phosphorylation results in activation of a camp ... · the journal of biological chemistry 0 1988...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 263, No. 21, Issue of July 25, PP. 10353-10358,1988 Printed in U.S.A.

Phosphorylation Results in Activation of a cAMP Phosphodiesterase in Human Platelets*

(Received for publication, June 30, 1987)

Colin H. MacpheeS, David H. ReifsnyderSO, Tom A. Moore$, Kenneth M. Lereav, and Joseph A. Beavo$ From the $Departments of Pharmacology and TBiochemistry, School of Medicine, University of Washington, Seattle, Washington 98195

Agents such as prostaglandins El and I2 which elevate cAMP levels in platelets also increase cAMP phosphodiesterase activity. Since much of the cAMP phosphodiesterase activity in human platelets is due to the cGMP-inhibited isozyme (Macphee, C. H., Harrison, S. A., and Beavo, J. A. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 6600-6663), we examined the regulation of this isozyme by prostaglandins El and IZ in intact platelets. Because this isozyme is a minor component of platelet protein, normally requiring several thousand- fold purification to achieve homogeneity, a specific monoclonal antibody (CGI-6) was utilized to identify and isolate the cGMP-inhibited phosphodiesterase activity. Treatment of intact platelets with the prostaglandins promoted an increase in the phosphorylation state of the cGMP-inhibited phosphodiesterase and a corresponding increase in phosphodiesterase activity. The effect on activity and phosphorylation of the cGMP-inhibited phosphodiesterase was observed within 2 min after intact platelets were exposed to the prostaglandins. The half-maximal effective dose for prostaglandin IZ (10 nM) was approximately 10-fold lower than that for prostaglandin El. The phosphorylated, cGMP-inhibited isozyme migrated as a 110-kDa peptide following sodium dodecyl sulfate gel electrophoresis. Direct in vitro phosphorylation of the platelet cGMP-inhibited phosphodiesterase by the catalytic subunit of CAMP-dependent protein kinase caused a similar increase in phosphodiesterase activity. Treat- ment with PKI peptide, a specific inhibitor of CAMP- dependent protein kinase, blocked the phosphorylation and the effect on activity. Taken together, the data strongly suggest that the effects of prostaglandins El and Iz on platelet phosphodiesterase activity are mediated by a direct CAMP-dependent protein kinase- catalyzed phosphorylation of the cGMP-inhibited phosphodiesterase isozyme.

Many hormones and neurotransmitters exert their physio- logical actions by activating adenylate cyclase and increasing the cytosolic levels of the second messenger molecule, CAMP. In order to control receptor-mediated elevations in CAMP,

* This work was supported by National Institutes of Health Grant AM21723 (to J. A. B.) and by National Research Service Awards AM07535 (to D. H. R.) and HL07397 (to K. M. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

5 To whom correspondence should be addressed Dept. of Phar- macology, SJ-30, University of Washington, Seattle, WA 98195.

cells contain several specific modulatory mechanisms by which the pre-stimulated state is returned with time. One of these mechanisms is at the level of receptor/adenylate cyclase desensitization (Sibley and Lefkowitz, 19851, whereas the other involves the removal of CAMP. The most important cellular mechanism for the removal of cAMP to the physio- logically inert 5’-AMP is degradation by specific cyclic nucleotide phosphodiesterases.

Most tissues contain one or more of several distinct phosphodiesterase isozymes which vary widely in their substrate affinities and modes of regulation (for a review, see Wells and Hardman, 1977; Beavo, 1988). These phosphodiesterase isozymes may be identified as belonging to one of several different isozyme families including: a large calcium/calmodulin- dependent isozyme family, a cGMP-dependent family in which cGMP stimulates the hydrolysis of CAMP, a “low K,,,” family of isozymes so designated since the K , for cAMP hydrolysis is 1 p~ or less, and a cGMP-specific family which includes the light-activated phosphodiesterase present in pho- toreceptor outer segments. Nearly all of these isozymes are trace proteins which are found in amounts of less than 0.01% in most tissues, and most of the isozymes are extremely sensitive to proteolytic degradation and subsequent activation. Therefore, most conventional methodologies for study- ing the effects of phosphorylation on a particular phosphodiesterase activity are ineffective. A number of studies have examined the regulation of low K,,, phosphodiesterase activity, due to the significance this may have in the homeostasis of cellular cAMP levels. However, very little definitive data is available on the actual mechanism(s) by which these hormone-induced alterations in phosphodiesterase activity occur within the intact cell. A major problem has been the lack of unambiguous identification of the phosphodiesterase isozyme which is under hormonal regulation. One method which has been particularly useful has been the utilization of specific monoclonal antibodies to isolate different phosphodiesterase isozymes.

Recently, we have demonstrated that the major platelet low K, cAMP phosphodiesterase appeared identical to the cGMP- inhibited phosphodiesterase isolated from bovine heart (Mac- phee et al., 1986). Since this isozyme represents more than 80% of the total low K , cAMP phosphodiesterase activity found in platelet extracts, it seemed likely to play an important role in regulating cAMP homeostasis in these cells. The potential importance of regulating phosphodiesterase activity is highlighted when one considers that platelet responsiveness depends on a critical balance between agents that alter cytosolic free calcium and cAMP (Feinstein et al., 1981).

Prostaglandins 12 and El, which are known to increase cytosolic cAMP levels, also produced an activation of low K,,, cAMP phosphodiesterase activity in platelets (Alvarez et al.,

10353

10354 Phosphorylation of cGMP-inhibited Phosphodiesterase

1981; Hamet et al., 1983). In the present study, a monoclonal antibody which recognizes the cGMP-inhibited phosphodiesterase was used to demonstrate a mechanism by which PGIz or PGE, increase the low K,,, phosphodiesterase activity in intact human platelets. This is the first report using an intact cell system which demonstrates a correlation between activation of a well characterized, specific phosphodiesterase isozyme and an increased state of phosphorylation. Further evidence that the regulation of phosphodiesterase activity may be directly mediated by CAMP-dependent protein kinase is shown by in vitro phosphorylation of the cGMP-inhibited platelet isozyme. Thus, the cGMP-inhibited phosphodiesterase activity appears to mediate a negative feedback mechanism which aids rapid restoration of elevated cellular cAMP levels and thereby allows the platelets to rapidly respond to aggregatory signals following stimulation of the prostaglandin receptor.

EXPERIMENTAL PROCEDURES

MateriaL~-[2,8-~H]cAMP (26 Ci/mmol) was purchased from ICN Pharmaceuticals, whereas both [32P]ATP (3000 Ci/mmol) and [3'P] orthophosphoric acid, carrier-free (5 mCi/ml), were from Du Pont- New England Nuclear. Crotalus atrox snake venom, CAMP, benzamidine HCl, aspirin, apyrase, PGIz, PGEI, and bovine serum albumin (fraction V) were obtained from Sigma. Leupeptin and pepstatin A were bought from Boehringer Mannheim and Triton X-100 from Pierce Chemical Co. DEAE-Sephadex A-25 was obtained from Phar- macia LKB Biotechnology, Inc. Heat-inactivated and formalin-fixed cells of the Cowan I strain of Staphylococcus aureus (Pansorbin) were purchased from Behring Diagnostics, and rabbit anti-mouse IgG antiserum was obtained from Miles Laboratories. PKI peptide (see Scott et al., 1986) and the catalytic subunit of CAMP-dependent protein kinase were a generous gift from the laboratory of Dr. E. G. Krebs, University of Washington, Seattle, WA.

Preparation of Human Platelets-Fresh human blood from healthy volunteers was collected in one-sixth volume of acid/citrate/dextrose (1.5% citric acid, 2.5% trisodium citrate, 2% dextrose). The antico- agulant also contained aspirin (100 pM, final concentration) to eliminate any thromboxane formation. Following 15-20 min at room temperature, platelet-rich plasma was prepared by centrifugation at 150 X g for 15 min. The plasma was carefully removed, and platelets were sedimented at 1100 X g for 15 min before resuspending and washing in 10 ml of a modified calcium and phosphate-free, isotonic buffer containing 10 mM Hepes, pH 6.2 (room temperature), 135 mM NaC1, 2.7 mM KC1, 1 mM MgC12, 12 mM NaHC03, 5.5 mM dextrose, 13 mM trisodium citrate, and 0.3% bovine serum albumin. This platelet suspension was then recentrifuged at 1100 X g for 10 min. The platelet-free medium was carefully aspirated and the platelet pellet resuspended in 2 ml of incubation buffer (1.5-2.5 X log platelets/ml). The incubation buffer was similar to the wash buffer except the pH was adjusted to 7.4 (37 "C) and 0.02 units/ml of apyrase replaced trisodium citrate.

Human Platelet Incubations-For phosphorylation experiments, the platelet suspension was incubated for 90 min at 37 'C in the presence of [32P]orthophosphate (400 pCi/ml). For consistency between studies, platelet suspensions which were to be used in phosphodiesterase activity studies underwent similar treatments except that no radioactivity was added. Following this 90-min incubation period, the platelets were centrifuged at 1100 X g for 10 min and resuspended to a concentration of 1-1.2 X log platelets/ml in fresh incubation buffer containing no radioactivity.

After a 15 min preincubation period at 37 "C, 0.25-ml aliquots of this suspension were removed and placed into pre-warmed 1.5-ml Eppendorf tubes already containing prostaglandin or vehicle. At appropriate times thereafter, tubes were removed and platelets pel-

The abbreviations used are: PGI,, prostacyclin; PGEI, prostaglandin El; CGI-5, monoclonal antibody directed against the cGMP- inhibited phosphodiesterase; ROS-1, monoclonal antibody directed against the rod outer segment phosphodiesterase; PKI peptide, a synthetic peptide (5-24) based on the active sequence of the specific heat-stable inhibitor of CAMP-dependent protein kinase; SDS, sodium dodecyl sulfate; Hepes, 4-(2-hydroxyethyl)-l-piperazineeth- anesulfonic acid.

leted in a Fisher microcentrifuge (10,000 X g, 6 s). The incubation was rapidly removed and replaced with 0.25 ml of ice-cold homogenization buffer before sonication for 15 s (Braun-Sonic 2000, 44% intensity). The time between centrifugation and sonication was pre- cisely 30 s and has been incorporated into all time points. The homogenization buffer for phosphorylation studies consisted of 50 mM Tris-HCI, pH 7.8 (4 "C), 50 mM benzamidine HCl, 50 mM NaF, 2.5 mM sodium pyrophosphate, 5 mM @glycerophosphate, 2 mM EDTA, 20 pg/ml leupeptin, 20 pg/ml pepstatin A, 50 nM PKI peptide, 1.5% Triton X-100, and 0.1% SDS. The homogenization buffer used in phosphodiesterase activity studies was the same as above except SDS was omitted. Both buffers solubilized greater than 95% of platelet low K, cAMP phosphodiesterase activity. The homogenized, sonicated platelet suspension was either analyzed by protein immunoblot analysis or immunoadsorbed to a specific or control antibody reagent as described below.

Protein Zmmunoblot Analysis-The sonicated platelet suspension was boiled in SDS-containing buffer and analyzed by SDS gel electrophoresis on a 10% gel (Laemmli, 1970). The transfer of the protein to nitrocellulose and the subsequent immunoblotting analysis using a monoclonal antibody directed against a cGMP-inhibited phosphodiesterase and visualization process using "'1-protein A were described previously (Harrison et al., 1986b).

fmmunoadsorption Analysis of the cGMP-inhibited Phosphodiester- ase-A specific monoclonal antibody (CGI-5, which recognizes the platelet low K,,, phosphodiesterase, see Macphee et al., 1986) or control monoclonal antibody (ROS-1, directed against a phosphodiesterase isolated from the bovine rod outer segment, see Hunvitz et al., 1984) were used in these studies. Solid-phase antibody reagents were prepared as described previously (Harrison et al., 1986a) using Pansorbin, a commercially available preparation of heat-killed Cowan I strain of S. aureus cells. To reduce any nonspecific binding to the solid-phase antibody reagent, 9 pl of Pansorbin, previously washed and resuspended in homogenization buffer, was added to each preparation of the sonicated platelet suspension (250 pl) and incubated on ice for 20 min. This was centrifuged (10,000 X g, 25 min), and 220 pl of the resulting supernatant was incubated with Pansorbin containing either CGI-5 or ROS-1. Occasionally, portions of this supernatant were boiled 1:l (v/v) in SDS-containing sample buffer prior to analysis by SDS gel electrophoresis. After 2.5 h incubation on a rotating wheel at 4 'C, solid-phase antibody complex was repelleted (10,000 X g, 2.5 min) and the supernatant removed. The antibody complex was then washed by resuspension/centrifugation in 0.75 ml of the following buffers: once in homogenization buffer; once in 50 mM Tris-HC1, pH 7.8 (4 "C), 0.5 M NaCl, and 0.15% Triton X-100; and, finally, once in 50 mM Tris-HC1 at pH 7.8 (4 "C). For analysis of the 32P content in the phosphodiesterase, the washed antibody complex was then resuspended and boiled in 100 pl of sample buffer containing SDS. Samples were analyzed by SDS gel electrophoresis on a 10% gel (Laemmli, 1970) and phosphorylated proteins visualized by autoradiography. The radioactivity incorporated into each band was quantitated by excising the band, incubating in 30% hydrogen peroxide at 55 "C for 10 h, and counting via liquid scintillation spectrometry. The back- ground radioactivity (less than 5% of counts observed in the phosphodiesterase band) was determined from the relevant area of the gel from control immunoadsorptions subtracted from all other values. For analysis of the cGMP-inhibited phosphodiesterase activity, after washing the antibody complex in buffers without SDS, a portion of the immunoadsorbed phosphodiesterase was removed and assayed as described below.

In Vitro Phosphorylution of cGMP-inhibited Phosphodiesterase- Washed platelets (1 X 10') were sonicated as described above in 1 ml of homogenization buffer without SDS. Following centrifugation, 0.9 ml of the supernatant was incubated with either ROS-1 or CGI-5 monoclonal antibody reagent. The resulting complexes were washed twice in 0.75 ml of 50 mM Tris-HC1, pH 7.8 (4 "c) , 0.5 M NaC1, 50 mM benzamidine HC1,20 pg/ml leupeptin, and 20 pg/ml pepstatin A before final resuspension in Tris-HC1, pH 7.8 (4 "C), containing 150 mM NaCl. Phosphorylation reactions were carried out at 30 "c 8s described previously (Harrison et al., 1986b) except 10 p M ATP was used. Following termination by cooling on ice, the antibody complexes were washed twice with 40 mM Tris-HC1, pH 7.8 (4 "c), 0.5 M NaC1, and 0.1% Triton X-100 before boiling in SDS-containing buffer. Samples were then loaded onto a 10% acrylamide gel and analyzed by SDS gel electrophoresis. Parallel experiments were conducted in the absence of radioactivity to determine the effect of phosphorylation on phosphodiesterase activity. In these experiments an aliquot of the reaction mix was removed at the appropriate time and immediately

Phosphorylation of cGMP-inhibited Phosphodiesterase 10355

assayed for phosphodiesterase activity. For convenience, the protein kinase and ATP were not removed prior to the phosphodiesterase assay since they were not found to interfere with the assay. This methodology allowed the determination of the effect of phosphorylation on activity without subjecting the samples to possible losses which may have occurred during the washing of the antibody-phosphodiesterase complex. Similar results were obtained when the phosphorylated-antibody complex was washed before removing an aliquot for the phosphodiesterase assay.

Phosphodiesterase Activity Assay-Phosphodiesterase activity was assayed a t 30 "C, using 1 p~ CAMP, as described previously (Harrison, 1986a). Unless otherwise state, 10 pl of platelet homogenate or 25 pl of the phosphorylated antibody-phosphodiesterase complex was assayed under conditions where the rates of hydrolysis were linear over the times used.

RESULTS

Effect of PGI, on Platelet CAMP Phosphodiesterase Actiu- ity-When intact platelets were incubated with 1 PM PGI,, a rapid increase in CAMP phosphodiesterase activity was observed in homogenates (Fig. lA). The increase in activity reached a maximum of 50-60% above controls within the first few minutes following PGI, addition and was maintained for a t least 10 min. Prostaglandin E, also promoted a 50% increase in activity (data not shown). These observations are in general agreement with a previous report which showed an increase in CAMP phosphodiesterase activity in response to prostaglandins in homogenized platelets (Alvarez et al., 1981). The increase in activity appeared not to be caused by proteolytic activation of the cGMP-inhibited phosphodiesterase. As shown in Fig. lB, no degradation of the 110-kDa protein band was seen after treatment of intact platelets with PG12.

We have previously shown that the major low K, phosphodiesterase present in platelets appeared to be very similar or identical to a recently purified 110-kDa isozyme from bovine heart (Macphee et al., 1986). This particular phosphodiesterase isozyme has been designated as cGMP-inhibited phosphodiesterase, since low concentrations of cGMP inhibit (Ki = 6 X lo-' M) CAMP hydrolysis (Harrison et al., 1986a). As shown in Fig. 2, the PGIz-induced increase in CAMP phosphodiesterase activity was due to an increase in the cGMP-inhibited phosphodiesterase activity. In comparison to the control ROS-1 monoclonal antibody, the CGI-5 monoclonal antibody immunoadsorbed greater than 80% of the CAMP phosphodiesterase activity and essentially all of the PG1,-stimulated activity. All of the adsorbed activity following incubation with

FIG. 1. Effect of PGII on human platelet cAMP phosphodiesterase (PDE) activity. A, washed platelets were exposed to 1 p~ PGI? (closed box) or vehicle (open box) a t 37 "C for the times indicated. Reactions were terminated by replacing the incubation medium with ice-cold homogenization buffer. After sonication, the homoge- nated suspension was assayed for phosphodiesterase activity using 1 p~ CAMP. The data represent mean S.E. for six separate experiments. The average zero time phosphodiesterase activity was 16.8 & 1.9 pmol/min/1oR platelets. R, after treating the washed platelets with 1 p~ PGI2 (lanes 1-3) or vehicle (lane 4 ) for 1, 5, 10, or 10 min, respectively, the platelets were sonicated and analyzed by protein immunoblot analysis using a monoclonal antibody directed against a cGMP-inhibited phosphodiesterase. Ap- proximately 2.5 X loR platelets were used per lane.

CGI-5 was recovered in the resuspended antibody pellets (Fig. 2).

Effect of PGI, and PGE, on the Phosphorylation of cGMP- inhibited Phosphodiesterase-Agents which increase platelet CAMP levels are known to increase the net phosphorylation of 22-, 24-, and 50-kDa proteins (see Haslam et al., 1979). A similar response was observed in the present study using intact human platelets (Fig. 3A). However, cGMP-inhibited phosphodiesterase could only be distinguished from the other phosphorylated proteins by enriching the phosphodiesterase through the use of monoclonal antibody immunoadsorption (compare Fig. 3, A and B) . Using this procedure, we demonstrated that the cGMP-inhibited phosphodiesterase was phosphorylated when platelets were exposed to PGI2 (Fig. 3B). The phosphorylated enzyme migrated as a 110-kDa peptide following SDS gel electrophoresis and was immunoadsorbed by CGI-5 but not by ROS-1 (compare lanes 5 and 7, Fig. 3B).

The increase in phosphorylation of the immunoadsorbed, cGMP-inhibited phosphodiesterase was rapid and reached a maximum of approximately 3-fold above controls within 1 min following PGI, addition (Fig. 4). The elevated phosphorylation state was maintained for a t least 30 min (data not shown). The time course of cGMP-inhibited phosphodiesterase phosphorylation as identified by immunoadsorption with CGI-5 (Fig. 4) was very similar to the time course observed for PGI,-induced activation of a CAMP phosphodiesterase present in homogenized platelet suspension cAMP phosphodiesterase activation (Fig. lA). In both instances, the effects of phosphorylation were observed within 1 min following incubation with PGI, and were maintained for at least 10 min. We were unable to determine the magnitude of the phosphorylation effect following incubation of PGI2 from 0 to 1 min, since 45 s of the initial time interval involves the processing of the sample as described under "Experimental Procedures." The dose dependence of the prostaglandin-induced effect on the incorporation of the phosphate into the 110-kDa peptide (Fig. 5) is identical to the prostaglandin- induced activation of low K,, CAMP phosphodiesterase as reported by Alvarez et al. (1981). The ECso value of the response was approximately 10 nM for PGI2 which was 10- fold lower than the ECso for PGE,. This difference in ECso values is consistent with the knowledge that both PGIz and PGE, share the same receptor with PGI, being 10-30 times more potent (Whittle et al., 1985).

B

kDa 1 2 3 4

I l . I . l . I . I . I

0 2 4 6 8 1 0

Incubation Time (min)

116 -

66-

45 -

29 -

Dye - Front

10356 Phosphorylation of cGMP-inhibited Phosphodiesterase an r

No Ab ROS-1 Ab Supn. ROS-1 Ab Pell. CGI-5 Ab Supn.

Control

FIG. 2. Immunoadsorption of the PGI2-mediated increase in cAMP phosphodiesterase activity. Washed platelets were incubated with 1 ~ I M PGI? or vehicle a t 37 “C for 5 min before termination as outlined under “Experimental Procedures.” Supernatant fractions from each incubation were incubated for 2.5 h with either ROS-1 or CGI-5 monoclonal antibody reagents. Following precipitation of the solid-phase reagents by centrifugation, the supernatants and resuspended pellets (unwashed) were assayed for phosphodiesterase activity a t 1 g M CAMP.

1

I n Vitro Phosphorylation and Activation of cGMP-inhibited Phosphodiesterase-Platelet extracts were incubated with a monoclonal antibody (CGI-5) to selectively isolate the low K,,, phosphodiesterase activity or a control antibody to measure nonspecific binding. After a 2-h immunoadsorption, the antibody pellet was washed and resuspended as described under “Experimental Procedures.” This method of in vitro phosphorylation on the solid-phase antibody complex was chosen to eliminate any endogenous kinase activity present in the platelet extract. The antibody complexes were then incubated with the catalytic subunit of CAMP-dependent protein kinase with or without the synthetic PKI inhibitor peptide. The isolated, cGMP-inhibited phosphodiesterase was readily phosphorylated by the catalytic subunit and migrated as a M, = 110,000 protein band following SDS gel electrophoresis (Fig. 6A) . No phosphorylation was observed in the absence of kinase (lane 1 ) or when the platelet preparation was immunoadsorbed by a control antibody of the same IgG subclass directed against the bovine rod outer segment phosphodiesterase (ROS-1, lane 4 ) . Addition of the synthetic PKI inhibitor peptide prevented the CAMP-dependent protein kinase from phosphorylating the isolated cGMP-inhibited phosphodiesterase (lane 3) .

Phosphorylation of the cGMP-inhibited phosphodiesterase resulted in approximately a 50% increase in the phosphodiesterase activity (Fig. 6B). In these experiments, the antibody- phosphodiesterase complex, prepared as described above, was incubated with the catalytic subunit of CAMP-dependent protein kinase with or without the synthetic PKI inhibitor peptide. Immunoadsorption of the phosphodiesterase activity present in the platelets was only observed when the platelets were treated with the CGI-5 monoclonal antibody reagent (Fig. 6B, open bars). Less than 5% of the control activity was nonspecifically adsorbed by the ROS-1 antibody (solid bar). The synthetic PKI peptide also prevented the increase in activity which accompanied phosphorylation (Fig. 6B).

DISCUSSION

Incubation of intact platelets with PGI, and PGE,, which activate adenylate cyclase, is known to rapidly increase low K,,, cAMP phosphodiesterase activity (Alvarez et al., 1981; Hamet et al., 1983; Goldberg et al., 1984). However, it is not

116-

66-

45 - e

k I > a 1 2 3 4 5 6 7

116-

66 -

45 -

29 -

Dye - Front

FIG. 3. Effect of PC12 on phosphorylation of proteins and the cGMP-inhibited phosphodiesterase isolated from human platelets. A, washed platelets, maintained at 37 “C, were preincubated with [R2P]orthophosphate for 90 min before incubation with 1 PM PGIl for the following time intervals: lane 1 , 0 min control; lane 2, 45 s PGIZ; lane 3, 1.5 min PGI2; lane 4, 5 min PGL; lune 5, 10 min PGI,; lane 6, 10 min control. Following termination, the solubilized platelet proteins were analyzed directly by SDS gel electrophoresis. B, after the platelets were treated with PG12, the solubilized proteins were incubated with CGI-5 monoclonal antibody, and the antibody- phosphodiesterase complex was analyzed by SDS gel electrophoresis. Lanes 1-6 corresponded to the same incubation treatment as in A. A control monoclonal antibody (ROS-1, lane 7) also was incubated with the platelets following the 10-min treatment with PGI2.

known which of the phosphodiesterase isozymes contributes to this increased activity, and the evidence for CAMP-dependent protein kinase involvement remains indirect (Tremblay et al., 1985). Indeed, many reports have documented increased low K,,, cAMP phosphodiesterase activity within other cell types in response to hormones. The best known examples include incubation of adipocytes with insulin or lipolytic hormones (Loten and Sneyd, 1970; Pawlson et al., 1974; Weber and Appleman, 1982) and hepatocytes treated with insulin or glucagon (Loten et al., 1978). Several studies have attempted to both characterize the identity and mode of activation of the responsible phosphodiesterase(s). The most direct study showed the phosphorylation of a 52-kDa peptide and activation of a plasma membrane “high affinity” cAMP phosphodiesterase by insulin in a hepatocyte broken cell system (Marchmont and Housley, 1980; Houslay et al., 1984). A recent study by Gettys et al. (1988) indicates that the

Phosphorylation of cGMP-inhibited Phosphodiesterase 10357

0 2 4 6 8 1 0 Incubation Time (min)

FIG. 4. The time course of PGI2-mediated phosphorylation of cGMP-inhibited phosphodiesterase in intact human platelets. Washed platelets which had been preincubated with ["'Plortho- phosphate were exposed to 1 PM PGIZ (closed box) or vehicle (open box) a t 37 "C for the times indicated. Reactions were terminated and phosphorylated cGMP-inhibited phosphodiesterase immunoadsorbed and quantified as described under "Experimental Procedures." The data represent mean f S.E. for five separate experiments. The average zero time radioactivity was 126 f 19 cpm/lO* platelets.

0 -10 -9 -8 -7 -6 -5 Log [Prostaglandinl,M

FIG. 5. Concentration dependence of PG12 and PGEl-mediated phosphorylation of cGMP-inhibited phosphodiesterase in intact human platelets. Washed platelets which had been preincubated with ["'Plorthophosphate were incubated for 5 min at 37 "C with various concentrations of either PGIZ (closed box) or PGE, (open box). Reactions were terminated and phosphorylated cGMP-inhibited phosphodiesterase immunoadsorbed and quantified as described under "Experimental Procedures." Data shown are from representative experiments which have been replicated twice.

catalytic subunit of CAMP-dependent protein kinase may activate one of the low K,,, phosphodiesterase activities in adipocyte extracts. Although such studies do not clearly identify the actual phosphodiesterase involved, they have indicated a CAMP-dependent protein kinase-catalyzed activation of cAMP phosphodiesterase in liver, and fat cells. A general mechanism therefore appears to exist in many cells by which phosphodiesterase activation helps to restore hormone-elevated cAMP levels.

In this report, we have used intact human platelets as a model system to determine if, and subsequently how, cGMP- inhibited phosphodiesterase was activated following stimulation with agents that elevate cAMP levels. The data in Fig. 1 demonstrated that PGIz initiated a rapid phosphorylation and activation of platelet cGMP-inhibited phosphodiesterase activity. The observations with intact platelets were further substantiated by in uitro experiments using the platelet isozyme (isolated by a monoclonal antibody, CGI-5) and the catalytic subunit of CAMP-dependent protein kinase. In these studies the increase in cAMP phosphodiesterase activity was

116- YZ -

66-

4s-

- + + + KlnJw 1 2 3 4 L a n c

" + - PKI

I

I

- 2 + + + Kln:lw

1 4 Lmc

+ - PKI

FIG. 6. In vitro phosphorylation and activation of platelet CGI-5-purified phosphodiesterase. Phosphorylation of the phosphodiesterase adsorbed to CGI-5 monoclonal antibody (lanes 1-3) or ROS-1 monoclonal antibody (lane 4 ) was performed with purified catalytic subunit of CAMP-dependent protein kinase in the presence or absence of 5 nM PKI peptide. Reactions were incubated for 30 min with 10 p~ ATP. The immune complex was analyzed by SDS gel electrophoresis and visualized by autoradiography ( A ) or assayed for phosphodiesterase activity using 1 PM cAMP ( B ) . The basal activity observed from four platelet preparations varied from 80 to 380 pmol/ min/ml. Each value represents the mean f S.E. from four platelet preparations.

related to the phosphorylation state of cGMP-inhibited phosphodiesterase.

The interpretation of the results presented in this paper depends on the presumption that the 110-kDa peptide seen following SDS gel electrophoresis is a cGMP-inhibited phosphodiesterase. Since purification to homogeniety and characterization of an unproteolyzed, 110-kDa cGMP-inhibited phosphodiesterase has not been demonstrated from any tissue including platelets, some uncertainty exists. The anti-pro- tease buffer described in a previous report (Macphee et al., 1986) can be utilized in rapid immunoadsorption experiments. Unfortunately, this buffer is not effective for longer term, large scale purification procedures. Nevertheless, the probable identification of the 110-kDa peptide as a low K, cGMP- inhibited phosphodiesterase activity has been strongly sug- gested (if not proven) by several different criteria. Two inde- pendently isolated monoclonal antibodies (CGI-2 and CGI-5) have been utilized to adsorb this isozyme from bovine heart (Harrison et al., 1986b) and human platelets (Macphee et al., 1986). The adsorbed activity coincided with the removal of a 110-kDa peptide band from the supernate and the appearance of a 110-kDa peptide in the pellet. Using immunoblotting, Macphee et al. (1986) showed that the low K,,,, cGMP-inhibited phosphodiesterase activity co-migrated with the 110-kDa peptide following DEAE chromatography of sonicated human platelets. A genetic mutant (designated K30a) of mouse lym- phoma cells also has been characterized as having an increased level of cGMP-inhibited phosphodiesterase activity (Brothers et al., 1982), and this activity can be immunoadsorbed by the CGI monoclonal antibodies (Reifsnyder et al., 1985). Furthermore, photoaffinity labeling of the K30a cell extracts with [32P]cGMP followed by SDS gel electrophoresis identified a protein band which was very similar to the cGMP- isozyme present in human platelets and this protein co- migrated with the cGMP-inhibitedphosphodiesterase activity

10358 Phosphorylation of cGMP-inhibited Phosphodiesterase

following chromatographic separation on DEAE-cellulose (Groppi et al., 1983). In light of the above observations, it seems very likely that a cGMP-inhibited phosphodiesterase activity does reside in the 110-kDa peptide. The ultimate proof will require characterization of the pure protein.

The use of monoclonal antibodies directed against cGMP- inhibited isozyme permits the rapid isolation of unproteolyzed phosphodiesterase-antibody complex. Grant and Coleman (1984) have isolated a similar phosphodiesterase activity from human platelets that migrated as a 61-kDa peptide after SDS gel electrophoresis and presumably represents a proteolyzed form of the native enzyme. We have observed a time-dependent proteolysis of the platelet preparation which can result in the generation of 80- and 60-kDa peptides (Macphee et al., 1986). In this report, activation of the cGMP-inhibited phosphodiesterase was not associated with the appearance of 80- and 60-kDa peptides, since only the 110-kDa peptide band was identified by immunoblotting (see Fig. 1B). Furthermore, since the antibody also can recognize smaller fragments of the cGMP-inhibited phosphodiesterase (Harrison et al., 1986b), activation was not associated with phosphorylation of 80- and 60-kDa peptides. The 110-kDa protein band was the only phosphorylated band identified following prostaglandin treatment (see Fig. 3B) or following in vitro phosphorylation (Fig. 6B).

The regulation of cGMP-inhibited phosphodiesterase by CAMP-dependent phosphorylation implies an important role for this isozyme in platelet function. For example, it is important for platelet homeostasis to have a mechanism for rapidly lowering CAMP caused by thrombin-induced release of PGI, from healthy endothelial cells. This would then allow the platelets to respond appropriately at the point of injury. Activation of the cGMP-inhibited phosphodiesterase isozyme mediated by the PGIn-induced increase in protein kinase activity would seem to be one such mechanism. The inhibitory effects of thrombin on adenylate cyclase may be another such regulatory mechanism (Aktories and Jakobs, 1984). It is in- teresting to note that a greater response was observed for phosphorylation relative to phosphodiesterase activity (compare Figs. lA and 4) . Similarly, the basal value of phosphodiesterase activity varied greatly between donors (see Fig. 6). These observations may be a reflection of the level of endogenous phosphate. Alternatively, additional processes may be involved in regulating this activity. Currently, we are inves- tigating whether cGMP-inhibited phosphodiesterase is under complex control through the interplay of different protein kinases.

REFERENCES Aktories, K., and Jakobs, K. H. (1984) Eur. J . Biochem. 145 , 333-

Alvarez, R., Taylor, A., Fazzari, J. J., and Jacobs, J . R. (1981) Mol.

Beavo, J. A. (1988) Adu. Second Messenger Phosphoprotein Res. 2 , in

Brothers, V. M., Walker, N., and Bourne, H. R. (1982) J. Biol. Chem.

Feinstein, M. D., Rodan, G. A., and Cutler, L. S. (1981) Platelets in Biology and Pathology 2 (Gordon, J. L., ed) pp. 437-472, North- Holland, Amsterdam

Gettys, T. W., Vine, A. J., Simonds, M. F., and Corbin, J. D. (1988) J. Biol. Chem. 263,10359-10363

Goldberg, N. D., Walseth, T. F., Eide, S. J., Krick, T. P., Kuehn, B. L., and Gander, J. E. (1984) Adu. Cyclic Nucleotide Protein Phos- phorylation Res. 16,363-379

Grant, P. G., and Coleman, R. W. (1984) Biochemistry 2 3 , 1801- 1807

Groppi, V. E., Steinberg, F., Kaslow, H. R., Walker, N., and Bourne, H. R. (1983) J. Biol. Chem. 258,9717-9723

Hamet, P., Franks, D. J., Tremblay, J., and Coquil, J. F. (1983) Can. J. Biochem. Cell Biol. 6 1 , 1158-1165

Harrison, S. A., Chang, M. L., and Beavo, J. A. (1986a) Circulation

Harrison, S. A., Reifsnyder, D. H., Gallis, B., Cadd, G. G., and Beavo,

Haslam, R. J.. Lvnham. J. A., and Fox. J. E. B. (1979) Biochem. J .

338

Phurmacol. 20,302-309

press

257,9349-9355

73, Suppl. 3,109-116

J. A. (1986b) Mol. Phurmacol, 2 9 , 506-514

178,397-406 " Houslay, M. D., Wallace, A. V., Marchmont, R. J., Martin, B. R., and

Heworth, C. M. (1984) Adu. Cyclic Nucleotide Protein Phospho-

. .

rylation Res. 16 , 159-176

Chem. 259,8612-8618 Hurwitz, R. L., Bunt-Milam, A. H., and Beavo, J. A. (1984) J. Biol.

Laemmli, U. K. (1970) Nature 227,680-685 Loten, E. G., and Sneyd, J. G. T. (1970) Biochem. J. 120, 187-193 Loten, E. G., Assimacopoulos-Jeannet, F. D., Exton, J. H., and Park,

Macphee, C. H., Harrison, S. A., and Beavo, J. A. (1986) Proc. Natl.

Marchmont, R. J., and Houslay, M. D. (1980) Nature 286,904-906 Pawlson, L. G., Lovell-Smith, C. J., Manganiello, V. C., and Vaughan,

M. (1974) Proc. Natl. Acad. Sci. U. S. A. 7 1 , 1639-1642 Reifsnyder, D. H., Harrison, S. A., Macphee, C. H., McCormack, M.

B., and Beavo, J. A. (1985) Fed. Proc. Am. Soc. Exp. Bid. 44,1816

Scott, J . D., Glaccum, M. B., Fischer, E. H., and Krebs, E. G. (1986) (abstr.)

Proc. Natl. Acad. Sci. U. S. A. 83, 1613-1616 Sibley, D. R., and Lefkowitz, R. J. (1985) Nature 317,124-129 Tremblay, J., Lachance, B., and Hamet, P. (1985) J. Cyclic Nucleotide

Weber, H. W., and Appleman, M. M. (1982) J. Biol. Chem. 257,

Wells, J. N., and Hardman, J. G. (1977) Adv. Cyclic Nucleotide Res.

Whittle, B. J. R., Hamid, S., Lidbury, P., and Rosam, A. C. (1985)

C. R. (1978) J. Biol. Chem. 253,746-757

Acad. Sci. U. S. A. 83,6660-6663

Protein Phosphorylation Res. 10, 397-411

5339-5341

8, 119-143

Adu. Exp. Med. Biol. 192,109-125

phosphorylation results in activation of a camp ... · the journal of biological chemistry 0 1988...

Documents