A3 adenosine receptor activation inhibits cell proliferation via phosphatidylinositol 3-kinase/Akt-dependent inhibition of the extracellular signal-regulated kinase 1/2 phosphorylation in A375 human melanoma cells.

Adenosine exerts its effects through four subtypes of G-protein-coupled receptors: A(1), A(2A), A(2B), and A(3). Stimulation of the human A(3) receptor has been suggested to influence cell death and proliferation. The phosphatidylinositide-3-OH kinase (PI3K)/Akt and the Raf/mitogen-activated protein kinase (MAPK/ERK) kinase (MEK)/mitogen-activated protein kinase (MAPK) pathways have central roles in the regulation of cell survival and proliferation. Due to their importance, the cross-talk between these two pathways has been investigated. Here, we show that the A(3) adenosine receptor agonist Cl-IB-MECA stimulates PI3K-dependent phosphorylation of Akt leading to the reduction of basal levels of ERK1/2 phosphorylation, which in turn inhibits cell proliferation. The response to Cl-IB-MECA was not blocked by A(1), A(2A), or A(2B) receptor antagonists, although it was abolished by A(3) receptor antagonists. Furthermore, the response to Cl-IB-MECA was generated at the cell surface, since the inhibition of A(3) receptor expression, by using small interfering RNA, abolished agonist effects. Using A375 cells, we show that A(3) adenosine receptor stimulation results in PI3K-dependent phosphorylation of Akt, leading to the reduction of basal levels of ERK1/2 phosphorylation, which in turn inhibits cell proliferation.

different mechanisms through which A 3 adenosine receptors are able to inhibit cell proliferation was found to involve inhibition of telomerase activity and a cell cycle arrest in the G 0 /G 1 phase, leading to a cytostatic effect (7,13,14). Furthermore, it has been demonstrated that the antigrowth signal exerted by A 3 receptors blocks cells into G 1 -late cell cycle phase (9). Prompted by the antiproliferative effect of A 3 adenosine receptor stimulation, adenosine derivatives have been examined in vivo for their ability to suppress tumor growth with success demonstrated in different experimental tumor models in mice (15,16).
The molecular pathway sustaining the antiproliferative action of A 3 receptor has been defined. The activation of this G i -coupled receptor increases Ca 2ϩ intracellular levels and decreases cAMP concentration (17,18), and recently, it has been demonstrated that it can activate the phosphatidylinositol 3-kinase (PI3K) 1 /Akt pathway (8). On the other hand, it has been reported that A 3 stimulation correlates with the protein kinase A, Akt, c-myc, and cyclin D down-regulation (19).
The other branch of the adenosine signaling cascade has been demonstrated with studies performed in Chinese hamster ovary cells transfected with the human A 3 adenosine receptor (CHO-A 3 ) (20 -23) and in microglia cells (24). These studies show that A 3 receptor signaling in CHO cells leads to stimulation of ERK1/2 phosphorylation and activity (20,21). In particular, A 3 receptor signaling to ERK1/2 depends on ␤␥ release from pertussis toxin-sensitive G proteins, PI3K, Ras, and * 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.
Together, the Ras-MEK-ERK1/2 and the PI3K-Akt routes form the two major branches of intracellular adenosine A 3 receptor signaling. It should be noted that PI3K might be upstream/downstream of Ras, thus regulating ERK1/2. As for A 3 receptors, it has been demonstrated that Ras, in CHO-A 3 cells, is activated downstream of PI3K (21).
In the present study, we focused on the regulation of both Akt and ERK1/2 by A 3 receptors, to gain more insight into adenosine signal transduction. We studied the dynamics between these two routes by use of specific inhibitors. We present a molecular mechanism able to explain the antiproliferative activity of a selective agonist of A 3 adenosine receptor, Cl-IB-MECA, by using the human melanoma cell line A375 (25). A375 cells were chosen because the expression and the active functional role of all adenosine receptor subtypes has been firmly established and, furthermore, the singular role of the A 3 receptor in cell survival and proliferation has been evaluated (9). Thus, we have inquired into the ability of the A 3 adenosine receptor to modulate the MEK/ ERK1/2 and the PI3K/Akt pathways, focusing our study on the cross-talk signaling that has been demonstrated to be present in several cellular systems (26 -29).
Using A375 cells, we show that A 3 adenosine receptor stimulation results in PI3K-dependent phosphorylation of Akt, leading to the reduction of basal levels of ERK1/2 phosphorylation levels, which in turn inhibits cell proliferation.

Trypan Blue Exclusion
Cells were collected and stained with 0.4% of trypan blue for 5 min at room temperature before being examined under the microscope. The number of viable cells was determined by trypan blue exclusion. The dead cells that stained blue were scored positive and counted against the total number of cells to determine the percentage of cell death.

MTT Assay
The number of living cells was determined by evaluating the mitochondrial dehydrogenase activity by using MTT that is converted into a formazan product in living cells. 10 5 cells were plated in 24-multiwell plates; 500 l of complete medium were added to each well with different concentrations of Cl-IB-MECA. The cells were then incubated for 24 h. At the end of the incubation period, 50 l of MTT solution (5 mg/ml) were added to each well. The plates were incubated for 2 h at 37°C, and then 550 l of an acid propanol solution (0.1 N HCl in isopropyl alcohol) were added to each well to dissolve the formazan. The optical density of each well was read on a spectrophotometer at 570 nm. For each experiment, four individual wells of each drug concentration were prepared. Each experiment was repeated three times.

ATPLite Assay
The intracellular ATP concentration was determined with a luminescent ATP detection kit (ATPLite-M; PerkinElmer Life Sciences) according to the manufacturer's directions. Light units generated by ATP in each sample were normalized to control (solution with known ATP concentration) and expressed as the absolute ATP levels.

[ 3 H]Thymidine Incorporation
Cell Proliferation Test-Cells were seeded in fresh medium with 1 Ci/ml [ 3 H]thymidine in Dulbecco's modified Eagle's medium containing 10% fetal calf serum, penicillin (100 units/ml), streptomycin (100 g/ml), L-glutamine (2 mM) and simultaneously treated with adenosine analogues. After different times of labeling, cells were trypsinized, dispensed in four wells of a 96-well plate, and filtered through Whatman GF/C glass fiber filters using a Micro-Mate 196 cell harvester (PerkinElmer Life Sciences). The filter-bound radioactivity was counted on Top Count Microplate Scintillation Counter (efficiency 57%) with Micro-Scint 20.
JAM Test-This assay measures cell death by quantifying the amount of fragmented DNA. Target cells were labeled with 1 Ci/ml [ 3 H]thymidine for 20 h in Dulbecco's modified Eagle's medium containing 10% fetal calf serum, penicillin (100 units/ml), streptomycin (100 g/ml), L-glutamine (2 mM). The cells were then washed and treated with new unlabeled medium containing adenosine analogues for 24 h. At the end of the incubation period, the cells were trypsinized and dispensed in four wells of a 96-well plate, filtered through Whatman GF/C glass fiber filters using a Micro-Mate 196 cell harvester (PerkinElmer Life Sciences). The filter-bound radioactivity was counted on Top Count Microplate Scintillation Counter (efficiency 57%) with Micro-Scint 20.
The amount of apoptotic and necrotic cells, measured as the loss of radioactivity associated with the loss of fragmented and degraded DNA, was detected by filtration and subsequent washing with a Micro-Mate 196 cell harvester followed by quantification with a Top Count Microplate Scintillation Counter.
The percentage of cell death is expressed as 100 ϫ (dpm (U) Ϫ dpm (T) )/ dpm (U) , where dpm (U) represents the radioactivity of untreated cells and dpm (T) is the radioactivity of treated cells (9).

Morphological Analysis
To recover all seeded cells, the adherent culture fraction was trypsinized and mixed with the supernatant fraction. Then the cell suspension was spun to a slide by Cytospin 3 cytocentrifuge (Shandon) at 250 rpm for 10 min. As previously described (30), cells were fixed in 4% paraformaldehyde for 10 min and permeabilized in PBS solution containing 0.1% Triton X-100, and the DNA was stained with 4Ј,6diamidino-2-phenylindole. Slides were mounted in DABCO glycerol-PBS and observed on a Zeiss Axiophot fluorescent microscope.

Flow Cytometry Analysis
A375 adherent cells were trypsinized, mixed with floating cells, washed with PBS, and permeabilized in 70% (v/v) ethanol/PBS solution at 4°C for at least 24 h. Cells were washed with PBS, and the DNA was stained with a PBS solution, containing 20 g/ml of propidium iodide and 100 g/ml of RNase, at room temperature for 30 min. Cells were analyzed with an EPICS XL flow cytometer (Beckman Coulter, Miami, FL), and the content of DNA was evaluated by the Cell-LISYS program (BD Biosciences). Cell distribution among cell cycle phases and the percentage of apoptotic cells were evaluated as previously described (30). Briefly, the cell cycle distribution is shown as the percentage of cells containing 2n (G 0 /G 1 phases), 4n (G 2 and M phases), 4n Ͼ x Ͼ 2n DNA amount (S phase) judged by propidium iodide staining. The apoptotic population is the percentage of cells with DNA content lower than 2n.

Treatment of Cells with siRNA
A375 cells were plated in 6-well plates and grown to 50 -70% confluence before transfection. Transfection of siRNA was performed at a concentration of 100 nM using RNAiFect TM Transfection Kit (Qiagen). Cells were cultured in complete media, and at 24, 48, and 72 h total RNA was isolated for real time RT-PCR analysis of A 3 receptor mRNA and for Western blot analysis of A 3 receptor protein. To quantify cell transfection efficiency, we used fluorescein-labeled siRNA (Qiagen). After 24 h of transfection, cells were tripsinized and resuspended in PBS for flow cytometry analysis. Fluorescence obtained from FITC-siRNA-transfected cells was compared with autofluorescence generated by untransfected control.

Real Time RT-PCR Experiments
Total cytoplasmic RNA was extracted by the acid guanidinium thiocyanate phenol method (33). Quantitative real time RT-PCR assay (34) of A 3 mRNA transcript was carried out using gene-specific double fluorescently labeled TaqMan MGB probe (minor groove binder) in an ABI Prism 7700 Sequence Detection System (Applied Biosystems, Warrington, Cheshire, UK) (35). The following primer and probe sequences were used for real time RT-PCR: A 3 forward primer, 5Ј-ATG CCT TTG GCC ATT GTT G-3Ј; A 3 reverse primer, 5Ј-ACA ATC CAC TTC TAC AGC TGC CT-3Ј; A 3 MGB probe, 5Ј-FAM-TCA GCC TGG GCA TC-TAMRA-3Ј (where the fluorescent reporter FAM and the quencher TAMRA are 6-carboxyl fluorescein and 6-carboxyl-N,N,NЈ,NЈ-tetramethylrhodamine, respectively). For the real time RT-PCR of the reference gene, the endogenous control human ␤-actin kit was used, and the probe was fluorescence-labeled with VIC TM (Applied Biosystems, Monza, Italy).

Signaling Pathway
Cells were treated for 30 min with metabolic inhibitors or with drug vehicle (Me 2 SO) prior to being challenged with adenosine analogues. U0126 was used as inhibitor of MEK-1 and MEK-2 to prevent ERK-1 and ERK-2 MAPK activation. LY294002 was used as an inhibitor of PI3K, U73122 as an inhibitor of PLC, and SH6 as an inhibitor of Akt (Akt-I; Vinci-Biochem).

Ras Activity Assay
Raf-1 Ras-binding domain (RBD) fragment (Raf-1 amino acid residues 1-149) fused to glutathione S-transferase was purchased from Upstate Biotechnology, Inc. Ras-GTP from various treated lysates was "pulled down" using the glutathione S-transferase fusion protein corresponding to the human Ras binding domain of Raf-1 bound to agarose. The activated Ras affinity precipitation assay was performed as described according to the manufacturer's protocol. Briefly, A375 melanoma cells were serum-starved for 24 h, treated with Cl-IB-MECA at different concentrations (0.1-10 M) for different times (from 5 to 60 min), washed twice with cold PBS, and lysed with Mg 2ϩ lysis/wash buffer. Then the lysate was incubated with 10 g of Raf-1 RBD-conjugated agarose at 4°C for 30 min. Raf-1 RBD-conjugated agarose specifically binds to and precipitates Ras-GTP from cell lysates. After washing the beads three times with Mg 2ϩ lysis/wash buffer, they were suspended in 2ϫ Laemmli sample buffer, subjected to SDS-PAGE and immunoblot analysis using 1 g/ml anti-Ras monoclonal antibody as a primary antibody (Upstate Biotechnology), and visualized using the ECL Western blotting detection system.

Densitometry Analysis
The intensity of each band in immunoblot assay was quantified using molecular analyst/PC densitometry software (Bio-Rad). Mean densitometry data from independent experiments were normalized to control. The data were presented as the mean Ϯ S.E. and analyzed by Student's t test.

Statistical Analysis
All values given throughout are expressed as means Ϯ S.E. from three independent experiments except where indicated. Data sets were examined by analysis of variance (ANOVA) and Dunnett's test (when required). A p value less than 0.05 was considered statistically significant.

RESULTS
Cl-IB-MECA Inhibits Cell Proliferation and ERK1/2 Activation-In this study, we investigated the functionality of A 3 adenosine receptors expressed in A375 human melanoma cells by using the selective adenosine analogue Cl-IB-MECA (25).  5), or A 3 receptor antagonist MRE 3008F20 (100 nM) (lane 6) for 1 h. Cellular extracts were prepared and subjected to immunoblot assay using an Anti-ACTIVE®MAPK antibody. The blot was then stripped and used to determine total ERK1/2 expression using an anti-ERK1/2 (pAb).
dose-dependent manner, with an IC 50 ϭ 1.2 Ϯ 0.3 M. Similar results were obtained with another A 3 adenosine receptor agonist, IB-MECA (data not shown). By using the MTT assay, we observed that Cl-IB-MECA produced small but reproducible antiproliferative effects (Fig. 1B). To evaluate whether the reduced number of viable cells quantified by [ 3 H]thymidine incorporation and MTT was due to a cell death or to a proliferative block, we determined the cell toxicity of the agonist Cl-IB-MECA on melanoma cells. To this aim, we treated A375 cells with 5 and 10 M Cl-IB-MECA for 24 h and determined the cell viability by the trypan blue exclusion method. Cl-IB-MECA treatment did not significantly affect cell viability under the experimental conditions (cell viability of Cl-IB-MECAtreated cells was 97 Ϯ 3% when compared with untreated cells), suggesting that the decrease in the cell number after the treatment with the A 3 receptor agonist Cl-IB-MECA was not caused by an increase of cell death. To confirm these results, we analyzed the effect of A 3 adenosine receptor stimulation on cell survival by the JAM test. A375 cells, previously labeled with [ 3 H]thymidine, were treated for 24 h with increasing concentrations of Cl-IB-MECA. A 3 receptor stimulation did not induce cell death, as shown in Fig. 1C. To further investigate the effect of A 3 receptor stimulation on cell viability, we measured the intracellular ATP content of A375 cells (12 Ϯ 1 nmol/10 6 cells). ATP content after Cl-IB-MECA treatment was determined and normalized to the remaining cell numbers. We found that Cl-IB-MECA did not modify the intracellular concentration of ATP (13 Ϯ 2 nmol/10 6 cells, 10 M for 24 h). We also investigated cellular morphology of A375 drug-treated cells. The morphological analysis revealed the absence of mitotic figures (data not shown). To further confirm the absence of cell death under these conditions, we studied a typical event of apoptosis, namely the presence of phosphatidylserine at the outer face of the plasma membrane, in cells cultured in the presence of 10 M Cl-IB-MECA. As shown in Fig. 1D, phosphatidylserine exposure was indeed revealed, after 24-h incubation with 10 M Cl-IB-MECA, by the binding of annexin V-fluorescein. However, Cl-IB-MECA did not increase apoptosis when compared with Me 2 SO-treated cells. Apoptosis was confirmed after staining of picnotic nuclei with Hoechst 33258 (data not shown). The cytometry investigation showed a clear arrest in G 0 /G 1 cell cycle phases of melanoma cells treated with Cl-IB-MECA (10 M) with respect to control cells. The accumulation of cells in the G 1 phase was increased by up to 32% with a corresponding decrease of cells in G 2 /M (up to 28%; p Ͻ 0.05) (Fig. 1E).
To investigate the signal transduction pathway by which Cl-IB-MECA inhibits melanoma cell proliferation, we studied the changes in ERK1/2 phosphorylation, by using a specific antibody targeting phosphorylated ERK1/2. Fig. 2, A and B, shows a Western blot assay of ERK1/2 after treatment of melanoma cells with Cl-IB-MECA (10 M) in a time course experiment. Interestingly, prolonged inhibition of ERK1/2 phosphorylation was induced after the treatment of melanoma cells with Cl-IB-MECA, and this effect was sustained for 2 h after Cl-IB-MECA treatment. The kinetics of ERK1/2 phosphorylation reduction was very rapid; Cl-IB-MECA appeared to impair ERK1/2 phosphorylation after 5 min. Furthermore, we have investigated the dose dependence of Cl-IB-MECA to inhibit ERK1/2 phosphorylation. Fig. 2C shows that the maximum inhibition of ERK1/2 phosphorylation was produced by 10 M Cl-IB-MECA with an EC 50 of 4.3 Ϯ 0.5 M and 2.2 Ϯ 0.3 M for ERK-2 and ERK-1, respectively. Fig. 2D shows the data obtained after densitometry analysis of the ERK-1 and ERK-2 phosphorylation levels.
Cl-IB-MECA Inhibits Cell Proliferation and ERK1/2 Activation via Adenosine A 3 Receptors-The family of adenosine receptors consists of four subtypes of G protein-coupled receptors, designed A 1 , A 2A , A 2B , and A 3 . We have previously demonstrated that all four adenosine receptors are expressed in human melanoma A375 cells (36,37). To evaluate the functional role of adenosine receptor subtypes on A375 melanoma cell proliferation and on ERK1/2 activation, we tested the effect of Cl-IB-MECA in combination with DPCPX (an A 1 receptor antagonist), SCH 58261 (a selective A 2A receptor antagonist), MRE 2029F20 (a selective A 2B receptor antagonist), and MRE 3008F20 (a selective A 3 receptor antagonist) (36 -39). MRE 3008F20 reduced the Cl-IB-MECA effect on cell proliferation in a dose-dependent manner, but concentrations of DPCPX, SCH 58261, and MRE 2029F20 sufficient to block A 1 , A 2A , or A 2B receptors, respectively, failed to block the Cl-IB-MECA effect on cell proliferation (Fig. 2E). Furthermore, although the A 1 , A 2A , and A 2B receptor antagonists were not able to prevent ERK1/2 inhibition induced by Cl-IB-MECA, the selective A 3 receptor antagonist MRE 3008F20 (0.1 M) abrogated the Cl-IB-MECA-induced inhibition of ERK1/2 activation (Fig. 2F). Therefore, we hypothesized that the A 3 adenosine receptor subtype may be responsible for Cl-IB-MECA-mediated inhibition of cell proliferation and of ERK1/2 activation in A375 melanoma cells.
A 3 Adenosine Receptor Signals through a Pathway Including PLC/PI3K/Akt-We next sought to examine the signaling pathway by which A 3 adenosine receptor activation inhibits A375 melanoma cell proliferation and ERK1/2 activation.
Many G i/o -coupled receptors are known to activate ERK1/2 in a PI3K-dependent manner (20,40,41). In particular, A 3 receptor activation has been shown to activate PLC and PI3K (20,24). We have investigated the involvement of PLC-PI3K-Akt cascade on the Cl-IB-MECA-induced impairment of cell proliferation. We studied the effect of U73122 (a membrane-permeable amino-steroid inhibiting PLC-dependent pathways) on Cl-IB-MECA-dependent inhibition of melanoma cell proliferation. Pretreatment of cells with 0.25, 0.5, and 1 M U73122 abrogated, in a dose-dependent manner, Cl-IB-MECA effect on cell proliferation (Fig. 3A), suggesting a critical role for PLC in A 3 receptor-dependent inhibition of cell proliferation.
These results indicate that the antiproliferative effect of Cl-IB-MECA is mediated by a PLC-PI3K-Akt signaling pathway. To verify that PLC-PI3K-Akt signal was the molecular pat-hway sustained by A 3 receptor stimulation, we also investigated the effect of U73122, LY294002, and the Akt inhibitor on ERK1/2 phosphorylation levels under Cl-IB-MECA treatment.
Pretreatment of A375 cells with 10 M LY294002 for 30 min impaired Cl-IB-MECA inhibition of ERK1/2 phosphorylation (Fig. 4C). Fig. 4D shows the data obtained after densitometry analysis of the phospho-ERK-2 and phospho-ERK-1 protein levels. Similar results were obtained after the pretreatment of A375 cells with 5 M Akt inhibitor (Fig. 4, E and F). These data suggest that the A 3 adenosine receptor signals through a pathway including PLC-PI3K-Akt.
To provide additional support for this possibility, we investigated the changes in Akt phosphorylation level in A375 cells after Cl-IB-MECA treatment. Akt is a well described downstream target of PI3K activity that was expected to be phosphorylated at serine 473 upon an increase in PI3K activity (42). A375 cells were starved for 24 h to reduce basal P-Akt levels and then were treated with Cl-IB-MECA (10 M). Fig. 5A shows the kinetics of Akt phosphorylation promoted by Cl-IB-MECA. A 3 receptor stimulation induced Akt phosphorylation in a time-dependent manner. The maximum induction of Akt phosphorylation was at 5 min after the treatment (Fig. 5B).  4 -6 of A, C, and E, respectively. Cellular extracts were prepared and subjected to immunoblot assay using an anti-ACTIVE®MAPK antibody. The blot was then stripped and used to determine total ERK expression using an anti-ERK1/2 (pAb). A, C, and E, lane 1, Me 2 SO vehicle. B, D, and F, the immunoblot signals were quantified using Molecular Analyst/PC densitometry software (Bio-Rad). Densitometric analysis of ERK1/2 phosphorylated isoforms is reported. Data were normalized; the unstimulated control (0) was set to 100% as well as the cells treated with the inhibitors alone. The error bars give S.E. of three independent experiments. *, p Ͻ 0.01 versus Me 2 SO-treated cells. Analysis was by ANOVA followed by Dunnett's test. EC 50 of 151 Ϯ 19 nM (Fig. 5, C and D). The addition of the selective antagonist of the human A 3 receptor, MRE 3008F20 (1 M) blocked Cl-IB-MECA (0.1 and 1 M)-induced increase of Akt phosphorylation (Fig. 5, E and F).
To verify whether any cross-talk exists between PI3K/Akt and Raf/MEK/ERK pathways in human A375 melanoma cells after A 3 receptor stimulation, we investigated Raf-1 inactivation levels by Western blot. For this purpose, we utilized an antibody able to recognize an inhibitory phosphorylation site on Ser 259 of Raf. Cl-IB-MECA induced Raf inactive levels in a dose-and time-dependent manner. Raf phosphorylation was increased by Cl-IB-MECA with an EC 50 value of 33 Ϯ 2 nM (Fig.  6, A and B). The kinetics of Raf phosphorylation was comparable with the time course of Akt and ERK1/2 phosphorylation, with a maximum level at 5 min (Fig. 6, C and D). The addition of the selective antagonist of the human A 3 receptor, MRE 3008F20 (1-5 M) blocked Cl-IB-MECA inhibition of Raf activity (Fig. 6, E and F). Pretreatment of A375 cells with 10 M LY294002 for 30 min impaired Cl-IB-MECA inhibition of Raf (Fig. 6, G and H). Similar results were obtained after the pretreatment of A375 cells with 5 M of Akt inhibitor (Fig. 6,  G and H). These results indicate that PI3K and Akt are involved in Cl-IB-MECA-induced Raf Ser 259 phosphorylation.
A 3 Receptor Gene Silencing Attenuates the Akt/MAPK Signal Transduction Pathway-To confirm the role of A 3 receptor stimulation in the Akt/MAPK signaling pathway in vitro, we tried to knock down A 3 receptor expression using siRNA, leading to a transient knockdown of the A 3 receptor gene. We designed four siRNAs from the human A 3 receptor gene sequence. Although there was a difference in silencing ability, all of the siRNAs were able to suppress endogenous A 3 receptor protein expression in human A375 cells (Fig. 7). Therefore, siRNA were useful to investigate the A 3 receptor stimulatory effect in A375 cells.
A375 cells were mock-transfected or transfected with small interfering RNAs that target A 3 receptor mRNA (siRNA A3 ) for degradation. To evaluate transfection efficiency, A375 cells were also transfected with a siRNA control labeled with fluorescein. By flow cytometry, we observed a transfection efficiency of 86 Ϯ 5% (Fig. 7A). After transfection, the cells were cultured in complete medium, and at 24, 48, and 72 h, total RNA was isolated for real time RT-PCR analysis of A 3 receptor mRNA and for Western blot analysis of A 3 receptor protein. As expected, A 3 receptor mRNA levels were significantly reduced in cells transfected with siRNA A3 (Fig. 7B). Furthermore, A 3 receptor protein expression was strongly reduced in siRNA A3treated cells (Fig. 7, C and D). A rescue experiment, extending the analysis of A 3 receptor protein expression at 96 and 120 h post-siRNA A3 transfection, was performed. Fig. 7C shows that A 3 receptor protein began to recover in A375 cells 96 h after siRNA A3 treatment. By 120 h, the protein had returned to base-line values.
Neither mock transfection nor transfection with an siRNA targeted to an irrelevant mRNA inhibited A 3 receptor mRNA or protein expression. To confirm the specificity of the siRNA A3mediated silencing of the A 3 receptor, we investigated the expression of A 2A receptor protein in siRNA A3 -treated cells (Fig. 7C). Fig. 7C demonstrates that treatment of A375 cells with siRNA A3 reduced the expression of A 3 protein but had no effect on the expression of the A 2A receptor. Therefore, at 72 h from the siRNA A3 transfection, A375 cells were exposed to increasing concentrations of the A 3 adenosine receptor agonist Cl-IB-MECA (1-10 M) for different times. Cell proliferation was evaluated by [ 3 H]thymidine incorporation assay, and total protein was harvested for Western blot analysis. As control, A375 cells were exposed to random siRNA. We found that the inhibition of A 3 receptor expression is sufficient to block Cl-IB-MECA-induced inhibition of cell proliferation and of ERK1/2 phosphorylation levels (Fig. 7, E and F, respectively). These results clearly show the connection between A 3 receptor stimulation and MEK/MAPK signaling in melanoma cells.  2-7, respectively). B, the immunoblot signals were quantified using Molecular Analyst/PC densitometry software (Bio-Rad). Densitometric analysis of Akt phosphorylated isoforms is reported. The unstimulated control (lane 1) was set to 100%. *, p Ͻ 0.05 with respect to unstimulated control; analysis was by ANOVA followed by Dunnett's test.  3 and 4, respectively) before determination of Akt phosphorylation. F, densitometric analysis of Akt phosphorylated isoform is reported. *, p Ͻ 0.05 with respect to untreated cells; analysis was by ANOVA followed by Dunnett's test. Tubulin shows equal loading protein.

Adenosine A 3 Receptors Do Not Activate Ras in A375
Melanoma Cells-To determine whether A 3 receptor stimulation could modulate activity of Ras, we performed a Ras activation assay using cell lysates of serum-starved A375 melanoma cells as described under "Experimental Procedures." The main component of the assay was a recombinant protein derived from Raf that contains a domain capable of binding only to GTPbound Ras (Raf-1 Ras binding domain). Active GTP-bound Ras was pulled down from cell lysates with the glutathione Stransferase-Raf-RBD coupled to glutathione-agarose, and the fraction of activated Ras was determined by immunoblotting with a Ras antibody. The active form of Ras (Ras-GTP) was increased in response to 10% fetal calf serum and to GTP␥S. On the contrary, Cl-IB-MECA (1-5-10 M) did not modulate the activation of Ras (Fig. 8).
Adenosine A 3 Receptor Stimulation Inhibits MEK1/2-We performed immunoblot analysis for MEK1/2 to assess their involvement in the signal transduction of A 3 receptor stimulation for inhibition of melanoma cell proliferation and of ERK1/2 phosphorylation levels. We found that MEK1/2 phosphoryla- tion was significantly inhibited in response to 5 and 10 M Cl-IB-MECA (Fig. 9). The inhibition of MEK1/2 phosphorylation via Cl-IB-MECA was abrogated by the A 3 receptor antagonist MRE 3008F20 (1 M), by the PI3K inhibitor LY294002 (10 M), and by the Akt inhibitor (5 M) (Fig. 9). DISCUSSION Recently, the importance of small autacoids, primarily thought to be neurotransmitters, in mediating a variety of biological activities in the skin (9,(43)(44)(45) has been demonstrated. In particular, several studies have indicated that adenosine, via stimulation of its receptors, is involved in cell proliferation and cell death. In agreement with these studies, we also found that A 3 receptor inhibits human melanoma A375 cell line proliferation.
Because it has been reported that the activity of Akt or MAPK or both is elevated in many cancer cells and is known to play critically important roles in cellular proliferation (46 -49), we tested the hypothesis that A 3 receptor stimulation regulates cell growth signaling via the ERK1/2 and/or Akt pathway in melanoma cells.
We found that serum-deprived A375 melanoma cells had no basal Akt phosphorylation, whereas Cl-IB-MECA treatment resulted in the phosphorylation of Akt at the Ser 573 phosphorylation site. Furthermore, Akt phosphorylation matched the phosphorylation of Raf at an inhibitory site (Ser 259 ). Surprisingly, serum-deprived A375 cells showed high basal levels of ERK1/2 phosphorylation (Fig. 2). The high levels of ERK1/2  (n ϭ 3). Cell lysates were incubated with glutathione S-transferase-Raf-1 RBD agarose-conjugated beads. Bound proteins were denaturated by boiling in reducing sample buffer and then resolved by SDS-PAGE followed by Western blotting analysis to detect Ras-GTP using an anti-Ras monoclonal antibody. phosphorylation in unstimulated A375 cells may reflect a neurospecific (melanocyte precursor cells derive from the neural crest) characteristic, since ERK1/2 is not usually phosphorylated after long periods of serum deprivation in cells of muscular and adipose origin (50,51). Interestingly, Cl-IB-MECA stimulation resulted in a time-and dose-dependent reduction in ERK1/2 phosphorylation (Fig. 2). It is suggested that this mechanism may be peculiar for melanoma cells, having a misregulation of proliferative pathways, since A 3 receptors increased ERK1/2 phosphorylation in CHO-A 3 cells in a dose-dependent manner (20) and induced a biphasic effect on the phosphorylation levels of ERK1/2 on microglia cells (24). Further studies in other different cell systems will enhance our understanding of the role of A 3 receptors in the modulation of mitogenic signaling.
There are two possible explanations for the Cl-IB-MECAinduced inhibition of ERK1/2 phosphorylation, either inhibition of ERK1/2 kinase or induction of a ERK1/2 phosphatase.
The possibility that an ERK1/2 kinase is inhibited has been examined by Rommel et al. (26). Recently, it was reported that the phosphorylation of Raf by Akt inhibits the activation of the ERK1/2 signaling pathway, suggesting the presence of crosstalk between the two signaling pathways (52). Furthermore, it has been shown that the Raf-MEK-ERK pathway can be inhibited by Akt in differentiated myotubes but not in their undifferentiated myoblast precursors. The authors suggested that regulation of a Raf/Akt interaction, underlying the ERK1/2 inhibition, might be mediated by stage-specific modification of these proteins or by stage-specific accessory proteins. This regulation might be intact in cells of neuronal origin also. To this end, we examined whether any cross-talk exists between ERK1/2 and Akt pathways in A375 melanoma cells. The clas-sical MAPK cascade leads from the Ras kinases to the MAPK kinase MEK1/2. There is evidence that Akt is able to phosphorylate Raf, thereby efficiently abrogating Raf activity on downstream substrates (27,28,52). We studied the effects of A 3 receptor stimulation on the proliferation of melanoma cells in the presence of specific inhibitors of the PI3K and Akt signal transduction pathways. We could effectively block the Cl-IB-MECA-induced reduction of ERK1/2 phosphorylation with an inhibitor of PI3K (Fig. 4). Indeed, application of Cl-IB-MECA in combination with PI3K inhibition resulted in a clear increase of ERK1/2 phosphorylation when compared with P-ERK1/2 in the presence of Cl-IB-MECA alone. These data suggest that the Ras-Raf-MEK-ERK pathway is normally activated by A 3 receptor stimulation, as is the PI3K-Akt route. It is clear that these apparently separate routes should actually interact.
In order to investigate the functionality of A 3 receptors expressed in melanoma cells, we used the selective adenosine analogue Cl-IB-MECA. It is not clear whether the growthinhibitory action of micromolar concentrations of the A 3 receptor agonist Cl-IB-MECA is due to its role as an extracellular ligand for cell surface receptors or whether it acts intracellularly as a second messenger. In particular, this agonist may, in high concentrations, activate A 1 receptors, which, however, when compared with A 3 receptors, are expressed at a low level in A375 cells (B max ϭ 23 Ϯ 7 and B max ϭ 291 Ϯ 50 fmol mg Ϫ1 of protein for A 1 and A 3 , respectively) (36). Thus, the effects we report here on melanoma cell proliferation and on ERK1/2 phosphorylation induced by Cl-IB-MECA using high (10 Ϫ6 M) concentrations of Cl-IB-MECA are almost certainly due to A 3 receptor stimulation. In particular, the effects of Cl-IB-MECA on cell proliferation and on ERK1/2 phosphorylation are not mediated by A 1 , A 2A , or A 2B receptors. In support of this conclusion, DPCPX, SCH 58261, and MRE 2029F20, adenosine receptor antagonists highly selective for A 1 , A 2A , and A 2B receptors, respectively, did not block the inhibitory effect of A 3 receptor stimulation on cell proliferation and on P-ERK1/2 modulation. On the contrary, the effects on cell proliferation and on P-ERK1/2 modulation were inhibited by the A 3 receptor antagonist, MRE 3008F20. Furthermore, the Cl-IB-MECA-induced effects on cell proliferation and ERK1/2 phosphorylation in human A375 melanoma cells were abolished in cells in which A 3 receptor protein was knocked down by si-RNA A3 treatment when compared with wild-type cells. These findings, together with the specificity of the agonist used, make us confident that the effects are due to the A 3 receptor subtype.
A 3 receptor stimulation inhibits the proliferation of melanoma cells partly by a PLC-sensitive mechanism. Pretreatment of cells with a PLC-␥ inhibitor strongly abrogated the Cl-IB-MECA effect on cell proliferation and on ERK1/2 phosphorylation, suggesting a critical role for PLC-␥ in A 3 receptor signaling. Furthermore, pretreatment of A375 cells with a PI3K inhibitor and an Akt inhibitor impaired Cl-IB-MECA-induced inhibition of cell proliferation and the effects of A 3 receptor stimulation on Raf, MEK1/2, and ERK1/2 phosphorylation. These data suggest that the A 3 adenosine receptor signals through a pathway including PI3K-Akt. On the contrary, Ras was not activated, at least when measured with the pull-down assay. These results confirm the hypothesis of this study; in A375 cells, A 3 receptors decrease MEK1/2-ERK1/2 phosphorylation and cell proliferation via the inhibition of Raf by a PI3K-Akt pathway without affecting Ras.
Our results indicate that Cl-IB-MECA acts extracellularly as a first messenger for cell surface receptors. In A375 cells, the inhibition of PLC-PI3K-Akt signaling is able to block the effect of A 3 receptor stimulation on cell proliferation, suggesting that in the melanoma cell system, an inhibitory connection between , respectively) before determination of MEK1/2 phosphorylation. LY294002 and Akt inhibitor are shown in lanes 7-9 and 10 -12, respectively. Cellular extracts were prepared and subjected to immunoblot assay using a P-MEK1/2 antibody. The blot was then stripped and used to determine total MEK expression using an anti-MEK1/2. Total MEK1/2 shows equal loading protein. MRE 3008F20, PI3K inhibitor LY 294002, and Akt inhibitor were added 30 min before the addition of Cl-IB-MECA. Cells were harvested after 30 min of treatment. B, the immunoblot signals were quantified using Molecular Analyst/PC densitometry software (Bio-Rad). Densitometric analysis of the MEK1/2 phosphorylated isoform is reported. Data were normalized; the unstimulated control (lane 1) was set to 100%. The error bars give S.E. of three independent experiments. *, p Ͻ 0.05 with respect to untreated cells; analysis was by ANOVA followed by Dunnett's test.
PLC-PI3K-Akt and ERK1/2 is present. Finally, we have described the molecular mechanism sustained by Cl-IB-MECA interfering with cell proliferation. Cl-IB-MECA via A 3 adenosine receptor binding activates a PLC-PI3K-Akt signaling that in turn reduces P-ERK1/2 levels necessary for cell proliferation. As a consequence, cells accumulated in G 0 /G 1 cell cycle phases, and a low level of DNA incorporation was observed.
Further definition of the pathways leading to ERK1/2 inactivation and translation into an in vivo model are required to clarify whether adenosine signaling in vivo has characteristics similar to those observed in this in vitro model. In this scenario, the regulation of ERK1/2 by A 3 receptor and PI3K would represent an important aspect of adenosine signaling.