Role of platelet plasma membrane Ca2+-ATPase in health and disease

Abstract

Platelets have essential roles in both health and disease. Normal platelet function is required for hemostasis. Inhibition of platelet function in disease or by pharmacological treatment results in bleeding disorders. On the other hand, hyperactive platelets lead to heart attack and stroke. Calcium is a major second messenger in platelet activation, and elevated intracellular calcium leads to hyperactive platelets. Elevated platelet calcium has been documented in hypertension and diabetes; both conditions increase the likelihood of heart attack and stroke. Thus, proper regulation of calcium metabolism in the platelet is extremely important. Plasma membrane Ca²⁺-ATPase (PMCA) is a major player in platelet calcium metabolism since it provides the only significant route for calcium efflux. In keeping with the important role of calcium in platelet function, PMCA is a highly regulated transporter. In human platelets, PMCA is activated by Ca²⁺/calmodulin, by cAMP-dependent phosphorylation and by calpain-dependent removal of the inhibitory peptide. It is inhibited by tyrosine phosphorylation and calpain-dependent proteolysis. In addition, the cellular location of PMCA is regulated by a PDZ-domain-dependent interaction with the cytoskeleton during platelet activation. Rapid regulation by phosphorylation results in changes in the rate of platelet activation, whereas calpain-dependent proteolysis and interaction with the cytoskeleton appears to regulate later events such as clot retraction. In hypertension and diabetes, PMCA expression is upregulated while activity is decreased, presumably due to tyrosine phosphorylation. Clearly, a more complete understanding of PMCA function in human platelets could result in the identification of new ways to control platelet function in disease states.

Key Words: Plasma membrane Ca²⁺-ATPase; Human platelets; Ca²⁺ transport; Signaling; Cytoskeleton; Phosphorylation; PDZ domain

INTRODUCTION

The normal physiological function of platelets is to maintain hemostasis, which is accomplished by platelet aggregation and initiation of clot formation at the site of damage to a blood vessel. Aberrant platelet function has severe pathological consequences because platelet-mediated thrombus formation leads to heart attacks and strokes. An increase in intracellular Ca²⁺ is a major signaling event in the activation of platelets, the plasma membrane Ca²⁺-ATPase (PMCA) prevents inappropriate activation by maintaining low cytoplasmic Ca²⁺. This review describes the role of PMCA in resting and activated platelets in health and disease.

PLATELET Ca²⁺ SIGNALING

An increase in intracellular calcium ([Ca²⁺]_i) is a major signal for platelet activation and accompanies activation by all agonists under physiological conditions[1]. Ca²⁺ is first released from intracellular stores termed the dense tubular system in response to formation of inositol trisphosphate from plasma membrane phosphatidyl inositol-(4,5)-bisphosphate. This step is mediated by activation of G-protein-coupled receptors and activation of phospholipase C. Release from intracellular stores is followed by influx of extracellular Ca²⁺via store-mediated Ca²⁺ entry and the P2X receptor. The increase in [Ca²⁺]_i is very rapid (milliseconds to seconds) followed by a slower return to lower levels brought about by PMCA and SERCA (Ca²⁺-pump located in the dense tubular system). In platelets, the plasma membrane Na⁺/Ca²⁺ exchanger and mitochondria do not contribute significantly to reduction in [Ca²⁺]_i[1], although the exchanger contributes to calcium influx in collagen-activated platelets[2]. Modulation of [Ca²⁺]_i has been shown to be an important regulator of platelet structural changes following the initial aggregation process[3], and of thrombus growth[4]. Jackson and colleagues have demonstrated that prolonged elevation of [Ca²⁺]_i and resultant calpain activation causes platelet fragmentation and formation of microbodies that limits the growth of the thrombus in collagen-activated platelets under flow conditions. PMCA is regulated by calpain, as described in more detail below. Furthermore, platelet-platelet interactions in the growing thrombus lead to Ca²⁺ signaling throughout the thrombus, which is required for individual platelets to remain bound to the thrombus. Thus, the rate at which PMCA pumps Ca²⁺ from the platelets after activation probably has important functional consequences for platelet activation, thrombus formation and thrombus maintenance.

PMCA IN HUMAN PLATELETS

PMCA is a P-type plasma membrane Ca²⁺-pumping ATPase that contains 10 transmembrane sequences. Most of its protein mass is in the cytoplasm. Importantly, PMCA is stimulated by Ca²⁺/calmodulin and therefore responds directly to [Ca²⁺]_i[5-7]. The accepted function of PMCA is to maintain low [Ca²⁺]_i by catalyzing ATP-dependent Ca²⁺ efflux, although new roles have recently been identified such as organization of protein complexes at the plasma membrane[8]. There are four independent genes from which alternatively spliced isoforms are expressed, which results in a total of approximately 30 varieties of PMCA[6,7]. Isoforms PMCA1b and PMCA4b have both been detected in human platelets[9,10], although the expression of PMCA1b appears to be very low[11]. Thus, PMCA4b is the major isoform that provides for Ca²⁺ efflux in the human platelets. Johansson et al[12] and more recently Rosado et al[13] have demonstrated the prominent role of PMCA in Ca²⁺ extrusion in human platelets and the lack of contribution of Na⁺/Ca²⁺ exchange. It has been shown that PMCA in human platelets is regulated by cAMP-dependent phosphorylation[9], tyrosine phosphorylation[9,14-17], and blood pressure[16-19], as shown in Table 1. It is clear that PMCA plays an important role in regulating platelet [Ca²⁺]_i and that control of PMCA activity has significant functional consequences[9,13] that are discussed in more detail.

Table 1 Modes of regulation of plasma membrane Ca²⁺-ATPase in human platelets.

Type of regulation	Effect on PMCA	Ref.
Ca2+/calmodulin	Stimulation	[20]
Protein-kinase-A-dependent phosphorylation	Stimulation	[9,12]
Tyrosine phosphorylation	Inhibition	[9,13,16,21]
Calpain-mediated cleavage (124 kDa)	Stimulation	[22]
Calpain-mediated cleavage (100 kDa and smaller)	Inhibition	[22]
Transcriptional/translational	Up- or downregulation depending on isoform	[23,24]

PMCA: Plasma membrane Ca²⁺-ATPase.

REGULATION OF PMCA BY PHOSPHORYLATION

cAMP

In 1992, Johansson et al[19] demonstrated that increased platelet cAMP led to an increased rate of Ca²⁺ extrusion, although direct phosphorylation of PMCA was not demonstrated. PMCA is indeed a substrate for protein kinase A, and phosphorylation results in activation of pump activity[25]. In 1997, Dean et al[9] demonstrated directly that treating platelets with prostaglandin E1, which raises platelet cAMP levels, resulted in direct phosphorylation of PMCA, which could explain the earlier observation of enhanced Ca²⁺ efflux[19].

Tyrosine phosphorylation

Dean et al[9] also have demonstrated that PMCA is phosphorylated at tyrosine residues during platelet activation by thrombin. The kinetics of tyrosine phosphorylation during platelet activation correspond with activation of focal adhesion kinase (FAK) and rearrangement of the cytoskeleton. Phosphorylation occurs on Tyr 1176 in PMCA 4b when purified PMCA is treated with the tyrosine kinase src in vitro, which results in significant inhibition of PMCA activity[9]. Rosado et al[13] have confirmed our observation of PMCA tyrosine phosphorylation during platelet activation and concomitant inhibition of Ca²⁺ efflux following agonist-mediated increases in [Ca²⁺]_i. Wan et al[14] have shown that Tyr 1176 is the only tyrosine residue that is phosphorylated on PMCA4b during platelet activation in vivo, and that the probable kinase responsible for this phosphorylation is FAK, based on the amino acid sequence that surrounds Tyr 1176 and the kinetics of FAK activation. Recently, Bozulic et al[21] have demonstrated that inhibition of PMCA4b tyrosine phosphorylation by introduction of an inhibitory peptide of the same sequence as the phosphorylation site on PMCA results in a significant decrease in [Ca²⁺]_i during platelet activation. Introduction of the peptide inhibited PMCA tyrosine phosphorylation by 60% and significantly delayed the onset of thrombin-mediated platelet aggregation. Taken together, these results demonstrate that tyrosine phosphorylation of PMCA4b during platelet activation inhibits PMCA-dependent Ca²⁺efflux and provides a positive feedback loop mechanism for enhancing the increase in [Ca²⁺]_i during platelet activation.

REGULATION OF PMCA BY CALPAIN

It is well established that calpain is activated by the increase in [Ca²⁺]_i that accompanies platelet activation, and that the proteolytic events that follow calpain activation significantly affect signaling during activation[3]. PMCA has been demonstrated to be a substrate for calpain both in vitro[26,27] and in human erythrocytes[28]. Calpain-dependent cleavage of PMCA first removes the C-terminal auto-inhibitory domain that results in formation of a 124-kDa PMCA fragment, whereas further calpain-dependent cleavage catalyzes formation of 100-kDa and smaller fragments[26,28]. The 124-kDa fragment is fully active and no longer regulated by calmodulin, whereas the 100-kDa and smaller fragments are inactive. Thus, calpain has the potential to activate and inhibit PMCA irreversibly. Brown et al[22] have demonstrated that PMCA is cleaved to smaller fragments during platelet activation with a time course similar to src and SNAP 23 cleavage by calpain[29,30]. Approximately 60% of PMCA is cleaved within 18 min after activation with thrombin or collagen to 124-kDa and smaller species. The significance of this event during platelet activation is that it maintains a portion of PMCA in a form that is fully active even as [Ca²⁺]_i is decreasing. In contrast to the 124-kDa species, the 100-kDa form is probably inactive[26,27,31]. Furthermore, the 100-kDa species is also likely to have the C terminus removed and thus become unable to interact with the cytoskeleton (see below). However, it has not been established whether N- or C-terminal fragments are removed to form the 100-kDa species during platelet activation[22]. This regulation of PMCA probably has important consequences for later stages of platelet activation such as aggregation and clot retraction.

PMCA ASSOCIATION WITH THE CYTOSKELETON IN HUMAN PLATELETS

PDZ domains

Several types of protein-protein interactions lead to complexes of membrane proteins with cytoplasmic proteins. One of these interactions utilizes the PDZ domain, a structural domain that was initially discovered in proteins localized to postsynaptic structures in neural tissue[32]. The PDZ domain binds tightly to specific tetrapeptide sequences at the C terminus of membrane-associated proteins. Membrane proteins that exhibit this C-terminal motif include ion channels, neurotransmitter receptors and PMCA[32-34]. Strehler and coworkers have shown that PMCA isoforms 2b and 4b interact with guanylate kinase family proteins via their PDZ domains[25,34]. As described below, we have demonstrated that PMCA4b binds to the cytoskeleton in activated platelets via its PDZ binding motif[20]. Platelets contain several PDZ-domain-containing proteins including SAP97[35], CLP-36[35,36] and PDZ-GEF1[37]. SAP97 and CLP-36 possess multiple protein interaction domains and thus are able to connect PMCA to the cytoskeleton by interacting with other proteins that are capable of binding to the cytoskeleton, such as band 4.1/spectrin (binds SAP97) and α-actinin (binds CLP-36).

Cytoskeletal rearrangement in platelet activation

Resting platelets exhibit well-defined cytoskeletal structures: cytoplasmic actin filaments and a membrane skeleton located just under the plasma membrane that consists of both actin filaments and microtubules[38]. Upon activation, there is extensive rearrangement of the cytoskeleton; the proportion of total actin in filaments increases rapidly from 30% to 70%[39]. Prior to platelet aggregation, cytoskeletal changes result in altered platelet morphology such as formation of filopodia, aid in secretion of stored contents from granules, and are associated with activation of the fibrinogen receptor, αIIbβ3 integrin[40]. Activation of the fibrinogen receptor results in fibrinogen binding that leads to cell-cell interactions and platelet aggregation. Tyrosine phosphorylation is intimately involved in the process of platelet aggregation and cytoskeletal rearrangement. Seconds after platelets bind an activator such as thrombin, src and syk non-receptor tyrosine kinases are activated. This results in a wave of tyrosine phosphorylation and association of phosphorylated proteins with the cytoskeleton[41]. Activation of αIIbβ3 integrin and subsequent binding of fibrinogen results in a second wave of tyrosine phosphorylation, including activation of FAK and the binding of additional proteins to the cytoskeleton including src, FAK[42] and PMCA[20,43].

PMCA translocation to filopodia

Many platelet proteins become associated with cytoplasmic actin cytoskeleton during platelet activation[38,40,41]. We have also shown this to be true for PMCA[20]. We showed that approximately 75% of the PMCA becomes associated with the cytoskeleton and remains associated long after PMCA tyrosine phosphates are removed. PMCA associated with the cytoskeleton retains Ca²⁺-ATPase activity which indicates that a change in cellular location could greatly affect local [Ca²⁺]_i. Attachment of ion transporters to the cytoskeleton has been shown to occur in several cell types[44], but our work has provided the first example of PMCA association with the cytoskeleton. We also have shown that association with the cytoskeleton is inhibited by introduction of a PMCA4b C-terminal peptide into platelets. This peptide contains the PDZ-domain-binding motif described above.

In a more recent publication[43], we have demonstrated, using immunofluorescence microscopy, that PMCA is translocated to filopodia in activated platelets and that this translocation is blocked by introduction of the PMCA4b C-terminal peptide. Although incorporation of the peptide has no effect on the rate and extent of platelet activation, it significantly increases the rate of clot retraction. These results indicate that PMCA translocation to filopodia is important for regulating late events in platelet thrombus formation such as clot retraction. We have speculated[43] that clot retraction rate is enhanced when PMCA translocation is inhibited because Ca²⁺ levels in filopodia are elevated due to lack of PMCA, and that this enhances the retraction process.

Mechanism of PMCA association with the cytoskeleton

In order to understand the mechanism of PMCA association with the cytoskeleton in activated platelets, co-immunoprecipitation assays coupled with immunoblotting and electrospray ionization tandem mass spectrometry have been used to identify proteins that interact with PMCA in resting platelets[35]. Our results have indicated that the LIM family protein, CLP-36, binds to PMCA in resting platelets and mediates binding of PMCA to the cytoskeleton during platelet activation. In addition, PMCA is associated with α-actinin and γ-actin in resting platelets. This implies that PMCA is already associated with small actin complexes in resting platelets, by means of PDZ domain interactions. PMCA then associates with the actin cytoskeleton during cytoskeletal rearrangement upon platelet activation (Figure 1). These observations suggest complex regulation of PMCA by interactions with anchoring and cytoskeletal proteins in addition to the reversible serine/threonine and tyrosine phosphorylation events we have previously described in human platelets.

Figure 1 Model for association of plasma membrane Ca²⁺-ATPase with the cytoskeleton and translocation to the filopodia during platelet activation. Plasma membrane Ca²⁺-ATPase (PMCA) is associated with CLP-36, actinin and actin in the resting platelet in small complexes. Upon platelet activation (arrow), actin polymerizes into long fibers resulting in formation of filopodia and translocation of PMCA into the filopodia where it remains active in the plasma membrane. Taken from reference 33 with permission from Schattauer GmbH, Stuttgart.

REGULATION OF PLATELET PMCA IN DISEASE

There is an extensive literature on changes in platelet [Ca²⁺]_i in a variety of diseases including hypertension, diabetes and coronary heart disease. In this review, I limit the discussion to effects of disease on platelet PMCA. In 1986, Resink et al[45] reported that a calmodulin-sensitive ATPase, presumably PMCA, had significantly increased activity in platelet membranes obtained from hypertensive individuals. A similar study published in 1992[46] reached the opposite conclusion, when it was shown that calcium efflux activity was significantly inhibited in hypertension. In 1994, Dean et al[15] reported that PMCA activity was decreased as a function of diastolic blood pressure in humans. We speculated that a factor present in the circulation of hypertensive individuals causes inhibition of PMCA and resultant increased [Ca²⁺]_i. After demonstrating that PMCA is inhibited by tyrosine phosphorylation[9], it has become clear that tyrosine phosphorylation of PMCA could explain the inhibition of PMCA in hypertension. Therefore, we measured the levels of tyrosine phosphorylation of PMCA in healthy volunteers as a function of blood pressure[16]. Separation of volunteers into normotensive (diastolic < 85 mmHg) and hypertensive (diastolic > 85 mmHg) has revealed a significantly higher level of PMCA tyrosine phosphorylation in hypertensive individuals. These results suggest that PMCA in platelets of hypertensive individuals is inhibited because of tyrosine phosphorylation leading to elevated [Ca²⁺ ]_i, hyperactive platelets, and enhanced risk of heart attack and stroke. A rat model of hypertension has indicated changes in the expression levels of PMCA1b and PMCA4b in rat platelets as a function of hypertension, but no activity measurements have been reported[10], therefore, the contribution to platelet calcium efflux in this model is unknown. Platelet [Ca²⁺]_i has also been reported to be increased in type 2 diabetes by several groups. Rosado et al[16] have shown that this increase in [Ca²⁺]_i is caused by tyrosine phosphorylation of PMCA, thus confirming our earlier work in hypertension. However, a more recent publication[23] has reached the conclusion that PMCA4b has enhanced expression in diabetes, although calcium transport studies were not undertaken, so that the overall effect on activity and [Ca²⁺ ]_i was not determined. The same group has published similar data with respect to PMCA expression in hypertension, which was increased, but in that study, PMCA activity was measured and found to be decreased[24]. The lack of correlation between protein expression and activity has been ascribed to tyrosine phosphorylation.

CONCLUSION

Ca²⁺ is a major second messenger in platelet activation, and elevated [Ca²⁺]_i leads to hyperactive platelets. Elevated platelet [Ca²⁺]_i has been documented in hypertension and diabetes; both conditions increase the likelihood of heart attack and stroke. Thus proper regulation of platelet calcium metabolism is extremely important. PMCA is a major player in platelet calcium metabolism because it provides the only significant route for calcium extrusion. In keeping with the important role of calcium in platelet function, PMCA is a highly regulated transporter. In human platelets, PMCA is activated by Ca²⁺/calmodulin, by cAMP-dependent phosphorylation and by calpain-dependent removal of the inhibitory peptide. It is inhibited by tyrosine phosphorylation and calpain-dependent proteolysis. In addition, the cellular location of PMCA is regulated by PDZ-domain-dependent interaction with the cytoskeleton during platelet activation. Rapid regulation by phosphorylation results in changes in the rate of platelet activation and secretion, whereas calpain action and interaction with the cytoskeleton appear to regulate later events in platelet function such as clot retraction. In hypertension and diabetes, PMCA expression is upregulated whereas activity is decreased, presumably due to tyrosine phosphorylation. Clearly, a more complete understanding of PMCA function in human platelets could result in identification of new ways to control platelet function in disease states.