Impaired function of TXA2 and/or calcium in the platelets from SHRSP at hypertensive ages

Aggregation of washed platelets from stroke-prone spontaneously hypertensive rats(SHRSP) was markedly reduced with the development of hypertension in comparison with age-matched normotensive Wistar Kyoto rats(WKY) (Tomita et al. Stroke 15, 70-75, 1984). The mechanism of the hypoaggregability of SHRSP platelets was studied. Ca2+-dependence of thrombin-induced aggregation and MDA formation, and Ionophore A23187-induced aggregation of the platelets from SHRSP at a hypertensive age(16-weeks) was similar to that of the aggregation of platelets from age-matched WKY. Optimum Ca2+ concentration for aggregation and MDA formation was 1-2 mM. There was no difference in aggregation in Ca2+-free medium between the two strains. The enhancement by Ca2+ of both thrombin- and Ionophore A23187- induced aggregation, however, was markedly less in SHRSP than in WKY, whereas their MDA formation was equally enhanced by Ca2+. At a prehypertensive age (4-weeks) the degree of enhancement of aggregation by Ca2+ did not differ in the two strains. The magnitude of phospholipid(PI, PC, PE) degradation, and MDA formation IN SHRSP at early- and late- hypertensive ages(11- and 17-weeks) were either the same as or greater than that of age-matched WKY. A linear correlation line between the amount of MDA formed and the degree of platelet aggregation of SHRSP shifted to the right of WKY. In addition, thrombin-induced thromboxane B2 formation in SHRSP platelets was similar to that in WKY in the concentration range of 0.22 - 0.44 U/ml, and became significantly higher at 0.65 U/ml despite severe hypoaggregability of SHRSP platelets in all the concentrations examined. The overproduction of MDA or thromboxane B2 appears to be a compensatory mechanism. These results suggest that abnormalities of SHRSP platelets at hypertensive ages are due to an impaired function of thromboxane A2 and/or calcium concerned in aggregation and secretion but not due to a defect in cyclo-oxygenase and thromboxane synthetase pathway.     

Mass contents of inositol 1,4,5-trisphosphate and 1,2-diacylglycerol in human platelets stimulated with a thromboxane analogue and thrombin.

Mass contents of inositol 1,4,5-trisphosphate (IP3) and 1,2-diacylglycerol (DG) were measured in U46619-stimulated human platelets. 1 microM of U46619 induced maximum responses in aggregation, 5-hydroxytryptamine (5HT) secretion and increase in intracellular free Ca2+ concentration ([Ca2+]i). Aggregation was almost comparable to that induced by maximal dose (1 U/ml) of thrombin, while 5HT release was almost half. The initial [Ca2+]i peak in response to U46619 was about half of thrombin stimulation. Production of IP3 and DG was, however, less than one tenth of that seen in thrombin stimulation. The profile (time course and concentration-dependency) of IP3 formation did not correlate with that of [Ca2+]i, suggesting that U46619 stimulates IP3-dependent and -independent Ca2+ mobilization. DG production was small but sustained for more than 5 min. These findings support the recent hypothesis that aggregation is regulated by a delayed accumulation of DG. The low level of 5HT secretion could be explained by the low production of second messengers, IP3 and DG.     

Hypoaggregability of washed platelets from stroke-prone spontaneously hypertensive rats (SHRSP)

The aggregation properties of washed SHRSP platelets were investigated in comparison with normotensive WKY platelets at prehypertensive (4 weeks), early hypertensive (11 weeks) and late hypertensive (17 weeks) ages in the absence of plasma factors. The number of platelets in SHRSP was markedly lower with the development of hypertension than that in WKY. The thrombin- and collagen-induced aggregation was markedly reduced in the platelets from 11 and 17 week old SHRSP compared with that of age-matched WKY, whereas the degree of platelet aggregation in 4 week old SHRSP showed a tendency to be even greater than that in WKY. The changes in blood pressure and platelet aggregability were correlated inversely. ADP did not induce aggregation in the same system used for thrombin and collagen stimulation but in another system it aggregated washed rat platelets. Aggregation responses to ADP and ionophore A23187 were also significantly lower in 14 week old SHRSP platelets than age-matched WKY platelets. Together with other evidence, these results suggest that defective Ca2+ function, rather than the presence of exhausted platelets, is responsible for hypoaggregability in SHRSP platelets.     

Loss of thrombin-induced Ca2+ mobilization in a subpopulation of platelets during storage.

Thrombin-induced changes in cytosolic free Ca2+ ([Ca2+]i) were studied in human platelets that had been stored for up to 6 days. Changes in [Ca2+]i were measured with Indo-1-loaded platelets and quantitated with two different methods: (i) measurement of the changes in total fluorescence; (ii) measurement of the [Ca2+]i changes in individual platelets in a flow cytometer, allowing the detection of non-responding platelets. The maximal concentration of [Ca2+]i after stimulation with 0.5 U of thrombin/ml decreased from 544 +/- 58 nM (mean +/- SEM, n = 6) on day 0, to 276 +/- 9 nM on day 3 and to 203 +/- 23 nM on day 6. The percentage of platelets responding to 0.5 U of thrombin/ml declined from 90 +/- 2% on day 0 to 72 +/- 4% on day 3, and to 47 +/- 8% on day 6. Nevertheless, also the responding platelets showed a decreased rise in [Ca2+]i. The study shows that during platelet storage a decrease in the rise in [Ca2+]i upon thrombin stimulation occurs. This decrease is partly due to the formation of a subpopulation of platelets that is completely unresponsive and partly due to a decreased responsiveness in the remainder of the platelets; it is not due to a gradual decline in [Ca2+]i rise in all platelets. This phenomenon provides new insight in the functional defect of stored platelets.     

Secreted dense granule adenine nucleotides promote calcium influx and the maintenance of elevated cytosolic calcium levels in stimulated human platelets

Evidence that secreted dense granule adenine nucleotides mediate part of the agonist-induced cytosolic calcium ([Ca2+]i) responses in human platelets was obtained from comparisons of fura-2-loaded platelets from normal subjects and from patients with a form of platelet storage pool deficiency (SPD) in which the secretory dense granules and their contents are virtually absent. SPD platelets had normal initial [Ca2+]i increases induced by thrombin and the endoperoxide analog U46619, but a significantly enhanced decay of elevated [Ca2+]i levels following the initial increases. With thrombin, this enhanced [Ca2+]i decay was associated with decreased Ca2+ influx, as measured by Mn2+ quench of fura-2 fluorescence. Addition of micromolar concentrations of ADP, alone or together with ATP, after stimulation reversed the enhanced [Ca2+]i decay and increased Mn2+ quench in SPD platelets, but had no effect on these responses in normal platelets, while addition of 100-fold higher concentrations of ATP or apyrase before stimulation increased [Ca2+]i decay and decreased Mn2+ quench in normal platelets, but had little effect in SPD platelets. ATP and alpha,beta-methylene ATP, a specific agonist for P2X1 receptors, at micromolar concentrations also increased Mn2+ quench, but to lesser extents than did ADP, in SPD platelets isolated and loaded with fura-2 in the presence of apyrase. Similar effects of ADP and excess ATP were seen in U46619-stimulated platelets, but decreased Ca2+ influx could not be measured directly in SPD platelets, presumably due to the very transient influx response seen with U46619. These results suggest that secreted dense granule ADP and ATP contribute to the maintenance of elevated [Ca2+]i levels, but not to the initial [Ca2+]i increases, in stimulated human platelets, most likely via a nucleotide-specific component of Ca2+ influx which may be mediated by interactions with both P2X1 and P2Y1 purinoceptors.     

Reduced functions of intracellular Ca in aggregation, secretion and protein phosphorylation of permeabilized platelets from stroke-prone spontaneously hypertensive rats

Aggregation, secretion and 47kDa protein (P47) phosphorylation by various agonists such as thrombin, ADP and ionophore A23187 were markedly reduced in platelets from stroke-prone spontaneously hypertensive rats (SHRSP) compared with those of age-matched Wistar Kyoto rat (WKY) platelets, suggesting defective functions of intracellular Ca²⁺ in SHRSP platelets (Tomita et al. Hypertension 1989: 14: 304–315). To clarify the mechanism of the platelet hypofunctions, saponin permeabilized platelets were prepared to compare the responses of platelets from both rats in varying concentrations of extracellular Ca²⁺. The leakage of lactate dehydrogenase from saponin (15 μg/ml)-treated platelets was approx. 5 % of total activity; the degree of the leakage in both platelets did not differ. In saponin-treated platelets, extracellular Ca²⁺ alone did not induce either aggregation or secretion in both strains. However, in the presence of 1-oleoyl-2-acetylglycerol (10 μg/ml), Ca²⁺ dose dependently stimulated both aggregation and secretion. Under this condition, Ca²⁺ sensitivity of aggregation, secretion and P47 phosphorylation in SHRSP platelets were significantly reduced compared with those in WKY platelets. These results strongly suggest that intracellular Ca²⁺ functions are impaired in SHRSP platelets.     

Temperature-dependence of LDL binding and activation of human platelets.

The effect of low density lipoprotein (LDL) on intracellular free calcium ion concentration ([Ca2+]i), taken as an index of the degree of platelet activation, was investigated in normal volunteers. At 37 degrees C LDL, in a dose of 20 micrograms of protein/ml, increased [Ca2+]i in all subjects tested (basal 57 +/- 11 to 113 +/- 19 nM). In contrast, when measurements were performed at 20 degrees C, no effect on [Ca2+]i was seen following LDL. Thrombin (0.2 U/ml) increased [Ca2+]i to 455 +/- 98 nM. When platelets had been exposed to LDL before thrombin stimulation, this increase was less pronounced (to 301 +/- 43 nM). Our finding of a temperature dependence of LDL induced increase in platelet [Ca2+]i supports the concept of a platelet-LDL receptor mediated mechanism. Furthermore, the lower thrombin response following LDL exposure suggests a LDL-thrombin interaction, possibly at the thrombin receptor level and/or calcium recruitment from the same stores.     

Cell shape change and cytosolic Ca2+ in human umbilical-vein endothelial cells stimulated with thrombin.

We quantified thrombin-induced endothelial cells shape change and investigated the role of Ca2+ in such shape change. We used the fluorescent Ca2+ indicator, fura2, to measure both shape change as cell size and intracellular free Ca2+ ([Ca2+]i), in cultured human umbilical-vein endothelial cells (HUVEC). Thrombin induced concentration-dependent decreases in cell size (percentage of cell size at 6 min after stimulation with 0.01 U/ml, 0.1 U/ml, or 1 U/ml thrombin) was 90.1 +/- 1.5%, 78.1 +/- 2.4%, and 40.9 +/- 2.4%, respectively. Thrombin also increased [Ca2+]i in a concentration-dependent manner. Both depletion of extracellular Ca2+, and also the addition of W5, a calmodulin antagonist, inhibited thrombin-induced size reduction. These results indicate an association between shape change and [Ca2+]i mobilization in human endothelial cells stimulated by thrombin.     

Mechanism of the cytoprotective effect of prostaglandin I2 and its analogue in human platelets

The responsiveness of washed human platelets to thrombin was well preserved up to 72 and 96 hours, when they were stored at 4 degrees C with prostaglandin I2 (PGI2) and its stable analogue OP-41483, respectively, while the washed platelets stored with sodium citrate completely lost their responsiveness to thrombin at 48 hours. The effect of PGI2 or the analogue on the resting level of intraplatelet Ca2+ concentration ([Ca2+]i) was investigated, utilizing fluorescent Ca2+ indicator quin 2. PGI2 lowered the resting [Ca2+]i in a dose related manner in the presence or absence of the extracellular Ca2+. The similar results were obtained when OP-41483 was added to the quin 2 loaded platelets. From these results, it was concluded that the lowering of the resting [Ca2+]i may be the prerequisite for the cytoprotective effect of PGI2 or OP-41483.     

On the significance of different aequorin loading techniques on intracellular aequorin discharge, baseline calcium, platelet aggregation and aequorin-indicated Ca(2+)-transients.

The study compares the decay of intracellular luminescence activity (Lmax), the levels of basal [Ca2+]i in resting platelets, and agonist-induced peak [Ca2+]i-signals in platelets loaded with aequorin using the EGTA-, DMSO- and hypoosmotic shock treatment (HOST)-techniques. The highest load of intracellular aequorin with almost unchanged luminescence activity during 4 h was achieved with HOST. Lmax decreased linearly in EGTA- and HOST-platelets, but the decay rate and the levels of basal [Ca2+]i were significantly lower in HOST-platelets. Platelet aggregation and aequorin-indicated [Ca2+]i-rise induced by thrombin and collagen were similar in EGTA- and HOST-platelets. In HOST-platelets, ADP-induced platelet aggregation was always accompanied by aequorin-signals, while at a similar time point, aequorin-signals were absent in 3 of 5 cases in EGTA-platelets. The initial aequorin loading was highest in DMSO-platelets, but Lmax described an exponential decay, which was most pronounced when DMSO-platelets were maintained in Ca(2+)-free buffer (R2 = 0.86). Agonist-induced platelet aggregation was significantly reduced in DMSO-platelets: thrombin-stimulation was accompanied by a significantly lower and delayed [Ca2+]i-rise and no aequorin-signal was obtained in response to ADP in 3 of 5 cases. The study shows that in addition of being a rapid loading-technique, the criteria of high intracellular aequorin load with low luminescence consumption, low basal [Ca2+]i and completely preserved platelet functions are most convincingly met by the HOST-method.     

Evidence that the rat is not an appropriate model to study the role of prostaglandins in normal or abnormal platelet aggregation

Abnormal platelet aggregation seen in experimentally induced diabetic, hypercholesterolemic and spontaneously hypertensive rats (SHR) has been linked with increased prostaglandin synthesis. The present study was conducted to examine the role of prostaglandins in rat platelet activation using normal Wistar Kyoto (WKY) and SHR rats. Up to 30 microM ADP did not induce secondary phase of platelet aggregation in rat PRP and up to 30 microM epinephrine did not produce any response in rat PRP. In other experiments ADP (1.0 microM) and epinephrine (2.0 microM) induced typical biphasic aggregation responses in human PRP. Up to 20 microM U46619, a stable analog of prostaglandin H2, did not induce platelet aggregation in rat PRP or washed rat platelets. In contrast 2.0 microM U46619 caused maximal aggregation in human PRP and washed human platelets. Arachidonic acid (1.5-2.0 mM) induced aggregation in washed rat platelets. However, this was associated with excessive (67% and 94%) loss of cytoplasmic LDH. The low concentrations of thrombin (0.04 and 0.05 U/ml), induced two to three-fold increase in aggregation response in SHR platelets as compared to WKY platelets. Higher concentrations of thrombin (0.1 and 0.3 U/ml) induced similar aggregation responses in SHR and WKY platelets. Thrombin (0.04-0.3 U/ml) induced serotonin secretion in a concentration dependent manner. The extent of secretion was the same in SHR and WKY platelets at all concentrations. Thrombin-induced synthesis of thromboxane A2 (TXA2) in WKY and SHR platelets was quantified using a radioimmunoassay for TXB2. Thrombin (0.04-0.3 U/ml) produced TXB2 in WKY and SHR platelets in a concentration dependent manner. The SHR platelets produced significantly larger amounts of TXB2 as compared to WKY platelets. In other experiments aspirin (500 microM) inhibited thrombin (0.05 U/ml) induced TXB2 synthesis by 75% in both WKY and SHR platelets but failed to inhibit aggregation or secretion in either WKY or SHR platelets. Based on these data it is suggested that: (a) rat platelets inspite of their ability to synthesize TXA2 do not require TXA2 for aggregation; and (b) the rat may not be an appropriate model to study the role of prostaglandins in normal or abnormal platelet aggregation.     

Ouabain enhances basal and stimulus-induced cytoplasmic calcium concentrations in platelets

Incubation of platelet-rich plasma (PRP) with ouabain, an inhibitor of sodium/potassium ATPase (Na+/K+ ATPase), induced a significant rise in basal platelet intracellular calcium concentration [( Ca2+]i) when measured using fura 2. Ouabain induced an enhanced aggregation response to low doses of collagen in both PRP and washed platelets loaded with aequorin. In aequorin loaded platelets this enhanced aggregation response was associated with an enhanced rise in [Ca2+]i such that the relationship between [Ca2+]i and aggregation was unchanged. As inhibition of plasma membrane Na+/K+ ATPase would lead to a raised intracellular sodium ion concentration [( Na+]i) the results suggest that in the platelet, [Na+]i can modulate [Ca2+]i and hence influence the response of platelets to stimuli such as collagen.     

Glutamate-induced calcium transient triggers delayed calcium overload and neurotoxicity in rat hippocampal neurons.

Glutamate-induced changes in intracellular free Ca2+ concentration ([Ca2+]i) were recorded in single rat hippocampal neurons grown in primary culture by employing the Ca2+ indicator indo-1 and a dual-emission microfluorimeter. The [Ca2+]i was monitored in neurons exposed to 100 microM glutamate for 5 min and for an ensuing 3 hr period. Ninety-two percent (n = 64) of these neurons buffered the glutamate-induced Ca2+ load back to basal levels after removal of the agonist; thus, the majority of cells had not lost the ability to regulate [Ca2+]i at this time. However, following a variable delay, in 44% (n = 26) of the neurons that buffered glutamate-induced Ca2+ loads to basal levels, [Ca2+]i rose again to a sustained plateau and failed to recover. The changes in [Ca2+]i that occur during glutamate-induced delayed neuronal death can be divided into three phases: (1) a triggering phase during which the neuron is exposed to glutamate and the [Ca2+]i increases to micromolar levels, followed by (2) a latent phase during which the [Ca2+]i recovers to a basal level, and (3) a final phase that begins with a gradual rise in the [Ca2+]i that reaches a sustained plateau from which the neuron does not recover. This delayed Ca2+ overload phase correlated significantly with cell death. The same sequence of events was also observed in recordings from neuronal processes. The delayed Ca2+ increase and subsequent death were dependent upon the presence of extracellular Ca2+ during glutamate exposure. Calcium influx during the triggering phase resulted from the activation of both NMDA and non-NMDA receptors as indicated by studies using receptor antagonists and ion substitution. Treatment with TTX (1 microM) or removal of extracellular Ca2+ for a 30 min window following agonist exposure failed to prevent the delayed Ca2+ overload. The delayed [Ca2+]i increase could be reversed by removing extracellular Ca2+, indicating that it resulted from Ca2+ influx. The three phases defined by changes in the [Ca2+]i during glutamate-induced neuronal toxicity suggest three distinct targets to which neuroprotective agents may be directed.     

Biogenesis of G-protein mediated calcium signaling in human megakaryocytes.

To understand how platelet signal transduction pathways develop during megakaryocytopoiesis, we isolated human stem cells from umbilical cord blood and cultured the cells in the presence of thrombopoietin (TPO). Based on the early expression of CD61 and late expression of CD42b, immature (CD61+/CD42b(low)) and mature (CD61+/ CD42b(high)) megakaryocytes were immunomagnetically purified and, together with stem cells (CD34+), characterized for Galpha-protein expression and agonist-induced [Ca2+]i increases. Megakaryocytopoiesis was accompanied by down-regulation of the 43 kDa and 46 kDa variants of G16alpha, constant expression of Gsalpha, and up-regulation of Gqalpha and Gialpha1/2. The increase in Gqalpha and Gialpha1/2 expression was accompanied by an increase in Ca2+ signaling triggered by thrombin and other agonists known to signal to Ca2+ via these G-proteins in platelets. The prostacyclin analog iloprost and TPO also induced [Ca2+]i increases, and the iloprost-induced Ca2+ response disappeared during maturation. These data reveal sharp changes in Ca2+ regulation during megakaryocytopoiesis.     

Effects of storage and 37 degrees C-incubation with fresh plasma on platelet calcium distribution, mobilization, influx and aggregability

Aggregability of platelet concentrate decreased progressively during storage at room temperature and recovered when these stored platelets were incubated with fresh plasma at 37 degrees C. During storage, the resting level of cytoplasmic free Ca2+ ([Ca2+]i) which was measured by fluorescence Ca2+ indicator, quin 2, increased. On the other hand, the amount of Ca2+ mobilization and influx after stimulation with ADP or thrombin decreased. After 37 degrees C-incubation of these stored platelets with fresh plasma for 2-3 hours, resting level of [Ca2+]i decreased, and the amount of Ca2+ mobilization and influx increased (recovered). The resting level of [Ca2+]i indicated reverse correlation to aggregability and to the amount of Ca2+ mobilization and influx. These results suggest that the platelet Ca2+ influx in resting or activation state change during storage and by 37 degrees C with fresh plasma. These changes cause the reduction and the restoration of platelet aggregability.     

Selective enhancement by basic fibroblast growth factor of NMDA receptor-mediated increase of intracellular Ca2+ concentration in hippocampal neurons.

The short-term effect of bFGF on intracellular Ca2+ concentration ([Ca2+]i) of hippocampal neurons was investigated using dissociated cell cultures. Changes in [Ca2+]i were measured by microfluorometrically monitoring the fluorescence intensities from individual neurons loaded with fura-2. Perfusion of bFGF (20 ng/ml) alone did not affect the basal level of [Ca2+]i in hippocampal neurons, but clearly enhanced the [Ca2+]i increase induced by NMDA. Quisqualate or KCl-induced [Ca2+]i increase was not influenced by bFGF. These results suggest that bFGF selectively enhances the NMDA receptor-mediated response in hippocampal neurons.     

Effect of the removal of extracellular Ca2+ on the response of cytosolic concentrations of Ca2+ to ouabain in carotid body glomus cells of adult rabbits.

Effect of the removal of extracellular Ca2+ on the response of cytosolic concentrations of Ca2+ ([Ca2+]i) to ouabain, an Na+/K+ exchanger antagonist, was examined in clusters of cultured carotid body glomus cells of adult rabbits using fura-2AM and microfluorometry. Application of ouabain (10 mM) induced a sustained increase in [Ca2+]i (mean+/-S.E.M.; 38+/-5% increase, n=16) in 55% of tested cells (n=29). The ouabain-induced [Ca2+]i increase was abolished by the removal of extracellular Na+. D600 (50 microM), an L-type voltage-gated Ca2+ channel antagonist, inhibited the [Ca2+]i increase by 57+/-7% (n=4). Removal of extracellular Ca2+ eliminated the [Ca2+]i increase, but subsequent washing out of ouabain in Ca2+-free solution produced a rise in [Ca2+]i (62+/-8% increase, n=6, P<0.05), referred to as a [Ca2+]i rise after Ca2+-free/ouabain. The magnitude of the [Ca2+]i rise was larger than that of ouabain-induced [Ca2+]i increase. D600 (5 microM) inhibited the [Ca2+]i rise after Ca2+-free/ouabain by 83+/-10% (n=4). These results suggest that ouabain-induced [Ca2+]i increase was due to Ca2+ entry involving L-type Ca2+ channels which could be activated by cytosolic Na+ accumulation. Ca2+ removal might modify the [Ca2+]i response, resulting in the occurrence of a rise in [Ca2+]i after Ca2+-free/ouabain which mostly involved L-type Ca2+ channels.     

A defect in collagen receptor-Ca2+ signaling system in platelets from cattle with Chediak-Higashi syndrome

Decreased platelet aggregation to collagen is a cause for bleeding diathesis of Chediak-Higashi syndrome (CHS). We investigated whether the collagen receptor-Ca2+ signaling system was impaired in platelets from cattle affected with CHS. A collagen-induced increase in cytosolic Ca2+ ([Ca2+]i) was depressed in CHS platelets, which was accompanied by a decrease in the production of inositol 1,4,5-trisphosphate. When the influences of endogenous arachidonic acid metabolites and ADP were excluded, convulxin or collagen-related peptide, which are specific agonists for the collagen receptor GPVI, increased [Ca2+]i in both normal and CHS platelets. In contrast, rhodocytin, which was thought to activate another collagen receptor GPIa/IIa, increased [Ca2+]i in CHS platelets to a lesser extent than in normal ones. Cytochalasin D, an actin polymerization inhibitor, depressed the response to collagen or rhodocytin but not the response to convulxin. Adhesion of CHS platelets to acid soluble type I collagen, which was mediated by GPIa/IIa, was similar to that of normal platelets. These results suggest that a defect in the rhodocytin-sensitive pathway is responsible for decreasing the response to collagen in CHS platelets. It remains to be determined which receptor is associated with the mechanism.     

Involvement of mitochondrial Na+-Ca2+ exchange in intracellular Ca2+ increase induced by ATP in PC12 cells

The involvement of mitochondrial Na+-Ca2+ exchange in Ca2+ responses to ATP was examined in rat pheochromocytoma (PC) 12 cells. Intracellular Ca2+ ([Ca2+]i) and Na+ concentrations ([Na+]i) were measured using fura-2 and SBFI, respectively. ATP caused concentration-dependent increases in [Ca2+]i and [Na+]i. High concentrations of ATP elicited a Ca2+ transient followed by a slow recovery of [Ca2+]i (a sustained phase) in 77% of PC12 cells. The sustained phase of Ca2+ response appeared only when the peak Ca2+ transient exceeded 500 nM. FCCP, a protonophore, greatly enhanced Ca2+ responses to ATP only in cells with the sustained phase but not without this phase. The sustained phase was decreased by clonazepam and CGP37157, mitochondrial Na+-Ca2+ exchange inhibitors, and extracellular Na+ removal but not by cyclosporin A, an inhibitor of permeability transition pores. The reintroduction of Na+ 3.5 min after ATP stimulation in the absence of Na+ caused Na+ concentration-dependent increases in [Ca2+]i and [Na+]i. The increase in [Na+]i was correlated with that in [Ca2+]i. FCCP caused a great increase in [Ca2+]i 4.5 min after ATP stimulation in the absence of extracellular Na+ but not in its presence, indicating that mitochondria retain Ca2+ in the absence of Na+. These results suggest that ATP causes a large increase in [Ca2+]i which was sequestered in mitochondria and that the sustained phase of Ca2+ response to ATP are mainly due to the release of mitochondrial Ca2+ through Na+-Ca2+ exchangers in PC12 cells.