Introduction

Abstract Introduction References

The purine nucleotide adenosine 5'-triphosphate (ATP) is found in every cell of the human body. In 1929, Drury and Szent-Gyorgyi (1) first showed that extracellular ATP exerts pronounced effects on the cardiovascular system, independent of its pivotal role in intracellular metabolism and cellular energetics. Since then, numerous studies have indicated that extracellular ATP acts as a physiologic regulator in various organs and tissues, acting through specific cell-surface receptors (2). These ATP-specific receptors, termed P2-purinoceptors, are structurally and functionally distinct from the P1-purinoceptors that mediate the actions of extracellular adenosine, a metabolite of ATP (2, 3). Two subfamilies of P2-purinoceptors have been defined: P2X and P2Y, each consisting of at least seven members (X_1-7 and Y_1-7). P2X-purinoceptors are ATP-gated ionic channels, and P2Y-purinoceptors are G protein-linked (5).

Asthma is a complex disease in which mast cells play a central role in both the initiation and maintenance of the inflammatory reaction (6, 7). Previous studies with rat peritoneal mast cells have demonstrated direct Ca²⁺- dependent histamine release induced by ATP (8), as well as potentiation of histamine release induced by either antigen or the ionophore A23187 (12). Subsequently, it was shown that the effects of ATP were attributable to a minor, fully ionized tetrabasic acid component of ATP termed ATP⁴, whose cytotoxic, permeabilizing effects are reversed by complexation with extracellular Ca²⁺ and Mg²⁺ (13). It is also now known that the effects of ATP⁴ are mediated through activation of the P2X₇- (previously termed P2Z-) purinoceptor expressed on the rat mast cell surface (14).

ATP is found at a concentration of 5-10 mM in every cell, with the exception of platelets, in which its concentration is far higher. It is released into the extracellular space from activated platelets, exercising muscle, ischemic cells, inflammatory cells, and necrotic/apoptotic cells, and also from nerve terminals, as a cotransmitter (17). Recently, inhalation of aerosolized ATP was shown to trigger bronchoconstriction in healthy and asthmatic human subjects; in the latter, ATP was 50-fold more potent than methacholine and 87-fold more potent than histamine in producing a 15% decrease in FEV₁ (26). Extracellular ATP could therefore serve as an important modulator of pulmonary mast cell function. Thus, the present study was aimed at determining the effects of ATP on human lung mast cells (HLMC) assessed through histamine release as a cell-specific marker. This study tested the hypothesis that ATP modulates anti-immunoglobulin (Ig)E- and ionophore A23187-induced histamine release in HLMC by activating P2-purinoceptors as distinct from P1(adenosine)-purinoceptors. The study data support this hypothesis and strongly suggest that the functional P2-purinoceptors mediating the action of ATP include the P2Y₁- and/or P2Y₂-purinoceptor subtypes.

	Materials and Methods

Materials

The following were purchased: Porcine elastase Type I, chymopapain, ATP, ,methylene-ATP (,mATP), ,methylene ATP (,mATP), 2-methylthio-ATP (2mSATP), and adenosine (Sigma Chemical Co., St. Louis, MO); collagenase (Worthington, Freehold, NJ); deoxyribonuclease (DNase), pronase (Calbiochem, San Diego, CA), and gelatin (Difco Laboratories, Detroit, MI). Monoclonal antihuman IgE antibody was generously provided by Dr. Robert Hamilton of Johns Hopkins University, Baltimore, MD. HMC-1 cells (27) were supplied by Dr. Joseph H. Butterfield of the Mayo Clinic, Rochester, MN.

Buffers

Lung fragments were washed with Tyrode's buffer containing (g/liters) NaCl, 8.0; KCl, 0.2; NaH₂PO₄, 0.05; and glucose, 1.0. The buffer was titrated to pH 7.2 by the addition of NaHCO₃. Mast cell isolation and elutriation were performed in Tyrode's buffer with (g/liters) gelatin (1.0), magnesium (0.25; 1 mM), and DNase (0.01) (TGMD). Histamine release experiments were done with Pipes- albumin (0.003%) buffer containing (g/liters) glucose (1.0), CaCl₂ · 2 H₂O, 0.14 (1 mM); and MgCl₂ · 6 H₂O, 0.2 (1 mM) (PAGCM).

HLMC

Mast cells were dispersed from human lung according to methods previously reported (28, 29). Briefly, lung specimens obtained at thoracotomy for bronchogenic carcinoma were finely minced and extensively washed in divalent cation-free Tyrode's buffer. Fragments were briefly incubated in a mixture of pronase (2 mg/ml) and chymopapain (0.5 mg/ml). Freed cells were harvested through Nytex nylon cloth (150 µ pore size). Residual fragments were further exposed to a mixture of collagenase (1 mg/ml) and elastase (10 U/ml). All incubations and washes were performed at 37°C; recovered cells were immediately washed three times in large volumes of TGMD. Mast cell purities in these human lung cell suspensions ranged from 1-8% as determined by alcian blue staining (30). HLMC were further purified by countercurrent elutriation, using previously reported methods (29). Mast cells were purified (80 to > 98%) by flotation of enriched elutriation fractions through a discontinuous Percoll gradient (31). Further mast cell purification was accomplished by immunomagnetic negative selection against CD2, CD3, CD4, CD8, CD14, CD16, CD21, and human leukocyte antigen-DR to ensure against contamination by T cells, B cells, natural killer cells, monocytes, and dendritic cells prior to mast cell stimulation, according to previously described methods (32, 33).

Histamine Release Assay

Mast cells (10-50 × 10³/tube) were preincubated for 15 min, either in buffer alone or in buffer solutions, each containing one of the test compounds, and were then challenged with buffer, anti-IgE, or calcium ionophore A23187 at 37°C in PAGCM. The concentrations of anti-IgE or ionophore A23187 that were used produce 30 to 70% of maximal histamine release. Twenty minutes after activation, mast cells were rapidly pelleted and the supernatants removed for histamine analysis. Histamine release was expressed as the net quantity of histamine released divided by the total histamine content × 100%. The total cellular histamine content was determined after cell lysis with 2% perchloric acid. Spontaneous histamine release was always < 2% of cellular histamine, and generally < 1%. Histamine measurements were made according to the automated spectrofluorometric method of Technicon (Tarrytown, NY). Variations between replicates were consistently < 5%. All assays were run in duplicate.

Functional Ectonucleotidase Activity

To determine whether ATP was metabolized by purified HLMC, 0.3 to 1.0 × 10⁵ HLMC in 250 µl PAGCM (n = 3) were preincubated with 10⁴ M ATP for 15 min and subsequently incubated with or without anti-IgE (3 µg/ml) for an additional 20 min. Control preparations consisted of HLMC without ATP, as well as a solution of 10⁴ M ATP in PAGCM. The supernatants were separated from the cells by centrifugation at 14,000 × g for 5 min, and were kept at 20°C until analyzed with high-pressure liquid chromatography (HPLC). The method of Stocci and colleagues (34) was used for the detection of purine compounds. The HPLC system consisted of a Waters 600E controller (Waters, Milford, MA), Waters Novapak 4 µm 3.9 × 150 mm C₁₈ column, and 990 photodiode array detector. The solvent consisted of 0.1 mM KH₂PO₄; 8 mM tetrabutylammonium hydrogen sulfate (TAHS; pH 6.0) (buffer A); and 0.1 mM KH₂PO₄, 8 mM TAHS (pH 6.0), with 30% (vol/vol) methanol (buffer B). The flow rate was 1 ml/min, with the following gradient program: 100% buffer A to 2.5 min, and a linear gradient to 20% buffer B at 5 min, 40% buffer B at 10 min, 100% buffer B at 13 min, and 100% buffer B to 30 min. An aliquot of 100 µl of supernatant (neat) was injected, and the separation was monitored at 254 nm over the 30-min run time. The ATP peak areas were calculated and compared among the conditions.

RNA Extraction and Polymerase Chain Reaction

Total cellular RNA (tcRNA) was isolated from HLMC with a purity  90% through a modified phenol-chloroform extraction technique adapted from Chomczynski and Sacchi (35). For positive controls, whole blood was processed by Ficoll-Hypaque gradient centrifugation to obtain peripheral blood mononuclear cells (32, 33), which were similarly treated for tcRNA. Purified mast cell tcRNA was treated with 10 U heparinase-I (Sigma) at room temperature for 2 h to neutralize the inhibitory effects of mast cell heparin on reverse transcription-polymerase chain reaction (RT-PCR) (36). Complementary DNA (cDNA) was synthesized from 1 µg tcRNA using oligo (deoxythymidine) primers and Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc., Grand Island, NY) at 37°C for 1 h in the presence of 20 U of ribonuclease inhibitor and with 10 nM of each deoxynucleotide triphosphate (Promega Corporation, Madison, WI). The following oligonucleotides (synthesized by OPERON Technologies, Alameda, CA) were used: For P2Y₁, sense: 5'-CCGCCGTCTCCTCGTCGTTCAAAT-3'; antisense: 5'-TTGAGGGGGTACACCACACCGCTGTA-3'. For P2Y₂, sense: 5'-GCCCTTTGCCGTCATCCTTGTC-3'; antisense: 5'-TCCAGCGAGCGGAAGGAGTAGTAG-3'. For P2Y₇, sense: 5'-TATCATCCTGCTGTCAGTGGCGCTG-3'; antisense: 5'-TAGCTTCTGGGACACAAAGGGGCG-3'. For P2X₇, sense: 5'-TGAAGTCTCTGCCTGGTGCCCCAT-3'; antisense: 5'-CGGAAAATGG-GACACTGTGGATTCTG-3'. For glyceraldehyde-3-phosphate dehydrogenase (GAPDH), sense: 5'-AGAAGGTGGTGAAGCAGGCGTCG-3'; antisense: 5'-CCTTGGAGGCCATGTGGGCC-3'.

PCR was done with a programmable thermal cycler (GeneAmp 9600; Perkin Elmer, Foster City, CA), using 1 U Taq DNA polymerase (Life Technologies) for 30 cycles (30 s at 94°C, 30 s at 60°C, 60 s at 72°C), followed by an additional product extension step (72°C for 5 min). PCR products were separated by agarose gel electrophoresis and visualized through ethidium bromide staining, with a digital image analysis system (Gel Doc 1000; Bio-Rad Laboratories, Hercules, CA). Amplified PCR products were 370 bp for P2Y₁, 197 bp for P2Y₂, 322 bp for P2Y₇, 203 bp for P2X₇ (P2Z), and 228 bp for GAPDH.

Statistical Analysis

Data are given as means ± SEM. Differences in the amounts of histamine released among groups were determined through repeated-measures analysis of variance and the paired Student's t test.

	Results

Effects of ATP on Anti-IgE

Incubation of purified HLMC with ATP at concentrations ranging from 10⁷ to 10³ M did not directly induce histamine release (n = 23). In 20 of 23 preparations in which HLMC responded to anti-IgE stimulation, ATP at 10⁴ M enhanced histamine release (from 10.9 + 2.7% to 19.2 + 2.9%, P < 0.01). In nine of these 20 anti-IgE-responsive preparations (control anti-IgE-induced histamine release was 10.1 + 3.4%, n = 9), we examined the dose-dependent effects of ATP at concentrations from 10⁵ M-10³ M (Figure 1). In six of the nine preparations, ATP was examine at 10⁶ M. In all nine experiments, ATP at both 10⁵ M and 10⁴ M enhanced histamine release (P < 0.05). In six experiments, ATP at 10⁶ M had a similar effect, but this failed to reach significance, possibly owing to the smaller number of experiments performed with this dose of ATP (i.e., n = 6 versus n = 9 for 10³ to 10⁵ M ATP). ATP at 10³ M enhanced anti-IgE-induced histamine release in seven of nine experiments, and inhibited release in two of nine. Overall, this enhancement of histamine release, to 14.0 + 2.4%, was not statistically significant (P > 0.05). In three of 23 preparations that failed to respond to anti-IgE alone, preincubation with ATP (10⁶ to 10³ M) was without effect.

View larger version (42K): [in this window] [in a new window]   Figure 1.   Dose-response relationship of ATP-modulated histamine release induced by anti-IgE in HLMC. ATP was added to cells 15 min prior to anti-IgE (3 µg/ml) challenge. ATP potentiated the release of histamine at all doses tested; this effect reached significance at 104 M and 105 M. *P < 0.05, n = 9, except for 106 M ATP, n = 6.

We next contrasted the relationship between the least and most anti-IgE-responsive preparations to the effects of ATP (10⁴ M) (Figure 2). Interestingly, ATP enhanced anti-IgE-induced histamine release by ~ 8-10% at both extremes. Therefore, in terms of percent enhancement, the effects of ATP were most striking when anti-IgE- induced histamine release was low. Specifically, in experiments with a low (< 3%) net anti-IgE-induced release of histamine (1.8 ± 0.4%; range: 0.5-2.9%; n = 6), ATP (10⁴ M) enhanced histamine release to 13.5 ± 2.7% (750% enhancement). Anti-IgE-induced histamine release of 24.2 ± 4.2% (range: 14.0-45.9%, n = 7) was enhanced by ATP (10⁴ M) to 32.9 ± 4.5%, representing only a 35% enhancement.

View larger version (19K): [in this window] [in a new window]   Figure 2.   Potentiation by ATP (104 M) of anti-IgE-induced histamine release in HLMC. The effect of ATP was inversely related to the efficacy of anti-IgE alone in inducing histamine release. ATP was relatively more potent in preparations in which anti-IgE induced release of histamine was < 3% ("low" responders) than in preparations in which anti-IgE-induced histamine release was  14% ("high" responders). Shown are results obtained in 13 of a total of 20 preparations representing extremes of response to anti-IgE.

Calcium Ionophore Challenge

In 10 HLMC preparations, we examined the effect of ATP (10⁴ M) on ionophore A23187-induced histamine release. The extent of histamine release, of 49 ± 5.4% in the presence of ATP (10⁴ M), was not different from that induced by ionophore alone, of 51.1 ± 6.6%. Of five experiments in which ionophore A23187 alone induced a lower release of histamine (1-30%), which was more comparable with the responses observed with anti-IgE, no enhancement by ATP was observed in three.

Adenosine Effects

Because ATP is degraded to adenosine by ectoenzymes (3), and adenosine modulates histamine release from rat and human mast cells and basophils (38), we also determined the effect of adenosine on histamine release from HLMC. In six dose-response experiments (Figure 3), previous observations (37) were confirmed: Adenosine alone did not directly induce histamine release from HLMC, but exerted a bimodal modulatory effect on anti-IgE-induced histamine release; adenosine at 10³ M inhibited anti-IgE-induced release of 10.3 ± 3.0% to 5.3 ± 1.9% (P < 0.05). At lower concentrations (10⁴ to 10⁵ M), adenosine enhanced histamine release to 11.2 ± 4.7% and 13.4 ± 5.6%, respectively, but neither effect was statistically significant (n = 6). In these same experiments, ATP at both 10⁴ M and 10⁵ M significantly enhanced anti-IgE-induced histamine release.

View larger version (35K): [in this window] [in a new window]   Figure 3.   Comparative modulatory effects of ATP and adenosine on anti-IgE-induced histamine release in HLMC. ATP potentiated histamine release, whereas adenosine exhibited a bimodal effect, inhibiting the release of histamine at 103 M, and potentiating its release at 104 M and 105 M, although the latter potentiating effect did not reach significance (n = 6).

Ectonucleotidase Activity

To determine whether the effects of extracellular ATP on purified HLMC might be partly mediated by degradation to adenosine, we used HPLC to examine the potential ectoenzymatic breakdown of ATP to adenosine. In three individual experiments, HLMC failed to exhibit functional ectoATPase activity (Figure 4). Human lung fragments under identical conditions exhibited conversion of ATP to adenosine over the 15-min incubation period (data not shown).

View larger version (26K): [in this window] [in a new window]   Figure 4.   Lack of degradation of extracellular ATP by HLMC activated by anti-IgE. Conditions are as defined in MATERIALS AND METHODS. (A) Anti-IgE-activated HLMC. (B) Anti-IgE- activated HLMC + 104 M ATP. (C) Control; 104 M ATP without HLMC. There is no noticeable decrease in the area under the ATP peak (arrow at 20 min) in (B) versus (C), and no additional peaks corresponding to ATP metabolites (i.e., ADP, AMP, adenosine) appear in (B). The early peak is the solvent-front artifact, and the low broad peaks are due to the change in solvent composition. Representative of three experiments.

ATP Analogues

P2-purinoceptors have been pharmacologically classified as P2X and P2Y on the basis of patterns of responsiveness to synthesized purine analogues (5, 50). In 10 experiments, we determined the effects of a series of ATP analogues on anti-IgE-induced histamine release. Anti-IgE-induced release of 9.9 ± 3.1% was enhanced by all members of this series of compounds. In eight of 10 experiments, ATP itself was the most potent enhancer (17.7 ± 4.1%). In two of 10, 2mSATP was the most potent analogue, and in five of 10 was the second most potent analogue (14.3 ± 3.9%, n = 10). Overall, the rank order of potency was ATP > 2mSATP > ,mATP > ,mATP, which is in accord with a P2Y-purinoceptor-mediated functional response (50).

Uridine Triphosphate Responsiveness

Because P2Y₂-purinoceptors have been shown to be widely expressed in immune cells (51), we compared ATP with uridine triphosphate (UTP), the preferred agonist for this receptor, for effects on anti-IgE-induced histamine release. In this group of six experiments, control anti-IgE- induced histamine release of 14.9 ± 3.9% was enhanced by ATP (10⁴ M) to 23.0 ± 4.7% (P < 0.05), as compared with 19.2 ± 5.0% (P < 0.05) in the presence of an equimolar concentration of UTP. Thus, UTP was less potent than ATP in modulating anti-IgE-induced histamine release.

Pyridoxalphosphate-6-Azophenyl-2',4'-Disulfonic Acid

To further exclude possible mediation by a functional P2X-purinoceptor(s) (50, 52), we examined the effect of pyridoxalphosphate-6-azophenyl-2',4'-disulfonic acid (PPADS), a putative selective P2X-purinoceptor antagonist (53), on the effect of ATP. In four experiments, anti-IgE-induced control release of histamine of 11.2 ± 5.3% was enhanced by ATP (10⁴ M) to 15.7 ± 7.1%. Preincubation of HLMC in PPADS (10⁴ M) for 15 min prior to addition of ATP (10⁴ M) produced no significant modulation of the ATP effect (16.1 ± 5.9% release).

P2Y-Receptor Expression

To corroborate the analogue and antagonist data, we examined purified HLMC preparations for constitutive expression of P2Y-purinoceptors, which has recently been reported for other human immune cells (51). In all of five experiments, HLMC expressed transcripts for P2Y₁, and in all of three experiments HLMC expressed transcripts for P2Y₂ (Figure 5). P2Y₇, a purinoceptor found in human cell systems, was undetected in four of five and was faintly expressed in one of five experiments. On the basis of reports implicating the P2X₇/P2Z purinoceptor in rodent mast cell histamine release, we examined HLMC for similar expression of this receptor. We failed to detect the receptor in any of five purified HLMC preparations. This transcript was, however, expressed robustly on cells of the HMC-1 line (Figure 6).

View larger version (65K): [in this window] [in a new window]   Figure 5.   HLMC express mRNA for P2Y1 and P2Y2, but not for P2Y7. HLMC were challenged with either buffer or anti-IgE for 2 h, followed by extraction of tcRNA. Shown are GAPDH amplified for 20 cycles, and P2Y1, P2Y2, and P2Y7 amplified for 30 cycles. P2Y1 and P2Y2 are expressed both constitutively and with anti-IgE challenge. In the experiment shown, there is apparent increased expression following the immunologic challenge. Pos. = positive control: PCR mixture including cDNA obtained from peripheral blood mononuclear cells. Neg. = negative control: PCR mixture without cDNA (i.e., H2O added).

View larger version (40K): [in this window] [in a new window]   Figure 6.   P2X7 mRNA is expressed by the human leukemic mast cell line HMC-1, but not by HLMC. HMC-1 (5 × 105/ml) were challenged with buffer, phorbol myristate acetate (50 ng/ml), or ionophore A23187 (1 µg/ml) for 2 h. HLMC (5 × 105/ml) were challenged with buffer or anti-IgE for 2 h. Following challenges, cells underwent tcRNA extraction. RT-PCR was performed for 30 cycles. GAPDH signal was readily detected in all cell samples at 20 cycles of PCR (data not shown). Positive and negative controls are indicated as previously described.

	Discussion

To date, adenosine, and not adenine nucleotides, has received considerable recognition for a potential role in human allergies and asthma (37). The findings of the current study show for the first time that extracellular ATP and related adenine nucleotides markedly potentiate histamine release from immunologically stimulated mast cells. This effect was not attributable to ectoenzymatic breakdown of adenine nucleotides to adenosine. Also, adenosine, in contrast to ATP, exerted a bimodal effect on anti-IgE-induced histamine release: At high concentrations it significantly inhibited and at lower concentrations potentiated histamine release, though not significantly. Furthermore, in absolute terms, the enhancing effects of ATP were greater than those of equimolar doses of adenosine. This action of ATP on HLMC is distinct from that described for rat peritoneal mast cells, in which ATP alone provokes histamine release (8). In rat cells, this response was attributed to activation of the ligand-binding channel receptor P2X₇/P2Z, for which the agonist is the tetrabasic acid form of ATP (ATP⁴) (13). In the present experiments, negligible amounts of ATP⁴ were present, owing to the inclusion of both Ca²⁺ and Mg²⁺ at millimolar concentrations in all assay buffers. Moreover, ATP challenge of HLMC in Ca²⁺- and Mg²⁺-free media failed to provoke histamine release (results not shown).

Evidence generated in the present study suggests that the subtype of P2-purinoceptor mediating the effects of ATP on HLMC differs from that described in the rat peritoneal mast cell. The HLMC response appears to be mediated by a member of the G protein-coupled P2Y-family of receptors, on the basis of the following lines of evidence: First, the pattern of pharmacologic responsiveness to stable analogues of ATP (ATP > 2mSATP > ,mATP > , mATP) is consistent with abundant support in the literature for P2Y-purinocepter- and not P2X-purinoceptor- mediated effects (50, 51). Second, in the absence of any available "selective" pharmacologic antagonists or antibodies to P2Y-purinoceptor, we examined the effects of the putative P2X-selective antagonist PPADS (52) on the effect of ATP on histamine release. The failure of this agent to influence ATP-induced enhancement of histamine release further supports P2X-independent mediation of such release. Moreover, in all preparations examined, HLMC expressed mRNA for both P2Y₁- and P2Y₂- purinoceptors. The mRNA for P2X₇/P2Z, the purinoceptor that mediates the cell-membrane permeabilization action of ATP (53) and histamine release from rodent mast cells (13), was either absent or at best weakly present in these HLMC preparations. Interestingly, this contrasts with the robust expression of P2X₇/P2Z found in the HMC-1 line, and points to differences between this cell line and HLMC for this and other gene transcripts (32, 33). P2Y₇, a purinoceptor found in human cell systems exhibiting a pharmacologic analogue profile compatible to that described earlier, was undetected in four of five preparations and was only faintly detected in one preparation. It should be noted here that a P2Y-like protein encoded by cDNA cloned from a human erythroleukemic cell line (54) and designated the P2Y₇ receptor has been recently identified as a receptor for leukotriene B₄ (55). Although we have shown that HLMC constitutively express P2-purinoceptors, it is not known how individual purinoceptors are modulated by conditions of culture, cell activation, and pharmacologic agents.

The effects of extracellular ATP on immunologically-mediated histamine release were consistently observed, even in experiments in which ATP failed to enhance calcium ionophore A23187-triggered histamine release. This suggests that the potentiating effect of extracellular ATP on anti-IgE-induced histamine release is due to a complex interaction between the two relevant signal-transduction pathways, rather than merely to an increase in Ca²⁺ influx induced by ATP.

The present data strongly suggest that a P2Y-purinoceptor signal-transduction pathway mediates the modulatory action of ATP on histamine release from HLMC. Stimulation of the P2Y-purinoceptor can result in the activation of two different pathways: one that involves pertussis toxin (PTX)-sensitive G protein(s) and adenylyl cyclase, and the other that involves PTX-insensitive G protein and phospholipase C. Because PTX failed in our hands to modify the effect of ATP on anti-IgE-induced histamine release (data not shown), it seems likely that the latter pathway mediates the effect of ATP. The exact mechanism by which this pathway operates remains to be elucidated, but it probably involves ATP-induced increased intracellular Ca²⁺ ([Ca²⁺]_i) levels. Indeed, in preliminary experiments with a fluorescent dye, we recorded such increases (Schulman and colleagues, unpublished observations). Thus, the lack of effect of ATP on calcium ionophore-induced histamine release could be explained by near maximal levels of [Ca²⁺]_i induced by the ionophore and which render the P2Y-purinoceptor-dependent effect of ATP on [Ca²⁺]_i inconsequential. It can be hypothesized that the anti-IgE- and ATP-activated pathways converge functionally at a step characterized by increased levels of [Ca²⁺]_i. The relative roles of extracellular Ca²⁺ influx and of Ca²⁺ released from internal stores in the action of ATP remain to be determined.

ATP is released into the extracellular space from ischemic cells, inflammatory cells, necrotic and apoptotic cells, activated platelets, and exercising muscle cells, as well as from nerve terminals as a cotransmitter (17). Thus, it can be hypothesized that endogenous ATP can exacerbate anti-IgE-induced histamine release in vivo. Furthermore, it seems that ATP could play an important mechanistic role in histamine release and therefore in bronchoconstriction in asthmatic patients. We have previously shown that extracellular ATP can stimulate vagal afferent nerve terminals in the lung (56). This could lead to both local axon and central vagal reflexes, which are known to play a major role in neurogenic inflammation and neurogenic bronchoconstriction (57). For example, ATP, stored in large amounts in platelets, is released into the extracellular fluid during platelet activation (23). Indeed, it has already been shown that both platelet activation (60, 61) and augmentation of vagal tone are associated with nocturnal asthma characterized by acute bronchoconstriction in the early morning (62, 63). Thus, ATP could trigger neurally mediated bronchoconstriction exacerbated by the direct action of ATP on mast cells. On the basis of this interpretation of our data, it can be conceptualized that there is an ATP axis in asthma, the modulation of which by pharmacologic agents could provide a new therapeutic target in this disease.