CatSper channels are regulated by protein kinase A

Mammalian sperm must undergo capacitation as a preparation for entering into hyperactivated motility, undergoing the acrosome reaction, and acquiring fertilizing ability. One of the initial capacitation events occurs when sperm encounter an elevated HCO3− concentration. This anion activates the atypical adenylyl cyclase Adcy10, increases intracellular cAMP, and stimulates protein kinase A (PKA). Moreover, an increase in intracellular Ca2+ concentration ([Ca2+]i) is essential for sperm capacitation. Although a cross-talk between cAMP-dependent pathways and Ca2+ clearly plays an essential role in sperm capacitation, the connection between these signaling events is incompletely understood. Here, using three different approaches, we found that CatSper, the main sperm Ca2+ channel characterized to date, is up-regulated by a cAMP-dependent activation of PKA in mouse sperm. First, HCO3− and the PKA-activating permeable compound 8-Br-cAMP induced an increase in [Ca2+]i, which was blocked by the PKA peptide inhibitor PKI, and H89, another PKA inhibitor, also abrogated the 8-Br-cAMP response. Second, HCO3− increased the membrane depolarization induced upon divalent cation removal by promoting influx of monovalent cations through CatSper channels, which was inhibited by PKI, H89, and the CatSper blocker HC-056456. Third, electrophysiological patch clamp, whole-cell recordings revealed that CatSper activity is up-regulated by HCO3− and by direct cAMP injection through the patch-clamp pipette. The activation by HCO3− and cAMP was also blocked by PKI, H89, Rp-cAMPS, and HC-056456, and electrophysiological recordings in sperm from CatSper-KO mice confirmed CatSper's role in these activation modes. Our results strongly suggest that PKA-dependent phosphorylation regulates [Ca2+]i homeostasis by activating CatSper channel complexes.

Chang and Austin (1,2) demonstrated independently that sperm must undergo capacitation in the female reproductive tract before being capable of penetrating and fertilizing the egg. During this process sperm undergo numerous physiological, biophysical, and biochemical changes. One of the first responses that occur during sperm capacitation is an increase in intracellular cAMP levels induced by the influx of bicarbonate that activates the atypical soluble adenylyl cyclase Adcy10 (also known as sAC) 3 (3,4), eventually leading to PKA-dependent signaling events that include an increase in protein tyrosine phosphorylation (5). The key role cAMP plays in capacitation and fertilization has been demonstrated using pharmacological (6) tools for gain and loss of function of cAMP-dependent pathways. The role of cAMP was confirmed using mice lacking either Adcy10 (7,8) or the spermspecific splicing variant of the PKA catalytic subunit C␣2 (9). Both these mice are sterile and their sperm cannot fertilize in vivo or in vitro. Furthermore, Jansen et al. (10) demonstrated that sperm from Adcy10 null mice expressing a photoactivated adenylyl cyclase recovered their motility and fertilizing capacity when light-stimulated.
In addition to cAMP, sperm capacitation is also associated with the modulation of [Ca 2ϩ ] i , and multiple data indicate that Ca 2ϩ plays a biphasic role in the regulation of cAMP homeostasis. First, Ca 2ϩ positively modulates Adcy10 activity by decreasing this enzyme's K m toward ATP (11). Second, Ca 2ϩ negatively regulates cAMP-dependent pathways activating cAMP degradation by phosphodiesterase 1 (PDE1) and by activating the Ser/Thr phosphatase PPP3C (also known as calcineurin) and therefore stimulating the hydrolysis of Ser/Thr phosphorylation sites downstream of PKA (11).
Work by different groups (12)(13)(14)(15)(16) indicates that sperm incubated in media with low extracellular Ca 2ϩ concentrations (Ͻ100 nM) do not undergo hyperactive motility. On the other hand, it has been shown that increasing the concentration of intracellular Ca 2ϩ ([Ca 2ϩ ] i ) using Ca 2ϩ ionophores such as A23187 resulted in complete loss of motility. However, A23187-treated sperm are not dead; if the ionophore is washed out, they recover motility and undergo hyperactivation (17). More interesting, sperm treated with A23187 as described above can overcome the inhibition of hyperactivation and fer- . The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This article contains Figs. S1 and S2 . 1 To whom correspondence may be addressed. E-mail: pvisconti@vasci.
tilization produced by blocking cAMP signaling with H89 (17). Altogether, these experiments suggest that, in sperm, Ca 2ϩ plays a role downstream of the activation of cAMP-dependent pathways. Despite these findings, how cAMP regulates [Ca 2ϩ ] i is not clear. During capacitation, [Ca 2ϩ ] i increases and it has been shown that the sperm-specific Ca 2ϩ channel CatSper contributes to the influx of this cation. This channel complex is modulated by changes in intracellular pH (pH i ) and in membrane potential (Em). Mice lacking the CatSper channel complex are sterile and their sperm are unable to undergo the capacitation-associated changes in the sperm motility pattern known as hyperactivation (14, 18 -20). CatSper is constituted by a complex that contains nine proteins: four that form a pore through which Ca 2ϩ can enter (CatSper 1-4), and five accessory proteins whose roles in hyperactivated motility are less clear (21). The localization of CatSper in the mouse sperm flagella is notable; it is found in four parallel tracks which generate unique Ca 2ϩ signaling domains along the principal piece of the flagella.
Although the roles of cAMP and CatSper in sperm capacitation are well established, the extent to which CatSper is modulated by a cAMP-dependent pathway is still controversial. On the one hand, it is well established that permeable cAMP agonists can induce an increase in [Ca 2ϩ ] i in sperm (22) and that this stimulatory effect is abolished in sperm from CatSper null mice (18). On the other hand, CatSper does not have any predicted cAMP-binding domain, and recent work by Strünker and co-workers presented evidence that the stimulation of human CatSper by these modified cAMP agonists is because of the chemical moieties modifying the cAMP molecule acting externally on CatSper (22,23).
To further evaluate this controversy, in the present work we used a combination of experimental strategies. First we showed that HCO 3 Ϫ and 8-Br-cAMP induced an increase in [Ca 2ϩ ] i in single cells. Both responses were blocked in the presence of the PKA antagonist PKI, as were the responses of 8-Br-cAMP with H89. Second, we used indirect recording of Em changes produced by chelating divalent cations with EGTA to measure CatSper activity (24). Using this methodology we showed that HCO 3 Ϫ stimulated CatSper and that this increased activity was blocked by the PKA inhibitors PKI and H89. Finally, the Em data were confirmed using whole-cell patch clamp ionic current recordings. Our results are consistent with a role of PKA in CatSper channel complex regulation. Notably, cAMP in the pipette solution, in whole-cell patch clamp recordings, strongly activates CatSper current and this effect is blocked by the PKA inhibitor PKI and H89. CatSper current was also stimulated by an alkalinization induced by NH 4 Cl, but this response was not sensitive, neither to PKI nor Rp-cAMPS, another specific PKA inhibitor. Consistent with these findings, HCO 3 Ϫ was unable to stimulate ionic currents in sperm from CatSper KO mice. Altogether our results strongly suggest that CatSper channels are activated by PKA by either direct or indirect phosphorylation, highlighting a connection between these two major signaling pathways required for sperm capacitation.

Bicarbonate increases [Ca 2؉ ] i and this response is inhibited by PKI, a PKA inhibitor
Capacitation involves cAMP production and Ca 2ϩ uptake. Because Adcy10 is regulated by HCO 3 Ϫ and permeable cyclic nucleotides have been shown to induce [Ca 2ϩ ] i increases in mouse sperm, we reexamined the [Ca 2ϩ ] i responses to these compounds. Addition of control vehicle in sperm loaded with the Ca 2ϩ -sensitive dye Fluo-3 did not elicit any Ca 2ϩ responses (Fig. 1A). In contrast, addition of HCO 3 Ϫ induced a reproducible increase in [Ca 2ϩ ] i (Fig. 1B). The effect of HCO 3 Ϫ upon sperm [Ca 2ϩ ] i can be explained by either an HCO 3 Ϫ induced stimulation of cAMP synthesis by Adcy10 (3,4) or by an HCO 3 Ϫ -dependent intracellular alkalinization with the consequent direct effect on CatSper activity, or furthermore by a combination of both pathways. Consistent with the first possibility, the HCO 3 Ϫ -induced increase in [Ca 2ϩ ] i was blocked by the myristoylated PKA peptide inhibitor PKI (Fig. 1C). Furthermore, in agreement with a role of a cAMP-dependent pathway, addition of 8-Br-cAMP (1 mM) induced an increase in [Ca 2ϩ ] i (Fig. 1D). Moreover, the effect of 8-Br-cAMP was also blocked by the PKA inhibitory peptide PKI (Fig. 1E). Average fluorescent traces and average maximal responses are summarized in Fig. 1, F and H and G and I, respectively. Although, as described previously (18,19,22), permeable cGMP also induces [Ca 2ϩ ] i increases ( Fig. 2A), this response was not blocked by PKI (Fig.  2B), suggesting that only the effect of the cAMP analogue is mediated by PKA. Average traces and average maximum responses are summarized in Fig. 2, E and G, respectively. Regarding changes in pH i , it is well established that pH i alkalinization can induce an increase in [Ca 2ϩ ] i . Consistently, addition of NH 4 Cl to sperm elicited an elevation in [Ca 2ϩ ] i (Fig. 2C); however, contrary to the HCO 3 Ϫ effect on [Ca 2ϩ ] i , the NH 4 Clinduced increase was not blocked by PKI (Fig. 2D). These findings are summarized in Fig. 2, F and H.

The 8-Br-cAMP induced [Ca 2؉ ] i increase is also blocked by H89, another PKA inhibitor
Considering that some undesired effects of PKI have been described (26) and to confirm the involvement of PKA, we tested another PKA inhibitor, H89, which is a ATP-binding site competitive inhibitor. As shown in Fig. S1, the increase in [Ca 2ϩ ] i caused by adding 1 mM 8-Br-cAMP ( Fig. S1A) was prevented by 10 M of H89 (Fig. S1B). Average fluorescent traces and average maximal responses are summarized Fig. S1, E and G, respectively. These findings agree with the hypothesis that PKA is involved in the cAMP-induced increases in [Ca 2ϩ ] i in mouse sperm. The [Ca 2ϩ ] i increase induced by 10 mM NH 4 Cl (Fig. S1C) was also inhibited by 10 M H89 (Fig. S1D). These observations are summarized in Fig. S1, F and H, respectively.

Bicarbonate stimulates the EGTA-induced mouse sperm Em depolarization in a PKA-dependent manner
the EGTA-induced depolarization was due mainly to this CatSper property and the amplitude of this response can be used to evaluate CatSper activity in a variety of conditions. Consistently, CatSper inhibitors blocked the EGTA-induced response (24,28). Although indirect, this assay has the advantage of being straightforward and it is useful to design more demanding electrophysiological experiments. Therefore, to evaluate CatSper activity, Em of noncapacitated sperm was mea-sured with a fluorescent cyanine dye before and after adding 3.5 mM EGTA to suddenly decrease external [Ca 2ϩ ] below 30 nM in the external media. As a result, the sperm resting Em (Ϫ45 mV) was significantly depolarized as Na ϩ influx occurs mainly through CatSper. As predicted, the depolarizing response to EGTA was Na ϩ -dependent (Fig. S2, A-D).
As HCO 3 Ϫ activates Adcy10, to analyze the extent by which the cAMP pathway was involved in CatSper regulation, we first

PKA activates CatSper
assayed the effect of adding 25 mM HCO 3 Ϫ to sperm on the EGTA-induced depolarization amplitude (Fig. 3A). Previously, we had shown that HCO 3 Ϫ addition induced Em hyperpolarization in sperm ( Fig. 3B) (4). Once fluorescence reached the new steady state, EGTA was added and the depolarization amplitude measured. The experiments show that the EGTAinduced depolarization was increased when sperm were previously exposed to HCO 3 Ϫ (Fig. 3B). Furthermore, 10 M of PKI ( Fig. 3C) and H89 (Fig. 3D), two PKA inhibitors, essentially eliminated the HCO 3 Ϫ stimulation. Confirming that EGTAinduced depolarization is mainly because of CatSper, 10 M HC-056456 (HC), a CatSper blocker (29), inhibited the HCO 3 Ϫ stimulated response (Fig. 3E). A summary of these experiments is presented in Fig. 3F showing the average differences between Em ϩ compound (Em D ) minus Em resting, corrected by the initial HCO 3 Ϫ induced hyperpolarization.

Bicarbonate increases CatSper activity by stimulating PKA
The results described above on the effects of HCO 3 Ϫ and cAMP on [Ca 2ϩ ] i , and of HCO 3 Ϫ on the amplitude of the EGTA-induced depolarization, as well as their inhibition by PKI and H89, suggest that PKA may be involved in the regulation of CatSper. To further evaluate this hypothesis, we conducted whole cell patch clamp electrophysiological measurements by attaching a patch pipette to the sperm plasma membrane cytoplasmic droplet (30). As indicated earlier, CatSper monovalent currents can be measured by reducing divalent cations from the recording media to less than 30 nM (30). These monovalent currents are because of the permeation of Na ϩ and Cs ϩ ions. For these experiments, we used a voltage ramp protocol from Ϫ80 to ϩ80 mV (Fig. 4A). In these conditions, we measured typical monovalent CatSper currents obtained in control conditions and after perfusing HCO 3 Ϫ (25 mM) in the same sperm (Fig. 4B). Results from five independent experiments indicate a significant HCO 3 Ϫ -induced increase in I CatSper currents (Fig. 4C). In addition, the HCO 3 Ϫ -induced increase of these currents was suppressed with the CatSper inhibitor HC (Fig. 4

, D and E).
It is known that the addition of HCO 3 Ϫ may lead to a pH i increase either directly or possibly mediated by the sperm-specific Na ϩ /H ϩ exchanger stimulated by hyperpolarization or by the rise in cAMP (4,31,32). To discard these possibilities I CatSper was recorded substituting external Na ϩ for Cs ϩ because Na ϩ /H ϩ exchangers do not operate with Cs ϩ (33) and, as in other experiments, including 20 mM HEPES in the internal solution for buffering pH i changes. Under these conditions, an Na ϩ -free medium, the possible pH i increase induced by HCO 3 Ϫ , is effectively eliminated; however, when the cell was exposed to HCO 3 Ϫ , I CatSper increased (Fig. 4F). These findings, summarized in Fig. 4G, indicate that the stimulation caused by HCO 3 Ϫ is not because of any pH i increase it could cause. Finally, the role of CatSper mediating the HCO 3 Ϫ -induced current stimulation was further analyzed using sperm from CatSper null mice. In sperm from the null mice, the stimulatory effect of HCO 3 Ϫ in current measurements was not observed Ϫ induced hyperpolarization (⌬Em (Em D Ϫ Em R )).

PKA activates CatSper
( Fig. 4, H and I). Altogether, these results indicate that the HCO 3 Ϫ -induced increase in currents are mediated by CatSper.

The intracellular cAMP/PKA pathway contributes to I CatSper stimulation
If HCO 3 Ϫ stimulates I CatSper through the action of Adcy10 and the cAMP pathway, its effect should be inhibited by PKI, which specifically eliminates PKA activity. Therefore, sperm were exposed to PKI (10 M) and then HCO 3 Ϫ was added. In these conditions, the HCO 3 Ϫ -induced I CatSper stimulation was completely inhibited (Fig. 5, A and B). On the other hand, as described by others (30,34), addition of NH 4 Cl (10 mM) significantly stimulated I CatSper currents (Fig. 5C). However, contrary to the addition of HCO 3 Ϫ , the NH 4 Cl stimulatory effect was not abrogated in the presence of PKI (10 M) (Fig. 5D). These loss of function experiments strongly suggest that the HCO 3 Ϫ effect on I CatSper is mediated by PKA.
To further test this possibility, gain of function experiments using cAMP agonists were conducted. It has been proposed that the effect of chemically modified permeable cyclic nucleotides on human CatSper is not because of the cyclic nucleotide part of the molecule but to the chemical moiety attached to the nucleotide acting externally on the channel and not through intracellular pathways. Therefore, to test that cAMP is acting inside, the nucleotide was directly introduced into the sperm through the patch clamp pipette. Initially, we tested 1 mM cAMP (not shown) and thereafter found that 100 M elicited a similar response. To prevent cAMP degradation, the bath solution contained the phosphodiesterase inhibitor IBMX (100 M). Under these conditions I CatSper was stimulated

PKA activates CatSper
by 100 M cAMP (Fig. 5, E and F). Again, if PKA mediates the cAMP stimulation of I CatSper , it would be expected that PKI would inhibit this response, and indeed it does (Fig. 5, E and F).
In another set of experiments we tested H89, another PKA inhibitor. The addition of 10 M of H89 inhibited I CatSper previously stimulated by 25 mM of HCO 3 Ϫ (Fig. 6, A and B). As expected, H89 also inhibited I CatSper stimulated by cAMP inside the pipette. Consistent with our results measuring [Ca 2ϩ ] i , adding H89 was also able to eliminate the I CatSper stimulated by alkalinization with 10 mM of NH 4 Cl (Fig. 6, C and D).
The unexpected inhibition of the NH 4 Cl-induced stimulation of I CatSper suggested an additional target (Fig. 6, E and F).
To contend with this lack of specificity we examined the effect of Rp-cAMPS, a membrane-permeable and specific inhibitor of the activation by cAMP of cAMP-dependent protein kinase I and II. This analog binds to the PKA regulatory subunit and prevents its dissociation from the catalytic subunit and is also resistant toward cyclic nucleotide phosphodiesterases (35,36). This mode of inhibition differs from that of PKI, which is a competitive inhibitor of the catalytic subunit.
We found that Rp-cAMPS inhibited the stimulation of the current by HCO 3 Ϫ (Fig. 7, A and B). Nevertheless, the addition of Rp-cAMPS did not prevent the stimulatory effect of NH 4 Cl (Fig. 7, C and D). This result strengthens our hypothesis that the PKA pathway participates in the activation of I CatSper .
During capacitation, sperm are exposed to high HCO 3 Ϫ concentrations that elevate cAMP via sAC stimulation and activate the PKA pathway. Here we presented evidence consistent with the proposal that PKA is able to stimulate CatSper directly or through another protein it phosphorylates (Fig. 8). The model shows that in addition to voltage and pH alkalinization, CatSper can also be activated by a PKA-dependent phosphorylation event.

Discussion
Sperm gain the ability to fertilize metaphase II-arrested eggs in the female tract in a process known as capacitation. At the molecular level, capacitation is initiated as soon as spermatozoa are exposed to HCO 3 Ϫ anions which stimulate Adcy10 and produce a fast elevation of cAMP levels. In addition to cAMP,

PKA activates CatSper
sperm capacitation is also associated with an increase in [Ca 2ϩ ] i . Although several Ca 2ϩ channels have been proposed to be present in mammalian sperm (13,37,38), only two of them have been conclusively shown to be present in these cells using electrophysiological recordings combined with knockout genetic models (30,39,40). One of them is the purinergic receptor P2X2 (40). The second is the Ca 2ϩ channel complex CatSper (18). Whereas P2X2 KO mice are fertile, lack of CatSper results in complete sterility (14,41,42). Recently, we have shown that the lack of hyperactivation and infertility phenotypes can be overcome elevating [Ca 2ϩ ] i pharmacologically by exposing CatSper KO sperm to A23187 for a short period of time (43). Although the role of CatSper is well established in sperm function, how this channel complex is regulated is still not fully understood.
As first described by Weyand's lab (44), it is well established that cGMP and cAMP permeable agonists induce [Ca 2ϩ ] i ele-vation in mammalian sperm. Moreover, in bovine sperm, caged cyclic nucleotides were shown to increase [Ca 2ϩ ] i (44,45). Originally, this effect was attributed to the presence of a testisspecific transcript of a cyclic nucleotide gated channel (46). Interestingly, the cyclic nucleotide-dependent stimulation of [Ca 2ϩ ] i increase was not observed in sperm from CatSper KO mice (18), indicating that this channel complex mediates cyclic nucleotide effects in [Ca 2ϩ ] i . However, none of the CatSper subunits contain consensus sequences for cyclic nucleotidebinding domains, suggesting that the effect of permeable cyclic nucleotides is not direct.
In a landmark manuscript, Kirichok et al. (30) showed that whole cell currents could be recorded in mouse sperm. CatSper was readily recorded in divalent-free external solutions where it efficiently transports monovalent cations such as Na ϩ and Cs ϩ . In this work, the authors did not observe current stimulation by 8-Br-cAMP and 8-Br-cGMP (Fig. S3 in Ref. 30). However, they

PKA activates CatSper
performed these experiments in the presence of divalent cations, a condition where the currents are very small. Furthermore, these tests were carried out at pH i 6, where CatSper is not very active or at pH i 8, where it is fully active before adding the cyclic nucleotides. Possibly these conditions explain why no changes in the current amplitude were observed.
Other electrophysiological and [Ca 2ϩ ] i imaging studies reported that 8-Brc-GMP and 8-Br-cAMP did stimulate mouse monovalent currents and Ca 2ϩ uptake. The 8-Br-cGMP response was sensitive to diltiazem, which the authors interpreted as indicating the participation of cyclic nucleotidegated (CNG) channels. The 8-Br-cGMP response was also inhibited by PKG antagonists, suggesting the involvement of this kinase in their regulation (47,48). Although small molecular weight inhibitors often have off-target effects on other kinases, these experiments suggest that cyclic nucleotide effects are mediated by phosphorylation.
More recently, the effect of permeable cyclic nucleotides was attributed to binding of the chemically modified cGMP and cAMP to extracellular human but not mouse CatSper domains (22,30) and not by modulation of intracellular cGMP-or cAMP-dependent pathways. The authors showed that although exogenous addition of 8-Br-cGMP or 8-Br-cAMP elevated [Ca 2ϩ ] i when cGMP or 8-Br-cGMP were directly introduced inside the sperm through a patch clamp pipette, these compounds failed to elevate [Ca 2ϩ ] i . A similar experiment introducing cAMP inside sperm was not reported. Considering that many other compounds with different chemical structure such as odorants, menthol, progesterone, and permeable cyclic nucleotides also stimulate CatSper currents (22); these authors suggested that CatSper can act as a polymodal sensor for chemical cues. Comparative studies between mouse and human CatSper have revealed species-specific differences. Although human CatSper currents can be stimulated indirectly by progesterone (23,49,50), this hormone does not activate the mouse CatSper channel complex (49). On the other hand, pH i alkalinization can directly regulate CatSper activity in both species. Despite these advances, the role of cAMP-dependent pathways in the regulation of CatSper is controversial.
As mentioned in the Introduction, there is also contrasting evidence regarding the ability of HCO 3 Ϫ to induced an increase in [Ca 2ϩ ] i (18,19). In this last paper, this response was considered not significant, although observing Fig. 2 in Ref. 19, a  During capacitation sperm are exposed to high HCO 3 Ϫ concentrations that stimulate, via sAC, cAMP increases which activate PKA that stimulates CatSper, directly or indirectly. NBC represents the Na ϩ /HCO 3 Ϫ cotransporter that transports HCO 3 Ϫ into the cell. We propose that in addition to voltage and alkaline pH i , the PKA pathway also contributes to CatSper activation.

PKA activates CatSper
change caused by HCO 3 Ϫ can clearly be seen which is not present in sperm from CatSper null mice.
The controversy surrounding the response to permeable analogs of cAMP and cGMP and HCO 3 Ϫ in mouse sperm motivated us to reevaluate these matters. To examine our hypothesis that the cAMP/PKA pathway regulates CatSper, this channel complex activity was tested using a battery of assays. First we examined how [Ca 2ϩ ] i responded to conditions that increase PKA activity such as addition of HCO 3 Ϫ and permeable cAMP agonists. As shown before by several groups (18, 25, 51, 52), 8-Br-cAMP elevated [Ca 2ϩ ] i and similar results were observed when HCO 3 Ϫ was added in single sperm loaded with Fluo-3, a Ca 2ϩ indicator dye. Consistent with our hypothesis, myristoylated PKI peptide, which specifically blocks PKA activity, inhibited both the HCO 3 Ϫ and the 8-Br-cAMP responses. On the other hand, PKI did not block the [Ca 2ϩ ] i elevation caused by 8-Br-cGMP or by the NH 4 Cl-induced pH i alkalinization. Importantly, the lack of PKI inhibitory effect when [Ca 2ϩ ] i elevation was induced by 8-Br-cGMP and NH 4 Cl, which do not activate PKA, indicates that the PKI effect is not unspecifically targeting CatSper. To strengthen the evidence indicating the involvement of PKA in the regulation of CatSper, we also used H89, another PKA inhibitor. As anticipated, H89 inhibited the [Ca 2ϩ ] i increase induced by 8-Br-cAMP, supporting our hypothesis that CatSper can be regulated through the cAMP/ PKA pathway.
CatSper efficiently transports monovalent cations in the absence of external Ca 2ϩ (32). Thus, when EGTA is added to sperm in noncapacitating media in sufficient amount to lower external Ca 2ϩ to Ͻ30 nM, a depolarization results which is indirectly related to CatSper's activity (24,27). This strategy conveniently allows sperm population studies in multiple conditions before embarking on more time-consuming electrophysiological recordings. Using this methodology, we determined that HCO 3 Ϫ enhances the EGTA-induced depolarization and that this response was blocked by either PKI, H89, or HC, a CatSper blocker (29). Our findings are consistent with the hypothesis that PKA-dependent phosphorylation pathways are involved in the regulation of CatSper activity.
To further examine this hypothesis, sperm exposed to noncapacitating conditions were patch clamped and currents measured. As mentioned earlier, CatSper is optimally studied electrophysiologically by recording monovalent currents when divalent cations are eliminated from the recording media (30). Basal currents were recorded from mouse sperm suspended under this condition by applying a voltage ramp. HCO 3 Ϫ addition following this first recording induced a significant increase in the observed currents that were blocked in the presence of either PKI, H89, or HC. Because unexpectedly we found H89 also inhibited the NH 4 Cl-induced I CatSper stimulation, indicating a certain lack of specificity, we also tested Rp-cAMPS, an additional specific PKA inhibitor. As this inhibitor blocked the HCO 3 Ϫ stimulation of I CatSper but did not influence its enhancement by NH 4 Cl, the results using Rp-cAMPS support the participation of PKA in the regulation of this channel. Indeed, two specific PKA inhibitors acting by distinct mechanisms, PKI and Rp-cAMPS, significantly diminished the HCO 3 Ϫ stimulation of I CatSper .
To further establish the target of the HCO 3 Ϫ stimulation leading to this current increase, experiments were carried out in sperm from CatSper null mice. As anticipated, this anion was unable to cause a current increase in these sperm suspended in divalent-free media, clearly suggesting that CatSper is the target. Therefore, pharmacological and genetic loss of function experiments are consistent with the hypothesis that PKA-dependent pathways up-regulate CatSper activity. In addition, gain of function experiments showed that introducing cAMP inside the sperm using the patch pipette also induced a significant increase in current amplitude that was blocked by PKI and H89.
All the findings presented here are consistent with our initial hypothesis that HCO 3 Ϫ enters sperm, activates sAC, and elevates cAMP which stimulates PKA. Once active, we show that this kinase directly or indirectly regulates CatSper activity by phosphorylation, enhancing its capacity to conduct and elevate [Ca 2ϩ ] i , a necessary condition for capacitation including hyperactivation and preparation for the acrosome reaction. Our findings support the central role cAMP/PKA and CatSper play in the maturation process required so sperm may fertilize the egg.

Materials and methods
Products and sources were as follows: Protein kinase A inhibitor (PKI) fragment 14 -22, myristoylated trifluoroacetate salt, and Rp-adenosine 3Ј,5Ј-cyclic monophosphorothioate triethylammonium salt (Rp-cAMPS) (Sigma-Aldrich), and H89 dihydrochloride (Cayman Chemical, Ann Arbor, MI), 3-isobutyll-methylxanthine hydrate (IBMX) (Calbiochem), CatSper blocker HC-056456 (VITAS-M Laboratory) (29), 3Ј,5Ј-cAMP sodium salt (cAMP) (Sigma-Aldrich), 8-bromoguanosine 3Ј,5Јcyclic monophosphate sodium salt monohydrate cGMP (Sigma-Aldrich). NaHCO 3 and all other salts were from Sigma-Aldrich. In all cases aliquots were diluted and added to the recording solutions and volumes chosen so that the maximum concentration of solvent was at most 1% (v/v). After 10 min, sperm in suspension were washed in 10 ml of the same medium by centrifugation at 800 ϫ g for 10 min at room temperature (room temperature ϭ 24°C). Sperm were then resuspended to a final concentration of 2 ϫ 10 7 cells/ml and diluted 10 times in the appropriate medium depending on the experiment performed. The pH was maintained at 7.4.

Intracellular calcium ([Ca 2؉ ] i ) measurements
Aliquots of motile noncapacitated sperm were loaded with 4 M of the fluorescent Ca 2ϩ indicator Fluo-3 AM in the presence of 0.05% pluronic acid for 30 min. Afterward, they were washed once at 2000 rpm/8 min and resuspended in noncapacitating TYH media. Sperm were attached on laminin (1 mg/ml) precoated cover slips, allowing their flagella to move continuously. The coverslip was mounted on a chamber (Harvard Apparatus) and placed on the stage of an inverted microscope (Eclipse TE 300; Nikon). Sperm were exposed to TYH medium alone (MR) for control experiments or to 1 mM 8-Br-cAMP, 10 M PKI, 1 mM 8-Br-cGMP, 10 mM NH 4 Cl for test experiments. Inhibitors such as PKI and H89 (10 M), were incubated 5 min before adding nucleotides or 10 mM NH 4 Cl. As vitality control 10 M ionomycin was added at the end of each experiment. In these experiments at least 150 individual mouse sperm were analyzed per treatment from each of three different mice. Fluo-3 loaded samples were excited with a Blue LED (3.15 A, Luminus Devices, Woburn, MA), with a band pass excitation (HQ 480/40X) filter, dichroic mirror (Q505lp), and emission (HQ 535/50M) filters (Chroma Technology, Bellows Falls, VT). The LED output was synchronized to the exposure out signal of an iXon 888 CCD camera via the control box to produce a single flash of 2-ms duration per individual exposure. The camera exposure time was set equivalent to the flash duration (2 ms). Images were collected every 500 ms using iQ software (Andor Technology). Images were processed and analyzed with macros written in ImageJ (Version 1.38, National Institutes of Health). Regions of interest were drawn on each sperm head and then analyzed for quantification. A plot was generated in Microsoft Office Excel 2007 (Microsoft). Fluorescence is expressed as F-F0/F0.

Membrane potential measurements
Em measurements were performed following the protocol described previously in detail (4,27). Briefly, mature sperm from caudal epididymides were collected, diluted in noncapacitating TYH medium, and exposed to a final concentration of 1 M Em-sensitive dye 3,3Ј-dipropylthiacarbocyanine iodide (DiSC 3 (5)) for 5 min. Mitochondrial membrane potential was dissipated with 500 nM carbonyl cyanide m-chlorophenylhydrazone (CCCP), and sperm were incubated for 2 additional minutes. After this period, 1.5 ml of the sperm suspension was transferred to a gently stirred cuvette at 37°C and the fluorescence monitored using an Ocean Optics USB4000 spectrofluorometer operated by Spectra Suite (Ocean Optics, Largo, FL ) at 620/670 nm excitation/emission wavelength pair (27). Cell hyperpolarization decreases the dye fluorescence. Recordings were initiated after reaching steady-state fluorescence (1-3 min) and were converted to Em as described previously (27). Calibration was performed by adding 1 M valinomycin and sequential additions of KCl. The equilibrium potential for K ϩ was calculated with the Nernst equation considering intracellular mouse sperm K ϩ is 120 mM.

Caudal sperm preparation for electrophysiology
Cauda epididymal sperm were obtained from CD-1 WT and from C57BL/6 CatSper1 KO mice for swim out in noncapacitating TYH medium. Sperm were stored in physiological solu-tion (HS) with the following composition in mM: 135 NaCl, 5 KCl, 2 CaCl 2 , 1 MgSO 4 , 10 lactic acid, 1 pyruvic acid, 5 glucose, 20 HEPES, pH 7.4, at 4°C until assayed. At the desired time, 100 l aliquots of the cell suspension were dispensed into a recording chamber (1 ml total volume) and subjected to electrophysiological recording.

Electrophysiology
Whole-cell macroscopic currents were obtained by patch clamping the sperm cytoplasmic droplet (30). All recordings were performed using a patch clamp amplifier (Axopatch 200, Molecular Devices) at room temperature (22°C). Pulse protocols, data acquisition, and data storage were performed using pCLAMP6 software (Molecular Devices) and data analysis was carried out with Clampfit 10.6 (Molecular Devices), Origin 7.5 (Microcal Software), and Sigma Plot 10 (Systat Software). Current records, unless indicated otherwise, were acquired at 20 -100 kHz and filtered at 2-5 kHz (internal four-pole Bessel filter) using a computer attached to a DigiData 1200 (Molecular Devices). Patch pipettes were pulled from borosilicate glass (Kimble; Queretaro, México) and had a final resistance between 15 and 20 megaohms. Initial experiments were carried out using physiological solutions (HS) in the bath and the pipette solution contained in mM: 110 Met-K, 30 KCl, 10 NaCl, 1 ATP-Mg, 1 CaCl 2 , 10 EGTA, 10 HEPES, pH 6.8. For recording CatSper currents we used a CatSper-recording solution, also called divalent cation-free solutions by Kirichok et al. (30); the bath solution contained in mM: 150 sodium gluconate, 2 Na 2 EDTA, 2 EGTA, 20 HEPES, pH 7.4, and the pipette solution contained in mM: 135 Cs-MeSO 3 , 5 CsCl, 5 Na-ATP, 10 EGTA, 20 HEPES, pH 7.0. The osmolarity of all solutions was adjusted with dextrose. The total currents under physiological conditions were recorded applying a voltage-step protocol from Ϫ100 mV to ϩ200 mV in 10-mV increments with a holding potential of 0 mV and lasting the time indicated in the figures. For CatSper current (I CatSper ) recording we used a conventional voltage-ramp protocol from Ϫ80 mV to ϩ80 mV with duration of 750 ms from a holding potential of 0 mV. Seals between the patch pipette and the cytoplasmic droplet in sperm were formed in HS bath solution and after achieving the whole cell configuration the bath solution could be changed for divalent cation-free solution.

Data analysis
Data are given as mean Ϯ S.E. In all cases, differences between raw experimental and control data were tested by Student's t tests if not otherwise stated. Differences were considered significant when *, p Ͻ 0.05; **, p Ͻ 0.01; and ***, p Ͻ 0.001.