|Home | About | Journals | Submit | Contact Us | Français|
Phospholipase A2 (PLA2) activity has been shown to be involved in the sperm acrosome reaction (AR), but the molecular identity of PLA2 isoforms has remained elusive. Here, we have tested the role of two intracellular (iPLA2β and cytosolic PLA2α) and one secreted (group X) PLA2s in spontaneous and progesterone (P4)-induced AR by using a set of specific inhibitors and knock-out mice. iPLA2β is critical for spontaneous AR, whereas both iPLA2β and group X secreted PLA2 are involved in P4-induced AR. Cytosolic PLA2α is dispensable in both types of AR. P4-induced AR spreads over 30 min in the mouse, and kinetic analyses suggest the presence of different sperm subpopulations, using distinct PLA2 pathways to achieve AR. At low P4 concentration (2 μm), sperm undergoing early AR (0–5 min post-P4) rely on iPLA2β, whereas sperm undergoing late AR (20–30 min post-P4) rely on group X secreted PLA2. Moreover, the role of PLA2s in AR depends on P4 concentration, with the PLA2s being key actors at low physiological P4 concentrations (≤2 μm) but not at higher P4 concentrations (~10 μm).
The acrosome reaction (AR)3 is an exocytotic and mandatory event occurring in the cumulus mass or on the zona pellucida of the oocyte and allowing the release of several types of sperm enzymes. The release of these enzymes leads to a partial hydrolysis of the layers surrounding the oocyte, allowing the sperm to reach the oolema where the fusion between the two gametes occurs. The physiological agonists of the AR and their respective downstream molecular pathways are the subject of intense investigation (1, 2). Two main physiological agonists were identified as follows: progesterone (P4), which is released by the cells of the cumulus mass, and ZP3, one of the glycoproteins of the zona pellucida (ZP) (3,–5). However, for most fertilizing mouse sperm, the AR is triggered before contact with the ZP (6), and acrosome-reacted sperm are able to freely cross both the cumulus and the ZP (7), strongly suggesting that AR occurs before reaching the ZP and that P4 may be the main physiological effector of mouse AR in vivo. As in most exocytotic processes, Ca2+ plays a central role, and blocking the increase of [Ca2+] completely inhibits the AR. It is accepted that the main targets of Ca2+ are specific calcium-binding proteins such as synaptotagmin (8, 9), even if Ca2+ also activates calmodulin-dependent enzymes such as PI3K, whose inactivation blocks the AR (10). Numerous agonist-dependent actors involved in [Ca2+] increase have been characterized, including Ca2+ influx through activation of voltage-dependent calcium channels and store-operated channels and calcium release from the acrosome itself through activation of PLCδ4 and the inositol 1,4,5-trisphosphate receptor (2, 11,–13). Contrary to neuronal exocytosis where [Ca2+] increase is followed within milliseconds by degranulation, sperm granule exocytosis spreads over 30 min in the presence of the agonists. Several molecular pathways are activated during the first few minutes that are mandatory for the sperm AR, including PLA2. These enzymes catalyze the hydrolysis of phospholipids at the sn-2 position and generate lysophospholipids and free fatty acids that are precursors of different lipid mediators. The involvement of PLA2 in AR was first suggested 30 years ago by experiments showing that a PLA2 activity was associated with mouse acrosomal membranes and was inhibited by nonspecific PLA2 inhibitors such as indomethacin, sodium meclofenamate, mepacrine, and p-bromophenacyl bromide (14). The role of PLA2 was reinforced by indirect results showing that fatty acids and lysophospholipids could trigger the acrosome reaction (15,–17). By using 14C-labeled fatty acids, Roldan and Fragio (18) showed that fatty acids are released during the AR of ram sperm and that RO314493, a PLA2 inhibitor, blocked both fatty acid release and the acrosome reaction. Similar results were found in guinea pig sperm, where progesterone or ZP-induced fatty acid release and the AR were blocked by aristolochic acid, another PLA2 inhibitor (19). Another study showed that the mouse AR is also blocked by ONO RS-82, a different PLA2 inhibitor (20). Taken together, these studies showed that the activation of PLA2 plays a central role in the AR and that the production of fatty acids and/or lysophospholipids is required for acrosomal and plasma membrane fusion during the AR.
At this time, however, the molecular identification of the PLA2 isoforms involved in the AR observed in these pharmacological studies was impossible due to the lack of data regarding the specificity of the above used PLA2 inhibitors. Indeed, the PLA2 superfamily comprises now 37 different enzymes, divided into 16 groups and many more subgroups. It includes six distinct major types of enzymes, namely the cytosolic PLA2s (cPLA2) and the Ca2+-independent PLA2s (iPLA2), both groups being intracellular, the secreted PLA2s (sPLA2), and more specific groups such as the platelet-activating factor acetylhydrolases, the lysosomal PLA2s, and the adipose PLA2 (21, 22). Both intracellular and secreted PLA2s have been identified in mouse and human sperm cells (23,–26). In mature mouse sperm, the group X sPLA2 (mGX) is located in the acrosome and is released during the AR (24). It specifically improves fertilization and early embryo development (24, 27, 28).
PLA2s are involved in numerous physiological processes (29, 30), and one of the challenging questions regarding PLA2s is their respective contribution to cellular functions. The pharmaceutical industry has made significant efforts to design specific PLA2 inhibitors with the highest possible affinity (22) to control PLA2 activity in different pathological contexts. Among the different emerging drugs, the following drugs with reported specificities represent valuable tools to decipher the role of each PLA2 isoform in the AR. For sPLA2, LY329722 is an interesting compound, which was developed from a lead compound, LY311727, a very potent inhibitor of human group IIA sPLA2 (31). LY329722 inhibits several subgroups of sPLA2 (I/II/V/X) and has an IC50 of 30 nm for the group X sPLA2 in vitro. For iPLA2, two inhibitors emerged, S-bromoenol lactone (BEL) and FKGK18. BEL is a selective and irreversible inhibitor of iPLA2β with a half-maximal inhibition of around 60 nm after preincubation (32). Although it is a potent inhibitor of iPLA2β, BEL also inhibits non-PLA2 enzymes (33). For this reason FKGK18, a recently characterized iPLA2β inhibitor presenting fewer off-target effects with a similar potency, was identified (33). Finally, pyrrolidine-1 (Pyr-1), a pyrrolidine-based inhibitor, was identified to inhibit cPLA2α (IC50 = 70 nm). In vitro assays showed that it is less potent toward cPLA2γ and iPLA2β (IC50 ~10 μm) and that 10 μm Pyr-1 has no inhibitory effect on group IIA, V, and X secreted PLA2s (34).
Because of the presence of a least three different PLA2s in sperm (23, 24, 26), and because a cross-talk between different PLA2s has been described (35), we assessed the possible involvement of several PLA2s during P4-induced AR in mouse sperm cells. If several PLA2s are involved, then their respective contributions and their time of action should be characterized. Herein, we show for the first time that two PLA2s are required during the mouse AR, iPLA2β and mouse group X sPLA2 (mGX). Conversely, cPLA2α does not appear to be involved. Interestingly, we showed that the overall contribution of PLA2 in the AR depends on P4 concentration.
All animal procedures were performed according to the French guidelines on the use of living animals in scientific investigations with the approval of the local Ethical Review Committee. mGX-KO mice (null for Pla2g10 gene) on a C57BL/6J background were obtained from Lexicon Inc. as described (36). cPLA2α KO mice (null for the Pla2g4a gene) were obtained as described (37). iPLA2β KO mice (null for the Pla2g6a gene) were obtained as described previously (23). All other mice (OF1 strain, 2–6 months old) were from Charles River Laboratories (Les Oncins, France).
Sperm was displayed over a slide, dried at room temperature, and then fixed in 75% ethanol for Harris-Schorr staining. Motility of sperm was assessed with computer-assisted motility analysis. Non-capacitated sperm suspension was immediately placed onto an analysis chamber (100 μm depth, Leja Products B.V., Netherlands) and kept at 37 °C for microscopic quantitative study of sperm movement. Sperm motility parameters were measured at 37 °C using a sperm analyzer (Hamilton Thorn Research, Beverly, MA). The settings employed for analysis were as follows: acquisition rate, 60 Hz; number of frames, 100; minimum contrast, 25; minimum cell size, 10; low static size gate, 2.4; high static size gate, 2.4; low static intensity gate, 1.02; high static intensity gate, 1.37; minimum elongation gate, 12; maximum elongation gate, 100; magnification factor, 0.70. The motility parameters measured were curvilinear velocity (VCL), average path velocity (VAP), and straight-line velocity (VSL). At least 100 motile sperm were analyzed for each assay. Motile sperm and progressive sperm were characterized by VAP >1 μm/s, by average path velocity >30 μm/s and straightness (VSL/VAP) >70%, respectively.
Eggs were collected from mature OF1 females, synchronized with 5 units of pregnant mare serum gonadotrophin and 5 units of human chorionic gonadotrophin. Sperm were capacitated for 35–55 min in M16 2% BSA (37 °C, 5% CO2) and introduced into droplets containing oocytes. Oocytes were incubated with 1.5 × 105 to 5 × 105 capacitated sperm/ml (37 °C, 5% CO2) in M16 medium, and unbound sperm were washed away after 4 h of incubation. Twenty four hours after fertilization, the different stages, i.e. unfertilized oocytes, aborted embryos (corresponding to fragmented oocytes or oocytes blocked after the extrusion of the second polar body), and 2-cell embryos (as an indication of successful fertilization) were scored.
Mouse sperm, obtained by manual trituration of caudae epididymides, were allowed to swim in M2 medium for 10 min and capacitated in M16 medium with 2% fatty acid-free BSA at 37 °C, 5% CO2 for 65 min. For progesterone (P4) treatment, capacitated sperm (35 min) were incubated with P4 in M16 medium at 37 °C for the last 30 min of capacitation (total duration of the experiment 65 min). Control experiments were always performed with the same concentration of DMSO in the medium (DMSO concentrations ranging from 0.01 to 0.1%).
Sperm were transferred in PBS solution and then fixed with 4% paraformaldehyde solution for 5 min. Sperm were washed (100 mm ammonium acetate, 2 min), wet mounted on slides, and air-dried. Slides were then rinsed with water and stained with Coomassie Blue (0.22%) for 2 min. Slides were analyzed, and at least 150 sperm were scored per condition.
Eggs were collected from mature OF1 females (6 weeks old) synchronized with 5 units of pregnant mare serum gonadotrophin and 5 units of human chorionic gonadotrophin. Superovulation was confirmed by plasma P4 concentrations determined to be around 150 nm (154.4 ± 35.12 nm, n = 5). Cumulus-oocyte complexes harvested from 10 mice were placed in culture Petri dishes containing 250 μl of M16 medium and incubated for 10 min, 1 and 2 h at 37 °C in a 5% CO2 atmosphere. Incubations were stopped by putting the dishes on ice and by centrifugation at 5,000 × g for 10 min at 4 °C. The supernatants and the cell pellets were frozen and used to determine the concentration of progesterone.
Progesterone concentrations were determined by radioimmunoassay using 125I-labeled progesterone (IBL International, Hamburg, Germany). The RIA detection limit was calculated to range between 0.1 and 0.5 ng/ml, i.e. between 0.3 and 1.6 nm depending upon the labeling lot. Cross-reactivity with other steroid hormones was reported to be minimal as follows: <3.5% for 20β-dihydroprogesterone and 5α-pregnane-3,20-dione; 1.5% for 17α-hydroxyprogesterone; 0.8% for 11-deoxycorticosterone; <0.4% for pregnenolone and corticosterone; and ≤0.1% for all other steroid hormones tested by the manufacturer. Of the steroids indicated, only 17α-hydroxyprogesterone is generated in significant amounts by cumulus cells. Its concentration is ~5-fold lower that of P4 in plasma during the luteal phase, and a 1.5% cross-reactivity with P4 can therefore not significantly bias the data. Briefly, after centrifugation the cell pellets were extracted with diethyl ether and centrifuged to remove cell debris. The ether extracts were evaporated under nitrogen, and the residues were re-dissolved in 250 μl of Tris buffer (10 mm, pH 7.4) containing 0.1% BSA. Culture medium (250 μl) and plasma samples were assayed directly according to the manufacturer's instructions. Progesterone levels were undetectable in the Tris/BSA buffer as well as in the M16 medium used for the incubations. For the calculation of intra-cumulus P4 concentrations, the total volume of the cumuli was estimated based on the number of oocytes per sample and on their average volume.
Washed sperm were resuspended in Laemmli sample buffer without β-mercaptoethanol and boiled for 5 min. After centrifugation, 5% β-mercaptoethanol was added to the supernatants, and the mixture was boiled again for 5 min. Protein extracts equivalent to 1–2 × 106 sperm were loaded per lane and subjected to SDS-PAGE. Resolved proteins were transferred onto polyvinylidene difluoride membranes (Millipore). Membranes were treated with 20% fish skin gelatin (Sigma) in PBS-T and then incubated for 1 h at room temperature with anti-phosphotyrosine antibody (clone 4G10, Millipore, Molsheim, France) (1:10,000); this was followed by 1 h of incubation with a horseradish peroxidase-labeled secondary antibody. Immunoreactivity was detected using chemiluminescence detection kit reagents and a ChimidocTM Station (Bio-Rad, Marnes-la-Coquette, France).
M2 medium, M16 medium, progesterone, and BSA were from Sigma (Lyon, France). Pregnant mare serum gonadotrophin and human chorionic gonadotrophin were from Intervet (Beaucouze, France), and BEL was from Interchim (Montluçon, France). Pyr-1 and LY329722 was provided by Prof. Michael Gelb (University of Washington, Seattle) and FKGK18 by George Kokotos (University of Athens, Greece).
n represents the number of biological replicates, and for each replicate, more than 100 sperm were assessed per condition. Statistical analyses were performed with SigmaPlot. t tests were used to compare the effects of various compounds on AR and fertility. Data represent mean ± S.E. Statistical tests with 2-tailed p values of ≤0.05 were considered significant.
We focused our study on iPLA2β, mouse group X sPLA2 (mGX), and cPLA2α for the following reasons. First, iPLA2β and mGX have been shown to be present in mouse sperm (23, 24). Second, cPLA2α is a constitutive enzyme with a ubiquitous expression and an important role in lipid mediator release and vesicle trafficking (38, 39), and more importantly its enzymatic activity and distribution are modified in patients presenting asthenozoospermia (26), suggesting that it may be important for male fertility. Deficient mice for these three proteins were previously described; the absence of the corresponding protein was confirmed, and their phenotype was studied (23, 36, 37). Nevertheless, the reproductive phenotype of mGX and iPLA2β KO males was partially reported (23, 24), but no data are available for cPLA2α KO males. In Table 1, we present for the first time a comparative table showing several reproductive parameters of males from these three deficient mice. Both iPLA2β and mGX KO males exhibit sperm defects with teratozoospermia and low motility (Table 1 and Fig. 1). iPLA2β-deficient males were clearly the most affected, presenting deficient spermatogenesis as witnessed by the significant level of teratozoospermia (45% of typical sperm morphology versus 92% for WT sperm) and low motility, leading to complete in vitro infertility. The sperm phenotype of mGX KO males was less severe, with 70% of typical sperm morphology and half-reduction of progressive sperm, leading to a significant decrease of in vitro fertilization outcome (Table 1 and Fig. 1). In contrast, cPLA2α KO males presented a normal spermatocytogram and fertility, although mobility was affected. These latter results suggest that cPLA2α is dispensable. However, its involvement in AR may be revealed only under specific conditions, as shown for Pkdrej KO mice for which the defective reproductive phenotype was revealed only under specific competing conditions (40). For this reason, the involvement of cPLA2α was also explored in the rest of this study.
In the literature, although numerous concentrations of P4 have been tested to induce AR, from sub-micromolar to tens of micromolars, P4 at 10–15 μm is usually used (20, 25, 41, 42). These concentrations may, however, not represent the physiological concentrations met by the sperm during their journey to oocytes embedded in the cumulus mass. To fit such physiological conditions, it was thus important to evaluate physiological P4 concentrations in both the cumulus mass and the tubal fluid. Cumuli of 10 superovulated females were harvested and incubated for 120 min in the fertilization medium, and [P4] was measured after centrifugation in both the cumuli and in the incubation medium by radioimmunoassay. We found similar P4 concentrations in the cumuli (0.71 ± 0.08 μm (n = 5)) and in the fertilization medium (1.13 ± 0.27 μm (n = 5)). Moreover, incubation of cumuli for various time points indicates that P4 synthesis is stable for at least 2 h (Fig. 2A). The value found for cumuli should be representative of the physiological value. This may not be the case for the fertilization medium because tubal fertilization is a dynamic process in which tubal fluid undergoes a continuous turnover. Indeed, P4 upon its synthesis is partly metabolized into 17OH-progesterone, a precursor of androstenedione, and partly released into the extracellular fluid by passive diffusion as it is a lipophilic hormone. Part of the progesterone that is released further diffuses and can be metabolized by adjacent cells, with the rest being cleared by the flux of tubal fluid and by the bloodstream. It is therefore difficult to reproduce in vitro these in vivo conditions. In our experimental conditions, the only clearance mechanism is metabolism of P4 by the cumulus cells, and therefore the measured concentration reflects the maximal attainable concentrations of P4 in vivo. One can therefore reasonably assume that the [P4] surrounding the cumuli does not exceed low micromolar levels. As an example, P4 concentrations in tubal fluid were reported to range from 2 to 10 nm in rabbit (43) and from 50 to 100 nm (44) in hamster depending upon the phase of the oestrous cycle. In super-ovulation, these concentrations can be assumed to be higher, due to the maturation of many more follicles. Thus, sperm meet a P4 concentration gradient during its journey to oocytes, from sub-micromolar concentrations in the tubal fluid to low micromolar concentrations in the cumulus mass. For this reason, two concentrations of P4 were tested as follows: 10 μm, a concentration that has been tested in various reports, and also 2 μm, a concentration closer to the physiological one. AR was assessed with the Coomassie Blue protocol (Fig. 2B) (45). We next performed a dose-response experiment of the AR of intact sperm at the end of the capacitation period as a function of P4 concentration ([P4]), testing P4 concentrations between 0.1 and 50 μm (Fig. 2C). The dose-response curve is bell-shaped, and 10 μm P4 elicited the highest response with 65.1 ± 1.9% (n = 5) acrosome-reacted sperm after 30 min of incubation, likely explaining why such concentrations are generally used in reports studying molecular pathways involved in the AR. Remarkably, physiological concentration of P4 at 2 μm was potent and triggered AR of ~55% of intact sperm.
We first evaluated the involvement of iPLA2β by measuring the inhibitory effect of BEL and FKGK18 on the P4-induced AR. These compounds were tested on sperm capacitated for 35 min and further incubated for 30 min with 2 or 10 μm P4, and the AR rate was compared with the level obtained in the absence of the tested drug. The two compounds are potent inhibitors of the P4-induced AR with similar inhibitory effects; at 2 μm P4 the induced AR is inhibited by ~25% (n = 3–6) and at 10 μm by ~20–30% (n = 3–6) (Fig. 3, A and B). These results suggest that iPLA2β is involved in the P4-induced AR in the mouse. To confirm this result, the acrosome reactions of sperm from KO males deficient in iPLA2β were analyzed and compared with those of control sperm from WT littermate males. Because these KO mice have a high level of teratozoospermia, we only took into account sperm exhibiting a normal head shape. The absence of the corresponding protein led to a decrease in the AR rate, and the percentage of decrease was close to those measured in pharmacological experiments (Fig. 3C), confirming the involvement of iPLA2β in P4-induced AR. Finally, the specificity of BEL was challenged by measuring its ability to inhibit the AR of sperm from iPLA2β KO mice. BEL was no longer effective, demonstrating that its inhibitory effect on AR is mediated by iPLA2β (Fig. 3D). Altogether, these results demonstrate that iPLA2β contributes to P4-induced AR.
Next, we evaluated the involvement of mGX sPLA2 by measuring the inhibitory effect of LY329722 on the P4-induced AR as described above. LY32972, a pan inhibitor of sPLA2, was a potent inhibitor of the P4-induced AR; at 2 μm the P4-induced AR is inhibited by ~30% (n = 3) and at 10 μm by ~35% (n = 3) (Fig. 4A). These results suggest that an sPLA2 is involved in the P4-induced AR in the mouse. We assumed that it was most probably the sPLA2 of group X (mGX), as we have previously shown that mGX is the only sPLA2 in mouse sperm (24). To confirm this result, the acrosome reactions of sperm from mGX KO males were analyzed and compared with those of control sperm from WT littermate males. The absence of mGX led to a decrease in the AR rate, and the percentage of decrease was close to that measured in pharmacological experiments (Fig. 4B), confirming the involvement of mGX in P4-induced AR. Finally, LY329722 was unable to inhibit the AR of mGX KO sperm, demonstrating that its inhibitory effect is mediated by mGX (Fig. 4C) and confirming that mGX is the only sPLA2 involved in mouse AR. Altogether, these results demonstrate that mGX contributes to P4-induced AR.
We finally evaluated the involvement of cPLA2α by measuring the inhibitory effect of Pyr-1 on the P4-induced AR as described for iPLA2β. Pyr-1 was a potent inhibitor of the P4-induced AR; at 2 μm the P4 induced AR was inhibited by ~30% (n = 3) and at 10 μm by ~35% (n = 3) (Fig. 5A), suggesting that cPLA2α is involved in the P4-induced AR in the mouse. However, Pyr-1 displayed the same inhibitory effect on both WT and cPLA2α KO sperm, suggesting that Pyr-1 is not specific for cPLA2α (Fig. 5B). This latter result, associated with the fact that the AR rates of sperm from cPLA2α KO and WT males were similar (Fig. 5C), demonstrates that cPLA2α is not involved in the mouse sperm AR and that at 1 μm, Pyr-1 inhibits the AR through an enzyme other than cPLA2α. Because this enzyme may be iPLA2β, the inhibitory effect of Pyr-1 was measured on sperm from iPLA2β KO mice; Pyr-1 was ineffective at blocking the AR (Fig. 5D), strongly suggesting that the target of Pyr-1 was actually iPLA2β.
We have shown herein that iPLA2β and mGX inhibition, either through pharmacological tools or genetically modified mice, led to a decrease of the total P4-induced AR rate, demonstrating that these enzymes contribute to AR. In contrast, cPLA2α does not participate with P4-induced AR in mouse.
Because the AR spreads over a long time period, the measure of the final AR rate 30 min after addition of P4 only gives a partial understanding of the mechanisms involved. To obtain better insight, the level of AR was plotted as a function of time from the beginning of the capacitation (time −35 min) to the end of the P4-induced AR period (up to 50 min), with time 0 corresponding to the addition of P4 into the capacitation medium (Fig. 6). After P4 addition, the AR rate was monitored every 5 min. This graph highlights two points. First, the level of AR of mouse epididymal sperm after collection was 14.08 ± 0.45 (n = 24). This ~15% likely corresponds to either poorly matured sperm or sperm stored in the epididymis for too long. Second, the kinetics revealed the presence of two types of AR, the spontaneous AR (SAR), which occurs during the capacitation period, and the P4-induced AR. Importantly, the AR kinetics highlight different sperm subpopulation; some sperm cells accomplish their AR very early after addition of P4 (<5 min), whereas others require longer incubation time with P4 (20–30 min). At 2 μm P4 (Fig. 6, green triangles, n = 12), analysis of the kinetics of the P4-induced AR suggests the involvement of different actors for the different subpopulations, because a significant speeding up of slope in the curve is observed between 10 and 15 min after P4 introduction to the medium compared with slope measured between 0 and 10 min (Fig. 6, highlighted by a green arrow). Moreover, the plateaux observed at 2 or 10 μm P4 show that a subpopulation of sperm (~45 and 30%, respectively) was insensitive to this agonist. Interestingly, at 10 μm P4 the kinetic of the AR is different (Fig. 6, red circles, n = 12); the initial AR increase (0–5 min) is stronger (as indicated by the comparison of slopes measured at 2 μm P4 versus that measured at 10 μm P4). Moreover, no change in the slope was observed between 0 and 30 min, suggesting that AR signaling may be different.
SAR, which is not well characterized yet, occurs in media supporting or not capacitation in the mouse and corresponds to sperm undergoing AR in the absence of any external stimulus. Although increased SAR was associated with bigger litter size in CD46 KO mice (46), its physiological role remains elusive, because its rate greatly varies among species. We previously showed that during the first 45 min of capacitation, SAR is independent of mGX (24, 27). Nevertheless, another PLA2 could be involved during the earlier phase of the SAR. To test this hypothesis, sperm were incubated between −35 and 0 min with the following PLA2 inhibitors, BEL and FKGK18 for iPLA2β and LY329722 for mGX. Among the three different inhibitors, only those targeting iPLA2β were able to inhibit the SAR (Fig. 7A), showing that iPLA2β is activated during capacitation and is necessary to trigger the SAR. We also measured the rate of SAR in similar conditions (35 min of capacitation) of sperm from iPLA2β, mGX and cPLA2α KO mice (Fig. 7B). For mGX and cPLA2α KO sperm, the absence of these proteins did not prevent the occurrence of SAR, confirming that these enzymes are not involved in SAR. For iPLA2β, the level of AR at t = −35 min was already high in KO mice (20.6 ± 0.6%, n = 4) compared with WT littermates (14.3 ± 0.8%, n = 4), due to defective spermatogenesis (Fig. 1). Thus, experiments with iPLA2β KO mice were not informative to determine the role of this protein in SAR. Finally, the inhibition of SAR observed with sperm incubated with BEL is unlikely to be due to an absence of capacitation because protein phosphorylation on tyrosine residues was observed with BEL, albeit slightly weaker (Fig. 7C).
We next assessed when and for how long iPLA2β is required during AR induced by 2 μm P4. Time window of its involvement was determined by measuring the kinetics of AR in the presence of the iPLA2β inhibitors BEL or FKGK18. As expected, during the capacitation (−35 to 0 min), the SAR is blocked, and in these conditions, the rate of AR at t = 0 is lower than that of sperm incubated in control conditions (Fig. 8A). For P4-induced AR, we focused our attention on two phases, corresponding to the fast responding subpopulation (0–5 min) and to the late responding subpopulation (20–30 min). The slopes of the two phases were measured allowing us to characterize the rise of AR in both subpopulations (Fig. 8B). The early phase was clearly inhibited by iPLA2β inhibitors, as shown by the significant differences of the slopes between control and treated sperm, whereas the late phase was unchanged by BEL and FKGK18 (Fig. 8B). Because iPLA2β inhibitors also inhibited the SAR (Fig. 7A), we wondered whether the inhibition of AR rise observed with BEL or FKGK18 at 0–5 min might be related to the SAR inhibition induced by the inhibitor treatment. To address this question, a new experimental design was tested, with BEL introduced at the end of the capacitation period (Fig. 8C). Remarkably, whenever the iPLA2β inhibitor BEL was introduced, the response to 2 μm P4 was clearly delayed, and no increase was observed during the first 5 min (Fig. 8D), demonstrating that BEL inhibition is independent on the level of SAR. It is worth noting that the speeding up of the slope observed with sperm in control conditions between 10 and 15 min (Fig. 6, green arrow) was abolished with BEL and FKGK18, and slopes measured at 10–15 min dropped from 2.51 ± 0.19 (n = 12) in control condition to 1.4 ± 0.11 and 1.33 ± 0.13 for BEL and FKGK18, respectively (n = 3); the difference between slopes being statistically different (p = 0.015 and 0.013, respectively). This result suggests that iPLA2β contributes to the AR during the first ~15 min after P4 addition. Altogether, the observed final inhibition of ~30% AR with BEL is thus due to the delayed response and the absence of slope increase between 10 and 15 min. Altogether, these experiments show that iPLA2β is necessary for the early phase (0–5 min), contributes to the second phase (5–15 min), and is dispensable to the late phase (20–30 min) of AR induced at physiological P4 concentration.
At 10 μm P4, contrary to what was observed for 2 μm P4, the inhibition of the early phase in the presence of iPLA2β inhibitors was not observed (Fig. 9A), and the slopes of AR measured in control conditions or with BEL were not statistically different (Fig. 9B), suggesting that at this P4 concentration iPLA2β contributes slightly to the onset of the AR. In contrast to what was observed at low P4 concentration, the late phase (20–30 min) was significantly inhibited (Fig. 9B). Similar results were obtained with iPLA2β KO mice (Fig. 10). At 2 μm P4, the slope of the initial rise (0–5 min) was significantly reduced in comparison with that obtained with WT sperm (p < 0.001), whereas the slope of the late phase of AR was unchanged (Fig 10, A and B). At 10 μm P4, no difference was observed for 0–5 min, although a significant difference of the slope of AR increase was observed between WT and KO sperm during the late phase, confirming the results obtained with BEL and FKGK18. Importantly, results obtained with pharmacological tools or deficient animal show that the time window of the involvement of iPLA2β during AR is P4 concentration-dependent.
We next assessed when and for how long mGX is required during P4-induced AR. The time course of activation was determined by measuring AR kinetics in the presence of LY329722 and by comparing the kinetics of AR of mGX KO and WT sperm. Contrary to what was observed when iPLA2β is inhibited, the AR of the fast responding sperm cells was not modulated by LY329722 or when mGX was genetically altered at low P4 concentration (Fig. 11A), and no statistical differences were found between slopes of P4-induced rise of AR in the different conditions tested between 0 and 5 min (Fig. 11B), confirming that mGX is dispensable to the early phase of AR. In contrast, the AR of late responding sperm cells (20–30 min) was blocked by LY329722 for WT sperm and inhibited for mGX KO sperm (Fig. 11, A and B). Moreover, from 10 min, the rise of AR was significantly reduced. These results thus show that mGX is necessary for the intermediate and late phase of AR, from 10 min.
Similar results were obtained at 10 μm P4; the early phase of AR (0–5 min) was not inhibited by LY329722 or changed with sperm from mGX-deficient mice, whereas the late phase was strongly inhibited in both conditions (Fig. 11, C and D).
Importantly, the overall inhibition measured at 30 min was more pronounced at 2 μm than 10 μm, with 68 and 56% of inhibition, respectively (Fig. 11, A versus B), suggesting that different pathways are activated at different P4 concentrations. To address this hypothesis, we measured AR inhibition by LY329722 as a function of P4 concentration. For P4 concentrations below 1 μm, LY329722 inhibition was very strong, and the AR was almost completely abolished (Fig. 12). In contrast, the potency of LY329722 decreased when P4 concentration increased above 1 μm, strongly suggesting that an alternative sPLA2-independent AR pathway is activated at high concentrations of P4.
We have clearly shown above that iPLA2β and mGX are sequentially involved at low P4 concentration. iPLA2β is necessary for the fast responding subpopulation (0–5 min) and contributes to sustain AR up to ~15 min. In contrast, mGX is dispensable for the early phase but is necessary for sperm accomplishing their AR after 10 min. To confirm these time windows of involvement, both iPLA2β and mGX were inhibited at the same time, either by pharmacological tools or by a combination of inhibitor and KO mice; WT sperm samples were treated with both LY329722 and BEL inhibitors, and mGX KO sperm were treated with BEL and iPLA2β KO sperm with LY329722. In the three conditions, both the early (0–5 min) and late phase (20–30 min) of AR are inhibited (Fig. 13A), and significant differences between slopes of P4-induced rise of AR measured in control conditions and those measured with inhibitors and/or KO sperm were found at both early and late phases (Fig 13B).
The importance of phospholipase activity in the sperm AR was demonstrated several decades ago (1). However, the molecular identities of the implicated PLA2s were not determined. Here, we have challenged the involvement of three different PLA2s as follows: iPLA2β, cPLA2α, and group X sPLA2 in the mouse AR by using specific inhibitors and the corresponding knock-out mice. Concerning the specificity of the used inhibitors, it is worth noting that the specificity of both iPLA2β and sPLA2 inhibitors was validated by their lack of inhibition in the corresponding KO sperm. Moreover, LY329722 does not inhibit iPLA2β, becauseLY329722 has no effect on SAR, contrary to BEL and FKGK18, two iPLA2β inhibitors. Similarly, BEL does not inhibit recombinant mGX sPLA2 (data not shown). Collectively, these results validate our pharmacological study performed with these two compounds. Moreover, we showed that Pyr-1 was not specific at 1 μm and also inhibited iPLA2β. Altogether, from pharmacological data and study of the AR of sperm from three different PLA2 knock-out mouse strains, we demonstrated that iPLA2β and group X sPLA2 are involved in the P4-induced AR in mouse sperm. In contrast, AR measured on sperm from cPLA2α mice show that this PLA2 is involved neither in spontaneous nor induced AR. Although the presence of other PLA2s cannot be ruled out, the probability of their contribution in the AR is very low. Concerning sPLA2s, we have previously shown that only mGX is present in mouse mature sperm (24). Concerning the other members of the iPLA2 (γ δ, ϵ, ζ, and η) and cPLA2 (β, γ δ, ϵ, and ζ) groups, no expression was found in the testes (22), and it is thus unlikely that they are involved in the AR. Nevertheless, the presence of other phospholipases generating fusogenic lipids is not excluded. Indeed, PLA1 is present in sperm (47), and PLB and PLD1 are activated during the AR (48, 49), possibly explaining why some sperm were still able to perform the AR although iPLA2β and mGX were inhibited (Fig. 13).
Most of the studies performed on the molecular pathways involved in the AR focus either on the first minutes of the AR (50,–52) or on the final rate obtained after 20–30 min (20, 48, 53). Only a few studies presented kinetic data, mainly on a single cell basis, thus focusing on fast responding cells (54,–56). However, sperm population is very heterogeneous, and the presence of sperm subpopulations with different degrees of maturity after capacitation completion is well documented in human and mouse (57,–61). At the population level, AR is a long lasting event occurring over 30 min (Fig. 6A), and only kinetic studies can address the question whether the AR is an homogeneous event. For this study, we chose to follow the full release of the acrosomal vesicle as evidenced by Coomassie Blue staining as an indication of AR outcome. We showed that this method can discriminate between different phases of the AR, revealing very sensitive subpopulations completing their AR in less than 5 min, and others requiring up to 30 min. Here, we show for the first time that these subpopulations do not rely on the same molecular pathways to achieve the AR. At 2 μm P4, a concentration close to the physiological concentration inside the cumulus, we showed that inhibition of iPLA2β by BEL blocked the fast responding subpopulation (Figs. 8 and and1010A). For this fast responding subpopulation, mGX activity was not necessary because LY329722 alone had no effect and the initial rise of AR rate in mGX KO sperm was similar to that of WT sperm (Fig. 11A). Interestingly, between 10 and 15 min, the slope of the AR rise in the presence of BEL was lower than that measured in control condition (Fig. 8A), showing that iPLA2β inhibition partially blocked the AR during this interval and thus suggesting that early lipid hydrolysis by iPLA2β was responsible for the speeding up of AR increase observed between 10 and 15 min in control conditions (Fig. 6, green arrow). In contrast, mGX inhibition by LY329722 led to a strong blocking of the AR after 5 min (Fig. 10), demonstrating the key role of this enzyme in the intermediate and late phase.
We can thus propose that at this low physiological P4 concentration, sperm achieving their AR early rely only on iPLA2β, whereas sperm achieving their AR late rely mainly on mGX sPLA2, and in between these two phases, both PLA2s are important.
Remarkably, the potencies of PLA2 inhibitors on the AR were dependent on P4 concentration. At low P4 concentration, the initial phase (0–5 min) was strongly dependent on the iPLA2β activity (Fig. 8). At high P4 concentration, inhibition of iPLA2β (Fig. 9) or its absence (Fig. 10B) only slightly decreased the level of AR. Similarly, mGX inhibition produced a near-complete inhibition (>95%) of the P4-induced AR at [P4] <1 μm and a weaker inhibition at 10 μm P4 (~50%) (Fig. 12). The AR induced by low P4 concentrations is thus strongly PLA2-dependent. Moreover, these results suggest that there are alternative pathways allowing a bypass of the lack of fusogenic lipids produced by iPLA2β or mGX at high P4 concentration. Ca2+ plays a central role in the AR, and Kobori et al. (62) showed that low or high P4 concentrations activate different Ca2+ responses; for [P4] <5 μm sperm exhibiting only transient calcium increase contrary to [P4] >10 μm, where both transient and long lasting calcium increases were observed. We can thus postulate that Ca2+ signaling triggered by low [P4] is able to initiate the AR in a primed sperm subpopulation only in conjunction with the production of fusogenic lipids by iPLA2β. At high [P4], AR onset relies on a likely stronger Ca2+ signaling that can trigger the AR without fusogenic lipids or with a different set of fusogenic lipids, which could be produced by PLA1 (47), PLB (48), or PLD1 (49). Alternatively, P4 may bind to a different receptor at high concentrations (63) and activate a different pathway.
It is worth noting that at low P4 concentrations (< 1 μm), inhibition of mGX sPLA2 by LY329722 led to an almost complete inhibition of the AR (Fig. 12). This result suggests that the expected synergy between P4-induced Ca2+ increase and fusogenic lipids produced by iPLA2β is not sufficient to trigger a full AR, but rather produce transitional states of acrosomal exocytosis, allowing only transient opening of pores. These pores can open and close before the final loss of the acrosome as suggested previously (64). During this transient period, sperm visually appear to be “acrosome-intact” sperm. The inhibition of mGX, located in the sperm acrosome (24), likely through these pores, may prevent synergistic effects of fusogenic lipids produced by both PLA2s leading to their complete opening and thus pores return to a closed state, leading to a blocking of the AR.
Importantly, we showed that physiological P4 concentration is closer to 1 than 10 μm (Fig. 2A), and our results clearly indicate that under physiological conditions both iPLA2β and group X sPLA2 are key players of the mouse sperm AR. Overall, our results indicate that different molecular pathways are activated by different P4 concentrations during the AR and thus that [P4] is a key factor, which should be taken into account in P4-induced AR studies.
We showed that the SAR occurring during the first 35 min of the capacitation period was inhibited only by iPLA2β inhibitors and not by a pan sPLA2 inhibitor (Fig. 7B), demonstrating that iPLA2β is required during capacitation.
Remarkably, comparison of the final rate of the AR at 30 min between the sperm treated before (Fig. 8A) or at the end of the capacitation (Fig. 8C) shows no statistical difference (39.0 ± 0.57 versus 39.6 ± 0.33, n = 3; p = 0.8), although the level of AR at t = 0 is different, due to SAR inhibition when iPLA2β is inhibited during capacitation (Fig. 7A). This result indicates that inhibition of iPLA2β during capacitation increased the slope of P4-induced AR rise, mean slopes being 0.73 and 1.03 (measured between 5 and 30 min after P4), whereas BEL is introduced at the beginning or at the end of the capacitation, respectively. This result thus suggests that the subpopulation of sperm, which should have achieved their AR during capacitation, was primed and waiting for a cofactor to achieve the AR, which may be Ca2+ influx triggered by P4.
In conclusion, the progesterone-induced AR is a long lasting event, spreading over 30 min in the mouse, and by performing kinetic analyses, we revealed new insight concerning the AR and showed the presence of different sperm subpopulations that do not rely on the same molecular pathways to achieve their AR. At low physiological P4 concentration, sperm achieving their AR early (0–5 min post progesterone) rely only on iPLA2β, whereas sperm achieving their AR late (20–30 min post progesterone) rely on group X sPLA2. We thus showed that the AR is not a homogeneous molecular process and instead changes over time. Moreover, we showed that PLA2 involvement depends upon P4 concentration, PLA2s being key actors at low P4 concentrations close to the physiological concentration (<2 μm) and less central at higher P4 concentrations above 10 μm.
C. A., S. H., and G. L. conceived and coordinated the study and wrote the paper. R. A. N., S. Y., J. P. H., J. E., and G. M. performed and analyzed AR studies. P. F. R. and T. K. designed primers and performed genotyping of KO mice. S. B. designed, performed, and analyzed the experiments shown in Fig. 1A. J. T. provided the iPLA2β KO mice. G. K. provided FKGK18. P. F. R. revised the article critically. All authors reviewed the results and approved the final version of the manuscript.
We thank Lexicon Genetics Inc. for providing mGX-deficient mice and Miriam Kolko and Charlotte Taul Braendstrup (University of Copenhagen, Denmark) for iPLA2β-deficient mice. We thank Prof. Michael Gelb (University of Washington) for providing LY329722 and pyrolline-1. We thank Marie Cristou-Kent for English correction.
*This work was supported by CNRS (to C. A. and G. L.), Agence Nationale de la Recherche Grant ANR-10-EMMA-042. The authors declare that they have no conflicts of interest with the contents of this article.
3The abbreviations used are: