Differential role of bovine serum albumin and HCO₃⁻ in the regulation of GSK3 alpha during mouse sperm capacitation

Abstract

To become fertile, mammalian sperm are required to undergo capacitation in the female tract or in vitro in defined media containing ions (e.g. ${\text{HCO}}_3^{\hspace{0.5em}-}$, Ca²⁺, Na⁺, and Cl⁻), energy sources (e.g. glucose, pyruvate) and serum albumin (e.g. bovine serum albumin (BSA)). These different molecules initiate sequential and concomitant signaling pathways, leading to capacitation. Physiologically, capacitation induces changes in the sperm motility pattern (e.g. hyperactivation) and prepares sperm for the acrosomal reaction (AR), two events required for fertilization. Molecularly, ${\text{HCO}}_3^{\hspace{0.5em}-}$ activates the atypical adenylyl cyclase Adcy10 (aka sAC), increasing cAMP and downstream cAMP-dependent pathways. BSA, on the other hand, induces sperm cholesterol release as well as other signaling pathways. How these signaling events, occurring in different sperm compartments and with different kinetics, coordinate among themselves is not well established. Regarding the AR, recent work has proposed a role for glycogen synthase kinases (GSK3α and GSK3β). GSK3α and GSK3β are inactivated by phosphorylation of residues Ser21 and Ser9, respectively, in their N-terminal domain. Here, we present evidence that GSK3α (but not GSK3β) is present in the anterior head and that it is regulated during capacitation. Interestingly, BSA and ${\text{HCO}}_3^{\hspace{0.5em}-}$ regulate GSK3α in opposite directions. While BSA induces a fast GSK3α Ser21 phosphorylation, ${\text{HCO}}_3^{\hspace{0.5em}-}$ and cAMP-dependent pathways dephosphorylate this residue. We also show that the HCO₃⁻-induced Ser21 dephosphorylation is mediated by hyperpolarization of the sperm plasma membrane potential (Em) and by intracellular pH alkalinization. Previous reports indicate that GSK3 kinases mediate the progesterone-induced AR. Here, we show that GSK3 inhibition also blocks the Ca²⁺ ionophore ionomycin-induced AR, suggesting a role for GSK3 kinases downstream of the increase in intracellular Ca²⁺ needed for this exocytotic event. Altogether, our data indicate a temporal and biphasic GSK3α regulation with opposite actions of BSA and ${\text{HCO}}_3^{\hspace{0.5em}-}$. Our results also suggest that this regulation is needed to orchestrate the AR during sperm capacitation.

Introduction

Mammalian sperm acquire the ability to fertilize in the female tract in a process known as capacitation. This process can be mimicked in vitro with an absolute requirement for ${\text{HCO}}_3^{\hspace{0.5em}-}$, Ca²⁺, energy nutrients, and a protein source, which is usually bovine serum albumin (BSA). Capacitation induces changes in the motility pattern known as hyperactivation and prepares the sperm for the acrosome reaction, a physiologically induced exocytosis. At the molecular level, capacitation is associated with an HCO₃⁻-dependent increase in cAMP synthesis by the atypical adenylyl cyclase Adcy10 (aka sAC), and the consequent activation of protein kinase A (PKA). These changes are followed by a series of signaling events that include increases in tyrosine phosphorylation, intracellular Ca²⁺ [(Ca²⁺)]_i, intracellular pH (pH_i), and hyperpolarization of the sperm plasma membrane (Em). Pharmacologic and genetic loss-of-function approaches have revealed that these capacitation-associated signaling events are essential for the sperm to become fertile (Gervasi and Visconti, 2016). BSA, on the other hand, has been shown to be involved in plasma membrane rearrangements by inducing cholesterol release (Go and Wolf, 1985; Tesařík and Fléchon, 1986; Osheroff et al., 1999; Visconti et al., 1999a,b; Travis and Kopf, 2002).

The complexity of capacitation at the molecular level is underscored by differences in kinetics as well as by the subcellular localization of the different signaling processes. As an example of temporal differences, although PKA is necessary and sufficient to induce tyrosine phosphorylation, contrary to the fast PKA activation (∼1 min incubation), tyrosine phosphorylation is a slow process starting at ∼45 min of incubation in capacitation media in mouse sperm (Visconti, 2009). Similarly, Em hyperpolarization follows different kinetics, starting at ∼15 min of sperm incubation in capacitating media (Demarco et al., 2003). Regarding localization, PKA activation is required for the preparation of the acrosome reaction in the sperm head (Visconti et al., 1995); however, immunofluorescence and PKA activity assays indicate that PKA catalytic subunit localizes exclusively to the sperm flagellum (Wertheimer et al., 2013). Therefore, the molecular connection between PKA activation and proteins involved in the acrosome reaction is not straightforward. One possibility is that events starting in the tail can signal processes in other sperm compartments. In this regard, Em changes are transmitted almost instantaneously throughout the cell and would be able to transfer information between different sperm compartments. Consistent with this hypothesis, the sperm-specific K⁺ channel Kcnu1 (aka SLO3) (Santi et al., 2010), the main channel responsible for Em changes during capacitation (Chávez et al., 2013), localizes to the flagellar principal piece. Despite SLO3 localization on the flagellum, sperm from SLO3 KO mice models are unable to undergo the acrosome reaction even when challenged with Ca²⁺ ionophores (Santi et al., 2010). This result suggests that there is information transfer between the tail and the head.

Evidence collected by our group (De La Vega-Beltran et al., 2012) and others (Balestrini et al., 2021) suggests that the capacitation-induced SLO3 activation together with the consequent Em hyperpolarization are necessary steps to prepare the sperm for the acrosome reaction. However, the signaling events connecting Em hyperpolarization with the acrosome reaction are not well understood. Recently, Dr Nixon’s group proposed that activation of glycogen synthase kinase 3 (GSK3) is necessary for an agonist-induced acrosome reaction and that its action is mediated by dynamin phosphorylation (Reid et al., 2015). GSK3 is encoded by two different genes, GSK3α and GSK3β (Park et al., 2023); in sperm, both isoforms are present (Choi et al., 2016) but only GSK3α is essential for fertilization (Maurin et al., 2013; Bhattacharjee et al., 2018). GSK3α and GSK3β are negatively regulated by phosphorylation of residues Ser21 and Ser9, respectively, in their N-terminal domain (Dey et al., 2020). Both GSK3 isoforms were shown to localize in the tail as well as in the head of mature sperm; however, GSK3α appears to be the predominant isoform in the head.

Here we present evidence that GSK3α-inactivating phosphorylation in Ser21 is differentially regulated during capacitation. On one hand, GSK3α Ser21 phosphorylation is up-regulated by a BSA-dependent pathway in minutes. On the other hand, activation of cAMP-dependent pathways induces GSK3α Ser21 dephosphorylation, an action mediated by Em hyperpolarization and by the increase in pH_i. Independent actions of BSA and the ${\text{HCO}}_3^{\hspace{0.5em}-}$/cAMP pathway have been reported (Visconti et al., 1995; Xia and Ren, 2009b). Overall, these data are the first to show antagonistic roles of these compounds during capacitation and suggest a possible mechanism by which Em hyperpolarization started in the flagellum can modulate processes in the head related to the acrosome reaction.

Materials and methods

Materials

Chemicals and other laboratory reagents were purchased from the following companies: Bovine Serum Albumin (BSA, fatty acid-free, cat# A0281, Sigma-Aldrich, St Louis, MO, USA) for media; Trizma base (cat# T-1503, Sigma-Aldrich); sodium chloride (NaCl, cat# S1679, Sigma-Aldrich), magnesium chloride (MgCl₂, cat# M0250, Sigma-Aldrich); potassium chloride (KCl, cat# P5405, Sigma-Aldrich); β-mercaptoethanol (βME, cat# 161-0710, Bio-Rad Laboratories, Hercules, CA, USA); sodium bicarbonate (NaHCO₃, cat# S5761, Sigma-Aldrich); and lithium chloride (LiCl, cat# L-9650, Sigma-Aldrich). Further, we used valinomycin (cat# V0627, Sigma-Aldrich); N⁶,2′-O-dibutyryladenosine 3′,5′-cyclic monophosphate sodium salt (dbcAMP, cat# D0627, Sigma-Aldrich) and 3-isobutyl-1-methylxanthine (IBMX, cat# I5879, Sigma–Aldrich); N-2-hydroxyethylpiperazine-N′2-ethanesulfonic acid (HEPES, cat# BP310, Fisher Scientific, Hampton, NH, USA); glycerol (cat# G33-500, Fisher Scientific); sodium dodecyl sulfate (SDS, cat#161-0302, Bio-Rad Laboratories); anti-GSK3 alpha monoclonal antibody (GSK3α cat# 4337S, Cell Signaling Technology, Danvers, MA, USA); anti-GSK3 beta monoclonal antibody (GSK3β cat# 9832S, Cell Signaling Technology); anti-phospho-GSK3 alpha/beta antibody (p-GSK3α/β, cat# 9331S, Cell Signaling Technology); anti-phospho-PKA substrate monoclonal antibody (pPKAs, cat# 9624S, Cell Signaling Technology); anti-alpha tubulin polyclonal antibody (cat# 2144S, Cell Signaling Technology); anti-pY antibody (clone 4G10) (cat# 05-321, Millipore, Burlington, MA, USA); GSK3 alpha/beta antibody (GSK3α/β,cat# PA5-99531, Thermo Fisher, Waltham, MA, USA); horse-radish peroxidase-conjugated anti-mouse immunoglobulin G (cat# 115-035-174, Jackson Immuno Research Laboratories, West Grove, PA, USA), horse-radish peroxidase-conjugated anti-rabbit immunoglobulin G (cat# 211-032-171, Jackson Immuno Research Laboratories). Small molecule inhibitors 6-[[2-[[4-(2,4-dichlorophenyl)-5-(5-methyl-1H-imidazol-2-yl)-2-pyrimidinyl]amino]ethyl]amino]-3-pyridinecarbonitrile (CHIR99021, cat# 4423; Tocris, Mineapolis, MN, USA) while TDI-10229, a sAC inhibitor, was from the laboratory of Jochen Buck and Lonny Levine (Weill Cornell Medicine, NY, USA). Finally, enhanced chemiluminescence (ECL) prime western blotting detection reagents (cat# RPN2232 GE Healthcare, Pittsburgh, PA, USA). A summary of the antibodies and reagents used in this study with the respective concentrations employed is presented in Supplementary Tables S1 and S2.

Animals and ethics approval

The present study was conducted on male mice (CD-1 retired breeders) obtained from Charles River Laboratories (Wilmington, MA, USA) aged 8–12 weeks. The study was performed following the specific guidelines laid by the Office of Laboratory Animal Welfare and approved by the Institutional Animal Use and Care Committee (IACUC), University of Massachusetts-Amherst (UMass IACUC Protocol # 2019-0008).

Transfer of epididymis and sperm preparation from SLO3 KO mice

Epididymides from SLO3 knockout (KO) mice were obtained from Dr Celia Santi’s laboratory (Washington University, St Louis, MO, USA) in the following way: whole epididymides were dissected out containing caput, corpus, and cauda regions and were immediately immersed in 1.5 ml Eppendorf tubes containing mineral oil. Each tube contained a maximum of two epididymides. The tubes were then sealed with parafilm to prevent any leakage, stored at 4°C and shipped at this temperature using ice packaged boxes. Upon receiving the epididymides, excess mineral oil was removed by placing the epididymides in soft tissue wipes and the caudal epididymal region was dissected out and treated as described below.

Culture media

The non-capacitating (NC) medium used in the current study consists of HEPES-based modified Toyoda Yokoyama-Hosi (m-TYH) medium consisting of 119.37 mM NaCl, 4.78 mM KCl, 1.19 mM KH₂PO4, 1.19 mM MgSO₄, 5.56 mM glucose, 1.71 mM CaCl₂, 20 mM HEPES, 0.51 mM Na-pyruvate. The media thus prepared was maintained at a pH of 7.2 with NaOH. Other media used in this work were prepared by supplementing the NC media with either 15 mM NaHCO₃ (HCO₃⁻-TYH media), 5 mg/ml of BSA (BSA-TYH media), or both compounds. In all cases, pH was maintained at 7.2.

Sperm preparation

Caudal epididymal spermatozoa were obtained by swim-out. Briefly, CD1 male mice were euthanized, both cauda epididymides dissected, fat tissue removed, and three incisions made with a sharp razor blade. The epididymides were then placed in 1 ml of NC media, pre-heated at 37°C, and spermatozoa allowed to swim-out for 10 min before removing the remaining epididymal tissue from the suspension. At this point, sperm concentration varies between 10 and 20 million per ml; after counting, the suspension was adjusted to 10 million/ml. Immediately after, 100 µl of the original suspension in NC media were added to the respective experimental media conditions (NC, HCO₃⁻-TYH, BSA-TYH, or CAP) containing or not inhibitors and/or activators of different pathways at the concentrations depicted for each particular experiment. To study the antagonistic effect of BSA and other pathways, sperm aliquots were added to BSA media and incubated for 5 min before adding other molecules such as ${\text{HCO}}_3^{\hspace{0.5em}-}$, cAMP agonists, NH₄Cl, or valinomycin; this condition was denoted as BSA5′+${\text{HCO}}_3^{\hspace{0.5em}-}$. Incubations in the respective media, with the addition or not of signaling pathways inhibitors or activators, were conducted at 37°C, pH 7.2 for different periods.

Caput sperm were obtained as described previously (Tourzani et al., 2018, 2021). Briefly, caput epididymides were dissected and squeezed in TYH NC media to allow immotile sperm to leave the tissue. Sperm were then purified using a Percoll gradient. Sperm were then incubated in either NC or CAP media as described above.

Sodium dodecyl sulphate–polyacrylamide gel electrophoresis and immunoblotting

Caput or cauda sperm obtained as described above were incubated in different experimental conditions. Four hundred microliters containing 10⁶ spermatozoa incubated under different conditions and for different time periods were centrifuged at 12 100 ×g for 5 min at room temperature. Then, sperm pellets were washed twice with 600µl PBS, supernatant removed, and the final sperm pellets extracted for sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and western blots. In some cases, the final sperm pellets were stored at −80°C until further use. Sperm pellets were resuspended in non-reducing sample buffer (2% SDS, Tris–HCl pH 7.0) (Visconti et al., 1996; Mohanty et al., 2023), heated in boiling water for 5 min and centrifuged for 5 min at 12 100 ×g. Supernatants were then supplemented with β-mercaptoethanol to a final 5% (v/v) concentration, boiled for 5 additional min and centrifuged once more at 12 100 ×g. The extracted proteins were then loaded onto a 10-well 8% gel to undergo SDS-PAGE using a system from Bio-Rad at constant amperage of 40 mA per gel. Following electrophoresis, proteins in the gel were transferred to PVDF (cat# IVPH00010, Millipore Sigma, MA, USA) membranes as described previously (Towbin et al., 1979). Transfers were conducted in Towbin buffer (25 mM Tris, 192 mM glycine, pH 8.3 supplemented with 20% (v/v) methanol) (Towbin et al., 1979) at 250 mV of constant voltage for 2 h on ice. PVDF membranes were then blocked for 1 h at room temperature in 5% milk in TBS-T buffer (100 mM Tris–HCl pH 7.5, 150 mM NaCl, 0.01% Tween-20) and then used for western blotting. The source, type, dilution, and buffer used for different antibodies are detailed above and in Supplementary Table S1. Briefly, blots were incubated at 4°C overnight with different antibodies diluted as indicated in Supplementary Table S1. The following day, blots were washed for 10 min with 10 ml of TBS-T three times, and then incubated for 1 h with the corresponding secondary HRP conjugated anti-mouse or anti-rabbit antibody (Jackson Immunoresearch Laboratories) at a dilution of 1:5000 for the respective GSK3 antibodies and, at 1:10 000 dilution for anti-phospho-PKA substrates (anti-pPKAs) and anti-phospho-tyrosine (anti-pY) antibodies prepared in TBS-T. Secondary antibodies were then washed three times in 15 ml of TBS-T buffer (5 min each wash); blots were then developed with chemiluminescence reagents using ECL prime (GE Healthcare, Pittsburg, PA, USA). After developing, in most experiments, antibodies were then removed/stripped as described below. Blots were then blocked once more either with milk, BSA or gelatin (Supplementary Table S1) and used with other antibodies. In general, this procedure was used for loading control, either with anti-GSK3α, anti-GSK3β, anti-GSK3α/β, or anti-tubulin.

Immunolocalization

For immunolocalization, aliquots of spermatozoa, incubated under conditions that support or not capacitation, were centrifuged at low speed (800 ×g) to maintain sperm integrity for 5 min, washed once with PBS and fixed for 10 min at room temperature with 4% paraformaldehyde (EMS, Hatfield, MA, USA). Samples were then centrifuged at 1000 ×g for 5 min followed by washing with PBS (1×) prior to mounting onto poly-l-lysine coated glass slides and air-dried. Membrane permeabilization was achieved with 0.5% Triton X-100 prepared in 1× PBS for 5 min at room temperature followed by washing three times in 1× PBS each for 5 min. The cells were then blocked with 3% BSA prepared in PBS-Tween 20 (0.01%) for 1 h at room temperature. Spermatozoa were then incubated with either anti-pGSK3α, anti-GSK3β, or anti-pGSK3α/β at 1:100 dilution prepared in 5% BSA in PBS-T and kept overnight in a humidifier chamber at 4°C. Spermatozoa were then washed with PBS-T the following day and incubated for 1 h at room temperature with the corresponding Alexa Fluor^TM 555-conjugated secondary anti-rabbit or anti-mouse antibody at 1:300 dilution in 5% BSA in PBS-T. For acrosome staining, the secondary antibody solution also contained Alexa Fluor^TM 568-conjugated peanut agglutinin (PNA) (1:100). Following that, samples were then washed with PBS-T three times (5 min each) and mounted on slides using vectashield added (Vector Laboratories, Newark, NY, USA) as mounting medium. For regular epifluorescence, image capture was performed using a 60× objective (Nikon, Melville, NY, USA; PlanApo, NA 1.49) in an inverted fluorescence microscope (Nikon Eclipse T300) coupled with a CMOS camera (Zyla 4.2, Andor Technology, Belfast, UK). For dual-color super-resolution structured illumination microscopy (SIM), images were captured with a 100× objective (Nikon, PlanApo, NA 1.49) in an A1R-SIMe microscope. Super-resolved images reconstruction was performed using NIS elements software (Nikon).

Acrosome reaction

To measure Ca²⁺ and the acrosome reaction in single cells, alternate measurements were conducted as described previously (Sánchez-Cárdenas et al., 2021). Briefly, mouse sperm were incubated with 2 µM Fluo 4 AM and 0.05% Pluronic acid for 40 min in NC medium and once loaded, cells were capacitated in media containing BSA and ${\text{HCO}}_3^{\hspace{0.5em}-}$. For recording, sperm were immobilized on mouse laminin (1 mg/ml)-coated cover slips and the chamber was filled with NC medium. Once attached, sperm were incubated in the recording chamber with 5 μM FM4-64, which was maintained in the medium during the entire recording. For imaging experiments, we used a chamber recording model MS-508D (ALA Scientific) at 37°C controlled by the heater controller model 202A (Harvard Apparatus, Holliston, MA, USA). Sperm were recorded with a microscope (Nikon Eclipse Ti-U) and observed with an oil immersion fluorescence objective (Nikon plan Apo TIRF DIC H/N2 60× /1.45 NA). As the light source, a Nikon Intensilight-CHGF1 was used and for Fluo 4 and FM4-64 detection, the fluorescence cubes used were: GFP 96343, D: 495, Exc: 470/40, barrier 525/50 (Nikon) and Wide Green 11007v2, D: 565 dcxt, excitation 535/50, emission 590 lpv2 (Chroma Technology Corporation, Bellows Falls, VT, USA), respectively. Images were acquired with an Andor Ixon 3 EMCCD camera model Du-8970-C00#B (Andor Technology) under protocols written in Andor iQ 1.10.2 software version 4.0. Fluorescence images of both dyes were acquired every 20 s with 100 and 20 ms exposure/illumination for Fluo 4 and FM4-64, respectively, for a period of 30 min. Regions of interest (ROIs) were drawn on the head of individual spermatozoon in the movie for quantification and a plot of the fluorescence intensity changes versus time was obtained in Origin 6.0 (Microcal Software, Piscataway, NJ, USA). Fluorescence is expressed as (F − F₀)/F₀. When brightness and contrast were adjusted, this was carried out equally in all images or movies taken under the same conditions.

Statistical analysis

Statistical analysis was performed with the aid of GraphPad Prism software (San Diego, CA, USA; www.graphpad.com). Briefly, the respective bands or sections in the blot were scanned and images analyzed using ImageJ software (National Institutes of Health, Bethesda, MD, USA). Then, the pixel ratio between pGSK3 and total GSK3α or (Fig. 1) the ratio between total GSK3α, β, or α/β and tubulin was obtained. In each case, ratios were obtained from the same blot which was stripped and re-probed up to three times as described above. In most cases, the obtained ratios were normalized against the NC condition ratio. In some cases, normalization was against the CAP condition. The most relevant observations obtained at 1 h were repeated at least 10 times with different mice; time curves and concentration curves were repeated at least three times. For comparisons, data were analyzed for normal distribution and homogeneity of variances. For simple comparisons between control and groups, data were subjected to parametric Student’s t-tests, while one-way ANOVA was performed for comparison between multiple groups. For data showing statistical differences between groups multiple comparisons were performed using Tukey’s test. Data were represented as the average ± SEM of the normalized values (*P < 0.05; **P < 0.01; ***P < 0.001). Statistical significance was considered a value of P < 0.05.

Figure 1.

GSK3α and β are present in mouse sperm and their quantity does not change during capacitation. Cauda epididymal sperm were isolated and incubated in conditions that support capacitation (CAP) or not (NC) as described in the Materials and methods. The respective sperm extracts were analyzed by western blots using either anti-GSK3α/β (A), anti-GSK3α (D), anti-GSK3β (F), or anti-α-tubulin (H) antibodies. Quantification for each antibody (B and C for GSK3α and GSK3α in blot A, respectively; E for blot D; and G for blot F) was conducted as described in the Materials and methods; briefly, blots were scanned and analyzed using ImageJ densitometry. Ratios between the respective GSK3 and tubulin pixels from the same blot were obtained and normalized to the respective NC ratio in each independent blot. A Student’s t-test was performed. No significant differences were observed (n = 3). CAP, capacitating media; NC, non-capacitating media; GSK3, glycogen synthase kinase 3.

Results

GSK3α and β are present in mature sperm

To confirm the presence of GSK3α and β in mouse sperm, to evaluate the extent to which their quantities changed during capacitation and to validate the antibodies used during this work, we first conducted western blots. These blots involved sperm incubated in conditions that support (CAP) or not (NC) capacitation using different anti-GSK3 antibodies and anti-tubulin as loading controls (Fig. 1). These antibodies recognized either GSK3α and β (Fig. 1A–C), only GSK3α (Fig. 1 D and E), only GSK3β (Fig. 1F and G), or α-tubulin (Fig. 1H). These western blots were run sequentially, starting with anti-GSK3α/β and finishing with anti-α-tubulin, with a stripping step between detections. For quantification, blots were scanned, analyzed using ImageJ software, and the ratio between the respective GSK3 signal and tubulin normalized to NC values as described in the Materials and methods (Fig. 1B, C, E, and G). No changes in GSK3α and β quantities were observed between NC and CAP conditions.

The specific antibodies, recognizing exclusively one isoform, were then used to localize GSK3α and GSK3β in sperm by immunofluorescence. These experiments showed differential localization for both kinases; GSK3α is primarily present in the sperm head; and with less intensity it is also observed decorating the whole flagellum (Fig. 2A and Supplementary Fig. S1A). On the other hand, GSK3β localizes to the mid-piece and the sperm neck with some signal also observed in the principal piece (Fig. 2B and Supplementary Fig. S1B). The GSK3α signal does not seem to overlap with PNA staining, in a position closer to the sperm nucleus. To further analyze this localization, we conducted SIM to enhance resolution (Supplementary Fig. S2A). Using SIM, the GSK3α and the PNA signal were clearly separated, with GSK3α immediately adjacent to PNA in the anterior head; altogether, these results suggest its localization to the inner acrosomal membrane or its proximity.

Figure 2.

GSK3α and GSK3β localize to different mouse sperm compartments. Cauda epididymal sperm were suspended in non-capacitating media and immediately processed for immunofluorescence as described in the Materials and methods. Immunofluorescence was conducted using specific anti-GSK3α (A) and anti-GSK3β (B) antibodies. In each case, preparations were also stained with Alexa 568-labeled PNA to visualize the sperm acrosome. Left panels represent DIC images and the white bar corresponds to 12 micrometers. GSK3, glycogen synthase kinase 3; DIC, differential interference contrast; PNA, peanut agglutinin

GSK3α Ser21 phosphorylation is differentially regulated during capacitation

Both GSK3α and GSK3β require dephosphorylation of Ser 21 and Ser 9, respectively, to become active (Rayasam et al., 2009). We used anti-phospho-Ser9/21 GSK3 antibodies, which recognize both GSK3α and GSK3β N-terminal phosphorylation, to follow the phosphorylation state of these serine residues in sperm incubated in conditions that support capacitation (CAP) or not (NC) (Fig. 3A and B). In cauda mature sperm, only phosphorylated GSK3α Ser21 was detected. To discard the possibility that this result was due to lack of specificity of the anti-phospho-antibody toward phospho-Ser9 in GSK3β, a similar experiment was conducted using caput sperm. In caput sperm, both GSK3α and GSK3β were found phosphorylated in their N-terminal Ser residues (Supplementary Fig. S3A–C). In addition, we obtained similar results in cauda and caput sperm using specific anti-phospho-Ser21 GSK3α (Supplementary Fig. S3D and E) and, specific anti-phospho-Ser9 GSK3β (Supplementary Fig. S3F and G) antibodies. Overall, these results indicate that GSK3β is completely dephosphorylated in sperm obtained from the cauda epididymis (Fig. 3A). We focused the remaining experiments on the regulation of GSK3α Ser21 phosphorylation in mature cauda sperm. In CAP conditions, GSK3α becomes dephosphorylated in a time-dependent manner (Fig. 3C). Quantification of western blot independent replicates (Fig. 3D) indicated that GSK3α Ser21 phosphorylation follows a biphasic kinetic response (note that in Fig. 3C and D, the NC condition is on the right). While in the first few minutes, Ser21 is significantly phosphorylated, at later times (between 15 and 30 min), phosphorylation is reduced to lower levels than the NC control. Consistent with the localization of GSK3α, immunofluorescence with anti-phospho-Ser21 antibodies show staining in the same sperm head region as GSK3α as well as in the tail (Fig. 3E and Supplementary Figs S2B and S4A). Once capacitated, while the fluorescent signal is maintained in the tail, it is no longer observed in the anterior head (Fig. 3F and Supplementary Fig. S4B).

Figure 3.

GSK3α Ser21 phosphorylation follows biphasic kinetics during mouse sperm capacitation. Sperm from cauda epididymis were incubated in conditions that support capacitation (CAP) or not (NC). (A) Sperm extracts were analyzed by western blot using anti-phospho-Ser21/Ser9 antibodies; after developing, the blots were stripped and re-probed with anti-GSK3α antibodies that recognized exclusively GSK3α. (B) Densitometric analysis of bands with ImageJ was carried out as described in the Materials and methods by performing ratios of phospho versus total GSK3α and normalizing the results to the respective NC ratio. Data are expressed as average ratio intensity ± SEM. A Student’s t-test was performed with P < 0.05 considered significant denoted as (*) (n = 3). (C) Western blots were conducted on extracts from sperm incubated for different times in conditions that support capacitation or in NC media for 60 min (notice that the NC condition is at the right of the blot). Once developed with anti-phospho-GSK3α Ser21/GSK3β phospho-Ser9 antibodies, blots were stripped and re-probed with anti-GSK3α antibody. (D) Quantification was conducted as in (C), normalizing against the NC ratio between phospho and total GSK3α. (E and F) Anti-phospho-GSK3α Ser21/phospho-GSK3β Ser9 antibodies were used for immunofluorescence in either sperm incubated for 60 min in NC (E) or CAP (F) conditions as described in the Materials and methods. In all cases, sperm were co-stained with Alexa 568-conjugated PNA. Data in (B) and (D) represent average ratios ± SEM. A Student’s t-test was performed and significance level denoted as: *P < 0.05; **P < 0.01; ***P < 0.001 (n = 3). CAP, capacitating media; NC, non-capacitating media; PNA, peanut agglutinin; GSK3, glycogen synthase kinase 3. The whole blots for the cropped blots presented here are shown in Supplementary Fig. S6.

The biphasic Ser21 phosphorylation kinetics response suggests involvement of different signaling pathways playing opposite roles in the regulation of GSK3α. The main difference between the NC and CAP conditions in our experiments is the presence of BSA and ${\text{HCO}}_3^{\hspace{0.5em}-}$ in the incubation media. Therefore, to dissect the action of these molecules, we followed Ser21 phosphorylation in sperm incubated in the presence of either BSA without ${\text{HCO}}_3^{\hspace{0.5em}-}$ (Fig. 4A and B) or ${\text{HCO}}_3^{\hspace{0.5em}-}$ without BSA (Fig. 4C and D). Each compound had the opposite effect on GSK3 Ser21 phosphorylation; while BSA increased significantly Ser21 phosphorylation immediately (∼1 min) and maintained Ser21 high phosphorylation for at least 60 min (Fig. 4A and B), ${\text{HCO}}_3^{\hspace{0.5em}-}$ incubation decreased Ser21 phosphorylation following relatively slow kinetics (Fig. 4C and D). These data are consistent with the biphasic kinetics observed during capacitation above (Fig. 3C and D). The BSA-induced GSK3α Ser21 phosphorylation was not mimicked by beta cyclodextrin analogs (Fig. 4E and F), suggesting that the increase in Ser21 phosphorylation is not mediated by cholesterol release. In Fig. 4E, it is possible to observe a higher MW band which corresponds to nonspecific staining of excess BSA. Although sperm are thoroughly washed by two centrifugation steps, the amount of BSA in the media is high (5 mg/ml) and sometimes was not completely eliminated. As a control, we confirmed the BSA nonspecific staining using a BSA concentration curve in media not containing spermatozoa (Supplementary Fig. S3H). When sperm are incubated for 5 min with BSA alone in the absence of ${\text{HCO}}_3^{\hspace{0.5em}-}$ to allow Ser21 phosphorylation to increase and, only then, ${\text{HCO}}_3^{\hspace{0.5em}-}$ is added to the sperm suspension, ${\text{HCO}}_3^{\hspace{0.5em}-}$ reduces phosphorylation in about the same period as observed in full capacitating media (∼15 min) (Fig. 4G and H). Panels B, D, F, and H represent quantification of the results after normalization of the ratio between the phospho-GSK3α Ser21 signal and total GSK3α for each independent replicate.

Figure 4.

GSK3α Ser21 phosphorylation is differentially regulated by BSA and ${\mathbf{HCO}}_3^{\hspace{0.5em}-}$ during mouse sperm capacitation. Cauda epididymal sperm were incubated in TYH media containing either: BSA, ${\text{HCO}}_3^{\hspace{0.5em}-}$, both compounds or neither BSA nor ${\text{HCO}}_3^{\hspace{0.5em}-}$. In some experiments to evaluate the effect of cholesterol release, 3 mM of either β-methyl cyclodextrin or β-hydroxymethyl cyclodextrin was used instead of 5 mg/ml of BSA. At different incubation times, sperm proteins were extracted and analyzed using anti-phospho-GSK3 Ser21 antibodies. (A and B) Time curve in media using BSA alone (BSA + NC), and compared with controls of NC and CAP sperm extracts that were added in their respective lanes. (C and D) Time curve in media using ${\text{HCO}}_3^{\hspace{0.5em}-}$ alone (${\text{HCO}}_3^{\hspace{0.5em}-}$ + NC), and compared with controls of NC and CAP sperm extracts that were added. (E and F) Evaluation of beta cyclodextrins. In these experiments, beta cyclodextrins were used to replace BSA. (G and H) Time curve upon adding ${\text{HCO}}_3^{\hspace{0.5em}-}$ to sperm incubated for 5 min only in the presence of 5 mg/ml of BSA. In this experiment, sperm were incubated for 5 min in the presence of BSA. At 5 min, ${\text{HCO}}_3^{\hspace{0.5em}-}$ was added to a final concentration of 15 mM and sperm incubated for different time periods as shown in the figure. In this experiment, controls of sperm incubated for 60 min in either NC, ${\text{HCO}}_3^{\hspace{0.5em}-}$ alone, BSA alone or CAP conditions were also evaluated. In all experiments, upon conducting immunodetection with anti-phospho-GSK3α Ser21 antibodies, blots were stripped and re-probed with total anti-GSK3α. A, C, E, and G: western blots were conducted with anti-phospho-GSK3α Ser21, followed by stripping and re-probing with anti-total GSK3α antibody. B, D, F, and H: quantification of the respective A, B, C, and D experiments. Quantification was with ImageJ software, as described in the Materials and methods, by obtaining ratios of the phosphorylated bands with those of the total GSK3α protein. Results corresponding to three independent western blots were normalized to the NC ratio and expressed as average ratio ± SEM (n = 3). A Student’s t-test was performed and significance level denoted as: *P < 0.05; **P < 0.01; ***P < 0.001. TYH, Toyoda Yokoyama-Hosi (m-TYH) medium; BSA, bovine serum albumin; ${\text{HCO}}_3^{\hspace{0.5em}-}$, bicarbonate; CAP, capacitating media; NC, non-capacitating media; GSK3, glycogen synthase kinase 3. The whole blots for the cropped blots presented here are shown in Supplementary Figs S6 and S7.

Cyclic AMP-dependent pathways downregulate GSK3α N-terminal phosphorylation

As part of capacitation, ${\text{HCO}}_3^{\hspace{0.5em}-}$ activates sAC (Buffone et al., 2014; Gervasi and Visconti, 2016); therefore, to test the hypothesis that ${\text{HCO}}_3^{\hspace{0.5em}-}$ action was mediated by this pathway, sperm were incubated in the absence (NC) or presence of BSA, ${\text{HCO}}_3^{\hspace{0.5em}-}$ or both compounds (CAP) with or without the addition of dibutyryl cAMP (dbcAMP) and IBMX. As shown above, in control conditions, while ${\text{HCO}}_3^{\hspace{0.5em}-}$ decreased (Fig. 5A, third lane), BSA increased (Fig. 5A, fourth lane) Ser21 phosphorylation on GSK3α. Addition of cAMP agonists had marginal effects when added to sperm incubated in the presence of ${\text{HCO}}_3^{\hspace{0.5em}-}$ (Fig. 5A, fifth lane). On the other hand, addition of either ${\text{HCO}}_3^{\hspace{0.5em}-}$ or cAMP agonists decreased the high Ser21 phosphorylation observed with BSA alone (Fig. 5A, sixth lane). Quantification of the results from three independent experiments was by normalizing ratios to NC data, as explained in the Materials and methods (Fig. 5B).

Figure 5.

The ${\mathbf{HCO}}_3^{\hspace{0.5em}-}$-induced GSK3α Ser21 dephosphorylation in mouse sperm is mediated by cyclic AMP-dependent pathways. (A and B) Effect of cAMP agonists on sperm incubated in different media. Cauda sperm were incubated in conditions that support capacitation (CAP) or not (NC) and in NC media supplemented with either 5 mg/ml BSA alone (BSA), or 15 mM ${\text{HCO}}_3^{\hspace{0.5em}-}$ alone (${\text{HCO}}_3^{\hspace{0.5em}-}$): the latter conditions were repeated in the presence of cAMP agonists (1 mM dbcAMP, 100 µM IBMX). (C and D) sAC inhibition with TDI10229 blocks the capacitation-induced GSK3α Ser21 dephosphorylation. Sperm were capacitated for 60 min in complete media in the presence or absence of different TDI10229 concentrations as indicated in the figure. As controls, sperm extracts were incubated in CAP media and in NC media containing BSA alone without the addition of ${\text{HCO}}_3^{\hspace{0.5em}-}$. (E and F) TDI10229 inhibition is overcome by the addition of cAMP agonists. Cauda sperm were incubated in media that support capacitation in the presence or absence of 10 µM TDI10229 and in the presence or absence of cAMP agonists (1 mM dbcAMP, 100 µM IBMX). A, C, and E: western blots were conducted with anti-phospho-GSK3α Ser21, followed by stripping and re-probing with anti-total GSK3α antibody. B, D, and F: quantification of the respective A, C, and E experiments. Quantification was with ImageJ software as described in the Materials and methods by obtaining ratios of the phosphorylated bands with those of the total GSK3α protein. Results of three independent western blots were normalized to the NC ratio (B) and CAP ratio (D and F) and expressed as average ratio ± SEM. A Student’s t-test was performed and significance level denoted as: *P < 0.05; **P < 0.01; ***P < 0.001 (n = 3). CAP, capacitating media; NC, non-capacitating media; BSA: bovine serum albumin; ${\text{HCO}}_3^{\hspace{0.5em}-}$, bicarbonate; sAC, soluble adenyl cyclase; cAMP, cyclic AMP; dbcAMP, dibutyryl cAMP; IBMX, 3-isobutyl-1-methylxanthine; GSK3, glycogen synthase kinase 3. The whole blots for the cropped blots presented here are shown in Supplementary Fig. S7.

Consistent with the role of cAMP, various concentrations of a recently developed sAC inhibitor (TDI 10229) (Balbach et al., 2021) completely blocked GSK3α Ser21 dephosphorylation in media that support capacitation (Fig. 5C and D). Moreover, TDI 10229 inhibited dephosphorylation of Ser21 upon addition of ${\text{HCO}}_3^{\hspace{0.5em}-}$ when sperm were primed for 5 min with BSA (Fig. 5C and D). As predicted, cAMP agonists overcame the TDI 10229 effects and induced GSK3α Ser21 dephosphorylation (Fig. 5E and F). Altogether, these results indicate that BSA and the ${\text{HCO}}_3^{\hspace{0.5em}-}$/sAC/cAMP-dependent pathways play opposite roles in GSK3α regulation.

The role of cAMP is mediated by changes in Em and pH_i

As shown above, dephosphorylation of the GSK3α N-terminus requires at least 15 min incubation in the presence of ${\text{HCO}}_3^{\hspace{0.5em}-}$ (Figs 3C and 4C). These slow kinetics are similar to those observed for the capacitation-induced hyperpolarization of the sperm plasma membrane potential (Em) (Demarco et al., 2003), which is another cAMP-dependent process (Ritagliati et al., 2018). To test the extent by which Em hyperpolarization was involved in GSK3α dephosphorylation, we induced hyperpolarization in BSA-treated sperm with 1 µM valinomycin (V), a well-characterized K⁺ ionophore (Su et al., 2019). In the presence of this compound, the sperm behaves as a K⁺ electrode and the Em is clamped at ∼−80 mV (Demarco et al., 2003). This compound induced Ser21 dephosphorylation with fast kinetics (Fig. 6A and B) in both BSA-treated (Fig. 6A and B) and in sperm incubated in NC conditions (Fig. 6C and D, first two lanes). Consistent with the hypothesis that hyperpolarization mediates GSK3α dephosphorylation, when the Em was clamped in a depolarized state by adding high K⁺ (70 mM) (De La Vega-Beltran et al., 2012), GSK3α dephosphorylation was not observed (Fig. 6C and D, lanes 6 and 8, without and with valinomycin, respectively). When the same experiments were repeated with sperm incubated under CAP conditions, we observed that K⁺ alone slightly increased phosphorylation (Fig. 6C and D, seventh lane) and the effect was even more pronounced when, in addition to high K⁺, valinomycin (V) was also added (Fig. 6C and D, ninth lane).

Figure 6.

The HCO₃⁻-induced GSK3α Ser21 dephosphorylation in mouse sperm is mediated by Em hyperpolarization and by alkalinization of pHi. (A and B) Cauda epididymal sperm were incubated in conditions that do not support capacitation without (NC) or with the addition of 5 mg/ml of BSA in the absence (BSA) or presence of 1 µM valinomycin (BSA + V). (C and D) Sperm were incubated in NC, CAP, or BSA alone media in the presence or absence of valinomycin (V) with or without the addition of high (70 mM) K⁺ ions (VK) as indicated in the figure. (E and F) Sperm were incubated in conditions that do not support capacitation without (NC) or with the addition of 5 mg/ml of BSA in the absence (BSA) or presence of 10 mM NH₄Cl (BSA + NH₄Cl). (G and H) Sperm were incubated in NC or BSA alone media in the presence or absence of valinomycin and high K⁺ (VK) without (VK) or with the addition of 10 mM NH₄Cl (VK + NH₄Cl) as indicated in the figure. A, C, E, and G: western blots were conducted with anti-phospho-GSK3α Ser21, followed by stripping and re-probing with anti-total GSK3α antibody. B, D, F and H: quantification of the respective A, C, E, and G experiments. Quantification was done with ImageJ as described in the Materials and methods by obtaining ratios of the phosphorylated bands with those of the total GSK3α protein. Results of three independent western blots were normalized to the NC ratio (B, D, and F) and BSA ratio (H) expressed as average ratio ± SEM. A Student’s t-test was performed and significance level denoted as: *P < 0.05; **P < 0.01; ***P < 0.001 (n = 3). CAP, capacitating media; NC, non-capacitating media; BSA, bovine serum albumin; Em, membrane potential; pHi, intracellular pH; GSK3, glycogen synthase kinase 3. The whole blots for the cropped blots presented here are shown in Supplementary Fig. S8.

Capacitation is also associated with the increase of pH_i (Matamoros-Volante and Treviño, 2020), another cAMP-dependent process (Nishigaki et al., 2014). To test whether the increase in pH_i was involved in GSK3α Ser21 dephosphorylation, NH₄Cl was added to sperm preincubated with BSA for 5 min. Alkalinization of pH_i with NH₄Cl immediately induced GSK3α dephosphorylation (Fig. 6E and F). Interestingly, NH₄Cl-induced dephosphorylation was not blocked by clamping the Em with high K⁺ (Fig. 6G and H); however, it is important to notice that, in this case, dephosphorylation followed slower kinetics. Therefore, conclusions regarding the hierarchy of these pathways are not straightforward. On the other hand, both valinomycin (Fig. 7A and B) and NH₄Cl (Fig. 7C and D) induced GSK3α Ser21 dephosphorylation in sperm incubated with TDI 10229 in both sperm incubated only with BSA or in CAP conditions. Altogether, these results suggest that both Em hyperpolarization and the increase in pH_i are downstream cAMP-dependent pathways.

Figure 7.

Valinomycin and NH₄Cl induced dephosphorylation of GSK3α Ser21 in the presence of the sAC inhibitor TDI10229 in mouse sperm. Cauda sperm were incubated in either capacitating (CAP) or BSA alone (BSA) conditions in the absence or presence of TDI10229 10 µM with or without 1 µM valinomycin (VAL) (A and B) or NH₄Cl 10 mM (C and D). A and C: western blots were conducted with anti-phospho-GSK3α Ser21, followed by stripping and re-probing with anti-total GSK3α antibody. B and D: quantification was as described in previous figures by normalizing to ratios obtained from phospho-Ser21 GSK3α to total GSK3α in CAP conditions (B and D). Results of three independent western blots were normalized to the CAP ratio (B and D) and expressed as average ratio ± SEM (*P < 0.05; **P < 0.01; ***P < 0.001) (n = 3). CAP, capacitating media; NC, non-capacitating media; BSA, bovine serum albumin; GSK3, glycogen synthase kinase 3; sAC, soluble adenyl cyclase. The whole blots for the cropped blots presented here are shown in Supplementary Fig. S9.

GSK3α dephosphorylation is mediated by SLO3

SLO3 is a testis-specific K⁺ channel responsible for sperm hyperpolarization (Santi et al., 2010); we used two approaches to test its participation in GSK3α regulation. First, we used a pharmacological strategy and showed that clofilium, an SLO3 inhibitor known to block the capacitation-induced Em hyperpolarization (Chávez et al., 2013), inhibited GSK3α Ser21 dephosphorylation (Fig. 8A and B). The clofilium inhibitory effect on GSK3α Ser21 dephosphorylation was overcome by the addition of valinomycin but not when high K⁺ was present (Fig. 8A and B). Second, because, in addition to SLO3, clofilium also blocks other K⁺ channels, we also used a genetic loss of function strategy. Experiments with wild-type sperm (Fig. 8C and D) rendered results similar to those reported above in Fig. 6. On the other hand, sperm from SLO3 KO mice incubated under conditions that support capacitation did not undergo GSK3α Ser21 dephosphorylation (Fig. 8E and F). Similar to wild type (Fig. 8C and D), when sperm from the SLO3 KO model (Fig. 8E and F) were treated with either valinomycin or NH₄Cl, GSK3α Ser21 was dephosphorylated. Also consistent with the results shown in Fig. 6, in the presence of high K⁺, valinomycin did not induce GSK3α Ser21 dephosphorylation and NH₄Cl was able to induce Ser21 dephosphorylation even when hyperpolarization was prevented by clamping the Em in a depolarized state with high K⁺.

Figure 8.

Dephosphorylation of GSK3α Ser21 is mediated by SLO3 K⁺ channels in mouse sperm. (A and B) Sperm were incubated in NC, BSA alone, or CAP media as described in previous figures. Incubations were conducted in the absence or the presence of 50 µM clofilium (CLO) with the addition of either 1 µM valinomycin (Val) or 1 µM valinomycin and 70 mM K⁺ (Val + K⁺). (C and D) Wild-type C57Bl6 and siblings SLO3 KO epididymides were shipped overnight from Washington University, St Louis, MO, USA. Sperm were obtained by swim-out and incubated in either NC, BSA alone, or CAP media in the absence or the presence of valinomycin (Val), valinomycin with high K⁺, or NH₄Cl (NH) as described in the figure. A, C, and E: western blots were conducted with anti-phospho-GSK3α Ser21, followed by stripping and re-probing with anti-total GSK3α antibody. B, D, and F: quantification was as described in previous figures by normalizing to ratios obtained from phospho-Ser21 GSK3α to total GSK3α in NC conditions (B and D). Results of three independent western blots were normalized to the NC ratio and expressed as average ratio ± SEM (*P < 0.05; **P < 0.01) (n = 3). CAP, capacitating media; NC, non-capacitating media; BSA, bovine serum albumin; GSK3, glycogen synthase kinase 3. The whole blots for the cropped blots presented here are shown in Supplementary Fig. S10.

GSK3 activation is required for the ionomycin-induced acrosome reaction

Previously, Nixon’s group showed that GSK3 (α and/or β) inhibition blocked the progesterone-induced acrosome reaction (Reid et al., 2015) suggesting a role of GSK3 in this exocytotic event. Because progesterone acts exclusively in capacitated sperm, the effect on the acrosome reaction could be due to the block of capacitation-induced pathways or, alternatively, direct effects on exocytosis. To evaluate these alternative possibilities, we used the Ca²⁺ ionophore ionomycin, a compound capable of inducing the acrosome reaction in the absence of capacitation (Sánchez-Cárdenas et al., 2021). To evaluate whether the GSK3 role is downstream or upstream of the required increase in [Ca²⁺]_i, 10 µM ionomycin was added to capacitated sperm and the acrosome reaction evaluated in single cells in the presence or absence of two well-characterized GSK3 inhibitors: lithium and CHIR99021 (Reid et al., 2015; Choi et al., 2016). Surprisingly, both compounds inhibited the ionomycin-induced acrosome reaction (Fig. 9) suggesting that the role of GSK3 is downstream the rise of [Ca²⁺]_i.

Figure 9.

The ionomycin-induced acrosome reaction in mouse sperm is blocked by GSK3 inhibitors CHIR99021 and lithium. Cauda sperm were incubated with 2 µM Fluo 4 AM and 0.05% pluronic acid for 40 min in NC medium. Once loaded, cells were attached to laminin-coated cover slips and the chamber was filled with NC medium and sperm incubated in the recording chamber with 5 μM FM4-64. (A) Representative graph indicating Fluo4 fluorescence (green line) in sperm that either react (right panel) or not (left panel), measured by FM4-64 fluorescence (red line) upon addition of 10 µM ionomycin. (B) Fluorescence image sequence of sperm before and after addition of ionomycin. Left images indicate Fluo4 fluorescence in green; right images show changes in FM4-64 fluorescence in red. The images are representative of three independent experiments. (C) Average percentage of responding sperm upon ionomycin addition for Fluo4 (left panel) or FM4-64 (right panel). Data represent the average ± SEM. A one-way ANOVA was performed and significance level denoted as: *P < 0.05 (n = 3). For each individual experiment, at least 10 spermatozoa were analyzed. GSK3, glycogen synthase kinase 3; IONO, ionomycin; N AR, non-acrosome reaction; NC, non-capacitating; CAP, capacitating; CRL, control; CHIR, CHIR99021.

Discussion

Sperm capacitation involves a series of biochemical processes occurring physiologically in the female tract and culminates in the fusion of the spermatozoon with the egg. This process can be mimicked in vitro by incubation of mature sperm, either ejaculated or from the cauda epididymis, in defined media consisting of a physiological buffer to maintain pH, ions such as (e.g. Na⁺, K⁺, Cl⁻, ${\text{PO}}_4^{\mathrm{\hspace{0.5em}3}-}$, ${\text{HCO}}_3^{\hspace{0.5em}-}$, and Ca²⁺), energy nutrients (e.g. glucose, pyruvate, and lactate), and serum albumin. Although the individual relevance of each of these compounds varies depending on the species, all of them play roles in the regulation of sperm signaling during capacitation. In this work, we focused on BSA and ${\text{HCO}}_3^{\hspace{0.5em}-}$ action because of their relevance in capacitation (Gervasi and Visconti, 2016). In less than 1 min, ${\text{HCO}}_3^{\hspace{0.5em}-}$ activates sAC with the consequent increase in cAMP followed by a fast activation of PKA and probably other cAMP-dependent molecules such as the sperm-specific Na⁺/H⁺ exchanger (Wang et al., 2007; Windler et al., 2020) or EPAC (Branham et al., 2006). The action of serum albumin is less understood; addition of BSA to sperm induced the release of cholesterol, a process that can be mimicked by the addition of beta-cyclodextrin analogues (Choi and Toyoda, 1998; Visconti et al., 1999a). However, other evidence indicates that BSA is also involved in the regulation of pathways independently of cholesterol release (Xia and Ren, 2009b; Takei et al., 2021). Despite the fast action of these molecules, downstream events are activated at different time periods. For example, although the evidence that cAMP-dependent pathways are necessary for the increase in tyrosine phosphorylation, this last event occurs about 30–45 min after the initial PKA activation. Similarly, Em hyperpolarization occurs downstream of the increase in cAMP; however, changes in Em are delayed by about 15 min. On the other hand, the kinetics of BSA-dependent processes are less understood. Although cholesterol release is thought to be a slow process (Visconti et al., 1999b), other works have shown BSA effects in minutes (Xia and Ren, 2009a). Despite the limited knowledge regarding the interaction of serum albumin and ${\text{HCO}}_3^{\hspace{0.5em}-}$ pathways in sperm, both compounds are required for successful capacitation and IVF. In most cases studied, ${\text{HCO}}_3^{\hspace{0.5em}-}$ and BSA can act either synergistically, or independently.

Because sperm are translationally inactive, signaling in sperm is mostly related to post-translational modification, among them phosphorylation and dephosphorylation (Gervasi and Visconti, 2017). It is therefore not surprising that several kinases have been proposed to play essential roles in sperm signaling. Among them, the regulation of PKA by the ${\text{HCO}}_3^{\hspace{0.5em}-}$/sAC/cAMP pathway is one of the best studied using both pharmacological (Visconti, 2009; Gervasi and Visconti, 2017; Balbach et al., 2021) and genetic approaches (Skålhegg et al., 2002). The necessity of other kinases, such as TSSK1, 2 (Shang et al., 2010; Salicioni et al., 2020), 3 (Nayyab et al., 2021), and 6 (Spiridonov et al., 2005; Salicioni et al., 2020), casein kinase II (Xu et al., 1999), and CaMKIV (Wu et al., 2000), have also been validated using KO genetic models. However, because most of these models present spermiogenesis defects, it has not been possible to establish if, in addition, they also are essential for sperm maturation or capacitation. Other relevant sperm kinases are GSK3α and GSK3β. Although in most tissues both isoforms can replace each other’s function, while the GSK3α KO is viable (Kaidanovich-Beilin et al., 2009), GSK3β KO presents an embryonic lethality phenotype (Hoeflich et al., 2000). On the other hand, specific deletion of either GSK3α or GSK3β using a Cre-lox strategy indicates that, while GSK3β is dispensable for reproduction, GSK3α KO mice are sterile (Bhattacharjee et al., 2018). As both kinases are in almost all cases functionally interchangeable, this result was surprising. One plausible hypothesis consistent with these results is that the two isoforms have differential localizations. Although both have been localized to the sperm head (Bhattacharjee et al., 2018) and tail, only GSK3α immunofluorescence has been validated with sperm from the respective GSK3α null mice (Bhattacharjee et al., 2015). In addition to investigations using genetically modified KO mice, pharmacological studies have suggested a role for either GSK3α or GSK3β in regulation of the progesterone-induced acrosome reaction (Reid et al., 2015). Considering that capacitation is essential to regulate acrosomal exocytosis, we hypothesized that either GSK3α or GSK3β are regulated during capacitation.

GSK3α and GSK3β catalytic domains have 98% homology in their catalytic site; however, the homology is reduced in both the N-terminal and C-terminal domains. Their regulation is complex and includes phosphorylation of different domains, protein–protein interactions with their substrates, and localization to different cell domains. Both kinases are blocked by phosphorylation of their N-terminal domain (Ser21 for GSK3α; Ser9 for GSK3β) and require dephosphorylation of these residues to become active. Conversely, phosphorylation of Tyr residues embedded in the catalytic domain is essential for both kinases to be fully activated (Tyr279 for GSK3α; Tyr216 for GSK3β). N-terminal phosphorylation events can be followed using specific anti-phospho-antibodies recognizing either GSK3α Ser21, GSK3β Ser9, or both phosphorylated residues.

As a summary of our results (Supplementary Fig. S5), we found that:

• while GSK3β Ser9 phosphorylation is observed in immature caput sperm, once the sperm matured in the epididymis, this post-translational modification is gone;

• on the other hand, GSK3α Ser21 remains phosphorylated throughout the maturation process;

• once sperm are released from the cauda epididymis, GSK3β Ser 9 remains dephosphorylated; however, GSK3α Ser 21 undergoes a biphasic regulation. While phosphorylation of this residue is significantly increased during the first 15 min of capacitation, GSK3α Ser21 then becomes dephosphorylated. This last result is consistent with previous reports indicating that capacitation conditions induced GSK3α N-terminal dephosphorylation (Dey et al., 2020);

• while BSA in the media causes sperm GSK3α Ser21 phosphorylation, ${\text{HCO}}_3^{\hspace{0.5em}-}$ induces dephosphorylation of this N-terminal Ser;

• the HCO₃⁻-induced dephosphorylation is blocked in the presence of specific sAC inhibitors (Balbach et al., 2021) and induced with cAMP agonists. Moreover, cAMP agonists also induced dephosphorylation when sAC is blocked with TDI 10229;

• in the absence of ${\text{HCO}}_3^{\hspace{0.5em}-}$ (and consequently, in the absence of cAMP), hyperpolarization of the sperm EM using valinomycin (1 µM) induced dephosphorylation of GSK3α Ser21, suggesting that the capacitation-associated hyperpolarization is downstream of cAMP-dependent pathways. Importantly, valinomycin did not induce dephosphorylation when used combined with high K⁺ to maintain the sperm Em clamped in a depolarized state;

• interestingly, increasing pH_i with addition of NH₄Cl also induces Ser21 dephosphorylation in control conditions as well as in the absence of ${\text{HCO}}_3^{\hspace{0.5em}-}$, when the cAMP pathway is blocked with TDI10229, or when the Em was clamped in a depolarized state. Altogether, these experiments suggest that, at least in part, pH_i increase is a downstream Em hyperpolarization. Despite this conclusion, multiple molecules are known to be involved in the regulation of sperm pH_i (Gardner and James, 2023) and the NH₄Cl-induced alkalinization bypasses the spatial-temporal regulation of all of these molecules. Therefore, these results are silent regarding other effects of alkalinization upstream of SLO3;

• SLO3 is a sperm-specific K⁺ channel that is essential to modulate changes in Em during capacitation. Therefore, to further evaluate the role of hyperpolarization, we evaluated the GSK3α Ser 21 status in sperm from SLO3 KO mice and using clofilium, an SLO3 inhibitor. In these experiments, BSA induced GSK3α Ser21 phosphorylation in sperm from SLO3 KO at similar levels to the ones observed for wild type sperm. However, in SLO3 KO sperm, as well as in those treated with clofilium, GSK3α Ser21 did not undergo dephosphorylation during capacitation. Consistent with previous results using wild-type sperm, both valinomycin, to induce Em hyperpolarization, as well as NH₄Cl, to increase pH_i, dephosphorylated Ser21 residues; and

• finally, GSK3 kinases had been proposed to modulate the progesterone-induced acrosome reaction (Reid et al., 2015). These experiments were done by incubating sperm with CHIR99021 to block GSK3 (α and β) and then inducing the acrosome reaction by adding progesterone. In these conditions, the GSK3 inhibitors block the exocytotic event. However, progesterone can act as an acrosome reaction agonist only in capacitated sperm. Therefore, the inhibitor could be acting either upstream (e.g. pointing to a role on other capacitation-associated events), or downstream (directly blocking exocytosis). To evaluate these possibilities, we induced the acrosome reaction with ionomycin, a Ca²⁺ ionophore that can bypass capacitation events and directly induce the acrosome reaction. For these experiments, we used CHIR99021, the same GSK3 inhibitor used in the previous work (Reid et al., 2015), as well as Li⁺ (Choi et al., 2016), another generic GSK3 blocker. Surprisingly, GSK3 inhibitors block the ionomycin-induced acrosome reaction; because GSK3α is the only GSK3 found in the anterior acrosome, these experiments suggest that GSK3α activity is required for acrosomal exocytosis.

Altogether, these results suggest a hypothetical mechanism by which GSK3α phosphorylation and dephosphorylation occur during capacitation. Briefly, once the sperm is released to an environment containing serum albumin, GSK3α Ser21 in the sperm head becomes phosphorylated with the resultant inactivation of the pathway. We can speculate that this mechanism can serve to avoid spontaneous acrosome reactions. Simultaneously, in sperm exposed to high ${\text{HCO}}_3^{\hspace{0.5em}-}$, sAC is activated, cAMP synthesis is upregulated, and cAMP-dependent pathways are activated (Balbach et al., 2023). Downstream of cAMP, several processes start including activation of PKA-dependent and PKA-independent pathways. While PKA activation is fast, other downstream signals follow slower kinetics; among them, an SLO3-mediated Em hyperpolarization. Although SLO3 is predicted to localize in the sperm flagellum principal piece (Navarro et al., 2008), contrary to other signaling pathways, changes in Em affect the whole cell. We have used valinomycin to induce hyperpolarization as well as valinomycin + 70 mM KCl to clamp the sperm Em in a depolarized state. On the other hand, we used NH₄Cl to temporally increase pHi; however, it is important to take into consideration that pH changes can be local. Therefore, although NH₄Cl induces GSK3α Ser21 dephosphorylation even when the sperm Em is depolarized, this observation is silent regarding the role of pHi in upstream events such as SLO3 regulation. Our model is not in disagreement with an early role for the cAMP-dependent increase in pHi upstream of SLO3 activation and the consequent Em hyperpolarization. Indeed, recent work using a genetic mouse model lacking the sperm-specific Na⁺/H⁺ exchanger (Hernández-Garduño et al., 2022), as well as electrophysiological evidence, clearly indicates an alkalinization-dependent modulation of SLO3 activity (Schreiber et al., 1998; Navarro et al., 2008). We hypothesize that while physiologically, during capacitation, multiple (at least two) sequential, independent, and compartmentalized alkalinization events occur, NH₄Cl induces alkalinization simultaneously in the whole cell.

Although N-terminal dephosphorylation is a necessary step for GSK3 activation, dephosphorylation of GSK3α Ser21 or GSK3β Ser9 is not sufficient to start specific GSK3-dependent phosphorylation. Many GSK3 substrates require a ‘priming phosphate’ that serves as a GSK3-binding site to undergo further phosphorylation by GSK3 (Rayasam et al., 2009). In sperm, Nixon’s group has proposed dynamin 1 priming as a necessary step for GSK3-dependent phosphorylation of this protein (Reid et al., 2015). Although the identity of the priming kinase has not been established, this group hypothesized that dynamin 1 phosphorylation by GSK3 is required for the acrosome reaction. Supporting this hypothesis, GSK3 inhibition completely blocked the progesterone-induced acrosome reaction (Reid et al., 2015). Because progesterone induces the acrosome reaction only on capacitated sperm, the reported inhibition could occur at any step before the induction of the acrosome reaction. In this work, we used the same inhibitors to test whether the GSK3 function is upstream or downstream of the increase in [Ca²⁺]_i produced by ionomycin. Ionomycin is a Ca²⁺ ionophore capable of inducing the acrosome reaction even in non-capacitated sperm. Surprisingly, the ionomycin-induced acrosome reaction was blocked when sperm were previously incubated with GSK3 inhibitors. This result is consistent with previous results from Nixon’s laboratory, suggesting that during capacitation, GSKα is needed for the acrosome reaction but its action can be observed only upon priming of dynamin 1 or another molecule required for exocytosis.

Finally, it is important to highlight that the present study only revealed a partial picture of GSK3 regulation during capacitation. Many questions remain unanswered; for example, which is the kinase involved in phosphorylating GSK3α upon BSA addition? Which is the phosphatase(s) involved in dephosphorylation of GSK3α downstream of the cAMP-dependent pathway? Regarding this last question, it has been shown that GSK3 dephosphorylation is regulated by [Ca²⁺]_i through activation of the Ca²⁺/calmodulin-dependent phosphatase calcineurin (Dey et al., 2020). Like pH_i, the increase in Ca²⁺ is temporally and spatially regulated; its role, as well as the role of Ca²⁺/calmodulin pathways, in GSK3α regulation warrant further investigation. Another remaining question is the paradoxical role of pH_i; on one hand, alkalinization is upstream of SLO3 activation; however, our findings indicate that alkalinization dephosphorylates GSK3α Ser21 even when the sperm Em is clamped in a depolarized state with high K⁺ and valinomycin.

Supplementary data

Supplementary data are available at Molecular Human Reproduction online.

Data availability

The data underlying this article are available in the article and in its online supplementary material.

Authors’ roles

Conceptualization, P.E.V. and G.M.; methodology, G.M., B.P., D.A.T., A.M.S., and M.G.G.; software, G.M.; validation, G.M., B.P., and C.S.-C.; formal analysis, G.M.; investigation, G.M., B.P., M.G.G., and D.A.T.; resources, P.E.V., A.D., C.M.S., and J.R.P.; data curation, G.M.; writing—original draft preparation, P.E.V. and G.M.; writing—review and editing, A.D., M.G.G., C.M.S., A.M.S., J.R.P., C.S.-C., and P.E.V.; visualization, G.M., M.G.G., and C.S.-C.; supervision, P.E.V. and G.M.; project administration, P.E.V.; funding acquisition, P.E.V., C.M.S., A.D., and J.R.P. All authors have read and agreed to the published version of the article.

Funding

The Eunice Kennedy Shriver NICHD, NIH (HD-038082 to P.E.V., HD088571 to J.B., L.L., P.E.V., and HD069631 to C.M.S); National Institute of Health and National Institute of Environmental Health Sciences (R01ES030942 to P.E.V. and J.R.P.); University Grants Commission, Government of India for the postdoctoral fellowship (F.15-1/2017/PDFWM-2017-18-ORI-48394(SA-II) to G.M.); Biotechnology Training Program (BTP) supported by National Research Service Award (T32 GM108556 to D.A.T.); Consejo Nacional de Humanidades Ciencia y Tecnología de México (CONACHYT, CF-2023-I-291); Dirección de Asuntos del Personal Académico/Universidad Nacional Autónoma de México (DGAPA/UNAM, IN200919 to A.D.).

Conflict of interest

The authors declare no conflicts of interest.