Adenosine A 2 Receptor Occupancy Regulates Stimulated Neutrophil Function via Activation of a Serine/Threonine Protein Phosphatase*

Adenosine modulates generation of superoxide anion by neutrophils via occupancy of specific adenosine A 2A receptors. However, the intracellular signal transduc- tion pathways by which occupancy of neutrophil adenosine A 2A receptors inhibits superoxide anion genera- tion (O 2 .) are not well understood. We, therefore, tested the hypothesis that signaling at polymorphonuclear leu- kocyte (PMN) adenosine receptors proceeds via activation of a serine/threonine protein phosphatase (pp). Both the specific pp1 inhibitor calyculin A (10 n M ) and the pp2A inhibitor okadaic acid (10 (cid:109) M ) enhanced O 2 . generation (185 (cid:54) 24 and 189 (cid:54) 35% of control, respectively, p < 0.0001 for both, n (cid:53) 8), as reported previously. Calyculin A, but not okadaic acid, completely reversed inhibition of stimulated O 2 . generation by the adenosine A2 receptor agonist 5 (cid:42) - N -ethylcarboxamidoadenosine (NECA; IC 50 (cid:53) 30 n M ; p < 0.0001, analysis of variance). Calyculin A also reversed the adenosine receptor-medi-ated desensitization of bound chemoattractant recep- tors in neutrophils. Treatment of PMNs with NECA increased the pp1 activity of crude membrane preparations in a time- and dose-dependent fashion (EC 50

NECA inhibited cytosolic protein phosphatase activity by 78 ؎ 12% (p < 0.003, n ‫؍‬ 6) but did not shift pp1 catalytic subunit from cytosol to plasma membrane. Similar changes were observed in neutrophil cytoplasts depleted of organelles and nucleus. Moreover, the selective protein kinase A inhibitor KT5720 (10 M) reversed the capacity of dibutyryl cAMP but not NECA to increase pp1 activity (p < 0.01, n ‫؍‬ 5) in keeping with its effects on O 2 . generation. Western blot analysis of PMN subcellular fractions demonstrated the presence of pp1␣ and pp1␥1 but not pp1␥ 2 isotypes in both cytosol and plasma membrane but not in azurophil or specific granules. We conclude from these studies that signal transduction by adenosine in PMN proceeds via a novel pathway: cAMP-independent activation of a serine/threonine protein phosphatase in the plasma membrane.
Adenosine, an autacoid released by many different cell types, regulates a variety of stimulated neutrophil functions including production of superoxide anion generation (1,2), ␤ 2 -integrin-and L-selectin-mediated adhesion to endothelial cells (3,4), and phagocytosis (5). Adenosine regulates these neutrophil functions by coupling with specific cell surface receptors (2) on the neutrophil (reviewed in Ref. 6).
Four different types of adenosine receptors (A 1 , A 2A , A 2B , and A 3 ) have been described, both at the molecular level and by pharmacological analyses (7). Based on pharmacologic data, it appears that the A 2A receptor on neutrophils mediates the inhibition of neutrophil superoxide anion generation, adhesion, and phagocytosis (6). Before their resolution at the molecular level, adenosine A 2 receptors were believed to modulate cellular function via activation of adenylyl cyclase, with cAMP as their intracellular messenger. As expected, occupancy of adenosine A 2 receptors stimulated the accumulation of cAMP in neutrophils, but unexpectedly, cAMP proved not to be the second messenger for inhibition of stimulated superoxide anion generation (8 -10).
Because adenosine A 2 receptor occupancy had previously been shown to increase cytosolic protein phosphatase activity in bovine adrenal chromaffin (PC12) cells (11), we tested the hypothesis that occupancy of neutrophil adenosine receptors stimulates an increase in the serine/threonine protein phosphatase activity of the plasma membrane where it would be situated to modulate function of the neutrophil NADPH-oxidase. We found that the protein phosphatase 1 (pp1) 1 inhibitor calyculin A completely reversed the effects of adenosine receptor occupancy on stimulated neutrophil generation of superoxide anion. Moreover, adenosine receptor occupancy stimulated an increase in plasma membrane-associated protein phosphatase activity, most likely pp1, and that activation of this phosphatase is independent of cAMP.
Isolation of Neutrophils-Human neutrophils were isolated from whole blood after centrifugation through Hypaque-Ficoll gradients, sedimentation through dextran (6% w/v), and hypotonic lysis of red blood cells. This procedure allows study of populations that are 98 Ϯ 2% neutrophils with few contaminating erythrocytes or platelets. Neutro-phils were suspended in HEPES-buffered saline supplemented with Mg 2ϩ (1.2 mM) and Ca 2ϩ (1.3 mM) (12).
Quantitation of Superoxide Anion Generation-Superoxide anion generation was monitored by determination of the superoxide dismutase inhibited reduction of ferricytochrome c. Neutrophils (20 ϫ 10 6 /ml) were preincubated for 30 min (37°C) in the presence of okadaic acid (10 M), calyculin A (10 nM), or medium. Duplicate reaction mixtures containing 2 ϫ 10 6 neutrophils and 75 nmol of ferricytochrome c, in the presence and absence of superoxide dismutase (1 g/ml), cytochalasin B (5 g/ml), and FMLP (0.1 M) as a stimulus were then prepared. After incubation (10 min at 37°C) the cells were then centrifuged at 4°C (500 ϫ g), and the supernatants were collected. Absorption at 550 nm was measured, and the nanomoles of superoxide anion generated were calculated as described previously (2,13).
Association of Chemoattractant Receptors with the Cytoskeleton-The association of chemoattractant receptors with the cytoskeleton was determined by a modification of the method of Jesaitis et al. (14,15), as we have described previously (16). Briefly, neutrophils (5 ϫ 10 6 ) were incubated with [ 3 H]FMLP (25 nM) in the presence and absence of adenosine receptor agonist (NECA, 1 M) after preincubation (30 min at 37°C) with medium alone or calyculin A (10 nM). The reaction was terminated by the addition of a 4-fold excess of ice-cold buffer. After washing with ice-cold buffer, the cells were lysed with Triton X-100 (0.5%), and then the cytoskeletons were isolated by centrifugation in a microcentrifuge for 1 min at 4°C. The cytoskeletal pellets were resuspended in scintillation fluid, and the radioactivity was quantitated. Replicate incubations with labeled FMLP were carried out in the presence of excess unlabeled FMLP (10 M, nonspecific binding), and the specific binding was calculated as the difference between the total and nonspecific binding.
Incubation of Neutrophils before Assay of Phosphorylase (Serine/ Threonine Phosphoprotein) Phosphatase Activity-Neutrophils were incubated in the presence of buffer or adenosine receptor agonists or antagonists for varying periods of time at 37°C. The cells were then washed and resuspended in lysis buffer before sonication. In preliminary experiments, we found that adenosine receptor agonists maximally increased neutrophil phosphorylase (serine/threonine phosphoprotein) phosphatase activity by 1 min of incubation and that increased phosphorylase phosphatase activity persisted unchanged for at least 5 min (data not shown); thus, all subsequent incubations were carried out for 5 min.
Separation of Crude Membranes (Particulate Fraction) from Cytosol-Following appropriate incubations, neutrophils were disrupted with a tissue homogenizer (Polytron; Brinkmann Instruments) in the presence of antiprotease complex (1 mM EDTA, 2.5 g/ml chymostatin, 1 g/ml pepstatin, 5 g/ml antipain, 2.5 g/ml N ␣ -p-tosyl-L-lysine chloromethyl ketone, 1 g/ml leupeptin, and 0.023 trypsin inhibitory units/ml aprotinin). Whole cells, nuclei, and other large debris were cleared by centrifugation at 300 ϫ g for 10 min at 4°C, the clear supernatant was collected, and a particulate fraction was separated by centrifugation at 20,000 ϫ g for 20 min at 4°C. The supernatant was collected ("cytosol"), and the pellet was washed once and resuspended in assay buffer. The protein content was determined by BCA protein assay (17).
Preparation of Neutrophil Cytoplasts-Neutrophils were suspended in 12.5% Ficoll (w/v) in the presence of cytochalasin B (20 M) and incubated for 5 min at 37°C. The suspension was then layered on a prewarmed gradient of 16% Ficoll (w/v) and centrifuged for 30 min at 33°C at 81,000 ϫ g in an ultracentrifuge. The neutrophil cytoplasts present at the interface between the 12.5 and 16% layers of Ficoll were collected and washed five times before use (18). Plasma membranes were separated from cytosol after sonication by centrifugation (200,000 ϫ g at 4°C).
Protein Phosphatase Activity-Protein phosphatase activity was measured as the calyculin A (10 nM) inhibited dephosphorylation of phosphorylase a, as described previously by Cohen et al. (19). [ 32 P]Phosphorylase a was generated from phosphorylase b (10 mg/ml) by treatment (incubated 1 h at 37°C) with phosphorylase kinase (0.2 mg/ml) in the presence of [ 32 P]ATP (0.2 mM, 10 6 cpm/pmol) in a buffer consisting of Tris (100 mM, pH 8.2), sodium glycerol 1-phosphate (100 mM), CaCl 2 (0.1 mM), MgCl 2 (10 mM), and sodium acetate (10 mM). The reaction was terminated by the addition of an equal volume of NH 4 SO 4 (475 g/l) and incubation on ice for 30 min. The iced mixture was centrifuged for 10 min at 20,000 ϫ g, and the supernatant was discarded. The pellet was resuspended (5 mM Tris, 0.1 mM EGTA, 10% glycerol, and 45% NH 4 SO 4 ) and dialyzed to remove excess 32 P. The dialyzed phosphorylase suspension was centrifuged at 15,000 ϫ g for 5 min, and the supernatant was discarded. The precipitate was then resuspended (50 mM Tris, 0.1 mM EGTA, and 0.1% dithiothreitol) and stored at 4°C until use that day. The enzyme source was incubated for 10 min at 37°C (50 mM Tris, 0.1 mM EGTA, 0.1% dithiothreitol, 0.03% Brij 35, and 1 mg/ml bovine serum albumin) suspended with substrate (3 mg/ml phosphorylase a) in the presence of caffeine (15 mM) in the presence or absence of calyculin A (10 nM) and incubated for 10 min at 37°C. The reaction was stopped by the addition of a volume of 20% trichloroacetic acid; then the mixture was incubated on ice for 30 min and centrifuged at 16,000 ϫ g for 2 min. The supernatant was collected, and radioactivity was quantitated in a scintillation counter. All assays were performed in duplicate, and replicates varied by less than 10%.
Preparation of Subcellular Fractions of Whole Neutrophils-Subcellular fractions were separated by a variation of the method of Philips et al. (20). Neutrophil suspensions were subjected to two hypotonic lysis steps, and then resuspended (1.0 -1.5 ϫ 10 9 ) in ice-cold relaxation buffer with protease inhibitors (leupeptin, chymostatin, pepstatin A, and antipain at 10 mg/ml; 2 mM L-phenylmethylsulfonyl fluoride; and 100 kallikrein inhibitory units/ml aprotinin). The neutrophil suspension was pressurized with nitrogen for 20 min at 350 psi with constant stirring in a nitrogen bomb (Parr Instruments Co., Moline, IL) at 4°C. The cavitate was then collected dropwise into EGTA, pH 7.4, sufficient for a final concentration of 1.25 mM. Nuclei and unbroken cells were pelleted by centrifugation (500 ϫ g for 10 min at 4°C). The supernatant was decanted and then loaded onto sucrose gradients (40%) that had been precooled to 4°C. Cytosol, plasma membrane, azurophil granule, and specific granule fractions were isolated by ultracentrifugation (150,000 ϫ g for 120 min at 4°C) in an SW41 rotor (Beckman Instruments). Membranes and granule fractions were subjected to seven cycles of freeze thawing in the presence of protease inhibitors, followed by centrifugation (200,000 ϫ g for 10 min at 4°) in a 100.3 rotor of a TL-100 ultracentrifuge (Beckman); pellets were resuspended in 200 -300 l of 20 mM Tris-HCl and protease inhibitors and then stored frozen at Ϫ70°C until ready for use.
Immunoblot-Proteins were subjected to SDS-polyacrylamide gel electrophoresis. Separated proteins were electroblotted on nitrocellulose in Dunn carbonate buffer. Blots were blocked with 5% nonfat dry milk in phosphate-buffered saline with 0.3% Tween 20. Blots were probed with specific antisera, washed, and developed with alkaline phosphatase-linked antisera (1:100 -1:1000 dilutions), followed by visualization by standard technique (22). Rat brain extract (5 g of protein) containing immunoreactive pp1 was used as a positive control in these experiments (data not shown).

Effect of Okadaic Acid (10 M) and Calyculin A (10 nM) on Stimulated Superoxide Anion Generation and Its Regulation by an Adenosine A 2 Receptor
Agonist-To determine whether activation of serine/threonine protein phosphatases is relevant to inhibition of neutrophil function by adenosine receptor agonists, we examined the effect of okadaic acid (10 M), a relatively specific inhibitor of protein phosphatase 2a (pp2a), or calyculin A (10 nM), a specific inhibitor of protein phosphatase 1 (pp1), on the capacity of the adenosine A 2 receptor agonist NECA to inhibit superoxide anion generation by neutrophils. Both okadaic acid and calyculin A enhanced superoxide anion generation in response to the chemoattractant FMLP (189 Ϯ 35 and 185 Ϯ 24% of control, respectively; p Ͻ 0.0001 for both, n ϭ 8), and as we and others have reported previously (6), NECA inhibited superoxide anion generation in response to FMLP (IC 50 ϭ 30 nM; Fig. 1). However, NECA failed to inhibit superoxide anion generation in neutrophils preincubated with calyculin A (Fig. 1). In contrast, okadaic acid treatment of neutrophils did not affect the ability of NECA (1 M) to inhibit superoxide anion generation (61 Ϯ 4 versus 62 Ϯ 4% inhibition in the absence and presence of okadaic acid; p ϭ not significant, n ϭ 7). These results are consistent with the hypothesis that adenosine receptor occupancy regulates superoxide anion generation via activation of a protein phosphatase, most likely pp1.

Effect of NECA on Association of Bound Chemoattractant
Receptors to the Cytoskeleton-We have demonstrated previously that adenosine A 2 receptor occupancy inhibits stimulated neutrophil function by promoting more rapid and complete association of chemoattractant receptors with the cytoskeleton. This phenomenon is associated with desensitization of chemoattractant receptors (9,10,16). We, therefore, determined whether inhibition of protein phosphatase activity could reverse the effect of NECA on association of bound FMLP receptors with the cytoskeleton. As demonstrated previously (9, 16), NECA promoted association of chemoattractant, bound to its receptor, with the cytoskeleton (177 Ϯ 21% of control association; p Ͻ 0.001, n ϭ 5). Calyculin A alone induced a small but significant increase in association of chemoattractant receptors with the cytoskeleton (120 Ϯ 15% of control association; p Ͻ 0.05, n ϭ 5). However, calyculin A partially reversed the effect of NECA on association of chemoattractant receptors with the cytoskeleton in parallel with its effects on inhibition by NECA of superoxide anion generation (142 Ϯ 6% of control association; p Ͻ 0.05 versus NECA alone, n ϭ 5). Effect of NECA on Protein Phosphatase Activity in Plasma Membrane and Cytosolic Fractions-We next examined protein phosphatase activity in soluble and particulate fractions of human neutrophils pretreated with NECA. In contrast to previous studies (11), we observed that occupancy of adenosine receptors by NECA (1 M) significantly decreased soluble protein phosphatase activity but increased protein phosphatase activity in the particulate fraction in neutrophils (Table I). To determine whether release of proteolytic enzymes from neutrophil granules during preparation might have contributed to the decreased protein phosphatase activity present in the soluble fraction, we studied the effect of NECA on protein phosphatase activity in neutrophil cytoplasts (neutrophil fragments free of nuclei and granules but containing cytosol and filamentous structures). Like whole neutrophils, incubation of cytoplasts with NECA for 5 min led to an increase in plasma membraneassociated protein phosphatase activity and a decrease in cytosolic activity (Table I). The effect of NECA on serine/threonine phosphatase activity in the particulate fraction of whole neutrophils was dose-dependent and occurred at concentrations similar to those that inhibit superoxide anion generation (EC 50 ϭ 40 nM; Fig. 2). Incubation of isolated plasma membranes, cytosol, or their combination with NECA did not induce any increase in protein phosphatase activity (92 Ϯ 5, 101 Ϯ 8, or 107 Ϯ 5% of control phosphatase activity, respectively; p ϭ not significant, n ϭ 6). Subcellular Localization of Protein Phosphatase 1 Isozymes-Since occupancy of adenosine receptors increased protein phosphatase activity in one cellular compartment (particulate fraction) but decreased protein phosphatase activity in a different cellular compartment (soluble fraction), we determined the subcellular localization of different protein phosphatase catalytic isozymes by Western blot analysis. Neutrophils, like other cell types tested (23,24), express pp1␣ and pp1␥1 but not pp1␥2 isozymes (Fig. 3). pp1␣ was expressed in both cytosol and plasma membrane, but significant expression of pp1␣ was not detected in granule fractions. Proteolytic cleavage of a terminal hexapeptide most likely accounts for the altered mobility of cytosolic pp1a detected by Western blot. 2

FIG. 2. NECA increases protein phosphatase activity in the particulate fraction of neutrophils in a dose-dependent fashion.
Neutrophils were incubated in the presence or absence of NECA at varying concentrations for 5 min (37°C) before washing, lysis, and separation of particulate from soluble fractions. Each point represents the mean (bars, S.E.) of five or more separate determinations performed in duplicate. Control phosphorylase phosphatase activity in these experiments was 1.54 Ϯ 0.14 nmol PO 4 Ϫ /mg protein/min. induce any change in the subcellular distribution of either of the pp1 isozymes (Fig. 4). We conclude from these experiments that increments in protein phosphatase activity in plasma membrane must be due to activation of the enzyme rather than translocation of catalytic subunits to the plasma membrane although the nature of the activation step(s) remains to be determined.
Effect of Dibutyryl cAMP on Protein Phosphatase Activity in a Particulate Fraction of Human Neutrophils-We have demonstrated previously that the protein kinase A inhibitor KT5720 (10 M) does not affect the capacity of NECA to inhibit superoxide anion generation but reverses inhibition by the cell-soluble cAMP agonist dibutyryl cAMP of stimulated superoxide anion generation (9). In parallel, inhibition of protein kinase A activity by KT5720 significantly reversed the effect of dibutyryl cAMP, but not NECA, on protein phosphatase activity in the particulate fraction of neutrophils (Fig. 5). The observation that inhibition of protein kinase A diminished stimulation of protein phosphatase activity by a cAMP agonist but not by an adenosine receptor agonist supports the hypothesis that occupancy of adenosine A 2A receptors inhibits superoxide anion generation by a cAMP-independent mechanism. DISCUSSION We report here evidence that adenosine receptors on human neutrophils modulate stimulated superoxide anion generation via cAMP-independent activation of a plasma membrane-associated serine/threonine protein phosphatase. Evidence to support this hypothesis includes: 1) reversal of adenosine receptormediated inhibition of superoxide anion generation by the protein phosphatase 1 inhibitor calyculin A (but not the protein phosphatase 2A inhibitor okadaic acid); 2) increased serine/ threonine protein phosphatase activity of particulate fractions from neutrophils treated with the adenosine A 2 receptor agonist NECA; and 3) reversal by the protein kinase A inhibitor KT5720 of dibutyryl cAMP-mediated, but not adenosine receptor-mediated, inhibition of superoxide anion generation and stimulation of protein phosphatase activity.
Previous studies have shown, in nonhematopoietic cells, that adenosine receptors signal via activation of serine/threonine protein phosphatases. Mateo et al. (11) have demonstrated that adenosine A 2 receptors activate a cytosolic protein phosphatase in neural cells, although the functional significance of this adenosine receptor-induced change is not clear. Gupta et al. (25) have reported that adenosine A 1 receptor agonists diminish isoproterenol-induced increases in protein phosphatase inhibitor-1 (ppi-1) activity in cardiac muscle and would, therefore, be expected to increase cytosolic protein phosphatase 1 activity. The studies reported here contrast sharply with those previous studies; the A 2 agonist NECA inhibited soluble (cytosolic) protein phosphatase activity but increased protein phosphatase activity in the particulate fraction. Moreover, the adenosine receptor-induced inhibition of superoxide anion generation correlates best with the enhanced protein phosphatase activity in the particulate fraction.
Signal transduction at adenosine receptors, like other members of the family of seven-transmembrane spanning receptors, is mediated by a family of heterotrimeric proteins which, in the active state, bind and hydrolyze GTP (G proteins). Adenosine , and azurophil granule (AG; 25 g of protein) fractions were subjected to SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and blotted for phosphoprotein phosphorylase isozymes, as described. Rat brain extracts (5 g of protein) containing immunoreactive protein phosphatase 1 were used as a positive control in these experiments. Similar results were observed in at least one other experiment for each isozyme.
FIG. 4. The adenosine A 2 agonist NECA does not affect subcellular localization of pp1 enzymatic subunits. After incubation with buffer or NECA, subcellular fractions of neutrophil cytoplasts were separated, and samples of plasma membrane (100 g/lane) and cytosol (50 g/lane) were subject to SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and blotted for phosphoprotein phosphorylase isozymes, as described. Similar results were observed in at least one other experiment for each isozyme in neutrophil cytoplasts and in three experiments with whole neutrophils.
FIG. 5. The protein kinase A antagonist KT5720 (10 M) reverses the effect of dibutyryl cAMP (DBcAMP) but not NECA on protein phosphatase activity in neutrophils. Neutrophils were incubated in the presence or absence of KT5720 for 30 min (37°C), washed, and then incubated with medium, NECA, or dibutyryl cAMP for 5 min (37°C) before washing, lysis, and separation of particulate from soluble fractions. Phosphorylase phosphatase activity was assayed as described under "Experimental Procedures." These results represent the mean of from 3 to 6 separate experiments performed in duplicate; bars, S.E. A 2 receptors are linked to G S␣␤␥ signal transduction proteins which, upon activation, dissociate to G ␣ and G ␤␥ subunits; in many cell types, the G ␤␥ proteins directly activate adenyl cyclase, leading to intracellular accumulation of cAMP. Since inhibition of protein kinase A blocks the effect of dibutyryl cAMP but not of the adenosine A 2 receptor agonist on protein phosphatase activity in the particulate fraction, it is likely that, as in the case of adenosine-mediated inhibition of O 2 . generation, cAMP is not the second messenger. Although originally described as a single receptor that signals via cAMP (26,27), it has been demonstrated previously that adenosine A 2 receptors signal some events via cAMP-independent pathways in the neutrophil (8,9). Similar observations have been made recently for mast cells as well (28). One explanation for the divergent effects on signaling at neutrophil adenosine receptors is that either G S␣ or G ␤␥ proteins directly activate plasma membrane protein phosphatase in addition to directly (G ␤␥ ) activating adenyl cyclase (29). Alternatively, adenosine receptor occupancy may activate other intermediate signaling proteins (e.g. proteases, transmethylases, kinases, or phospholipases) that stimulate phosphatase activity. Using immunoprecipitation techniques, we have been unable to demonstrate any change in phosphorylation status of either pp1 isozymes or associated proteins after treatment of neutrophils with adenosine receptor agonists 3 and other posttranslational modifications of the enzymatic or regulatory subunits of pp1 have yet to be investigated. The simplest explanation for the disparate effects of adenosine receptor occupancy on particulate and cytosolic protein phosphatase activity is that protein phosphatase enzymatic subunit is translocated from the cytosol to the plasma membrane. However, we were unable to demonstrate any change in the subcellular distribution of protein phosphatase catalytic subunits after stimulation with NECA. Alternatively, adenosine receptor occupancy may signal two separate events: activation of a protein phosphatase in the particulate fraction, and inhibition of cytosolic protein phosphatase activity. Indeed, previous studies have suggested that G ␣S -linked receptors activate a cytosolic protein phosphatase inhibitor (ppi-1) via a cAMP-dependent mechanism (30), and similar events may proceed in the cytosol of adenosine-treated neutrophils.
Previous studies have demonstrated that agents that inhibit protein phosphatases (okadaic acid and calyculin A) prevent termination of neutrophil superoxide anion generation and thereby increase the amount of superoxide anion generated (31)(32)(33)(34)(35), as found here. The effects of protein phosphatase inhibitors on stimulated neutrophil function correlate with inhibition of dephosphorylation of p47 phox , a protein involved in the assembly of the NADPH oxidase of the neutrophil; dephosphorylation of p47 phox and probably other proteins lead to termination of superoxide anion generation (32,33,35). Adenosine receptor occupancy leads to premature termination of superoxide anion generation (6) mediated, as shown here, via activation of a protein phosphatase. Adenosine receptor agonists have not previously been shown to increase the rate at which p47 phox is dephosphorylated, although careful kinetic studies have not been performed (9). It is tempting to speculate that adenosine receptor occupancy inhibits superoxide anion generation by activating a specific phosphatase that dephosphorylates p47 phox and thereby terminates the generation of superoxide anion by stimulated neutrophils. Alternatively, adenosine receptor-stimulated phosphatase(s) may dephosphorylate other, as yet unidentified, upstream signaling proteins that promote the association of bound chemoattractant receptors with the cytoskeleton, thereby desensitizing the chemoattractant receptors and terminating the generation of superoxide anion.
Regardless of the molecular targets for adenosine receptorstimulated protein phosphatase(s), the results presented here demonstrate, for the first time, that adenosine A 2a receptors on human neutrophils signal via activation of a phosphoprotein phosphatase located in the particulate fraction (plasma membrane). Our results further suggest that the protein phosphatase activated is protein phosphatase 1 and that activation of this phosphatase is independent of cAMP and its downstream effector protein kinase A.