Portal de Periódicos da CAPES

Plasma Membrane Ca2+-ATPase Associates with the Cytoskeleton in Activated Platelets through a PDZ-binding Domain

2001; Elsevier BV; Volume: 276; Issue: 18 Linguagem: Inglês

10.1074/jbc.m009850200

1083-351X

Martin Zabe, William L. Dean,

Platelet Disorders and Treatments

The plasma membrane Ca2+-ATPase (PMCA) plays an essential role in maintaining low cytosolic Ca2+ in resting platelets. During platelet activation PMCA is phosphorylated transiently on tyrosine residues resulting in inhibition of the pump that enhances elevation of Ca2+. Tyrosine phosphorylation of many proteins during platelet activation results in their association with the cytoskeleton. Consequently, in the present study we asked if PMCA interacts with the platelet cytoskeleton. We observed that very little PMCA is associated with the cytoskeleton in resting platelets but that ∼80% of total PMCA (PMCA1b + PMCA4b) is redistributed to the cytoskeleton upon activation with thrombin. Tyrosine phosphorylation of PMCA during activation was not associated with the redistribution because tyrosine-phosphorylated PMCA was not translocated specifically to the cytoskeleton. Because PMCA b-splice isoforms have C-terminal PSD-95/Dlg/ZO-1 homology domain (PDZ)-binding domains, a C-terminal peptide was used to disrupt potential PDZ domain interactions. Activation of saponin-permeabilized platelets in the presence of the peptide led to a significant decrease of PMCA in the cytoskeleton. PMCA associated with the cytoskeleton retained Ca2+-ATPase activity. These results suggest that during activation active PMCA is recruited to the cytoskeleton by interaction with PDZ domains and that this association provides a microenvironment with a reduced Ca2+concentration. The plasma membrane Ca2+-ATPase (PMCA) plays an essential role in maintaining low cytosolic Ca2+ in resting platelets. During platelet activation PMCA is phosphorylated transiently on tyrosine residues resulting in inhibition of the pump that enhances elevation of Ca2+. Tyrosine phosphorylation of many proteins during platelet activation results in their association with the cytoskeleton. Consequently, in the present study we asked if PMCA interacts with the platelet cytoskeleton. We observed that very little PMCA is associated with the cytoskeleton in resting platelets but that ∼80% of total PMCA (PMCA1b + PMCA4b) is redistributed to the cytoskeleton upon activation with thrombin. Tyrosine phosphorylation of PMCA during activation was not associated with the redistribution because tyrosine-phosphorylated PMCA was not translocated specifically to the cytoskeleton. Because PMCA b-splice isoforms have C-terminal PSD-95/Dlg/ZO-1 homology domain (PDZ)-binding domains, a C-terminal peptide was used to disrupt potential PDZ domain interactions. Activation of saponin-permeabilized platelets in the presence of the peptide led to a significant decrease of PMCA in the cytoskeleton. PMCA associated with the cytoskeleton retained Ca2+-ATPase activity. These results suggest that during activation active PMCA is recruited to the cytoskeleton by interaction with PDZ domains and that this association provides a microenvironment with a reduced Ca2+concentration. plasma membrane Ca2+-ATPase PSD-95/Dlg/ZO-1 homology domain phenylmethylsulfonyl fluoride Triton X-100 Elevation of cytosolic free Ca2+ is essential for efficient activation of platelets. Platelet agonists such as thrombin promote Ca2+ release from internal stores and entry through plasma membrane channels (1Rink T.S. Sage S.O. Annu. Rev. Physiol. 1990; 52: 431-449Crossref PubMed Scopus (321) Google Scholar). The predominant mechanism for maintenance of low intracellular Ca2+ in unstimulated platelets and the removal of Ca2+ after intracellular store depletion is efflux catalyzed by the plasma membrane Ca2+-ATPase (PMCA)1 (2Haynes D.H. Platelets. 1993; 4: 231-242Crossref PubMed Scopus (49) Google Scholar, 3Rosado J.A. Sage S.O. J. Biol. Chem. 2000; 275: 19529-19535Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). PMCA is a highly regulated enzyme, in terms both of diversity of gene expression and direct modification of enzyme activity (4Carafoli E. FASEB J. 1994; 8: 993-1002Crossref PubMed Scopus (368) Google Scholar). In the platelet, two isoforms are expressed, PMCA1b and PMCA4b (5Paszty K. Kovacs T. Lacabaratz-Porret C. Papp B. Enouf J. Filoteo A.G. Penniston J.T. Enyedi A. Cell Calcium. 1998; 24: 129-135Crossref PubMed Scopus (27) Google Scholar, 6Dean W.L. Chen D. Brandt P.C. Vanaman T.C. J. Biol. Chem. 1997; 272: 15113-15119Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar), and enzyme activity has been shown to be modulated by cAMP-dependent phosphorylation and tyrosine phosphorylation (6Dean W.L. Chen D. Brandt P.C. Vanaman T.C. J. Biol. Chem. 1997; 272: 15113-15119Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar). Recently, Rosado and Sage (3Rosado J.A. Sage S.O. J. Biol. Chem. 2000; 275: 19529-19535Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar) confirmed that PMCA is phosphorylated on tyrosine residues during platelet activation and showed that small GTPases of the Ras family participate in the signaling pathway in thapsigargin- and ionomycin-stimulated platelets. In addition to PMCA, many other platelet proteins are phosphorylated on tyrosine residues during platelet activation. This phosphorylation promotes association of membrane proteins such as the integrin αIIbβ3 and tyrosine kinases such as p60src, p125FAK with the cytoskeleton (7Fox J.E. Thromb. Haemostasis. 1993; 70: 884-893Crossref PubMed Scopus (207) Google Scholar, 8Fox J.E. Shattil S.J. Kinlough-Rathbone R.L. Richardson M. Packham M.A. Sanan D.A. J. Biol. Chem. 1996; 271: 7004-7011Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar) leading to cytoskeletal rearrangement and a dramatic change in platelet shape. Thus PMCA tyrosine phosphorylation could potentially promote association of PMCA with the cytoskeleton. Another potential mechanism for interaction of PMCA with the cytoskeleton has been described recently. The last four residues of PMCA4b (ETSV) contain the E(T/S)XV motif that has been identified as a target for binding to PDZ domains (9Saras J. Heldin C.H. Trends Biochem. Sci. 1996; 21: 455-458Abstract Full Text PDF PubMed Scopus (224) Google Scholar, 10Kim E. DeMarco S.J. Marfatia S.M. Chishti A.H. Sheng M. Strehler E.E. J. Biol. Chem. 1998; 273: 1591-1595Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar). PMCA1b as well as the -2b and -3b splice variants exhibit a similar C-terminal sequence of ETSL. Kim et al. (10Kim E. DeMarco S.J. Marfatia S.M. Chishti A.H. Sheng M. Strehler E.E. J. Biol. Chem. 1998; 273: 1591-1595Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar) demonstrated that C-terminal peptides (10 amino acids in length) and glutathioneS-transferase fusion proteins containing C-terminal sequences of both PMCA2b and PMCA4b interact with PDZ domains, whereas a PMCA2a peptide does not. PDZ domains are central organizers of protein complexes at the plasma membrane (10Kim E. DeMarco S.J. Marfatia S.M. Chishti A.H. Sheng M. Strehler E.E. J. Biol. Chem. 1998; 273: 1591-1595Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar, 11Garner C.C. Nash J. Huganir R.L. Trends Cell Biol. 2000; 10: 274-280Abstract Full Text Full Text PDF PubMed Scopus (482) Google Scholar) and mediate interactions with the cytoskeleton (12Bunn R.C. Jensen M.A. Reed B.C. Mol. Biol. Cell. 1999; 10: 819-832Crossref PubMed Scopus (164) Google Scholar, 13Wu H. Reuver S.M. Kuhlendahl S. Chung W.J. Garner C.C. J. Cell Sci. 1998; 111: 2365-2376Crossref PubMed Google Scholar). Thus both of the PMCA isoforms found in the platelet, PMCA1b and PMCA4b, could interact with PDZ domains in the cytoskeleton or other organized structures. In this study, we investigated the association of platelet PMCA with the cytoskeleton of resting and thrombin-activated platelets. We compared the cellular distribution of PMCA during platelet activation with that of p125FAK and investigated whether platelet aggregation and tyrosine phosphorylation of PMCA are required for this association. Finally, we asked if the C-terminal region containing the PDZ domain-binding residues is involved in cytoskeletal association. Antipain, apyrase (grade V), aspirin, leupeptin, sodium orthovanadate, thapsigargin, human thrombin, phenylmethylsulfonyl fluoride (PMSF), the synthetic peptide Arg-Gly-Asp-Ser (RGDS), and saponin were purchased from Sigma, and calpeptin was from Calbiochem. Chemiluminescence reagents were from PerkinElmer Life Sciences. Anti-focal adhesion kinase p125FAK and p60src were from Upstate Biotechnologies (Lake Placid, NY), anti-phosphotyrosine PY20 (horseradish peroxidase, conjugated and nonconjugated) was from BD Transduction Laboratories (Lexington, KY), and anti-PMCA monoclonal antibody (5F10) was from Affinity Bioreagents (Neshanic Station, NJ). A monoclonal antibody specific for PMCA4, JA9, was provided kindly by John T. Penniston (Mayo Clinic, Rochester, MN). Horseradish peroxidase-conjugated sheep anti-mouse IgG antibody was from Amersham Pharmacia Biotech. Platelet concentrates were obtained from the American Red Cross (Louisville, KY) and prepared as described previously (14Blankenship K.A. Dawson C.B. Aronoff G.R. Dean W.L. Hypertension. 2000; 35: 103-107Crossref PubMed Scopus (34) Google Scholar). Briefly, platelets were isolated by centrifugation and resuspended in Tyrode's-HEPES buffer containing 138 mm sodium chloride, 2.9 mm potassium chloride, 12 mm sodium bicarbonate, 0.36 mm sodium phosphate, 5.5 mmglucose, 1.8 mm calcium chloride, 0.4 mmmagnesium chloride, 10 mm HEPES, pH 7.4, 0.1 mmaspirin, and 0.2 unit/ml apyrase. Platelet suspensions were activated with 1.0 NIH unit of human thrombin/ml at room temperature. In some experiments platelets were preincubated for 5 min with the synthetic peptide RGDS (0.5 mm) in Hanks'-HEPES buffer (138 mm sodium chloride, 5.4 mm potassium chloride, 12 mm sodium bicarbonate, 0.36 mm sodium phosphate, 5.5 mm glucose, and 10 mm HEPES, pH 7.4). Platelets were lysed according to Fox et al. (15Fox J.E. Reynolds C.C. Boyles J.K. Methods Enzymol. 1992; 215: 42-58Crossref PubMed Scopus (51) Google Scholar) by addition of an equal volume of ice-cold lysis buffer containing 2% Triton X-100 (TX100), 10 mm EGTA, 100 mmTris-HCl, pH 7.4, 2 mg/ml leupeptin, 10 μg/ml antipain, 10 μg/ml pepstatin A, 1 mm dithiothreitol, 1 mm sodium orthovanadate, and 1 mm PMSF. Lysates were centrifuged at 16,000 × g at 4 °C for 5 min to sediment the cytoplasmic cytoskeletal components (low speed pellet). The supernatant then was centrifuged at 100,000 × g for 2.5 h at 4 °C in a TL100 rotor (Beckman Instruments) to sediment membrane cytoskeleton components (high speed pellet). Sedimented low speed material was washed once with 1 volume of lysis buffer diluted 1:1 with H2O and then resuspended in sample dilution buffer containing 62.5 mm Tris-HCl, pH 6.8, 2% SDS, 5% β-mercaptoethanol, 11% glycerol, 0.002% bromphenol blue, and 1 mm PMSF. Aliquots of the supernatant from the 100,000 × g centrifugation were solubilized with 3 volumes of sample dilution buffer, and the high speed pellet was resuspended in sample dilution buffer. All samples were heated to 42 °C for 20 min. Boiling was omitted to prevent aggregation of PMCA. The samples were analyzed by SDS-polyacrylamide gel electrophoresis immediately or stored at −20 °C. A peptide corresponding to the C-terminal 10 residues of isoform 4b of human PMCA (hPMCA4b) with the sequence SSLQSLETSV (10Kim E. DeMarco S.J. Marfatia S.M. Chishti A.H. Sheng M. Strehler E.E. J. Biol. Chem. 1998; 273: 1591-1595Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar) and a peptide with a scrambled sequence of the same amino acid composition with the sequence VTSLQSLSES were synthesized chemically in the Macromolecular Structure Analysis Facility of the University of Kentucky (Lexington, KY). The amino acid sequences were confirmed by liquid phase sequencing. To introduce the peptides into platelets, we used saponin to permeabilize the cells according to Authi et al. (16Authi K.S. Rao G.H. Evenden B.J. Crawford N. Biochem. J. 1988; 255: 885-893Crossref PubMed Scopus (16) Google Scholar). Platelets were isolated as described above, centrifuged for 2 min at 6,000 × g, resuspended in Hanks'-HEPES buffer, and centrifuged again. The cells then were resuspended in prewarmed (37 °C) buffer containing 140 mm KCl, 1 mm glucose, 1 mmMgCl2, 0.42 mm NaH2PO4, 6 mm NaHCO3, and 10 mm HEPES, pH 7.4. Saponin (20 μg/ml) and the peptides were added and the platelet suspension was incubated for 2 min. The platelets then were activated with thrombin (1 unit/ml) under constant stirring, and the reaction was stopped by adding an equal volume of lysis buffer. The low speed cytoskeletal material was isolated as described above. Samples were analyzed on 7.5% SDS-polyacrylamide gels according to Laemmli (17Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (212382) Google Scholar). After electrophoresis, proteins were transferred electrophoretically to 0.45-μm nitrocellulose membranes (Bio-Rad). Membranes were blocked for 1 h by incubation in 5% (w/v) nonfat dry milk (Carnation) in Tween 20-containing Tris-buffered saline (7.5 mm Tris, pH 7.5, 37.5 mm NaCl, and 0.1% (v/v) Tween 20) at room temperature. Membranes were rinsed twice and washed in Tween 20-containing Tris-buffered saline twice for 10 min each. The membrane was divided into two parts and PMCA was probed in the upper part of the membrane (above 90 kDa) with a dilution of 1:1,000 anti-PMCA antibody (5F10 or JA9) in antibody dilution buffer (2% bovine serum albumin in Tween 20-containing Tris-buffered saline). To probe for tyrosine-phosphorylated proteins or focal adhesion kinase p125FAK, the same portion of the membrane was stripped with buffer containing 62.5 mm Tris, pH 6.8, 2% SDS, and 0.7% β-mercaptoethanol and probed either with a 1:2,500 dilution of PY20-horseradish peroxidase or with 0.2 μg/ml anti-p125FAK monoclonal antibody in antibody dilution buffer for 1 h. The lower portion of the same membrane (below 90 kDa) was probed with 0.5 μg/ml anti-p60src monoclonal antibody. In the case of monoclonal nonconjugated antibodies (5F10 and JA9), the membranes were rinsed and incubated with a dilution of 1:10,000 of sheep anti-mouse-horseradish peroxidase antibody. Membranes were rinsed as above and then analyzed with chemiluminescence reagents. The signal was captured on x-ray films (Eastman Kodak Co.) and quantified after scanning on an Hewlett-Packard flat-bed scanner using Un-Scan-It software (Silk Scientific). Total signal was determined by adding the pixels from supernatant and low and high speed pellets after correcting for the volume of these three platelet fractions. Units of the total signal were in pixels/ml of platelet sample. Analysis of statistical significance was performed using the Student's unpaired ttest. Ca2+-ATPase activity was assayed using a system in which ATP hydrolysis is coupled to NADH oxidation by pyruvate kinase using phosphoenolpyruvate and lactate dehydrogenase (18Dean W.L. Methods Enzymol. 1988; 157: 351-360Crossref PubMed Scopus (19) Google Scholar). The low speed pellets obtained as described above were washed once in buffer containing 20 mm HEPES, pH 7.5, 130 mmNaCl, 1 mm MgCl2, 5% glycerol, 1 mm dithiothreitol, and 1 mm PMSF. The pellets were resuspended with sonication for 15 s on ice in 500 μl of the same buffer with 10 μg/ml leupeptin instead of PMSF. The material was assayed for Ca2+-ATPase activity in buffer containing 20 mm HEPES, pH 7.5, 130 mm NaCl, 10 mm MgCl2, 1 mm ATP, 0.01 mm CaCl2, 50 nm calmodulin, and the coupled assay components. The difference in the rate of ATP hydrolysis in the presence and absence of 5 mm EGTA was used to calculate the Ca2+-ATPase activity. Protein content was determined with the BCA protein assay kit (Pierce) using bovine serum albumin as a standard. Most of the cytoplasmic actin filaments in unstimulated platelets are aggregated into large complexes that sediment from TX100-lysed platelets at 16,000 × g (low speed pellet) (19Fox J.E. Taylor R.G. Taffarel M. Boyles J.K. Goll D.E. J. Cell Biol. 1993; 120: 1501-1507Crossref PubMed Scopus (134) Google Scholar). However, the membrane-associated skeletal fragments and associated proteins such as integrin αIIbβ3, spectrin, vinculin, and p60src require 100,000 × g for sedimentation from TX100-lysed unstimulated platelets (high speed pellet) (20Fox J.E. Lipfert L. Clark E.A. Reynolds C.C. Austin C.D. Brugge J.S. J. Biol. Chem. 1993; 268: 25973-25984Abstract Full Text PDF PubMed Google Scholar). To investigate the redistribution of PMCA during platelet activation, platelets were lysed in a buffer containing 1% TX100, and the cytoplasmic cytoskeletal components were isolated by centrifugation at 16,000 ×g (low speed pellet) followed by centrifugation at 100,000 × g to isolate the membrane cytoskeleton (high speed pellet). Analysis of PMCA distribution after activation with thrombin was compared for unstirred platelets, stirred platelets, and stirred platelets in the presence of RGDS, an inhibitor of platelet aggregation, to determine whether PMCA localization was dependent on cell-cell interaction as has been demonstrated for integrin αIIbβ3 and p60src (21Shattil S.J. Haimovich B. Cunningham M. Lipfert L. Parson J.T. Ginsberg M.H. Brugge J.S. J. Biol. Chem. 1994; 269: 14738-14745Abstract Full Text PDF PubMed Google Scholar). As shown in Fig.1 a, the majority of PMCA in unstimulated platelets (0 min) was associated with the TX100-soluble supernatant (70%) and the high speed pellet (20%) with only 10% in the low speed pellet. Thrombin-mediated activation resulted in redistribution of PMCA to the low speed pellet; 30 min after addition of thrombin, 80% of the PMCA was redistributed to the low speed pellet. Stirring enhanced the rate and extent of PMCA redistribution (Fig. 1 b) resulting in ∼75% of total PMCA being associated with the low speed pellet 5 min after thrombin addition. Inhibition of aggregation with RGDS (Fig. 1 c) inhibited this redistribution with only 20% of total PMCA associated with the low speed pellet 30 min after thrombin addition. The amount of PMCA detected in the immunoblots of the supernatant fractions in Fig. 1,a and b (see insets) is small because of the large volume of supernatant and resultant dilution compared with the much smaller volumes and higher concentrations of low speed and high speed pellets. The second band in the inset labeledPMCAfrag is apparently a proteolytic fragment of PMCA because it was eliminated by the addition of calpeptin (30 μm) to thrombin-activated platelets (data not shown). These results demonstrate extensive redistribution of PMCA to the cytoskeleton that requires cell-cell contacts as shown by stimulation with stirring and inhibition with RGDS. These results are nearly identical to those reported for αIIbβ3 and p60src by Fox et al. (20Fox J.E. Lipfert L. Clark E.A. Reynolds C.C. Austin C.D. Brugge J.S. J. Biol. Chem. 1993; 268: 25973-25984Abstract Full Text PDF PubMed Google Scholar). Similar results were obtained when redistribution of p125FAK was analyzed as shown in Fig.2. Comparison of Figs. 1 and 2 indicates that both PMCA and p125FAK are redistributed to the cytoskeleton in an aggregation-dependent reaction initiated by thrombin-dependent activation under the conditions used in our experiments. Because we reported earlier that platelets contain both PMCA1b and PMCA4b (6Dean W.L. Chen D. Brandt P.C. Vanaman T.C. J. Biol. Chem. 1997; 272: 15113-15119Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar), it is important to determine which isoforms are associating with the cytoskeleton. Kim et al. (10Kim E. DeMarco S.J. Marfatia S.M. Chishti A.H. Sheng M. Strehler E.E. J. Biol. Chem. 1998; 273: 1591-1595Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar) demonstrated that both PMCA1b and PMCA4b bind to PDZ domains, although PMCA4b was observed to have significantly higher affinity. To determine the relative distribution of the two isoforms in the cytoskeleton, monoclonal antibodies 5F10 and JA9 were employed (22Borke J.L. Caride A. Verma A.K. Penniston J.T. Kumar R. Am. J. Physiol. 1989; 257: F842-F849Crossref PubMed Google Scholar, 23Brini M. Bano D. Manni S. Rizzuto R. Carafoli E. EMBO J. 2000; 19: 4926-4935Crossref PubMed Google Scholar); 5F10 recognizes both isoforms, whereas JA9 only binds PMCA4. As shown in Fig. 3, we used the ratio of the signals obtained with JA9 and 5F10 to estimate the relative amounts of these two isoforms in whole platelets and the cytoplasmic cytoskeletal fraction of activated platelets. The JA9/5F10 ratio is equivalent tox(PMCA4b)/(y(PMCA1b) + x(PMCA4b)), where x and y depend on antibody affinity, antibody dilution, and chemiluminescence development conditions. Because x and y are unknowns caused by lack of pure standards, the ratios can be used only to compare relative amounts of the two isoforms on a single immunoblot where x andy are the same for all samples on the blot. The results show that compared with the ratio in whole platelets (Whole Plt), there is not a significant change in the ratio of PMCA4b to total PMCA in the pellet (Pellet) fractions in thrombin-activated platelets. This suggests that both PMCA1b and PMCA4b become associated with the cytoskeleton during platelet activation. Another possibility is that there is little PMCA1b in the platelet as reported by Pasztyet al. (5Paszty K. Kovacs T. Lacabaratz-Porret C. Papp B. Enouf J. Filoteo A.G. Penniston J.T. Enyedi A. Cell Calcium. 1998; 24: 129-135Crossref PubMed Scopus (27) Google Scholar) so that even if PMCA1b were not associated with the low speed pellet, the JA9/5F10 ratio would not change significantly. Because p60src and p125FAK associate with the cytoskeleton as a result of tyrosine phosphorylation, we investigated the relative levels of PMCA tyrosine phosphorylation in the TX100 fractions after lysis of thrombin-activated platelets. Tyrosine phosphorylation was assessed using anti-phosphotyrosine immunoblotting with the PY20 antibody, whereas the extent of PMCA tyrosine phosphorylation was estimated by comparing the phosphotyrosine and PMCA signals at the migration position of PMCA. Thus the extent of PMCA phosphorylation is presented as the PY20/PMCA ratio. As shown in Fig. 4, thrombin addition under stirred conditions promoted PMCA tyrosine phosphorylation in the supernatant (PY20/PMCA ratio of ∼3.0 at 5–30 min) and in the high speed pellet (PY20/PMCA ratio of ∼6.0 at 5–15 min) from an initial ratio (0 min) of ∼1.0. No increase in the PY20/PMCA ratio was observed in the low speed pellet. Although in theinsets (one example of the signal with PY20 for each fraction) it appears that the majority of tyrosine phosphorylation occurs in the low speed pellet, this is also the location of the majority (85%) of the PMCA in activated platelets thus resulting in a relatively low PY20/PMCA ratio (∼1.0). These results suggest that unlike p60src and p125FAK, PMCA redistribution to the cytoskeleton is not dependent on tyrosine phosphorylation of PMCA, because the PY20/PMCA ratio is not increased in the low speed pellets. Another possible mechanism for PMCA association with the cytoskeleton is interaction of the C-terminal PDZ-binding domain in PMCA1b and PMCA4b with cytoskeletal PDZ domain-containing proteins. Kim et al. (10Kim E. DeMarco S.J. Marfatia S.M. Chishti A.H. Sheng M. Strehler E.E. J. Biol. Chem. 1998; 273: 1591-1595Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar) demonstrated that a peptide representing the C terminus of PMCA4b, SSLQSLETSV, bound with high affinity to the PDZ1+2 domain of the human homolog ofDrosophila discs-large protein. We therefore tested the ability of this peptide to disrupt PMCA-cytoskeletal interactions in saponin-permeabilized thrombin-activated platelets. A scrambled peptide composed of the same amino acids, VTSLQSLSES, was used as a negative control. The results presented in Fig.5 a show that the C-terminal peptide (PDZ) prevented association of PMCA with the low speed pellet in permeabilized thrombin-stimulated platelets, whereas the scrambled peptide (scr.) had a much smaller effect on PMCA association with the cytoskeleton. Addition of saponin alone had no effect on the redistribution of PMCA to the cytoskeleton as shown by the lack of effect of saponin addition in the absence of added peptides. Although the C-terminal peptide prevented PMCA association with the cytoskeleton, it had very little effect on the translocation of p60src and p125FAK to the cytoskeleton compared with the effect of the scrambled peptide (Fig. 5 b). This was expected because p60src and p125FAK associate with the cytoskeleton by means of tyrosine phosphates rather than through PDZ domain interactions. Fig. 5 c shows that the PDZ peptide significantly inhibited PMCA association with the cytoskeleton in four different platelet preparations, whereas the effect of the scrambled peptide was not significantly different from the control. Having established that PMCA becomes associated with the actin cytoskeleton during platelet activation by means of binding to cytoskeletal PDZ domains, it became important to establish a physiological role for this phenomenon. If association of PMCA with the cytoskeleton causes localization of the pump in specific areas of the activated platelet, then it is essential that the pump retains its biological activity. We therefore measured the Ca2+-ATPase activity in the isolated cytoskeleton. As shown in Fig. 6, cytoskeletal Ca2+-ATPase activity was measured as a function of time after thrombin addition (shaded bars) in the presence of 200 nm thapsigargin to inhibit sarco(endo)plasmic reticulum Ca2+-ATPase-type ATPases. Maximal activity was observed 15 min after the addition of thrombin, and the specific activity is in the range observed for platelet plasma membranes, 10–40 nmol/min/mg (24Dean W.L. Quinton T.M. Cell Calcium. 1995; 17: 65-70Crossref PubMed Scopus (23) Google Scholar). This activity correlates well with the amount of PMCA associated with cytoskeleton (open bars). The enzymatic characteristics of the cytoskeletal Ca2+-ATPase activity were examined and the results are presented in TableI. The Ca2+-ATPase activity in the low speed pellet was inhibited 43% by 10 μmorthovanadate, an indication that a portion of the ATPase activity is due to PMCA because this ATPase is inhibited strongly at 10 μm vanadate, whereas sarco(endo)plasmic reticulum Ca2+-ATPase-type ATPases require 100 μmvanadate for complete inhibition (25Caroni P. Carafoli E. J. Biol. Chem. 1981; 256: 3263-3270Abstract Full Text PDF PubMed Google Scholar). Furthermore, the ATPase activity is partially stimulated by calmodulin (18%), an activator of PMCA. This relatively modest level of stimulation may indicate the presence of endogenous calmodulin in the isolated cytoskeleton because no attempt was made to strip the pellet of calmodulin. The low level of inhibition by thapsigargin (11%) indicates low abundance of sarco(endo)plasmic reticulum Ca2+-ATPase in the pellet. Taken together these data indicate that a portion of the Ca2+-ATPase activity in the isolated cytoskeleton is due to the presence of PMCA.Table IProperties of Ca2+-ATPase activity in the low speed pelletAssay conditionsSpecific activityInhibition(nmol/min/mg) ± S.E.%Complete assay28 ± 20− Calmodulin23 ± 218+ Calmodulin+ Thapsigargin (200 nm)25 ± 211+ Calmodulin+ Vanadate (10 μm)16 ± 143The low speed pellet was prepared from thrombin-activated platelets and the Ca2+-ATPase activity measured as described under "Experimental Procedures." A preparation of low speed pellet from a single bag of platelets was assayed in quadruplicate and activities are averages of four assays ± S.E. The complete assay contained ATPase assay components plus 50 nm calmodulin. Open table in a new tab The low speed pellet was prepared from thrombin-activated platelets and the Ca2+-ATPase activity measured as described under "Experimental Procedures." A preparation of low speed pellet from a single bag of platelets was assayed in quadruplicate and activities are averages of four assays ± S.E. The complete assay contained ATPase assay components plus 50 nm calmodulin. PMCA is the major Ca2+ efflux mediator in human platelets, and it therefore plays a fundamental role in platelet intracellular Ca2+ homeostasis (2Haynes D.H. Platelets. 1993; 4: 231-242Crossref PubMed Scopus (49) Google Scholar, 3Rosado J.A. Sage S.O. J. Biol. Chem. 2000; 275: 19529-19535Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). Changes in the activity and/or distribution of PMCA might have a crucial impact on platelet functions. Our earlier finding that platelet PMCA is phosphorylated on tyrosine residues and inhibited in hypertensive individuals resulting in increased cytosolic Ca2+demonstrates the importance of this protein for proper platelet function (14Blankenship K.A. Dawson C.B. Aronoff G.R. Dean W.L. Hypertension. 2000; 35: 103-107Crossref PubMed Scopus (34) Google Scholar). After the initial increase of intracellular Ca2+ after activation by strong agonists such as thrombin, platelet shape changes dramatically because of rearrangements of the cytoskeleton. The association of many proteins with the actin cytoskeleton such as the integral membrane protein integrin αIIbβ3and cytosolic nonreceptor tyrosine kinases such as p60src and p125FAK requires tyrosine phosphorylation (20Fox J.E. Lipfert L. Clark E.A. Reynolds C.C. Austin C.D. Brugge J.S. J. Biol. Chem. 1993; 268: 25973-25984Abstract Full Text PDF PubMed Google Scholar). However, in this report we show that tyrosine phosphorylation of PMCA does not promote association of this protein with the cytoskeleton because the greatest extent ((tyrosine phosphates)/(PMCA)) of PMCA tyrosine phosphorylation in thrombin-activated platelets is either in the supernatant or high speed pellet (Fig. 4). Thus it is the more active, less phosphorylated form that associates with the cytoplasmic cytoskeleton. This implies that the purpose of redistribution of PMCA is to direct the active Ca2+ pump to specific locations in the activated platelet, possibly to reduce intracellular Ca2+ in focal contact regions thus preventing proteolytic degradation by the Ca2+-activated proteinase calpain (19Fox J.E. Taylor R.G. Taffarel M. Boyles J.K. Goll D.E. J. Cell Biol. 1993; 120: 1501-1507Crossref PubMed Scopus (134) Google Scholar). The concomitant inhibition of the pump not associated with the cytoskeleton by tyrosine phosphorylation presumably allows for higher accumulation of cytosolic Ca2+ in the bulk of the activated platelet. We showed that the C-terminal PDZ-binding domain of PMCA is involved in the redistribution of PMCA1b and PMCA4b to the actin cytoskeleton (Fig.5). The C-terminal peptide of PMCA4b specifically inhibited PMCA association but had no effect on the tyrosine phosphorylation-mediated association of p60src and p125FAK with the cytoskeleton. Furthermore, a scrambled peptide composed of the same amino acid residues had no effect on PMCA association with the low speed pellet. This provides strong evidence for the role of the C terminus of b-spliced PMCA isoforms in association with the platelet cytoskeleton during activation. Kim et al. (10Kim E. DeMarco S.J. Marfatia S.M. Chishti A.H. Sheng M. Strehler E.E. J. Biol. Chem. 1998; 273: 1591-1595Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar) speculated that the role of the PMCA PDZ-binding domain in neuronal cells may be to direct the pump into dendritic spines and restrict signaling by theN-methyl-d-aspartate receptor. We propose a similar redistribution of platelet PMCA into focal contacts, possibly in filopodia, to regulate cytosolic Ca2+ in these microdomains. PDZ domains have been shown to be involved in multiprotein complexes often linking plasma membrane proteins to stable subcellular structures (11Garner C.C. Nash J. Huganir R.L. Trends Cell Biol. 2000; 10: 274-280Abstract Full Text Full Text PDF PubMed Scopus (482) Google Scholar, 12Bunn R.C. Jensen M.A. Reed B.C. Mol. Biol. Cell. 1999; 10: 819-832Crossref PubMed Scopus (164) Google Scholar). In contrast, PMCA in the platelet only becomes associated with the cytoskeleton after activation and cytoskeletal rearrangement. The mechanism for this behavior could be similar to that described for neuronal F-actin binding proteins such as neurabin-I and -II (26Satoh A. Nakanishi H. Obaishi H. Wada M. Takahashi K. Satoh K. Hirao K. Nishioka H. Hata Y. Mizoguchi A. Takai Y. J. Biol. Chem. 1998; 273: 3470-3475Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar). These PDZ domain-containing F-actin binding proteins become associated with polymerized actin, but dissociate upon depolymerization of the cytoskeleton. Thus, PMCA could associate with the cytoskeleton in activated platelets by interacting with a PDZ domain containing F-actin binding protein that binds to the reorganized cytoskeleton after activation. Although neurabins have not been reported in platelets, a protein with similar properties, CLP-36, has been described recently (27Bauer K. Kratzer M. Otte M. Luber K. Essler M. Hagmann J. Arnold G. Eckerskorn C. Lottspeich F. Siess W. Blood. 2000; 96: 4236-4245Crossref PubMed Google Scholar). This PDZ domain-containing protein binds to α-actinin in the cytoskeleton of activated platelets by means of a distinct intervening sequence thus leaving the PDZ domain available for interaction with PDZ domain-binding proteins such as PMCA. In conclusion, we have demonstrated that PMCA is not associated with the cytoskeleton in resting platelets and that association increases dramatically upon activation with thrombin. Tyrosine phosphorylation does not promote the translocation, but the C-terminal PDZ-binding domain is required for PMCA redistribution. This is the first direct demonstration of a physiologically relevant role for the C-terminal PDZ binding sequence in PMCA4b.