ATP-activated Ca(2+)-permeable channels in rat peritoneal macrophages

Involvement of P2 receptors in hematopoiesis and hematopoietic disorders, and as pharmacological targets

Several reports have shown the presence of P2 receptors in hematopoietic stem cells (HSCs). These receptors are activated by extracellular nucleotides released from different sources. In the hematopoietic niche, the release of purines and pyrimidines in the milieu by lytic and nonlytic mechanisms has been described. The expression of P2 receptors from HSCs until maturity is still intriguing scientists. Several reports have shown the participation of P2 receptors in events associated with modulation of the immune system, but their participation in other physiological processes is under investigation. The presence of P2 receptors in HSCs and their ability to modulate this population have awakened interest in exploring the involvement of P2 receptors in hematopoiesis and their participation in hematopoietic disorders. Among the P2 receptors, the receptor P2X7 is of particular interest, because of its different roles in hematopoietic cells (e.g., infection, inflammation, cell death and survival, leukemias and lymphomas), making the P2X7 receptor a promising pharmacological target. Additionally, the role of P2Y12 receptor in platelet activation has been well-documented and is the main example of the importance of the pharmacological modulation of P2 receptor activity. In this review, we focus on the role of P2 receptors in the hematopoietic system, addressing these receptors as potential pharmacological targets.

P2 purinoceptors and pyrimidinoceptors of catecholamine-producing cells and immunocytes

ATP is a neuronal (co)transmitter. In addition, both ATP and UTP may exit damaged cells and thereby function as extracellular signal molecules. The targets of signalling may be the P2 (for ATP and UTP) and P1 (for the degradation product adenosine) receptors of, for instance, neurons and immunocytes. UTP may also act at separate pyrimidinoceptors. Catecholamine-producing cells (adrenal chromaffin cells and peripheral and central noradrenergic neurons) possess P2X and P2Y purinoceptors. ATP appears to be a fast excitatory neuro-neuronal transmitter of the noradrenergic coeliac and locus coeruleus neurons. This effect is mediated by P2X purinoceptors. P2Y purinoceptor-mediated slow excitatory synaptic potentials have not yet been demonstrated either in the peripheral or central nervous system. On the other hand, after neuronal injury microglial cells (brain immunocytes) are engaged in a process called 'synaptic stripping', i.e. the displacement of synaptic boutons from the neuronal surface. During this process microglial cells are in direct contact with the (co)transmitter ATP. Activation of P2X, P2Z and P2Y purinoceptors results in an elevated intracellular Ca2+ concentration in microglia and macrophages. Various functions of these cells are regulated by intracellular Ca2+ (e.g. cytokine production, phagocytosis) and may therefore be modulated by nucleotides. Since neuronal damage leads to the transformation of microglial cells to macrophages and, at the same time, to the efflux of nucleotides from the damaged cells, the requirements for a modulatory interaction are fulfilled.

Molecular physiology of P2X receptors

P2X receptors are membrane ion channels that open in response to the binding of extracellular ATP. Seven genes in vertebrates encode P2X receptor subunits, which are 40-50% identical in amino acid sequence. Each subunit has two transmembrane domains, separated by an extracellular domain (approximately 280 amino acids). Channels form as multimers of several subunits. Homomeric P2X1, P2X2, P2X3, P2X4, P2X5, and P2X7 channels and heteromeric P2X2/3 and P2X1/5 channels have been most fully characterized following heterologous expression. Some agonists (e.g., alphabeta-methylene ATP) and antagonists [e.g., 2',3'-O-(2,4,6-trinitrophenyl)-ATP] are strongly selective for receptors containing P2X1 and P2X3 subunits. All P2X receptors are permeable to small monovalent cations; some have significant calcium or anion permeability. In many cells, activation of homomeric P2X7 receptors induces a permeability increase to larger organic cations including some fluorescent dyes and also signals to the cytoskeleton; these changes probably involve additional interacting proteins. P2X receptors are abundantly distributed, and functional responses are seen in neurons, glia, epithelia, endothelia, bone, muscle, and hemopoietic tissues. The molecular composition of native receptors is becoming understood, and some cells express more than one type of P2X receptor. On smooth muscles, P2X receptors respond to ATP released from sympathetic motor nerves (e.g., in ejaculation). On sensory nerves, they are involved in the initiation of afferent signals in several viscera (e.g., bladder, intestine) and play a key role in sensing tissue-damaging and inflammatory stimuli. Paracrine roles for ATP signaling through P2X receptors are likely in neurohypophysis, ducted glands, airway epithelia, kidney, bone, and hemopoietic tissues. In the last case, P2X7 receptor activation stimulates cytokine release by engaging intracellular signaling pathways.

Fundamental Ca2+ signaling mechanisms in mouse dendritic cells: CRAC is the major Ca2+ entry pathway

Although Ca(2+)-signaling processes are thought to underlie many dendritic cell (DC) functions, the Ca(2+) entry pathways are unknown. Therefore, we investigated Ca(2+)-signaling in mouse myeloid DC using Ca(2+) imaging and electrophysiological techniques. Neither Ca(2+) currents nor changes in intracellular Ca(2+) were detected following membrane depolarization, ruling out the presence of functional voltage-dependent Ca(2+) channels. ATP, a purinergic receptor ligand, and 1-4 dihydropyridines, previously suggested to activate a plasma membrane Ca(2+) channel in human myeloid DC, both elicited Ca(2+) rises in murine DC. However, in this study these responses were found to be due to mobilization from intracellular stores rather than by Ca(2+) entry. In contrast, Ca(2+) influx was activated by depletion of intracellular Ca(2+) stores with thapsigargin, or inositol trisphosphate. This Ca(2+) influx was enhanced by membrane hyperpolarization, inhibited by SKF 96365, and exhibited a cation permeability similar to the Ca(2+) release-activated Ca(2+) channel (CRAC) found in T lymphocytes. Furthermore, ATP, a putative DC chemotactic and maturation factor, induced a delayed Ca(2+) entry with a voltage dependence similar to CRAC. Moreover, the level of phenotypic DC maturation was correlated with the extracellular Ca(2+) concentration and enhanced by thapsigargin treatment. These results suggest that CRAC is a major pathway for Ca(2+) entry in mouse myeloid DC and support the proposal that CRAC participates in DC maturation and migration.

Procedures to characterize and study P2Z/P2X7 purinoceptor: flow cytometry as a promising practical, reliable tool

The expression of P2Z/P2X7 purinoceptor in different cell types is well established. This receptor is a member of the ionotropic P2X receptor family, which is composed by seven cloned receptor subtypes (P2X1 - P2X7). Interestingly, the P2Z/P2X7 has a unique feature of being linked to a non-selective pore which allows the passage of molecules up to 900 Da depending on the cell type. Early studies of P2Z/P2X7 purinoceptor were exclusively based on classical pharmacological studies but the recent tools of molecular biology have enriched the analysis of the receptor expression. The majority of assays and techniques chosen so far to study the expression of P2Z/P2X7 receptor explore directly or indirectly the effects of the opening of P2Z/P2X7 linked pore. In this review we describe the main techniques used to study the expression and functionality of P2Z/P2X7 receptor. Additionally, the increasing need and importance of a multifunctional analysis of P2Z/P2X7 expression based on flow cytometry technology is discussed, as well as the adoption of a more complete analysis of P2Z/P2X7 expression involving different techniques.

Coexistence of purino- and pyrimidinoceptors on activated rat microglial cells

1. Nucleotide-induced currents in untreated (proliferating) and lipopolysaccharide (LPS; 100 ng ml(-1)) treated (non-proliferating) rat microglial cells were recorded by the whole-cell patch-clamp technique. Most experiments were carried out on non-proliferating microglial cells. ATP (100 nM-1 mM), ADP (10 nM-10 mM) and UTP (1 microM-100 mM), but not uridine (100 microM-10 mM) produced a slow outward current at a holding potential of 0 mV. The effect of UTP (1 mM) did not depend on the presence of extracellular Mg2+ (1 mM). The outward current response to UTP (1 mM) was similar in non-proliferating and proliferating microglia. 2. In non-proliferating microglial cells, the ATP (10 microM)-induced outward current was antagonized by suramin (300 microM) or reactive blue 2 (50 microM), whereas 8-(p-sulphophenyl)-theophylline (8-SPT; 100 microM) was inactive. By contrast, the current induced by UTP (1 mM) was increased by suramin (300 microM) and was not altered by reactive blue 2 (50 microM) or 8-SPT (100 microM). 3. The current response to UTP (1 mM) disappeared when K+ was replaced in the pipette solution by an equimolar concentration of Cs+ (150 mM). However, the effect of UTP (1 mM) did not change when most Cl- was replaced with an equimolar concentration of gluconate (145 mM). The application of 4-aminopyridine (1 mM) or Cs+ (1 mM) to the bath solution failed to alter the UTP (1 mM)-induced current. UTP (1 mM) had almost no effect in a nominally Ca2+-free bath medium, or in the presence of charybdotoxin (0.1 microM); the inclusion of U-73122 (5 microM) or heparin (5 mg ml(-1)) into the pipette solution also blocked the responses to UTP (1 mM). By contrast, the effect of ATP (10 microM) persisted under these conditions. 4. I-V relations were determined by delivering fast voltage ramps before and during the application of UTP (1 mM). In the presence of extracellular Cs+ (1 mM) and 4-aminopyridine (1 mM) the UTP-evoked current crossed the zero current level near -75 mV. Omission of Ca2+ from the Cs+ (1 mM)- and 4-aminopyridine (1 mM)-containing bath medium or replacement of K+ by Cs+ (150 mM) in the pipette solution abolished the UTP current. 5. Replacement of GTP (200 microM) by GDP-beta-S (200 microM) in the pipette solution abolished the current evoked by UTP (1 mM). 6. When the pipette solution contained Cs+ (150 mM) instead of K+ and in addition inositol 1,4,5,-trisphosphate (InsP3; 10 microM), an inward current absolutely dependent on extracellular Ca2+ was activated after the establishment of whole-cell recording conditions. This current had a typical delay, a rather slow time course and did not reverse its amplitude up to 100 mV, as measured by fast voltage ramps. 7. A rise of the internal free Ca2+ concentration from 0.01 to 0.5 microM on excised inside-out membrane patches produced single channel activity with a reversal potential of 0 mV in a symmetrical K+ solution. The reversal potential was shifted to negative values, when the extracellular K+ concentration was decreased from 144 to 32 mM. By contrast, a decrease of the extracellular Cl- concentration from 164 to 38 mM did not change the reversal potential. 8. Purine and pyrimidine nucleotides act at separate receptors in rat microglial cells. Pyrimidinoceptors activate via a G protein the enzyme phospholipase C with the subsequent release of InsP3. The depletion of the intracellular Ca2+ pool appears to initiate a capacitative entry of Ca+ from the extracellular space. This Ca2+ then activates a Ca2+-dependent K+ current.

Calcium-mobilizing purine receptors on the surface of mammalian articular chondrocytes

If we are to fully understand mechanisms of cartilage homeostasis, it is essential that we know the full catalogue of receptors present on the surface of a chondrocyte and the pathways regulated by ligands that bind to these receptors. In this study, we describe chondrocyte responses to adenosine 5'-triphosphate and related molecules. Adenosine 5'-triphosphate stimulated a statistically significant, dose-dependent, transient rise in the concentration of calcium ions in Fura 2-loaded, differentiated, primary chondrocytes. The increase occurred in the absence of extracellular calcium, indicating a mobilization from intracellular stores. The increase in concentration of cytoplasmic calcium ions induced by adenosine 5'-triphosphate was mimicked by uridine 5'-triphosphate but not by 2-methylthioadenosine 5'-triphosphate, cytidine 5'-triphosphate, or adenosine. Heterologous desensitization experiments demonstrated that chondrocytes showed no subsequent response to uridine 5'-triphosphate after initial stimulation with adenosine 5'-triphosphate nor did they respond to adenosine 5'-triphosphate in inverse conditions, thereby indicating competition for the same receptor site. Together, these results are consistent with the presence of a P2U receptor on the cell surface of chondrocytes. Purine-induced calcium mobilization in passaged chondrocytes showed the same pharmacological profile with respect to agonist sensitivity, but responses were of greater magnitude than responses in primary differentiated chondrocytes, suggesting upregulation of the receptor with time in culture. Adenosine 5'-triphosphate and uridine 5'-triphosphate (1-100 microM) did not alter cartilage matrix synthesis as measured by rate of incorporation of [35S]sulfate into glycosaminoglycan by cartilage explants or primary chondrocytes. Matrix degradation, measured by release of glycosaminoglycan from cartilage explants, was also unaltered by adenosine 5'-triphosphate or uridine 5'-triphosphate (1-100 microM). Production of prostaglandin E2 was upregulated by incubation with either adenosine 5'-triphosphate or uridine 5'-triphosphate. These data demonstrate the presence of a functional P2U-like purine receptor on the surface of primary articular chondrocytes and support the hypothesis that altered concentrations of extracellular purines may influence chondrocyte metabolism.

Effect of intracellular calcium on ATP-activated, GTP-dependent calcium channels in rat macrophages

1. In order to study the effect of intracellular free Ca2+ concentration ([Ca2+]i) on the activity of ATP-activated, GTP-dependent Ca2+ channels in rat macrophages, experiments were performed using the inside-out configuration of the patch-clamp technique. 2. Channel activity was observed in the cell-attached mode when 100 microM ATP was added to the pipette solution containing 105 mM Ba2+, but it disappeared rapidly after patch excision. The activity could be restored by the application of 100 microM GTP or GTP gamma S onto the internal surface of the plasma membrane. 3. The properties of the GTP gamma S-evoked channels are identical to those of channels activated by extracellular application of ATP. The channels exhibited four current sublevels with conductances of about 3.5, 7, 10 and 15 pS when 105 mM Ba2+ was the only permeant cation. The extrapolated reversal potentials were similar for all the sublevels and averaged about +40 mV. 4. Elevation of [Ca2+]i within the range 0.01-1 microM resulted in a decrease in mean inward current. The half-maximal value of the mean current was about 0.08 microM. 5. This decreases in mean current resulted from a redistribution of sublevel occupancies: the 1st sublevel tended to be come more abundant with elevation of [Ca2+]i, while the relative weights of the high-conductance 3rd and 4th sublevels decreased. 6. The open-channel current fell with an increase in [Ca2+]i as quickly as the mean current did, indicating that the sublevel redistribution alone is sufficient to produce the revealed decrease in net inward current. 7. It is concluded that [Ca2+]i elevation does not fix the channel in a closed state but rather decreases the ability of the channel to operate in high-conductance states.

A cation non-selective channel induced by extracellular ATP in macrophages and phagocytic cells of the thymic reticulum

Extracellular ATP4- can bind to P2Z purinergic receptors including depolarization and cytoplasmic membrane permeabilization to small molecular weight solutes in macrophages, thymocytes, mast cells, phagocytic cells of the thymic reticulum and other cell types. An ATP(4-)-induced cation current has been described in whole-cell records of some of these cells but it is currently not clear whether these currents and the phenomenon of membrane permeabilization are a consequence of only one type of P2Z-associated channel/pore or two different phenomena triggered by one or more receptors. Here we use the outside-out patch-clamp technique to describe a single channel associated with this cation current in two murine phagocytic cells: intraperitoneal macrophages and phagocytic cells of the thymic reticulum. Multi channel currents could be readily observed in 77% of the outside-out patches of macrophages. Single channels of 7.8 pS could usually be resolved only in tail currents. Reversal potential measurements and ion replacement experiments indicated a lack of cation selectivity, similarly to what has already been described for the ATP(4-)-induced whole-cell inward current. No large-conductance channels that could explain the permeabilization to small molecular weight studies solutes was observed under our experimental conditions. A single channel of approx. 5 pS was also observed in phagocytic cells of the thymic reticulum under similar conditions. We conclude that the channel here described is the main carrier of cation current usually associated with the binding of ATP4- to P2Z receptors in whole-cell and outside-out patch-clamp experiments.

ATP-operated calcium-permeable channels activated via a guanine nucleotide-dependent mechanism in rat macrophages

1. To elucidate the possible involvement of a G protein in ATP-evoked Ca(2+)-permeable channel activity, membrane currents of rat peritoneal macrophages were recorded using inside-out and cell-attached configurations of the patch clamp technique. 2. In inside-out experiments with a pipette solution containing 105 mM Ba2+, application of 100 microM GTP or GTP gamma S to the internal surface of the membrane elicited a rise in channel activity. This effect was observed in 49% of the patches investigated (n = 69). The mean value of NPo (N, number of open channels; Po, channel open probability) was equal to 0.49 +/- 0.27 (mean +/- S.E.M.; n = 16). The delay in the activity development was 21 +/- 8 s (n = 18) with 200 microM ATP added to the pipette solution and about 4 min (n = 5) without agonist in the pipette. Similar results were obtained with 10 mM Ca2+ as the only permeant cation. 3. Properties of GTP gamma S-evoked channels were identical to those of channels activated by extracellular application of ATP. The channels exhibited at least four conductance sublevels, the 4th one being the least frequent. With 105 mM Ba2+ as a permeant cation, sublevel conductances were 3.5, 7, 10 and 15 pS. Corresponding values for 10 mM Ca2+ were about 4, 9, 13 and 17 pS. Extrapolated reversal potential (Er) values were about +40 and +25 mV for Ba2+ and Ca2+, respectively. 4. The activity of channels with similar characteristics could be induced by the extracellular application of fluoride in cell-attached experiments without any agonist in the pipette solution.(ABSTRACT TRUNCATED AT 250 WORDS)

ATP-activated inward current and calcium-permeable channels in rat macrophage plasma membranes

1. To study mechanisms of receptor-operated Ca2+ influx in non-excitable cells, membrane currents of rat peritoneal macrophages were recorded using whole-cell cell-attached and outside-out configurations of the patch clamp technique. Under whole-cell recording conditions, ATP applied in micromolar concentrations elicited an inward current response when the bath solution contained Ba2+, Ca2+ or Na+ as the only permeant cations. 2. Increasing the Mg2+ concentration had an inhibitory effect on the ATP-induced inward current indicating that the active form of ATP responsible for the cation entry is ATP4-. The nucleotide potency order was ATP > ATP gamma S > ADP. UTP was completely ineffective (n = 19). The data obtained are consistent with the ATP receptor being of the P2Z type. 3. The macrophage plasma membrane was impermeable to Tris+ during the ATP-induced current at ATP4- concentrations varying from 0.07 to 500 microM. At higher concentrations, ATP produced a large inward steady-state current, which could be attributed to membrane permeabilization. 4. Activity of single channels was recorded when ATP was applied to the external surface of the patch membrane both in cell-attached and outside-out experiments. A specific property of the channels appeared to be the existence of at least four conductance sublevels. With 105 mM Ba2+ as the permeant cation, the conductance sublevels were 3.5, 7, 10 and 15 pS. With 10 mM Ca2+ the sublevel conductances were equal to 4, 9, 13 and 17 pS. 5. The unitary conductance estimated from the whole-cell current noise analysis (3.5-4.5 pS for 105 mM Ba2+) was significantly lower than that obtained from single channel measurements at the main (3rd) current level, but it was very close to the conductance of the minimum (1st) level. 6. Extrapolated reversal potential values estimated from current-voltage curves for predominant conductance levels were equal to +40 and +26 mV for 105 mM Ba2+ and 10 mM Ca2+, respectively. The permeability ratios fell in the sequence: PCa:PBa:PK = 71.:29:1. Thus, ATP-activated channels in the macrophage membrane are rather selective for divalent vs. monovalent cations, with the predominant permeability being for Ca2+.

Nonselective cationic currents elicited by extracellular ATP in human B-lymphocytes

Adenosine 5'-triphosphate-(ATP)-induced whole-cell currents were studied in human B-lymphocytes, transformed by the Epstein-Barr virus, by means of the tight-seal voltage-clamp technique. During bath application of ATP, the membrane conductance was increased. The change of membrane conductance occurred within milliseconds. The dose response relationship for the ATP(4-)-elicited membrane current (Ip) was fitted by the Hill function with a Hill coefficient of 1 and a KD value of 0.2 mmol/l. Adenosine, as well as the Mg(2+)-bound form of ATP, did not effect the membrane conductance. Ip did not desensitize within 1 min and could be evoked repeatedly up to 100 times in 1 cell in the presence of the G-protein blocker Guanosine 5'-o-(2-thiodiphosphate) (GDP [beta S]). Therefore, it seems that ion channels in form of P2Z-purinoceptors are involved in the observed effects. The permeability (P) sequence for cations carrying Ip was PCa:PK:PCs:PNa:PTRIS = 35:2:1.2:1:0.1. The reversal potential of IP was not changed by substitution of intracellular Cl- for aspartate, indicating that anions are not involved in the purinoceptor-dependent conductance. A single-channel conductance of P2Z-receptor-dependent ion channels of about 3 pS was determined by noise analysis of Ip.

Activation of murine macrophages by hydrogen peroxide

Hydrogen peroxide at concentrations from 0.1 to 20 microM enhances phagocytosis and oxidative burst of murine peritoneal macrophages. The activation of these macrophage functions is paralleled by prolonged hyperpolarization and a transient increase in cytoplasmic free calcium concentration. All the effects are dose- and time-dependent. The results obtained for H2O2 are compared with those for a natural activator, peptide N-formyl-methionyl-leucyl-phenylalanine. The data demonstrate the ability of small doses of hydrogen peroxide to stimulate macrophages through the intracellular mechanisms of ion transduction.

Characterization and possible function of adenosine 5'-triphosphate receptors in activated rat microglia

1. Purinoceptor agonist-induced currents in untreated (proliferating) and lipopolysaccharide (LPS; 100 ng ml-1)-treated (non-proliferating) rat microglial cells in culture were recorded by the whole-cell patch-clamp technique. These cells have two preferred resting membrane potentials, one at -35 mV and another one at -70 mV. 2. Most experiments were carried out in non-proliferating cells. ATP, ATP-gamma-S and alpha,beta-MeATP (1-1000 microM in all cases) evoked an inward current at a holding potential of -70 mV, followed, in some experiments, by an outward current. At -70 mV 2-methylthio ATP (1-1000 microM) evoked an inward current, whereas at -35 mV it produced an outward current only. 3. When K+ was replaced in the pipette solution by an equimolar concentration of Cs+ (150 mM), the main outward component of the ATP-gamma-S (10 microM) induced response disappeared. Instead, an inward current was obtained. Replacement of K+ by Cs+ did not affect the inward current evoked by 2-methylthio ATP (300 microM). 4-Aminopyridine (1-10 mM), however, almost abolished this current and unmasked a smaller outward current. 4. The rank order of agonist potency was 2-methylthio ATP > ATP > alpha,beta-MeATP. Adenosine and UTP were inactive. Suramin (300 microM) and reactive blue 2 (50 microM) antagonized the effect of 2-methylthio ATP (300 microM). 5. I-V relations were determined by delivering fast voltage ramps before and during the application of 2-methylthio ATP (300 microM). In the presence of extra- (1 mM) and intracellular (150 mM) Cs+, the 2-methylthio ATP-evoked current crossed the zero current level near 0 mV. When both Cs+ (1 mm) and 4-aminopyridine (1 mM) were present in the bath medium, the intersection of the 2-methylthio ATP current with the zero current level was near - 75 mV.6. 2-Methylthio ATP (1-1I000 MicroM) induced the same inward current both in proliferating and nonproliferating microglia. However, the depolarizing response to 2-methylthio ATP (300 MicroM) was larger and longer-lasting in the proliferating cells. When the free Ca2+ concentration in the pipettes was increased from the standard 0.01 to 1 MicroM, the amplitude and duration of this depolarization was increased in non-proliferating cells. 4-Aminopyridine (1 mM) enhanced the duration, but not the amplitude of responses.7. ATP and its structural analogues stimulate microglial purinoceptors of the P2Y-type. This leads to the opening of non-selective cationic channels and potassium channels. Depending on the resting membrane potential, depolarization or hyperpolarization prevails. Although the inward current produced by 2-methylthio ATP is of similar amplitude in proliferating and non-proliferating microglia, the resulting depolarization is smaller in the latter cell type because of the presence of voltage-sensitive, outwardly rectifying potassium channels.

Metabotropic ATP receptor in hippocampal and thalamic neurones: pharmacology and modulation of Ca2+ mobilizing mechanisms

Changes in cytoplasmic free Ca2+ concentration, [Ca]i, elicited by ATP, were studied in neurones cultured from rat hippocampus and thalamus. ATP evoked [Ca]i increases in about 30% of all cells tested and suppressed [Ca]i transients in responsive cells. The number of responses to ATP markedly increased after pretreatment of cells with inhibitors of protein kinase C, H-7 or staurosporine. The potentiation was blocked by a phorbol ester and by dioleylglycerol. In pretreated cells both once peak [Ca]i and the number of successive trials were augmented by an [ATP] increase. The former effect can be described by the Michaelis-Menten equation whereas the latter one has a steeper, leftward-shifted dependence. Both concentration dependences are explained with a model, describing Ca2+ release as a threshold phenomena. ATP analogues had the rank of potency: ATP approximately ADP >> AMP > alpha, beta-MeATP. A single ATP application depleted internal Ca2+ stores which could be replenished by brief membrane depolarization with high-K+. ATP- and caffeine-induced [Ca]i transients were independent, indicating two non-overlapping Ca2+ storage sites. Only caffeine effects were potentiated at an elevated [Ca]i level, showing a Ca(2+)-induced Ca2+ release. Inhibitors of the Ca2+ pump in internal stores, ryanodine and sulphydryl reagents suppressed the ATP-induced [Ca]i transients, acting via different mechanisms.