The brain distribution and functional role of glial P2X7 receptors are broader and more complex than initially anticipated. We characterized P2X7 receptors from cerebellar astrocytes at the molecular, immunocytochemical, biophysical, and cell physiologic levels. Mouse cerebellar astrocytes in culture express mRNA coding for P2X7 receptors, which is translated into P2X7 receptor protein as proven by Western blot analysis and immunocytochemistry. Fura-2 imaging showed cytosolic calcium responses to ATP and the synthetic analog 3′-O-(4-benzoyl)benzoyl-ATP (BzATP) exhibited two components, namely an initial transient and metabotropic component followed by a sustained one that depended on extracellular calcium. This latter component, which was absent in astrocytes from P2X7 receptor knockout mice (P2X7 KO), was modulated by extracellular Mg2+, and was sensitive to Brilliant Blue G (BBG) and 3-(5-(2,3-dichlorophenyl)-1H-tetrazol-1-yl)methyl pyridine (A438079) antagonism. BzATP also elicited inwardly directed nondesensitizing whole-cell ionic currents that were reduced by extracellular Mg2+ and P2X7 antagonists (BBG and calmidazolium). In contrast to that previously reported in rat cerebellar astrocytes, sustained BzATP application induced a gradual increase in membrane permeability to large cations, such as N-methyl-d-glucamine and 4-[3-methyl-2(3H)-benzoxazolylidene)-methyl]-1-[3-(triethylammonio)propyl]diiodide, which ultimately led to the death of mouse astrocytes. Cerebellar astrocyte cell death was prevented by BBG but not by calmidazolium, removal of extracellular calcium, or treatment with the caspase-3 inhibitor, benzyloxycarbonyl-Asp(OMe)-Glu(OMe)-Val-Asp(OMe)-fluoromethylketone, thus suggesting a necrotic-type mechanism of cell death. Since this cellular response was not observed in astrocytes from P2X7 KO mice, this study suggests that stimulation of P2X7 receptor may convey a cell death signal to cerebellar astrocytes in a species-specific manner.
The P2X7 receptor is a piece of the complex puzzle of nucleotide signaling, which involves various membrane receptors, ectoenzymes, and membrane transporters (Volonté and D’Ambrosi, 2009). Among P2X nucleotide receptors, the ATP-gated cation channels, P2X7 stands out for its peculiar characteristics (North, 2002). This receptor exhibits low affinity for ATP, requiring at least a 100-fold higher nucleotide concentration for activation than the other P2X receptors, and is potentiated in divalent cation-free solutions. In addition, repeated or sustained P2X7 receptor stimulation can be associated with a progressive increase in plasma membrane permeability to large molecules (up to 900 Da). Structurally, P2X7 is distinguished from other members of the ionotropic family by a long intracellular C-terminal tail containing multiple protein and lipid interaction motifs (Costa-Junior et al., 2011). The first evidence suggesting the presence of P2X7 receptors in the nervous system came from the work of Deuchars et al. (2001), who observed that 3′-O-(4-benzoyl)benzoyl-ATP (BzATP), a P2X7 receptor agonist, depolarized the membrane and increased the firing rate of spinal cord glutamatergic neurons; hence, the authors proposed that P2X7 receptors could promote transmitter release from nerve terminals. The presence of presynaptic P2X7 receptors was corroborated by many studies carried out in synaptosomes, purified neuronal cultures, and tissue slices (Sperlágh et al., 2002; Miras-Portugal et al., 2003). In addition to neurons, functional P2X7 receptors were also identified in glial cells, including astrocytes from several brain areas (Kukley et al., 2001; Panenka et al., 2001; Duan et al., 2003), Müller and Bergmann glia (Pannicke et al., 2000), oligodendrocytes (James and Butt, 2001), and microglia (Hide et al., 2000). In astroglia, P2X7 receptors have also been related to the release of gliotransmitters such as glutamate and ATP, as well as to the propagation of intercellular calcium waves (Suadicani et al., 2006; Bennett et al., 2009). P2X7 expression appears to be increased in brain injury related to ischemia and to inflammation, as well as in neurodegenerative disorders such as Alzheimer’s disease (Parvathenani et al., 2003), Huntington’s disease (Díaz-Hernández et al., 2009), multiple sclerosis (Matute, 2011), and depression (Lucae et al., 2006). In contrast to their accepted contribution to neurodegeneration, neuroprotective actions of P2X7 receptors have also been reported (Skaper et al., 2010). In rat cerebellar granule neurons, P2X7 receptors activate a survival pathway alternative to that of the classic neurotrophic factors, depending on protein kinase C activation but converging on glycogen synthase kinase-3 inhibition to promote neuronal survival (Ortega et al., 2010). P2X7 receptors could also be involved in neuronal development. Both the P2X7 receptor and the ectoenzyme tissue-nonspecific alkaline phosphatase are expressed in mouse hippocampal neurons at early stages of differentiation, thereby controlling, in a coordinated way, the elongation of the growth cone (Díez-Zaera et al., 2011).
From the above-described, a question arises regarding whether the plethora of P2X7 receptor–mediated actions involves a single type of receptor or whether several types are needed (Sánchez-Nogueiro et al., 2005). We decided to gain insight into this issue by investigating differences in P2X7 receptor signaling between cerebellar astrocytes from two rodent species, namely the rat and the mouse. We previously demonstrated that rat cerebellar astrocytes express P2X7 receptors coupled to both ionotropic and metabotropic signaling mechanisms (Carrasquero et al., 2009, 2010). Moreover, treatment with BzATP for 30 minutes induced complex changes in cell morphology but no signs of cell death. Here, we carried out the characterization of P2X7 receptors in cerebellar astrocytes from mice. Our results show that the stimulation of mouse astrocytes with millimolar ATP or micromolar BzATP concentrations elicits fast cytosolic calcium responses and plasma membrane ion currents mediated by functional P2X7 receptors. Sustained P2X7 receptor stimulation leads to a progressive increase in plasma membrane permeability to N-methyl-d-glucamine (NMDG+) and to an enhanced uptake of 4-[3-methyl-2(3H)-benzoxazolylidene)-methyl]-1-[3-(triethylammonio)propyl]diiodide (Yo-Pro-1), which ultimately compromised the cell viability. Interestingly, nucleotide-induced cell death could not be elicited in astrocytes from P2X7 receptor–deficient mice. These results indicate that P2X7 receptor orthologs mediate a fundamentally opposed effect at the cell level (morphologic changes or cell death), thus pointing to differences in the signaling mechanisms distal to ion channel opening.
Materials and Methods
Chemicals and reagents were purchased from Sigma Chemicals (Alcobendas, Spain), Gibco (Invitrogen Barcelona, Spain), Molecular Probes (Invitrogen, Barcelona, Spain), and Merck (Madrid, Spain) unless otherwise noted.
All experiments were carried out at the Complutense University of Madrid (Madrid, Spain) following the International Council for Laboratory Animal Science guidelines. The assays were designed to minimize the number of mice while maintaining statistical validity.
Littermates (male and female) of wild-type C57B1/6J (WT) and P2X7 receptor knockout (P2X7 KO) mice were used as the source of primary astrocyte cultures. P2X7 KO mice were obtained from Pfizer (Groton, CT) as generated by Solle et al. (2001). P2X7 KO mice have a deletion in the region from T1527 to T1607, located at exon 13, resulting in the disruption of the C terminus of the P2X7 receptor. These mice were fertile.
Primary Cell Cultures.
Primary astrocyte cultures were prepared from cerebella of WT and P2X7 KO mice aged 4–5 days. Six to eight cerebella were immersed in isolation buffer [140 mM NaCl, 4 mM KCl, 1.2 mM Na2HPO4, 1.5 mM MgSO4, 15 mM glucose, 10 mM HEPES, and 50 µM bovine serum albumin (BSA), pH 7.4, containing 50 U/ml penicillin and 50 μg/ml streptomycin] and cleaned of other tissues. After mechanical disaggregation, cerebellar tissue was digested in a mixture of trypsin (0.17 mg/ml) and isolation buffer over 10 minutes at 37°C, at which time the reaction was stopped with 5 ml of isolation buffer containing 0.25 mg/ml soybean trypsin inhibitor and 10 U/ml DNase I. Cells then were dispersed by gentle mechanical pipetting. Astrocytes were purified as previously described by Jiménez et al. (1999). In brief, cultures were depleted of microglial cells by 16 hours of orbital shaking, and cerebellar cells were plated at a density of 105 cells/cm2 in Dulbecco’s modified Eagle’s medium containing 10% (v/v) fetal bovine serum, 30 mM glucose, 50 U/ml penicillin, 50 μg/ml streptomycin, 100 μg/ml kanamycin, and 2.5 μg/ml amphotericin. Cells were maintained in culture at 37°C in a humidified atmosphere of 5% CO2/95% air until reaching confluence (approximately 10 days). The medium was replaced every 3 days. The purity of the cultures was determined by morphologic examination under phase-contrast microscopy and by selective staining of glial fibrillary acidic protein (GFAP) performed by immunocytochemistry. Cells were harvested by treating the culture flask with 0.03% trypsin and 0.0013% EDTA in phosphate-buffered saline (PBS). For cytosolic calcium imaging, electrophysiological recordings, and immunocytochemical experiments, cells were plated onto 15-mm diameter coverslips precoated with poly(l-lysine) (Biochrom AG, Berlin, Germany) in 35-mm Petri dishes at a density of 104 cells/cm2. For cell viability assays, cells were plated on 24-multiwell plates at a density of 9 × 104/well, whereas for real-time polymerase chain reaction (RT-PCR) and Western blot studies, cells were plated onto 10-cm diameter Petri dishes at a density of 3 × 106 cells/plate. With the exception of calcium imaging and electrophysiological and immunocytochemical experiments, cells were routinely used 48 hours after plating.
For immunofluorescence assays, cells attached to coverslips were fixed with cold 4% (w/v) paraformaldehyde for 4 minutes, washed three times with PBS, and subsequently incubated at 37°C for 1 hour in blocking solution [PBS containing 3% (w/v) BSA, 0.1% (v/v) Triton X-100, and 5% (v/v) normal goat serum]. Rabbit antibody against the C terminus of the P2X7 receptor (1:100 dilution; Alomone Labs, Jerusalem, Israel) and mouse anti-GFAP IgG (1:200 dilution; Sigma Chemicals) were used as the primary antibodies. After overnight incubation at 4°C with these antibodies and three washes in blocking solution, cells were exposed for 1 hour at 37°C to anti-mouse fluorescein isothiocyanate–conjugated secondary antibody (1:200 dilution; Sigma Chemicals) and anti-rabbit-Cy3-conjugated secondary antibody (1:400; The Jackson Laboratory, Bar Harbor, ME), and washed again three times. Cell nuclei were revealed with 4′,6-diamidino-2-fenilindol (1 µM, 5 minutes). Coverslips, mounted according to standard procedures with Prolong antifade medium (Invitrogen, Barcelona, Spain), were viewed with a TE-200 microscope (Nikon, Melville, NY) equipped with a Fluor-S ×60/1.4 oil objective and fluorescein, rhodamine, and 4′,6-diamidino-2-fenilindol filters. Images were obtained using a Kappa ACC1 camera controlled by Kappa image base control software from Kappa Optronics GmbH (Gleichen, Germany).
Quantitative RT-PCR experiments were conducted as previously described (Gómez-Villafuertes et al., 2009). Total RNA was extracted from mouse cerebellar astrocyte cultures using an RNeasy Plus Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. After digestion with DNase, total RNA was quantified and reversed transcribed using M-MLV reverse transcriptase, 6 μg of random primers, and 350 μM dNTPs (all from Invitrogen). Quantitative real-time PCR was performed using gene-specific primers and TaqMan MGB probes for mouse P2X1–5 and P2X7 receptors, as well as β-actin (all from Applied Biosystems, Foster City, CA). For the P2X6 receptor, the primers were described previously and the probe design was FAM-5′-CTTCCGTTCCTCTGGC-3′-MGB. Fast thermal cycling was performed using a StepOnePlus Real-Time PCR System (Applied Biosystems) as follows: denaturation for one cycle of 95°C for 20 seconds, followed by 40 cycles each of 95°C for 1 second and 60°C for 20 seconds. The results were normalized by parallel amplification of the endogenous control β-actin.
Western blot experiments were performed on lysates of cultured cerebellar astrocytes. Cells were washed with normal Locke’s solution (140 mM NaCl, 4.5 mM KCl, 2.5 mM CaCl2, 1.2 mM KH2PO4, 1.2 mM MgSO4, 1.5 mM glucose, and 10 mM HEPES, pH 7.4) for 1 hour at 37°C and were subsequently scraped from the dishes in 500 µl of load sample buffer containing 12.5% (v/v) glycerol, 30 mM (hydroxymethyl)aminomethane (Tris), 1% (w/v) SDS, 0.02% (w/v) bromphenol blue, and 5% (v/v) β-mercaptoethanol (pH 6.8) for a total protein concentration of approximately 1 mg/ml. The resulting cell extracts were denatured at 99°C over 4 minutes and processed (aliquots of 30 μg of protein) by SDS gel electrophoresis on 10% acrylamide gels in electrophoresis buffer (25 mM Tris, 200 mM glycine, and 0.1% SDS, pH 8.3) at 4°C. The resolved proteins were then transferred to polyvinylidene fluoride membranes (Amersham Biosciences, Piscataway, NJ) at 240 mA for 80 minutes at 4°C in transfer buffer containing 25 mM Tris, 192 mM glycine, and 20% methanol, pH 8.3. Membranes were incubated for 1 hour at room temperature with 1% Tween 20 in Tris buffer solution (100 mM NaCl and 10 mM Tris, pH 7.5) plus 5% nonfat dry milk. Blots were incubated overnight at 4°C with primary antibodies against the C terminus of the P2X7 receptor (1:100 dilution; Alomone Labs) or glyceraldehyde 3-phosphate dehydrogenase (1:1000 dilution; Ambion, Austin, TX). Secondary anti-mouse (Santa Cruz Biotechnology, Santa Cruz, CA) or anti-rabbit (Dako, Carpinteria, CA) horseradish peroxidase–conjugated antibodies were used at a 1:3000 dilution. Signals were visualized with the Super Signal Substrate (Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s recommendations and quantified by a Fluo-S Imager (Bio-Rad, Hercules, CA). To validate the suitability of the P2X7 receptor antibody, similar experiments were carried out on total lysates obtained from mouse peritoneal macrophage cultures, prepared as described by Velasco et al. (1997).
Cytosolic Calcium Imaging.
Cytosolic calcium imaging experiments were conducted basically as previously described (Carrasquero et al., 2005). Astrocytes from WT and P2X7 KO mice attached to coverslips were incubated in Locke’s solution (see “Western Blotting”) supplemented with 1 mg/ml BSA and 5–7 μM fura-2/AM (Invitrogen) for 45 minutes at 37°C. Once washed in fresh Locke’s solution, coverslips were placed in a superfusion chamber and the cells were imaged through a Nikon TE-200 microscope with a Plan Fluor ×20/0.5 objective. The incoming light was alternately set at 340 and 380 nm. Exposure time was 100 ms and the change in the wavelength from 340 to 380 nm was carried out in less than 5 ms. Images were obtained using an ORCA-ER C 47 42-80 camera from Hamamatsu (Hamamatsu City, Japan) controlled by MetaFluor 6.2r and PC software (Universal Imaging Corp., Cambridge, UK). Fluorescence data were acquired at a frequency of 2 Hz as the average signal from an elliptical region within each cell. For each wavelength, the background signal was subtracted and the ratio F340/F380 calculated.
Specific reagents were dissolved in Locke’s solution and applied to the cells by superfusion at 37°C. Cells were challenged with pulses of variable duration (30–120 seconds) of either ATP or BzATP. The increase in F340/F380 over basal values (ΔF340/F380) was used to define the amplitude of the nucleotide-induced cytosolic calcium responses. Concentration-response curves were built by plotting the ΔF340/F380 values against the log concentration of the nucleotide. Curve fitting was done by nonlinear regression with Origin-Pro 8 software (OriginLab Corporation, Northampton, MA). The effect of several P2X receptor antagonists and modulators was also assayed. Astrocytes were pretreated with 10 µM Brilliant Blue G (BBG) or 3-(5-(2,3-dichlorophenyl)-1H-tetrazol-1-yl)methyl pyridine (Donnelly-Roberts and Jarvis, 2007), two P2X7 selective antagonists, or with 30 μM pyridoxalphosphate-6-azophenyl-2′,4′-disulfonic acid (PPADS), a nonselective P2 antagonist (Bianchi et al., 1999; Jiang et al., 2000) for 3 minutes, and then superfused with 3 mM ATP in the presence of the antagonist. Calcium responses were also evaluated after the addition of 3 mM MgSO4 to the normal Locke’s solution (final concentration of 4.2 mM) and in the absence of either extracellular magnesium or calcium (by the addition of 6 mM Tris/6 mM EGTA to achieve a free calcium concentration of ~200 nM).
Patch-clamp experiments were performed as described previously (Carrasquero et al., 2009). Cells seeded at low density (104 cells/cm2) were used for patch-clamp recording so that avoiding space clamp problems attributable to gap junctions could be avoided. Electrophysiological recordings were carried out with an EPC9 patch-clamp amplifier using Pulse/PatchMaster software (HEKA Electronic, Lambrecht, Germany). Pipettes were pulled from Kimax borosilicate glass (Witz Scientific, Holland, OH) and subsequently wax-coated and fire-polished to obtain a final resistance of 2–3 MΩ when filled with standard solutions. The extracellular (bath) solution contained the following: 140 mM NaCl, 1 mM CaCl2, 10 mM HEPES, and 10 glucose (pH 7.2, adjusted with NaOH; approximately 300 mOsm). Recording pipettes were filled with a solution containing the following: 140 mM NMDG+, 5 mM EGTA, and 10 mM HEPES (pH 7.2, adjusted with HCl; approximately 290 mOsm). Cells attached to glass coverslips were transferred to a recording chamber placed in the stage of an inverted Zeiss Axiovert 100 microscope (Carl Zeiss, Oberkochen, Germany) and continuously superfused with bath solution (perfusion rate of 1 ml/min).
Membrane currents were measured in the whole-cell configuration of the patch-clamp technique, filtered at 3 kHz, and sampled at 10 KHz. Once established electrical access to the cytoplasm, cells were held at a voltage (Vh) of −70 mV. Series resistance (5–10 MΩ) was compensated by 80% and monitored throughout the experiment together with the cell membrane capacitance. Considering that NMDG+ has a Mr of 195.21 Da, a series resistance of 10 MΩ, and an astrocyte’s membrane capacitance of 17.6 ± 1.42 pF (n = 51 cells), we estimated in approximately 10 minutes the time required by NMDG+ to equilibrate between the patch pipette and the cytoplasm of the recorded cell (Pusch and Neher, 1988). Because of this, drug application started at least 10 minutes after obtaining the whole-cell configuration, and experiments in which series resistance changed by more than 20% or holding current exceeded 20 pA were not analyzed. All recordings were obtained at room temperature (21–25°C).
Ligand-gated currents were activated by P2X receptor agonists applied onto the cell under investigation by means of a puffer glass-pipette (3–5 µm tip diameter, MPCU; List Electronics, Darmstadt, Germany) or a gravity-driven perfusion system with five independent lines controlled by electronic valves (The Lee Company, Westbrook, CO). Stock solutions of drugs were diluted daily in extracellular saline and incorporated into the perfusion system a few minutes before starting the experiments.
Time-dependent changes in the relative permeability of NMDG+ with respect to Na+ (PNMDG/PNa) of P2X7 receptors were computed from changes in the reversal potential (Vrev) of BzATP-induced currents, assuming ion activities of 0.75 for both cations (Nörenberg et al., 2011), by a transform of the Goldman-Hodgkin-Katz equation for bi-ionic conditions. Vrev values from individual cells were corrected for the liquid junction potential between the pipette’s solution and the extracellular solution, which was calculated to be −6.37 mV by using the Patcher’s Tools module included in Igor-Pro software (WaveMetrics, Inc., Lake Oswego, OR).
Measurements of Yo-Pro-1 uptake were performed in astrocyte cell populations. Astrocytes were collected by trypsinization from confluent cultures, washed, and resuspended in Mg2+-free Locke’s solution. Measurements were made in a Perkin-Elmer LS-50B fluorometer from 1.5-ml samples containing approximately 300,000 cells and 1 µM Yo-Pro-1 iodide (Molecular Probes) in stirred cuvettes at 37°C. Fluorescence was excited with 490-nm light and detected through a 510-nm centered slit. BzATP was added to the cuvettes from at least 100-fold concentrated stock solutions to avoid large volume variations. When the effect of P2X7 receptor antagonists was tested, cells were preincubated with the antagonist for 3 minutes before the addition of BzATP.
Cell Viability Assays.
Cell viability was evaluated with the LIVE/DEAD Viability/Cytotoxicity Kit for mammalian cells according to the manufacturer’s instructions. The kit combines two fluorescent probes, calcein AM, and ethidium homodimer-1, allowing the simultaneous determination of live and dead cells. Following cleavage by intracellular esterases, the polyanionic dye calcein is retained within live cells, producing an intense uniform green fluorescence, whereas ethidium homodimer-1 enters cells with damaged membranes and upon binding to nucleic acids produces a bright red fluorescence. Confluent WT or P2X7 KO cerebellar astrocytes, seeded in 24-well plates, were washed for 15 minutes with 1 ml normal Locke’s solution before being incubated with 3 mM ATP or 500 μM BzATP for 30 minutes at 37°C. At the end of the stimulation period, the incubation medium was replaced with fresh Locke’s solution, and the labeled cells were viewed under a fluorescence microscope equipped with a conventional fluorescein long-pass filter. All experiments were done in duplicate and repeated three times.
Cell viability was also determined by the 3-(4,5-dimethythiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay, which assesses mitochondrial function. Confluent astrocytes were first washed for 15 minutes with 1 ml of normal Locke’s solution warmed at 37°C. Cells were then treated with 5 mM ATP or 500 µM BzATP for 60 minutes. Nonstimulated cells were used as controls. After drug treatment, incubation medium was removed and the tetrasodium salt MTT (Sigma Chemicals) was added to a final concentration of 0.5 mg/ml in Locke’s solution and maintained for 2 hours at 37°C. An equal volume of MTT solubilization solution (10% Triton X-100 plus 0.1 N HCl in anhydrous isopropanol) was then added, and the mix was kept for 30 minutes at room temperature with orbital shaking. Cellular extracts were collected and measured spectrophotometrically at 570 nm. In some wells, the effects of P2X7 antagonists (BBG or calmidazolium) or the caspase 3 inhibitor II benzyloxycarbonyl-Asp(OMe)-Glu(OMe)-Val-Asp(OMe)-fluoromethylketone (Z-DEVD-FMK), administered 3 minutes before and during nucleotide incubation, were also evaluated. Values were normalized with respect to those obtained in nonstimulated cells.
Results are expressed as means ± S.E.M. calculated from at least three experiments usually performed on cells from different cultures. Statistical differences between two groups of data were assessed by the t test, and a P value < 0.05 was taken as the limit of significance. When multiple comparisons were made, one-way analysis of variance was used and Dunnett’s post test analysis was applied only when a significant (P < 0.05) effect was indicated by one-way analysis of variance (GraphPad Prism 5; GraphPad Software Inc., San Diego, CA).
Characterization of Cerebellar Astrocyte Cultures from Neonatal Mice.
Our study began by determining the age of mice rendering optimal cerebellar astrocyte cultures. Cerebellar cells obtained from P6/P7 mice grew slowly and tended to form dense multilayered aggregates. The confluent state was reached between 15 and 20 days in vitro, when most cells showed a dark-gray appearance and marked cytoplasmic granularity. In addition, P7 cells displayed small and inconsistent cytosolic calcium responses to purinergic agonists. These drawbacks were solved by using younger pups (P4–P5) and by plating the cells at higher density (105 cells/cm2). Because tissues from younger animals were more delicate, it was also necessary to use a low trypsin concentration (0.17 mg/ml) to avoid cell damage. Astrocyte cultures from P4 pups were found to have a good morphologic appearance and exhibited prominent cytosolic calcium responses to nucleotides. As for P6/P7 cultures, growing cells from P4 mice tended to form discrete multicellular groups. Smaller cell groups were easier to grow compared with isolated cells, which had slower growing and higher dying rates. The astrocytes used in this study formed a confluent layer of process-bearing cells after 10–12 days in culture. More than 98% of these cells expressed the astrocyte-specific marker GFAP (Fig. 1, A and C).
The same protocol was used to obtain purified astrocytic cultures from P2X7 KO mice. We did not observe any significant differences between WT and P2X7 KO cultures with regard to growing rate and morphology.
P2X7 Receptors Are Expressed in Mouse Cerebellar Astrocytes.
The expression of P2X7 receptor was investigated at the transcriptional and protein levels. Figure 1B shows that WT cerebellar astrocytes express mRNAs coding for P2X4, P2X5, and P2X7 receptor subunits. P2X4 receptor mRNA was the most abundant. Other ionotropic purinergic receptor subunits (P2X1, 2, 3, and 6) were not detected in mouse cerebellar astrocytes. The presence of P2X7 protein was analyzed by immunocytochemistry (Fig. 1C). Most of the GFAP-positive astrocytes exhibited a punctate immunolabeling, which appeared to be located at specific zones of the plasma membrane. The expression of P2X7 subunits was also corroborated by Western blot experiments (Fig. 1D). Immunotransference experiments in WT cerebellar astrocytes revealed a 77-kDa band that was not detected in extracts from P2X7 KO cultures. To validate the anti-P2X7 receptor antibody used in immunocytochemical experiments and the specificity of the band detected in WT cerebellar astrocyte extracts, we checked its presence in native immune cells, which express high levels of P2X7 receptors. As expected, a similar band (77 kDa) was detected on total lysates from mouse peritoneal macrophages, which would correspond to the full length of P2X7 subunits (isoform P2X7A).
Nucleotide-Evoked Calcium Responses in Mouse Cerebellar Astrocytes: Involvement of P2X7 Receptors.
Once the presence of P2X7 protein in WT cerebellar astrocytes was established, its functional properties were assessed. Cells were stimulated with the nucleotides α,β-methylene-ATP (α,β-meATP), ATP, CTP, ATP, UTP, and BzATP and the corresponding calcium responses were recorded at the single-cell level. Neither α,β-meATP (100 μM), which is selective for P2X1/3 subunits, nor CTP (100 μM), which activates P2X4 subunit-containing receptors, were able to modify the cytosolic free calcium concentration in mouse cerebellar astrocytes (data not shown), whereas practically all tested cells responded to 100 μM ATP, UTP, or BzATP stimulation, with a rise in cytosolic calcium (Figs. 2 and 3). Long stimulations (120 seconds) with 100 μM ATP produced an initial elevation of cytosolic calcium (ΔF340/F380 of 0.48 ± 0.01; n = 300 cells) that decayed with a biphasic time course (Fig. 2A). Interestingly, a 10-fold higher ATP concentration (1 mM) elicited calcium responses with more complex kinetics, consisting of an initial peak (ΔF340/F380 of 0.65 ± 0.01; n = 300 cells) followed by a sustained phase or by a slow increase in cytosolic calcium. The amplitude of the initial peak and the maximum value of the second phase of the calcium responses to different ATP concentrations (100 nM to 5 mM) were used to obtain the corresponding concentration-response curves. The EC50 value for the initial transient phase was 33 ± 3 μM, whereas that of the second component was 1.09 ± 0.11 mM (Fig. 2B). A similar phenotype was observed in the calcium responses evoked by BzATP, a synthetic agonist that displays greater potency than ATP at the P2X7 receptors (Fig. 3A). Therefore, the EC50 values for the two components of the calcium response to BzATP were 39 ± 3 and 230 ± 21 μM, respectively (Fig. 3B).
A hallmark of the P2X7 receptor is its sensitivity to extracellular divalent cations (Virginio et al., 1997; Acuña-Castillo et al., 2007). To determine whether nucleotide calcium responses were mediated by the P2X7 receptor, we first assessed their dependence on extracellular calcium. Removal of this cation from the perfusion medium ablated the second phase of calcium responses elicited by either ATP (3 mM) or BzATP (500 µM), whereas the initial transient component was significantly preserved (Fig. 4). Modifying the concentration of extracellular Mg2+, either by removing or increasing the Mg2+ content in the Locke’s solution, also produced a change in the shape of the calcium responses evoked by ATP (3 mM). Removing Mg2+ from the perfusion solution was associated with a potentiation of the second component of the response, which occluded the initial transient one. On the contrary, elevating the concentration of Mg2+ from 1.2 to 4.2 mM produced a marked reduction of the second component (Fig. 5A). These results strongly suggest the involvement of P2X7 receptors in ATP-induced calcium responses.
We next analyzed the effect of P2X7 receptor antagonists on calcium responses evoked by 3 mM ATP. Preincubation of cerebellar astrocytes for 3 minutes with 10 μM BBG abolished the second component of calcium responses, whereas the transient component (ΔF340/F380 = 0.43 ± 0.01; n = 300 cells) was mostly preserved (Fig. 5B). Another P2X7 antagonist, 3-(5-(2,3-dichlorophenyl)-1H-tetrazol-1-yl)methyl pyridine (A438079; 10 μM), also depressed the persistent component of the calcium response to ATP while barely affecting the initial transient component. These data corroborated that the extracellular calcium-dependent sustained component of the calcium response to ATP was mediated by P2X7 receptors. We also tested the effect of the broad P2 antagonist, PPADS. Preincubation with PPADS (30 μM) strongly depressed ATP calcium responses by reducing both the initial and the sustained phases of the response (Fig. 5B).
P2X7 Receptors Mediate Membrane Current Responses Evoked by Nucleotides in Mouse Cerebellar Astrocytes.
To get direct evidence supporting the involvement of P2X7 receptors in nucleotide-evoked calcium responses from mouse cerebellar astrocytes, we conducted patch-clamp experiments to detect the ion flux mediated by this sort of ionotropic receptor. In the first series of experiments, cells (n = 9) were consecutively challenged with 100 µM BzATP, 100 µM ATP, and 1 mM ATP (3-second pulses, at 2-minute intervals). As expected from calcium imaging results, 100 µM BzATP and 1 mM ATP applications were associated with the appearance of inwardly directed currents in all cells tested, whereas 100 µM ATP was not able to evoke any net inward current, thus implying a pure metabotropic action at this concentration on mouse cerebellar astrocytes (Fig. 6A). Currents evoked by 1 mM ATP were less than half (72.85 ± 21.2 pA; n = 9 cells) of those activated by BzATP in the same cells (171.90 ± 36.47 pA), suggesting a lower potency of the endogenous agonist at this ionotropic purinergic receptor (Young et al., 2008; Nörenberg et al., 2010; Oliveira et al., 2011). Peak currents evoked by BzATP in the entire group of cells tested were of 170.11 ± 18.30 pA (n = 42 cells), which considering an average membrane capacitance of 16.8 ± 0.9 pF, implies a mean current density of approximately 10 pA/pF. The contribution of P2X7 receptors to BzATP-induced currents was assayed by evaluating their sensitivity to BBG (10 µM), calmidazolium (1 µM), and Mg2+ (1 mM) added to the perfusion medium. Each of these three maneuvers (applied 2 minutes before and during agonist administration) reduced by approximately 75% the amplitude of the nucleotide-evoked current, hence pointing to the P2X7 receptor as the underlying channel (Fig. 6B).
Sustained Stimulation of P2X7 Receptors Increases Plasma Membrane Permeability to Large Cations in Mouse Cerebellar Astrocytes.
The P2X7 receptor is remarkable from the biophysical point of view in that brief stimulation opens a nonselective channel permeable to small ions (Na+, K+, Ca2+) (Carrasquero et al., 2009), but prolonged stimulation leads to a progressive increase in membrane permeability with the opening of a pore through which large cations, such as NMDG+ (Mr = 196) and Yo-Pro-1 (Mr = 629), can pass in and out of the cells (Chaumont and Khakh, 2008; Richler et al., 2008). Therefore, we set out to investigate whether an increase in membrane permeability to NMDG+ occurred during prolonged stimulation (90 seconds) of P2X7 receptors with BzATP (100 µM) by monitoring the shift in Vrev under bionic conditions (Na+ and NMDG+ being the only charge carriers in the extracellular and intracellular solution, respectively). Vrev was estimated from the current response to voltage ramps (from −100 to +50 mV over 1 second) repetitively applied every 3 seconds. Current ramps were leak-subtracted with the averaged current to 3 voltage ramps generated prior to BzATP application, and the measured Vrev were corrected by the liquid junction potential between the pipette’s solution and the extracellular solution (−6.37 mV) (Fig. 6C). In the continuous presence of BzATP, Vrev shifted from 29.71 ± 2.09 mV to a steady state value of 15.87 ± 0.16 mV, and this corresponding to a change in the permeability ratio PNMDG/PNa from 0.26 ± 0.02 to 0.49 ± 0.003 (n = 8 cells). Thus, long-lasting stimulation of P2X7 receptors is associated with the opening of a large-cation-permeable pore in mouse cerebellar astrocytes.
An alternative way of testing the increase in large-cation permeability brought about by a sustained stimulation of P2X receptors is to measure the cellular uptake of cationic dyes, such as Yo-Pro-1. Fig. 7, A and B, shows that mouse cerebellar astrocytes took up Yo-Pro-1 upon stimulation with 500 µM BzATP in a time-dependent manner. To confirm the involvement of P2X7 receptors in Yo-Pro-1 uptake, we examined the effect of BBG and calmidazolium. As shown in Fig. 7C, 10 µM BBG had no effect on basal uptake but completely blocked Yo-Pro-1 uptake elicited by BzATP. Interestingly and in contrast with its inhibitory effect on BzATP-evoked currents (Fig. 6B), 1 µM calmidazolium did not affect Yo-Pro-1 uptake in mouse cerebellar astrocytes, indicating that this cation dye was entering into the cell through nonselective membrane pores, which were formed after P2X7 receptor stimulation.
To further prove that Yo-Pro-1 uptake is an event secondary to P2X7 receptor stimulation, we investigated nucleotide responses in P2X7 KO cerebellar astrocytes. Figure 8 shows calcium responses and Yo-Pro-1 uptake obtained in cerebellar astrocytes from WT and P2X7 KO mice. As could be expected from Western blot experiments (Fig. 1D), P2X7 KO cerebellar astrocytes bathed in normal Locke’s solution exhibited cytosolic calcium responses to ATP (3 mM) or BzATP (500 µM) that displayed only a transient component (Fig. 8A). Peak values were similar for ATP (ΔF340/F380 of 0.43 ± 0.01; n = 100 cells) and BzATP stimulations (ΔF340/F380 of 0.40 ± 0.01; n = 120 cells), and compared well with those observed in WT cells (0.45 ± 0.02 for 3 mM ATP and 0.48 ± 0.03 for 500 µM BATP; n = 150 cells). Likewise, P2X7 KO astrocytes responded to UTP with cytosolic calcium increases similar to those observed in WT cells (results not shown). In P2X7 KO cells, application of BzATP (500 µM, 60 minutes) did not induce Yo-Pro-1 uptake. No difference was observed between data obtained from nonstimulated cells and cells treated with the nucleotide. These results further support the involvement of the P2X7 receptor in changes in plasma membrane permeability triggered by BzATP in mouse cerebellar astrocytes.
P2X7 Receptors Mediate Cell Death in Mouse Cerebellar Astrocytes.
Considering that prolonged activation of P2X7 receptors increased the membrane permeability to large molecules in mouse cerebellar astrocytes, we decided to test whether P2X7 receptor stimulation could also affect the viability of the cultures. As shown in Fig. 8C, treatment of WT cerebellar astrocytes with 3 mM ATP or 500 μM BzATP for 30 minutes considerably decreased cell viability as determined with the LIVE/DEAD Viability/Cytotoxicity Assay Kit. Importantly, similar treatments did not affect the viability of P2X7 KO cells, hence indicating that P2X7 receptors were engaged in the nucleotide-induced death of WT cerebellar astrocytes.
To more easily quantify the cell death response to nucleotides, we used the MTT assay, which estimates the number of viable cells by measuring the cellular metabolic state. As shown in Fig. 9A, a 60-minute treatment of astrocytes with high concentrations of ATP (5 mM) or BzATP (500 µM) led to a 40% reduction in cell viability. Importantly, the P2X7 antagonist BBG (10 µM), administered 3 minutes before and during the subsequent 60-minute incubation, prevented cell death induced by 500 µM BzATP, and markedly attenuated the diminution in cell viability produced by 5 mM ATP. At variance, calmidazolium (1 µM) was unable to inhibit cell death induced by either BzATP or ATP. It should be noted that neither BBG nor calmidazolium affected cerebellar astrocytes viability in the absence of nucleotides. Cell death elicited by 500 µM BzATP did not depend on either extracellular calcium or the activation of caspase 3 since it was not modified by removing calcium from the Locke’s solution or incubation with 30 µM Z-DEVD-FMK, a caspase-3 inhibitor (Fig. 9B). Altogether, our results point to a necrotic-type mechanism for the cell death elicited by P2X7 receptor stimulation in mouse cerebellar astrocytes (Fig. 9B).
The central message of this study is that mouse cerebellar astrocytes express functional P2X7 receptors, which are ultimately responsible for a nucleotide-induced cell death effect. This conclusion is supported by the following findings. First, the P2X7 subunit 77-kDa protein was detected by immunocytochemistry and Western blot analysis. Second, both ATP and BzATP induced membrane ionic currents and ionotropic cytosolic calcium responses, with BzATP being a more potent agonist. Third, nucleotide-induced membrane currents and inotropic calcium responses were sensitive to extracellular Mg2+ and to P2X7 receptor antagonists. Fourth, prolonged exposure of cerebellar astrocytes to high concentrations of nucleotides increased the permeability of plasma membrane to large cations and reduced cell viability, and both effects were prevented by treatment with BBG. Finally, cellular responses attributed to P2X7 receptors were absent in P2X7 KO cerebellar astrocytes.
This study also reports the preparation of purified astrocyte cultures from mice cerebella. In our hands, P4 was the best age from which cerebellar cells could be dissociated to initiate astrocyte cultures. Isolated cells typically reached confluence after 10–12 days in culture, and thereafter consistently exhibited cytosolic calcium responses to purinergic agonists. Likewise, we did not observe any obvious difference between WT and P2X7 KO cultures.
Mouse cerebellar astrocytes responded to ATP and BzATP stimulations with cytosolic calcium elevations whose kinetics depended on the nucleotide concentration used. Low concentrations of nucleotides (≤100 µM) induced transient calcium responses, which declined in the presence of the agonist, but biphasic calcium responses were obtained when the agonist concentrations were increased. The two phases of the calcium response evoked by ATP differed in their sensitivity to external divalents (Ca2+ and Mg2+) and to P2X7 receptor antagonists. The initial transient phase was mainly metabotropic in nature and was practically insensitive to P2X7 receptor antagonists, which indicates the involvement of a P2Y metabotropic receptor, at which ATP and BzATP were equipotent. In fact, similar calcium responses were elicited by UTP (Wildman et al., 2003; León-Otegui et al., 2011). By contrast, the second, sustained or slowly rising phase of the calcium response, observed only at high (millimolar) concentrations of ATP, depended on extracellular Ca2+, and was inhibited by extracellular Mg2+ and P2X7 receptor antagonists (BBG and A438079); all of these features support the involvement of P2X7 receptors in its generation (Virginio et al., 1997; Hibell et al., 2001; Acuña-Castillo et al., 2007). Moreover, the EC50 values estimated for this component of the calcium response (1.09 mM and 230 µM for ATP and BzATP, respectively) agree well with those reported for the cloned mouse P2X7 receptor (Chessell et al., 1998; Hibell et al., 2001; Young et al., 2007). Confirmatory evidence for the participation of P2X7 receptors was obtained in astrocytes from P2X7 KO mice: nucleotide calcium responses recorded from P2X7 KO cells lack the second component, such that their shape resembled calcium responses obtained in WT astrocytes in the absence of extracellular calcium.
Biphasic nucleotide calcium responses observed in mouse astrocytes were similar to those elicited by BzATP in rat cerebellar astrocytes (Carrasquero et al., 2009). In the rat model, however, ATP was unable to evoke a sustained calcium elevation like that elicited by the synthetic P2X7 agonist. Considering that the rat P2X7 receptor is more sensitive to ATP than the mouse ortholog (Young et al., 2007), our results suggest a reduced expression level of P2X7 receptors in rat cerebellar astrocytes compared with their murine counterparts.
The electrophysiological results also corroborate the involvement of P2X7 receptors in nucleotide responses. In mouse cerebellar astrocytes, nondesensitizing currents were elicited by the two nucleotides, and, as expected, BzATP displayed higher potency than ATP (Duan et al., 2003; Nörenberg et al., 2010, 2011). The magnitude and the current density of BzATP-induced currents in mouse cerebellar astrocytes (approximately 30 pA/pF) were much higher than those from rat cerebellar astrocytes (approximately 3 pA/pF) (Carrasquero et al., 2009). Another distinctive feature of P2X7 receptor–mediated currents from mouse cerebellar astrocytes is the progressive increase in the fractional permeability to NMDG+ during prolonged BzATP stimulation, which contrasts with the lack of such an increase observed in HEK-293 cells expressing rat P2X7 receptors when the extracellular solution contains the normal Na+ concentration (Jiang et al., 2005; Yan et al., 2008). This difference likely rests on the divergence in the sequence of the C-terminal tail region between the rat and mouse P2X7 receptor orthologs and on the accepted role of this intracellular domain in controlling plasma membrane permeability after P2X7 receptor stimulation (Chessell et al., 1998; Costa-Junior et al., 2011). Species difference between rat and mouse P2X7 receptors may also be responsible for differences in the ability to accumulate Yo-Pro-1 between rat and mice cerebellar astrocytes (Smart et al., 2003). Our results clearly show that Yo-Pro-1 uptake develops rapidly in mice cerebellar astrocytes upon BzATP stimulation. In contrast, we found no dye uptake in rat cerebellar astrocytes under the same experimental conditions (Carrasquero et al., 2009). Most likely, opening of a membrane pore permeable to large molecules underlies the cytotoxic effect of prolonged P2X7 receptor stimulation in mice cerebellar astrocytes. Three complementary lines of evidence support this assertion. First, neither accumulation of Yo-Pro-1 nor reduction of cell viability was observed in P2X7 KO cerebellar astrocytes upon incubation with BzATP. Second, in WT cerebellar astrocytes, BBG inhibited BzATP-induced ionic currents, as well as Yo-Pro-1 uptake and cell death. Third, calmidazolium, which is known to block P2X7 receptor–associated small cation channels without affecting the formation of large membrane pores (Virginio et al., 1997), had no effect on both Yo-Pro-1 uptake and cell death despite inhibiting BzATP-activated ionic currents in WT astrocytes. Hence, altogether these results point to a close relationship between P2X7 pore formation and cytotoxicity of mice cerebellar astrocytes. Moreover, cell death evoked by P2X7 receptor stimulation appears to be independent of extracellular Ca2+ and presumably Ca2+ entry as well and is not related to caspase 3 activation, thereby suggesting a necrotic-type mechanism of cell death. Necrotic death has long been considered an unregulated form of cell death, resulting in rupture of the plasma membrane and a rapid lysis of the cell. However, recent studies revealed that necrosis can also be regulated through specific intrinsic cellular programs, involving a new serine/threonine kinase family, the receptor interacting protein kinases (Festjens et al., 2007). In this respect, signaling pathways initiated by P2X7 receptor stimulation and leading to cell death in mouse cerebellar astrocytes will certainly deserve additional investigations.
Results obtained in mouse cerebellar astrocytes also contrast with those reported for other noncerebellar rat astrocytes. In rat cortical astrocytes, stimulation of P2X7 receptors induces a state of reversible growth arrest rather than cell death (Neary et al., 2008). Likewise, injection of BzATP into rat brain did not cause a decrease in the number of either GFAP-positive cells or GFAP/BrdU-proliferating cells (Franke et al., 2007). As suggested by our studies in rat and mouse cerebellar astrocytes, different cellular consequences of P2X7 receptor stimulation may be due to different receptor expression levels (Hickman et al., 1994; Narcisse et al., 2005; Nagasawa et al., 2009), but certainly also to differences in amino acid sequence at the C-terminal tail domain with ensuing differences in the ability to induce membrane permeabilization either directly or through the interaction with other signaling proteins expressed in an species-dependent manner (Costa-Junior et al., 2011).
Mouse models (WT) and protein-null animals (KO mice) are extensively used to better understand the function and pathophysiological implications of proteins. Studies carried out with P2X7 KO mice have revealed that P2X7 receptors play important roles under physiologic and pathologic conditions (Solle et al., 2001; Chessell et al., 2005; Fulgenzi et al., 2008), and have also suggested the existence of analogous proteins that escape from gene inactivation in the nervous system. Previous studies conducted in mouse cerebellar granule neurons from P2X7 KO mice revealed the existence of functional P2X7-like receptors, which probably correspond to a functional splice variant that obviates gene inactivation (Sánchez-Nogueiro et al., 2005; Nicke et al., 2009). The data presented here clearly demonstrate that WT cerebellar astrocytes express functional P2X7 receptors, which are absent from P2X7 KO cells. P2X7 KO cerebellar astrocytes do not express the P2X7 receptor protein nor exhibit any response that could be attributed to this receptor. This validates the P2X7 KO mouse model in astroglia (Suadicani et al., 2006).
A question that still remains open is the role of P2X7 receptors in astrocyte physiology. Although high extracellular ATP concentrations sufficient to activate P2X7 receptors can be reached at the extracellular space, the crucial issue is how long they remain high enough. Under physiologic conditions, ectoenzymes present in the same or surrounding cells will rapidly breakdown ATP, thereby terminating their actions on P2X7 receptors. However, under pathologic situations, such as brain ischemia or trauma there is a long-lasting ATP accumulation, leading to persistent activation of their receptors (Verkhratsky et al., 2009; Franke et al., 2012). Undoubtedly, brief stimulation or persistent activation of P2X7 receptors has different cellular implications. Data reported here show that persistent P2X7 stimulation might result in a species-related manner in a cell death signal to mouse cerebellar astrocytes. In addition to provide novel insight into P2X7 receptor signaling in cerebellar astroglia, our results may thus serve as a reminder of the importance of considering species differences when selecting experimental models for pathophysiological and drug-screening research programs. In this regard, one should keep in mind that heterologously expressed human, mouse, and rat P2X7 receptors all have been reported to show differences at the associated ion channel level with respect to agonist and antagonist potencies and with regard the kinetics of formation of the large membrane pores, which could be translated into larger differences in cellular responses as a function of the particular tissue in which they are naturally expressed (Chessell et al., 1998).
Participated in research design: Carrasquero, Artalejo, Miras-Portugal, Delicado.
Conducted experiments: Salas, Carrasquero, Olivos-Oré, Bustillo, Delicado.
Performed data analysis: Salas, Carrasquero, Olivos-Oré, Bustillo, Artalejo, Miras-Portugal, Delicado.
Wrote or contributed to the writing of the manuscript: Salas, Artalejo, Miras-Portugal, Delicado.
- Received September 6, 2013.
- Accepted October 7, 2013.
This research was supported by the Spanish Ministry of Economy and Competitiveness [Grants BFU2011-26253 and BFU2011-24743] and the Spanish Ion Channel Initiative [Grant CSD2008-00005]. E.S. was the recipient of a research fellowship at the University of Costa Rica and the Fundación Carolina.
- 3-(5-(2,3-dichlorophenyl)-1H-tetrazol-1-yl)methyl pyridine
- Brilliant Blue G
- bovine serum albumin
- glial fibrillary acidic protein
- 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
- phosphate-buffered saline
- pyridoxalphosphate-6-azophenyl-2′,4′-disulfonic acid
- real-time polymerase chain reaction
- Copyright © 2013 by The American Society for Pharmacology and Experimental Therapeutics