Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Angiogenesis is the process by which blood vessels develop from the preexisting vasculature with endothelial cell proliferation being a crucial component of this event. Vascular endothelial growth factor (VEGF), an endothelial cell mitogen, is a potent stimulus for angiogenesis (Tischer et al., 1991; Ferrara and Davis-Smyth, 1997). VEGF also has been demonstrated to protect endothelial cells from experimentally induced apoptosis (Gerber et al., 1998). VEGF is secreted by several cell types, including vascular smooth muscle cells (Li et al., 1995; Pedram et al., 1997), cardiomyocytes (Levy et al., 1995a), and tumor cells (Ryuto et al., 1996; Ferrara and Davis-Smyth, 1997). Alternative splicing of a single gene results in the expression of multiple VEGF isoforms with VEGF₁₂₁ (121 amino acids) and VEGF₁₆₅ (165 amino acids) being the predominant secreted isoforms (Tischer et al., 1991; Houck et al., 1992).

The importance of VEGF in developmental neovascularization has been demonstrated in mice as genetic disruption of VEGF (Carmeliet et al., 1996) or its tyrosine kinase receptors flt-1 (VEGFR-1; Fong et al., 1995) and flk-1/KDR (VEGFR-2; Shalaby et al., 1996) results in embryonic lethality. Although the precise phenotype varies with the specific gene targeted for deletion, examination of these knockout embryos reveals abnormalities in blood vessel formation. In the adult, VEGF appears to have a role in the progression of pathologies that have an angiogenesis-dependent component, including diabetic retinopathy (Aiello et al., 1994) and cancer (Hanahan and Folkman, 1996; Ferrara and Davis-Smyth, 1997). In cancer progression, angiogenesis is necessary for primary tumor growth and may be required for metastasis to secondary sites (Hanahan and Folkman, 1996). VEGF, in addition to factors such as angiopoietin-2, appears necessary for vascular development of the tumor (Holash et al., 1999). Procedures that disrupt VEGF signaling such as administration of antibodies directed against VEGF (Kim et al., 1993) and expression of a dominant negative form of flk-1 (Millauer et al., 1994) retard tumor growth in animals. Conversely, therapeutically induced vascular development achieved through VEGF-mediated angiogenesis may be beneficial in pathologies of vascular insufficiency (Isner and Asahara, 1999).

Based on the role of VEGF in development and its detrimental or beneficial involvement in disease, regulation of VEGF expression is being extensively investigated. VEGF levels are elevated by multiple stimuli, including hypoxia (Levy et al., 1995a,b), growth factors (Ryuto et al., 1996; Nauck et al., 1997), and cytokines (Li et al., 1995; Ryuto et al., 1996). Depending on the stimulus, up-regulation of VEGF mRNA may occur at the transcriptional or post-transcriptional level. PC12 rat pheochromocytoma cells have been used as a model to study regulation of VEGF expression as well as that of other hypoxia-responsive genes. Levy et al. (1995b) described the hypoxic up-regulation of VEGF mRNA in PC12 cells and identified a 28-base pair (bp) element in the promoter of the VEGF gene that was responsible for the majority of this response. This region structurally and functionally resembles the hypoxia-inducible factor-binding site in other hypoxia-sensitive genes. Subsequently, this group identified discrete segments of the 3' untranslated region of the rat VEGF gene that are apparently involved in the hypoxia-induced stabilization of VEGF mRNA (Levy et al., 1996).

In that VEGF up-regulation occurs in response to hypoxia and that extracellular adenosine levels are frequently elevated during hypoxia, the effect of adenosine receptor (AR) activation on VEGF expression has been examined. These studies have described varying effects of adenosine or its analogs on VEGF expression in different cell types (Hashimoto et al., 1994; Fischer et al., 1995; Takagi et al., 1996; Kobayashi and Millhorn, 1999). In several of these reports, the adenosine receptor subtype mediating the response was not identified due to the existence of multiple receptor subtypes on the cell under examination and the use of nonselective concentrations of ligands. Signaling pathways and molecular mechanisms that link adenosine receptor activation to VEGF regulation have not been extensively examined. This study was undertaken to examine regulation of VEGF expression by adenosine receptor activation in PC12 rat pheochromocytoma cells. In addition to their usefulness in examining VEGF gene regulation, PC12 cells have been used to study adenosine receptor signal transduction and regulation of receptor expression (Lai et al., 1997; Kobayashi et al., 1998). In this study, it is observed that down-regulation of VEGF expression occurs specifically on activation of the A_2AAR. This down-regulation appears to involve signaling via a G_s-dependent pathway that ultimately results in a reduction in the VEGF gene transcription rate.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. NECA (5'-N-ethylcarboxamidoadenosine) and CGS21680 {2-[4-[(2-carboxyethyl)phenyl]ethylamino]-5'-N-ethylcarboxamido-adenosine} were purchased from Research Biochemicals International (Natick, MA). ZM241385 [4-(-2-[7-amino-2-{2-furyl}{1,2,4}triazolo{2,3-a}{1,3,5}triazin-5-yl-amino]ethyl)phenol] was kindly provided by Zeneca Pharmaceuticals (Macclesfield, UK). All radiochemicals were obtained from DuPont NEN (Boston, MA). Cell culture supplies, oligonucleotides, pertussis toxin, and cholera toxin were purchased from Life Technologies (Gaithersburg, MD).

Cell Culture and Treatment Conditions. PC12 rat pheochromocytoma cells originally obtained from the American Type Culture Collection (CRL-1721; Manassas, VA) were supplied by the Duke University Cell Culture facility (Durham, NC). PC12 cells were maintained in RPMI 1640 medium supplemented with 10% heat-inactivated equine serum/5% fetal bovine serum (FBS), 5 mM glutamine, and 100 U/ml penicillin + 100 µg/ml streptomycin, and were grown on collagen-coated 100-mm dishes. Cells were typically kept in a cell culture incubator maintained at 37°C with a humidified atmosphere of 5% CO₂, 20% O₂. Before analysis of VEGF mRNA levels, culture medium was replaced with fresh serum-containing medium 30 to 60 min before introduction of AR agonist or antagonist ligands dissolved in dimethyl sulfoxide (DMSO). Control cells were treated with the appropriate volume of DMSO. Conditions for treatment with bacterial toxins are provided in Results. For certain experiments, a hypoxic environment was obtained by placing cells in a modular incubation chamber (Billups-Rothenberg, Del Mar, CA) that was flushed with a 5% CO₂, 95% N₂ gas mixture for 20 min and then sealed and placed in the 37°C cell culture incubator for the indicated time period.

Northern Blotting. Total RNA was isolated from cells with Trizol reagent (Life Technologies). Approximately 20 to 25 µg of total RNA was electrophoresed on 1% agarose gels containing 2.2 M formaldehyde. RNA was transferred to Zeta-Probe nylon membranes (Bio-Rad, Richmond, CA) and cross-linked via UV-irradiation in a Stratalinker (Stratagene, La Jolla, CA). Prehybridization of membranes was conducted at 42°C for 4 h in a buffer consisting of 50% formamide, 5× SSC (1× = 150 mM sodium chloride and 15 mM sodium citrate), 1% SDS, 5× Denhardt's solution, and 200 µg/ml sheared salmon sperm DNA. Hybridization was conducted in the same buffer with the concentration of salmon sperm DNA reduced to 100 µg/ml and with a 600-bp fragment of murine VEGF₁₆₅ random prime labeled with [³²P]dCTP as probe. Hybridization was conducted at 42°C for 12 to 14 h. Membranes were sequentially washed in 2× SSC (30 min; 42°C); 2× SSC, 0.5% SDS (30 min; 42°C); 0.3× SSC, 0.5% SDS (30 min; 42°C); and finally in 0.3× SSC, 0.5% SDS (30 min; 55°C). After autoradiography, membranes were stripped and reprobed with a [³²P]ATP-labeled 24-bp oligonucleotide specific for 28S ribosomal RNA to assess gel loading and transfer. Signals were quantitated by laser densitometry (Bio-Rad model 620 densitometer) and VEGF mRNA levels normalized to those obtained for 28S rRNA. For CGS21680 concentration-response data, individual experiments were analyzed with a computer-modeling program. Data are presented as mean ± S.E.

RNase Protection Assays. Probes for RNase protection assays (RPAs) were generated as described by Levy et al. (1995a) in an analysis of VEGF isoform regulation in rat neonatal myocytes. Briefly, oligonucleotide primers complementary to rat VEGF exon 5 (5'-AGACCAAAGAAAGATAGAACAAAG-3') and exon 8 (5'-TAATACGACTCACTATAGGGAGGGGTGAGAGGTCTAGTTCCCGA-3' with 23 bases representing T7 promoter sequence) were used in a polymerase chain reaction reaction using reverse transcribed PC12 cell RNA as template. The amplified segments of VEGF sequence of 239 bp (probe A) and 107 bp (probe B) were isolated from agarose gels and their identity confirmed by sequencing. For normalization of RPAs, DdeI-digested pTRI-GAPDH (Ambion, Austin, TX) was used as template to generate a 125-bp construct that protected a 70-bp segment of rat GAPDH. Radioactively labeled RNA probes were prepared with [³²P]UTP and an in vitro transcription kit (MAXI Script; Ambion).

Western Blotting. To analyze expression of VEGF protein, approximately equal numbers of PC12 cells were maintained in 100-mm dishes in serum-free RPMI medium for 12 h. Cells were then refed with either 3.5 ml of serum-free RPMI or RPMI containing 10% equine serum/5% FBS with or without 1 µM CGS21680. Cells were maintained at 37°C for an additional 12 h at which time medium was collected and centrifuged to pellet any detached cells. Medium was then incubated with a heparin-agarose suspension (Sigma Chemical Co., St. Louis, MO) for 2 h at 4°C with gentle rotation. The amount of conditioned medium incubated with heparin-agarose was normalized according to the protein concentration of lysates prepared from each dish of cells. The heparin-agarose was washed twice with a buffer consisting of 20 mM Tris, pH 7.4, at 5°C and 400 mM NaCl. Proteins were eluted by boiling samples for 5 min in SDS-polyacrylamide gel electrophoresis sample buffer and aliquots were electrophoresed on 12% polyacrylamide gels that were run under denaturing conditions. Proteins were transferred onto Protrans (Schleicher & Schuell, Keene, NH) nitrocellulose membranes. After blocking, membranes were incubated at 4°C for 12 h with a 1:100 dilution of VEGF147 antisera (Santa Cruz Biotechnology, Santa Cruz, CA). Membranes were then incubated with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody and signals visualized by chemiluminescence detection (Renaissance reagent; DuPont NEN). The amount of VEGF was quantitated by laser densitometry.

Analysis of mRNA Stability. Control and 1 µM CGS21680-treated PC12 cells were maintained for 4.5 h. At that time, all cells were exposed to 5 µg/ml actinomycin D (Sigma Chemical Co.) and total RNA isolated for both treatment groups immediately and at 15, 30, 60, 120, and 240 min postactinomycin D addition. Northern blotting was then performed as described above.

Nuclear Run-On Assays. PC12 cells in complete RPMI were treated with 1 µM CGS21680 or DMSO for 6 h at which time nuclei were isolated after cell homogenization in an Nonidet P-40 lysis buffer (Ausubel et al., 1997). For transcription reactions, approximately equal numbers of nuclei (3-5 × 10⁷ nuclei/reaction) were incubated with 5.5 mM each of GTP, ATP, and CTP along with 150 µCi of [-³²P]UTP. After a 30-min incubation at 30°C, to each reaction tube was added RNase-free DNase (30 U final) and CaCl₂ (1 mM final) and the incubation extended for an additional 10 min at 30°C. Samples were then treated with proteinase K for 30 min at 37°C. Samples were processed by guanidine thiocyanate and phenol-chloroform extractions followed by ethanol precipitation of RNA. Samples were resuspended in DEPC-H₂0, heated at 95°C for 5 min and an equal number of counts per minute was hybridized to prepared membranes (Zeta-Probe). Denatured target DNAs cross-linked to the membranes were murine VEGF₁₆₅ in pBluescript, murine -actin in pBluescript, and empty pBluescript vector. All target DNAs had been linearized with EcoRI. Hybridization was conducted in the same buffer at 42°C for 12 to 16 h. Blots were washed sequentially in 2× SSC/0.1% SDS (42°C), 2× SSC/0.1% SDS (55°C), and 0.2× SSC/0.1% SDS (65°C) with 30 min for each wash. Signals were analyzed with a PhosphorImager with ImageQuant software. VEGF signals were normalized to those obtained for -actin after subtraction of any signal for pBluescript vector alone.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

To study the regulation of VEGF expression after AR activation in PC12 cells, VEGF mRNA levels were quantitated via Northern blotting. In initial experiments, VEGF mRNA could be detected in PC12 cells maintained in normoxia and the level of VEGF mRNA was substantially decreased after a 6-h treatment with either AR agonist CGS21680 or NECA. Treatment of PC12 cells with 1 µM CGS21680 and 1 µM NECA reduced VEGF mRNA to levels that were and 17.5 ± 3.0% (n = 4) and 17.1 ± 6.6% (n = 3) of untreated cells, respectively. In that NECA is a nonselective AR agonist and CGS21680 at a concentration of 1 µM may activate multiple AR subtypes, complete CGS21680 concentration-response experiments were performed to define the AR subtype involved in the response. As shown in Fig. 1A, CGS21680 induced a concentration-dependent reduction in VEGF mRNA levels with maximal response at ~10 nM and an EC₅₀ value of 0.47 ± 0.17 nM (n = 7). The potency of CGS21680 in VEGF mRNA down-regulation strongly suggests that activation of specifically the A_2AAR is responsible for this response. AR involvement in the decrease in VEGF mRNA expression also is indicated by the sensitivity of CGS21680-induced down-regulation to the AR antagonist ZM241385 (Poucher et al., 1995). In PC12 cells exposed to 0.1 µM CGS21680 for 6 h, the level of VEGF mRNA was 41.3 ± 3.8% of that observed in control cells (Fig. 1B). However, in cells pretreated for 10 min with 1 µM ZM241385 before introduction of 0.1 µM CGS21680, VEGF mRNA expression was found to be increased 2.2 ± 0.4-fold above basal (n = 7). Interestingly, in PC12 cells treated with 1 µM ZM241385 alone, the level of VEGF mRNA was elevated 2.8 ± 0.6-fold above that in control cells (n = 3).

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Pharmacologic characterization of AR-mediated down-regulation of VEGF mRNA. A, concentration dependence of VEGF mRNA down-regulation by CGS21680. PC12 cells were treated with the indicated concentrations of CGS21680 for 6 h before isolation of RNA. VEGF mRNA levels were determined by Northern blotting. Top, representative Northern blot with positions of 28S and 18S rRNA as indicated to the left. Bottom, same blot probed with an oligonucleotide specific for 28S rRNA. Also shown is the concentration-response curve for the CGS21680-mediated down-regulation of VEGF mRNA. Northern blots were quantitated and VEGF mRNA normalized to the 28S rRNA signal. Points represent the mean of seven experiments with bars representing the S.E.M. EC50 value for CGS21680 was 0.47 ± 0.17 nM. B, inhibition of CGS21680-mediated down-regulation of VEGF mRNA by ZM241385. PC12 cells were treated with 0.1 µM CGS21680 for 6 h with or without a 10-min pretreatment with 1 µM ZM241385 as indicated. Certain cells were treated with 1 µM ZM241385 only for 6 h. Top, representative Northern blot with positions of 28S and 18S rRNA indicated to the left. Bottom, same blot probed with an oligonucleotide specific for 28S rRNA. Also shown is the graphical representation of Northern blot data after quantitation of VEGF signals by laser densitometry and normalization to 28S rRNA signal. Data are the mean of seven (CGS21680 and ZM241385 + CGS21680) or three (ZM241385 alone) experiments with bars representing S.E.M.

CGS21680-induced down-regulation of VEGF mRNA occurred in a time-dependent fashion. In five experiments, 1 µM CGS21680 produced an ~40% decrease in VEGF mRNA levels at 2 h with maximal down-regulation (18.1 ± 6.4% of basal) observed at 6 h of treatment. VEGF mRNA remained decreased to a level that was 30.8 ± 10.2% that of basal throughout 20 h of exposure to CGS21680. The CGS21680-induced down-regulation of VEGF mRNA was at least partially reversible. In a separate set of experiments, treatment of PC12 cells with 1 µM CGS21680 for 6 h reduced VEGF mRNA levels to ~17% of that observed in untreated cells. When cells were washed extensively with PBS and refed with complete medium, VEGF mRNA returned to a level that was 44.8 ± 7.6% of that in untreated cells in 6 h.

As a result of the organization of the VEGF gene and alternative splicing, multiple isoforms of VEGF exist (Tischer et al., 1991; Houck et al., 1992) with possible differences among isoforms in bioavailability and certain biological activities (Ferrara, 1999). To determine whether the CGS21680-induced down-regulation of VEGF mRNA observed in Northern blotting analysis results from a differential regulation of distinct VEGF mRNA transcripts, RNase protection assays were performed. As described in Materials and Methods, two DNA fragments were amplified from reverse-transcribed PC12 cell RNA and used to detect individual VEGF transcripts. Hybridization of probe A to VEGF₁₆₅ results in protection of a 239-bp fragment and hybridization to VEGF₁₈₉ results in protection of 209- and 30-bp fragments due to the inclusion of a segment of exon 6 in VEGF₁₈₉ (Levy et al., 1995a). Probe B consists of a 107-bp fragment and provides full-length protection of VEGF₁₂₁ (Levy et al., 1995a). After an overnight incubation of PC12 cells in serum-free medium, reintroduction of serum produced a similar fold increase in these three major VEGF isoforms (Fig. 2). Inclusion of CGS21680 inhibited expression of transcripts for VEGF₁₆₅ and VEGF₁₈₉ to an identical degree, 0.5 ± 0.1-fold and 0.5 ± 0.2-fold of the basal value, respectively. The serum-induced induction of VEGF₁₂₁ also was completely abolished by CGS21680 (0.9 ± 0.2-fold of basal). Thus, the three major isoforms of VEGF are very similarly down-regulated after activation of the A_2AAR.

View larger version (37K): [in this window] [in a new window]   Fig. 2.   Regulation of multiple VEGF transcripts by CGS21680. Serum-starved PC12 cells were treated with fresh serum-free medium (serum-free) or medium containing 10% FBS and 5% equine serum without (+serum) or with 1 µM CGS21680 (+serum; +CGS21680). RNase protection assays were performed as described in Materials and Methods. Hybridization of probe A with VEGF165 results in protection of a 239-bp fragment, whereas hybridization with VEGF189 protects a 209-bp fragment. Hybridization of probe B with VEGF121 results in protection of a 107-bp band. A, representative RPA with probe A and probe B. Positions of protected fragments are shown. Analysis of mRNA for GAPDH was used for normalization. B, graphical representation of the mean ± S.E. of four (VEGF165 and VEGF189) and seven (VEGF121) experiments. Results are reported relative to cells maintained in serum-free medium.

The above-mentioned findings indicate that activation of the A_2AAR produces a marked down-regulation of constitutive VEGF mRNA expression in PC12 cells maintained in complete medium and normoxia. The basal level of VEGF expression in PC12 cells may represent that which exists in the presence of components of the cell culture medium such as various growth factors of which several are known to increase VEGF expression (Ryuto et al., 1996; Nauck et al., 1997). Indeed, in the RPAs, reintroduction of serum to cells maintained in a serum-deficient medium up-regulated VEGF mRNA. To determine the effects of CGS21680 on the level of VEGF mRNA induced by an additional physiologically relevant stimulus, PC12 cells that were maintained in the presence of complete medium were exposed to a hypoxic environment. The precise mechanisms contributing to hypoxic induction of VEGF mRNA probably differ from those responsible for the up-regulation of VEGF by growth factors and cytokines as distinct transcription factors may be involved. VEGF mRNA expression is increased in response to hypoxia due to both transcriptional activation involving a hypoxia-inducible factor-1 binding site in the VEGF gene promoter as well as enhanced mRNA stability (Levy et al., 1995b, 1996). As described in Materials and Methods, PC12 cells were maintained in hypoxia for 3 h with or without a 16-h pretreatment with 1 µM CGS21680. As shown in Fig. 3, hypoxia induced a 2.1 ± 0.1-fold (n = 6) increase in the level of VEGF mRNA compared with cells maintained in normoxia. In cells pretreated with 1 µM CGS21680 before initiation of hypoxia, the VEGF mRNA level was reduced by 85.7 ± 3.4% compared with untreated hypoxic cells.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   CGS21680-induced inhibition of hypoxic induction of VEGF mRNA. PC12 cells were exposed to either normoxia (control) or hypoxia for 3 h without (hypoxia) or with a 16-h pretreatment with 1 µM CGS21680 (CGS21680hypoxia) as described in Materials and Methods. VEGF mRNA was analyzed via Northern blotting. A, representative Northern blot probed for VEGF mRNA and 28S rRNA. B, graphical representation of Northern blot data after quantitation of VEGF mRNA by laser densitometry and normalization to 28S rRNA. The data are the mean of six experiments with bars representing S.E.M.

To determine whether the A_2AAR-mediated down-regulation of VEGF mRNA is reflected in a decrease in protein expression, conditioned medium was collected from PC12 cells that had been incubated in serum-free medium and then exposed to serum-containing growth medium either in the absence or presence of 1 µM CGS21680. As shown in Fig. 4, reintroduction of serum markedly increased the secretion of a protein that was detected by antisera raised against VEGF. The size of the detected protein, M_r of ~25,000 Da, is consistent with the monomeric form of VEGF₁₆₅. Treatment of cells with CGS21680 produced a 58.6 ± 6.8% (n = 7) inhibition of the serum-induced up-regulation of VEGF protein.

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Inhibition of VEGF protein secretion by CGS21680. PC12 cells were serum-starved for 12 h and then refed with serum-free medium (serum-free), or medium containing 10% equine serum/5% FBS in the absence (+serum) or presence of 1 µM CGS21680 (+CGS21680; serum). Medium was collected after 12 h and prepared for Western blotting as described in Materials and Methods. Shown is a representative Western blot with positions of molecular mass standards indicated to the left.

To begin to explore the signal transduction pathway that couples A_2AAR activation to VEGF mRNA down-regulation, the identity of the distinct G protein(s) that may mediate this response was examined (Fig. 5). PC12 cells were treated with either 1 µM CGS21680 or 100 ng/ml cholera toxin for 6 h and VEGF mRNA levels were quantitated. Cholera toxin produces a direct activation of the G_s subunit as a result of the disruption of the GTPase activity of the protein. In five experiments, VEGF mRNA was similarly reduced by CGS21680 (40.3 ± 12.5% of control) and cholera toxin (31.4 ± 8.4% of control). Next, to assess any possible involvement of G_i/o proteins in the A_2AAR-mediated response, the effect of 1 µM CGS21680 was determined in control cells and cells that had been treated with pertuss toxin (200 ng/ml × 18 h) to ablate receptor-G_i/o coupling. In control cells, CGS21680 reduced VEGF mRNA to a level that was 23.2 ± 5.2% of basal (n = 4). In parallel experiments with cells treated with pertussis toxin, the basal level of VEGF mRNA was 76.5 ± 8.8% of the basal value in control cells. In the presence of pertussis toxin, CGS21680 down-regulated VEGF mRNA to a level that was 44.6 ± 4.4% of that in cells that had not been exposed to pertussis toxin. This represents a level of VEGF mRNA of 59.0 ± 3.5% of that observed in cells that were exposed to pertussis toxin but not treated with CGS21680.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   G protein involvement in the A2AAR-mediated down-regulation of VEGF mRNA. A, PC12 cells were treated with 1 µM CGS21680 or 100 ng/ml cholera toxin for 6 h as indicated at which time RNA was isolated and VEGF mRNA content analyzed via Northern blotting. Results are reported as the percentage of VEGF mRNA relative to untreated cells (basal). Shown is mean ± S.E. of five experiments. B, PC12 cells were untreated or exposed to pertussis toxin (200 ng/ml × 18 h) before the addition of 1 µM CGS21680. Results are reported as the percentage of VEGF mRNA relative to cells not exposed to pertussis toxin. Shown is mean ± S.E. of four experiments.

To examine the molecular level at which activation of the A_2AAR may down-regulate the steady-state level of VEGF mRNA, the effect of CGS21680 on the stability of the VEGF transcript and VEGF gene transcription rate was determined. To address possible post-transcriptional regulation of VEGF mRNA by CGS21680, RNA was isolated from control and CGS21680-treated PC12 cells at various intervals after introduction of the transcription inhibitor actinomycin D. VEGF mRNA levels were quantitated and normalized to the 28S rRNA content. In four experiments (Fig. 6), the half-life of VEGF mRNA in control and CGS21680-treated PC12 cells was 78.2 ± 6.9 and 78.0 ± 6.1 min, respectively, indicating that A_2AAR activation does not increase the lability of the VEGF transcript.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Effect of CGS21680 on VEGF mRNA stability. Northern blotting was performed with RNA obtained from control PC12 cells () and cells treated with 1 µM CGS21680 () for 4.5 h. RNA was isolated at the times shown after addition of actinomycin D (5 µg/ml). The VEGF mRNA signal was normalized to that for 28S rRNA, calculated as a percentage of that determined at time 0 and plotted on a log scale versus time. Points represent the mean of four experiments and bars indicate S.E.M.

The effect of A_2AAR activation on the VEGF gene transcription rate was next examined in nuclear run-on experiments. After a 4-h treatment with 1 µM CGS21680 or vehicle, nuclei were isolated from PC12 cells and newly transcribed RNA was quantitated as described in Materials and Methods. Compared with control cells, PC12 cells treated with CGS21680 displayed a reduction in VEGF gene transcription, whereas that for -actin was little affected (Fig. 7). In six experiments, the rate of VEGF gene transcription compared with that of -actin was reduced by 48.9 ± 10.2% in cells treated with the A_2AAR agonist.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Effect of CGS21680 on VEGF gene transcription. Nuclei were isolated from control PC12 cells and cells treated with 1 µM CGS21680 for 4 h. Nuclear run-on experiments were performed as described in Materials and Methods. A, representative blot showing hybridization of labeled RNA to immobilized VEGF cDNA in pBluescript. -Actin in pBluescript was used for normalization and empty pBluescript vector was used to assess background signals. B, graphical representation of the results (mean ± S.E.) from six experiments. The VEGF/actin ratio is expressed as a percentage of that observed in untreated cells (control).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

VEGF is an endothelial cell mitogen that has a critical role in angiogenesis during embryonic development and in disease (Ferrara and Davis-Smyth, 1997). The present findings demonstrate that activation of the A_2AAR in PC12 cells markedly down-regulates VEGF expression at the mRNA and protein levels. Four AR subtypes, A₁AR, A_2AAR, A_2BAR, and A₃AR, are recognized (Olah and Stiles, 1995). PC12 cells do not express the A₁AR (van der Ploeg et al., 1996) and radioligand-binding studies and adenylyl cyclase assays indicate that the A₃AR is not functionally expressed in these cells (M. E. Olah, unpublished data). However, both the A_2AAR and A_2BAR have been characterized in PC12 cells (van der Ploeg et al., 1996). The potency of CGS21680 in down-regulating VEGF mRNA (EC₅₀ of ~0.5 nM) strongly suggests that the response is mediated specifically by the A_2AAR. The cloned rat A_2AAR expressed in COS-6 M cells displayed a K_d of 38 nM for [³H]CGS21680 (Fink et al., 1992). Conversely, 10 µM CGS21680 failed to stimulate adenylyl cyclase in COS-6 M cells expressing the cloned rat A_2BAR (Stehle et al., 1992). Involvement of an AR in the CGS21680-mediated down-regulation of VEGF mRNA also is indicated by the ability of the antagonist ZM241385 to abolish the reduction of VEGF mRNA produced by the agonist. Interestingly, ZM241385 increased VEGF mRNA expression by ~2.5-fold. Up-regulation of VEGF mRNA by ZM241385 may occur due to the inhibition of adenosine present in the culture medium that may tonically down-regulate VEGF via activation of the A_2AAR. Similarly, a possible selective blockade of the A_2AAR by ZM241385 may unmask an adenosine-induced up-regulation of VEGF mRNA that is mediated by the A_2BAR. Indeed, a role for the A_2BAR in the induction of VEGF in human retinal endothelial cells has recently been described (Grant et al., 1999). It would be of interest to determine any effect of selective activation of the A_2BAR on VEGF expression in PC12 cells. However, in preliminary experiments assessing adenylyl cyclase activation, it was not possible to define concentrations of various AR antagonists, including ZM241385, that abolished signaling through the A_2AAR while producing minimal antagonism at the A_2BAR. Markedly selective A_2BAR agonists do not exist.

Varying effects of AR activation on VEGF expression have been reported. In human lymphocytic U-937 cells (Hashimoto et al., 1994) and porcine microvascular endothelial cells (Fischer et al., 1995), adenosine or its analogs up-regulated VEGF mRNA. However, the use of the AR agonists NECA and ()-(R)-N⁶-(phenylisopropyl)adenosine at high concentrations made it difficult for the AR subtype mediating the response to be identified. Additionally, AR antagonists were used at nonselective concentrations in these studies (Hashimoto et al., 1994; Fischer et al., 1995). While this manuscript was in preparation, the ability of 10 µM NECA to down-regulate VEGF mRNA expression presumably via activation of the A_2AAR in PC12 cells was reported (Kobayashi and Millhorn, 1999). Additionally, exposure of PC12 cells to an AR antagonist during hypoxia resulted in enhanced VEGF mRNA levels relative to untreated PC12 cells maintained in hypoxia (Kobayashi and Millhorn, 1999). The AR ligands analyzed in the latter study do not discriminate between the A_2AAR and A_2BAR at the concentrations tested.

Takagi et al. (1996) reported that CGS21680 at concentrations relatively selective for the A_2AAR up-regulated VEGF mRNA by ~1.5-fold in bovine retinal pericytes. Hypoxic induction of VEGF protein was nearly abolished in these cells by pretreatment with an AR antagonist (Takagi et al., 1996). The qualitatively differential effect of A_2AAR activation on VEGF expression in retinal pericytes versus PC12 cells may occur for several reasons. First, as described below, signaling initiated by A_2AAR activation may vary among cell types. Second, the ability of a specific second messenger molecule to regulate VEGF mRNA may differ in a cell-dependent fashion. For example, elevations in intracellular cAMP levels have been shown to increase (Claffey et al., 1992; Hashimoto et al., 1994; Takagi et al., 1996), decrease (Fischer et al., 1995), or have no effect (Levy et al., 1995a) on VEGF mRNA levels in varying cell types. In this context, the transformed versus untransformed state of primary retinal pericytes versus PC12 cells may be relevant. A reduction in VEGF expression in PC12 cells differentiated toward a neuronal phenotype by treatment with nerve growth factor was described (Claffey et al., 1992). Studies are currently examining the regulation of VEGF expression by the A_2AAR and other G protein-coupled receptors in additional cell types in which signaling pathways similar to those in PC12 cells may exist.

VEGF is up-regulated by multiple stimuli, including hypoxia (Levy et al., 1995a,b), growth factors (Nauck et al., 1997), and cytokines (Li et al., 1995; Ryuto et al., 1996). Conversely, few agents have been shown to reduce VEGF expression as described presently for A_2AAR agonists. Glucocorticoids inhibit the up-regulation of VEGF induced by stimuli such as serum, platelet-derived growth factor, phorbol esters, and platelet-activating factor in different cell types (Finkenzeller et al., 1995; Heiss et al., 1996; Nauck et al., 1997). Interestingly, dexamethasone down-regulated VEGF mRNA levels in NIH 3T3 cells stimulated with phorbol-12-myristate-13-acetate or platelet-derived growth factor, but did not affect hypoxic induction of VEGF mRNA (Finkenzeller et al., 1995). It was demonstrated that atrial natriuretic peptide inhibits both endothelin- and hypoxia-induced up-regulation of VEGF in human umbilical vein smooth muscle cells (Pedram et al., 1997). This effect was independent of cGMP generation and occurred, at least in part, via inhibition of VEGF gene transcription (Pedram et al., 1997). In the present study, VEGF mRNA expression under hypoxic conditions was down-regulated by CGS21680. However, it is not currently possible to define this repression as specific for hypoxia-response elements in the VEGF gene. To examine the susceptibility of hypoxia-specific VEGF induction to CGS21680, in additional experiments PC12 cells were maintained in 0.1% FBS during hypoxia. However, even in this markedly reduced serum concentration, the normoxic VEGF mRNA level was down-regulated by CGS21680 (data not shown). Nonetheless, in the relevant context of a reduced oxygen environment, VEGF mRNA remains substantially down-regulated in response to A_2AAR activation. A_2AAR activation in PC12 cells attenuates the hypoxia-induced inhibition of voltage-sensitive potassium currents and subsequent calcium influx (Kobayashi et al., 1998). These findings and those of the present study suggest that elevated adenosine levels that may exist during hypoxia may act via the A_2AAR in a negative feedback loop to modulate hypoxia-induced responses.

The mechanism by which A_2AAR activation results in VEGF mRNA down-regulation has begun to be explored. This response appears to occur in large part due to coupling of the A_2AAR to G_s because cholera toxin reproduced the down-regulation of VEGF mRNA observed with CGS21680. Coupling of the A_2AAR to G_s activation with subsequent adenylyl cyclase stimulation is established in several cell types, including PC12 cells (van der Ploeg et al., 1996). However, activation of the A_2AAR may produce responses independent of elevated intracellular cAMP levels (Lai et al., 1997; Sexl et al., 1997). Relevant are findings obtained in PC12 cells that indicate the A_2AAR couples to activation of a novel protein kinase C isoform, possibly via a G_i/o protein (Lai et al., 1997). The present study provides evidence that G_i/o activation also may partially mediate the A_2AAR-mediated response. In pertussis toxin-treated PC12 cells, the ability of CGS21680 to reduce VEGF mRNA content was slightly blunted compared with control cells. However, the role of G_i/o proteins in VEGF regulation may be complex because prolonged treatment with pertussis toxin alone slightly decreased VEGF mRNA. Experiments are currently underway to define the intracellular signaling cascades coupling A_2AAR stimulation to VEGF mRNA down-regulation and to examine the possibility as suggested by the above-mentioned data that activation of multiple signaling pathways may be involved.

Finally, it appears that activation of the A_2AAR and ensuing signaling cascade decreases the steady-state VEGF mRNA level by reducing the VEGF gene transcription rate. VEGF mRNA up-regulation can occur due to enhanced gene transcription and/or a stabilization of VEGF mRNA (Levy et al., 1995b, 1996; Li et al., 1995; Ryuto et al., 1996). Experiments using actinomycin D indicated that the half-life of the VEGF transcript was unaffected by CGS21680 when the A_2AAR agonist was used for a duration that markedly reduces VEGF mRNA levels. Conversely, nuclear run-on experiments revealed a substantial reduction in the VEGF gene transcription rate in cells treated with CGS21680. Inhibition of VEGF gene transcription is apparently reflected in a decreased expression of transcripts for VEGF₁₂₁, VEGF₁₆₅, and VEGF₁₈₉.

In summary, stimulation of the A_2AAR in PC12 cells substantially reduces VEGF mRNA expression and VEGF protein secretion. Regulation of VEGF secretion by the A_2AAR or perhaps other G protein-coupled receptors on selected targets may represent a means to positively or negatively regulate angiogenesis for therapeutic benefit. However, cell-specific responses to adenosine and the associated underlying mechanisms require further exploration.