Address correspondence and reprint requests to Luísa V. Lopes, Neurosciences Unit, Instituto de Medicina Molecular, Faculdade de Medicina de Lisboa, Av. Professor Egas Moniz, 1649-028 Lisbon, Portugal. E-mail: email@example.com
In situations of hypoxia, glutamate excitotoxicity induces neuronal death. The release of extracellular adenosine is also triggered and is accompanied by an increase of the stress mediator, corticotrophin-releasing factor (CRF). Adenosine A2A receptors contribute to glutamate excitoxicity and their blockade is effective in stress-induced neuronal deficits, but the involvement of CRF on this effect was never explored. We now evaluated the interaction between A2A and CRF receptors (CRFR) function, upon glutamate insult. Primary rat cortical neuronal cultures (9 days in vitro) expressing both CRF1R and CRF2R were challenged with glutamate (20–1000 μM, 24 h). CRF1R was found to co-localize with neuronal markers and CRF2R to be present in both neuronal and glial cells. The effects of the CRF and A2A receptors ligands on cell viability were measured using propidium iodide and Syto-13 fluorescence staining. Glutamate decreased cell viability in a concentration-dependent manner. Urocortin (10 pM), an agonist of CRF receptors, increased cell survival in the presence of glutamate. This neuroprotective effect was abolished by blocking either CRF1R or CRF2R with antalarmin (10 nM) or anti-Sauvagine-30 (10 nM), respectively. The blockade of A2A receptors with a selective antagonist SCH 58261 (50 nM) improved cell viability against the glutamate insult. This effect was dependent on CRF2R, but not on CRF1R activation. Overall, these data show a protective role of CRF in cortical neurons, against glutamate-induced death. The neuroprotection achieved by A2A receptors blockade requires CRF2R activation. This interaction between the adenosine and CRF receptors can explain the beneficial effects of using A2A receptor antagonists against stress-induced noxious effects.
corticotrophin-releasing factor (formerly known by CRH for corticotrophin-releasing hormone)
glial fibrillary acidic protein
microtubule-associated protein 2
phosphate buffer saline
cAMP-dependent protein kinase A
protein kinase C
Adenosine is a neuronal modulator that binds to different G-protein coupled receptors (Fredholm et al. 2001). Among them, the adenosine A2A receptors are attractive pharmacological targets because of their contribution to neuronal excitability, by increasing the release of glutamate (Lopes et al. 2002). In situations, where the release of glutamate is exacerbated, neuronal death either by apoptotic or necrotic processes can be detected (Nicotera et al. 1999).
In noxious brain conditions, such as accumulation of amyloid-β peptide (Aβ), hypoxic events or upon aging, there is an increase of both extracellular adenosine and adenosine A2A receptors levels in the brain (Latini and Pedata 2001; Rebola et al. 2005a; Cunha et al. 2006). In such situations, the blockade of A2A receptors can prevent the evoked neurotoxicity (Monopoli et al. 1998; Chen et al. 1999; Dall'Igna et al. 2003; Canas et al. 2009). Moreover, memory impairments caused by Aβ are prevented by A2A receptors antagonists (Canas et al. 2009). However, the mechanism by which the blockade of A2A receptors is effective in reverting these stressful effects remains unknown.
Stress response, in mammals, is dependent on the activation of the hypothalamic–pituitary–adrenal (HPA) axis. In the sequence of a stress stimulus, such as neurotoxicity induced by glutamate, the corticotrophin-releasing factor (CRF) is released from the hypothalamus, leading to HPA axis activation (Vale et al. 1981). CRF, a 41 amino acid peptide, has also important effects in extrahypothalamic sites, namely in thalamus, amygdala, hippocampus, frontal cerebral cortex, striatum, and cerebellum (Fischman and Moldow 1982; Swanson et al. 1983). In the hippocampus, CRF is released from inhibitory interneurons (Chen et al. 2001), binds to CRF1R abundant in dendritic spines of pyramidal neurons (Chen et al. 2004a, b), and modulates neuronal function and cognition (Radulovic et al. 1999). CRF1R is expressed in forebrain glutamatergic and γ-aminobutyric acid-containing (GABAergic) neurons as well as in midbrain dopaminergic neurons. CRF also binds to CRF2R (Chen et al. 1993; Lovenberg et al. 1995). These are predominantly Gs-coupled proteins that use cAMP as intracellular signaling molecule, but they also signal through Gi/o and Gq proteins, with minor involvement (Chen et al. 1986; Grammatopoulos et al. 2001).
CRF receptors are expressed in several brain regions that include hypothalamic and extrahypothalamic areas (Chalmers et al. 1995; Bittencourt et al. 1999). The actions of CRF in extrahypothalamic areas are still poorly explored. Although CRF1R receptor expression is very high in neocortical, cerebellar, and sensory relay structures, CRF2R receptor expression is generally confined to subcortical structures. The highest levels of CRF2R receptor mRNA in brain are evident within the lateral septal nucleus, the ventromedial hypothalamic nucleus and the choroid plexus. CRF2R -expressing cells are also evident, albeit in much lower density in the hippocampal formation and anterior and lateral hypothalmic areas. This heterogeneous distribution of CRF1R and CRF2R receptor mRNA suggests distinctive functional roles for each receptor in CRF-related systems (Chalmers et al.1995).
Noteworthy, CRF has a similar effect to that achieved by A2A receptor blockade, by counteracting classic neuronal insults as excitatory amino acids, hypoxia or amyloid-β25-35 peptide in cortical neurons (Fox et al. 1993; Pedersen et al. 2001), but neither the mechanism or subtype of receptor involved are known.
In this study, we investigated the relationship between the neuroprotective effects of adenosine A2A receptors blockade and activation of the two subtypes of CRF receptors. This was assessed using a glutamate insult, a major stress condition that reproduces the excitotoxic events following hypoxia/ischemia or during stroke. We found that, in cultured cortical neurons, CRF avoids cell death induced by glutamate, an effect dependent on the simultaneous activation of both CRF receptors, CRF1R and CRF2R. Blockade of A2A receptors is neuroprotective and requires activation of CRF2R. This interaction between the adenosine and CRF receptors can explain the beneficial effects of using A2A receptor antagonists against stress-induced noxious effects.
Urocortin and Antalarmin were purchased from Sigma (Madrid, Spain). 4-[2-[[6-Amino-9-(N-ethyl-β-d-ribofuranuronamidosyl)-9H-purin-2-yl]amino]ethyl] benzene propanoic acid (CGS 21680), 2-(2-Furanyl)-7-(2-phenylethyl)-7H-pyrazolo[4,3-e][1,2,4]triazolo[1,5-c]pyrimidin-5-amine (SCH 58261), l-Glutamic Acid, Anti-sauvagin-30 and Astressin were purchased from Tocris (UK). Aβ25-35 peptide was from Bachem (Bubendorf, Switzerland). These drugs were diluted in the assay solution from stock aliquots made in water or dimethyl sulfoxide stored at −20°C. All other reagents used were of the highest purity available and suitable for cell culture.
Primary rat cortical cultures
Cortical neurons were cultured from 18 days Sprague Dawley rat (Harlan, Barcelona, Spain) embryos according to Pedersen et al. 2002. Briefly, pregnant rats were handled according to the Portuguese law on animal care and European Union guidelines (86/609/EEC), and decapitated under deep anesthesia with Halothane. The embryos were collected in Hanks' Balanced Salt Solution and rapidly decapitated. Meninges and white mater were removed and whole cortices were incubated for 15 min in Hanks' Balanced Salt Solution (Calcium 1 mM and Magnesium 1 mM) and 0.025% trypsin. Cells were centrifuged three times and washed with Hanks' Balanced Salt Solution (with Calcium 1 mM and Magnesium 1 mM, 10% fetal bovine serum) and finally re-suspended in Neurobasal Medium. After counted, cells were plated on poly-l-lysine-coated coverslips in 24-well plates at density of 8 × 104 cells/well. Neurons were grown for 9 days at 37°C in a 5% CO2 humidified atmosphere in Neurobasal medium with 2% B-27 supplement, glutamate 25 μM, glutamine 0.5 mM, and 2 U/mL Penicillin/Streptomycin, in the absence of any positive selection for neurons. Medium was totally replaced at day 4 (without glutamate) and 60 min before drug treatment (without glutamate and B-27 supplement). Pure neuronal cultures were obtained by addition at day 3 in culture of 2 μM cytosine arabinoside (Ara-C) and used at day 9.
Quantitative real time RT-PCR (qPCR) was performed using RNA extracts from pure neuronal cultures. Briefly, neuronal cultures were washed with phosphate buffer saline (PBS) scraped and collected in 0.5 mL Eppendorf vials containing lysis buffer for RNA extraction. Total RNA was extracted using the RNAspin Mini RNA isolation kit (GE Healthcare, Buckinghamshire, UK) according to the manufacturer's instructions. RNA quantification was determined with nanoDrop 2000 software (Thermo scientific, Wilmington, DE, USA). Reverse transcription was carried out with the SuperScript™ First-Strand Synthesis System for RT-PCR (Invitrogen, Life technologies, Carlsbad, CA, USA) according to the manufacturer's instructions, using both 50 ng/μL random hexamers and 0.5 μg/μL oligo(dT)12-18 in a final volume of 20 μL. Negative controls were made without reverse transcriptase and confirmed the absence of signal. qPCR was carried out with SYBR green PCR Master mix (Applied Biosystems, Warrington, UK), using 2 μL of cDNA in a final volume of 25 μL (4.6 ng/μL of total cDNA), containing 0.3 μM of A2AR primer, 0.2 μM of β-actin primer (reference gene), or 0.4 μM of CRF2 primer. qPCR was performed with a Rotor-Gene 6000 Real Time Rotary Analyzer (Corbett Research, Cambridge, UK) for 45 cycles of 95°C for 20 s, 58°C for 5 s, and a final step of 10 s at 72°C.
The primers used in qPCR include: forward 5′-AACGGCATCAAGTACAACACGAC-3′ and reverse 5′-CGATTCGGTAATGCAGGTCATAC-3′for CRF2 (product size 142 bp, Invitrogen), forward 5′-ATTCCACTCCGGTACAATGG-3′ and reverse 5′-AGTTGTTCCAGCCCAGCAT-3′ for A2AR (product size 115 bp, Invitrogen) and forward 5′-AGCCATGTACGTAGCCATCC-3′ and reverse 5′-CTCTCAGCTGTGGTGGTGAA-3′ for β-actin (product size 228 bp, Invitrogen). The qPCR products were analyzed by electrophoresis on a 2% agarose gel containing GelRed™ Nucleic Acid Gel Stain (Biotium, Hayward, CA, USA).
To characterize the primary cortical neuronal cultures with 9 days in vitro, cell medium removed, cells were washed with phosphate buffer saline (PBS: NaCl 137 mM, KCl 2.7 mM, KH2PO4 1.8 mM and Na2HPO4 10 mM, pH 7.4) and fixed for 10 min at 20–23°C with 4% paraformaldehyde in PBS. After washing with PBS, cells were permeabilized with 0.1% Triton-X in PBS, blocked for 30 min with 0.25% gelatine in PBS, washed with PBS 0.05% Tween-20 and incubated for 1 h at 20–23°C with primary antibodies diluted in PBS 0.1% gelatine (mouse anti-MAP2 1 : 200, Millipore-Billerica, MA, USA-MAB3418, rabbit anti-GFAP 1 : 100, Sigma G9269, mouse CD11b 1 : 250, Serotec-Oxford, UK-MCA275R, Goat anti-CRF1R 1 : 25, Santa Cruz Biotechnology-Santa Cruz, CA, USA-sc 12381 and rabbit anti-CRF2R 1 : 25, Novus biologicals-Cambridge, UK-nbp1-00767, mouse anti-A2AR 1 : 50, Santa Cruz sc 32261. After washes (PBS 0.05% Tween-20), cells were incubated with secondary antibodies diluted in PBS 0.1% gelatine (anti-mouse Alexa Fluor 568 and anti-rabbit Alexa Fluor 488, both from Invitrogen). 4′,6-diamidino-2-phenylindole (DAPI, 70 μg/mL; Sigma) was used to label cell nucleus (5 min incubation). Coverslips were mounted with MOWIOL (Sigma), and cells were observed either with an Axiovert 200 fluorescence microscope (Carl Zeiss light microscopy, Gottingen, Germany) or Zeiss confocal LSM 710 microscope (Carl Zeiss MicroImaging GmbH, Jena, Germany).
Glutamic acid (or glutamate) was used as neurotoxic insult, concentrations ranging from 20 to 1000 μM were applied for 24 h in primary cultured neuronal cells with 8 days in vitro (Tamura et al. 1993). Amyloid-β25-35 peptide (Aβ, 25 μM) was used as positive control for apoptosis (Estus et al. 1997). CRF and adenosine A2A receptors antagonists were applied 15 min before initiating cell insult, while their agonists were added to cell medium immediately prior to glutamate. A2A receptors ligands were used as in Rebola et al. (2005b), while CRF receptors agonist and antagonists' use was based on Pedersen et al. (2002), Elliott-Hunt et al. (2002), and Gulyas et al. (1995) reports. Dimethyl sulfoxide concentration in cell medium was always kept below 0.001%.
Propidium iodide and Syto-13 uptake assay
Cells were washed with KREBS-HEPES (NaCl 117 mM, KCl 3 mM, Glucose 10 mM, NaHCO3 26 mM, Na2HPO4 1.25 mM, HEPES 10 mM, CaCl2 2 mM, MgCl2 1 mM), incubated with Syto-13 (4 μM, emits preferentially at 509 nm when excited at 488 nm) and propidium iodide (PI, 5 μg/mL, absorbing preferentially at 535 nm and emitting at 617 nm) for 3 min at room temperature and directly observed on Axiovert 200 fluorescence microscope. Three to four arbitrary photographs from each coverslip were shot and an average of 1400 cells was counted per condition in each experiment. Viable cells presented homogeneous cell body labeled with Syto-13, whereas primary and secondary apoptotic cells showed fragmented or condensed nucleus (labeled with Syto-13 or PI). Necrotic cells appeared as diffuse blots, emitting in propidium iodide range (Canas et al. 2009). After cell counting (see Fig. 1c), cell viability is presented as the ratio between the number of living cells and the number of total cells counted.
Cells from primary cultures with 9 days in vitro were washed with cold PBS. Using NP-40 lysis buffer pH 8.0 (1% Nonidet P40, 150 mM NaCl, 50 mM Tris-base, 1 mM EDTA, 5 mM dithiothreitol, proteases inhibitors - Complete, EDTA-free Protease Inhibitor cocktail tablets; Roche, Manheim, Germany) cells were mechanically scarped. The resulting solution was centrifuged at 16 000 g during 10 min at 4°C and pellet was discarded and the supernatant used for western blot. The protein concentration was determined using a BioRad DC Protein assay Kit (based on Lowry et al. 1951) because of the high levels of detergents present in the sample. The appropriate volume of each sample was diluted in water and sample buffer (350 mM Tris pH 6.8, 30% glycerol, 10% sodium dodecyl sulfate, 600 mM dithiothreitol and 0.012% Bromophenol blue). The samples were denatured at 95°C for 5 min.
Based on the protocol of Towbin et al. (1979), samples and molecular weight markers were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (10% for resolving and 5% for stacking gels) in denaturing conditions and electro-transferred to polyvinylidene difluoride membranes (Millipore). Membranes were blocked with 5% non-fat dry milk for 1 h, washed with TBS-T 0.1% (Tris Buffer Saline with 0.1% Tween-20 solution, 200 nM Tris, 1.5 M NaCl), and incubated with primary antibody (diluted in TBS-T, 3% Bovine Serum Albumin and 0.1% NaN3) overnight at 4°C. After washing again with TBS-T for 30 min, the membranes were incubated with horseradish peroxidise (HRP, EC 18.104.22.168) conjugated secondary antibody (in 5% non-fat dry milk) for 1 h at 20–23°C, (primary and secondary antibodies: mouse anti-A2A receptor (Upstate/Millipore 05-717), goat anti-CRF1R (Santa Cruz Biotechnology sc 12381), rabbit anti-CRF2R (Novus biologicals nbp1-00767), rabbit anti-caspase-3 (Santa Cruz Biotechnology sc-7148), rabbit anti-tubulin (abcam, Cambridge, UK, ab4074), goat anti-rabbit-HRP (Santa Cruz Biotechnology sc-2004), goat anti-mouse-HRP (Santa Cruz Biotechnology sc-2005), and donkey anti-goat-HRP (Santa Cruz Biotechnology sc-2020). After 40 min of washing with TBS-T, chemoluminescent detection was performed with ECL-PLUS western blotting detection reagent (GE Healthcare) using X-Ray films (Fujifilm, Dusseldorf, Germany). Optical density was determined with Image-J software (NIH, Bethesda, MD, USA).
The values presented are mean ± SEM of n independent experiments, corresponding to different cell cultures. In statistical tests between three or more conditions, a one-way anova was used, followed by a Bonferroni's multiple comparison post-hoc test. Values of p < 0.05 were considered to be statistically significant.
Characterization of the primary neuronal cell cultures
Primary cortical cultures with 9 days in vitro were labeled with anti-MAP2 (microtubule-associated protein 2 – neuronal marker) and anti-GFAP (glial fibrillary acidic protein – mature astrocytic marker) antibodies (Fig. 1a). Approximately, 50% of cells were labeled red (MAP2 expression). The presence of astrocytes was less than 20% (green labeled GFAP positive cells). The remaining cells are microglia that express CD11b, a selective marker (labeled in red, Fig. 1b). The cell culture expresses both CRF1R and CRF2R, and adenosine A2A receptors (Fig. 1d). CRF1R are mainly co-localized with neuronal markers (Fig. 1e see also figure S1 panel A), whereas CRF2R are present in both neuronal and glial cells (astrocytes, Fig. 1f see also figure S1 panel B). We also detected co-expression of mRNA encoding for A2A receptors and CRF2R in pure neuronal cultures and further confirmed the co-localization of the respective receptor protein by immunocytochemistry (Fig. 1g–h, see also figure S1 panel C).
Cell viability upon glutamate insult
Cell survival upon glutamate insult was evaluated by simultaneous labeling with propidium iodide and Syto-13. The cultures were treated for 24 h with five different glutamate concentrations (20, 50, 100, 500, and 1000 μM) covering several degrees of cell injury which can induce either apoptosis or necrosis (Lipton and Rosenberg 1994; Bonfoco et al. 1995). Although propidium iodide is incorporated through the disrupted membrane of dying or dead cells exclusively, emitting red light, Syto-13 is capable of labeling all cells (dead or living cells, emitting green light). Apoptotic cells present either condensed or fragmented nucleus and are labeled in green or red; necrotic cells appear as characteristic red dots. Cell viability was accessed by the percentage of cells that does not present any apoptotic or necrotic markers (see Fig. 1c).
As shown in Fig. 2, incubation for 24 h with glutamate resulted in a reduction of cell viability, in a concentration-dependent manner, with the lowest viability of 66.5 ± 4.23% of control (p < 0.001, n = 4) produced by the higher glutamate concentration used (1000 μM). This insult was considerably lower when compared to amyloid-β25-35 peptide (Aβ, 25 μM), which increases cell death by apoptosis (Estus et al. 1997; Kemppainen et al. 2012) that caused a 43.1 ± 2.53% reduction in cell viability. We further confirmed that cell death in our experimental conditions occurs mainly in neurons, as illustrated in Fig. 2b (see also figure S2). The density of MAP2 positive cells is decreased in glutamate treated cultures, and 90% of the remaining neurons (350 out of 400 counted neurons) present clear neuronal atrophy visible by a decrease in the number and length of the neuritis.
As observed in Fig. 3, the apoptotic marker caspase-3 increased with increasing glutamate concentrations, reaching a maximum level at 50 μM (468 ± 20.3% of control, p < 0.001, n = 3). For higher concentrations of glutamate, caspase-3 levels slightly decreased in relation to the maximum level.
Effect of corticotrophin-releasing factor on glutamate neurotoxicity
Urocortin, a peptide belonging to the CRF family, activates both subtype 1 and 2 of CRF receptors, CRF1R and CRF2R (Vaughan et al. 1995). Urocortin 10 pM was applied to cell medium immediately before glutamate (20 to 1000 μM range). In a similar model, this urocortin concentration provides a maximum neuroprotective effect against an Aβ stimulus (Pedersen et al. 2002). As shown in Fig. 4, urocortin increased cell survival in the presence of 50 μM (from 77.8 ± 0.95% to 88.5 ± 0.97%, p < 0.05, n = 4) and 100 μM (from 78.3 ± 1.37% to 88.7 ± 1.43%, p < 0.001, n = 8) glutamate. For lower (20 μM) and higher concentrations of glutamate (500 and 1000 μM), urocortin was not able to improve cell viability. Activation of CRF receptors in the absence of glutamate insult did not alter cell viability by itself (94.0 ± 2.18% of CTR, p > 0.05, n = 4, Fig. 4). Glutamate 100 μM was used in subsequent experiments because of the higher degree of neuroprotection induced by urocortin in this condition.
To distinguish the subtype of CRF receptors underlying the protection afforded by urocortin, the selective CRF1R and CRF2R antagonists, antalarmin 10 nM (Ant), and anti-Sauvagine-30 10 nM (a-Sau), as well as a non-selective antagonist for CRF receptors, astressin 10 nM (Ast), were used. As shown in Fig. 5a, the protection by urocortin was lost by blocking either both receptors simultaneously (with astressin, from 88.7 ± 1.43% to 74.5 ± 1.94%, p < 0.01, n = 4) or by blocking selectively CRF1R with antalarmin (from 88.7 ± 1.43% to 74.0 ± 4.46%, p < 0.01, n = 8) or CRF2R with anti-Sauvagine-30 (from 88.7 ± 1.43% to 72.0 ± 2.50%, p < 0.01, n = 8). Both antalarmin and anti-Sauvagine-30 decrease cell viability in cells treated with glutamate (from 78.3 ± 1.37% to 68.1 ± 2.81% and 68.2 ± 4.97%, respectively, p < 0.05, n = 3). None of these drugs significantly altered cell viability, when applied in the absence of glutamate (p > 0.05, n = 3, Fig. 5b).
Involvement of adenosine A2A receptors on CRF neuroprotection
In view of the involvement of adenosine A2A receptors in excitotoxicity phenomena, we then evaluated the possible interaction between the neuroprotection mediated by A2A receptors blockade and the consequence of activating CRF receptors.
The blockade of A2A receptors by its selective antagonist, SCH 58261 (50 nM), did not change the neuroprotection induced by urocortin (n = 6, Fig. 6). However, SCH 58261 (50 nM) alone prevented the cell death induced by glutamate (78.3 ± 1.37% to 88.5 ± 1.60%, p < 0.01, n = 4). Interestingly, neuroprotection obtained by the CRF receptor agonist, urocortin, was not additive with that achieved by the A2A receptor antagonist, SCH 58261 (Fig. 6) suggesting a common mechanism of action. In contrast, A2A receptor activation with the selective agonist, CGS 21680 (30 nM), did not altered cell death alone (76.4 ± 3.62%, p > 0.05, n = 6), or in the presence of urocortin (from 88.7 ± 1.43% with urocortin, to 82.4 ± 2.45% with CGS 21680, n = 8, Fig. 6).
Furthermore, as presented in Fig. 7, the neuroprotective effect achieved by the blockade of A2AR (in the presence of urocortin) was lost by selective CRF2R blockade (71.5 ± 2.52%, p < 0.001 compared with SCH 58261 plus urocortin, n = 6) or the blockade of both CRF receptors simultaneously (70.4 ± 2.34%, p < 0.001 compared with SCH 58261 plus urocortin, n = 4). Selective blockade of CRF1R did not affect neuroprotection obtained with SCH 58261 plus urocortin (86.0 ± 1.57%, n = 6).
The release of adenosine that occurs as a consequence of hypoxic events (Andiné et al. 1990) is accompanied by an increase in the levels of the stress regulator, corticotrophin-releasing factor (CRF) in the brain (Chen et al. 2004a, b). In addition, the in vivo modulation of adenosine A2A receptors is responsible for reversion of stress-induced effects in the hippocampus (Batalha et al. 2012), suggesting an involvement of these receptors in the stress response system. This raises the question whether A2A receptors regulate the levels or the function of the main stress mediators, either CRF or glucocorticoids. We now explored the neuroprotective effects of A2A receptors blockade and activation of CRF receptors, under stressful conditions (glutamate insult), to disclose a possible interaction between the mechanisms operated by both receptors.
The major finding of this work is that the protective action of urocortin, a CRF agonist, against a glutamate insult is dependent on both CRF receptor subtypes, CRF1R and CRF2R. Moreover, we show for the first time that the neuroprotection achieved by blockade of A2A receptors is effective only if CRF2R are active.
In hypoxic or ischemic events, the massive release of glutamate induces damage to the surrounding neuronal population, both by necrosis and apoptosis (Nicotera et al. 1999). As expected, glutamate increased cell death in a concentration-dependent manner for concentrations higher than 20 μM, which is in accordance with previous reports (Choi et al. 1987). Urocortin, a peptide belonging to the CRF family that has equivalent affinity to CRF1R and CRF2R, prevented cell death induced by glutamate (50–100 μM). For higher glutamate concentrations, urocortin seems to be inefficient in reverting cell death most probably because of the severe cell necrosis that occurs for such concentrations. On the other hand, for lower glutamate concentrations, the evoked cell death is probably not enough to allow a significant and measurable increase in cell viability induced by urocortin. The concentration of glutamate (100 μM) used in most of the experiments was enough to induce measurable cell death without causing cell detachment because of necrosis, as known to occur for high glutamate concentrations (Ankarcrona et al. 1995). On the other hand, this glutamate concentration is able to induce apoptosis, as can be concluded by the increase in apoptotic marker caspase-3.
CRF1R and CRF2R are present in neuronal cells (Pedersen et al. 2002). CRF1R but not CRF2R were shown to be involved in the neuroprotection against oxidative stress, by testing various insults in cortical neurons (Pedersen et al. 2002). Urocortin 2, a CRF2R agonist, is not able to revert cell death caused by radical oxygen species, whereas urocortin, a non-specific CRF receptor agonist is able to protect neuronal cells from an equivalent insult, by activating CRF1R (Pedersen et al. 2002). The different subtypes of CRF receptors have different functions in HPA axis and neuroprotection (Pedersen et al. 2002; Rissman et al. 2007). To determine which CRF receptor subtype is underlying the urocortin-induced neuroprotection against the glutamate insult, we analyzed the neuroprotective effect mediated by selective agonists and by urocortin in the presence of selective antagonists for each receptor subtype. As we show here, CRF receptor activation by urocortin is neuroprotective against glutamate insults only when both CRF1R and CRF2R are active, as the blockade of either of the receptors selectively was sufficient to prevent the neuroprotective effect of urocortin. This data suggest a common mechanism of action of both CRF receptors in neuroprotection against a glutamate insult and that both of them are required to afford neuroprotection.
Adenosine A2A receptors are modulatory targets against neurologic insults, as A2A receptor blockade is known to be neuroprotective (reviewed by de Mendonça et al. 2000). Besides the common intracellular signaling pathways (Blank et al. 2003; Fredholm et al. 2005), CRF receptors also share an ability to interfere with neuroprotection, (Fox et al. 1993). Therefore, we explored a possible interaction between these two receptors by studying their functional role in the prevention of cell death induced by a glutamate insult. Blockade of A2A receptors by SCH 58261 reverted cell death induced by glutamate (100 μM), as previously observed (Popoli et al. 2003). We then tested the combined actions of CRF receptors activation and A2A receptors blockade upon the glutamate insult. Remarkably, the neuroprotection afforded by A2A receptor blockade, was not additive to that provided by urocortin, suggesting a common downstream signaling pathway, shared by both receptor subtypes. Both A2A receptors and CRF receptors are able to alter gene expression. Whereas blockade of A2A receptor is neuroprotective, probably by reducing cAMP-dependent protein kinase A (PKA) and protein kinase C (PKC) phosphorylation activity (Fredholm et al. 2005), CRF receptor activation leads to an increase in PKA activity (Bayatti et al. 2005). Indeed, PKA and PKC phosphorylation leads to insertion of NMDA and AMPA receptors in the cell membrane (Tan et al. 1994; Leonard and Hell 1997; Dias et al. 2012), and this may exacerbate the excitotoxicity induced by glutamate (Leveille et al. 2008). Involvement of p38 in the neuroprotection afforded by A2A receptors blockade has also been reported (Melani et al. 2006), whereas a mechanism by which CRF receptors may affect p38 is presently unknown and awaits further investigation.
In the presence of urocortin, the neuroprotective effects of A2A receptor blockade by SCH 58261 were abolished by CRF2R but not by CRF1R blockade, suggesting that the A2A receptors role in the control of cell viability is dependent on CRF2R, but independent of CRF1R. This least investigated receptor, CRF2R, has been linked to neuroprotection against glutamate insults in retinal cells (Szabadfi et al. 2009). This is the first report regarding neuroprotection mediated by CRF2R, in the brain. Depending on the type of insult CRF may require different contributions of CRF1R and/ or CRF2R to the neuroprotection. Earlier reports in cell cultures have focused on necrosis-inducing insults, in which CRF1R seem to be the main contributors to the neuroprotection achieved by urocortin. On the contrary, a non- CRF1R dependent protection by urocortin was only observed when apoptosis-mediated cell death also occurred (see Pedersen et al. 2002). The neuroprotective role of urocortin herein described could clearly be ascribed to CRF2R activation.
Both CRF2R and A2A receptors were found to be expressed in neurons and to co-localize in the same cells in our mixed cell cultures (Fig. 1g–h). These data suggest that receptor cross talk is the most likely mechanism, but we cannot fully preclude the possibility that A2A receptor activation by glutamate-induced adenosine release could modulate CRF levels. However, the absence of significant changes in cell viability upon direct activation of A2A receptors with the agonist CGS 21680, does not seem to favor the latter possibility.
Overall, the data points to a new role of CRF against glutamate-induced neuronal death either by direct activation of CRF receptors, or modulation of A2A receptor-mediated actions. The observed neuroprotection achieved by A2A receptor blockade requires CRF2R activation. This interaction between the adenosine and CRF receptors can explain the beneficial effects of using A2A receptor antagonists against stress-induced noxious effects.
This work was supported by Fundação para a Ciência e Tecnologia. We are thankful to André Jerónimo-Santos for helpful discussions, to Ana Patricia Simões (CNC, Coimbra) for help in propidium iodide labeling technique, and to João Baião, Carla Batalha, and António Temudo for technical support. The authors declare that this research was conducted in the absence of any commercial or financial relationships that could be seen as a potential conflict of interest. All authors have contributed to and approved the final manuscript. Designed research: AMS, MJD, and LVL. Performed research: JSV, DGF, VLB, and JEC. Analyzed data: JSV, VLB, RG, and LVL. Wrote the manuscript: JSV, VLB, AMS, MJD, and LVL.