A Requirement of Bicarbonate for Ca2’-induced Elevations of Cyclic AMP in Guinea pig Spermatozoa*

Ca4+ causes less than 2-fold elevations of guinea pig sperm cyclic AMP concentrations when cells are incubated in a minimal culture medium in the absence of bicarbonate (Hcoa-). However, in the presence of HCOJ-, Ca2+ increases cyclic AMP by as much as 25- fold within 1 min. The (ea2+, HCOs-)-induced elevations occur in either the presence or absence of the permeant anions, pyruvate and lactate. In the absence of extracellular ea2+, HC03- elevates cyclic AMP only slightly. The effect of HCO3- is concentration-dependent, with maximal responses obtained at concentrations of greater than 25 mM. Ca”’ (25 mM HCOs-) at concentrations of less than 100 p~ causes one-half-maximal ele- vations of cyclic AMP. The (Ca”, HCOs-)-induced elevations of cyclic AMP are observed at various extra- cellular pH values (7.5-8.5) and in the presence or absence of extracellular Na+ 01- K+. NEW1 does not elevate sperm cyclic AMP concentrations and does not greatly alter the (Ca2+, HCOs-)-induced elevations. The putative Caz+ transport antagonist, D-600 pM), completely blocks the (ea2+, HCOs-)-induced elevations of cyclic AMP. A23187, in the presence but not in the absence of extracellular ea3+, increases sperm cyclic AMP but does not further elevate cyclic AMP in HC03” treated cells. These studies establish that Ca2+-dependent elevations

it has been suggested that increased Ca" uptake results in the direct or indirect activation of the sperm adenylate cyclase (5, 7 , 9).
In 1979, we demonstrated (10) that Ca2+ could cause marked elevations of guinea pig sperm cyclic AMP concentrations, and subsequent work demonstrated that Ca2+ (micromolar concentrations) could activate the Mg'+-dependent adenylate cyclase of guinea pig spermatozoa (11). Ca" also has been reported to cause elevations of cyclic AMP in hamster spermatozoa (12). Since sperm cells may not contain a functional GTP regulatory component (5,13), the regulation of adenylate cyclase by Ca2" could represent a primary mechanism of enzyme regulation in these cells.
In studies designed to determine requirements for the Cazcinduced elevations of cyclic AMP in guinea pig spermatozoa, we discovered that Ca2+ failed to elevate cyclic nucleotide concentrations in the absence of bicarbonate (HCO.y-). We suggest here that bicarbonate ion is a requisite for the specific uptake or membrane binding of low concentrations of Ca2'; this uptake or binding results in large elevations of cyclic AMP and could account, in part, for the reported stimulation of sperm motility by HCO;" in some animals (14-16).

EXPERIMENTAL PROCEDURES
~ateria~s-Common chemicals were of the highest reagent grade available and were purchased from Sigma Chemical Co. or Fisher Scientific Co. D-600 was from Knoll A-G, and I-metbyl-3-isobutyixanthine was from Dr. J. N. Wells (Department of Pharmacology, Vanderbilt University Medical Center).
Media-A solution containing 102 m%f NaCI, 1.7 mM CaCIz, 0.25 mpx sodium pyruvate, 20 mM sodium lactate, 1 mM MgC12, and 25 mM triethanolamine buffer at pH 8.0 was used as the basic minimal cuiture medium. The medium was modified in various experiments to contain various concentrations of NaHCOa, 10 mM glucose for p pvate and lactate, 25 mM NaKC03 to replace the triethano~amine of CaC12. When '%'2aa* was added, approximately 4 X 10'' dpm of buffer, 1 mM l-methyl-3-~obuty~anthine, or various concentrations 45ca2c were added to each incubation vessel. For short incubation time periods (1 min), the pH of the various media remained at pH 8.0. When media were gassed with 955 (air/CO& the pH was maintained at approximately 7.7 (25 m~ NaHCOd. In some cases, choline chloride was substituted for NaCI, pyruvic acid for sodium pyruvate, lactic acid for sodium lactate, and KHCOJ for NaHC03. Sperm Preparation and Incubation Procedure-Spermatozoa were prepared and washed as previously described (10) except in some cases where choline chloride (Na+-free experiments) was used to wash the sperm cells.
Reaction vessels were incubated with either air or air/COe (955) as the gas phase. In general, cells collected in 154 m~ NaCJ were added immediateIy after collection to an incubation vessel containing the minimal culture medium (total volume = 1 ml). After incubation at 37 "C, the reactions were stopped (cyclic AMP) with 1 ml of 0.5 N perchloric acid. When 45Ca2' uptake was measured, reactions were stopped by rapid filtration on Whatman GF/C glass fiber filters.
In some cases, cells were allowed to preincubate with the buffer for various periods of time at 37 "C prior to the addition of CaCl?.
cyclic Nucleotide Determinations-The acidified cell suspensions were frozen and thawed 3-5 times and the cyclic AMP was then purified on acid alumina and Dowex 50-(H') columns as previously described (17). The concentration of cyclic AMP was determined by radioimmunoassay (18) as modified by Harper and Brooker (19).
Cu2+ Uptake-At the end of the incubation, cells were diluted with 2 ml of 154 m~ NaCl and rapidly added to GF/C glass fiber filters. The cells were washed 3 times with the saline solution by vacuum filtration and the fiters were then counted for radioactivity.
HCOs-/C02 Equilibria-In the experiments where HCOJ-was added to the incubation vessels with air as the gas phase, the other species present would be COn and HzC03. However, in most cases, the time of incubation was less than 1 min and, therefore, the only major species present was HCOa-. Similarly, in some experiments, cells were gassed with air/CO? (955) in the presence of 25 m~ HCOaor in the presence of triethanolamine buffer (no HCOs-added). When the triethanolamine buffer was used, the incubation times were short such that the pH of the medium did not appreciably change. Thus, under these conditions, only small amounts of HCOlC were present.

RESULTS
Requirement of HCOs"-When spermatozoa were added to a minimal culture medium (pH 8.0) containing Ca2+ and HC0,-(25 mM), cyclic AMP concentrations were elevated by as much as 25-fold within 30 s (not shown). The concentrations were maximal between 15 s and 1 min and then declined rapidly toward basal concentrations. If Ca2+ was omitted from the culture medium, cyclic AMP concentrations were elevated less than 2-fold. These observations are similar to those reported previously (10). Unexpectedly, however, when HC03was omitted from the buffer, cyclic AMP concentrations were elevated less than 2-fold in response to Ca"+. Thus, the marked elevations of cyclic AMP in guinea pig spermatozoa appeared to require both extracellular Ca2+ and HC03-. Even when the sperm cell concentration was increased to 2 X 10' ceUs/d, both Ca'+ and HC03-were still required for the marked elevations of cyclic AMP (not shown).
The requirement of HC0,-for Ca"-induced elevations of cyclic AMP was concentration dependent (Fig. 1) and did not require the presence of exogenous substrates (Fig. 2). In other experiments, glucose also replaced pyruvate and lactate in the minimal culture medium without alteration of the HC03response curve. In these experiments, where the incubation time was 1 min or less, the pH of the medium was maintained m~ lactate, and 2.5 mM CaC12 at the pH values given on the abscissa. NaHCOS was added to give 0, 5, 10, or 25 mM concentrations and sperm (100 pl, to give 2 X IO6 sperm/ml) were then immediately added. Reactions (total volume = 1 ml at 37 "C) were stopped with perchloric acid at 30 s. The pH values of duplicate reaction vessels were determined at the end of the incubation. The pH varied less than 0.06 unit between the different HCO:; concentrations. constant at pH 8.0 in an air atmosphere. However, in other experiments, cells were incubated for up to 1 h in a 95:5 (air/ CO,) atmosphere a t pH 7.7 and were shown to remain completely respondent to Ca'+ with respect t o elevations of cyclic AMP (not shown). Thus, cells remain in a Ca"-responsive state when incubated in a HC03"based buffer, and HC0:3-, itself, cannot duplicate the effect of Ca2+ even after long time periods of incubation. When cells were incubated for up to 1 h in a minimal culture medium in the absence of HCOa-, they also did not acquire CaEt sensitivity (not shown).
Based on these results, it was postulated that the functional role of HC03-was either t o increase intracellular pH which might then be a requirement for a Ca'+-dependent response or to serve as a specific activator or counter-ion for Ca" uptake (or Ca'+ binding to the membrane). CO, rather than HCOBalso remained as a potential mediator of the Ca2+ response.
Extracellular pH and HC03--The effect of HC03-on cyclic AMP as a function of extracellular pH is shown in Fig.  3. In these experiments, the gas phase was air. The various Sperm Cyclic AMP HCOs-concentrations altered the initial pH (indicated on the abscissa) less than 0.06 unit over the 30-s time period of incubation. Although an apparent diminished response to HC0.3-was observed at the higher pH (8.5), the basal cyclic AMP concentrations also were slightly lower, and, thus, nearly equivalent responses were observed when expressed as percentages. Extracellular pH, then, appeared to have only slight effects on the (Ca'+-HC03-)-induced elevations of cyclic AMP. Intracellular pH and HC03"-C02 or HC0,is known to alter intracellular pH (20), and it, therefore, seemed possible that the effect of Ca" on cyclic AMP might require an altered intracellular pH. The penneant anions, pyruvate and lactate, would be expected to lower intracellular pH (20), but as already indicated, Ca2+ failed to elevate cyclic AMP concentrations in either the presence or absence of these substances when HC03-was absent. In other experiments, cells were buffered with 25 mM triethanolamine and gassed for 1-5 min with air/COz (95:5) followed by the addition of Ca2'. Under these short incubation conditions, where the amount of HCO:lformed was low and the Con should decrease intracellular pH, Ca" failed to significantly increase cyclic AMP (not shown). Thus, neither lowered intracellular pH nor COS in the absence of significant amounts of HC03-appeared capable of intereacting with Ca2+ to elevate cyclic AMP.
NH,', in contrast, causes increases in intracellular pH (20). Cells incubated in a minimal culture medium containing 10 mM NH&1 and 25 m~ NaHC03 for 10 min followed by the addition of Ca" showed slight increases in responsiveness to Ca2' (Fig. 4). These data could support the concept of an increased intracellular pH being required in order for Ca" to elevate cyclic AMP. However, when NH,CI was added in the absence of NaHC03-, it failed to elevate cyclic AMP concentrations in a Ca2'-dependent manner (Table I).
Monovalent Cations and HCO.?--Cells collected and washed in a sodium-free or a potassium-free, choline-substituted medium were much less responsive to Ca2+ and HCO3than cells prepared in physiological saline. Nevertheless, a Ca2'-dependent elevation of cyclic AMP was still observed ( Table I). The percent elevation of cyclic AMP caused by Ca2+ and HC03-was comparable for all treatments. In other experiments, the diuretic furosemide at concentrations of 50 p~ also failed to block the (Ca2+, HCOS-)-induced elevations of cyclic AMP (not shown). Adenylate Cyclase-To determine whether the Mg'+-dependent adenylate cyclase was affected by HC03-, cells were broken and washed as described previously (11). Enzyme was then incubated with various concentrations of Ca2+ (0-1.0 mM) and with various concentrations of NaHC03-(0-10 mM) for various periods of time. Although Ca2+ stimulated the adenylate cyclase 3-to 5-fold as reported in other studies ( l l ) , NaHCOa-did not alter the Ca2+ stimulation of the enzyme (not shown).
Calcium Transport-Since apparent alterations of intracellular pH did not appear to account for the HC03-requirement, the possibility of a HCO3"activated Ca" uptake system was investigated.
Low concentrations of extracellular Cap+ were capable of causing the large elevations of cyclic AMP (Fig. 5). CaZc at concentrations of less than 10 PM caused a positive response, 11.9 f 2.6 Picomoles of cyclic AMPIIO' cells. and concentrations of slightly less than 100 ELM caused onehalf-maximal elevations of cyclic AMP. D-600, a putative Ca2+ transport antagonist, completely blocked the (Ca", HCOs-)-induced elevations of cyclic AMP a t 5C-100 pM concentrations (Fig. 6).
A23187, the divalent cation ionophore, also was capable of causing elevations of cyclic AMP, although the elevations %"I 2y NoC\ were generally only 2-to 3-fold (Fig. 7). Nevertheless, its effects on cyclic AMP required extracellular Ca'+ and A23187 failed to elevate cyclic AMP further in the presence of HC03- (Fig. 7).
In final experiments, the uptake of 45Ca2f was estimated in the presence or absence of HCOa-. The effect of HC03-on Ca2+ uptake was variable and statistically not significant (not shown). A23187, in contrast, caused 2to 4-fold increases in 46Ca2+ uptake (not shown) under similar conditions.

DISCUSSION
Ca" was previously shown to cause marked elevations of cyclic AMP and to activate the Mg2+-dependent adenylate cyclase of guinea pig spermatozoa (IO,11) and to activate motility and elevate cyclic AMP of hamster spermatozoa (12). In the studies reported here, it is shown that one other component, HC03-, is absolutely required in order for Ca'+ to elevate cyclic AMP in intact guinea pig spermatozoa. The effect of HC03-does not appear to be directly on adenylate cyclase since it does not stimulate the adenylate cyclase in broken cell preparations. Instead, the function of HC03-appears to be related to the transport of Ca2+ since D-600 blocks the (Ca2+, HCOS-)-induced elevations of cyclic AMP, and A23187, at least in part, reproduces the effect of HCOs-. These data also suggest that multiple Ca2+ uptake systems exist in spermatozoa since Singh et al. (21) have demonstrated that D-600 can facilitate 45Ca" uptake by guinea pig spermatozoa. An argument against an effect of HC03-on Ca'+ transport is the failure to detect HCOs--induced increases in 45Ca2+ uptake. However, certain potential problems with the uptake experiments must be considered. 1) Sperm immediately bind approximately 75 pmol of 45Ca2'/107 cells when added to the Ca'+-containing medium; and 2) the uptake of Ca2+ (assuming it occurs) is probably very rapid and in low amounts. Thus, binding of 45Ca2+ to the sperm membrane itself may represent amounts high enough to prevent detection of a low capacity uptake system. Recently, Hefner and Storey (22) observed a positive effect of Ca2+ on the motility of mouse spermatozoa and also failed to detect Ca2+ uptake by three different methods. It is possible, as suggested by Hefner and Storey ( 2 3 , that a physiologically active Ca2+ pool exists directly on specific membrane sites. In this case, HCOs-might facilitate the binding of Ca2+ to those sites. The failure to observe significant HCO3"induced increases in *%a'+ uptake, however, also could be explained by a concomitant negative effect of HCO:,on '%a2+ uptake by mitochondria (23, 24).
The site of Ca" action to elevate cyclic AMP has not been firmly established, although the sperm adenylate cyclase appears to represent the most likely target. Methylxanthines synergize with Ca" to elevate cyclic AMP (IO) and Ca'+ activates the Mg"-dependent adenylate cyclase in guinea pig broken sperm cell preparations (11). Sperm cells contain high quantities of calmodulin (25-29) and adenylate cyclase is known to be activated by the Ca2+-calmodulin complex (30).
Nevertheless, we have not successfully activated the calmodulin-deficient adenylate cyclase of guinea pig spermatozoa (11). Recently,Lotersztajn et al. (31) have suggested that a Ca'+-dependent activator of a (Ca"', Mg'+)-ATPase from rat liver may not be calmodulin; this also could be the case for the sperm adenylate cyclase.
The results reported here emphasize the importance of buffer selection in intact cell studies. In the absence of HC03-, guinea pig sperm cells survive in vitro, but at least one response, the elevation of cyclic AMP, does not occur. Female oviduct fluids have been reported to contain high concentrations (35-90 mM) of HC03-(32, 33), and large elevations of cyclic AMP in response to Ca2+, therefore, normally would be expected. Bicarbonate also has been suggested to be an important component of the "natural" capacitation process (32) even though HC03-does not appear required for ea2+ uptake by capacitated spermatozoa (10).
Phosphodiesterase inhibitors and cyclic AMP are known to stimulate sperm metabolism and motility in a wide variety of animals including the guinea pig ( 5 ) . HC03-, likewise, has been reported to stimulate sperm metabolism and motility in various animals (14-16, 34). Although these effects of HC03could be due in part to fixation of COz into the tricarboxylic acid cycle, at least one report (16) has suggested that HC03stimulates spermatozoa without direct entry into metabolic pathways. If this is true, HC03-may stimulate spermatozoa by virtue of its primary effect on Ca2+ uptake. Subsequent events such as the elevations of cyclic AMP could explain the motility or metabolic activations. It also has been suggested that small increases in cytosolic Ca2' (35) or eazf binding to specific membrane sites (22) may directly stimulate sperm motility. In at least one case, hamster spermatozoa, extracellular Ca'+ appears to be absolutely required for motility (36).