Activation of Protein Kinase A and Atypical Protein Kinase C by A2A Adenosine Receptors Antagonizes Apoptosis Due to Serum Deprivation in PC12 Cells*

We found in the present study that stimulation of A2A adenosine receptors (A2A-R) prevents apoptosis in PC12 cells. This A2A-protective effect was blocked by protein kinase A (PKA) inhibitors and was not observed in a PKA-deficient PC12 variant. Stimulation of PKA also prevented apoptosis, suggesting that PKA is required for the protective effect of A2A-R. A general PKC inhibitor, but not down-regulation of conventional and novel PKCs, readily blocked the protective effect of A2A-R stimulation and PKA activation, suggesting that atypical PKCs (aPKCs) serve a critical role downstream of PKA. Consistent with this hypothesis, stimulation of A2A-R or PKA enhanced nuclear aPKC activity. In addition, the A2A-protective effect was blocked by a specific inhibitor of one aPKC, PKCζ, whereas overexpression of a dominant-positive PKCζ enhanced survival. In contrast, inhibitors of MAP kinase and phosphatidylinositol 3-kinase did not modulate the A2A-protective effect. Dominant-negative Akt also did not alter the A2A-protective effect, whereas it significantly reduced the protective action of nerve growth factor. Collectively, these data suggest that aPKCs can function downstream of PKA to mediate the A2A-R-promoted survival of PC12 cells. Furthermore, the results indicate that different extracellular stimuli can employ distinct signaling pathways to protect against apoptosis induced by the same insult.

Adenosine, which is released from metabolically active cells by facilitated diffusion or is generated extracellularly by degradation of released ATP, is a potent biological mediator (1). It is well known that adenosine modulates the activity of numerous cell types including various neuronal populations, platelets, neutrophils, and smooth muscle cells (1). To date, four adenosine receptors (A 1 , A 2A , A 2B , and A 3 ) have been identified. These receptors all contain seven transmembrane domains and belong to the G protein-coupled receptor (GPCR) 1 family (2).
We previously cloned the cDNA and the gene for the rat A 2A adenosine receptor (A 2A -R; see Refs. 3 and 4). In the central nervous system, the rat A 2A -R gene is heavily expressed by striatal neurons and colocalizes with the D 2 dopamine receptor in GABAergic striopallidal neurons (5). Low level A 2A -R expression is also observed in the cortex, hippocampus, cerebellum, and other areas of the brain (6). Importantly, A 2A -R has been regarded as a potential therapeutic target in protecting against neurodegeneration (e.g. Parkinson's disease and Huntington's disease; see Refs. 7 and 8) and neuronal trauma (e.g. hypoxia/ischemia; see Ref. 9). Moreover, stimulation of A 2A -R delays apoptosis in human neutrophils (10) and protects the hippocampus from excitotoxicity in a model of kainate-induced neuronal cell death (11). The molecular mechanisms underlying the protective effect of adenosine acting at A 2A -R remain largely uncharacterized in neuronal cells.
The rat pheochromocytoma cell line PC12 displays phenotypic traits associated with both adrenal chromaffin cells and sympathetic neurons and is a useful model for studying the actions of neurotrophic factors and neurotransmitters. In the past decade, this cell line has also served as a popular model system for studying the functions of various survival factors, including nerve growth factor (NGF; see Ref. 12). We previously demonstrated that stimulation of A 2A -R increases intracellular cAMP formation and activates novel protein kinase C isozymes in PC12 cells (13,14). In the present study, we found that activation of A 2A -R prevents apoptosis in serum-deprived PC12 cells. This protective mechanism involves transient enhancement of PKA activity and subsequent activation of atypical PKCs (aPKCs) and requires the activity of a serine-threonine protein phosphatase (PPase).
Cell Culture-PC12 cells were maintained in DMEM supplemented with 10% v/v horse serum and 5% v/v fetal bovine serum. A123, a cAMP-dependent protein kinase (PKA)-deficient variant of PC12 cells (15), was kindly provided by Dr. J. A. Wagner (Cornell University Medical College, New York). A123 cells were maintained in DMEM supplemented with 5% v/v horse serum and 10% v/v fetal bovine serum. Novel PKC-dominant-negative PC12 variants (16) were maintained in DMEM supplemented with 10% v/v horse serum, 5% v/v fetal bovine serum, and G418 (50 g/ml).
DNA Fragmentation-Cells were plated at the density of 3 ϫ 10 6 cells per 100-mm plate. After 24 h, cells were treated with the indicated reagent(s) for another 24 h and harvested by centrifugation, resuspended in 100 l of lysis buffer (10 mM EDTA, 50 mM Tris-HCl, 0.5% Sarkosyl, and 0.5 mg/ml proteinase K), and incubated at 50°C for 3 h. RNase (2 mg/ml) was added to the lysate for another 15 h. The lysate was extracted with 200 l of phenol/chloroform and then centrifuged again for 5 min. DNA fragments present in the supernatant were separated using a 2% agarose gel.
MTT Assay-Cells grown on 150-mm plates were washed twice with phosphate-buffered saline (PBS) and resuspended in DMEM. The resuspended cells were plated on 96-well plates (1.5 ϫ 10 4 cells/well) and treated with the indicated reagent(s) for 24 h. MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] was then added to the medium (1 mg/ml), and cells were incubated at 37°C for 3 h. Me 2 SO (100 l) was then applied to the medium to dissolve the formazan crystal derived from mitochondrial cleavage of the tetrazolium ring of MTT. The absorbency at 570 nm in each well was measured on a micro-enzyme-linked immunosorbent assay plate reader. None of the reagents used in this study interfered with the MTT values.
PKC Activity Assay-PKC activity was measured as described previously (14) with slight modifications. To measure atypical PKC activity, PC12 cells were first treated with PDD (1 M) for 20 h to downregulate conventional and novel PKCs. Cells were then washed with twice DMEM and incubated with the indicated reagents for the desired period of time. Different fractions of cells were then collected as described below. PKC activity was measured in a 40-l reaction containing 136 mM NaCl, 5.4 mM KCl, 0.3 mM Na 2 HPO 4 , 0.3 mM KH 2 PO 4 , 10 mM Mg 2 SO 4 , 25 mM ␤-glycerophosphate, 5 mM EGTA, 2.5 mM CaCl 2 , 1 mM glucose, 0.5% Triton X-100, and 25 mM HEPES, pH 7.2. Reactions were started by adding 150 M of substrate (⑀ peptide, Upstate Biotechnology Inc.) and 100 M of [␥-32 P]ATP (2 Ci/mmol). After incubation for 15 min at 30°C, the reaction was terminated by adding 10 l of 25% (w/v) trichloroacetic acid. The samples were centrifuged at 7,500 ϫ g for 10 min. The supernatants were then spotted on 2 ϫ 2-cm phosphocellulose squares (Whatman P-81), washed three times using 75 mM phosphoric acid, and once using 75 mM sodium phosphate, pH 7.5. Radioactivity retained on the P-81 papers were measured by scintillation counting. PKC activity was assayed as described above except that a PKC-specific pseudosubstrate peptide (sequence 113-129; SIYRR-GARRWRK-LYRAN) was added during the assay to block the PKC activity. PKC activity in PC12 cells was determined as the difference between the PKC activity assayed in the absence and in the presence of 300 M PKC-specific pseudosubstrate peptide (17). PKC activity increased linearly for up to 30 min using up to 30 g of protein.
Isolation of Membrane, Cytosol, and Nuclear Fractions-Membrane, cytosol, and nuclear fractions were isolated as described by Zhou et al. (18). Briefly, PC12 cells were collected by centrifugation (1000 ϫ g, 2 min), resuspended, and incubated in 1 ml of PKC sonication buffer (2 mM Tris, pH 7.6, 50 mM 2-mercaptoethanol, 1 mM EDTA, 1 mM EGTA, 1 mM PMSF, 100 M leupeptin, 10 M aprotinin, and 1 mM NaF) at room temperature for 2 min, and chilled on ice for 5 min. Nonidet P-40 was added to 1% (v/v) final concentration. Samples were forced once through a 20-gauge needle and then MgCl 2 was added to 5 mM. The samples were centrifuged at 600 ϫ g for 5 min to collect the nuclear fractions in the pellets. The supernatants were collected as the non-nuclear fraction or were further centrifuged at 100,000 ϫ g for 45 min to separate the cytosol and membrane fractions. The pellets (i.e. the membrane fractions) were resuspended in 300 l of PKC sonication buffer containing 0.1% Triton X-100. The nuclear fractions were resuspended in 150 l of buffer C (20 mM HEPES, pH 8, 425 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM dithiothreitol, 80 g/ml PMSF, 1 mM NaVO 4 , 20 mM NaF, 100 nM okadaic acid, and 25% glycerol) and incubated for 30 min on ice. The nuclear samples were centrifuged at 3,000 rpm for 10 min at 4°C to collect nuclear extracts in the supernatants. Protein concentrations were measured using the Bio-Rad Protein Assay Dye Reagent.
Western Blot Analysis-PC12 cells were rinsed with ice-cold PBS and lysed in ice-cold lysis buffer (20 mM HEPES, 1 mM dithiothreitol, 20 mM EGTA, 10% glycerol, 50 mM ␤-glycerophosphate, 10 mM NaF, 1% Triton X-100, 1 mM PMSF, 1 mM Na 3 VO 4 , 2 M aprotinin, 100 M leupeptin, 2 M pepstatin, and 0.5 M OKA). Cell debris was removed by centrifugation at 7,500 ϫ g for 10 min. The supernatant was utilized for the Western blot analysis. Protein concentrations were determined using the Bio-Rad Protein Assay Dye Reagent. Equal amounts of sample were separated by SDS-polyacrylamide gel electrophoresis using 10% polyacrylamide gels. The resolved proteins were then electroblotted onto Immobilon polyvinylidene difluoride membranes (Millipore, Bedford, MA). Membranes were blocked with 1% bovine serum albumin and incubated with the desired primary antibody for 1 h at room temperature, followed by the corresponding secondary antibody for 1 h at room temperature. Blots were then washed and immunoreactive bands were detected by enhanced chemiluminescence (Pierce) and recorded using Kodak XAR-5 film.
Transfection and Cell Viability Determinations-All plasmids used in transient transfection experiments were prepared by CsCl purification. Cells were transfected using Tfx TM (Promega), following the manufacturer's protocol, and then harvested between 48 and 72 h posttransfection. Transfection efficiency was typically between 10 and 15%. For survival analyses, cells were transiently transfected with a control vector or with vectors encoding the gene of interest along with oneseventh of the molar amount of an expression construct (pEGFP, CLONTECH; Palo Alto, CA) encoding green fluorescent protein (GFP), as indicated. Two days post-transfection, cells were subjected to serum deprivation for 24 h. Transfected cells were identified by GFP expression. Survival was determined as percentage of GFP-expressing cells by counting GFP-expressing cells in photomicrographs taken using a fluorescent microscope and normalized to the total number of cells counted from the corresponding photomicrographs of cells examined by phase contrast. An average of 1500 to 4000 total cells was counted for each experimental condition. Alternatively, GFP-expressing cells were quantified by flow cytometry as indicated below. These two methods produced similar results for the percentage of GFP-expressing cells. The survival index of 100% is designated as the percentage of GFP-expressing cells transfected with a control vector under the indicated treatment conditions.
Flow Cytometry-PC12 cells were transiently transfected with the indicated plasmid construct plus one-seventh the amount of GFP vector. Forty eight hours post-transfection, serum was withdrawn for 24 h, and cells were analyzed for the expression of GFP by gently removing the cells from plates with PBS containing trypsin (0.15%) and EDTA (0.53 mM) and then analyzing cell samples by flow cytometry with a Becton Dickinson FACScan. The transfected cells were identified by the expression of GFP that was detected using the FL-1 channel (excitation, 488 nm; emission, 530/30 nm). For each transfected plasmid, 30,000 cells were analyzed.
Statistics-Unless indicated otherwise, results were analyzed by one-way analyses of variance. Differences between means were assessed by the Student-Newman-Keuls method and were considered significant where p Ͻ 0.05.

Effects of an A 2A -R-selective Agonist, CGS 21680, on Serum-
deprived Apoptosis-Serum deprivation for 24 h resulted in significant DNA fragmentation in PC12 cells (Fig. 1A). Serum deprivation also decreased cell survival, as measured by the MTT assay (Fig. 1C). Addition of an A 2A -R-selective agonist, CGS21680 (CGS, 0.1 M), reversed the DNA fragmentation and cell death induced by serum deprivation. Addition of CGS (0.1 M) also significantly reduced phosphorylation of the stressactivated kinases JNK1 and JNK2 (Fig. 1B), which are implicated in apoptosis (12). Such protection by A 2A -R required new protein synthesis, because a protein synthesis inhibitor blocked the prevention of apoptosis by activation of A 2A -Rs in a dosedependent manner (Fig. 1C).

The Role of PKA in the A 2A -protective Effect in PC12 Cells-
Since activation of A 2A -R led to a transient increase in cAMP in PC12 cells (13), we first examined whether PKA plays an important role in preventing apoptosis due to serum deprivation. As shown in Fig. 2A, two PKA inhibitors (H-89 and KT-5720) reduced the protective effect of CGS and forskolin (FK) in serum-deprived apoptosis. In addition, CGS and FK exerted no effect on serum-deprived apoptosis in a PKA-deficient PC12 variant (A123, Fig. 2B), further supporting our hypothesis that PKA is critical for the protective effect of A 2A -R against apoptosis. Furthermore, the effect of CGS was markedly reduced by an A 2A -R-selective antagonist, 8-(3-chlorostyryl)caffeine (CSC) ( Fig. 2A). Thus, the effect of CGS is mediated by A 2A -Rs.
Phosphatases and PKC Mediate A 2A -R-evoked Protection-Because stimulation of A 2A -R activates a serine/threonine PPase in neutrophils (19), we examined whether a PPase was involved in prevention of apoptosis by A 2A -R activation. As shown in Fig. 3, two serine/threonine PPase inhibitors, okadaic acid (OKA) and calyculin A (Caly A), blocked the protective effect of A 2A -R activation in a dose-dependent manner. Maximal inhibition of the A 2A -protective effect by Caly A and OKA occurred at 1 and 10 nM, respectively. Caly A and OKA at these concentrations also markedly reduced the protective effect of FK (Fig. 3). Caly A (1 nM) or OKA (10 nM) alone did not markedly affect the survival of PC12 cells in the absence or presence of serum. Thus, a serine-threonine PPase appears to act downstream of PKA to facilitate survival of serum-starved cells upon A 2A -R stimulation.
Activation of A 2A -R has been shown to stimulate the ERK/ MAPK pathway in several cell types, including PC12 cells (20). Therefore we examined whether this pathway is involved in the protection against apoptosis by A 2A -R activation. As shown in Fig. 4A, treatment with CGS or FK increased phosphorylation of ERK1/ERK2 without altering protein levels. A MAPK kinase inhibitor (PD98059) blocked the FK-and CGS-mediated activation of ERK. However, PD98059 did not prevent the protective effect of CGS in serum-deprived cells (Fig. 4B). Therefore, activation of ERK is not required for A 2A -R-mediated protection against apoptosis. This finding is consistent with the observation that ERK is not important for cAMP-or NGF-mediated survival of primary sympathetic neurons (21).
We previously showed that stimulation of A 2A -Rs activates novel PKCs (14). Therefore, we used a PKC inhibitor bisindolylmaleimide I-HCl (BiM) to examine whether PKC is involved in the protective effect of A 2A -R. As shown in Table I, BiM markedly reduced the protective effect of CGS and FK. PKC therefore might be involved in A 2A -R-mediated protection and exert its effect downstream of PKA.
Atypical PKCs Mediate A 2A -R Prevention of Serum-deprived Apoptosis-PKC is a family of serine/threonine protein kinases that is composed of three subfamilies as follows: conventional, novel, and atypical. We previously demonstrated that two novel PKC isozymes (␦ and ⑀) play significant roles in the desensitization of A 2A -R-induced cAMP formation in PC12 cells (14). Moreover, two atypical PKC isozymes (aPKCs; / and ) and two conventional PKC isozymes (␣ and ␥) were also observed in our line of PC12 cells (Fig. 5, B and C). To identify the PKC isozymes involved in the protective effect of A 2A -R, we treated PC12 cells with a PKC-stimulating phorbol ester, PDD (100 nM), for 20 h to induce proteolysis and down-regulation of conventional and novel PKCs. This treatment caused downregulation of the conventional PKCs (␣ and ␥; Fig. 5C) and novel PKCs (␦ and ⑀; see Ref. 14). Because aPKCs ( and and ) are insensitive to diacylglycerols and phorbol esters, long term PDD treatment did not decrease levels of aPKCs in PC12 cells (Fig. 5B). Most interestingly, long term PDD treatment did not alter the response to A 2A -R stimulation (Fig. 5A). Moreover, as shown in Table II, stimulation of A 2A -R using CGS exerted a similar protective effect in PC12 variants expressing dominant-negative fragments of PKC⑀ or PKC␦. Taken together, these data strongly suggest that conventional and novel PKCs are not involved in A 2A -R-mediated protection against apoptosis in PC12 cells.
We next considered whether aPKCs are important for the A 2 -R-mediated protection by measuring aPKC activity after A 2A -R stimulation in PC12 cells treated with PDD for 20 h. In these cells, CGS enhanced aPKC activity in both nuclear and non-nuclear fractions (Fig. 6, A and B). The increase in aPKC activity was much greater in the nuclear fraction as compared with the non-nuclear fraction. Western blot analysis showed that PKC and PKC/ immunoreactivities were markedly increased in the nuclear fraction following treatment with CGS (Fig. 6C). FK also enhanced aPKC activity in the nuclear fraction with a time course comparable to that observed with CGS (Fig. 7). These results suggest that PKA mediates increases in nuclear aPKCs during A 2A -R stimulation.
PKC has been implicated in survival following serum deprivation in PC12 cells (22). Therefore, we used a cell-permeable   (Fig. 8B). Comparable results were obtained when control cells transfected with empty vector and the GFP-expressing construct were treated with CGS. Stimulation of A 2A -R using CGS enhanced the number of GFP-expressing cells by 51 Ϯ 20% (mean Ϯ S.E.; p Ͻ 0.05, Student's t test, seven independent experiments). This observation is consistent with the protective effect of A 2A -R assessed by the MTT assay (Fig. 1C). These findings indicate that PKC specifically regulates A 2A -R-mediated survival in PC12 cells.
To examine whether stimulation of A 2A -Rs regulates PKC, and whether this regulation is PKA-dependent, we next determined the activity of PKC under our experimental conditions. We found that nuclear PKC activity was increased by CGS (Fig. 6) and FK (Fig. 7). Moreover, transient overexpression of PKC ϩ enhanced the survival of serum-deprived A123 cells (a PKA-deficient PC12 variant) (Fig. 8C). Collectively, these results suggest that the protective effect of A 2A -R stimulation in serum-deprived PC12 cells requires the sequential activation of PKA and PKC.

The PI3K/Akt Pathway Is Not Involved in the A 2A -R-mediated Protection against Apoptosis in Serum-deprived PC12
Cells-Data from the above experiments suggest that PKC mediates the A 2A -protective effect in PC12 cells. Interestingly, the myristoylated PKC pseudosubstrate inhibitor also partially suppressed the protective effect of NGF in serum-deprived PC12 cells (Fig. 8A). Thus, distinct anti-apoptotic signals may converge on PKC in PC12 cells. Because phosphatidylinositol 3-kinase (PI3K) has been implicated in the NGF-induced translocation of PKC to the nucleus (24) and in suppression of apoptosis (25), we next investigated if the PI3K pathway was involved in the protective effect of A 2A -R stimulation. Although a PI3K inhibitor (LY294002) abolished the protective effect of NGF against apoptosis in a dose-dependent manner, it did not reduce CGS-mediated survival (Fig. 9A).
We next examined whether A 2A -R stimulation activates Akt, one of the downstream targets of PI3K implicated in cell survival. We examined the phosphorylation levels of the two activating residues (26), threonine 308 and serine 473 of Akt, by using Western blot analysis (Fig. 9, B and C). Our data show that NGF markedly increases Akt phosphorylation, whereas stimulation of A 2A -R does not. We next employed a kinase-dead Akt (dnAkt) that contains a point mutation in its catalytic domain (K179M; see Ref. 27) to determine whether Akt is involved in the A 2A -R-protective effect. This K179M-Akt mutant is unable to transmit a signal downstream and effectively decreases upstream signals mediated by 3Ј-phosphorylated phosphoinositides and PDK1, which activate Akt (28). Overexpression of dnAkt reduced NGF-mediated cell survival during serum deprivation but did not alter CGS-mediated survival (Fig. 9D). Collectively, these results indicate that although the PI3K/Akt pathway plays a role in NGF-mediated survival, it is not activated by A 2A -R stimulation nor is it required for the protective effect of A 2A -R activation in serum-starved PC12 cells.

DISCUSSION
In the present study, we found that A 2A -R activation protects PC12 cells from apoptosis induced by serum deprivation. This protective effect requires PKA activation, since it is blocked by two different PKA inhibitors (H-89 and KT 5720), and is absent in PKA-deficient PC12 cells. In contrast to NGF-mediated survival, A 2A -mediated protection does not require activation of PI3K or Akt. Although MAPK was activated by stimulation of A 2A -R, blocking the MAPK pathway did not alter A 2A -R-mediated protection against apoptosis. Compared with the well characterized mechanisms involving ERK/MAPKs and PI3K underlying NGF-mediated survival in PC12 cells, the results of the present study demonstrate that A 2A -R utilizes a distinct set of signaling pathways to activate key downstream mediators (e.g. PKC) of the anti-apoptotic processes (Fig. 10).
In PC12 cells, cyclic AMP has been implicated in mitosis, apoptosis, and differentiation. Long lasting elevation of cellular cAMP, either by treatment with cAMP analogs or FK, rescues PC12 cells from cell death (29). Treatment of PC12 with cAMP analogs also eventually leads to neuronal differentiation, but this requires exposure to these agents for several days (30). Although activation of A 2A -R increases cAMP, this response is transient (13) and does not stimulate neural differentiation of PC12 cells. 2 Results in the present study suggest that transient activation of the cAMP/PKA pathway is sufficient to protect against cell death due to serum deprivation.
Our data also suggest that downstream of PKA, aPKCs me-diate the protective effect of A 2A -R against apoptosis in PC12 cells. This conclusion is based on the following evidence. First, the protective effect of A 2A -R can be reversed by a general PKC inhibitor (BiM , Table I) and by a PKC-specific inhibitor (Fig.  8A), but not by down-regulation of conventional and novel PKCs ( Fig. 5 and Table II). BiM is widely used as a selective inhibitor of PKCs. Since BiM acts as a competitive inhibitor of ATP binding, it may also inhibit PKA at high concentrations (31). Although the in vitro K i value of BiM for PKC is 10 nM, the effective concentration for inhibiting PKC in cultured cells appears to be higher. For example, in 3T3 fibroblasts, maximal inhibition of PKC-dependent phosphorylation by BiM occurs at 5 M, which is a concentration that does not inhibit PKAmediated phosphorylation in those cells (31). In addition, in rat basophilic leukemia (RBL-2H3) cells, 10 M BiM blocks PKCbut not PKA-evoked phosphorylation of phospholipase C (32). BiM has therefore been routinely used to block PKC-mediated responses at concentrations ranging from 1 to 10 M in studies employing various types of cultured cells (33,34). Moreover, in our study, we demonstrated that in addition to BiM, a PKCspecific pseudosubstrate peptide inhibitor also blocked the protective effect of A 2A -R, thus confirming the involvement of PKC in this process (Fig. 8A).
The second line of evidence supporting a role for aPKCs in A 2A R-mediated survival comes from our observation that stimulation of A 2A -R increased nuclear aPKC activity and the amount of nuclear PKC and / immunoreactivity in PC12 cells (Fig. 6). This process appears to be mediated by PKA, since FK also increases nuclear aPKC activity (Fig. 7). Increased nuclear aPKC may be important for antagonizing apoptosis since it has been observed following treatment with other mitogenic and differentiating factors that promote cell survival (24,35). Finally, overexpression of a dominant-positive PKC enhanced the survival of both wild-type and PKAdeficient PC12 cells (Fig. 8, B and C). These results indicate that aPKCs lie in a signal transduction pathway downstream of PKA that mediates that protective effect of A 2A -R stimulation and that at least one aPKC, PKC, is critical for the prevention of serum-deprived apoptosis in PC12 cells.
This protective effect of aPKCs is consistent with previous studies suggesting that aPKCs are potent anti-apoptotic kinases (36). For example, low concentrations of ceramide transiently activate PKC in conjunction with NF-B and promote survival of PC12 cells during serum deprivation (22). NGF, which promotes survival of PC12 cells, also increases the abundance of nuclear PKC in PC12 cells (24). Results in the present study demonstrate that the myristoylated PKC-selective inhibitor partially suppressed the protective effect of NGF (Fig.  9A), further suggesting that PKC is an important anti-apoptotic factor in PC12 cells. Since PKC has been implicated in the regulation of NF-B (37), NF-B might act downstream of PKC to inhibit apoptosis due to serum deprivation. PKC has been shown to protect human leukemia cells against druginduced apoptosis (38). We found that nuclear PKC is increased following A 2A -R stimulation, suggesting that PKC may also contribute to important nuclear events that protect against apoptosis in PC12 cells. Further studies are required to clarify the role of PKC/ in the protective effect of A 2A -R stimulation.
The involvement of the PI3K/Akt pathway in preventing apoptosis has been well established (25). Various GPCRs activate Akt in either a PI3K-dependent or -independent manner (39,40). In NGF-differentiated PC12 cells, PI3K plays a critical role in cell survival (41). We considered whether PI3K is important for A 2A -mediated survival, since PKC can be activated in a PI3K-dependent manner (42). However, we found that 2 C. H. Chen and Y. Chern, unpublished observations. whereas a PI3K inhibitor blocked the ability of NGF to promote survival upon serum deprivation, the protective effect of A 2A -R stimulation was PI3K-independent ( Fig. 9A and Fig. 10). In addition, Akt, a kinase important for NGF-mediated survival (43), was not involved in the protective effect of A 2A -R stimulation. These findings indicate that A 2A -R agonists and NGF utilize different signaling pathways to prevent apoptosis. A similar situation exists in sympathetic neurons, where the PI3K/Akt pathway plays a critical role in depolarization-mediated survival, but not in survival mediated by the cAMP/PKA pathway (44). Our results also indicate that in PC12 cells these two pathways converge on at least one common anti-apoptotic factor, PKC. The mechanism by which PKA activates PKC independent of PI3K requires further study.
Our findings also implicate serine/threonine protein phosphatases in the protective effect of A 2A -R stimulation. Activation of A 2A -R increases the activity of a serine/threonine PPase in neutrophils (19). In the present study, we found that two serine/threonine PPase inhibitors (Caly A and OKA) blocked the protective effect of CGS and FK at nM concentrations. Maximal inhibition was achieved using 1 nM of Caly A or 10 nM of OKA (Fig. 3). Caly A is a selective inhibitor of PP 1 and PP 2A (K i ϭ 0.5-2 nM). In contrast, OKA is a relatively specific inhibitor of PP 2A with K i values of 0.2 nM for PP 2A and 20 nM for PP 1 . Because 10 nM of OKA completely blocked A 2A -R-mediated protection against apoptosis (Fig. 3B), the PPase most likely involved in this process is a member of the PP 2A family. Several important proteins involved in apoptosis, such as IB kinase, MAP kinases, and cell cycle regulators are substrates of PP 2A (45). In a cell-free model of apoptosis, OKA suppresses caspase-3 activation and Akt cleavage, two key events in cell death (46). It remains to be determined whether stimulation of A 2A -R activates a PP 2A -like activity downstream of PKA that protects PC12 cells from apoptosis or whether basal PP 2A activity is merely necessary for A 2A -R-mediated survival.
A 2A -Rs are expressed in many areas of the brain (6) and in various peripheral tissues (47). Previous work has suggested a role for A 2A -Rs in protection against cell death. Kobayashi and Millhorn (9) reported that expression of A 2A -Rs is increased by hypoxia. Stimulation of A 2A -Rs in PC12 cells partially protects against cell death induced by hypoxia (9). In human neutrophils, stimulation of A 2A -Rs delays apoptosis, presumably via a PKA-dependent mechanism (10). Our present study demonstrates that multiple mechanisms downstream of PKA underlie the action of A 2A -Rs in preventing apoptosis of serum-deprived PC12 cells. Our findings provide the first clear evidence that PKC is a key downstream component of a PKA-dependent, anti-apoptotic signaling pathway activated by a GPCR. Further knowledge about protective mechanisms evoked by A 2A -R stimulation may help to facilitate the clinical application of A 2A -R agonists in the treatment of neurodegeneration-associated nervous system trauma and neurological disease.