A2B Adenosine and P2Y2 Receptors Stimulate Mitogen-activated Protein Kinase in Human Embryonic Kidney-293 Cells

Mitogen-activated protein kinase (MAPK) cascades underlie long-term mitogenic, morphogenic, and secretory activities of purinergic receptors. In HEK-293 cells,N-ethylcarboxamidoadenosine (NECA) activates endogenous A2BARs that signal through Gs and Gq/11. UTP activates P2Y2 receptors and signals only through Gq/11. The MAPK isoforms, extracellular-signal regulated kinase 1/2 (ERK), are activated by NECA and UTP. H-89 blocks ERK activation by forskolin, but weakly affects the response to NECA or UTP. ERK activation by NECA or UTP is unaffected by a tyrosine kinase inhibitor (genistein), attenuated by a phospholipase C inhibitor (U73122), and is abolished by a MEK inhibitor (PD098059) or dominant negative Ras. Inhibition of protein kinase C (PKC) by GF 109203X failed to block ERK activation by NECA or UTP, however, another PKC inhibitor, Ro 31-8220, which unlike GF 109203X, can block the ζ-isoform, and prevents UTP- but not NECA-induced ERK activation. In the presence of forskolin, Ro 31-8220 loses its ability to block UTP-stimulated ERK activation. PKA has opposing effects on B-Raf and c-Raf-1, both of which are found in HEK-293 cells. The data are explained by a model in which ERK activity is modulated by differential effects of PKC ζ and PKA on Raf isoforms.

are divided into two major subfamilies, the P2X receptors that are ligand-gated channels, and the P2Y receptors that are G protein-coupled (2). The activation of G protein-coupled purinergic receptors has acute functional effects on all tissues that can be attributed to G protein-mediated effects on enzymes and ion channels. In addition, recent evidence indicates that purinergic receptor activation produces more slowly developing mitogenic, morphogenic, and secretory activities (3,4).
Recent studies have suggested that A 2B ARs, in addition to coupling to G s and cyclic AMP accumulation, appear to be responsible for triggering acute Ca 2ϩ mobilization and degranulation of canine mast cells (5) as well as a delayed interleukin-8 release from human HMC-1 mast cells (6). A role for mast cell A 2B ARs in asthma is suggested by the therapeutic efficacy of theophylline and enprofylline. Both of these xanthines were found to block human A 2B ARs in the therapeutic dose range, and enprofylline was found to be a selective antagonist of human A 2B ARs (7). Stimulation of adenylyl cyclase probably cannot account for A 2B AR-mediated degranulation and stimulation of interleukin-8 synthesis from human HMC-1 mast cells, and in fact cyclic AMP has been found to be inhibitory to rodent mast cell degranulation (8,9). In mast cells, activation of IgE receptors and adenosine receptors produces a synergistic interaction to trigger degranulation (10). IgE receptors are known to activate MAPK in mast cells (11,12), but little is known about the regulation of this signaling pathway by adenosine receptors. The study of mast cell adenosine receptors is complicated by the fact that individual cells express multiple adenosine receptor subtypes. In addition, different adenosine receptor subtypes appear to be functionally predominant in different mast cell lines (5,13,14). For this reason we decided to initially characterize functional effects of the endogenous A 2B AR in HEK-293 cells where it is the only adenosine receptor expressed.
ERK1/2 are 44-and 42-kDa isoform members of the MAPK family that regulate gene expression, protein synthesis, cell growth, secretion, and differentiation (15,16). MAP kinase signaling was initially shown to be activated by single-transmembrane receptor protein tyrosine kinases, such as the EGF and platelet-derived growth factor receptors. In recent years, a number of mitogenic G protein-coupled receptor (GPCR) agonists including lysophosphatidic acid (17), angiotensin II (18), endothelin (19), thromboxane A 2 (20), and bombesin (21) have been shown to be capable of potently activating ERK. In contrast to receptor tyrosine kinases, the intermediate steps linking GPCRs to the activation of ERK are poorly understood, and significant heterogeneity and complexity exist in the signaling pathways utilized by various GPCRs (22). It is now widely believed that the mechanism of ERK activation by GPCRs varies among cell types and individual receptors (23). * This work was supported by National Institutes of Health Grants RO1-HL37942 (to J. L.) and GM 47332 (to M. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
In the present study we show that activation of HEK-293 cell adenosine receptors stimulates adenylyl cyclase, Ca 2ϩ mobilization, and ERK1/2 activation. ERK activation is Ras-dependent, but is not blocked by inhibitors of protein kinase C (PKC) or tyrosine kinases, and differs from ERK activation elicited by UTP acting on a P2Y 2 receptor. We also demonstrate that A 2B ARs are principally responsible for initiating a sustained ERK activation in canine mastocytoma cells.
Cell Culture and Transfection-HEK-293 cells were obtained from the American Type Culture Collection and maintained in Dulbecco's modified Eagle's medium/F-12 medium supplemented with 10% fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin at 37°C in a humidified 5% CO 2 atmosphere. Canine BR mastocytoma cells were maintained in low-glucose Dulbecco's modified Eagle's medium supplemented with 2% donor calf serum, 1.5 mM L-histidine, 100 units/ml penicillin, and 100 g/ml streptomycin. Transient transfection of HEK-293 cells was performed on 90% confluent monolayers in 100-mm plates by means of LipofectAMINE according to the manufacturer's protocol. Empty pcDNA3 vector was added to keep the total mass of DNA added per plate constant. ERK1/2 activation assays were performed approximately 30 h after transfection.
Cyclic AMP Assays-HEK-293 cells were washed twice and resuspended in serum-free Dulbecco's modified Eagle's medium/F-12 containing 15 mM HEPES, pH 7.4, 1 unit/ml adenosine deaminase, and 20 M of the phosphodiesterase inhibitor, Ro 20-1724, and then aliquoted into test tubes. Compounds in 50-l aliquots were added to 200 l of cell suspension and transferred to a 37°C shaker bath for 15 min. Assays were terminated by the addition of 500 l of 0.15 N HCl. Cyclic AMP in the acid extract (500 l) was acetylated and quantified by automated radioimmunoassay (26).
Measurement of Intracellular Ca 2ϩ -Monolayers of HEK-293 cells were loaded with 1 M Fura-2/AM in buffer containing 100 mM NaCl, 5 mM KCl, 1 mM MgSO 4, 1 mM KH 2 PO 4 , 25 mM NaHCO 3, 0.5 mM CaCl 2 , 2.7 g/liter D-glucose, 20 mM HEPES, pH 7.4, and 0.25% bovine serum albumin for 45 min. Cells were washed and resuspended in the same buffer without bovine serum albumin, plus 1 unit/ml adenosine deaminase. Fluorescence was monitored at an emission wavelength of 510 nm and excitation wavelengths of 340 and 380 nM using an SLM spectrofluorimeter in a thermostable cuvette.
ERK1/2 Activation Assay-Prior to stimulation, HEK-293 cells or canine BR cells were serum-starved for about 18 h. Assays were carried out on monolayers of HEK-293 cells in serum-free Dulbecco's modified Eagle's medium/F-12 medium in a 37°C, 5% CO 2 incubator or on suspended canine BR cells in complete Tyrode's buffer in a 37°C shaking water bath. The reactions were terminated by placing the cells on ice and washing with ice-cold phosphate-buffered saline. The cells were then lysed in Triton lysis buffer (50 mM Tris-HCl, pH 7.5, 100 mM NaCl, 50 mM sodium fluoride, 5 mM EDTA, 1% (v/v) Triton X-100, 40 mM ␤-glycerophosphate, 40 mM p-nitrophenyl phosphate, 200 M sodium orthovanadate, 100 M phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml pepstatin A, 1 g/ml aprotinin). The lysate was mixed and clarified by centrifugation (15 min, 14,000 rpm, 4°C) in an Eppendorf microcentrifuge. The supernatant was subjected to SDS-polyacrylamide gel electrophoresis followed by transfer to nitrocellulose and immunoblotting. For co-transfection experiments, FLAG-tagged ERK2 was immunoprecipitated from the cell lysate (ϳ400 g) using the anti-FLAG M2 gel according to the manufacturer's instruction (Kodak) before resolution by SDS-polyacrylamide gel electrophoresis. Phosphorylation and activation of ERK1/2 was detected by immunoblotting using rabbit polyclonal anti-phospho-ERK1/2 antibody and visualized by enhanced chemiluminescence with horseradish peroxidase-conjugated goat antirabbit IgG as the secondary antibody (1:10,000 dilution). The membranes were then stripped by incubating in stripping buffer (62.5 mM Tris-HCl, 2% SDS, and 100 mM ␤-mercaptoethanol, pH 6.7, at 65°C) in a shaking water bath, and re-probed with mouse monoclonal anti-ERK2 antibody to quantify the total ERK2 loaded onto each lane. For quantification of ERK1/2 phosphorylation, films were scanned by a laser densitometry (Molecular Dynamics) and volume integration was performed using Image Quant TM software (Molecular Dynamics).

Endogenous A 2B ARs Evoke cAMP Accumulation and Ca 2ϩ
Mobilization in HEK-293 Cells-G s -coupled A 2B ARs are widely expressed in tissue culture lines. To detect A 2B ARs in cultured HEK-293 cells, we performed cAMP accumulation assays using the non-selective adenosine receptor agonist NECA. As shown in Fig. 1A, NECA produces a concentration-dependent increase in intracellular levels of cyclic AMP with an EC 50 of 2.7 Ϯ 0.9 M. Fig. 1B shows that the response to NECA (1 M) is substantially attenuated by the A 2B AR-selective antagonist enprofylline (100 M) as well as by the non-selective AR antagonist theophylline (100 M), but not by the A 1 AR-selective antagonist, WRC0571. In binding assays both enprofylline and theophylline block recombinant human A 2B with K I values of 7 M (7). When added at 1 M, CPA, IB-MECA, or CGS21680, agonists that are selective for A 1 , A 3 , A 2A adenosine receptors, respectively, had little effect on intracellular cAMP in HEK-293 cells (Fig. 1B). Only at a very high concentration (100 M) did the A 2A -selective compound CGS21680 induce a small increase in intracellular cAMP (Fig. 1B). Also, we could not detect A 1 , A 2A , or A 3 receptors by subtype-selective radioligand binding to HEK-293 cell membranes (data not shown). These find-ings are consistent with the observation that mRNA transcripts for A 2B , but not for A 1 , A 2A , or A 3 adenosine receptor subtype have been detected in HEK-293 cells by Northern analysis (27). Collectively, these data suggest that the predominant endogenous adenosine receptors found on HEK-293 cells are A 2B ARs that are functionally coupled to G s to stimulate adenylyl cyclase.
We next sought to identify and characterize other signaling pathways mediated by A 2B ARs in HEK-293 cells. We found that NECA (1 M) triggers transient intracellular Ca 2ϩ mobilization ( Fig. 2A), which is blocked by both enprofylline and theophylline but not by WRC0571, whereas 1 M CPA, IB-MECA, or CGS21680 failed to provoke such a response (Fig.  2B). Ca 2ϩ mobilization also is elicited in response to UTP (via a P2Y 2 receptor) or lysophosphatidic acid, as described in previous studies (2,28). Overnight pretreatment of HEK-293 cells with 100 ng/ml pertussis toxin had no effect on the NECA-or UTP-induced increase of intracellular Ca 2ϩ level (data not shown), but a 15-min pretreatment with 10 M U73122, a specific phospholipase C inhibitor (29), completely abolished both the NECA-and UTP-induced responses (Fig. 2C). The action of NECA to increase cyclic AMP and Ca 2ϩ signaling cannot be attributed to acute cross-talk between these two signaling pathways since forskolin elevates cyclic AMP and not Ca 2ϩ , and UTP elevates Ca 2ϩ , but not cyclic AMP (data not shown). Based on these findings, we conclude that endogenous-A 2B adenosine receptors in HEK-293 cells couple to cAMP accumulation via G s , and to Ca 2ϩ mobilization via a pertussis toxin-insensitive G protein, probably G q /11. Endogenous A 2B ARs Activate ERK1/2 in HEK-293 Cells-Many G i -, G q -, and some G s -coupled receptors have been shown to elicit ERK activation in a variety of tissues and cultured cells, but little is known about the regulation of this pathway by adenosine receptors. We next set out to determine if endogenous A 2B ARs also couple to ERK activation in HEK-293 cells. HEK-293 cells were serum-starved overnight prior to stimulation with NECA, and ERK1/2 activation was then monitored by Western analysis using phospho-specific ERK antibodies, which only recognize activated and dually phosphorylated (Thr 183 and Tyr 185 ) ERK1/2. As shown in Fig. 3 4B). These data suggest that NECA-induced ERK1/2 activation in HEK-293 cells is mediated by the endogenous A 2B AR.
A 2B AR-induced ERK1/2 Activation Is MEK1/2-and Ras-dependent-To investigate the mechanism of A 2B AR activation of ERK1/2, we first examined the effect of a highly specific inhibitor of MEK, PD098059 (30). The ERK activation cascade is thought to proceed through Raf, which phosphorylates and activates MEK1/2. The MEKs phosphorylate ERK1/2 on both Thr and Tyr residues. PD098059 inhibits the activation of both MEK1 (IC 50 ϭ 5-10 M) and MEK2 (IC 50 ϭ 50 M). As shown in Fig. 5A, ERK1/2 activation in response to 10 M NECA stimulation was completely abolished by pretreatment for 20 min with 50 M PD098059, suggesting that MEK1/2 are involved in A 2B AR-mediated ERK1/2 activation.
Next, we investigated the involvement of p21 ras (Ras) in NECA-induced ERK1/2 activation. Both Ras-dependent and independent pathways have been reported for GPCR-mediated ERK activation (31). HEK-293 cells were transiently transfected with FLAG-tagged ERK2 together with either dominantnegative Ras-N17 or empty vector pcDNA3. Consistent with the well known involvement of Ras in receptor protein tyrosine kinase-mediated ERK activation, overexpression of Ras-N17 (confirmed by Western analysis using anti-Ras antibodies, Fig.  5B, bottom blot) completely inhibited the ERK activation by EGF. Also inhibited were the NECA-and UTP-induced ERK activation. These data suggest that the signaling from the A 2B AR or the P2Y 2 receptor to ERK activation requires functional Ras in HEK-293 cells.
Insensitivity of A 2B AR-mediated ERK1/2 Activation to the Tyrosine Kinase Inhibitor Genistein-Ras activation in response to EGF or ligands for G-protein coupled receptors such as the lysophosphatidic acid receptor in Rat-1 fibroblasts (32) generally requires tyrosine kinase activation, which, in most cases, can be blocked by the tyrosine kinase inhibitor, genistein. Although A 2B AR and UTP receptor activation of ERK1/2 is Ras-dependent, neither response is affected by preincubation of cells with 100 M genistein for 20 min, whereas under the same conditions, EGF-induced ERK1/2 activation was greatly reduced (Fig. 6). This suggests that NECA-and UTP-induced ERK1/2 activation in HEK-293 cells may utilize genisteininsensitive tyrosine kinases or be independent of tyrosine kinase activity.
Effect of Elevated Intracellular Cyclic AMP on A 2B AR-mediated ERK1/2 Activation-Since A 2B ARs are positively coupled to adenylyl cyclase, we set out to determine if increased cAMP contributes to A 2B AR-induced ERK1/2 activation. Depending on the cell type, cAMP can have either a stimulatory (via B-Raf) or an inhibitory (via c-Raf-1) impact on ERK activation (33). Western analysis reveals the presence of both B-Raf and c-Raf-1 in HEK-293 cells (data not shown). Forskolin (10 M) increased cyclic AMP and induced a transient ERK1/2 activation in HEK-293 cells with a time course similar to that produced by NECA (data not shown). However, the magnitude of ERK activation in response to forskolin (10 M) was about 35% lower than the activation induced by NECA (Fig. 7). The in-crease in intracellular cAMP in response to a 5-min simulation with forskolin (10 M) is about 2-fold higher than that induced by NECA (10 M, data not shown). These data indicate that cAMP accumulation can contribute to but may not fully account for the NECA-stimulated ERK activation. On the other hand, we investigated the effect of the protein kinase A inhibitor, H-89, on NECA-and forskolin-stimulated ERK activation. In a series of experiments, pretreatment of cells with H-89 (10 M, 30 min) abolished the forskolin-induced ERK1/2 activation, whereas it only slightly decreased NECA-or UTP-induced ERK1/2 activation (Fig. 7). Taken together, these data are consistent with the hypothesis that cyclic AMP may have both stimulatory and inhibitory inputs on A 2B AR-mediated ERK activation.
Effect of the Phospholipase C Inhibitor U73122 on A 2B ARmediated ERK1/2 Activation-Since A 2B ARs appear to couple to G q /11 and activation of this pathway stimulates phospholipase C activity, we next set out to determine if phospholipase C is involved in NECA-stimulated ERK1/2 activation. As shown in Fig. 8, preincubation of HEK-293 cells with the specific phospholipase C inhibitor, U73122 (10 M for 15 min), significantly attenuates (Ͼ50%) but does not eliminate NECAand UTP-stimulated ERK1/2 activation, suggesting NECAand UTP-induced ERK1/2 occurs at least in part via phospholipase C activation.
We were particularly struck by the differential effect of Ro 31-8220 on NECA-and UTP-stimulated ERK activation. Since A 2B ARs signal through G s and G q /11, and P2Y 2 receptors signal through G q /11 only, we set out to determine if the differential effect of Ro 31-8220 is due to the elevated cAMP accompanied by G s activation. We reasoned that through simultaneous application of both UTP and forskolin to the cell, it would be possible to mimic the cellular effect of A 2B AR activation by NECA. Fig. 9C shows that in fact the combination of forskolin and UTP does mimic the NECA response and is not inhibited by Ro 31-8220. In addition, ERK activation in response to forskolin alone is enhanced by Ro 31-8220 pretreatment. These data suggest that the lack of an apparent inhibitory effect of Ro 31-8220 on A 2B AR-induced ERK1/2 activation may be due to the enhancement of cyclic AMP-mediated responses by Ro 31-8220.
A 2B ARs Initiate Sustained ERK1/2 Activation in Canine BR Mast Cells-A 2B ARs have recently been shown to play an important role in regulating degranulation and cytokine release from canine and human mast cells (5). We next determined if the A 2B AR-mediated ERK1/2 activation that occurs in HEK-293 can also be observed in canine BR mast cells. As shown in Fig. 10, NECA elicited ERK1/2 activation in canine BR mast cells. In contrast to the transient ERK1/2 activation in HEK-293 cells, the response in canine BR cells, peaked by 1 min (the earliest time point assayed), and was sustained for at least 60 min (Fig. 10A). This response was also completely blocked by the MEK1/2 inhibitor PD098059 (50 M, data not shown). Compared with NECA (1 M), CPA, IB-MECA, or CGS21680 (1 M) were relatively weak activators of ERK1/2 in canine BR mast cells (Fig. 10B). Furthermore, the NECA (1 M)-induced response was blocked by enprofylline (100 M). These data suggest that the A 2B AR is principally responsible for initiating the sustained ERK1/2 activation in canine BR mast cells. DISCUSSION The activation of purinergic receptors produce various acute G protein-mediated responses, e.g. changes in muscle tone, neuronal firing, immune function, and secretion of various hormones and cytokines. Recent studies also suggest that purines may trigger more slowly acting signal transduction cascades to mediate changes in cellular proliferation (36,37), growth and differentiation (38), and apoptosis (39,40). MAPK cascades may regulate the latter responses. In the present study we show that stimulation of endogenous A 2B ARs in HEK-293 cells evokes three responses: cyclic AMP accumulation, Ca 2ϩ mobilization, and activation of ERK1/2. This newly characterized A 2B AR-mediated ERK1/2 activation and a P2Y 2 receptor-mediated response elicited by UTP are dependent on Ras and MEK1/2. Both responses are attenuated by U73122, an inhibitor of phospholipase C, are completely insensitive to genistein, an inhibitor of certain tyrosine kinases, and are only minimally affected by the PKA inhibitor, H-89. In this study, we have demonstrated for the first time the existence of an interesting interaction between PKA and PKC in regulating ERK activity by endogenous GPCRs in HEK-293 cells. We also have discovered that both B-Raf and c-Raf-1 exist in HEK-293 cells, and we discuss below how these Raf isozymes may participate in cross-talk between the PKA and PKC pathways to influence ERK activity. These results provide a novel mechanistic insight into pathways linking GPCRs to ERK activation. Our data with endogenous purinergic receptors show both similarities and differences (discussed below) with previous stud-ies that utilized transiently overexpressed receptors. A caveat to the use of overexpression is the possibility that it may result in abnormal coupling.
One interesting observation in this study is the differential effects of the two closely related bisindolylmaleimide PKC inhibitors, GF 109203X and Ro 31-8220. Whereas both inhibitors effectively blocked ERK activation by phorbol 12-myristate 13acetate, only Ro 31-8220 attenuated P2Y 2 -and A23187-mediated responses. Both compounds have been reported to be potent and selective PKC inhibitors. Nevertheless, it has been recently noted that these two PKC inhibitors have differential actions and distinct pharmacological properties (41)(42)(43)(44). Ro 31-8220 is a much more potent inhibitor of PKC (106 -169 nM versus 5800 nM) than is GF 109203X (45)(46)(47)(48), whereas they are almost equipotent as inhibitors of other PKC isoforms. This raises the interesting possibility that G q /11-coupled receptors may selectively activate PKC to regulate ERK1/2 activity in HEK-293 cells. The presence of PKC in HEK-293 cell membranes has recently been demonstrated by Western blotting (49). Of note, in this regard, is the finding that in rat astrocytes, ERK activation by endogenous P2Y receptors, was also inhibited by Ro 31-8220, but not by Gö 6976, an inhibitor of PKC ␣ and ␤1 isozymes which does not affect PKC ␦, ⑀, or (50). It has been reported that in vascular smooth muscle cells, PKC , mediates Ras-dependent ERK1/2 activation induced by Ang-II (51). Involvement of Ras in PKC activation both in vivo and in vitro has also been reported (52). In a previous study (23), it has been concluded that GPCRs in HEK-293 cells do not activate ERK via PKC because these responses are insensitive to GF 109203X. Based on our data, we propose that PKC is the primary PKC isozyme that contributes to ERK stimulation due to activation of A 2B AR and P2Y 2 receptors in HEK-293 cells (Fig. 11).
Another interesting aspect of this study is evidence of crosstalk between PKA and PKC in modulating ERK activation by A 2B ARs. We show that ERK activation by the P2Y 2 receptor, but not by A 2B AR, is inhibited by Ro 31-8220. The major difference between P2Y 2 receptors and A 2B ARs is that only A 2B ARs couple via G s to activate adenylyl cyclase. This suggests that a common G q /11 and PKC pathway utilized by A 2B ARs and P2Y 2 receptors may be modulated by cyclic AMP. We show that cyclic AMP has several effects on ERK signaling in HEK-293 cells (Fig. 11). ERK is activated by forskolin and this response is abolished by the PKA inhibitor, H-89, and enhanced by the PKC inhibitor Ro 31-8220. The addition of forskolin converts UTP-induced activation of ERK from being attenuated, to being unaffected by Ro 31-8220. According to our model (Fig. 11) Ro 31-8220 can inhibit PKC -mediated activation of ERK, but enhance a cyclic AMP-dependent B-Raf-induced activation. Hence, when cyclic AMP is elevated by NECA or UTP plus forskolin, Ro 31-8220 has little net effect on ERK activation.
According to this scenario, one would expect that cyclic AMP should contribute to ERK activation mediated by A 2B ARs. However, the PKA inhibitor, H-89, although inhibitory to forskolin-induced activation of ERK, only marginally affects NECA-induced responses. To account for this, our model draws on previous studies which show cyclic AMP can have either an inhibitory (e.g. in Rat-1 or NIH3T3 cells) or stimulatory effect (in PC12 cells) on ERK activation depending on the cell type involved. Work by Vossler et al. (33) suggested that cyclic AMP decreases ERK stimulation by inhibiting c-Raf-1 activation, whereas it increases ERK activity by activating B-Raf. Consistent with this notion, we have detected by Western blotting both B-Raf and c-Raf-1 in the HEK-293 cells used in this study. Our demonstration that forskolin can activate ERK in HEK-293 cells, likely via B-Raf, is confirmatory of previous work by Daaka et al. (53). The lack of an effect of elevated cAMP on A 2B AR-mediated ERK activation might result from a balance between opposing effects of cyclic AMP on ERK activation, i.e. stimulation via B-Raf activation and inhibition via c-Raf-1 inactivation. We note that the magnitude of cyclic AMP accumulation induced by NECA (10 M) is lower than that induced by forskolin (10 M). This may also contribute to the small apparent contribution of cAMP to A 2B AR-mediated ERK activation, along with the fact that PKC activation by NECA counteracts the effect of PKA on B-Raf.
In the present study, we show that ERK activation by both A 2B ARs and P2Y 2 receptors requires functional Ras. The mechanisms of Ras activation by various GPCRs remain poorly characterized, particularly for G q -coupled receptors. By overexpressing the carboxyl terminus of the ␤-adrenergic receptor kinases 1 (␤ARKct), a scavenger of G␤␥ released from activated G proteins, it has been shown that Ras activation by G i -coupled receptors is mediated by G␤␥ (54). However, in the case of G q -coupled receptors, the involvement of G␤␥ is controversial (54,55). We found that transfection of HEK-293 cells with ␤ARKct has little effect on co-transfected epitope-tagged ERK activation by A 2B ARs and P2Y 2 receptors in HEK-293 cells (data not shown). This suggests that G␣, but not G␤␥, is principally responsible for Ras-dependent MAPK activation following the stimulation of purinergic receptors. It is possible that G i -coupled receptors release more and/or different ␤␥ subunits than G q or G s .
We also demonstrate herein that Ras-dependent ERK1/2 activation by either A 2B ARs or UTP receptors in HEK-293 cells does not require genistein-sensitive tyrosine kinases. Although this is somewhat unexpected, it is not without precedent. G icoupled m2 muscarinic receptors expressed in Rat-1 fibroblasts activate ERK in a Ras-and Raf-dependent manner, and this response is insensitive to inhibition by genistein (56). The same holds true for G i -coupled 5-HT 1A receptors expressed in Chi-nese hamster ovary cells (22). In contrast, ERK activation by transiently transfected G i -coupled m2 muscarinic receptors in COS cells, G i -coupled lysophosphatidic acid receptors in Rat-1 fibroblast (57), or G q /11-coupled ␣ 1B -adrenergic receptors in HEK-293 cells (23) require both Ras and genistein-sensitive tyrosine kinases. It is possible that overexpression of these receptors influences coupling to ERK. Although this discrepancy cannot be definitively explained at present, these findings support the current view that GPCRs modulate ERK activity in a cell-type, receptor-specific, and possibly a receptor-density dependent manner.
The issue of how ERK signaling specificity is achieved, especially in the case of multiple GPCRs apparently coupled to the same type of G-protein in the same cell type, is addressed in a recent study by Mitchell et al. (58) describing the activation of phospholipase D by various G q /11-coupled receptors. These investigators identified a specific structural motif (NPXXY versus DPXXY) in a subset of G q /11-coupled receptors, which is important for Rho-mediated phospholipase D activation. A similar scenario is conceivable for ERK activation by GPCRs coupled to the same G proteins. G protein-coupled receptors may couple to additional cellular constituents other than G proteins to transduce signals and influence ERK activation in a receptor-specific way. Daaka and co-workers (59, 60) recently presented evidence to address the importance of GPCR endocytosis in the activation of ERKs. In view of the various pathways used for GPCR internalization, it is possible that specific mechanisms of ERK activation by various GPCRs might also reflect differences in receptor endocytosis.
In summary, all of our data fit the scheme depicted in Fig.  11. P2Y 2 receptors couple via G q /11 only, and activate ERK via pathways including genistein-insensitive tyrosine kinases, phospholipase C, PKC , Ras, and Raf. In contrast, A 2B ARs couple to both G s and G q /11, and the G s signaling branch exerts both stimulatory and inhibitory effects on G q /11-mediated ERK activation via cyclic AMP-dependent PKA. G q /11-mediated FIG. 11. Model for the regulation of ERK1/2 activity by A 2B AR and P2Y 2 receptors in HEK-293 cells. Both A 2B ARs and P2Y 2 receptors activate ERK by a pathway that is thought to include G q /11, a genistein-insensitive tyrosine kinase, Ras, B-Raf, c-Raf-1, and MEK. Forskolin and the A 2B AR elevate cyclic AMP and can activate ERK via a pathway that is attenuated by PKC (). ERK activation by endogenous purinergic receptors is stimulated in part by PKC (). The model accounts for the effects of NECA, UTP, forskolin, H-89 and Ro 31-8220 (see text).
PKC activation has inhibitory effects on cAMP-dependent ERK activation. Hence, there exists an interesting cross-talk between PKC and PKA signaling pathways in regulating ERK activity by G s -and G q /11-coupled receptors in HEK-293 cells. It is notable that extracellular ATP released from cells is rapidly broken down to adenosine by ectonucleotidases. Since both P1 and P2 receptors are simultaneously activated, cross-talk between the ERK activation pathways mediated by different purinergic receptors assumes an even broader role in a physiological context.
In the present study, we also demonstrate that A 2B ARs trigger sustained ERK activation in canine BR mast cells. Although this response is initiated by A 2B adenosine receptors, the sustained phase of activation may not be solely maintained by A 2B ARs. The mediators released in response to A 2B AR activation of mast cells may play a role in promoting sustained ERK activation. The cellular implications of transient versus sustained ERK activation are different. In PC12 neuronal cells, sustained ERK activation by NGF leads to differentiation and is associated with ERK translocation from the cytosol to the nucleus, whereas EGF-mediated transient ERK activation leads to cell proliferation but not differentiation or ERK nuclear translocation (16). What might be the function of A 2B ARmediated ERK activation in mast cells? It was reported that in the rat mast cell line (RBL-2H3), activation of IgE receptors triggers ERK activation, and this signaling is responsible for the release of arachidonic acid and the regulation of cytokine gene expression but not the release of secretory granules (which contains histamine, ATP, etc.) (11,12). Activation of PKC and an increase in intracellular Ca 2ϩ provide sufficient signals for mast cell degranulation (61). Based on these observations, we hypothesize that ERK activation may be responsible for AR-mediated release of arachidonic acid and promotion of cytokine production (such as A 2B AR-mediated interleukin-8 synthesis in human mast cell line HMC-1 (6)). The sustained phase ERK activation may also be important for AR regulation of mast cell proliferation and differentiation.