level of AR at 0 (control), 10, 20, and 30 min of coincubation of caudal sperm and VGS. Table 1 shows that whereas at 0 min the rate of AR sperm (distinguished in Figure 5a) for the PBS (control) and VGS reconstituted in PBS at 0 min (control) were similar (32% and 31%,), they were higher (83.9–95.3%) for the reconstituted VGS, with fold change differences of 3.3 (10 min), 3.0 (20 min), and 2.6‐fold (30 min), compared with the PBS. The grand total of acrosome‐reacted sperm coincubated in PBS and VGS averaged 29.2% and 75.7% (2.8‐fold increase), respectively. The fold‐changes represent significant (p < 0.001) increases of AR sperm at each of the time points and the overall group, compared with the control. When the frequencies of acrosome‐reacted sperm were investigated following coincubation of sperm at various concentrations of VGS (undiluted, diluted 2×, 5×, 10×, 20×) for 10 min, a dose‐response relationship was detected. However, at 20× (the lowest concentration) and 10×, the rates of acrosome‐reacted sperm were significantly (p < 0.05; p < 0.01) higher than that seen for sperm coincubated in PBS (Figure 5b).

Table 1.

Rates of acrosome‐reacted sperm after coincubation of sperm with PBS (as control) or VGS re‐constituted in PBS

Incubation time (min)	Number (%)
	Status	Control	VGS
0	AR	115 (32.03)	107 (31.02) a , b , f
	No AR	306	316
	Total	421	423
10	AR	91 (25.78) d	405 (83.97) a , b
	No AR	328	88
	Total	419	493
20	AR	104 (30.98) e	429 (92.37) c , d
	No AR	324	49
	Total	428	478
30	AR	123 (28.12) f	403 (95.36) e , f
	No AR	348	20
	Total	471	423
Grand Total	AR	433 (29.23) g	1344 (75.68) g
	No AR	1306	473
	Total	1739	1817

Note. AR: Acrosome reaction. Identical letters refer to significant differences.

^a p < 0.001: between before and after 10 min coincubated with VGS.

^b p < 0.001: between before and after 20 min coincubated with VGS.

^c p < 0.001: between before and after 30 min coincubated with VGS.

^d p < 0.001: between coincubation with PBS as a control and VGS for 10 min.

^e p < 0.001: between coincubation with PBS as a control and VGS for 20 min.

^f p < 0.001: between coincubation with PBS as a control and VGS for 30 min.

^g p < 0.001: between the grand control and coincubated with VGS.

Figure 5.

Sperm‐VGS interaction in vitro and in vivo increases the frequency of AR sperm: in vitro response is dose‐dependent. (a) The acrosomal status of sperm was detected via Coomassie Blue staining after VGS coincubation. Sperm with an intact acrosomal cap is shown by black arrows, whereas a missing cap after the acrosome reaction is shown by the red arrow. (b) Sperm coincubated in varying concentrations of the VGS (undiluted, 2×, 5×, 10×, 20x diluted VGS) or PBS for 30 min show a dose‐response relationship. Using χ2 analysis, there is a significant (*p < 0.05, **p < 0.01, and ***p < 0.001) increase in AR between samples coincubated with PBS and 20×/10x dilution. (c) Different numbers (12, 24, 36, 48 × 10⁵) of sperm were deposited in the vaginas of superovulated mice and the rate of acrosome reaction compared with the ex vivo control: †C (48 × 10⁵). Chi‐squared (χ ²) analysis shows a significant (p < 0.001) increase between the samples and control. (d) Sperm (36 × 10⁵) were deposited in the vaginas for different time periods (0–30 min). The percentages of AR sperm seen after 10 min and beyond were significantly (p < 0.01) increased, compared with that of the start point (0 min). Increases between 10 and 30 min were insignificant. AR: acrosome‐reacted; PBS: phosphate‐buffered saline; VGS: vaginosomes [Color figure can be viewed at wileyonlinelibrary.com]

In addition, sperm was exposed to the vaginal environment in vivo. When various numbers of caudal sperm were artificially deposited in the vagina for 30 min and sperm retrieved after the sacrifice of the females, exposure to the vaginal environment resulted in significant (p < 0.001) increases in the frequency of acrosome‐reacted sperm, compared with the unexposed control held at 37°C (Figure 5c). Similarly, when the same number of sperm were exposed to the vaginal environment for 10, 15 20, or 30 min there was a significant (p < 0.001) increase in the frequency of acrosome‐reacted sperm for 10 min (60%), compared with the PBS control (28%) with the longer exposures having higher frequencies (Figure 5d).

To confirm the rates of AR seen in sperm exposed to VGS and the PBS control, as reflected in the staining of the acrosomal cap, we assayed the rate of sperm with an intact acrosomal matrix, as detected by fluorescently labeled PNA lectin which is a biochemical marker of the acrosomal matrix (Guyonnet, Egge, & Cornwall, 2014). In Figure 6 we show PNA‐positive and –negative sperm. Of 212 sperm coincubated in PBS 143 or 67% were PNA‐positive and 33% PNA‐negative. On the other hand, 96 of 229 or 42% were PNA‐positive and 58% were PNA‐negative for sperm coincubated in VGS. Chi‐squared (χ²) analysis of a 2 × 2 contingency table, with Yates correction, showed the difference to be highly significant (p < 0.01); and indicates that the rate of PNA‐negative sperm, which is acrosome‐reacted, is dependent on the presence of VGS during coincubation.

Figure 6.

Detection of AI and AR sperm via FITC‐PNA staining of the acrosomal matrix. (a) An AI sperm is seen with the green FITC‐PNA staining (arrowed) of the acrosomal matrix. (b) An AR sperm is seen without the acrosomal matrix staining. Phase contrast images are seen below sperm with the acrosomal matrix present/absent, followed by the merged images. AI: acrosome‐intact; AR: acrosome‐reacted; FITC: fluorescein isothiocyanate; PNA: peanut agglutinin [Color figure can be viewed at wileyonlinelibrary.com]

4. DISCUSSION

Our results show that, like other female biofluids (Al‐Dossary et al., 2013; Griffiths Miller et al., 2008), the murine VLF contains membrane‐bound vesicles that are released from the epithelial lining of the vagina. The detection of VGS in situ in vaginal sections is consistent with their release from the tissue into the lumen. These EVs, similar in sizes to those detected in the uterus and oviducts (Al‐Dossary et al., 2013; Griffiths, Galileo et al., 2008), carry the characteristic biochemical markers (CD9 and HSC70) as well as sperm regulatory proteins including the PMCAs and tyrosine‐phosphorylated proteins. As vaginal exosomes are conserved in humans where they were shown to inhibit an early step of the HIV‐1 life cycle (Smith & Daniel, 2016), it is highly likely that the sperm regulatory proteins detected in the mouse exosomes are also conserved in humans.

During mating in the mouse, sperm is deposited in the vagina where vaginal plugs are formed, promoting transport of the sperm to the uterus (Suarez & Pacey, 2006). We show that in vitro exposure of caudal sperm to VGS in the VLF for 30 min, compared with PBS, resulted in significantly higher numbers of sperm with elevated levels of PMCA1 (27‐fold), PMCA1/4 (8.2‐fold), SPAM1 (16.4‐fold), and tyrosine‐phosphorylated proteins (3.3‐fold). When caudal sperm and VGS/VLF were coincubated for 3 hr, there was only a 3‐fold increase in PMCA4, compared with the PBS control. It is likely that this lower level of uptake at 3 hr reflects a lower level of sperm viability which is required for uptake as fluorescence recovery after photobleaching experiments show that uptake occurs in live cells (Al‐Dossary et al., 2015).

As the mechanism of EV cargo delivery to sperm has been shown to be fusogenic (Al‐Dossary et al., 2015; Schwarz et al., 2013), sperm‐VGS interaction would be expected to enrich the sperm membrane with cholesterol and lipids. This enrichment would decrease its fluidity and consequently result in the prevention of premature capacitation and premature AR, as shown for human sperm and their interaction with prostasomes (Arienti, Carlini, De Cosmo et al., 1998a; Arienti, Carlini, Polci, Cosmi, & Palmerini, 1998a; Carlini, Palmerini, Cosmi, & Arienti, 1997; Pons‐Rejraji et al., 2011).

The levels of protein uptake in sperm after coincubation may be related to their physiological roles. After 30 min coincubation, the large increases of the PMCAs (8.2‐ and 27‐folds) which are Ca²⁺ efflux pumps are consistent with their role in keeping intracellular Ca²⁺ levels low and thus preventing premature capacitation and premature AR. In fact PMCA4, the major Ca²⁺efflux pump in murine sperm (Wennemuth et al., 2003), is likely to be stimulated by the decapacitation factor reported by Adeoya‐Osiguwa and Fraser (1996). On the other hand, the 3.3‐fold increase in tyrosine‐phosphorylated proteins which are known to stimulate capacitation (Baker et al., 2006; Kwon et al., 2014; Visconti et al., 1995) is an order of magnitude lower than the increases seen for the PMCAs and SPAM1 (16.4‐fold) after 30 min of coincubation. Whereas SPAM1 is a hyaluronidase that facilitates sperm penetration in the cumulus matrix and the perivitelline space (Martin‐DeLeon, 2011), the lower increase of tyrosine‐ phosphorylated proteins are of physiological consequence in incrementally stimulating capacitation. Evidence for this comes from our finding that a 10‐min period of coincubation of sperm with VGS in vitro, or in vivo (following exposure to the vaginal environment) resulted in the induction of the AR, a hallmark of capacitated sperm. Table 1 and Figure 5d show that significantly (p < 0.001; p < 0.01) more sperm were acrosome‐reacted after coincubation of sperm with VGS, compared with PBS.

That significantly more sperm were acrosome‐reacted after coincubation with VGS, compared with PBS, is consistent with the presence in VGS of the 95 kDa phosphorylated protein that is associated with the AR (Kwon et al., 2014; Leyton & Saling, 1989), and the delivery of these proteins to sperm. It should be noted that the AR in this study was progesterone‐induced. Progesterone has been shown to induce an increase in cytosolic calcium in sperm and their fusion with prostasomes which contain calcium enhances the calcium increase (Arienti, Nicolucci, Santi, Carlini, & Palmerini, 2001). As we have detected calcium in OVS (unpublished observation), it is highly likely to be present in VGS which after fusion with sperm would enhance the intracellular calcium induced by progesterone to a level that would prime sperm for the AR.

A significant increase in acrosome‐reacted sperm was also detected when porcine sperm were coincubated with EVs isolated from seminal plasma (Siciliano et al., 2008). Thus VGS appear to play similar roles to prostasomes, with respect to inhibiting premature capacitation and premature AR as well as promoting them. Based on the significantly increased number of sperm with elevated levels of proteins delivered by VGS, it appears that VGS play a more predominant role in inhibiting these physiological states than in promoting them. It should also be noted that SPAM1 is present in murine prostasomes (Zhang, Morales, Badran, El‐Alfy, Martin‐DeLeon, 2004) which are also likely to carry PMCA4 which is present in human prostasomes (Andrews et al., 2015). It is thus conceivable that the high levels of PMCAs and SPAM1 that are delivered to sperm by VGS/VLF are accommodated by caudal sperm, because they have not yet been exposed to the prostatic/seminal fluid. Thus in the in vivo situation, ejaculated sperm might be likely to acquire lower levels of these proteins from VGS, by virtue of their interaction with, and prior uptake from, prostasomes. It is also likely that in vivo the length of time of exposure to the vaginal environment might be shorter than that used in the study.

Our results show a dose‐response relationship with respect to the frequency of acrosome‐reacted sperm after coincubation with various concentrations of VGS. It is possible that the physiological concentration of VGS to which sperm are exposed in vivo is less than we have used in vitro because all of the VGS collected might not have been naturally released in the lumen. Thus we investigated the effect of sperm exposure to various dilutions of the VGS/VLF and detected that even for the most dilute concentration (20×) there was a significant increase in the number of acrosome reaction sperm. Further, in vivo exposure of sperm to the vaginal environment also showed a significant increase in AR rates. Thus the observation that tyrosine‐phosphorylated proteins, which are increased in sperm during capacitation (Kwon et al., 2014), can be delivered to sperm by VGS may provide an explanation for the recent findings of the acrosomal status of mouse sperm during their migration in the oviduct (Perez‐Cerezales et al., 2018). Within recent years, the widely held view that the fertilizing spermatozoon begins its AR on the zona pellucida of the oocyte and that sperm that has undergone this reaction earlier have lost their fertilizing potential, has been challenged (Perez‐Cerezales et al., 2018).

Inoue, Satouh, Ikawa, Okabe, and Yanagimachi (2011) have shown that mouse oocytes are fertilizable by sperm that completed the AR before entering the cumulus oophorus. They reported that sperm which acrosome reaction long before contacting the zona pellucida maintained their ability to traverse it and to effect fertilization (Inoue et al., 2011). Further, Hino et al. (2016) have shown that mouse sperm acrosome‐react during sperm migration in the lower and upper isthmus of the oviduct. Interestingly, it has been reported that almost all fertilizing mouse sperm within the oviduct begin to acrosome‐react before their ascent to the ampulla from the isthmus, although nothing is known of the inducing factor(s) involved (Hino et al., 2016; Jin et al., 2011; La Spina et al., 2016; Muro et al., 2016). Our findings for the transfer of tyrosine‐phosphorylated proteins to sperm by VGS, and their presence in murine UTS and OVS (Figure 4A) as recently shown by Bathala et al. (2018), allow us to propose that murine sperm are primed to acrosome‐react during their transit in the vagina, uterus, and oviduct by the acquisition of tyrosine‐phosphorylated proteins from EVs present in the luminal fluids of all three organs. This conclusion is consistent with the reported delivery of tyrosine‐phosphorylated proteins by uterosomal‐like vesicles to human sperm (Franchi et al., 2016). Our data, therefore, provide an answer to a question that has puzzled investigators for some time, namely “What induces the AR of mouse sperm within the isthmus” (Hino et al., 2016)? Thus, the present findings fill an important gap in our knowledge of the induction of murine sperm capacitation and the resulting AR.

Our results also open the possibility of treating some types of infertility in which there is a deficiency of fertility‐modulating sperm proteins. As EVs are currently being explored for clinical applications, the possibility exists that engineered EVs could be used to deliver missing proteins to sperm in patients who carry genetic mutations (Barkalina, Jones, Wood, & Coward, 2015). As humans are like mice, with respect to sperm delivery in the vagina which is readily accessible, engineered EVs could be provided in the vaginal fluid or in the form of a suppository, to allow the delivery of relevant proteins to sperm.

Supporting information

Supporting information

Supporting information