Activation of receptor-operated cation channels via P2X1 not P2T purinoceptors in human platelets.

We have investigated the purinoceptor subtypes responsible for calcium signaling in human platelets, which previous studies have shown to involve both Ca influx via receptor-operated cation channels and release of Ca from intracellular stores. Fura-2 measurements of [Ca] in stirred platelet suspensions showed that both ADP (40 μM) and the non-hydrolyzable ATP analogue αβ-meATP (α,β-methyleneadenosine 5′-triphosphate, 10 μM) activated a rapid Ca influx whereas only ADP mobilized Ca from internal stores. In “nystatin” whole-cell patch clamp recordings, ATP, ADP, and the non-hydrolyzable ATP analogues, α,β-meATP and ATPS (adenosine 5′-O-(3-thiotriphosphate), all activated a cation channel permeable to both monovalent and divalent cations with a single-channel conductance of 11 picosiemens in NaCl saline. The current response to ATP (40 μM) was activated within 20 ms and desensitized with a time constant of 47-107 ms in the continued presence of agonist, which are characteristics of P receptors in other tissues. We conclude that human platelets possess a P purinoceptor, which mediates a rapid phase of ADP- or ATP-evoked Ca entry via a cation channel, whereas one or more separate ADP-selective P purinoceptors evoke release of calcium from intracellular stores.

We have investigated the purinoceptor subtypes responsible for calcium signaling in human platelets, which previous studies have shown to involve both Ca 2؉ influx via receptor-operated cation channels and release of Ca 2؉ from intracellular stores. Fura-2 measurements of [Ca 2؉ ] i in stirred platelet suspensions showed that both ADP (40 M) and the non-hydrolyzable ATP analogue ␣␤-meATP (␣,␤-methyleneadenosine 5triphosphate, 10 M) activated a rapid Ca 2؉ influx whereas only ADP mobilized Ca 2؉ from internal stores. In "nystatin" whole-cell patch clamp recordings, ATP, ADP, and the non-hydrolyzable ATP analogues, ␣,␤-meATP and ATP␥S (adenosine 5-O-(3-thiotriphosphate), all activated a cation channel permeable to both monovalent and divalent cations with a single-channel conductance of 11 picosiemens in NaCl saline. The current response to ATP (40 M) was activated within 20 ms and desensitized with a time constant of 47-107 ms in the continued presence of agonist, which are characteristics of P 2X1 receptors in other tissues. We conclude that human platelets possess a P 2X1 purinoceptor, which mediates a rapid phase of ADP-or ATP-evoked Ca 2؉ entry via a cation channel, whereas one or more separate ADP-selective P 2 purinoceptors evoke release of calcium from intracellular stores.
In many cell types, extracellular ATP and ADP interact with a family of P 2 purinoceptors (1). Two subgroups have been recently cloned: ionotrophic (P 2X1 , P 2X2 , P 2X3 , P 2X4 ) and Gprotein-coupled (P 2Y , P 2U ) purinoceptors (2)(3)(4)(5)(6)(7)(8). Data suggest putative P 2D and P 2Z purinoceptors (9, 10) also exist. Human platelets are reported to possess a unique ADP-selective purinoceptor, termed P 2T , which mediates shape change and aggregation (11). Major actions of ADP in platelets include mobilization of intracellular calcium stores (13) and activation of a non-selective cation channel (12). In cell-attached patch clamp recordings (14), ADP evoked single channel activity if included in the pipette but not when added to the bath saline, demonstrating that this channel is activated by a direct receptor-operated or G-protein-linked mechanism rather than via a diffusible second messenger. Whether the actions of ADP in Ca 2ϩ signal generation in human platelets are mediated by one or multiple purinoceptors is, however, uncertain. In the present study we have used both whole-cell patch clamp recordings and fura-2 intracellular calcium measurements to investigate the purinoceptor subtypes involved in human platelet calcium signaling.
Isolation of Platelets-Venous blood was donated by healthy volunteers with their informed consent and platelet-rich plasma (PRP) prepared as described previously (13). Apyrase (20 g/ml) and aspirin (100 M) were added to minimize platelet activation by spontaneously released adenosine nucleotides and thromboxane, respectively. For electrophysiological recordings, 1-ml aliquots of PRP were spun for 1 min at low speed in a Microcenteur microcentrifuge (MSE Scientific Instruments, UK), and the pellet was resuspended in nominally calcium-free standard platelet saline containing 20 g/ml apyrase and 0.1% bovine serum albumin. The addition of bovine serum albumin further reduced spontaneous platelet activation. In the case of fluorescence measurements, platelets were loaded while in PRP with the fluorescent indicator, fura-2, as described previously (13).
Fluorescence Recordings-Fura-2 fluorescence measurements were made from stirred platelet suspensions using a Cairn spectrophotometer system (Cairn Research Ltd., UK). [Ca 2ϩ ] i was calculated from the 340/380 ratio based on calibration in the presence of 50 M digitonin as described by Grynkiewicz et al. (15). Dose-response curves were fitted by a four-parameter logistic function using Sigmaplot (Jandel Scientific), from which pEC 50 values were derived.
Electrophysiology-Whole-cell patch clamp recordings, formed by nystatin permeabilization of cell-attached patches, were made with a List EPC7 amplifier, as described previously (13). Patch pipettes (filled resistances of 5-10 megaohms) were pulled from borosilicate-filamented glass tubing (Clark Electromedical Instruments, UK). Membrane currents were filtered at 3 kHz and sampled at 100 s using a Digidata 1200 interface and pClamp6 software (Axon Instruments, CA). Agonists were applied from a closely apposed pressure injection pipette, and the bath was continuously counterperfused with saline containing 20 g ml Ϫ1 apyrase to minimize purinoceptor desensitization. All experiments were performed at the ambient temperature (23-25°C). Data were expressed as the mean Ϯ S.E. with number of observations (n) in parentheses.

RESULTS AND DISCUSSION
Intracellular Calcium Responses Evoked by ATP, ␣,␤-meATP, and ADP-Fura-2-loaded human platelets in stirred suspension were used to assess Ca 2ϩ influx and mobilization from intracellular calcium stores evoked by adenosine nucleotides and a range of related analogues as well as by other nucleotides. As previously reported (13), in the presence of 1 mM external calcium, 40 M ADP evoked a peak [Ca 2ϩ ] i rise of 313 Ϯ 12 nM (n ϭ 9), which in Ca 2ϩ -free saline was reduced to 151 Ϯ 50 nM (n ϭ 10) (Fig. 1a). Our results now demonstrate for the first time that ATP (40 M) also evokes an elevation in [Ca 2ϩ ] i , which peaked 138 Ϯ 45 nM (n ϭ 8) above basal levels in the presence of external Ca 2ϩ (Fig. 1b). This response was reduced to a gradual increase in cytosolic Ca 2ϩ of 57 Ϯ 37 nM (n ϭ 6) in Ca 2ϩ -free saline (Fig. 1b). Since the adenosine nucleotidase, apyrase, was necessary in these experiments to minimize desensitization of platelet purinoceptors by spontaneous release of endogenous ATP and ADP, the ATP-evoked signal could arise from ADP generated by this enzyme. Hence we investigated the effect of a non-hydrolyzable analogue of ATP. As shown in Fig. 1c, ␣,␤-meATP (10 M) elevated [Ca 2ϩ ] i by 90 Ϯ 9 nM (n ϭ 4) from resting levels in the presence of 1 mM external calcium but had no effect in Ca 2ϩ -free saline. This is to be expected for an agonist considered selective for P 2X1 purinoceptors, which do not activate phospholipase C (2).
Dose-response curves (Fig. 1d) were constructed for ADP (10 nM to 30 M) and ␣,␤-meATP (10 nM to 10 M) in the presence of 1 mM external calcium. In the case of ADP, a calcium response was detectable at 10 nM, was maximal at around 100 M, and had a pEC 50 of Ϫ6.7 Ϯ 0.2. The dose-response curve could also be fitted by a Hill plot with a Hill coefficient of 0.76 (not shown), which could be explained by the presence of two purinoceptors with differing affinities for ADP. In contrast, ␣,␤-meATP increased [Ca 2ϩ ] i with a threshold concentration of 0.3 M, a maximal response at 3 M, and a pEC 50 of Ϫ6.1 Ϯ 0.13. The ␣,␤-meATP response curve was also fitted by a Hill plot with a Hill coefficient of 1.5 (not shown). The pharmacological properties of this response are similar to those for ␣,␤-meATP-activating P 2X1 purinoceptors in smooth muscle (16) and provide further evidence for the presence of this class of purinoceptor in human platelets. Neither the purine nucleotide, AMP (100 M, n ϭ 3), nor the pyrimidine nucleotide, UTP (100 M, n ϭ 3), had an effect on [Ca 2ϩ ] i (data not shown).
ATP Rapidly Activates a Non-selective Cation Current in Whole-cell Recordings-Single human platelet responses to adenosine nucleotides were studied using the "nystatin" whole-cell patch clamp technique. Application of ATP, at a holding potential of Ϫ70 mV, evoked a transient inward current with a delay of Ͻ20 ms and a peak in the range of 25.5-106 pA (n ϭ 6) (Fig. 2a). The decay time of the current was fitted by a single exponential with a time constant in the range of 47-107 ms (n ϭ 6). The single channel current, when resolved from wholecell current records (12) at Ϫ70 mV, was 0.79 Ϯ 0.04 pA (mean Ϯ S.D., n ϭ 6) in standard NaCl platelet saline and 0.77 Ϯ 0.05 pA (mean Ϯ S.D., n ϭ 3) in 110 mM BaCl 2 saline. Permeability to Ca 2ϩ is predicted from this result since Ba 2ϩ has proved a good surrogate for Ca 2ϩ in studies of Ca 2ϩ -permeable channels including the platelet ADP receptor-operated channel (12,14). Following a 5-s application of ATP, reapplication of agonist within 1-2 min resulted in a much reduced inward current (Fig. 2b), and recovery of the response required several minutes washing in normal saline (not shown). The phenomenon of tachyphylaxis following application of agonist is characteristic of P 2X1 purinoceptors (2). As previously reported (12), ADP activates a non-selective cation current in "nystatin" whole-cell recordings from human platelets. This was shown to be via a direct receptor-operated (rather than second messenger-operated) channel in cell-attached recordings (14). In this study, ADP (40 M) activated a peak inward current in the range of 5-60 pA (n ϭ 4) at Ϫ70 mV (not shown). The single channel current was 0.71 Ϯ 0.07 pA (mean Ϯ S.D., n ϭ 3), a value comparable with that determined in earlier studies (12). From the maximal peak currents recorded, we estimate that ADP and ATP are capable of activating 80 -130 channels in a single platelet.
Response to ATP␥S and ␣,␤-meATP in Whole-cell Recordings-To ensure that the observed ATP-evoked currents were not due to breakdown products of the adenosine nucleotide, the effects of non-hydrolyzable analogues were investigated. A large transient inward current was activated by 40 M ATP␥S in standard platelet saline (Fig. 1c). This current had a similar delay time (Ͻ20 ms) to the ATP-evoked current, but the decay time was greater, being fitted by a single exponential with a time constant of 236 -372 ms in standard platelet saline (Fig. 2,  a and c). The prolonged time course of the ATP␥S-induced membrane current is an intrinsic property of the receptor as opposed to the inability of apyrase to catabolize ATP␥S, as all agonists were superfused in apyrase-free saline. Furthermore, activation of P 2X1 purinoceptors by ATP␥S in smooth muscle also evoked a prolonged inward current (17). The unitary current-voltage relationships for both ATP␥S and ␣,␤-meATP showed inward rectification between the measured range of Ϫ34 mV and Ϫ104 mV (Fig. 2d). These relationships were comparable with that for the current activated by ADP in standard platelet saline (Fig. 2d). Therefore, ADP and ATP activated the same cationic current in whole-platelet patch recordings.
From these observations we conclude that at least two different purinoceptor subclasses mediate calcium responses in human platelets. One or more purinoceptors are responsible for release of calcium from intracellular stores. This may include the novel platelet purinoceptor that has been designated P 2T (18). Our results demonstrate for the first time a P 2X purinoceptor activated by ATP and ADP, which is coupled to a nonselective cation channel, with pharmacological and kinetic properties similar to the P 2X1 purinoceptor recently cloned from vas deferens (2). Since damaged endothelial cells will release both ADP and ATP, rapid Ca 2ϩ influx via P 2X1 -coupled cation channels is likely to be one of the earliest events by which platelet activation occurs at sites of vascular injury. Furthermore, platelets secrete ADP and ATP from their dense storage granules (19), which will activate yet more platelets to increase the size of a developing thrombus. Therefore, Ca 2ϩ influx via the P 2X1 purinoceptor may represent a more important pathway in platelet activation than has previously been recognized.