|Home | About | Journals | Submit | Contact Us | Français|
During mammalian fertilization, the contact between sperm and egg triggers increases in intracellular Ca2+ concentration ([Ca2++]i) in sperm. Voltage-gated Ca2+ channels (CaVs) are believed to mediate the initial phase of [Ca2+]i increases in sperm induced by egg coat (zona pellucida [ZP]) glycoproteins, while store depletion-activated Ca2+ entry is thought to mediate the sustained phase. Using patch-clamp recording and Ca2+ imaging, we show herein that CaV channel currents, while found in spermatogenic cells, are not detectable in epididymal sperm and are not essential for the ZP-induced [Ca2+]i changes. Instead, CATSPER channels localized in the distal portion of sperm (the principal piece) are required for the ZP-induced [Ca2+]i increases. Furthermore, the ZP-induced [Ca2+]i increase starts from the sperm tail and propagates toward the head.
Increases in intracellular Ca2+ concentration ([Ca2+]i) have fundamental roles in sperm function such as capacitation, motility change, the acrosome reaction (AR), and egg penetration [1, 2]. Proteins of several Ca2+-permeable ion channels have been found in mammalian sperm. These include voltage-gated Ca2+ channels (CaVs), transient receptor potential channels, cyclic nucleic gated channels, and CATSPER channels . A major pathway for Ca2+ entry into sperm is believed to be through CaV channels, similar to that in neurons and other excitable cells [3–5]. Following the interaction between egg coat proteins (zona pellucida [ZP]) and sperm, for example, sperm CaVs are believed to be activated by membrane depolarization; the subsequent Ca2+ influx through the channels is proposed to lead to an initial [Ca2+]i rise and a sustained one later [1–6]. Consistent with a role of CaVs in sperm, a low voltage-activated (T type) CaV current has been recorded from spermatogenic cells (spermatocytes and spermatids, the precursors of sperm) [7–9] and from early-stage testis sperm . Because sperm do not normally have the capability to synthesize new proteins, the T-type CaV channels are believed to be responsible for the CaV-dependent Ca2+ influx in mature sperm .
Despite the seemingly overwhelming evidence supporting the roles of CaVs in the Ca2+ influx, functional CaV has not been directly analyzed from epididymal sperm, partially because of the difficulties in applying whole-cell patch-clamp technique to sperm until recently . In addition, the in vivo role of CaVs in sperm Ca2+ signaling and mammalian fertilization is questioned recently by the lack of obvious fertilization-specific phenotype in more than a dozen CaV gene knockouts, despite profound phenotypes in the nervous, muscle, and other systems [12, 13]. In this study, we directly test the existence of CaV currents and their role in ZP-induced Ca2+ entry in epididymal mouse sperm using whole-cell patch-clamp recording and Ca2+ imaging. Surprisingly, CaV current is not detectable in epididymal sperm, nor are the CaV channels essential for the ZP-induced Ca2+ influx. Instead, CATSPER channels are required for the ZP-induced [Ca2+]i increases.
Reagents were from Sigma (St. Louis, MO) unless otherwise stated. Fluo-4 AM, Fura-2 AM, and pluronic F-127 were from Molecular Probes (Invitrogen, Eugene, OR). Cell-Tak was from BD Biosciences (Bedford, MA). Protease inhibitor cocktail and DNase were from Roche Diagnostics (Indianapolis, IN). Pertussis toxin (PTX) and ionomycin were from Calbiochem (Gibbstown, NJ). Coomassie blue G-250 and mounting medium (Permount) were from Fisher Scientific (Pittsburgh, PA).
All procedures described herein were reviewed and approved by the University of Pennsylvania Institutional Animal Care and Use Committee and were performed in accord with the Guiding Principles for the Care and Use of Laboratory Animals. The Catsper1 knockout mutant strain had been backcrossed to C57BL/6J for more than 10 generations . The GFP-Catsper1 transgenic mice carry a transgene encoding a fusion protein between GFP and CATSPER1 in the Catsper1-null background . Unlike the Catsper1-null mice, the male transgenic mice are fertile, suggesting that the fusion protein functionally replaced the wild-type CATSPER1. Experiments involving mutant/wild-type paired preparations were done in a blind manner in which the experimentalist did not know the genotype during experiments.
The ZP were prepared using the Percoll gradient method . Ovaries from 25–35 mice (~3 wk old) were homogenized at room temperature in a glass tissue grinder in 2 ml of homogenization buffer (HB) buffer containing 25 mM triethanolamine,150 mM NaCl, 1 mM CaCl2, 1 mM MgCl2 (pH adjusted to 8.5 with 1 N HCl), 1 mM PMSF, DNase (0.2 mg/ml), and protease inhibitor cocktail with 2 mM ethylene. The homogenate was mixed with detergents (0.2 ml of 10% Nonidet P-40 and sodium deoxycholate) loaded onto a discontinuous three-step Percoll gradient in HB buffer with 2% (2 ml), 10% (2 ml), and 22% (3 ml) Percoll in a siliconized 15-ml plastic tube and centrifuged at 400 × g for 2 h at 4°C in a swinging bucket rotor. The 10% Percoll layer, a major ZP-containing fraction, was diluted with 45 ml of HB buffer and centrifuged to pellet the ZP at 20000 × g for 20 min at 4°C. Pelleted samples were washed with HB buffer followed by a wash with HS medium  (135 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 30 mM HEPES, 10 mM glucose, 10 mM lactic acid, and 1 mM pyruvic acid; pH adjusted to 7.4 with NaOH) or with divalent-free (DVF) buffer [11, 18] (150 mM Na-gluconate, 20 mM HEPES, and 5 mM HEDTA [N-(2-hydroxyethyl) ethylenediamine-N,N′,N′-triacetic acid], pH 7.4 with NaOH), each with a 20-min centrifucation at 20000 × g. The final ZP pellet was resuspended in 0.1–0.2 ml cold HS medium or DVF buffer, and the number of ZP was counted under the microscope. The ZP were solubilized by reducing the pH to 2.5 with 1 N HCl, followed by incubation at 37°C for 15 min and centrifugation at 20000 × g for 10 min at room temperature. The supernatant was neutralized to pH 7.4 with 1 N NaOH, stored in aliquots at −70°C, and used within 3 months. Negative controls included buffer alone (without the starting ovary materials) that underwent similar initial procedures, preparation from the 22% Percoll layer in the gradient spinning, and flow-through from solubilized ZP applied to a concentrator with a 10-kDa molecular weight cutoff (YM-10; Millipore, Bedford, MA) (the molecular weights of the mouse ZP are >60 kDa ). As expected, these controls did not induce obvious [Ca2+]i changes.
Caudal epididymides were excised and rinsed with HS medium. Sperm were released from three small incisions at 37°C, 5% CO2, for 15 min into HS medium supplemented with 5 mg/ml of bovine serum albumin and 15 mM NaHCO3. Released sperm were concentrated to 5 × 106 to 1×107/ml by centrifugation for 4 min at 300 × g, followed by capacitation in suspension for 90 min at 37°C, 5% CO2 in the same medium. During the last 25 min of capacitation, cells were loaded with 10 μM Fluo-4 AM and 0.05% Pluronic F-127, followed by two washes in imaging medium (HS supplemented with 15 mM NaHCO3), each with a 4-min spin at 300 × g. Washed sperm were plated onto coverslips coated with Cell-Tak. Small-volume imaging chambers (~1 cm diameter [90 μl]) were formed with Sylgard (Dow Corning, Midland, MI) on coverslips. Cells were allowed to attach for 10 min. A monochromator (DeltaRAM V; PTI, Birmingham, NJ) with a 75-W xenon lamp was used to generate the excitation at 488 nm. A 60× objective and a 1.6× adaptor on an inverted microscope (IX-71; Olympus, Tokyo, Japan) were used for imaging. Emissions (515–565 nm) were bandpass filtered (HQ540/50; Chroma, Rockingham, VT) and collected with a cooled charge-coupled device camera (CoolSNAP HQ; Roper Scientific, Tucson, AZ) for 25 milliseconds in every 0.5 sec for fast recording or for 100 milliseconds in every 6 sec for slow recording. Online control, data collection, and image processing were done using commercial software (ImageMaster 3; PTI). [Ca2+]i changes are presented as ΔF:F0 ratios after background subtraction, where ΔF is the change in fluorescence signal intensity and F0 is the baseline as calculated by averaging the 10 frames before stimulus application. For imaging of sperm from GFP-Catsper1 transgenic mice, the ratiometric measurement with Fura-2 AM (5 μM for loading) was used because of fluorescence from GFP. [Ca2+]i changes are presented as the ratio of F340:F380 after background subtraction. Calcium imaging experiments were done at 37°C unless otherwise stated. Cells with uneven dye loading were excluded from analysis. Motile sperm (identified by comparing multiple-image frames) that had one or two points attached to the coverslip were used for analysis. Cells with peak changes of >50% in ΔF:F0 (for Fluo-4 AM) or >0.1 in F340:F380 (for Fura-2 AM) after application of ZP were counted as responsive. To detect the Ca2+ responses at “clamped” membrane potential, K+ ionophore valinomycin (1 μM) was added to the imaging buffer, with additional K+ as indicated to replace equal molar concentration of Na+. The equilibrium (Nernst) potential for K+ (EK) was calculated based on the assumption of an intracellular K+ concentration of 120 mM [19, 20].
All recordings were done at room temperature. Liquid junction potentials, calculated using Clampex software (Molecular Devices, Sunnyvale, CA), were corrected. Sperm were obtained from 4- to 8-mo-old mice and recorded without capacitation. Whole-cell current recordings with seals between glass pipettes (7–10 MΩ resistance) and the sperm cytoplasmic droplets followed previously described methods . Unless otherwise stated, a modified HS solution (135 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgSO4, 20 mM HEPES, 5 mM Glucose, 10 mM lactic acid, 1 mM sodium pyruvate; pH adjusted to 7.4 with NaOH) was used as bath . Pipette solution contained 135 mM Cs-methanesulfonate, 5 mM CsCl, 5 mM Hepes, 10 mM ethyleneglycoltetracetic acid, 5 mM Na2ATP, and 0.5 mM Na2-glutamyl transpeptidase (pH adjusted to 7.2 with CsOH). Pipette solutions containing CsF  were also used in some recordings; no voltage-gated channel currents were detected in epididymal sperm (n = 9) (data not shown).
Patch-clamp recordings using spermatocytes and spermatids followed previously described methods [7, 14]. Cells were extruded by manual trituration from lightly dissociated seminiferous tubules with forceps and were filtered through a 70-μm cell strainer. Data from pachytene spermatocytes and round spermatids used for patch-clamp recordings were pooled.
Detection of the AR with Coomassie blue G-250 was performed as previously described [7, 21]. At least 300 sperm (>100 per sample) from three pairs of mutant/wild-type mice were counted as acrosome reacted (no staining in the acrosomal region) or as acrosome intact (dark blue staining over the acrosomal region).
Data analyses were performed using ImageMaster3 (Photon Technology International, Monmouth Junction, NJ), Excel (Microsoft, Redmond, WA), and Origin (Microcal Software Inc., Northampton, MA). Student t-test and single-factor ANOVA were used for statistical comparison between different treatments. P < 0.05 was considered statistically significant.
Ion channel currents have been difficult to characterize in epididymal sperm because of technical difficulties of whole-cell patch-clamp recordings in sperm until recently . To test directly whether epididymal sperm have functional CaVs, we recorded from mouse corpus epididymal sperm using the whole-cell configuration . As a positive control for sperm channel recording, the CATSPER channel current (ICATSPER) was readily detectable (n = 9) (Fig. 1A). However, we detected no CaV currents during the depolarizing pulses or on repolarization to −103 mV for tail current detection in any of the sperm cells (n = 23) (Fig. 1B). Using the same bath and pipette solutions, T-type CaV currents were detected in all 13 spermatocytes and spermatids (peak current at −53 mV; range, −13 to −137 pA [mean ± SEM, −55 ± 11 pA]) (n = 13) (Fig. 1C). Similar to T-type CaV channels in somatic cells , the channels in spermatogenic cells were easily inactivated even at hyperpolarized membrane potentials (half-maximum inactivation, −93 mV) (n = 5) (Fig. 1D). When spermatocytes and spermatids were held at −33 mV, the channels became completely inactivated, and the current was undetectable (n = 6) (Fig. 1E).
To maximize CaV current detection sensitivity, we also recorded from spermatogenic cells and sperm in a DVF bath. Like CaVs in other cells, the spermatogenic cell T-type CaV channels became nonselective and had a large peak inward current in the DVF bath (range, −85 to −2582 pA [mean ± SEM, −636 ± 393 pA]) (n = 6). Sperm CaV current recordings in the DVF bath were performed using Catsper1-null sperm because large inward current at the holding potential (−103 mV) through the wild-type CATSPER channel in the absence of Ca2+ made the recordings unstable (data not shown) (Fig. 1A) . Again, we detected no significant CaV currents (n = 9). These data suggest that spermatogenic cells possess T-type CaV channels, like neurons, muscle cells, and fibroblasts, but functional CaV currents are undetectable in the epididymal sperm.
The lack of detectable CaV current in the epididymal sperm was surprising because T-type CaV currents were clearly present in spermatogenic cells and were recorded from some testis sperm . When CaV currents recorded from spermatogenic cells and different stages of sperm were analyzed, CaV current sizes gradually decreased as testis sperm became more mature; no current was detectable in sperm before leaving the testis (Supplemental Fig. S1 and all Supplemental Data are available online www.biolreprod.org).
It is possible that small CaV currents are present in epididymal sperm but that the sizes of the currents fall below our threshold for detection even with a DVF bath. Furthermore, we recorded from incapacitated corpus sperm because the more mature, capacitated caudal sperm used for fertilization are inaccessible to robust whole-cell patch-clamp recording . To further test whether T-type CaV channels, if present, are required for the ZP-induced Ca2+ entry in mature sperm, we imaged the [Ca2+]i changes in the head region of capacitated caudal sperm using the Ca2+ indicator Fluo-4 AM. Consistent with other observations [23, 24], bath application of solubilized ZP (2 ZP/μl) led to increases in [Ca2+]i in the sperm head (represented as normalized fluorescence changes ΔF:F0) in 66% (106 of 160) of the cells. Most of the responses appeared within 20 sec of ZP application (Fig. 2A), and some (37%) also had a delayed phase clearly separated from the initial phase (Fig. 3A). The responses were disrupted by inhibiting the G protein with PTX (100 ng/ml [18% responsive cells] [Fig. 2B]) or phospholipase C with neomycin (1 mM [15% responsive cells] [Fig. 2C]), suggesting involvement of G proteins in the ZP-induced [Ca2+]i changes [7, 23, 25]. As a control, cell-permeable cGMP (8-Br-cGMP) induced an increase in [Ca2+]i  that was not blocked by PTX or neomycin treatment (Fig. 2, B and C).
We then increased the extracellular potassium concentration ([K+]o) from 5 mM to 18.5 mM and to 56.8 mM such that the EK values were −50 mV and −20 mV, respectively (assuming [K+]i = 120 mM [see Materials and Methods]). Because of the prominent roles of K+ in determining the resting membrane potential in sperm cells , increasing [K+]o should depolarize the cells and inactivate T-type CaV channels. Surprisingly, application of ZP still robustly led to increases in [Ca2+]i (Fig. 2, D and F). Neither the percentages of responsive cells nor the response magnitudes were significantly reduced (Fig. 2F). We then included the K+ ionophore valinomycin (1 μM) in the baths to clamp the membrane potential to the EK at −50 mV or −20 mV. The T-type CaV channels, if existing, would have been inactivated at those holding potentials (Fig. 1). Again, application of ZP increased [Ca2+]i (Fig. 2, E and F). Taken together, these data suggest that CaV channel activation is not required for the ZP-induced [Ca2+]i increases. Consistent with this notion, Zn2+ (10 μM) effectively blocked the T-type channel current in spermatocytes (n = 4) (data not shown) but did not significantly inhibit the ZP-induced Ca2+ responses (72% [18 of 25 cells] responsive; mean ± SEM ΔF:F0 peak, 362% ± 57%) (n = 18). See Figure 3 (A and E) for a comparison with responses in the absence of Zn2+.
Another family of Ca2+-permeable channels in sperm is the CATSPER . All four CATSPER family members have restricted expression in testis and sperm. The channel is activated by intracellular alkalization and is weakly affected by membrane potentials but, unlike CaV channels, is not inactivated by depolarization (Fig. 1A) . Sperm deficient in CATSPER channels can fertilize zona-free eggs but not zona-intact ones, presumably because of the mutant sperm's inability to penetrate the zona layer [14, 18, 26, 27]. We compared the ZP-induced [Ca2+]i changes in sperm prepared from wild-types vs. Casper1-null mutants, which have intact T-type CaV channel currents in spermatocytes . In contrast to the wild-type (Fig. 3A), sperm deficient in CATSPER1 lacked any detectable [Ca2+]i changes within 2 min of ZP stimulation (Fig. 3, B and E). Delayed responses were present in 18% of sperm (Fig. 3C). ZP-induced [Ca2+]i changes were restored by a transgene encoding a GFP-CATSPER1 fusion protein in the Catsper1-null background (Fig. 3, D and E).
The dependence of the ZP-induced [Ca2+]i increase in sperm head on CATSPER channels was unexpected, as CATSPER proteins [14, 18, 26] and currents through CATSPER channels  are localized selectively to the distal portion of sperm flagella (the principal piece). Findings from earlier studies using Ca2+-sensitive dyes with lower sensitivity (calcium green-1/rura-red ) or imaging at a slower frame rate (0.5 frames/sec ) suggested that the ZP-induced [Ca2+]i increase starts in the head. To further examine the spatial-temporal kinetics of the [Ca2+]i increases in the cellular subdomains along the sperm, we imaged the [Ca2+]i dynamics with the highly sensitive Fluo-4 AM Ca2+ indicator at a higher frame rate (2 frames/sec). Similar to previous findings , the absolute fluorescence changes (ΔF) observed in the tail regions were small compared with those in the head, presumably because of the smaller cellular volume in the tail regions. The normalized changes (ΔF:F0), however, clearly started from the principal piece (Figs. 4A and and55 and Supplemental Movie). At 37°C, the delay between the principal piece and head was a mean ± SEM of 2.9 ± 0.3 sec (n = 20). Lowering the temperature from 37°C to 18°C increased the interval to a mean ± SEM of 4.4 ± 0.3 sec (n = 12). There was also a mean ± SEM delay of 1.1 ± 0.1 sec between the [Ca2+]i increases in regions 5 μm apart within the midpiece (n = 20). In contrast, no significant delay was detected between regions with similar distance within the principal piece (Fig. 4B), the region where CATSPER channels are localized [14, 18, 26]. Calcium ionophore ionomycin (10 μM [Fig. 4A]) and A23187 (2 μM [Supplemental Fig. S2]) led to [Ca2+]i increases simultaneously in all subregions.
In this study, we directly analyzed ion channel currents from epididymal sperm using whole-cell patch-clamp recording. We detected no CaV currents in the sperm cells, although CATSPER channel currents were readily detectable. CaV channels, even if present in sperm, did not seem to be required for the ZP-induced [Ca2+]i changes, as the [Ca2+]i rises were not blocked by raising [K+]o such that the channels would have been inactivated. Finally, mutation in the Catsper1 channel gene disrupted a ZP-induced [Ca2+]i change, and such disruption could be rescued by a GFP-Catsper1 transgene.
How do these data reconcile with previous findings implicating CaVs in the Ca2+ influx in sperm? First, sperm membrane depolarization and intracellular alkalization, which can be induced by ZP proteins, were shown to lead to Ca2+ influx that can be inhibited by CaV channel blockers with variable efficacies [7, 29–32]. However, later findings revealed that the alkaline depolarization-induced Ca2+ influx required CATSPER channels . It is possible that some of the inhibition on the ZP-induced Ca2+ responses by nonspecific T-type CaV channel blockers such as Cd2+ and Ni2+ was due to the drug's blocking effects on CATSPER channels . Second, mRNAs of several CaV pore-forming subunits (α1) and the associated auxiliary subunits have been detected in testis and proteins detected in sperm [3, 9, 17, 34, 35]. However, the functional significance of these proteins in sperm physiology is not clear, as none of more than a dozen CaV channel gene knockouts seems to have significant defect in sperm function or male fertility [12, 13]. In contrast, disruption in any of the four Catsper genes leads to profound defect in sperm hyperactivated motility and to complete male infertility [14, 18, 26, 27]. Third, as for the possible role of the T-type CaV current recorded in spermatocytes and spermatids, a targeted disruption in a CaV gene (CaV3.2, official symbol Cacna1h) eliminates all detectable CaV current in spermatocytes but does not lead to defect in the ZP-induced [Ca2+]i changes or male fertility , despite pronounced smooth muscle phenotypes in the mutant . Taken together, the simplest explanation of these data is that the ZP-induced [Ca2+]i increase starts from Ca2+ entry via CATSPER channels in the tail instead of T-type CaV channels in the head.
The pH-sensitive CATSPER channels are presumed to be tetramers formed by all four CATSPER proteins , but the biological activators of the channels are unknown. Our data suggest that the egg coat proteins are potential activators, most likely through ZP-induced intracellular alkalization [11, 30]. Consistent with indirect activation, we did not observe direct potentiation of ICATSPER by ZP application during whole-cell patch-clamp recordings (n = 6) (Supplemental Fig. S3). Similar indirect activation of CATSPER channels was observed with cyclic nucleotides, which induce CATSPER-dependent [Ca2+]i increases  but do not seem to activate the channel directly . Given the localization of CATSPER proteins in the principal piece [14, 18], Ca2+ ions entering through the channels may diffuse toward the head but perhaps more likely are used as a signal to trigger the eventual [Ca2+]i rise in the midpiece and head . A “tail-to-head” sequence has also been described in the sperm [Ca2+]i increases induced by odorant bourgeonal in human , by cyclic nucleotides in mouse , and by egg peptide speract in sea urchin . How activation of the ZP receptors, presumably located in sperm head, is coupled to channel opening in the principal piece remains to be further examined, along with how Ca2+ entry through the channel in the tail leads to the eventual [Ca2+]i rises in the midpiece and head.
The [Ca2+]i increases seen in sperm can be divided into two temporal phases: a CATSPER-dependent initial phase that lasts minutes and a delayed phase that apparently does not have an absolute requirement for CATSPER but may occur via a store depletion-activated mechanism . The ZP-induced [Ca2+]i increases are believed to be important for the AR, a Ca2+-dependent exocytotic process by which the sperm cell releases its acidic acrosomal contents and prepares for its final fusion with the egg. We assayed the ZP-induced AR rates and found no significant difference between the Catsper1-null and wild-type sperm (Supplemental Fig. S4). The Ca2+ requirement for the AR may be from the delayed phase of the [Ca2+]i rises that is present in the mutant sperm (Fig. 3), consistent with the slow time course of the AR (lasting minutes). The Catsper1 mutant's ability to undergo the AR and inability to penetrate ZP-intact eggs suggest that the [Ca2+]i increase initiated in the principal piece and propagating along the whole sperm may be involved specifically in egg penetration. Future studies with freely moving sperm will be needed to further dissect the role of the CATSPER channels under physiological conditions.
We thank Drs. Mariano G. Buffone and George Gerton for performing pilot studies; Bayard Storey for showing us ZP preparation and AR detection; Betsy Navarro, Yuriy Kirichok, David Clapham, and George Gerton for critically reading early versions of the manuscript; and Alberto Darszon, Pablo Martinez Lopez, and Harvey Florman for useful discussion.
1Supported by NIH grants 1R01HD047578 and 1R03HD045290.