Calmodulin-stimulated cyclic nucleotide phosphodiesterases in plasma membranes of bovine epididymal spermatozoa

ARCHIVESOF BIOCHEMISTRYAND BIOPHYSICS Vol. 262, No. 2, May 1, pp. 439-444, 1988

Calmodulin-Stimulated Cyclic Nucleotide Phosphodiesterases Membranes of Bovine Epididymal Spermatozoa’

in Plasma

PREM S. CHAUDHRY AND EDMUND R. CASILLAS’ Department of Chemistry, New Mexico State University, Las Cmces, New Mexico 88003 Received July lo,1987

Cyclic nuoleotide phosphodiesterase in the plasma membranes of bovine epididymal spermatozoa was stimulated by added Ca2+and calmodulin. The rate of hydrolysis and responsiveness toward calmodulin was greater for CAMP than for cGMP. The kinetic analysis of the activity revealed two forms of phosphodiesterase with apparent Km values of 7.5 and 95 PM for CAMP. Calmodulin stimulated both of the activities by increasing the V,,, without affecting the K,‘s. The activity response with respect to Ca2+ concentration appears to be biphasic in both the absence and presence of added calmodulin. Trifluoperazine inhibited the Ca2+-and calmodulin-sensitive enzyme activity in a dose-dependent manner. The calmodulin-stimulated phosphodiesterase activity in the sperm plasma membranes can be solubilized and absorbed to a Calmodulin-Sepharose affinity column in the presence of Ca2+. 0 1988 Academic press, rnc.

Cyclic nucleotide phosphodiesterases exist in multiple molecular forms in different tissues, each with its own kinetic properties, substrate specificity, and cellular distribution (l-3). One form is activated by calmodulin (2-5). The well-characterized form is a cytosolic enzyme, catalyzes the hydrolysis of both cyclic AMP and cyclic GMP, but exhibits a lower apparent Km for cyclic GMP than for cyclic AMP (2, 3, 5). In tissues where Ca2+/calmodulin regulates phosphodiesterase activity, calmodulin appears to be present in amounts greater than that required to fully activate the calmodulin-dependent phosphodiesterase activity. Since spermatozoa contain high levels of calmodulin (6-10) and multiple forms of cyclic nucleotide phosphodiesterase (ll-15), it is likely that one or more of the forms is activated by Ca2+/calmodulin. Earlier attempts to demonstrate this effect were mainly unsuccessful (11,12,15).

Recently, Wasco and Orr (16) reported that a particulate cyclic nucleotide phosphodiesterase, associated with heads and tails of detergent-demembranated rat spermatozoa, is stimulated by Ca2+ and calmodulin at low micromolar concentrations of CAMP. Cyclic nucleotide phosphodiesterase activity in the membrane fraction was not calmodulin sensitive. Here we provide evidence, for the first time, that calmodulin-stimulated cyclic nucleotide phosphodiesterase is associated with plasma membranes of bovine epididymal spermatozoa. The calmodulin-sensitive enzyme(s) are tightly bound to the membranes and it appears that both high and low substrate Km forms are present.

Preparation of plasma membranes. Spermatozoa were collected from bovine caudal epididymides as described by Henle and Zittle (17) in a medium of pH 7.2 with the following composition: 10 mM Hepes,

‘Supported by PHS HD10664, RR-08126, RR08136, and RR-07154. ’ To whom correspondence should be addressed.

aAbbreviations used: Hepes, 4-(2-hydroxyethyl)1-piperazineethanesulfonic acid; CaM, calmodulin.

METHODS

439

0003-9861/88 $3.00 Copyright All rights

0 1388 by Academic Press, Inc. of reproduction in any form reserved.

440

CHAUDHRY

AND CASILLAS

150 mM NaCl, 1 mM benzamidine. After the cells were washed twice and suspended in the same medium, plasma membranes were isolated by the method described by Noland et aZ. (18). The membrane pellet thus obtained was suspended and washed extensively with the hypotonic buffer described under Results. The final plasma membrane pellet was suspended in hypotonic buffer (5 mM Tris-HCl, pH 7.0, 0.5 mM EDTA, 1 mM mercaptoethanol) and stored in small aliquots at -70°C. Solubilizatim and pur$icatim of cyclic nucleotide phosphodtiterase from plasma membranes. Extensively washed membranes were suspended in 20 mM Tris-HCl, pH 7.4, containing 0.1 mM benzamidine. Detergent (NP-40) was added to a final concentration of 0.5%. This suspension was then incubated for 2 h at 4”C, mixed occasionally, and then centrifuged at 105,000~ for 2 h. The supernatant was adjusted to give a final concentration of 1 mM MgCIZ, 0.5 mM CaCl*, 1 mM dithiothreitol, 0.1 mM benzamidine and was then added to a lo-ml column packed with CaMSepharose. The affinity column was previously equilibrated with a solution containing 20 mM Tris-HCl, pH 7.4,l mM MgC12,0.5 mM CaC12,1 mM dithiothreitol, 0.1 mM benzamidine, and 0.01% NP-40. After the sample was loaded, 50 ml of the equilibration solution was passed through the column followed by about 50 ml of the same solution, but containing 0.2 M NaCI. Fractions obtained from this latter step were monitored for protein content and phosphodiesterase activity and elution was continued until these components were no longer detectable. Affinity bound, cyclic nucleotide phosphodiesterase activity was eluted by the addition of 25 ml of equilibration solution that also contained 0.2 M NaCl and 2 mM EGTA. CaC& was added to the eluate to yield a 3 mM concentration and the sample was then immediately concentrated to 1 ml by ultrafiltration. The sample was then stored in 50% glycerol at -20°C. Cyclic nucleotide phosphodiesterase assay. The activity was measured as described by Thompson et al

(19). Reaction mixtures contained 50 mM Tris-HCI, pH ‘7.5;5 mM Mg-acetate; 1 mM dithiothreitol; 0.2 or 1.0 mM either [@HIcAMP or [S-‘H]cGMP; and 4-10 pg membrane protein in a final volume of 200 ~1. Calmodulin Ca”, or EGTA were added as described under Results. RESULTS

When plasma membranes were washed seven times with hypotonic buffer (5 mM Tris-HCl, pH 7.0, 0.5 mM EDTA, 1 mM mercaptoethanol), cyclic nucleotide phosphodiesterase activity was increased 200% by added Ca2+ but was unresponsive to added calmodulin. Although this washing procedure reportedly removes calmodulin from membrane preparations obtained from other tissues (20), calmodulin concentrations in sperm membranes presumably were still in excess of those required for maximal activation of the phosphodiesterase. However, when the sperm membranes were washed seven times with an alternative solution (10 InM Tris-HCl, pH 7.0,2 mM EDTA, 2 InM EGTA, 10 mM mercaptoethanol, 2 InM benzamidine) phosphodiesterase activity was responsive to both added Ca2+and to added calmodulin (Table I). The calmodulin-stimulated specific activity was approximately the same as that of the Ca2+-stimulated activity of the unwashed membranes. Although the alternative washing procedure was used in experiments described below, subsequent studies have shown that only 2 mM EGTA is absolutely required for the removal of calmodulin from the mem-

TABLE I EFFECT OF Gas+ AND CALMODULIN ON PLASMA MEMBRANE CYCLIC NUCLEOTIDE PHOSPHODIESTERASE(S) Cyclic nucleotide phosphodiesterase activity (nmol cyclic nucleotide hydrolyzed/min/mg protein) Cyclic AMP phosphodiesterase 7%’ Stimulation

Cyclic GMP phosphodiesterase

% Stimulation

% Stimulation

% Stimulation

Additions

0.2 mM

200 MMEGTA 200 @dEGTA, 350 PM Ca’+ 200 PM EGTA, 350 PM Ca*‘, 10 pg CaM

2.14 4.43

107

3.27 8.35

155

0.86 1.20

40

0.87 1.58

82

6.54

206

11.93

265

1.80

110

2.11

143

1.0 mM

0.2 mM

1.0 mM

PHOSPHODIESTERASES

FROM

BOVINE

branes. However, since broken sperm preparations always contain active proteases, the inclusion of benzamidine, or other protease inhibitors, is recommended. In addition, activity in the membranes hydrolyzed both CAMP and cGMP; the rate of hydrolysis and responsiveness to calmodulin were both greater for CAMP than for cGMP under identical conditions (Table I). A kinetic analysis of the phosphodiesterase activity present in the plasma membrane reveals two regions in the substrate concentration vs activity relationship that yields straight lines upon Lineweaver-Burk analysis (Fig. 1). One region displays an apparent Km for CAMP of 7.5 ~.LMand the other an apparent Km of 95 PM. Activity in both regions is stimulated by exogenous calmodulin and the effector alters the activity by increasing the V,,,,, without affecting the Km. The effect of added Cazf on phosphodie&erase activity, measured in the presence of EGTA to buffer Cazf concentrations, was studied both with and without added calmodulin. The activity response with respect to Ca2+ concentration is bi-

FIG. 1. Double-reciprocal plot showing the effect of added bovine brain calmodulin on the membrane phosphodiesterase activity. Activity was assayed in the presence of 200 NM EGTA, 350 pM Ca’+, and 10 FM calmodulin (0). Basal (CaM-insensitive) activity was measured in the presence of 200 PM EGTA (0).

EPIDIDYMAL

100

441

SPERMATOZOA

200

300

400

500

rca2j+M FIG. 2. Effect of Ca2+ on membrane phosphodiesterase activity in the absence (0) or presence of 10 pg CaM (0). All the incubation mixtures contain 200 pM EGTA and varying concentrations of exogenously added Ca2’.

phasic in both the absence and presence of added calmodulin (Fig. 2). Results from other experiments showed that enzyme activity increases with increasing amounts of added calmodulin and maximal stimulation, in the presence of Ca2’, occurs when 10 pg calmodulin is added (Fig. 3). Half-maximal stimulation occurred with 1 pg of calmodulin. In addition the stimulatory effect of calmodulin was observed at all incubation periods for at least 30 min (Fig. 4). The calmodulin antagonist, trifluoperazine, inhibited in a concentration-dependent manner, the Ca2+-stimulated activity in both the presence and absence of calmodulin (Fig. 5). Half-maximal inhibition occurs with 60 PM trifluoperazine. The basal (EGTA-insensitive) phosphodiesterase activity is not affected by trifluoperazine. Finally, we also attempted to detergent-solubilize and then purify phosphodie&erase activity from washed plasma membranes. About 40% of the original activity was made soluble with a solution containing 0.5% NP-40. Activity, presumably calmodulin sensitive, was further purified by CaM-Sepharose affinity chromatography. The purified enzyme was stimulated about 400% by micromolar levels of

442

CHAUDHRY

AND CASILLAS

50 2

4

6

6

10

12

14

16

CALMODULIN.@G

FIG. 3. Effect of calmodulin on membrane phosphodiesterase activity. Activity measured in the presence of 200 PM EGTA represents basal activity (CaM-insensitive activity); activity measured in the presence of 200 PM EGTA and 350 pM Ca*’ represents activity due to endogenous CaM in the membrane preparations. Calmodulin-dependent activity was assayed in the presence of 200 pM EGTA, 350 pM Ca*+, and the indicated concentrations of CaM from bovine brain.

Ca2+but was not stimulated by calmodulin. Others have also reported loss of sensitivity to calmodulin in similar studies (21,22). Possible explanations for the lack of stimulation by added calmodulin in-

.E z? >e !-=6z;

100

150

TRIFLU0PERAZINE.pt.I

FIG. 5. Effect of trifluoperazine on CaM-stimulated membrane phosphodiesterase. The activity was assayed in the presence of 200 pM EGTA, 350 pM Ca*+, and 10 pg CaM (0) or 200 PM EGTA (0) and the indicated concentrations of trifluoperazine.

elude the chance that the purified enzyme preparation already contained calmodulin, leached from the column, as reported by Westcott et al. (21), or that residual detergent inhibits the response to the effector (22). In addition, residual phosphodiesterase activity remaining in the detergent-extracted membranes was no longer responsive to added calmodulin. For these reasons it is not possible to calculate recoveries or specific activities of the purified calmodulin-responsive enzyme. However, if one uses responsiveness to Ca2+ as a criterion for the presence of the enzyme, the specific activity of the purified activity was 11 times greater than that in the washed plasma membranes. DISCUSSION

Although the function of cyclic AMP in spermatozoa is not known for sure, substantial evidence indicates that it helps regulate motility and there are indications that it might be involved in other sperm 25 5 10 15 20 c 30 processes such as maturation, energy meTIMEMIN tabolism, and the acrosome reaction FIG. 4. Time course for CAMP hydrolysis by plasma (23-26). These processes might be regumembrane phosphodiesterase. Assays were perof formed in the presence of 200 PM EGTA (0 - 0); 200 lated, in part, by the concentration CAMP in localized areas in the cell and PM EGTA and 350 pM Ca2+ (0-- - 0); or 200 PM EGTA, 350 PM ca*+, and 10 pg CaM (0 - 0). Other assay these concentrations are determined by the relative activities of adenylate cyclase conditions are described under Methods.

PHOSPHODIESTERASES

FROM

BOVINE

and CAMP phosphodiesterase. Adenylate cyclase in mammalian sperm is reportedly localized both in the plasma membrane and in the axoneme (27, 28) while cyclic nucleotide phosphodiesterase activity is associated with the plasma membrane, the cytosolic fraction (16, 28, 29), and the particulate fraction of demembranated spermatozoa (16). It is possible that one or more of the latter enzymes is regulated by Ca’+/calmodulin. Our present results clearly demonstrate the presence of Ca2+/calmodulin-stimulated cyclic AMP phosphodiesterase activity in sperm plasma membranes. This activity was observed only after extensive washing of the membranes in buffers containing 2 mM EGTA. The responsiveness of the activity to Ca2+,without added calmodulin, suggests that even our extensive washing procedure did not remove all of the endogenous calmodulin. Recently Noland et al. (30) have shown that calmodulin is tightly associated with sperm plasma membranes. This indication that calmodulin is tightly bound to the plasma membranes may account for the failure of others (16) to show stimulation of phosphodiesterase activity in sperm plasma membranes by added calmodulin. Alternatively, this failure to demonstrate the effect might have been due to the presence of detergent which was used to remove the membranes from the cells. Our efforts to purify the enzyme reported here also support the identification of the calmodulin-sensitive enzyme in the membranes. As is the case for the enzyme from other cells, detergent-solubilized phosphodiesterase activity binds to a calmodulin affinity column in the presence of Ca2+ and is released by elution with EGTA-containing buffer. We intend to continue our efforts to obtain calmodulinfree phosphodiesterase from the membranes. It will then be possible to study the effect of “calmodulin-specific” inhibitors on the interaction between calmodulin and the soluble, purified phosphodiesterase(,s). Although chlorpromazine, trifluoperazine (31), and Smethyl-omethoxy-:l-methyl-3-isobutylxanthine (32) clearly inhibit the calmodulin-stimu-

EPIDIDYMAL

SPERMATOZOA

443

lated enzymes from other cells, studies have shown that chlorpromazine and trifluoperazine also interact with membrane molecules other than calmodulin and produce nonspecific alterations in the structure of biological membranes (33). One interpretation of our kinetic analysis of the enzyme activity (Fig. 1) is that two forms of calmodulin-stimulated phosphodiesterase are present in the membranes. These activities display apparent Km’s of 7.5 and 95 pM for CAMP and are both stimulated by added calmodulin which affects the V,,, of the activity but not the K,‘s. The biphasic stimulation of the activity by Ca2+, both with and without added calmodulin (Fig. 2) also supports the two enzyme hypothesis. As indicated in the legend (Fig. 2), the incubation mixture contained 200 pM EGTA and varying amounts of exogenously added Ca2+. Although free Ca2+ concentrations can easily be calculated, a discussion of the physiological significance of the results of such calculations is premature. In this regard, we know that sperm plasma membranes contain several calcium-binding proteins, other than calmodulin (34). In any case, our results suggest that CAMP degradation by one or more phosphodiesterases might be regulated by fluctuations in intracellular Ca2+concentrations. Finally, our results with the phosphodiesterase activity in plasma membranes satisfy all the criteria suggested for categorizing calmodulin-sensitive activities (35). It is stimulated by added calmodulin and micromolar concentrations of Ca2+ and inhibited by calmodulin antagonists and by Ca2+chelators in a reversible fashion. Recently calmodulin stimulation of both nonmammalian (36) and mammalian (37) sperm adenylate cyclase activity has been reported. Although, at first glance, it seems paradoxical that enzymes that catalyze the synthesis and breakdown of an intracellular messenger would both be stimulated by a common effector, Piascik et al. have described such a system for brain CAMP (38). In that case, it is apparent that although both the cyclase and phosphodiesterase are stimulated by calmodulin, their response to Ca2+ concen-

444

CHAUDHRY

trations is much different. It may be that a similar regulatory system obtains in spermatozoa as well.

AND

CASILLAS

19. THOMPSON, W. J., TERASAKI, W. L., EPSTEIN, P. M., AND STRADA, S. J. (1979) Adv. Cyclic

Nucleotide Res 10,69-92. 20. SCHULMAN, H., AND GREENGARD, P. (1978) Proo.

NatL Acad Sci USA 75,5432-5436. REFERENCES 1. APPLEMAN, M. M., THOMPSON, W. J., AND RusSELL, T. R. (1973) Adv. Cyclic Nucleotide Res. 3,

65-98. 2. WELLS, J. N., AND HARDMAN, J. G. (1977) Adv.

Cyclic Nucleotide Res. 3.119-143. 3. APPLEMAN, M. M., ARIANO, M. A., TAKEMOTO, D. J., AND WHITSON, R. H. (1982) Ha&b. Exp. Pharnnzool 58,261-300. 4. LIN, Y. M., CHEUNG, W. Y. (1980) in Calcium and Cell Function (Cheung, W. Y., Ed.), Vol. 1, pp. 79-10’7, Academic Press, New York. 5. CHEUNG, W. Y., AND STORM, D. R. (1982) Handb.

Exp. Ph.armacol 58,301-323. 6. BROOKS, J. C., AND SIEGEL, F. L. (1973) B&hem.

Biophys. Res. Commun 55,710-716. 7. JONES, H. P., BRADFORD, M. M., MCRORIE, R. A., AND CORMIER, M. J. (1978) B&hem Biophys.

Res. Commun 82,1264-1272. 8. GARBERS, D. L., HANSBROUGH, J. R., RADANY, E. W., HYNE, R. V., AND KOPF, G. S. (1980) J.

Reprod FertiL 59,377-381. 9. FEINBERG, J., WEINMAN, J., WEINMAN, S., WALSH, M. P., HARRICANE, M. C., GABRION, J., AND DEMAIUE, J. G. (1981) Biochim Biophys.

Actu 673,303-311. 10. BURGESS, W. H. (1982) J. Biol. Chem. 257, 1800-1804. 11. WELLS, J. N., AND GARBERS, D. L. (1976) BioL Reprod 15,46-53. 12. TASH, J. S. (1976) J. ReprocL FertiL 47,63-72. 13. BHATNAGAR, S. K., CHAUDHRY, P. S., AND ANAND, S. R. (1979) J. Reprod. Fertil. 56, 133-139. 14. STEPHENS, D. T., WANG, J. L., AND HOSKINS, D. D. (1979) BioL Reprod 20,483-491. 15. BHATNAGAR, S. K., AND ANAND, S. R. (1982) Bio-

chim Biophys. Ada 716,212-224. 16. WASCO, W. M., AND ORR, G. A. (1984) B&hem.

Biophys. Res. Commun 118,636+X2. 17. HENLE, G., AND ZITTLE, C. A. (1942) Amer. J. PhysioL 136,70-74. 18. NOLAND, T. D., OLSON, G. E., AND GARBERS, D. L. (1983) Biol Reprod 29,987-998.

21. WESTCOTT, K. R., LA PORTE, D. C., AND STORM, D. R. (1979) Proc. NatL Acad. Ski. USA 76, 204-408. 22. SHARMA, R. K., AND WANG, J. H. (1981) Biochem Biophys. Res. Commun 100,710-715. 23. HOSKINS, D. D., AND CASILLAS, E. R. (1974) in Advanced in Sex Hormone Research (Singhal, R. L., and Thomas, J. A., Eds.), Vol. 1, pp. 283-324, University Park Press, Baltimore. 24. HOSKINS, D. D., AND CASILLAS, E. R. (1975) in Handbook of Physiology, Section on Endocrinology (Astwood, E. B., and Greep, R. O., Eds.), pp. 453-460, American Physiological Society. 25. GARBERS, D. L., AND KOPF, G. S. (1980) Adv. Cyclic Nucleotide Res. 13,251-306. 26. TASH, J. S., AND MEANS, A. R. (1983) BioL Repral. 28,75-104. 27. PETERSON, R. N., RUSSELL, L., HOOK, L., BUNDMAN, D., AND FREUND, M. (1980) J. Cell Sci. 43, 93-103. 28. BHATNAGAR, S. K., CHAUDHRY, P. S., AND ANAND, S. R. (1982) B&him. Biophys. Acta 716,424-430. 29. SANO, M. (1976) Biochim. Biophys. Acta 428, 525-531. 30. NOLAND, T. D., VAN ELDIK, L. J., GARBERS, D. L., AND BURGESS, W. H. (1985) Gamete Res. 11, 297-303. 31. LEVIN, R. M., AND WEISS, B. (1976) Mol. PhurmawL 12,581-589. 32. LORENZ, K. L., AND WELLS, J. N. (1983) Mol.

PharmacoL 23,424-430. 33. ROUFOGALIS, B. D. (1982) in Calcium and Cell Function (Cheung, W. Y., Ed.), Vol. 3, pp. 129-159, Academic Press, New York. 34. CHAUDHRY, P. S., AND PETERSON, R. N. (1987)

BioL Reprod. 36,144. 35. CHEUNG, W. Y. (1980) Science 207,19-27. 36. KOPF, G. S., AND VACQUIER, V. D. (1984) J. BioL

Ch.

259,7590-7596.

37. GROSS, M., TOSCANO, D., AND TOSCANO, W. A. (1985) Fed Proc. 44,1692. 38. PIASCIK, M. T., LEWIS WISLER, P., JOHNSON, C. L., AND POTTER, J. D. (1980) J. BioL Chem. 255, 4176-4181.