Phosphoprotein Phosphatase Inhibits Flagellar Movement of Triton Models of Sea Urchin Spermatozoa

CELL STRUCTURE AND FUNCTION 10, 327-337 (1985) C by Japan Society for Cell Biology

Daisuke Takahashi1, Hiromu Murofushi, Koichi Ishiguro2, Jun Ikeda and Hikoichi Sakai Department of Biophysics and Biochemistry, Faculty of Science, University of Tokyo, Bunkyo-ku, Tokyo 113, Japan

ABSTRACT. Phosphoprotein phosphatase prepared from bovine cardiac muscle was used to study the roles of axonemal phosphoproteins in the flagellar motility of sea urchin spermatozoa. When isolated axonemes were incubated with cyclic AMP-dependent protein kinase, ƒÁ[32P]ATP and cyclic AMP, more than 15 polypeptides were phosphorylated. Most were dephosphorylated by treatment with phosphoprotein phosphatase. When Triton models of sea urchin spermatozoa were treated with phospho-

protein phosphatase followed by an addition of ATP, the flagellar motility of the models was drastically reduced in comparison with that of the untreated models. The motility of the phosphatase-treated Triton models was partially restored by an addition of cyclic AMP and cyclic AMP-dependent protein kinase. These data give strong support to the idea that the motility of eukaryo- tic flagella is controlled by a protein phosphorylation-dephosphorylation system.

For several years much attention has been paid to the control of flagellar and ciliary movement in eukaryotes by the cyclic nucleotide system. Morton et al. (24) reported that flagellar movement of hamster cauda epididymal spermatozoa could be induced by diluting the semen with a medium containing calcium and that the initiation of beat was accompanied by an increase in the concentration of cyclic AMP within the cells. Garbers et al. (8) and Hoskins (14) reported that cyclic AMP or inhibitors of cyclic nucleotide phosphodiesterase added exogenously to bovine epididymal spermatozoa not only stimulated energy metabolism but activated flagellar motility as well. In addition, the spermatozoa of a wide variety of animals have been shown to contain enzymes related to the cyclic nucleotides such as adenylate cyclase (2, 7, 12, 24), guanylate cyclase (5, 12), cyclic nucleotide phosphodiesterase (2, 12, 24) and phosphoprotein phosphatase (28). An especially large amount of cyclic AMP- dependent protein kinase is included in sperm cells (6, 15, 19, 20, 25). To investigate the role of the cyclic nucleotide system in the regulation of flagellar motility, we used the Triton models of sea urchin spermatozoa developed by Gibbons

1 Present address: Department of Radiation Biophysics , Faculty of Medicine, University of Tokyo, Bunkyo-ku, Tokyo 2 Present address: Mitsubishi-Kasei Institute of Life Science , 11 Minamioya, Machida, Tokyo

327 328 D. Takahashi, et al. and Gibbons (9). This system is of particular advantage because macromolecules related to cyclic nucleotide system, such as protein kinase and protein phosphatase, are easily introduced to the motile apparatus. We found that the cyclic AMP-dependent protein kinase and a protein factor prepared from a detergent-extract of sea urchin or starfish spermatozoa reactivated the motility of the Triton models of sea urchin spermatozoa (16, 26). A similar protein factor was found by Tash et al. (29). It was phosphorylated by cyclic AMP-dependent protein kinase and induced the movement of Triton models of dog spermatozoa. The need for protein phosphorylation to initiate the flagellar beat also has been shown in Triton models of bull (21), fish (23) and tunicate (27) spermatozoa. If phosphorylation is required to initiate or activate flagellar movement, dephosphorylation of phosphoproteins in the flagellum should affect motility. No one has yet done this type of experiment which is necessary for confirmation of the significance of protein phosphorylation in flagellar motility. We used the phosphoprotein phosphatase of bovine cardiac muscle (4) for our experiments because it is easily be purified and gives a higher yield than the enzyme present in spermatozoa (28). We found that protein dephosphorylation of the axonemal phosphoproteins drastically reduced the motility of freshly prepared Triton models of sea urchin spermatozoa and that the original motility was partially restored by an addition of cyclic AMP and protein kinase. This result supports the belief in the significance of protein phosphorylation in the control of flagellar motility in spermatozoa and sheds light on unknown functions of the phosphoprotein phosphatase present in spermatozoa.

MATERIALS AND METHODS

Preparation of Triton models of sea urchin spermatozoa. Spermatozoa from the sea urchins, Pseudocentrotus depressus and Hemicentrotus pulcherrimus were collected by injecting 0.5 M KC1 into the coelom and were used within several hours. About 30 eul of "dry" sperm was suspended in 0 .5 ml of calcium-free sea water. A 40 ƒÊl portion of the suspension was added to 0.5 ml of 0.04 % (v/v) Triton X-100 in 50 mM Tris-HCl (pH 8.1)- 0.1 M KCl-2 mM MgCl2-0.1 mM EDTA (TKME) then gently swirled at room temperature for 30 s. The demembranated spermatozoa were precipitated by centrifugation at 1,000 g for 5 min after which they were suspended carefully in 0.5 ml of TKME to make a suspension of Triton models free from Triton X-100 and spermatozoal materials that have been made soluble by the detergent. The models were reactivated by mixing 40 ƒÊl of the suspension with 2 ml of 1 mM ATP in TKME. The percent motility of the Triton models, defined as (the number of motile Triton models/total Triton models) •~ 100, was checked with a microscope equipped with a dark-field condenser and a 100 W Hg arc lamp. Fresh Triton models that had a percent motility greater than 60 % (usually 60-90 %) were used routinely in our experiments.

Preparation of axonemes and purification of dynein. Axonemes were prepared from spermatozoa of the sea urchin Pseudocentrotus depressus. Spermatozoa that had been washed in calcium-free sea water were homogenized with 0.5 % (v/v) Triton X-100 in TKME to remove the membranes and separate the tails from the heads. The demembranated heads and tails were sedimented by centrifugation at 10,000 g for 10 min. After discarding the supernatant, the upper, white pellet consisting mainly of tails was suspended in TKME leaving the lower, hard pellet that contained mostly heads. The tail suspension was centri- Control of Flagellar Motility 329

fuged and its upper pellet was washed several times in TKME, which gave an axoneme fraction free from the contamination by heads. Dynein was prepared from Hemicentrotus spermatozoa by the method of Gibbons and Fronk (11) with some modification. Dynein was extracted from the axonemes by use of a solution containing 0.5 M KCl, 50 mM Tris-HCl (pH 8.1), 2 mM MgCl2 and 0.1 mM EDTA. It was purified by ultracentrifugation through a 5-20 % (w/v) sucrose density

gradient in 0.5 M KCl-50 mM Tris-HCl (pH 8.1)-2 mM MgCl2-0.1 mM EDTA. The specific activity of the purified dynein was about 1.1 ƒÊmol/mg/min, evidence that activation had already occurred during purification. No latent type of dynein has yet been obtained from the sperm of this sea urchin. Purification of phosphoprotein phosphatase from bovine cardiac muscle. Phosphoprotein

phosphatase was purified from bovine cardiac muscle according to the method of Chou et al. (4). The purified enzyme was dialyzed against TKME then stored at -80•Ž. Its activity was measured with casein as the substrate. A 0.2 ml mixture containing 7 mg/ml casein

(Merck) and phosphoprotein phosphatase in TKME was incubated at 30•Ž for 10 min, after which 0.2 ml of 15 % (w/v) trichloroacetic acid was added to terminate the reaction. The precipitates formed were removed by centrifugation, and the amount of inorganic

phosphate in the supernatant was measured by the method of Chen et al. (3). The specific activity of the purified enzyme was 1.9 ƒÊmol of Pi released/min/mg protein under our assay conditions. Purification of cyclic AMP-dependent protein kinase from sea urchin spermatozoa. Cyclic AMP-dependent protein kinase was purified from spermatozoa of the sea urchin Anthocidaris crassispina according to the method of Ishiguro et al. (16). The specific activity of the purified enzyme was 13 nmol of Pi transferred/min/mg protein. Electrophoresis and autoradiography. Electrophoresis was performed after Laemmli

(18) using 5-15 % (w/v) acrylamide density gradient gels. Proteins were stained with Coomassie Blue R-250. After being destained in 7 % (v/v) acetic acid solution, the gels were dried then autoradiographed using Kodak X-Omat S films. Other procedures. The Mg-ATPase activity of the dynein was assayed by the procedure described elsewhere (13). Proteinase activity was assayed in a reaction mixture containing 7 mg/ml of casein and 0.1 mg/ml of phosphoprotein phosphatase in TKME. This mixture was incubated at 30•Ž for 1 h, after which trichloroacetic acid was added to precipitate the

proteins in the mixture. The UV absorbance of the supernatant then was measured. Protein concentration was determined by the method of Lowry et al. (22) with bovine serum albumin as a standard.

RESULTS

Phosphorylation and dephosphorylation of axonemal proteins by protein kinase and phosphoprotein phosphatase. Chou et al. (4) reported that the phosphoprotein phosphatase prepared from bovine cardiac muscle dephosphorylates the regulatory subunit of cyclic AMP-dependent protein kinase, histone, casein and phosphorylase. They did not, however, examine the substrate specificity of the enzyme toward axonemal proteins. Therefore, before applying this enzyme to the Triton models of spermatozoa, we examined whether the enzyme is capable of dephosphorylating axonemal proteins.

We prepared 32P-labeled axonemes using ƒÁ-[32P]ATP and the cyclic AMP- dependent protein kinase purified from sea urchin spermatozoa. When the axonemes 330 D. Takahashi, et al.

Fig. 1. In vitro phosphorylation of axonemal proteins by protein kinase and phosphoprotein

phosphatase. A 200-ƒÊl portion of the axoneme suspension (prepared from Pseudocentrotus sper-

matozoa) consisting of 1 mg/ml axonemal proteins and 50ƒÊM ƒÁ-[32P]ATP (1 .9 •~ 108 cpm/ml) in TKME was incubated at 20•Ž for 5 min with, or without , 10 ƒÊg/ml of cyclic AMP-dependent protein kinase and 10 ƒÊM cyclic AMP. At the end of incubation, the suspension was chilled in an ice-water bath

Axonemes were collected by centrifugation at 10 ,000 g for 5 min, washed once in TKME then suspended in 0.2 ml of TKME. The resulting axoneme suspension was divided in two parts . Each part was incubated at 20•Ž for 15 min with, or without, 65 ƒÊg/ml of the phosphoprotein phosphatase.

The reaction was terminated by adding an equal volume of 0 .125 M Tris-HC1 (pH 6.8)-4 % (w/v) sodium dodecyl sulfate-20 % (v/v) glycerol-10 % (v/v) 2-mercaptoethanol , after which the mixture was heated at 80°C for 5 min.

A 60-ƒÊg portion of each protein sample was electrophoresed according to the method of Laemmli

(18). Lanes 1-5, protein staining. Lanes 6-9, autoradiogram. Lane 1, marker proteins (a, myosin heavy chain; b, phosphorylase; c, bovine serum albumin; d , tubulin; e, actin; f, glyceraldehyde-3- phosphate dehydrogenase; g, carbonic anhydrase; h, soybean trypsin inhibitor; i, cytochrome c).

Lanes 2 and 6, minus protein kinase and cyclic AMP, minus phosphoprotein phosphatase treatment . Lanes 3 and 7, minus protein kinase and cyclic AMP , plus phosphoprotein phosphatase treatment. Lanes 4 and 8, plus protein kinase and cyclic AMP, minus phosphoprotein phosphatase treatment . Lanes 5 and 9, plus protein kinase and cyclic AMP , plus phosphoprotein phosphatase treatment. The polypeptide band with the apparent molecular weight of 34 Kd that is present in lanes 3 and 5

was carried over from the phosphoprotein phosphatase preparation .

were incubated only with radioactive ATP without exogenous protein kinase, several proteins were phosphorylated, presumably by the kinase bound to the axoneme

(Fig. 1, lane 6). Phosphorylation was enhanced greatly by the addition of cyclic AMP-dependent protein kinase (lane 8), more than 15 polypeptides being phosphorylated by the enzyme. The apparent chain weights of the major phosphoproteins were 260, 250, 235, 205, 170, 96, 61 and 21 Kd. Control of Flagellar Motility 331

Fig. 2. Effects of varying the concentration of phosphoprotein phosphatase on the percent motility of Triton models. Triton model suspensions prepared from Pseudocentrotus (a) and Hemi- centrotus (b) spermatozoa were incubated at 20•Ž for 10 min with various concentrations of phosphoprotein phosphatase. At the end of incubation the Triton models were precipitated by centrifugation at 1,000 g for 5 min. The models then were suspended in 2 ml of 1 mM ATP in TKME and the percent motility was measured.

When the phosphorylated axonemes were incubated with phosphoprotein phosphatase, most of the radioactive phosphate was released except from the phosphoproteins with apparent chain weights of 110, 51 and 46 Kd (lane 9). Our data indicate that the phosphoprotein phosphatase prepared from bovine cardiac muscle can dephosphorylate most of the phosphoproteins present in the axoneme of sea urchin spermatozoon and that it can be used to study the roles of these axonemal phosphoproteins in flagellar movement. The phosphoprotein phosphatase fraction contained neither ATPase nor proteinase activities. ATPase activity in our purified phosphoprotein phosphatase fraction was measured in an incubation mixture of TKME, 2 mM ATP and 50 ƒÊg/ml of phosphoprotein phosphatase. Incubation of this mixture at 30•Ž for 30 min showed there was no release of inorganic phosphate when assayed by the procedure used to measure dynein ATPase activity. Furthermore, no proteinase activity contaminated the purified phosphoprotein phosphatase preparation when casein was the substrate. The fact that no degradation of proteins took place when axonemes were incubated with the phosphoprotein phosphatase also shows the lack of contamination (Fig. 1, lanes 2-5). Effects of phosphoprotein phosphatase on the motility of Triton models. When a Triton model suspension was incubated with phosphoprotein phosphatase, all the sperm models remained stationary with their tails in a straight line. On the addition of ATP, almost all were immotile in a form of "quiescence" (10). Experiments with various enzyme concentrations showed that more than 12 ƒÊg/ml of protein phosphatase was necessary to block the motility of Triton models of Pseudocentrotus spermatozoa (Fig. 2a); whereas, Triton models prepared from Hemicentrotus spermatozoa were less sensitive to phosphoprotein phosphatase treatment. About 70 ƒÊg/m1 of the enzyme was needed to render most of the Triton models immotile

(Fig. 2b). 332 D. Takahashi, et al.

When Triton models were incubated for more than 10 min at temperatures higher than 25•Ž, the percent motility of the control models decreased to less than 50 %. In subsequent experiments, therefore, enzyme treatment took place at 20•Ž for 10 min with 12 ƒÊg/ml (Pseudocentrotus) and 70 ƒÊg/ml (Hemicentrotus) of phosphoprotein phosphatase, added to 0.2-0.3 mg/ml of the Triton models. Under these conditions we consistently found that phosphoprotein phosphatase treatment drastically reduced the percent motility of the Triton models in comparison with the values found for control models incubated in the absence of the enzyme (7 experiments with Pseudo- centrotus and 10 with Hemicentrotus). In contrast to the more than 50 % motility found for the controls, 5-20 % (8 experiments) or less than 5 % (9 experiments) was found in the phosphoprotein phosphatase-treated models.

The extent to which the percent motility of the Triton models was decreased by this enzyme treatment differed from one model preparation to another. But, it was always found that phosphoprotein phosphatase treatment lowered the motility of the models far less than it did that of the controls. Also, the beat frequency decreased in the phosphatase-treated models even when the percent motility was not markedly lowered.

When the phosphoprotein phosphatase preparation was heated in a boiling water bath for 2 min, its activity was completely lost. Also, it caused the total loss of enzyme's ability to block the motility of Triton models. Potassium pyrophosphate at 7 mM, which inhibited this enzyme activity by 60 when 32P-labeled casein was the substrate, suppressed the action of protein phos-

phatase on the Triton model, the percent motility being as high as that of the control. Restoration by cyclic AMP and protein kinase of motility in phosphoprotein phosphatase-treated Triton models. To confirm whether the dephosphorylation of phosphoproteins renders Triton models immotile, immobilized models were reactivated by protein phosphorylation. After treatment with phosphoprotein phos-

phatase, the Triton models were collected by centrifugation in order to remove the enzyme. The models then were suspended in TKME containing ATP, after which the percent motility was checked. Cyclic AMP and cyclic AMP-dependent protein kinase then were added successively to the Triton model suspension, and the percent motility was measured after each addition. Some deviations in results were found in Triton models obtained from different preparations, the data from seven independent experiments falling into two categories: 1) Addition of cyclic AMP increased the

percent motility of the treated models, but a further addition of protein kinase had little effect. 2) Cyclic AMP reactivated the phosphatase-treated models, an even higher percent motility being obtained on the addition of the protein kinase (Table 1).

Full restoration of the motility of phosphatase-treated models by the protein-

phosphorylating system has not been achieved ; but, the percent motility was raised in most cases by an addition of cyclic AMP with, or without, the protein kinase. Cyclic GMP at 10µM had no effect (data not shown). We therefore conclude that it was not any contaminating factor(s) in the phosphatase preparation, but the

phosphoprotein phosphatase itself, that deprived the Triton models of their motility. Effects of phosphoprotein phosphatase on the ATPase activities of Triton models, axonemes and purified dynein. Clearly, dephosphorylation of phosphoprotein(s) in Triton models blocks flagellar motility. What kind of protein(s) must be phos-

phorylated if the flagellum is to be motile? One possibility is that dynein ATPase, like myosin (1), is regulated by the phosphorylation-dephosphorylation system. Control of Flagellar Motility 333

TABLE 1. RESTORATION OF THE MOTILITY OF PHOSPHOPROTEIN PHOSPHATASE-TREATED TRITON MODELS BY CYCLIC AMP-DEPENDENT PROTEIN KINASE

TABLE 2. EFFECTS OF PHOSPHOPROTEIN PHOSPHATASEON ATPASE ACTIVITY IN TRITON MODELS, ISOLATEDAXONEMES AND PURIFIED DYNEIN

Therefore, we investigated the effect of phosphoprotein phosphatase on the ATPase activity of dynein. When Triton models first had been incubated with the enzyme, their ATPase activities were decreased to about half the control's value. But, when axonemes were 334 D. Takahashi, et al.

Fig. 3. Effects of phosphoprotein phosphatase on ATPase activity in isolated axonemes assayed at various pHs. Axonemes (from Pseudocentrotus spermatozoa) were suspended in 0.2 ml of 0.1 M KCl-2 mM MgCl2-0.1 mM EDTA-1 mM DTT-50 mM buffer solution at a protein concentration of 10 ƒÊg/ml.The buffers used were MES (2-(N-morpholino)ethanesulfonic acid) for pHs 6 and 7,

Tris for pHs 8 and 9, and CAPS (3-cyclohexylaminopropanesulfonic acid) for pH 10. The axoneme suspension was incubated at 20•Ž for 15 min with, or without, 5 ƒÊg/ml of phosphoprotein phosphatase, after which ATP (final 1 mM) was added. ATP hydrolysis was allowed to continue for 10 min at 20•Ž, then an equal volume of 15 % (w/v) trichloroacetic acid was added. After centrifugation the amount of inorganic phosphate in the supernatant was determined by the method of Chen et al. (3).

Closed circles, control ; open circles, phosphoprotein phosphatase-treated axonemes. treated with protein phosphatase, no decrease in ATPase activity was detected (Table 2). We considered the possibility that the dephosphorylation of dynein shifts the pH profile of the enzyme activity as it does that of phosphorylase kinase (17); but, the protein phosphatase scarcely affected the ATPase activity of the axoneme over a wide pH range (Fig. 3). Furthermore, it did not inhibit the ATPase activity of the purified dynein (Table 2).

DISCUSSION Our study makes clear that the flagellar motility of Triton models of sea urchin spermatozoa is suppressed by phosphoprotein phosphatase purified from bovine cardiac muscle and that this suppression is reversed by a protein-phosphorylating system. Some protein(s) must be in phosphorylated form to maintain the continuous movement of the sperm's flagella. There are many proteins that are phosphorylated by protein kinase in the axonemes of sea urchin sperm (Fig. 1). Do they all need to be phosphorylated for flagellar motility, or is phosphorylation of only some necessary for movement? From the results reported here it is not possible to answer to this question. We need to determine what are suitable conditions for the selective de- Control of Flagellar Motility 335

phosphorylation of proteins in the axoneme by this phosphoprotein phosphatase and by different protein phosphatases with different substrate specificities. We also must determine the relation between the extent of the phosphorylation of axonemal proteins and the degree of flagellar motility of the Triton models. In our experiments to restore the flagellar motility of phosphatase-treated sperm models by a protein-phosphorylating system, an addition of cyclic AMP alone, without any supplement of the flagellar cyclic AMP-dependent protein kinase, sometimes reactivated flagellar motility. This suggests that a cyclic AMP-dependent protein-phosphorylating system functions in the axonemes of Triton models. We found that the cyclic AMP-dependent protein kinase in sea urchin spermatozoa is made soluble by detergent (25). But we also found that substantial cyclic AMP- dependent protein kinase activity remains in the axoneme preparation and that this enzyme can be made soluble in a solution containing a high concentration of KC1 or NaCl (unpublished observation). This bound-form of protein kinase may have an important function in reactivating the flagellar motility of phosphatase-treated Triton models. In the last investigation reported here, dynein was postulated as the possible target of the phosphoprotein phosphatase, and several experiments were performed to examine this possibility. We found that the heavy chain(s) and some intermediate chains of the purified dynein were phosphorylated by the cyclic AMP-dependent protein kinase purified from sea urchin spermatozoa (unpublished data). We here have shown that high molecular weight polypeptides that might be dynein heavy chains were phosphorylated by the protein kinase (Fig. 1). We also showed that the apparent ATPase activity in the Triton models is decreased by phosphoprotein phosphatase treatment. This may correspond to the large decrease in the flagellar movement of the models produced by this treatment. In contrast, phosphoprotein phosphatase did not affect the ATPase activity of isolated axonemes and purified dynein. Possibly, the target proteins of the phosphoprotein phosphatase are not dynein polypeptides ; the phosphorylated target(s) regulates dynein ATPase activity. Only the axonemes of Triton models would contain these regulatory protein(s); isolated axonemes and purified dynein would be devoid of this mechanism. Another possibility is that the dynein molecule, which is bound to the axoneme of the Triton model, preserves its characteristic of regulation by phosphorylation and dephosphorylation and that this characteristic is lost when the axonemes are isolated or when dynein is purified. Further investigation of the possible significance of the phosphorylation of dynein by the cyclic AMP-dependent protein kinase present in spermatozoa is needed. We have shown elsewhere that cyclic AMP-dependent protein kinase and a protein factor partially purified from the detergent-extract of spermatozoa reactivate the flagellar motility of Triton models prepared from aged spermatozoa (16, 26). We tested this protein factor's ability to reactivate the flagellar motility of phosphatase- treated models. Complete movement could not be restored when this factor fraction was added to phosphatase-treated models after the addition of cyclic AMP and protein kinase. In the experiments reported here, we used fresh "dry" sperm to obtain Triton models with high percent motilities. Possibly the protein factor may be present in the axonemes of Triton models. To determine whether this is so, more detailed studies must be made of the function of this protein factor that is necessary for the reacti- 336 D. Takahashi, et al. vation of Triton models prepared from aged spermatozoa. From our previous study of the significance of protein phosphorylation in the reactivation of Triton models of sea urchin spermatozoa and our present data on the suppression of these sperm models' motility by phosphoprotein phosphatase, it is clear that the protein phosphorylation-dephosphorylation system has a major role in the regulation of the flagellar motility of spermatozoa.

Acknowledgement.This study wassupported in part by Grants-in-Aid(Nos. 58540454,57440004 and 57380016)from the Ministryof Education,Science and Culture of Japan.

REFERENCES

1. ADELSTEIN,R.S., M.D. PATO, J.R. SELLERS,M.A. CONTI and C.R. EATON. Regulation of contraction by reversible phosphorylation of myosin and myosin kinase. In Protein phosphorylation. Cold Spring Harbor Conferences on Cell Proliferation vol. 8, eds. M. Rosen and E.G. Krebs, Cold Spring Harbor Laboratory, New York, pp. 811-822, 1981 2. CASCIERI,M., R.P. AMANNand R.H. HAMMERSTEDT.Adenine nucleotide change at initiation of bull sperm motility. J. Biol. Chem. 251, 787-793, 1976 3. CHEN, P.S. Jr., T.Y. TORIBARAand H. WARNER. Microdetermination of phosphorus. Anal. Chem. 28, 1756-1761, 1956 4. CHOU, C.K., J. ALFANOand O.M. ROSEN. Purification of phosphoprotein phosphatase from bovine cardiac muscle that catalyzes dephosphorylation of cyclic AMP-binding protein com- ponent of protein kinase. J. Biol. Chem. 252, 2855-2859, 1977 5. GARBERS,D.L., E.L. DYER and J.G. HARDMAN. Effects of cations on guanylate cyclase of sea urchin sperm. J. Biol. Chem. 250, 382-387, 1975 6. GARBERS,D.L., N.L. FIRST and H.A. LARDY. Properties of adenosine 3',5'-monophosphate- dependent protein kinse isolated from bovine epididymal spermatozoa. J. Biol. Chem. 248, 875-879, 1973 7. GARBERS, D.L. and J.G. HARDMAN. Factors released from sea urchin eggs affect cyclic nucleotide metabolism in sperm. Nature (Lond.) 257, 677-678, 1975 8. GARBERS,D.L., W.D. LUST, N.L. FIRST and H.A. LARDY. Effects of phosphodiesterase inhibitors and cyclic nucleotides on sperm respiration and motility. Biochemistry 10, 1825- 1831, 1971 9. GIBBONS,B.H. and I.R. GIBBONS. Flagellar movement and adenosine triphosphatase activity in sea urchin sperm extracted with Triton X-100. J. Cell Biol. 54, 75-97, 1972 10. GIBBONS,B.H. and I.R. GIBBONS. Calcium-induced quiescence in reactivated sea urchin serm. J. Cell Biol. 84, 13-27, 1980 11. GIBBONS,I.R. and E. FRONK. A latent adenosine triphosphatase form of dynein 1 from sea urchin sperm flagella. J. Biol. Chem. 254, 187-196, 1979 12. GRAY, J.P., G.I. DRUMMOND,D.E.T. LUK, J.G. HARDMANand E.W. SUTHERLAND. Enzymes of cyclic nucleotide metabolism in invertebrate and vertebrate sperm. Arch. Biochem. Biophys. 172, 20-30, 1976 13. HOSHINO, M. Interaction of Tetrahymena dynein with microtubule protein. Tubulin-induced stimulation of dynein ATPase activity. Biochim. Biophys. Acta 462, 49-62, 1976 14. HOSKINS,D.D. Adenine nucleotide mediation of fructolysis and motility in bovine epididymal spermatozoa. J. Biol. Chem. 248, 1135-1140, 1973 15. HOSKINS,D.D., E.R. CASILLUSand D.T. STEPHENS. Cyclic AMP-dependent protein kinase of bovine epididymal spermatozoa. Biochem. Biophys. Res. Commun. 48, 1331-1338, 1972 16. ISHIGURO,K., H. MUROFUSHIand H. SAKAI. Evidence that cAMP-dependent protein kinase and a protein factor are involved in reactivation of Triton X-100 models of sea urchin and starfish spermatozoa. J. Cell Biol. 92, 777-782, 1982 17. KREBS,E.G., J.T. STULL, P.J. ENGLAND,T.S. HUANG, C.O. BROSTROMand J.R. VANDENHEEDE. Control of Flagellar Motility 337

The regulation of muscle metabolism and function by protein phosphorylation. In Protein Phosphorylation in Control Mechanisms, eds. F. Huijing and E.Y.C. Lee, Acad. Press, New York, pp. 31-45, 1973 18. LAEMMLI,U.K. Cleavage of structural proteins during the assembly of the head of bacterio- phage T4. Nature (Lond.) 227, 680-683, 1970 19. LEE, M.Y.W. and R.M. IVERSON. Protein kinase in sea urchin gametes and embryos. Exp. Cell Res. 75, 300-304, 1972 20. LEE, M.Y.W. and R.M. IVERSON. An adenosine 3' : 5' monophosphate-dependent protein kinase from sea urehin spermatozoa. Biochim. Biophys. Acta 429, 123-136, 1976 21. LINDEMAN,C.B. A cAMP-induced increase in the motility of demembranated bull sperm models. Cell 13, 9-18, 1978 22. LOWRY,0.H., N.J. ROSEBROUGH,A.L. FARR and R.J. RANDALL. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275, 1951 23. MORISAWA,M. and M. OKUNO. Cyclic AMP induces maturation of trout sperm axoneme to initiate motility. Nature (Lond.) 295, 703-704, 1982 24. MORTON, B., J. HARRIGAN-LUM,L. ALBAGLIand T. Joss. The activation of motility in quiescent hamster sperm from the epididymis by calcium and cyclic nucleotides. Biochem. Biophys. Res. Commun. 56, 372-379, 1974 25. MUROFUSHI,H. Protein kinases in sea urchin spermatozoa. Ph.D. Dissertation, University of Tokyo, 1975 26. MUROFUSHI,H., K. ISHIGUROand H. SAKAI. Involvement of cyclic AMP-dependent protein kinase and a protein factor in the regulation of the motility of sea urchin and starfish spermatozoa. In Biological Functions of Microtubules and Related Structures, eds. H. Sakai, H. Mohri and G.G. Borisy, Acad. Press, New York, pp. 163-176, 1982 27. OPRESKO,L.K. and C.J. BROKAW. cAMP-dependent phosphorylation associated with activation of motility of Ciona sperm flagella. Gamete Res. 8, 201-218, 1983 28. TANG, F.Y. and D.D. HosKINs. Phosphoprotein phosphatase of bovine epididymal spermatozoa. Biochem. Biophys. Res. Commun. 62, 328-335, 1975 29. TASH, J.S., S.S. KAKAR and A.R. MEANS. Flagellar motility requires the cAMP-dependent phosphorylation of a heat-stable NP-40-soluble 56 kd protein, axokinin. Cell 38, 551-559, 1984

(Received for publication, August 23, 1985)