QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: mol.104.001107v1 66/5/1147    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sands, W. A. Articles by Palmer, T. M. Search for Related Content PubMed PubMed Citation Articles by Sands, W. A. Articles by Palmer, T. M.

Specific Inhibition of Nuclear Factor-B–Dependent Inflammatory Responses by Cell Type-Specific Mechanisms upon A_2A Adenosine Receptor Gene Transfer

Molecular Pharmacology Group, Division of Biochemistry and Molecular Biology, Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow, Scotland, United Kingdom

Received April 6, 2004; accepted July 30, 2004

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Adenosine is a potent inhibitor of inflammatory processes, and the A_2A adenosine receptor (A_2AAR) plays a key nonredundant role as a suppresser of inflammatory responses in vivo. In this study, we demonstrate that increasing A_2AAR gene expression suppressed multiple inflammatory responses in both human umbilical vein endothelial cells (HUVECs) and rat C6 glioma cells in vitro. In particular, the induction of the adhesion molecule E-selectin by either tumor necrosis factor (TNF) or Escherichia coli lipopolysaccharide (LPS) was reduced by more than 70% in HUVECs, whereas inducible nitric-oxide synthase (iNOS) induction was abolished in C6 cells after exposure to interferon- in combination with LPS and TNF, suggesting that the receptor inhibited a common step in the induction of each of these pro-inflammatory genes. Consistent with this hypothesis, A_2AAR expression inhibited the activation of NF-B, a key transcription factor whose proper function was essential for optimal iNOS and E-selectin induction. However, although NF-B binding to target DNA was severely compromised in both cell types, the mechanisms by which this occurred were distinct. In C6 cells, A_2AAR expression blocked IB degradation by inhibiting stimulus-induced phosphorylation, whereas in HUVECs, A_2AAR expression inhibited NF-B translocation to the nucleus independently of any effect on IB degradation. Together, these observations suggest that A_2AAR-mediated inhibition NF-B activation is a critical aspect of its anti-inflammatory signaling properties and that the molecular basis of this inhibition varies in a cell type-specific manner.

However, despite numerous observations describing the anti-inflammatory effects of A_2AAR activation in a variety of cellular systems, relatively little is known about the molecular mechanisms by which they occur. In this study, we demonstrate that elevating A_2AAR protein expression in glioma cells and vascular ECs is sufficient to markedly inhibit inflammatory responses to maximally effective concentrations of pro-inflammatory stimuli. In addition, we demonstrate that the receptor can exert these effects by specifically blocking activation of the critical transcription factor NF-B via distinct cell type-specific mechanisms.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Materials. The human A_2AAR gene was generously donated by Dr. Andrea Townsend-Nicholson (Royal Free Hospital, London, U.K.). GeneJuice transfection reagent was from Novagen (Madison, WI). HUVECs plus appropriate growth medium and supplements were from Cambrex Bio Science Nottingham Ltd. (Nottingham, UK). Nucleofection reagents were from Amaxa (Cologne, Germany). Antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) (IB, RelA/p65, 12CA5), Serotec (Oxford, UK) (E-selectin monoclonal antibody 1.2B6), BD Biosciences (Palo Alto, CA) [inducible nitric-oxide synthase (iNOS)], and Cell Signaling Technology Inc. (Beverly, MA) (Ser³²Ser³⁶-phosphorylated IB, Ser⁶³-phosphorylated c-jun, and Thr¹⁸⁰Tyr¹⁸²-phosphorylated and total p38). The NF-B oligonucleotide probe was from Santa Cruz Biotechnology. Escherichia coli 0111:B4 LPS, rat and human TNF, and rat interferon- (IFN) were from Sigma-Aldrich. A 3-B enhancer conA-luciferase reporter plasmid, sheep nonimmune antobodies, and anti-p50 and -p65 antibodies used for supershift experiments were generously donated by Prof. Ron Hay (University of St. Andrews, UK). A control pSV--galactosidase expression construct was from Promega. A bacterial expression plasmid that directed expression of GST-c-jun(1–79) and rabbit anti-JNK1,2 polyclonal antibody were generously donated by Prof. David Gillespie (Cancer Research UK Beatson Laboratories, Glasgow, UK). A control adenovirus (AV) that directed the expression of GFP alone, termed AV/GFP, was generously donated by Prof. Miles Houslay (University of Glasgow, UK). pcDNA3/HA-tagged wild type (WT) IB was generously donated by Prof. Warner Greene (Gladstone Institute of Virology and Immunology, University of California, San Francisco, CA). Sources of other materials have been described elsewhere (Palmer and Stiles, 1999).

Cell Culture. HUVECs were propagated in ECM-2 medium supplemented with 2% (w/v) fetal bovine serum, hydrocortisone, ascorbate, and recombinant growth factors as recommended by the supplier (Cambrex Bio Science Nottingham Ltd.). Human embryonic kidney 293 and C6 glioma cells were propagated in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum, 1 mM L-glutamine, 100 units/ml penicillin and 100 µg/ml streptomycin. The generation and characterization of the C6 cell lines used in this study have been described elsewhere (Palmer and Stiles, 1999). All cells were grown at 37°C in a humidified atmosphere containing 5% (v/v) CO₂.

Generation of a Myc Epitope-Tagged Human A_2AAR. Myc epitope (EQKLISEEDL) and His₆ sequences were added to the carboxyl terminus of the human A_2AAR by PCR using pCMV5/human A_2AAR cDNA as the template. The primers were 5'-TAGCAGAGCTCGTTTAGT-3', which anneals within pCMV5 upstream of the A_2AAR initiating Met, and 5'-TGATTTCTAGAGGACACTCCTGCTCCATCCTG-3', which was designed to remove the A_2AAR stop codon and add a XbaI site (in bold). This was to allow in-frame fusion of the A_2AAR upstream of the myc epitope and His₆ sequences in pcDNA3/mycHisA (Invitrogen) after ligation of HindIII/XbaI-digested vector and similarly digested PCR product. The integrity of the generated open reading frame was verified by DNA sequencing.

Generation of Recombinant Serotype 5 AV Encoding myc-humA_2AAR. The myc-humA_2AAR open reading frame was excised from pcDNA3/mycHisA by HindIII/PmeI digestion and subcloned into HindIII/EcoRV-digested shuttle vector pAdTrackCMV: the subcloning procedure destroyed the EcoRV and PmeI sites. The resulting pAdTrackCMV/myc-humA_2AAR construct was then linearized at the remaining PmeI site within the vector to expose the inverted terminal repeats, and cotransformed with pAdEasy1 into E. coli strain BJ5183 by electroporation in 2.0 mm cuvettes at 2500V, 200, and 25 µF. Successful pAdEasy1/myc-humA_2AAR recombinants were identified by PmeI digestion and PCR using myc-humA_2AAR-specific primers, and expanded in E. coli XL1 Blue. Plasmid DNA prepared from chosen clones was then digested with PacI and transfected using LipofectAMINE into low-passage human embryonic kidney 293 cells to initiate viral production. Large-scale amplification and titration of AV/GFP and AV/myc-humA_2AAR viral stocks were performed as described by He et al. (1998).

For infection, HUVECs were washed in regular growth medium and then incubated overnight with the same medium supplemented with recombinant AV at the m.o.i. indicated under Results and figure legends. The next day, the virus-containing medium was aspirated and replaced with normal medium. Cells were used for analysis 24 h later.

Luciferase Assays of Reporter Gene Transcription in C6 Glioma Cells. C6 cells at 70 to 80% confluence in six-well dishes were transfected with 1 µg/well each of 3-B enhancer conA-luciferase and pSV-galactosidase plasmids using 7 µl of GeneJuice according to the manufacturer's instruction. Forty-eight hours later, cells were treated as described in the figure legends. Reactions were terminated by placing on ice and washing twice with ice-cold PBS before harvesting by scraping into 50 µl of lysis buffer]50 mM potassium phosphate, pH 7.8, 0.2% (v/v) Triton X-100, and 0.5 mM DTT] that was clarified by centrifugation at 48,000g for 15 min. Fifteen microliters of the supernatant was used for determination of -galactosidase activity according to manufacturer's instructions; 25 µl was added to luciferase assay buffer [50 mM Tris phosphate, pH 7.8, 16 mM magnesium chloride, 2 mM DTT, 1.8% (v/v) Triton X-100, and 30% (v/v) glycerol] for determination of luciferase activity in a 96-well luminometer plate. Samples were assayed in triplicate, and luciferase activity was normalized to -galactosidase activity.

Transient Transfection of HUVECs. Endotoxin-free cDNA expression constructs were prepared using the Wizard Purefection plasmid DNA purification system (Promega). These were introduced into HUVECs using the Amaxa nucleofection kit according to the manufacturer's instructions. In brief, 1 x 10⁶ HUVECs per sample were resuspended in 100 µl of nucleofection buffer containing 1 µgof pMaxGFP (to assess transfection efficiency) and 4 µg of either pcDNA3 (vector control) or pcDNA3/HA-tagged WT IB. Plasmid DNA was introduced into the nucleus of the cells using an Amaxa Nucleofector set at program U-O1. Cells were then seeded into 96-well plates, and E-selectin induction was assessed 48 h later by ELISA.

¹²⁵I-ZM241385 Synthesis and Saturation Binding. ¹²⁵I-ZM241385 was synthesized, purified by HPLC, and used in radioligand binding studies on freshly isolated HUVEC membranes as described by Palmer et al. (1995).

Monocyte Adhesion Assay. 1 x 10⁴ HUVECs/well were seeded into a 24-well plate and grown overnight. The cells were then infected for 24 h at a m.o.i. of 25 pfu/cell with recombinant AVs. After treatment as indicated in the figure legends, the medium was aspirated, and HUVEC monolayers were overlaid with 1 x 10⁵ U937 promonocytic cells/well. The cells were allowed to adhere for 1 h at 37°C before removal of the medium and washing of monolayers three times with 1 ml/well serum-free Dulbecco's modified Eagle's medium to remove any nonadherent U937 cells. The cells were then fixed in 0.5 ml/well 4% (w/v) paraformaldehyde in 5% (w/v) sucrose/PBS, pH 7.2, and analyzed using a combination of fluorescence and bright-field microscopy to determine the number of U937 cells adhering to GFP-expressing HUVECs. At least 300 GFP-expressing cells were counted in three to five separate fields to calculate the average number of adherent cells per 100 GFP-expressing HUVECs for any given treatment.

ELISA for Cell Surface E-Selectin Expression. 1 x 10⁵ HUVECs/well were seeded in a 96-well plate and grown to 70% confluence. The cells were then infected for 24 h at a m.o.i. of 25 pfu/cell with recombinant AVs. The following day, cells were treated with TNF for the times indicated in the figure legends. The incubation was stopped by transferring the plate to ice, and cells were washed gently three times with 0.2 ml/well ice-cold PBS. Cells were then fixed by overnight incubation at 4°C with 0.1 ml/well 4% (w/v) paraformaldehyde in 5% (w/v) sucrose/PBS, pH 7.2. The next day, each well was washed gently three times with PBS and then incubated for 1 h at room temperature in PBS containing 0.1% (w/v) BSA to block nonspecific antibody binding sites. After three more washes with PBS, 0.1 ml of anti-E-selectin antibody diluted 1:1000 in PBS/0.1% (w/v) BSA was added to each well. A parallel set of isotype control incubations using anti-hemagglutinin antibody 12CA5 at 1:1000 dilution was included to assess nonspecific antibody binding. After a 2-h incubation at room temperature, antibody solution was removed and the cells were washed three times with PBS before the addition of 0.1 ml/well horseradish peroxidase-conjugated anti-mouse IgG diluted 1:1000 in PBS/0.1% (w/v) BSA and incubation for 1 h at room temperature. Wells were then washed five times with PBS before the addition of 0.1 ml/well 3,3',5,5'-tetramethylbenzidine. The reaction was allowed to develop at room temperature before A₆₀₀ determination using a plate reader.

Griess Assay of Nitrite Accumulation. Confluent C6 cells in six-well plates were incubated in normal medium with the stimuli indicated in the figure legends. Synthesis of nitric oxide was determined by assay of culture supernatants for nitrite, a reaction product of nitric oxide and molecular oxygen. In brief, 0.2 ml of culture supernatant was allowed to react with 0.2 ml of Griess reagent (Hevel and Marletta, 1994) and incubated at room temperature for 15 min. After transfer to a 96-well plate, sample absorbance at 570 nm was determined using a plate reader. Fresh culture media served as the blank in all experiments, and nitrite concentrations from experimental samples were calculated from a standard curve derived from the parallel reaction of known amounts of sodium nitrite.

Immunoblotting. Confluent HUVECs or C6 cells in six-well plates were treated as described in the figure legends before washing in ice-cold PBS and solubilization by scraping into 0.2 ml/well electrophoresis sample buffer (50 mM Tris-HCl, pH 6.8, 10% (v/v) glycerol, 2% (w/v) SDS) at room temperature. After brief probe sonication, insoluble material was pelleted by microcentrifugation, and the supernatant was assayed for protein content using a bicinchoninic acid assay. Samples equalized for protein content (typically 50–70 µg/sample) were fractionated by SDS-PAGE on 10 or 12% (w/v) resolving gels. After transfer to nitrocellulose, membranes were blocked for 1 h at room temperature in blocking solution [5% (w/v) skimmed milk in PBS containing 0.1% (v/v) Tween 20]. Membranes were then incubated overnight at 4°C with primary antibody diluted in fresh blocking buffer. Primary antibodies were each used at used at a final concentration of 1 µg/ml. After three washes in blocking solution, membranes were incubated for 1 h at room temperature with appropriate horseradish peroxidase-conjugated secondary antibody at a 1 in 1000 dilution. After further washes with blocking buffer and PBS, immunoreactive proteins were visualized by enhanced chemiluminescence. For phosphospecific antibodies, a similar protocol was used except that the primary antibodies were diluted in Tris-buffered saline containing 1% (w/v) BSA and 0.1% (v/v) Tween 20, and all washes were with Tris-buffered saline/0.1% (v/v) Tween 20. Quantitation was by densitometric scanning of exposed films using Totallab imaging software (Phoretix; Nonlinear Dynamics, Newcastle upon Tyne, UK).

Pull-Down Assay of c-Jun N-Terminal Kinase Activation. Confluent C6 cells in six-well dishes were treated as indicated in the figure legends. Incubations were terminated by placing the cells on ice and washing in ice-cold PBS. All subsequent procedures were performed at 4°C unless indicated otherwise. Cells were solubilized in 0.3 ml/well c-Jun N-terminal kinase (JNK) lysis buffer (25 mM HEPES, pH 7.7, 20 mM -glycerophosphate, 0.3 M sodium chloride, 1.5 mM magnesium chloride, 10 mM sodium fluoride, 0.1 mM sodium vanadate, 0.1 mM PMSF, and 10 µg/ml each of soybean trypsin inhibitor and benzamidine). After isolation of soluble fractions by microcentrifugation, samples equalized for protein content and volume were incubated overnight with rotation with 10 µg of recombinant GST-c-jun(1–79) immobilized to glutathione-Sepharose beads after induction and purification from transformed E. coli BL21 cultures. Complexes were isolated by brief microcentrifugation, washed three times with wash buffer [20 mM HEPES, 50 mM sodium chloride, 2.5 mM magnesium chloride, 0.1 mM EDTA and 0.05% (v/v) Triton X-100], and then resuspended in 30 µl of kinase assay buffer (25 mM HEPES, pH 7.5, 10 mM magnesium chloride, 20 mM -glycerophosphate, and 75 mM sodium vanadate) supplemented with 100 µM ATP and 2 mM DTT. Immobilized JNK was activated by incubation at 30°C for 30 min. Reactions were stopped by the addition of 10 µl of 4x electrophoresis sample buffer containing 12% SDS and heating to 95°C for 5 min. Samples were fractionated by SDS-PAGE on 12% (w/v) polyacrylamide resolving gels and transferred to nitrocellulose for immunoblotting with anti-Ser⁶³-phosphorylated c-jun antibody for visualization of GST-c-Jun(1–79) phosphorylation. Protein-equalized samples from each treatment were also processed in parallel for immunoblotting with anti-JNK1,2 antibody.

Electrophoretic Mobility Shift Assay. After treatment as described in the figure legends, confluent 10-cm dishes of HUVEC or C6 cells were washed three times with ice-cold PBS before scraping. Cells were pelleted by centrifugation at 400g and lysed by the addition of 0.4 ml/tube buffer A [10 mM HEPES, pH 7.9, 10 mM potassium chloride, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.625% (v/v) Nonidet P-40, 0.5 mM PMSF, and 10 µg/ml each of soybean trypsin inhibitor and benzamidine]. The samples were then centrifuged at 48,000g for 30 s at 4°C, and the supernatant was removed. The pellet was resuspended in 50 µl of buffer B (20 mM HEPES, pH 7.9, 0.45 M sodium chloride, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.5 mM PMSF, and 10 µg/ml each of soybean trypsin inhibitor and benzamidine). Samples were agitated for 15 min at 4°C, centrifuged at 48,000g for 5 min, and the protein content of the supernatant was determined using a Bradford assay. Five-microliter samples from each nuclear extract were then added to a ³²P-labeled double-stranded DNA probe (10,000 cpm/sample) containing the consensus B binding sequence GGGGACTTTCCC to give a final reaction volume of 25 µl containing 10 mM sodium HEPES, pH 7.9, 0.1 mM magnesium chloride, 0.1 mM EDTA, 0.5 mM DTT, 10% (v/v) glycerol, 50 mM sodium chloride, and 0.625 µg/ml poly(dIdC). After a 30-min incubation at room temperature, samples were analyzed by fractionation on a nondenaturing 6% (w/v) polyacrylamide gel containing 0.5x Tris-borate/EDTA buffer (45 mM Tris-borate, 1 mM EDTA) followed by autoradiography. For supershift analysis, 1 µl/sample of sheep polyclonal anti-p65, anti-p50 antibodies, or nonimmune serum controls were added after the initial binding of the probe, and samples were incubated for a further 30 min at room temperature before electrophoresis.

Laser-Scanning Confocal Microscopy. 1 x 10⁵ HUVECs were plated onto glass coverslips in six-well plates and cultured overnight. The next day, cells were treated as outlined in the figure legends, permeabilized by the addition of 0.4% (v/v) Triton X-100 in PBS for 3 min, then washed with PBS supplemented with 0.1% (v/v) goat serum and 0.2% (w/v) gelatin (PBSGG). Coverslips were then incubated with a 1:200 dilution of RelA/p65 antibody. After washing with three changes each of PBSGG and PBS, coverslips were incubated with at room temperature for 1 h in a 1:200 dilution of Alexa594-conjugated goat anti-rabbit IgG. After sequential washing with PB-SGG and PBS as described above, coverslips were mounted onto glass slides using 40% (v/v) glycerol in PBS for dual-label confocal microscopy using a Zeiss AxioVERT 100 confocal microscope. Images were collected using dual excitation (488 and 543 nm) and emission (515–540 nm for GFP, and 590–610 nm for Alexa594) filter sets. Single labeled controls ensured that there was no interchannel bleed-through under these conditions. All collected images were analyzed using Metamorph software (Universal Imaging Corporation, Downingtown, PA).

Statistical Analysis. Data are presented in the text as means ± S.E. for the number of experiments indicated, whereas representative experiments are presented in the figures. Statistical significance was assessed using Student's t test or analysis of variance with a probability of 0.05.

	Results

Top Abstract Materials and Methods Results Discussion References

 
Effect of A_2AAR Expression on iNOS Induction in C6 Glioma Cells. Glial cell inflammation plays an important role in the pathophysiology of Parkinson and Alzheimer diseases (Hirsch et al., 2003; Monsonego and Weiner, 2003). To determine the ability of the A_2AAR to regulate inflammatory responses in this cell type, rat C6 glioma cells stably expressing the canine A_2AAR to a range between 2 and 3 pmol/mg, as determined by saturation binding with the A_2AAR-selective antagonist radioligand ¹²⁵I-ZM241385, were used (Palmer and Stiles, 1999). The accumulation of nitrite and the associated induction of iNOS were used as inflammatory markers in this system, because iNOS-derived nitric oxide release contributes to the ability of activated glial cells to promote neuronal cell death (Brown and Bal-Price, 2003). Although no significant nitrite accumulation was observed when they were added individually, maximally effective concentrations of LPS, TNF, or a combination of the two with IFN produced a marked induction of nitrite accumulation in control C6 cells that was abolished in A_2AAR-expressing C6 cells (nitrite accumulation in response to LPS/TNF/IFN treatment for 24 h was reduced by 95 ± 9% in C6/A_2AAR cells versus control C6 cells, n = 10, p < 0.05; Fig. 1A). This was associated with an abolition of iNOS induction in response to LPS/TNF/IFN even in the absence of the AR agonist NECA (Fig. 1B). The constitutive inhibition of iNOS induction observed is unlikely to be caused by the accumulation of extracellular adenosine acting on the recombinant receptor for two reasons. First, all cellular incubations presented in this study were performed in the presence of 3 units/ml adenosine deaminase to convert any released adenosine to inosine, which is inactived at the A_2AAR (Jin et al., 1997). Second, inhibition of iNOS induction could not be reversed significantly by inclusion of the A_2AAR-selective antagonist ZM241385 at a concentration that blocks specific binding to the A_2AAR [LPS/TNF/IFN treatment produced a 12 ± 12% increase in nitrite accumulation in vehicle pretreated C6/A_2AAR cells versus 5 ± 15% in 1 µM ZM241385-pretreated cells, n = 3, p > 0.05 (not significant); Fig. 1C]. Therefore, it is possible that constitutive agonist-independent signaling from the A_2AAR is sufficient to block iNOS induction in C6 cells.

View larger version (17K): [in this window] [in a new window]   Fig. 1. Effect of A2AAR expression on nitrite accumulation and iNOS induction in C6 glioma cells. A, control and A2AAR-expressing C6 cells were treated for 24 h with 1 µg/ml LPS (L), 10 ng/ml TNF (T), or 10 units/ml IFN (I) individually or in combination (LPS/TNF/IFN) (C) before analysis of nitrite accumulation as described under Materials and Methods. This is one of multiple similar experiments. B, control and A2AAR-expressing C6 cells were treated for 24 h with or LPS/TNF/IFN in the absence of presence of 5 µM NECA. Detergent-soluble extracts from treated cells were then normalized for protein content before fractionation by SDS-PAGE and assessment of iNOS and A2AAR expression by immunoblotting. This is one of multiple similar experiments. C, control and A2AAR-expressing C6 cells were treated for 24 h with or without LPS/TNF/IFN in the absence of presence of 1 µM ZM241385 as indicated. Samples were then analyzed for nitrite accumulation as described under Materials and Methods. This is one of multiple similar experiments.

The A_2AAR Acts Intracellularly to Inhibit iNOS Induction. In theory, it was possible that A_2AAR expression was leading to constitutive induction and secretion of anti-inflammatory cytokines or other signaling molecules that were acting extracellularly to inhibit iNOS induction. However, medium taken from A_2AAR-expressing C6 cells failed to inhibit LPS/TNF/IFN induction of iNOS expression in control C6 cells (Fig. 2). In fact, medium from A_2AAR-expressing C6 cells actually potentiated iNOS induction by 62 ± 18% (p < 0.05, n = 4) compared with medium from control C6 cells (Fig. 2). Therefore, the A_2AAR must be acting intracellularly to block the signaling pathways leading to iNOS induction.

View larger version (26K): [in this window] [in a new window]   Fig. 2. Effect of conditioned medium from A2AAR-expressing C6 glioma cells on iNOS induction in control C6 glioma cells. Fresh medium was added to confluent dishes of control and A2AAR-expressing C6 cells. After 24 h, the medium from each dish was removed, supplemented with LPS/TNF/IFN or not, and added to confluent monolayers of control C6 cells for a further 24 h as described in Fig. 1. Protein-normalized soluble extracts were then prepared and fractionated by SDS-PAGE for immunoblot analysis of iNOS induction. This is one of multiple experiments.

Effect of A_2AAR Expression on the Activation of the NF-B, JNK, and p38 Pathways. Three key inflammatory signaling pathways activated in response to LPS and TNF include the NF-B pathway and the JNK and p38 stress-activated protein kinase signaling pathways (Manning and Davis, 2003; Kumar et al., 2003; Karin et al., 2004). Immunoblotting of extracts from TNF-treated control and A_2AAR-expressing C6 cells with an antibody versus active phosphorylated p38 demonstrated that in both cell types, TNF produced a transient activation of this kinase that peaked by 5 min and returned to basal by 30 min (Fig. 3A). Expression of the A_2AAR produced a 25 ± 6% (p < 0.05, n = 3) decrease in the maximal extent to which p38 was activated after normalization to total p38 expression. In vitro kinase assays of JNK activation demonstrated that A_2AAR expression had no significant effect on either the maximal extent of TNF stimulation [maximal activation reduced by 9 ± 6% at 15 min, p > 0.05 (not significant), n = 3; Fig. 3B] or total JNK expression levels (Fig. 3B). Consistent with a minimal effect on JNK activation, TNF stimulation of AP-1 activity, as determined after transient transfection of an AP1-luciferase reporter construct, was also not significantly different between control and A_2AAR-expressing C6 cells (data not shown).

View larger version (54K): [in this window] [in a new window]   Fig. 3. Effect of A2AAR expression on TNF stimulation of p38 phosphorylation and JNK activation in C6 glioma cells. A, control and A2AAR-expressing C6 glioma cells were treated with or without 10 ng/ml TNF or 5 µg/ml anisomycin for the indicated times before preparation of soluble cell extracts. After normalization for protein content, duplicate samples were fractionated by SDS-PAGE and transferred to nitrocellulose for parallel immunoblotting with Thr180/Tyr182 phospho-specific and total anti-p38 antibodies as described under Materials and Methods. B, control and A2AAR-expressing C6 glioma cells were treated with or without 10 ng/ml TNF or 5 µg/ml anisomycin for the indicated times before preparation of soluble cell extracts. After normalization for protein content, samples were assayed for JNK activation by incubation with glutathione-Sepharose-bound GST-c-Jun(1–79) and assay of the pulled down kinase. Phosphorylation of GST-c-Jun(1–79) was assessed by SDS-PAGE and immunoblotting with a Ser63-phospho-specific c-Jun antibody and parallel immunoblotting of treated cell extracts with anti-JNK1,2 antibody as described under Materials and Methods.

Activation of the NF-B pathway was assessed by two criteria. First, EMSAs were used to determine the ability of NF-B to bind to target DNA. Stimulation with LPS and TNF, but not IFN, for 60 min stimulated DNA binding in control C6 cells, but this was reduced by 85 ± 16% (TNF) and 90 ± 11% (LPS) in A_2AAR-expressing cells (both p < 0.05 versus control C6 cells, n = 3; Fig. 4A). Supershift analysis revealed that both bands constituted p50-p65 heterodimers (data not shown). Second, to determine the functional consequences of reduced DNA binding, control and A_2AAR-expressing C6 cells were transfected with a 3-B enhancer conA-luciferase reporter construct to assess NF-B–mediated transcription. Although TNF treatment for 6 h produced a 53.4 ± 3.6-fold increase in luciferase activity (n = 5) in control C6 cells, the fold-stimulation in A_2AAR-expressing C6 cells was only 4.3 ± 2.7 (p < 0.05 versus control C6 cells, n = 5; Fig. 4B). Therefore, A_2AAR expression selectively attenuated NF-B-dependent gene transcription, and this was associated with a parallel reduction in NF-B binding to target DNA.

View larger version (37K): [in this window] [in a new window]   Fig. 4. Effect of A2AAR expression on NF-B activation in C6 glioma cells. A, control and A2AAR-expressing C6 glioma cells were treated with or without 10 ng/ml TNF, 1 µg/ml LPS, or 10 units/ml IFN for 60 min before isolation of nuclei for electrophoretic mobility shift assay analysis with a B-selective oligonucleotide probe as described under Materials and Methods. This is one of multiple experiments. B, control and A2AAR-expressing C6 glioma cells were cotransfected with a B-luciferase reporter construct and pSV-galactosidase. Forty-eight hours after transfection, cells were treated with or without TNF as indicated, and extracts were prepared for assay of luciferase and -galactosidase activities and normalization with respect to transfection efficiency and protein content as described under Materials and Methods. This is one of multiple experiments.

Effect of NF-B Inhibition on iNOS Induction. Several studies examining the promoter elements that control iNOS gene transcription have revealed the presence of multiple B binding sites and demonstrated their importance for proper induction in response to inflammatory stimuli (Xie et al., 1994; Taylor and Geller, 2000). Given that the effects of A_2AAR expression on p38 and JNK activation were minimal compared with the effect on NF-B function, it was likely that the inhibition of iNOS induction observed (Fig. 1) could be accounted for by the effects of A_2AAR on the NF-B pathway. To test this hypothesis, C6 cells were preincubated with the NF-B inhibitor pyrrolidine dithiocarbamate (PDTC) before 24-h treatment with or without LPS/TNF/IFN and assessment of nitrite accumulation in the medium (Dai et al., 2003). Preincubation with PDTC attenuated nitrite accumulation by 55 ± 7% (p < 0.05, n = 3) compared with C6 cells pretreated with an antioxidant control compound [N-acetylcysteine (NAcCys)] (Fig. 5A). An alternative strategy was also employed to assess the effects of NF-B inhibition on nitrite accumulation. In particular, control C6 cells were transfected with a wild-type IB expression construct before stimulation with LPS/TNF/IFN (Fig. 5B). Expression of recombinant IB inhibits NF-B function by sequestering liberated NF-B dimers, thereby preventing their nuclear translocation and subsequent binding to target DNA (Sun et al., 1996). Compared with vector-transfected C6 cells, nitrite accumulation was reduced by 68 ± 15% in IB-transfected cells (n = 3, p < 0.05, Fig. 5B). Together, these data suggest that inhibition of NF-B caused by A_2AAR expression could at least partly explain the receptor's ability to attenuate iNOS induction.

View larger version (17K): [in this window] [in a new window]   Fig. 5. Effect of NF-B inhibition on nitrite accumulation in C6 glioma cells. A, control C6 glioma cells were pretreated with or without 100 µM PTDC or NAcCys for 30 min before the addition of vehicle or LPS/TNF/IFN for a further 24 h as indicated. Medium from treated cells was then analyzed for nitrite accumulation as described under Materials and Methods. This is one of multiple experiments. B, control C6 glioma cells were transfected with either pcDNA3 vector or pcDNA3/HA-tagged WT IB as indicated. Twenty-four hours after transfection, cells were treated with LPS/TNF/IFN for a further 24 h as indicated. Medium from treated cells was then analyzed for nitrite accumulation as described under Materials and Methods. This is one of multiple experiments.

Effect of A_2AAR Expression on IB Regulation in C6 Cells. TNF and LPS each activate NF-B by initiating the degradation of IB, which allows p50-p65 heterodimers to translocate to the nucleus and initiate transcription after a series of phosphorylation and acetylation reactions on p65 (Karin et al., 2004). IB phosphorylation on serines 32 and 36 by the IB kinase (IKK) complex triggers its polyubiquitination by the SCF^TrCP complex and subsequent degradation by the 26S proteosome. In control C6 cells, TNF was able to initiate a marked reduction in IB protein with kinetics similar to those observed by other investigators (Wang et al., 1999; Fig. 6A). However, in A_2AAR-expressing cells, the observed degradation of IB in response to TNF was significantly reduced (at 30 min, IB reduction in control C6 cells was 84 ± 12% versus 20 ± 5% in A_2AAR-expressing C6 cells, p < 0.05, n = 3; Fig. 6A). Immunoblotting with an antibody that specifically recognized Ser³²Ser³⁶-phosphorylated IB revealed that this was associated with almost complete abolition of IB phosphorylation (phosphorylation after a 15-min exposure to TNF reduced by some 83 ± 8% in A_2AAR-expressing C6 cells versus controls, p < 0.05, n = 3, Fig. 6A). A similar inhibition of IB degradation and phosphorylation was observed in response to LPS treatment, with Ser³²Ser³⁶-phosphorylation of IB after 15 min being reduced by 75 ± 11% in A_2AAR-expressing C6 cells versus controls, p < 0.05, n = 3, (Fig. 6B). Hence, the ability of the A_2AAR to inhibit NF-B activation in response to TNF and LPS is probably caused by inhibition of IB phosphorylation by the upstream IKK complex.

View larger version (42K): [in this window] [in a new window]   Fig. 6. Effect of A2AAR expression on agonist-stimulated IB phosphorylation and degradation in C6 glioma cells. A, control and A2AAR-expressing C6 glioma cells were treated with or without 10 ng/ml TNF for the indicated times before preparation of detergent-soluble extracts. Samples equalized for protein content were fractionated by SDS-PAGE for immunoblotting with anti-IB (upper) or anti-Ser32Ser36-phospho-IB (lower) antibodies as described under Materials and Methods. This is one of multiple experiments. B, control and A2AAR-expressing C6 glioma cells were treated with or without 1 µg/ml LPS for the indicated times before preparation of detergent-soluble extracts. Samples equalized for protein content were fractionated by SDS-PAGE for immunoblotting with anti-IB (upper) or anti-Ser32Ser36-phospho-IB (lower) antibodies as described above. This is one of multiple experiments.

Effect of A_2AAR Activation on Monocyte Adhesion to TNF-Stimulated HUVECs in Vitro. To determine whether the molecular mechanisms by which the A_2AAR inhibited inflammatory responses in C6 cells were conserved between different cell types, the effects of A_2AAR gene transfer were also assessed in HUVECs, a cell line whose responses to stimuli such as TNF and LPS are well characterized. Several reports have suggested that ECs express low levels of endogenous A_2AARs (Sexl et al., 1997; Feoktistov et al., 2002; Wyatt et al., 2002). To assess the ability of these receptors to inhibit inflammatory responses in HUVECs in vitro, cells were pretreated with a maximally effective concentration of the A_2AAR-selective agonist CGS21680before treatment with increasing concentrations of TNF and quantitation of U937 monocyte adhesion. Whereas maximally effective concentrations of TNF increased U937 adhesion by 148 ± 25-fold (n = 5 experiments) with an EC₅₀ of 0.37 ± 0.21 ng/ml, pretreatment with 5 µM CGS21680increased the EC₅₀ to 2.00 ± 0.88 ng/ml, thereby reducing monocyte adhesion at submaximal concentrations of TNF (adhesion observed at 1 ng/ml TNF reduced by 51 ± 14% n = 3, by 5 µM CGS21680versus vehicle-treated HUVECs, p < 0.05; Fig. 7, A and B). Pharmacological analysis of this effect demonstrated that whereas the effect of CGS21680was blocked completely by the A_2AAR-selective antagonist ZM241385, it could not be mimicked by the A₁AR-selective agonist CPA, which is indicative of an A_2AAR-mediated phenomenon (Fig. 7B).

View larger version (18K): [in this window] [in a new window]   Fig. 7. Effect of endogenous A2AAR activation on monocyte adhesion to TNF stimulated HUVECs in vitro. A, confluent HUVEC monolayers were pretreated for 30 min with 5 µM CGS21680before incubation with increasing concentrations of TNF for 6 h as indicated. U937 promonocytic cells were then overlaid and the number of adherent cells quantitated as described under Materials and Methods. This is one of multiple experiments. B, confluent HUVEC monolayers were pretreated for 30 min with 5 µM CGS21680(CGS), 5 µM CGS21680 and 1 µM ZM241385 (CGS+ZM) or 5 µM N6-cyclopentyladenosine (CPA) as indicated before treatment with or without 10 ng/ml TNF for 6 h and assessment of U937 adhesion as described under Materials and Methods. This is one of multiple experiments.

Effect of A_2AAR Gene Transfer on Monocyte Adhesion to TNF-Stimulated HUVECs in Vitro. To assess whether it was possible to potentiate the anti-inflammatory effects of the A_2AAR, HUVECs were infected with recombinant AV encoding a myc epitope-tagged human A_2AAR with the aim of increasing receptor expression. The AV also expressed GFP from a separate open reading frame to facilitate identification of successfully infected cells by fluorescence microscopy. By this approach, it was possible to achieve recombinant protein expression in approximately 90% of infected HUVECs at optimal m.o.i. (Fig. 8A) with A_2AAR expression reaching 0.34 ± 0.06 pmol/mg of protein (n = 3 experiments) as determined by saturation binding analysis using ¹²⁵I-ZM241385 (Fig. 8B). It is interesting that despite unequivocal identification of functional A_2AARs (Fig. 7), no specific binding of ¹²⁵I-ZM241385 was detectable in either uninfected or control AV/GFP-infected HUVECs. A similar inability to detect low level endogenous functional A_2AARs by radioligand binding has also been reported for T-lymphocytes (Armstrong et al., 2001).

View larger version (27K): [in this window] [in a new window]   Fig. 8. Adenovirus-directed expression of the human A2AAR in HUVECs. A, HUVECs were infected with AV/myc-human A2AAR at the indicated m.o.i. as described under Materials and Methods. After the isolation of detergent-soluble extracts, samples were equalized for protein content and fractionated by SDS-PAGE for immunoblotting with anti-myc epitope antibody 9E10 to specifically identify recombinant receptor. Blots were then stripped and reprobed with an antibody versus Gi-2, to ensure equivalent protein loading in each lane. Determination of "% cells infected" was derived from counting the number of GFP-expressing cells by fluorescence microscopy and the total number of cells by light microscopy within five different fields. This is one of multiple experiments. B, membranes prepared from HUVECs infected with 25 pfu/cell AV/myc-human A2AAR were subjected to saturation binding analysis using the A2AAR-selective antagonist radioligand 125I-ZM241385 as described under Materials and Methods. This is one of multiple experiments.

Experiments performed in C6 cells suggested that A_2AAR expression was sufficient to inhibit inflammatory responses even in the absence of agonist. To determine whether this was also a feature of endothelial cells, U937 adhesion assays were performed on TNF-treated HUVECs infected with either AV/GFP or AV/myc-humA_2AAR (Fig. 9). Whereas TNF strongly promoted monocyte adhesion in control GFP-expressing cells, expression of the A_2AAR inhibited this effect by 84 ± 8% (n = 3, p < 0.05 versus TNF-treated HUVECs infected with control AV/GFP), and this was not improved substantially by the presence of the AR agonist NECA (89 ± 8% inhibition, n = 3, p < 0.05 versus TNF- and NECA-treated HUVECs infected with control AV/GFP; Fig. 9).

View larger version (17K): [in this window] [in a new window]   Fig. 9. Effect of A2AAR gene transfer on monocyte adhesion to TNF-pretreated HUVECs in vitro. HUVECs were infected with AV/myc-human A2AAR at an m.o.i. of 25 pfu/cell as described under Materials and Methods. After pretreatment for 30 min with or without 5 µM NECA, cells were incubated with 10 ng/ml TNF for 6 h as indicated. U937 cells were then overlaid and the number of adherent cells quantitated as described under Materials and Methods. This is one of multiple experiments.

Effect of A_2AAR Gene Transfer on E-Selectin Induction in HUVECs. Adherence of leukocytes to inflamed endothelium is mediated by specific adhesion molecules on both cell types. E-selectin is a key molecule on the EC surface responsible for initial recruitment of leukocytes out of the circulation before their arrest and diapedesis into underlying tissue (Ley, 2001). In agreement with several other studies, exposure of control AV-infected HUVECs to either TNF for 6 h induced E-selectin expression at the cell surface (Oitzinger et al., 2001; Fig. 10A). A_2AAR gene transfer reduced the TNF-mediated increase in cell surface E-selectin by 76 ± 5% (n = 3, p < 0.05 versus TNF-treated AV/GFP-infected HUVECs; Fig. 10A). Moreover, this effect was not restricted to TNF, because induction of E-selectin after a 4-h stimulation with LPS was similarly reduced (80 ± 12%, n = 3, p < 0.05 versus LPS-treated AV/GFP-infected HUVECs; Fig. 10B) as determined by immunoblotting. Thus, A_2AAR expression is sufficient to inhibit both E-selectin induction and monocyte adhesion.

View larger version (26K): [in this window] [in a new window]   Fig. 10. Effect of A2AAR gene transfer on TNF- and LPS-mediated induction of E-selectin expression. A, HUVECs were infected with either AV/GFP (control AV) or AV/myc-human A2AAR at an m.o.i. of 25 pfu/cell as described under Materials and Methods. After pretreatment with or without 10 ng/ml TNF for 6 h, cell surface E-selectin expression was determined by ELISA as described under Materials and Methods. This is one of multiple experiments. B, HUVECs were infected with either AV/GFP (control AV) or AV/myc-human A2AAR at an m.o.i. of 25 pfu/cell as described under Materials and Methods. After treatment in the absence (0) or presence of 1 µg/ml LPS either alone (LPS) or with 5 µM NECA (LPS+N) for 4 h, cell surface E-selectin expression was determined by immunoblotting as described under Materials and Methods. This is one of multiple experiments.

Effect of NF-B Inhibition on E-Selectin Induction and Monocyte Adhesion. Analysis of the human E-selectin promoter has identified four positive regulatory domains (termed PDI to -IV) responsible for conferring induction in response to inflammatory stimuli (Read et al., 1997). PDs I, III, and IV are B-binding sites, whereas PDII is an AP-1 site at which an ATF2/c-jun heterodimer is constitutively bound. Thus, maximal induction arises from occupation of the B-binding sites by NF-B as well as p38- and JNK-mediated phosphorylation of ATF2 and c-jun, respectively (Read et al., 1997). The ability of TNF to promote E-selectin induction in HUVECs was found to be dependent on B binding rather than AP-1 activation, as HUVEC pretreatment with the NF-B inhibitor PDTC inhibited induction by 87 ± 10% under conditions in which the p38 inhibitor SB203580 had no significant effect [7 ± 16% increase versus nonpretreated cells, p > 0.05 (not significant), n = 3; Fig. 11A). It is important that the effect of PDTC could not be explained by its antioxidant properties, because parallel treatment with the anti-oxidant NAcCys had minimal effects on E-selectin induction [6 ± 10% increase versus nonpretreated cells, p > 0.05 (not significant), n = 3, Fig. 11A]. Consistent with other investigators (Read et al., 1997), inhibition of the extracellular signal-regulated kinase signaling pathway using the mitogen-activated protein kinase kinase inhibitor U0126 also had no effect on induction [6 ± 7% increase versus nonpretreated cells, p > 0.05 (not significant), n = 3; Fig. 11A]. Again, to confirm a role for NF-B, HUVECs were transfected with a WT IB expression construct before stimulation with TNF and measurement of E-selectin induction. These experiments demonstrated that compared with vector-transfected cells, IB expression reduced TNF-stimulated E-selectin induction by 72 ± 7% (n = 3, p < 0.05 versus vector-transfected controls; Fig. 11B). It is noteworthy that the observed reduction in E-selectin observed after inhibition of NF-B was associated with a parallel inhibition of U937 adhesion to TNF-stimulated HUVECs in vitro (PDTC produced a 88 ± 9% versus 25 ± 10% inhibition of adhesion in NAcCys-treated cells, n = 4, p < 0.05; Fig. 11C).

View larger version (15K): [in this window] [in a new window]   Fig. 11. Effect of NF-B inhibition on TNF-mediated induction of E-selectin and monocyte adhesion in HUVECs. A, confluent HUVECs in a 96-well plate were incubated with or without 10 ng/ml TNF after a 30-min preincubation in the absence or presence of 10 µM SB203580, 10 µM U0126, 100 µM PDTC, or 100 µM NAcCys. Quantitation of E-selectin induction was by ELISA as described under Materials and Methods. This is one of multiple experiments. B, HUVECs were transfected with pMaxGFP and either pcDNA3 vector or pcDNA3/HA-tagged WT IB, seeded into a 96-well plate and cultured for 48 h as described under Materials and Methods. Cells were then treated with or without 10 ng/ml TNF for 6 h before quantitation of E-selectin induction by ELISA. This is one of multiple experiments. C, confluent HUVEC monolayers were pretreated for 30 min with either 100 µM PDTC or NAcCys as indicated before stimulation with 10 ng/ml TNF for 6 h. U937 cells were then overlaid and the number of adherent cells quantitated as described the Materials and Methods. This is one of multiple experiments.

Effect of A_2AAR Gene Transfer on NF-B Regulation in HUVECs. The above experiments demonstrated that, similarly to C6 glioma cells, the anti-inflammatory effects of the A_2AAR in HUVECs could be largely mimicked by inhibition of NF-B signaling. To determine whether A_2AAR gene transfer could also inhibit NF-B binding to target DNA in HUVECs, EMSAs were performed on TNF-treated cells after infection with either AV/GFP or AV/myc-humA_2AAR (Fig. 12A). These demonstrated that the ability of TNF to promote p50-p65 heterodimer binding to target DNA was reduced by 86 ± 15% upon A_2AAR gene transfer (n = 5, p < 0.05 versus TNF-treated AV/GFP-infected HUVECs; Fig. 12A).

View larger version (35K): [in this window] [in a new window]   Fig. 12. Effect of A2AAR gene transfer on the regulation of NF-B in HUVECs. A, confluent HUVECs were treated for 60 min in the absence of presence of 10 ng/ml TNF before isolation of nuclei for incubation with a B-selective oligonucleotide probe and the indicated antibodies for electrophoretic mobility shift assay analysis as described under Materials and Methods. The asterisk indicates a nonspecifically labeled band. This is one of multiple experiments. B, confluent HUVECs monolayers were infected at an m.o.i. of 25 pfu/cell with either a control GFP-encoding AV or AV/myc-human A2AAR. After treatment with or without 10 ng/ml TNF for the indicated times, soluble cell extracts were prepared and equalized for protein content before SDS-PAGE and immunoblotting with anti-IB antibody. This is one of multiple experiments. C, HUVECs plated onto glass coverslips were infected at an m.o.i. of 25 pfu/cell with either control AV/GFP or GFP AV/myc-human A2AAR, which also expresses GFP from a separate open reading frame. After treatment with or without 10 ng/ml TNF for 60 min as indicated, cells were fixed and permeabilized, and p65 was identified by staining with a p65-specific antibody and Alexa594-conjugated anti-rabbit IgG and by dual label confocal microscopy as described under Materials and Methods. HUVECs expressing recombinant protein are visible as green from GFP-derived fluorescence, whereas endogenous p65 is visible as red. This is one of multiple experiments.

Translocation of NF-B to the nucleus is triggered by the phosphorylation-dependent degradation of the IB proteins that maintain NF-B in the cytoplasm (Karin et al., 2004). Expression of the A_2AAR in C6 cells severely attenuated IB phosphorylation on Ser³² and Ser³⁶, the two sites whose phosphorylation by the IKK complex is required for polyubiquitination and subsequent degradation by the proteosome (Fig. 6). However, despite unequivocal evidence of AV-directed A_2AAR expression in HUVECs, the ability of TNF to trigger the degradation of IB was unaffected [degradation compared with vehicle-treated control at 15 min was 94 ± 6% (AV/GFP) versus 89 ± 10% (AV/myc-humA_2AAR), n = 3, p > 0.05 (not significant); Fig. 12B], suggesting that the mechanisms by which the A_2AAR inhibits NF-B signaling differ between C6 glioma and HUVEC systems.

To determine whether the nuclear translocation of NF-B was altered despite normal IB degradation, confocal laser-scanning microscopy was performed to analyze the subcellular distribution of p65 in AV-infected HUVECs after treatment with or without TNF (Fig. 12C). Under unstimulated conditions, p65 resided in the cytoplasm in both sets of AV-infected cells. However, although stimulation with TNF promoted the nuclear translocation of p65 in controls, p65 was excluded from the nucleus of A_2AAR-expressing HUVECs (Fig. 12C). Hence, A_2AAR gene transfer in HUVECs profoundly inhibits the ability of NF-B p50-p65 heterodimers to bind target DNA and this is associated with a parallel reduction in p65 translocation to the nucleus in the absence of any effect on IB degradation

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
The A_2AAR has been identified as a protective anti-inflammatory receptor protein not only from numerous pharmacological studies (reviewed in Linden, 2001) but also from characterization of inflammatory responses in mice in which both copies of the A_2AAR gene have been deleted (Ohta and Sitkovsky, 2001). Gene dosage studies have provided evidence to show that, at least in T-lymphocytes, there is no A_2AAR reserve (Armstrong et al., 2001). Therefore, pathophysiological conditions that alter A_2AAR expression, such as the onset of hypoxia (Kobayashi and Millhorn, 1999) and EC exposure to Th1 cytokines (Nguyen et al., 2003), are likely to alter cellular responsiveness to inflammatory stimuli. However, despite unequivocal evidence of its potent anti-inflammatory properties across different cell types, the molecular basis for the anti-inflammatory effects of A_2AAR have not been examined in detail. To begin addressing some of these issues, we have focused on characterizing the effects of increasing A_2AAR expression on glial and EC model systems, each of which is associated with multiple inflammatory diseases.

iNOS induction is a well characterized marker of glial cell inflammation (Brown and Bal-Price, 2003). Stable expression of the A_2AAR in C6 glioma cells was sufficient to abolish induction of iNOS expression and NF-B activation. Moreover, this abolition did not require the presence of agonist and was not reversed by preincubation with the A_2AAR-selective antagonist ZM241385, suggesting that in this system, the receptor displays some agonist-independent basal activity. A similar phenomenon was also observed for monocyte adhesion, E-selectin induction, and NF-B activation in HUVECs upon A_2AAR expression to a level of only 0.3 to 0.4 pmol/mg, suggesting that agonist-independent signaling may be an intrinsic property of the A_2AAR. Consistent with this hypothesis, constitutive A_2AAR activation of either cAMP production or the extracellular signal-regulated kinase pathway in the absence of agonist has been observed in other studies (Ledent et al., 1992; Klinger et al., 2002). However, both this study and others describing agonist-independent signaling from the A_2AAR have been cell systems in which the recombinant receptor is overexpressed. Therefore, the physiological significance in vivo of any basal activity of endogenous A_2AARs expressed at low levels is unclear and will require the development of an A_2AAR-selective inverse agonist. Despite this caveat, when considered in the context of our observations, the increases in A_2AAR expression and function seen upon cytokine treatment or hypoxia in multiple cell types (Kobayashi and Millhorn, 1999; Khoa et al., 2001; Nguyen et al., 2003) probably constitute important negative feedback mechanisms that prevent the development of an excessive and inappropriate inflammatory response. The availability of A_2AAR-deficient mice can now allow quantitative assessment of the presence and importance of such a negative feedback loop in individual inflammatory cells.

A major finding from our experiments was that the pronounced inhibition of TNF- and LPS-mediated inflammatory responses could be explained by a suppressive effect of A_2AAR expression specifically on the NF-B pathway rather than the parallel p38 and JNK stress-activated kinase modules. However, whereas expression of the A_2AAR in glial cells blocked the IKK-mediated IB phosphorylation on Ser³² and Ser³⁶ that triggers its degradation, IB degradation was unaffected upon A_2AAR expression in endothelial cells. Thus, the A_2AAR is able to inhibit the NF-B pathway at multiple postreceptor loci common to TNF and LPS/Toll-like receptor 4 (TLR4) signaling. In C6 cells, this locus must reside at a site at which TLR4 and TNF receptor signaling converge to activate NF-B. The most likely candidates include components of the IKK complex (particularly IKK and IKK/NEMO) as well as any IKK kinases activated by both LPS and TNF.

In terms of receptor pharmacology, we found that activation of endogenous A_2AARs on HUVECs by the selective agonist CGS21680was able to significantly inhibit monocyte adhesion, whereas the nonselective AR agonist NECA was not (Figs. 7 and 9). At the concentration used (5 µM), NECA activates all four AR subtypes (Fredholm et al., 2001). Several studies have demonstrated that both A_2A- and A_2BARs are predominantly expressed in endothelial cells. Although both receptors share an ability to elevate cAMP levels via interaction with G_s, the A_2BAR can also couple to G_q/11 to mobilize intracellular calcium (Feoktistov and Biaggioni, 1997). This additional signaling capacity accounts for its reported ability to induce the expression of angiogenic factors from endothelial cells under conditions in which the A_2AAR does not (Feoktistov et al., 2002). Therefore, upon exposure to NECA, A_2BAR activation may counteract the simultaneous effects of A_2AAR activation under the same conditions, and this may explain why only CGS21680reduces the adhesion observed at maximally effective concentrations of TNF.

Some of the findings in HUVECs presented here are consistent with the recent observations of Majumdar and Aggarwal (2003), who described an adenosine-mediated inhibition of NF-B in several cell types, including HeLa and cultured endothelial cells. Pharmacological analysis ruled out a contribution of A₁ and A₃ARs but did not address the specific involvement of A_2A or A_2BARs in each of the cell types under investigation (Majumdar and Aggarwal, 2003). However, the data presented here would suggest that the A_2AAR subtype may be largely responsible for the inhibition of NF-B observed by these investigators. A particularly important finding from Majumdar and Aggarwal (2003) was that adenosine specifically inhibited TNF-induced NF-B activation and nuclear translocation in the absence of any effect on IB degradation, which is consistent with our observations in A_2AAR-expressing HUVECs. However, in contrast to their results, we found that A_2AAR expression blocked NF-B–mediated induction of E-selectin by both TNF and LPS, suggestive of an effect of the A_2AAR on the NF-B nuclear translocation process that is common to both TNFR1 and TLR4 signaling cascades. This could be either reduced nuclear import of p50-p65 heterodimers and/or accelerated CRM1-mediated export of IB/NF-B complexes out of the nucleus (Arenzana-Seisdedos et al., 1997).

A_2AAR activation elevates intracellular cAMP levels in HUVECs via interaction with G_s and stimulation of adenylyl cyclase (Klinger et al., 2002). Consistent with other investigators, we have noted that elevation of cAMP in endothelial cells is sufficient to ameliorate B-dependent inflammatory responses, including adhesion molecule induction and subsequent monocyte adhesion (Morandini et al., 1996; W. A. Sands and T. M. Palmer, unpublished observations). However, the mechanisms by which this occurs seem to vary depending on the nature of the cAMP-elevating stimulus. For example, cyclic AMP-responsive element binding protein, which is phosphorylated on Ser¹³³ and activated upon elevation of intracellular cyclic AMP levels, can directly block NF-B-induced transcription without altering nuclear translocation or DNA binding by recruiting the transcriptional coactivator cAMP-responsive element binding protein-binding protein, thereby reducing the number of transcriptionally competent cAMP-responsive element binding protein-binding protein/NF-B complexes (Parry and Mackman, 1997). In contrast, the cAMP-elevating hormone adiponectin inhibits NF-B responses in ECs by inhibiting IKK-mediated IB phosphorylation and degradation (Ouchi et al., 2000). The data from this study and that of Majumdar and Aggarwal (2003) argue that despite being able to elevate cAMP levels in HUVECs (Klinger et al., 2002), the A_2AAR inhibits NF-B in these cells via a unique mechanism distinct from those employed by forskolin and adiponectin.

It should be noted, however, that despite the profound inhibitory effects of A_2AAR on NF-