P2X7 Receptor Activation Induces Reactive Oxygen Species Formation and Cell Death in Murine EOC13 Microglia

The P2X7 purinergic receptor is a ligand-gated cation channel expressed on leukocytes including microglia. This study aimed to determine if P2X7 activation induces the uptake of organic cations, reactive oxygen species (ROS) formation, and death in the murine microglial EOC13 cell line. Using the murine macrophage J774 cell line as a positive control, RT-PCR, immunoblotting, and immunolabelling established the presence of P2X7 in EOC13 cells. A cytofluorometric assay demonstrated that the P2X7 agonists adenosine-5′-triphosphate (ATP) and 2′(3′)-O-(4-benzoylbenzoyl) ATP induced ethidium+ or YO-PRO-12+ uptake into both cell lines. ATP induced ethidium+ uptake into EOC13 cells in a concentration-dependent manner, with an EC50 of ~130 μM. The P2X7 antagonists Brilliant Blue G, A438079, AZ10606120, and AZ11645373 inhibited ATP-induced cation uptake into EOC13 cells by 75–100%. A cytofluorometric assay demonstrated that P2X7 activation induced ROS formation in EOC13 cells, via a mechanism independent of Ca2+ influx and K+ efflux. Cytofluorometric measurements of Annexin-V binding and 7AAD uptake demonstrated that P2X7 activation induced EOC13 cell death. The ROS scavenger N-acetyl-L-cysteine impaired both P2X7-induced EOC13 ROS formation and cell death, suggesting that ROS mediate P2X7-induced EOC13 death. In conclusion, P2X7 activation induces the uptake of organic cations, ROS formation, and death in EOC13 microglia.


Introduction
Microglia are the resident innate immune cells of the central nervous system (CNS) and play an important role in immune surveillance [1] and in the pathogenesis and progression of a number of CNS disorders [2]. Microglia are constantly mobile, spending time scanning the extracellular space of the CNS [1]. In response to brain injury or immunological stimuli, these cells become activated and undergo dramatic morphological and functional changes, which are highly dependent on the context of their activation [3]. Activated microglia phagocytose debris and peptides, present antigens, and produce a number of soluble factors. ese factors may be in�ammatory, regulatory, or cytotoxic in nature and include reactive oxygen species (ROS), nitric oxide (NO), proin�ammatory and anti-in�ammatory cytokines, prostaglandins, and growth factors [4,5]. Microglia can be both neuroprotective or neurotoxic when activated, depending on the factors they produce and the quantity and context in which they are released, with prolonged or excessive activation of these cells being associated with neuroin�ammation and the progression of a number of CNS disorders [6]. e mechanisms behind enhanced microglial activation in these disorders and the features determining the balance between neuroprotection and neurotoxicity are not fully understood.

Cell
Lines. e murine macrophage J774 cell line, the murine microglial EOC13 cell line, and the murine lymphoblast LADMAC cell line, all originally obtained from the American Type Culture Collection (Manassas, VA), were kindly provided by Jasmyn Dunn (University of Queensland, Brisbane, Australia) (J774) and Iain Campbell (University of Sydney, Sydney, Australia) (EOC13 and LADMAC). J774 cells were maintained in RPMI-1640 medium containing 10% FBS and 2 mM GlutaMAX (complete RPMI medium). EOC13 cells were maintained in DMEM/F12 supplemented with 10% FBS, 2 mM GlutaMAX, and 20% LADMAC conditioned medium (complete DMEM medium). Cell lines were maintained at 37 ∘ C and 95% air/5% CO 2 and passaged every 3-4 days. Quarterly mycoplasma testing was carried out using the MycoAlert Mycoplasma Detection Kit (Lonza, Rockland, ME), as per the manufacturer's instructions. For experiments, cells were harvested by cell scraping unless otherwise stated.

Fluorescent Cation Dye Uptake
Assay. Cells were washed in NaCl medium (300 × for 5 min), resuspended in NaCl medium, and equilibrated at 37 ∘ C for 5 min (1 × 10 5 cells/1 mL/tube). Cells were then incubated with 25 M ethidium + (or 1 M YO-PRO-1 2+ where indicated) in the absence or presence of the P2X7 agonists ATP or BzATP (as indicated) for 5 min. In some experiments, ATP-induced cation uptake was assessed with cells suspended in KCl medium (150 mM KCl, 5 mM glucose, and 10 mM HEPES, pH 7.4) or in NaCl medium containing 1 mM CaCl 2 or 100 M EGTA. In other experiments, cells were preincubated in the absence or presence of P2X7 antagonists or the ROS scavenger NAC (as indicated) for 15 and 30 min, respectively, prior to cation and ATP addition. Incubations with nucleotides were stopped by the addition of an equal volume of ice-cold NaCl medium containing 20 mM MgCl 2 (MgCl 2 medium) followed by centrifugation (300 × for 5 min). Cells were washed once with NaCl medium and events collected using a LSR II �ow cytometer (BD Biosciences, San Diego, CA) (excitation 488 nm, emission collected with 575/26 and 515/20 band-pass �lters for ethidium + and YO-PRO-1 2+ , resp.). e mean �uorescence intensity (MFI) of relative cation uptake was determined using FlowJo soware (Tree Star, Ashland, OR).

Cell Surface P2X7 Protein Detection by Flow Cytometry.
Cells in NaCl medium containing 10% NHS and 0.02% NaN 3 (1 × 10 5 cells/200 L/tube) were incubated with anti-P2X7 or rat IgG2b isotype control mAb (5 g/mL) at room temperature for 30 min. Cells were then washed twice with NaCl medium (300 × for 5 min) and incubated with APCconjugated anti-rat IgG Ab (1.3 g/mL) and 7AAD (to exclude dead cells) for 30 min protected from light. Cells were washed once as above. Events were then collected using a LSR II �ow cytometer (excitation 633 nm, emission collected with 660/20 band-pass �lter for APC; excitation 488 nm, emission collected with 695/40 band-pass �lter for 7AAD). Relative cell-surface P2X7 was determined using FlowJo soware and is expressed as the difference in the MFI of speci�c mAb labelling and isotype control labelling.

P2X7 Protein Detection by Confocal Microscopy. EOC13
or J774 cells in their respective complete culture medium were plated into 24-well plates with 13 mm glass coverslips (5 × 10 4 cells/0.5 mL/well) and incubated at 37 ∘ C, 95% air/5% CO 2 overnight to allow time to adhere. e following day, cells were �xed with 4% paraformaldehyde in PBS at room temperature for 15 min and then washed three times with PBS over 10 min. Cells were incubated with permeabilisation solution (PBS containing 0.1% DMSO, 2% NHS, and 0.1% Triton X-100) at room temperature for 10 min and washed three times with PBS. Cells were then blocked with 20% NHS in PBS at room temperature for 20 min. Cells were incubated at 4 ∘ C overnight with anti-rat P2X7 Ab (5 g/mL; preincubated for 1 h in the absence or presence of blocking peptide as per the manufacturer's instruction) in PBS containing 1% BSA, 0.2% NHS, and 0.05% NaN 3 . Cells were then washed as above and incubated at room temperature for 1 h with Cy3-conjugated anti-rabbit IgG Ab (15 g/mL) in PBS containing 0.2% NHS. Cells were washed as above and then the coverslips mounted onto slides with 50% (v/v) glycerol gelatin in PBS. Coverslips were sealed with nail varnish. Cells were visualised using a DM IBRE inverted microscope and TCS SP confocal imaging system (Leica, Mannheim, Germany) (excitation 488 nm, emission collected at 560-600 nm). Images were captured using Leica Confocal Soware.

ROS Formation
Assay. EOC13 cells in complete DMEM medium were plated into 24-well plates (5 × 10 4 cells/ 0.5 mL/well) and incubated at 37 ∘ C, 95% air/5% CO 2 overnight. Cells were then incubated with NaCl medium containing 10 M H 2 DCFDA (0.5 mL/well) at 37 ∘ C, 95% air/5% CO 2 , protected from light for 30 min. e medium was removed, and cells were further incubated in NaCl medium (containing 1 mM CaCl 2 ) in the absence or presence of 2 mM ATP at 37 ∘ C, 95% air/5% CO 2 for 15 min. Incubations were stopped by the addition of an equal volume of ice-cold MgCl 2 medium. Cells were harvested using 0.05% trypsin (5 min, 37 ∘ C) and were washed once with NaCl medium. Events were collected using a LSR II �ow cytometer (excitation 488 nm, emission collected at 515/20 nm) and the MFI of relative dichloro�uorescein (DCF) determined using FlowJo soware.
In some experiments, ATP-induced ROS formation was assessed in KCl medium, in NaCl medium in the absence of 1 mM CaCl 2 or presence of 100 M EGTA, or in complete DMEM medium in the absence or presence of 10 M AZ10606120 (15 min preincubation, prior to ATP addition). As free Ca 2+ lowers the concentration of ATP 4− [15], cells incubated in the absence of 1 mM Ca 2+ were incubated with 1.4 mM ATP to provide equimolar ATP 4− concentrations (575 M), as calculated using the Bound and Determined Program [16]. In other experiments, cells were preincubated in the absence or presence of AZ10606120, NAC, or DPI (as indicated) for 15, 30, and 30 min, respectively, prior to ATP addition. Cells prior to harvesting were also visualised by differential interference contrast (DIC) imaging using an Eclipse TE2000 inverted microscope (Nikon, Tokyo, Japan) to examine cell morphology, and DIC images were captured using Image-Pro AMS (Version 6.1) (Media Cybernetics, Rockville, MD).

NO Formation
Assay. EOC13 cells in complete DMEM medium were plated into 24-well plates (5 × 10 4 cells/0.5 mL/ well) and incubated at 37 ∘ C, 95% air/5% CO 2 overnight. Cells were then incubated with NaCl medium containing 10 M DAF-FM DA (0.5 mL/well) at 37 ∘ C, 95% air/5% CO 2 , protected from light for 30 min. e medium was removed, and the cells were washed once. Cells were then preincubated with NaCl medium in the absence or presence of 10 M AZ10606120 at 37 ∘ C, 95% air/5% CO 2 for 15 min. Following this, cells were further incubated in the absence or presence of 1.4 mM ATP for 15 min. Incubations were stopped by the addition of an equal volume of ice-cold MgCl 2 medium. Cells were harvested using 0.05% trypsin (5 min, 37 ∘ C) and were washed once with NaCl medium. Events were collected using a LSR II �ow cytometer (excitation 488 nm, emission collected at 515/20 nm) and the MFI of relative benzotriazole derivative determined using FlowJo soware.
2.10. Cell Death Assay. EOC13 cells in complete DMEM medium were plated into 24-well plates (5 × 10 4 cells/0.5 mL/ well) and incubated at 37 ∘ C, 95% air/5% CO 2 overnight. Cells were then incubated with �lter sterile ATP (as indicated) at 37 ∘ C, 5% CO 2 for 24 h. In some experiments, cells were preincubated in the absence or presence of 10 M AZ10606120 or 40 mM NAC for 15 or 30 min, respectively, prior to ATP addition. In other experiments, cells were preincubated in the absence or presence of 40 mM NAC for 90 min, with 2 mM ATP added in the �nal 15�60 min, and then the medium replaced and cells incubated at 37 ∘ C, 95% air/5% CO 2 for a further 24 h. Following the 24 h incubations, cells were harvested from wells using 0.05% trypsin (5 min, 37 ∘ C) and washed once with Annexin-V binding medium (NaCl medium containing 5 mM CaCl 2 ). Cells were then incubated with Annexin-V binding medium containing Annexin-V-Fluorescein and 7AAD at room temperature protected from light for 15 min. Events were collected using a LSR II �ow cytometer (excitation 488 nm, emission collected with 515/20 and 695/40 band-pass �lters for Annexin-V-Fluorescein and 7AAD, resp.) and the MFI of Annexin-V-Fluorescein and 7AAD determined using FlowJo soware. Quadrant markers were used to determine the percentage of Annexin-V + /7AAD − , Annexin-V − /7AAD + , and Annexin-V + /7AAD + cells. In some experiments, cells prior to harvesting were visualised by DIC imaging to examine cell morphology, and DIC images captured as outlined in Section 2.8.

Data Presentation and Statistical
Analyses. Data is presented as the mean ± SD. Differences between multiple treatments were compared by ANOVA paired with Tukey's HSD posttest using Prism 5 for Windows (Version 5.04) (GraphPad Soware, San Diego, CA), with differences considered signi�cant for . Concentration response and inhibition curves were �tted using Prism 5 and assuming a variable slope, with normalised and nonnormalised response curves, respectively, selected to obtain the best �t. Estimates of EC values and half maximal inhibitory concentrations (IC ) were obtained from individual �ts of these plots.

P2X7 Antagonists Inhibit ATP-Induced Ethidium + Uptake into J774 Macrophage Cells in a Concentration-Dependent
Manner. e murine macrophage J774 cell line is well known to express functional P2X7 [17]. Moreover, our group has demonstrated the presence of functional P2X7 in various cell types using a �xed-time �uorescent cation uptake assay (e.g., [14,18]). erefore, this technique was used to con�rm the presence of P2X7 in J774 cells and to validate the use of this cell line as a positive control. Incubation of J774 cells with the P2X7 agonist ATP and the most potent P2X7 agonist BzATP induced signi�cant ethidium + uptake into these cells compared to cells incubated in the absence of nucleotide (Figure 1(a)). In addition, incubation of J774 cells with ATP induced signi�cant YO-PRO-1 2+ uptake compared to cells incubated in the absence of ATP (Figure 1(b)). However, ATP-induced YO-PRO-1 2+ uptake was signi�cantly lower than ATP-induced ethidium + uptake (Figure 1(b)).

EOC13 Microglial Cells Express P2X7.
To determine whether EOC13 microglial cells express P2X7, a series of experiments using J774 cells as a positive control were performed. Firstly, RNA was isolated from EOC13 and J774 cells and ampli�ed by RT-PCR using primers for P2X7. Similar to J774 cells, EOC13 cells were found to express P2X7 mRNA, as evident from the 230 base pair band corresponding to the size of the predicted product ( Figure  2(a)). e presence of total P2X7 protein was determined by probing separated whole lysates of both cell lines with an anti-P2X7 Ab. Immunoblotting revealed one major protein band of 75 kDa for both EOC13 and J774 cells (Figure 2(b)), corresponding to the predicted size of glycosylated P2X7. Moreover, both cell lines were incubated with an anti-P2X7 mAb and the presence of cell-surface P2X7 determined by �ow cytometry. Immunolabelling demonstrated cell-surface P2X7 on both EOC13 and J774 cells, with MFIs of 13 ± 2 and 14 ± 4, respectively ( 3) (Figure 2(c)). Finally, both cell lines were stained with an anti-P2X7 Ab and analysed by confocal microscopy. is similarly demonstrated the presence of cell-surface P2X7, as well as intracellular P2X7, with bright staining observed on all cells (Figure 2(d)). Preincubation of the anti-P2X7 Ab with blocking peptide completely abrogated the detection of P2X7 in both cell lines (data not shown). Together, these results indicate that P2X7 is expressed in EOC13 cells.

EOC13 Microglial Cells Express Functional P2X7.
To determine whether P2X7 was functional in EOC13 cells, the �xed-time ethidium + uptake assay was performed. Both ATP and BzATP were found to induce signi�cant ethidium + uptake into EOC13 cells compared to cells incubated in the absence of nucleotide (Figure 3(a)). Next, EOC13 cells were incubated with increasing concentrations of ATP. ATP induced ethidium + uptake in a concentration-dependent manner, with maximal uptake obtained at an ATP concentration of 1 mM and with an EC 5 of ± M (Figure 3(b)). Subsequent characterisations of P2X7 in EOC13 microglia were performed using this maximal concentration of ATP (1 mM).

Mediators of In�ammation
To determine if the observed ATP-induced ethidium + uptake into EOC13 cells was mediated by P2X7, cells were preincubated in the absence or presence of P2X7 antagonists at inhibitory concentrations optimal for 1 mM ATPinduced ethidium + uptake in J774 cells, as demonstrated above (Figure 1(c)). Preincubation of EOC13 cells with 30 M BBG, 100 M A438079, 10 M AZ10606120, and 30 M AZ11645373 resulted in signi�cant impairment of ATP-induced ethidium + uptake by 75 ± 2, 9 ± , ± , and 99 ± %, respectively (Figure 3(c)). None of the P2X7 antagonists except AZ11645373 signi�cantly altered the basal ethidium + uptake into EOC13 cells. Again with the exception of AZ11645373, which reduced the amount of gated viable cells by ∼30%, cell viability (as assessed by forward and side scatter) was similar between treatments (data not shown).

�ediators of In�ammation
Relative P2X7 expression Relative cell number To determine if P2X7 activation could induce the uptake of a second organic cation into EOC13 cells, cells were preincubated in the absence or presence of AZ10606120, and ATPinduced YO-PRO-1 2+ uptake examined. Similar to ethidium + uptake, 1 mM ATP induced signi�cant YO-PRO-1 2+ uptake into EOC13 cells compared to cells incubated in the absence of ATP (Figure 3(d)). Moreover, preincubation of cells with 10 M AZ10606120 resulted in complete inhibition of ATPinduced YO-PRO-1 2+ uptake (Figure 3(d)). Incubation with AZ10606120 did not signi�cantly alter the basal YO-PRO-1 2+ uptake (Figure 3(d)). Furthermore, cell viability (as assessed by forward and side scatter) was similar between treatments (data not shown). Collectively, these results indicate that P2X7 is functional in EOC13 cells.

P2X7 Activation Induces ROS Formation in EOC13
Microglial Cells. P2X7 activation has been reported to induce ROS formation in a number of cell types, including primary microglia [24,25]. us, ATP-induced ROS formation in the EOC13 cell line was investigated using the ROS sensitive probe DCF. Cells loaded with H 2 DCFDA (which is converted to DCF inside cells) were incubated in the absence or presence of ATP, and the subsequent ROS formation analysed by �ow cytometry. As extracellular Ca 2+ has been reported to be important for P2X7-induced ROS formation in a number of cell types [24,26,27], assays were initially conducted in the presence of 1 mM Ca 2+ . However, due to the inhibitory action of Ca 2+ on P2X7 [15], assays were initially conducted with 2 mM ATP. Incubation with 2 mM ATP induced signi�cant ROS formation in EOC13 cells compared to cells incubated in the absence of ATP (MFI of ROS formation 6 ± 6 and 5 8 ± 6, resp., , ). Furthermore, preincubation of cells with 10 M AZ10606120 resulted in complete inhibition of ATP-induced ROS formation ( Figure  4(a)), indicating that this process is mediated by P2X7 activation. As for cation uptake (Figure 3), AZ10606120 did not signi�cantly alter the basal ROS formation (Figure 4(a)) or cell viability (as assessed by forward and side scatter) (data not shown).
P2X7 is a ligand-gated cation channel [7]; therefore the possible roles for cation �uxes in P2X7-induced ROS formation were next investigated. P2X7-induced ROS formation has been reported to be partially dependent on Ca 2+ in�ux in human promyelocytes [26] and rat submandibular gland cells [27]. us, to determine if Ca 2+ in�ux is required for P2X7-mediated ROS formation in EOC13 cells, ATPinduced ROS formation in the absence and presence of Ca (a-d) Incubations were stopped by the addition of MgCl 2 medium and centrifugation. Mean �uorescence intensities (MFI) of (le panels) DCF (ROS formation) or (right panels) ethidium + uptake (pore formation) were determined by �ow cytometry and results shown as means ± SD, ; * * * or * * compared to corresponding basal; † † † compared to corresponding ATP.
to similarly treated cells in the absence of ATP ( Figure  4(b)). Cells incubated in the absence of Ca 2+ had significantly higher ATP-induced ROS formation compared to those incubated in the presence of Ca 2+ . In contrast, ATPinduced ethidium + uptake (P2X7 function) was similar in cells incubated in the absence or presence of Ca 2+ ( Figure  4(b)), indicating that the differences in ATP-induced ROS formation were not due to altered P2X7 function. NaCl medium may contain nominal amounts of Ca 2+ . us, to further exclude a role for Ca 2+ in P2X7-mediated ROS formation in EOC13 cells, ATP-induced ROS formation was investigated in the absence and presence of the Ca 2+ chelator EGTA. Incubation with 1.4 mM ATP induced sig-ni�cant but similar amounts of ROS formation in both the absence and presence of 100 M EGTA compared to similarly treated cells in the absence of ATP (Figure 4(c)). Again, ATPinduced ethidium + uptake was similar in cells incubated in the absence or presence of EGTA (Figure 4(c)).
Finally, the role of K + in P2X7-induced ROS formation in EOC13 cells was investigated. Both ROS and K + efflux have been reported to be involved in interleukin-1 (IL-1 ) release from monocytes, although whether these downstream processes are linked is yet to be established [28]. us, to determine if K + efflux is involved in P2X7-mediated ROS formation in EOC13 cells, ATP-induced ROS formation was compared with cells in NaCl medium and KCl medium, which prevents the loss of intracellular K + . Incubation with 1.4 mM ATP induced signi�cant ROS formation in both NaCl and KCl media, with similar levels of ROS formation in both media (Figure 4(d)). Likewise, ATP-induced ethidium + uptake was similar in NaCl and KCl media (Figure 4(d)), indicating that the inability of high extracellular K + to impair ATP-induced ROS formation was not due to altered P2X7 function.
To con�rm that P2X7 activation induced ROS formation in EOC13 microglia, DCF-loaded cells in NaCl medium were preincubated in the absence or presence of the ROS scavenger NAC, before incubation in the absence or presence of ATP. As above (Figure 4), 1.4 mM ATP induced signi�cant ROS formation ( Figure 5(a)). Preincubation with 40 mM NAC inhibited ATP-induced ROS formation by 7 7± % ( Figure  5(a)). Basal ROS formation ( Figure 5(a)) and cell viability (as assessed by forward and side scatter) (data not shown) were similar between treatments. Preincubation of cells with 40 mM NAC inhibited ATP-induced ethidium + uptake by ± 2% (Figure 5(b)). us, the inhibitory effect of NAC on P2X7-induced ROS formation may be partially attributed to inhibition of P2X7 itself. However, incubation with NAC and ATP, but not either compound alone, reduced the amount of gated viable cells by ∼40% in the ethidium + uptake assay (as assessed by forward and side scatter) (data not shown). is suggests that the inhibitory action of NAC on ATP-induced ethidium + uptake may be a result of cytotoxicity in this assay.
To con�rm that NAC did not induce morphological changes or cause cytotoxicity under the conditions used for  the ROS assay, DIC images of adherent cells were acquired following incubation in the absence or presence of ATP. Cells incubated in the absence or presence of NAC (without ATP) displayed discrete cell bodies with long, spindled shaped processes ( Figure 5(c)), as previously observed [13]. Cells incubated in the absence or presence of NAC (with ATP) also displayed discrete cell bodies, but with retracted and branched processes ( Figure 5(c)), typical of ATP causing membrane changes [29]. erefore, in the ROS assay, EOC13 cell morphology was not altered by NAC when compared to the corresponding treatment. DCF-loaded cells were also preincubated in the absence or presence of the broad-spectrum ROS inhibitor DPI and the ATP-induced ROS formation investigated. However, a 30 min preincubation with DPI at various concentrations (5-40 M) led to high amounts of cell death (data not shown), and thus this compound was not examined further.
To further verify that P2X7 activation induces the formation of reactive species in EOC13 cells, ATP-induced NO formation was investigated using the NO sensitive probe DAF-FM DA. Cells loaded with DAF-FM DA (which reacts with NO to form a �uorescent ben�otria�ole) were preincubated in the absence or presence of AZ10606120, followed by incubation in the absence or presence of ATP, and the subsequent NO formation analysed by �ow cytometry. Incubation with 1.4 mM ATP induced signi�cant NO formation in EOC13 cells compared to cells incubated in the absence of ATP ( Figure 6). Furthermore, preincubation of cells with 10 M AZ10606120 inhibited ATP-induced NO formation by 82 ± % (Figure 6), indicating that this process is mediated by P2X7 activation. Again, AZ10606120 did not signi�cantly alter the basal NO formation (Figure 6) or cell viability (as assessed by forward and side scatter) (data not shown).

P2X7 Activation Induces Cell Death in EOC13 Microglial
Cells. P2X7 activation results in the death of various cell types [11,12]. To determine whether ATP induces the death of EOC13 microglia, cells in complete DMEM medium were incubated in the absence or presence of ATP for 24 h, and then the percentage of Annexin-V + /7AAD − , Annexin-V − /7AAD + , and Annexin-V + /7AAD + cells examined by �ow cytometry (Figure 7(a)). Cell death is expressed as the total of dying (Annexin-V + /7AAD − ) and dead (Annexin-V − /7AAD + and Annexin-V + /7AAD + ) cells. Incubation with either 2 or 3 mM ATP but not 1 mM ATP resulted in signi�cantly higher percentages of total cell death compared to cells incubated in the absence of ATP (Figure 7(a)). Next, to determine if the ATP-induced EOC13 death was mediated by P2X7 activation, cells were preincubated in the absence or presence of AZ10606120 and then incubated in the absence or presence of ATP for 24 h. As above (Figure 7(a)), 2 mM ATP induced signi�cant cell death, with higher percentages of total cell death compared to cells incubated in the absence of ATP (Figure 7(b)). Preincubation with 10 M AZ10606120 completely inhibited ATP-induced cell death (Figure 7(b)), indicating that this process is mediated by P2X7 activation.
P2X7-induced death of murine RAW264.7 macrophages is mediated by ROS formation [30]. erefore, a role for ROS in P2X7-induced death of EOC13 microglia was investigated. To con�rm that P2X7 induced ROS formation under conditions used to induce cell death, DCF-loaded EOC13 cells in complete DMEM medium were incubated in the absence or presence of ATP, and then subsequent ROS formation determined by �ow cytometry. Similar to ATP-induced cell death (Figure 7(a)), incubation with 2 or 3 but not 1 mM ATP induced signi�cant ROS formation in EOC13 cells compared to cells incubated in the absence of ATP ( Figure  7(c)). e requirement for higher ATP concentrations to induce cell death (Figure 7(a)) or ROS formation ( Figure  7(c)) compared to pore formation (Figure 3(b)) is in line with the inhibitory action of divalent cations [15], which are present in the culture medium but not in the NaCl medium used. To con�rm that this ROS formation was mediated by P2X7, cells were preincubated with AZ10606120 and the amounts of ATP-induced ROS formation determined. To parallel the cell death experiments, 2 mM ATP was utilised. Again, ATP induced signi�cant ROS formation compared to cells incubated in the absence of ATP (Figure 7(d)). Preincubation with 10 M AZ10606120 completely inhibited this ATP-induced ROS formation (Figure 7(d)).
Finally, the effect of NAC on P2X7-induced cell death was investigated. EOC13 cells were preincubated in the absence or presence of NAC followed by ATP for 24 h. However, 24 h incubation with 40 mM NAC in the absence of ATP led to �ediato�� o� �n�a��ation  Figure  7(e)). In contrast, a 75 and 30 min preincubation with NAC, followed by 15 and 60 min ATP treatment, respectively, had no effect on the percentage of cell death compared to cells incubated for the same time length with ATP in the absence of NAC (Figure 7(e)).
To further con�rm that NAC did not induce morphological changes or cause cytotoxicity under the conditions used for the cell death assay, DIC images of adherent cells were acquired following the 24 h incubation. As above ( Figure  5(c)), cells incubated in the absence or presence of NAC (without ATP) displayed discrete cell bodies with long, spindled shaped processes (Figure 7(f)). In addition, cells incubated in the presence of NAC and ATP displayed a similar morphology to that of cells incubated in the absence of ATP (Figure 7(f)). In contrast, cells preincubated with ATP alone displayed rounded and granular cell bodies with no or few processes (Figure 7(f)), characteristic of cell death.
Furthermore, wells preincubated with ATP alone had a high amount of nonadherent cells compared to the other treatments (data not shown). us, preincubation with NAC prevented the morphological changes associated with ATP incubation, but NAC alone had no effect on cell morphology.

Discussion
e current study demonstrates for the �rst time that the murine microglial EOC13 cell line expresses functional P2X7. Firstly, the presence of P2X7 mRNA and protein was established using RT-PCR and immunoblotting techniques. In addition, the presence of cell-surface P2X7 was demonstrated using immuno�uorescence staining. Moreover, P2X7 on EOC13 microglia was functional, as the P2X7 agonists ATP and B�ATP induced signi�cant ethidium + or YO-PRO-1 2+ uptake into these cells. In these experiments, ATP induced ethidium + uptake with an EC value which falls within the typical range for ATP-induced cation �uxes mediated by recombinant murine P2X7 [31]. Furthermore, pretreatment of cells with P2X7 antagonists inhibited ATPinduced organic cation uptake. Lastly, ATP could induce ROS formation in and the death of EOC13 cells, and both of these events, which are oen associated with P2X7 activation [10][11][12], could be inhibited by the P2X7 antagonist AZ10606120. e presence of functional P2X7 on EOC13 microglia is consistent with the presence of this receptor on primary microglia and microglial cell lines (including N9, N13, BV-2, and NTW8 cells) [32][33][34][35].
P2X7 activation induced ROS formation in EOC13 microglia. P2X7-induced ROS formation has been reported in primary microglia [24,25], and a role for this process in microglia has been highlighted by several studies. P2X7 activation induces the production of the ROS, superoxide, in primary rat microglia, while this receptor is upregulated in a transgenic mouse model of Alzheimer's disease [25]. Although a direct link between P2X7, superoxide, and Alzheimer's disease was not established, the authors proposed a link between these molecules and this disease. is link is supported by subsequent observations by others, where �brillar -amyloid peptide, which is associated with Alzheimer's disease, caused ATP release and autocrine activation of P2X7 leading to ROS formation in primary rat microglia [24]. In addition, another group demonstrated that P2X7-induced superoxide release from primary rat microglia induced injury of rat cortical neurons [36]. Collectively, this data indicates that P2X7-induced ROS formation from microglia may be involved in various neuroin�ammatory and neurodegenerative disorders. is may be of particular importance in diseases where microglial P2X7 is reported to be upregulated such as in Alzheimer's disease, multiple sclerosis, and amyotrophic lateral sclerosis [25,37]. It should be noted, however, that DCF, as employed in the current study and as widely used by others to detect ROS, can also propagate ROS formation [38]. Nevertheless, our observation that P2X7 activation also induces the formation of NO in EOC13 cells supports a role for this receptor in the formation of reactive species.
e current study excluded an essential role for Ca 2+ in�ux in P2X7-induced ROS formation in EOC13 microglia. is �nding is similar to other observations with other murine cell types, including submandibular glands [39] and erythroid cells [40]. In contrast, P2X7-induced ROS formation in primary rat microglia [24] and rat submandibular glands [27] is dependent on an in�ux of Ca 2+ . e reason for this difference between these two species remains unknown but may re�ect differences in experimental protocols or differences in signalling molecules between mice and rats. e current study also excluded an essential role for K + efflux in P2X7-induced ROS formation in EOC13 microglia. Both ROS formation and K + efflux are involved in IL-1 release from monocytes, although whether these downstream processes are linked has not been established [28]. us, our results indicate that P2X7-induced ROS formation does not require K + efflux and that ROS formation and K + efflux may be independent events in IL-1 release from myeloid cells.
P2X7 activation also induced cell death in EOC13 microglia. Use of an Annexin-V/7AAD assay suggested that this process was mediated by apoptosis. However, in the absence of other markers of apoptosis and necrosis, this remains to be established, especially since P2X7 activation induces both apoptosis and necrosis in the microglial N13 cell line [41]. Nevertheless, our observations support previous studies in which P2X7 activation induced death in primary microglia and other microglial cell lines [41,42]. e physiological role of P2X7-induced microglia death is unclear. Further obscuring this is the known role of P2X7 activation in inducing the proliferation of microglia [43]. is paradoxical role of P2X7 is thought to be related to the relative ATP concentration, with high concentrations promoting cell death and low concentrations promoting cell proliferation [44]. In support of this, our study observed that ATP only induced EOC13 cell death at 2 or 3 but not 1 mM ATP. Moreover, our data also showed that a transient incubation with ATP of 30-60 but not 15 min induced cell death in EOC13 microglia. is suggests that transient ATP release and subsequent P2X7 activation may be sufficient to kill microglia in vivo.
e current study examined a potential link between P2X7-induced ROS formation and death in EOC13 microglia. A previous study demonstrated that the ATPinduced death of murine RAW264.7 macrophages was mediated by ROS derived from NADPH oxidase downstream of P2X7 activation [30]. is contrasts with another study, which found that P2X7-induced ROS formation, but not death, was attenuated in primary macrophages from NADPH oxidase de�cient mice [45]. Our data using the ROS scavenger NAC supports a role for ROS formation in the P2X7-induced death of EOC13 microglia. e capacity of NAC to prevent P2X7-induced EOC13 microglia death was dependent on the preincubation time with NAC, as well as the total incubation time with ATP, with only 45-60 min preincubations with NAC preventing cell death induced by transient 30-45 min exposures to ATP. In contrast, 24 h incubation with NAC induced signi�cant amounts of EOC13 microglia death, equivalent to that induced by ATP alone. is cytotoxicity of NAC may have occurred due to increased toxic metabolic byproducts such as reduced glutathione [46]. Alternatively, scavenging of ROS by NAC may indicate that low amounts of ROS are important for EOC13 cell homeostasis. Of note, the ROS inhibitor DPI also induced the death of EOC13 microglia, albeit over a much faster time course. Finally, it should be noted that NAC inhibition of P2X7-induced death and ROS formation in EOC13 microglia may have been partly due to direct inhibition of P2X7. NAC inhibited ATP-induced pore formation by 30% compared to a 74 and 99% inhibition of ATP-induced ROS formation and cell death, respectively. is direct inhibition of P2X7 by NAC was not due to an acidic pH, which is known to impair P2X7 function [47], as the NAC-containing solutions were adjusted to pH 7.4 before each assay. us, our results indicate that either cellular signalling involving ROS may modulate P2X7 activation in EOC13 microglia or that NAC may directly impair P2X7 at 40 mM. e concentration of NAC used in these experiments (40 mM) is 4-8-fold higher than that used in a number of similar studies (e.g., [48]). e requirement for this high concentration of NAC remains unknown but may re�ect a reduced ability of NAC to cross the plasma membrane or to be converted to glutathione in EOC13 cells.
e presence of functional P2X7 on J774 macrophage cells was con�rmed in the current study. P2X7 is present in this cell line [17], and activation of P2X7 leads to the release of mature IL-1 [49], the formation of macrophage-derived multinucleated giant cells [50][51][52], and cell death [53]. In this study, the potency of four P2X7 antagonists against 1 mM ATP, the ATP concentration most commonly used to study P2X7, was determined. e IC 50 values for BBG, A438079, and AZ11645373 (1.8, 7.9, and 1.5 M, resp.) were within one log range of those published for recombinant murine P2X7 [54,55]. In contrast, the IC 50 value for AZ10606120 has not been reported for murine P2X7, although this compound has been shown to speci�cally bind to and inhibit rat and human P2X7 [20]. In the current study, this highly speci�c P2X7 antagonist also completely impaired ATP-induced ethidium + uptake, ROS formation, and death of murine EOC13 cells. us, AZ10606120 will be useful for future studies of murine P2X7.
In the CNS, extracellular ATP acting through P2X7 on microglia is an important mediator of neuroin�ammation [9]. ATP acts as a neurotransmitter and is released from neurons during synaptic transmission and from dying cells [56]. Under normal physiological conditions, extracellular ATP concentrations in the CNS are estimated to be in the nanomolar to micromolar range, depending on the balance between ATP release and degradation, while intracellular microglial ATP concentrations are in the millimolar range [57]. Aer CNS injury, however, extracellular ATP concentrations increase and can reach as high as the millimolar range [57]. Furthermore, it is hypothesised that ATP may act on microglial P2X7 at very close range where the concentration of ATP may be quite high. Activation of P2X7 on primary microglia and microglial cell lines leads to the release of proin�ammatory IL-1 and tumour necrosis factor- [33,58] and ROS formation [24]. Although proin�ammatory factors are important for immunity [1], prolonged or inappropriate release of such factors from chronically activated microglial can be highly toxic to neurons and can promote neuroin�ammation and neurodegeneration [5]. ere are a number of diseases in the CNS characterised by the presence of activated microglia, including Alzheimer's disease, prion infection, cerebral ischemia, multiple sclerosis, and amyotrophic lateral sclerosis. In such diseases, P2X7 has also been reported to be upregulated [25,37,59,60]. is raises questions of possible roles for P2X7 in mediating inappropriate microglial responses in CNS disorders.

Conclusions
is study demonstrates that EOC13 microglial cells express functional P2X7. Activation of this receptor by ATP resulted in organic cation uptake, ROS formation, and death in these cells. Moreover, the EOC13 cell line may be useful for investigating P2X7-mediated events in microglia and the role of this receptor in microglia-mediated in�ammatory disorders.