P2X7 mediates superoxide production in primary microglia and is up-regulated in a transgenic mouse model of Alzheimer's disease.

Primary rat microglia stimulated with either ATP or 2'- and 3'-O-(4-benzoylbenzoyl)-ATP (BzATP) release copious amounts of superoxide (O(2)(-)*). ATP and BzATP stimulate O(2)(-)* production through purinergic receptors, primarily the P2X(7) receptor. O(2)(-)* is produced through the activation of the NADPH oxidase. Although both p42/44 MAPK and p38 MAPK were activated rapidly in cells stimulated with BzATP, only pharmacological inhibition of p38 MAPK attenuated O(2)(-)* production. Furthermore, an inhibitor of phosphatidylinositol 3-kinase attenuated O(2)(-)* production to a greater extent than an inhibitor of p38 MAPK. Both ATP and BzATP stimulated microglia-induced cortical cell death indicating this pathway may contribute to neurodegeneration. Consistent with this hypothesis, P2X(7) receptor was specifically up-regulated around beta-amyloid plaques in a mouse model of Alzheimer's disease (Tg2576).

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
were collected by centrifugation (200 ϫ g for 10 min) and used the same day. The purity of the cultures was 98 -100% as determined by immunostaining with ED-40 antibody. Isolation of Cortical Neurons-Primary cortical cell cultures were prepared from embryos of timed pregnant Sprague-Dawley rats at E14 (18). Briefly, the cortex triturated in DNase/Protease dissociation buffer was centrifuged and resuspended in PC-1 SF medium (BioWhittaker). The cells (2 ϫ 10 5 /ml) were plated onto poly-L-ornithine-coated 24-well plates, and 4 days later the media were replaced with Neurobasal Medium containing B-27 supplement (Invitrogen), 1% penicillin/streptomycin, and 10 mM L-glutamine. Neuronal cells constituted 90 -95% of the total cells and were used on day 10 for experiments.
Isolation of Neutrophils-Neutrophils were isolated from peripheral blood of healthy human donors as reported previously (19).
Measurement of Superoxide Production-Superoxide (O 2 . ) was measured indirectly through the detection of hydrogen peroxide (H 2 O 2 ) by the method of Mohanty et al. (20). O 2 . production was measured in initial experiments by O 2 . -dependent superoxide dismutase-sensitive reduction of ferricytochrome c (21). However, microglia released very little O 2 . , and this procedure required a large number of cells. In subsequent experiments the more sensitive method of H 2 O 2 detection using conversion of 10-acetyl-3,7-dihydroxyphenoxazine (Amplex Red) to highly fluorescent resorufin in the presence of horseradish peroxidase was followed (20). Briefly, 5 ϫ 10 5  Measurement of Calcium-Isolated rat microglia were plated onto 384-well plates (Falcon, part 353961) at ϳ50% confluence. Cells were loaded with Fluo-4,AM (5 M) in HBSS containing 10 mM HEPES (pH 7.4) for 1 h before the experiment at room temperature and washed with the buffer. ATP and BzATP were used to stimulate the [Ca 2ϩ ] i signal. In the experiments where oATP and PPADS were used, cells were pretreated with the inhibitors for 2 h at 37°C and then loaded with Fluo-4,AM. The fluorescent signal from ϳ10 4 cells per well was measured using a fluorometric plate reader (FLIPR, Molecular Devices). Fluo-4 was excited at 488 nm, and fluorescence was measured at 510 nm in a time-resolved mode (1-Hz frequency). Relative f/f 0 intensity (in counts/ms) was used as an indication of [Ca 2ϩ ] i signal. Data acquisition and preliminary analysis were done using FLIPR software (Molecular Devices). All calcium measurements were done at room temperature.
Membrane Translocation of p67 phox -Fractionation of microglia was performed according to the method of Zhao et al. (23). Briefly, microglia (14 ϫ 10 6 cells/ml) in HBSS were treated with or without 500 M BzATP for 5-10 min. Cells were centrifuged, and the cell pellet was resuspended in 0.5 ml of relaxation buffer (100 mM KCl, 3 mM NaCl, 3.5 mM MgCl 2 , 1.25 mM EGTA, 10 mM PIPES (pH 7.3), 500 M phenylmethylsulfonyl fluoride, and 1:100 dilution of protease inhibitor mixture), sonicated (3 ϫ 10 s, 4°C using a microprobe sonicator), and centrifuged (500 ϫ g for 10 min) to remove nuclei and unbroken cells. The postnuclear lysates were then ultracentrifuged (100,000 ϫ g for 60 min, 4°C), and the resulting supernatant was designated the cytosolic fraction. The membrane/particulate pellet was resuspended in 200 l of relaxation buffer containing 1% Triton X-100. Protein concentration was estimated using the Bio-Rad D C Protein Assay, and 25 g of protein (for both the cytosolic and membrane/particulate fraction) was loaded onto a gel.
A spontaneously occurring rat microglial cell line was used for the above experiment because we were unable to generate sufficient numbers of primary microglia required for this experiment. The spontaneously occurring rat microglial cell line was isolated from primary rat microglia growing in LADMAC-conditioned media (ATCC, Manassas, VA). The cells were propagated in media (Dulbecco's modified Eagle's medium, 1% penicillin/streptomycin, 10 mM L-glutamine, 0.1 mM nonessential amino acids, 10% fetal bovine serum) containing 20% LAD-MAC-conditioned media. The cells were positive for ED-1, a microglial marker. The microglial cells responded to both LPS and BzATP as demonstrated by the generation of TNF␣ with LPS and ROIs with BzATP (data not shown).
Immunoblotting-Immunoblotting was performed as described previously (18). Briefly 25 g of protein was fractionated on a 10% SDS-PAGE gel, transferred to polyvinylidene difluoride membrane, and blocked in 5% milk/Tris-buffered saline containing 0.1% Tween 20 for 2 h. The membrane were washed and incubated overnight with antibodies specific for phospho-p42/44 MAPK (Thr-202/Tyr-204), phospho-p38 MAPK (Thr-180/Tyr-182) diluted 1:1000 in TBST containing 5% bovine serum albumin. Membranes were washed with TBST and incubated with an horseradish peroxidase-conjugated secondary antibody (1:2000) for 2 h. The membrane was washed extensively, and bands were detected using LumiGLO. The membranes were stripped using RESTORE Western blot stripping buffer (Pierce), washed several times, and blocked for 1 h. Membranes were incubated with antibodies specific for either unphosphorylated p42/44 MAPK or p38 MAPK diluted 1:1000 in blocking buffer. The next day membranes were incubated with the secondary antibody and visualized using LumiGLO.
The P2X 7 (1:500), p67 phox (1:500), and actin (1:750) antibodies were used according to the manufacturer's recommendation. In some experiments a P2X 7 control peptide corresponding to amino acid 576 -595 of rat P2X 7 (the immunogen used to generate the antibody) was utilized to determine specificity of the bands. The P2X 7 antibody was preincubated with the control peptide at a 1:1 dilution (v/v) for 1 h at room temperature prior to the addition to the membrane.
Neurotoxicity Assay-Primary rat microglia (1 ϫ 10 5 ) in Neurobasal Medium containing B-27, 1% penicillin/streptomycin, and 10 mM Lglutamine were seeded into a 48-well plate containing 1 ϫ 10 5 primary cortical neurons. The cells were allowed to settle for 2 h prior to the addition of stimuli. After a 72-h incubation, the supernatant was assayed for lactate dehydrogenase (LDH). Microglia and cortical cells were also independently cultured for 72 h in the presence of stimuli, and LDH released from microglia alone Ϯ stimuli were subtracted out from the values obtained from the combination of microglia and cortical neurons. The LDH was measured with a commercial kit obtained from Promega (WI). In one experiment a WST-1 cell survival assay was performed on the cells remaining in the well with a commercial kit obtained from Roche Diagnostics. WST-1 is a modified 3-(4,5dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium assay. The WST-1 assay enables colorimetric measurement of cell viability based on the cleavage of tetrazolium salts by mitochondrial dehydrogenase in viable cells.
Nitrite Assay-Nitrite assay was performed in a 96-well plate using modified Griess Reagent. In brief, 100 l of Griess Reagent was added to 100 l of supernatant in a 96-well plate. Samples were read at 540 nm, and values were calculated against a sodium nitrite standard curve.
Tissue Processing and Immunohistochemistry-The mice were sedated, perfused with 4% paraformaldehyde and decapitated. The brains (2-year Tg2576 mice and aged-matched controls) were removed and fixed in 4% paraformaldehyde. The brains were dehydrated in graded alcohol solutions followed by Histoclear and embedded in paraffin. Longitudinal serial sections were cut at 6-m thickness.
The sections were deparaffinized in Histoclear, rehydrated through a series of graded alcohols, washed in deionized water, and incubated in 1:5 methanol/water solution containing 3% H 2 O 2 for 30 min to quench endogenous peroxidase activity. The slides were rinsed in deionized water for 5 min followed by blocking in 5% normal goat serum in phosphate-buffered saline containing 0.01% Triton-X-100 for 1 h. Sections were incubated with primary antibody (Pan A-␤ 1:1000 (QCB), 4G8 1:1000 (Signet), P2X 7 1:100 (Pharmingen), CD45 1:200 (Serotec), or GFAP 1:1000 (Chemicon)) in 1% normal goat serum in phosphatebuffered saline overnight at 4°C. Immunohistochemistry was completed with appropriate biotinylated secondary antibody (1:500) in 2% normal goat serum/phosphate-buffered saline followed by avidin-biotin complex and visualized by diaminobenzidine development (Vector Laboratories). Primary or secondary antibodies were omitted from some sections to serve as negative controls. After the enzyme substrate (Vector Laboratories) was added, the manufacturer's protocol was followed. The slides were washed in water and counter-stained with Shandon-Lippshaw hematoxylin stain for 2 min at room temperature. Sections were washed in water for 1 min, incubated for 10 s in 50% ethanol/ H 2 O ϩ 1% HCl to remove residual hematoxylin, and then washed in water for 1 min. The slides were then dehydrated in a series of alcohol washes and sealed with a coverslip using DPX mounting medium.
Immunofluorescence was used to detect dual antigen labeling. Tissue sections were deparaffinized via a series of xylenes and alcohols. Sections were blocked in 10% donkey serum for 1 h and then incubated in primary antibodies using 1% serum in Tris-buffered saline overnight at 4°C. Primary antibodies (1:1000) specific for GFAP and P2X 7 were pooled for dual immunofluorescence. Slides were then incubated in pooled secondary antibodies using donkey anti-mouse Cy-2 (1:100) and donkey anti-rabbit Cy-3 (1:400) in 2% normal donkey serum in Trisbuffered saline for 1 h in the dark. Slides were mounted and coverslip sealed with Vectashield (Vector Laboratories, Burlingame, CA). Images were captured using a Zeiss Axiovert S100TV microscope with the Zeiss KS400 imaging system.
Generation of Hippocampal Lysates-Hippocampi from Tg2576 mice and age-matched controls (19 months, 1 male and 2 female) were excised and snap-frozen in liquid nitrogen and stored at Ϫ80°C. Tissues were homogenized for 20 s on ice in TNE buffer (50 mM Tris, 150 mM NaCl) at 20% weight/volume using a Polytron homogenizer. Samples were then diluted 1:1 in TNE buffer containing 2% SDS, 1% Nonidet P-40, and 1% deoxycholate, and sonicated (2 ϫ 15 s, 4°C). Protein concentration was estimated using the Bio-Rad D C Protein Assay, and 25 g of protein from each hippocampus was loaded onto a gel.
Statistics-Student's t test was performed to determine group differences (p Ͻ 0.05).  level of intracellular calcium (Fig. 2). ATP stimulated a transient increase of intracellular calcium ( Fig. 2A). BzATP caused a sustained increase in the level of intracellular free calcium ([Ca 2ϩ ] i ) that was maintained for more than 6 min (Fig. 2B). To determine whether ATP and BzATP were mobilizing intracellular or extracellular sources of Ca 2ϩ or both, additional experiments were carried out in Ca 2ϩ -free media containing 1,2-bis-(o-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid (Ca 2ϩ chelator). In the absence of extracellular Ca 2ϩ , the BzATP response was completely blocked indicating that BzATP was mobilizing only extracellular Ca 2ϩ (Fig. 2C). However, with ATP, the initial transient peak was reduced by about 75%, suggesting that ATP mobilizes both intracellular (via inositol 1,4,5-trisphosphate-induced Ca 2ϩ release) and extracellular sources of Ca 2ϩ (possibly via capacitative Ca 2ϩ influx) (24,25).

Generation of Reactive Oxygen Intermediates by Microglia-
Effect of Extracellular Calcium on ROI Production-Because ATP appeared to stimulate Ca 2ϩ release from intracellular stores and Ca 2ϩ influx from extracellular sources, whereas BzATP appeared to stimulate only Ca 2ϩ influx from extracellular sources, the effect of removal of extracellular Ca 2ϩ on H 2 O 2 production was examined. Both ATP-and BzATP-stimulated H 2 O 2 production was blocked to below control levels in the absence of extracellular Ca 2ϩ (Fig. 3). These results suggest that despite the differences in Ca 2ϩ mobilization, both ATP and BzATP required only extracellular Ca 2ϩ to generate H 2 O 2 .
Receptors Involved in the Generation of ROI-The ability of BzATP, an agonist of P2X receptors, to stimulate H 2 O 2 production and the requirement of extracellular Ca 2ϩ for this response suggest P2X receptors mediate the production of H 2 O 2 in microglia. To determine whether the production of H 2 O 2 was mediated through the P2X 7 receptor, two selective inhibitors of P2X 7 , PPADS and oATP, were tested. Both PPADS and oATP blocked H 2 O 2 production by BzATP treatment (Fig. 4A) suggesting that BzATP activates H 2 O 2 production primarily through the P2X 7 receptor. Further support for the role of P2X 7 in H 2 O 2 production was obtained by treating cells with Brilliant Blue G, a potent and highly selective inhibitor of P2X 7 , at nanomolar concentrations (26). Brilliant Blue G (500 nM) inhibited BzATP (250 M)-induced H 2 O 2 production by more than 80% (Fig. 4A).
To determine whether oATP and PPADS affected Ca 2ϩ responses similarly, Ca 2ϩ changes were measured in cells pretreated with oATP and PPADS in the presence or absence of BzATP. oATP (100 M) inhibited BzATP-induced Ca 2ϩ flux to near control levels (Fig. 4B). Similar results were obtained with PPADS (Fig. 4B). These results suggest that P2X 7 is the primary receptor stimulated by BzATP to generate H 2 O 2 .
Source of ROI-Several sources can contribute to the production of ROI. These include the classical NADPH oxidase, the mitochondrial respiratory chain, and microsomal enzymes. Pharmacological inhibitors of the NADPH oxidase were used to determine whether P2X 7 receptor activates NADPH oxidase. Three selective inhibitors with different mechanisms of action, diphenyleneiodonium chloride (DPI), apocyanin, and 4-(2-aminoethyl)-benzenesulfonyl fluoride (AEBSF) were used (27,28). As shown in Fig. 5A, all three inhibitors completely inhibited BzATP-induced H 2 O 2 release from microglia. To confirm the activation of NADPH oxidase by BzATP in microglia, a functional change in NADPH oxidase was examined. A critical step in the activation of the NADPH oxidase is the translocation of p67 phox from the cytosol to the membrane. In BzATP-stimulated microglia, p67 phox , which is primarily cytosolic, rapidly translocated to the particulate/membrane fraction (Fig. 5B). These results suggest that BzATP stimulates the release of ROI in microglia via the activation of the NADPH oxidase.
Signal  BzATP (Fig. 6). Nonetheless, PD98059 did attenuate BzATPinduced TNF␣ release suggesting that p42/44 MAPK was involved in cytokine signaling but not in the generation of H 2 O 2 (data not shown). LY294002, a selective phosphatidylinositol 3-kinase inhibitor (PI3-K) also significantly inhibited H 2 O 2 stimulated by BzATP. LY294002 at 50 and 10 M inhibited H 2 O 2 production by 74.7 Ϯ 2 (n ϭ 3) and 47 Ϯ 5% (n ϭ 3), respectively. The inhibitors used were not toxic to the cells at the concentrations used. These results suggest that whereas both p38 MAPK and p42/44 ERK are rapidly activated in microglia stimulated with BzATP, only the inhibition of p38 MAPK attenuates H 2 O 2 production. Furthermore, PI3-K may play a more important role in the release of H 2 O 2 than p38 MAPK because the PI3-K inhibitor blocks H 2 O 2 release to a greater extent than the p38 MAPK inhibitor.
Microglia Stimulated with ATP or BzATP Are Neurotoxic-To determine whether activation of microglia with ATP and BzATP is neurotoxic, a co-culture system using highly purified primary rat cortical neurons and primary rat microglia was employed. Stimulation of microglia with LPS resulted in massive production of TNF␣ and nitric oxide but no H 2 O 2 release (Figs. 1 and 7, B and C). LPS did not stimulate LDH release from microglia/cortical neurons co-cultures (Fig. 7A). Conversely, ATP or BzATP stimulated very little TNF␣ and nitric oxide production but induced neurotoxicity at 72 h (Fig. 7). There was no significant neurotoxicity up to 48 h post-stimulation with either ATP or BzATP. The amount of neurotoxicity was much greater with BzATP compared with ATP, which is consistent with the observation that BzATP is a more potent stimulus compared with ATP in the generation of H 2 O 2 . The values shown in Fig. 7A represent LDH released from both microglia and cortical cells minus LDH released from microglia alone. A WST-1 cell survival assay was used to confirm that the LDH release is a measure of neuronal toxicity. Neither ATP nor BzATP had a significant neurotoxic effect on cortical neurons (Fig. 7A). Results comparing LPS-treated co-cultures to ATP/ BzATP-treated co-cultures show that factors other than TNF␣ or nitric oxide are contributing to toxicity in our system.
Up-regulation of P2X 7 Receptor in a Transgenic Mouse Model of Alzheimer's Disease-By having demonstrated that ATPand BzATP-stimulated microglia release H 2 O 2 and kill cortical neurons in vitro, we examined the brains of a transgenic mouse model of AD (Tg2576 carrying a APP (K670N,M671L) double mutation) (32) to determine whether the mechanism is important in vivo. In 24-month-old transgenic mouse brains but not in aged-matched control brains, plaques were evident in the hippocampus and surrounding outer cortical region when stained with two different antibodies for A␤ (Fig. 8, A, B, and E; data not shown). P2X 7 immunostaining gave a ring-like pattern around plaques only in transgenic mice suggesting that cells staining for P2X 7 were surrounding the plaques (Fig. 8, C, D,  and F). This staining was not evident in the absence of the primary antibody indicating that the staining was not due to nonspecific binding of the secondary antibody to the plaques. There was some basal staining with the P2X 7 antibody in both control and transgenic animals suggesting low levels of P2X 7 are expressed in the brain.
To determine whether the increased staining of P2X 7 in the transgenic mice was due to increased expression of P2X 7 receptor, lysates from the hippocampi (region with higher concentration of plaques) of three 19-month Tg2576 mice and age-matched controls were separated by SDS-PAGE, electrophoretically transferred, and probed with a polyclonal antibody specific for P2X 7 (Fig. 8G). The signal for the 55-kDa isoform was higher in the transgenic mice compared with age-matched controls. Preincubation of the P2X 7 antibody with the P2X 7 peptide (amino acid 576 -595) resulted in the disappearance of the P2X 7 band and had no effect on the actin band, demonstrating the specificity of the antibody (Fig. 8G).
To determine the identity of cells expressing P2X 7 around the plaques, serial sections were stained with markers selective for either pan-A␤ (Fig. 9A), P2X 7 (Fig. 9, B and E), microglia (CD45 (Fig. 9C)), astrocytes (GFAP (Fig. 9F)), or P2X 7 and GFAP (Fig. 9D). Microglial and astrocytic staining were seen surrounding plaques in the transgenic animals (Fig. 9, C and F,  respectively). It could not be determined if P2X 7 immunoreactivity was restricted specifically to either microglia or astrocytes; however, very little co-localization was detected by dual immunofluorescence with anti-GFAP and anti-P2X 7 (assessed by the number of dual labeled yellow cells (Fig. 9D)). These results show that P2X 7 is up-regulated in Tg2576 mice around amyloid plaques and is regionally localized with activated microglia and astrocytes. DISCUSSION The P2X 7 receptor has been implicated in the activation of transcription factors, apoptosis, and in the release of pro-inflammatory substances like TNF␣ and interleukin-1␤ in microglia (8,9,14,33). In this report we demonstrate that P2X 7 is the primary receptor involved in H 2 O 2 production in primary rat microglia stimulated with ATP or BzATP. The P2X-selec-tive agonist BzATP was a more potent stimulus than ATP, a P2Y/P2X agonist. Functionally, the activation of microglia with ATP or BzATP induced cell death in primary cortical neurons. In vivo there was a striking association of P2X 7 receptor-positive cells around plaques in a transgenic mouse model of Alzheimer's disease.
No detailed study has profiled the expression of P2Y and P2X receptors in microglia, but microglial expression of both P2X and P2Y is supported by electrophysiological studies (13,34). Our results demonstrating Ca 2ϩ changes induced by ATP or BzATP also support the existence of functional P2X and P2Y receptors on primary rat microglia.
Several lines of evidence point to P2X receptors and P2X 7 , in particular, as the primary receptor involved in the generation of H 2 O 2 in ATP-or BzATP-stimulated microglia. Stimulation of H 2 O 2 production by P2X-selective agonist BzATP (35,36) provides the first line of evidence for the involvement of P2X receptors. P2X receptors mobilize only extracellular Ca 2ϩ , consequently the experiments that demonstrate BzATP mobilizes only extracellular Ca 2ϩ provide further evidence for the involvement of this receptor. The inability of BzATP to stimulate H 2 O 2 production in the absence of extracellular Ca 2ϩ provides a direct link between Ca 2ϩ mobilization and H 2 O 2 production. The inhibition of H 2 O 2 production and Ca 2ϩ influx by P2Xselective antagonist PPADS, P2X 7 -selective antagonist oATP (17), and P2X 7 -selective inhibitor Brilliant Blue-G in BzATPstimulated cells provide additional lines of evidence for the involvement of P2X 7 receptors. The contribution of other purinergic receptors cannot be excluded, and can be evaluated only by measuring H 2 O 2 production in microglia lacking P2X 7 receptors. The levels of ATP required to stimulate the P2X 7 receptor in vitro suggests that the low concentrations of ATP found in the extracellular milieu of the brain would not be sufficient for microglia to induce neurotoxicity. However, recent reports (6) suggest that low concentrations of ATP can act as a chemoattractant for microglia directing them to a region of injury. Furthermore, ATP released from activated astrocytes has been shown recently (37) to activate microglia. Finally, a recent report (38) has documented the significance, the source, and needed high concentrations of ATP to activate P2X 7 receptors in the brain, as well as the role of ATP in neurodegeneration.
Endogenous  Twoyear-old Tg2576 (B and D-F) and agematched control mice (A and C) were stained for ␤-amyloid with a pan-A␤ antibody (A, B, and E) and P2X 7 with a polyclonal anti-P2X 7 antibody (C, D, and F). Magnification bar ϭ 300 m for A-D and 50 m for E-F. G, hippocampal lysates from three 19-month-old Tg2576 mice or age-matched controls were probed with an antibody against P2X 7 . The specificity of the P2X 7 band was confirmed by the use of a blocking peptide as described under "Materials and Methods." The blot was then stripped and reprobed with an antibody against actin. and ROIs, toxicity is likely due to a combination of factors (41)(42)(43). The inability of LPS-activated microglia to induce cell death in cortical neurons despite large increases in TNF␣ and inducible nitric-oxide synthase shows that these products do not solely induce cell death in our microglia/cortical neuron co-culture system. Published reports support this observation. Klegeris et al. (44,45) demonstrated that LPS-stimulated microglia are not neurotoxic but in combination with IFN␥ induces neurotoxicity. IFN␥ can prime microglia to generate O 2 .
in the absence and presence of TNF␣ suggesting that TNF␣ could be indirectly inducing cell death by generating O 2 . (46,47). Moreover, both IFN␥/LPS or IFN␥/TNF␣ can up-regulate P2X 7 expression in THP-1 cells and monocytes (48,49). Given that ROIs can directly induce neurotoxicity, it will be important to determine the exact mechanism by which ATP and BzATP are inducing cortical cell death. The link between LPS, IFN␥, and P2X 7 expression has yet to be examined in detail in microglia.
Currently there are no reports on the expression profile of P2X 7 in Alzheimer's disease. Ours is the first paper demonstrating a remarkable difference in the staining pattern for P2X 7 in brain slices of a transgenic (Tg2576) mice model. This intense staining for P2X 7 around plaques can be the result of up-regulation of the P2X 7 receptor and/or aggregation of glia around plaques. The fact that P2X 7 message and receptor is up-regulated in monocytes treated with LPS/IFN␥ or LPS/ TNF␣ (48,50) raises the possibility that P2X 7 indeed can be up-regulated in the mouse model of AD. The increased P2X 7 immunoreactivity in immunoblots comparing Tg2576 hippocampi lysates to age-matched control hippocampi lysates and the presence of P2X 7 -immunopositive cells around plaques supports this theory. However, the identity of the P2X 7 immunopositive cells around plaques is still not clear, even though activated microglia and astrocytes are found in the same vicinity. Although a correlation has been established between increased P2X 7 immunoreactivity and amyloid plaques, the question of cause and effect is beyond the scope of this paper. Because P2X 7 knock out mice have been made, it will be interesting to see if microglia from these mice generate H 2 O 2 in response to BzATP or if a P2X 7 knock out mouse crossed with a Tg2576 mouse would have any alteration in plaque deposition. It is possible that receptor antagonists of P2X 7 could have therapeutic utility in treatment of AD by regulating pathologically activated microglia.