Cellular/Molecular

Production and Release of Neuroprotective Tumor Necrosis Factor by P2X7 Receptor-Activated Microglia

Tomohisa Suzuki, Izumi Hide, Katsutoshi Ido, Shinichi Kohsaka, Kazuhide Inoue and Yoshihiro Nakata
Journal of Neuroscience 7 January 2004, 24 (1) 1-7; DOI: https://doi.org/10.1523/JNEUROSCI.3792-03.2004

Abstract

After a brain insult, ATP is released from injured cells and activates microglia. The microglia that are activated in this way then release a range of bioactive substances, one of which is tumor necrosis factor (TNF). The release of TNF appears to be dependent on the P2X7 receptor. The inhibitors 1,4-diamino-2,3-dicyano-1,4-bis[2-amino-phenylthio]butadiene (U0126), anthra[1,9-cd]pyrazol-6(2H)-one (SP600125), and 4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)IH-imidazole (SB203580), which target MEK (mitogen-activated protein kinase kinase), JNK (c-Jun N-terminal kinase), and p38, respectively, all potently suppress the production of TNF in ATP-stimulated microglia, whereas the production of TNF mRNA is strongly inhibited by U0126 and SP600125. SB203580 did not affect the increased levels of TNF mRNA but did prevent TNF mRNA from accumulating in the cytoplasm. The ATP-provoked activation of JNK and p38 [but not extracellular signal-regulated kinase (ERK)] could be inhibited by brilliant blue G, a P2X7 receptor blocker, and by genistein and 4-amino-5-(4-chlorophenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine, which are general and src-family-specific tyrosine kinase inhibitors, respectively. Most important, we found that treatment of the microglia in neuron-microglia cocultures with the P2X7 agonist 2′-3′-O-(benzoyl-benzoyl) ATP led to significant reductions in glutamate-induced neuronal cell death, and that either TNF-α converting enzyme inhibitor or anti-TNF readily suppressed the protective effect implied by this result. Together, these findings indicate that both ERK and JNK are involved in the regulation of TNF mRNA expression, that p38 is involved in the nucleocytoplasmic transport of TNF mRNA, and that a PTK (protein tyrosine kinase), possibly a member of the src family, acts downstream of the P2X7 receptor to activate JNK and p38. Finally, our data suggest that P2X7 receptor-activated microglia protect neurons against glutamate toxicity primarily because they are able to release TNF.

Introduction

ATP is released from damaged cells as a result of ischemia or inflammation and serves as a cell-to-cell mediator through cell surface P2 receptors, which are widely distributed throughout the nervous system (Inoue, 2002). P2 receptor subfamilies have been divided into two subtypes: P2X and P2Y. P2X receptors (P2X1-P2X7) are coupled to nonselective cation channels, allowing influx of Na+ and Ca2+, leading to transient cell depolarization, whereas P2Y receptors (P2Y1, P2Y2, P2Y4, P2Y6, P2Y11, P2Y12, P2Y13, and P2Y14) are G-protein coupled, and their activation leads to inositol lipid hydrolysis, intracellular Ca2+ mobilization, or modulation of adenylate cyclase activation (North and Surprenant, 2000; Sak and Webb, 2002). Microglia have been shown to express multiple P2 receptor subtypes, including P2X7, P2Y2, and P2Y12 (Norenberg et al., 1994; James and Butt, 2002), indicating that ATP may be a critical regulator of microglial cell function. In fact, ATP strongly induces microglial chemotaxis via the Gi- and Go-coupled P2Y12 receptor (Honda et al., 2001) and stimulates the release of plasminogen, interleukin-6 (IL-6), and IL-1β (Ferrari et al., 1997; Inoue et al., 1998; Shigemoto-Mogami et al., 2001) by means of different types of P2 receptor and intracellular signals. ATP also stimulates the de novo synthesis and release of tumor necrosis factor (TNF) from rat microglia that flows from the activation of P2X7 receptors, extracellular signal-regulated kinase (ERK), and p38 mitogen-activated protein (MAP) kinase (Hide et al., 2000). The precise roles of each of these MAP kinases in P2X7 receptor-mediated TNF release have not yet been determined. Moreover, although the ATP-mediated activation of ERK and p38 by ATP occurs in the absence of extracellular Ca2+, it is still not known precisely how MAP kinases are activated via P2X7 receptors.

TNF is a proinflammatory cytokine that is upregulated in the brain in response to various insults or injury. This cytokine is mainly expressed by microglia and astrocytes around the injured area. The function of TNF, however, remains controversial. Although TNF has been implicated in the acceleration of injury and the pathology of neurodegenerative diseases, recently emerging evidence suggests that TNF may also serve a protective role. The precise mechanisms involved in these two seemingly conflicting activities are still undetermined, as is the role of the TNF that is released by P2X7 receptor-activated microglia.

In the present study, we investigated the signaling mechanism by which the activation of microglial P2X7 receptors regulates TNF release via ERK, JNK (c-Jun N-terminal kinase), and p38 MAP kinase. We also examined whether the TNF released by microglia is toxic or protective for neurons. Our data demonstrate that ERK, JNK, and p38 contribute in different ways to the ATP-provoked production of TNF. In addition to outlining a new mechanism for the activation of JNK and p38 via the P2X7-src-family protein tyrosine kinase pathway, we provide the first evidence that P2X7 receptor-activated microglia protect neurons against glutamate neurotoxicity because they release TNF.

Materials and Methods

Reagents. All reagents for cell culture, ATP, and 2′-3′-O-(benzoyl-benzoyl) ATP (BzATP) were obtained from Sigma (St. Louis, MO), and a rat TNF ELISA kit was obtained from Biosource International (Camarillo, CA). Antibody kits for p42-p44 (ERK), JNK, and p38 MAP kinase were obtained from Cell Signaling Technology (Beverly, MA). All other reagents were purchased from commercial sources and were of the highest available purity. We checked all reagents used for endotoxin contamination.

Cell culture. Microglia were obtained from primary cell culture of neonatal rat brains as described previously (Nakajima et al., 1989, 1992). After 7-16 d in culture, microglia were prepared as a floating cell suspension. Aliquots (1.5-2.0 × 105 cells) were transferred to the wells of a 24 well plate and allowed to adhere at 37°C for 45 min. Unattached cells were removed by rinsing with serum-free DMEM. Neuron-enriched cultures were prepared from primary cell culture of neonatal rat brain cortex. For neuron-microglia cocultures, cortical neurons were seeded into 24 well plates, and microglia were transferred in Transwell coculture inserts (Costar, Cambridge, MA).

TNF assay. For the assay of TNF release, microglia were incubated with 0.4 ml of serum-free DMEM with or without drugs for 3 hr. For the assay of intracellular TNF, microglia were incubated for 2 hr and then solubilized with 0.1% Triton X-100. TNF was assayed in 50 μl samples using a rat TNF ELISA kit according to the instructions of the manufacturer.

RNA extraction. Microglia were plated in 60 mm dishes (7.5 × 105 cells per 3 ml per dish) and stimulated with ATP or the P2X7 agonist BzATP for 1 hr. The cells were dissolved in Trizol LS reagents (Invitrogen, Gaithersburg, MD) before total RNA extraction. Briefly, samples were ethanol precipitated, the pellets were dissolved in RNase-free water, and RNA concentrations were determined spectrophotometrically.

Real-time reverse transcription-PCR assay. Total TNF mRNA was measured by real-time quantitative reverse transcription (RT)-PCR (ABI Prism model 7700 sequence detection system; PerkinElmer Applied Biosystems, Foster City, CA). RT-PCR was performed using the TaqMan one-step RT-PCR master mix reagents kit according to the protocol of the manufacturer (PerkinElmer Applied Biosystems). The sequences of the forward and reverse primers were 5′-ACAAGGCTGCCCCGACTAC-3′ and 5′-TCCTGGTATGAAATGGCAAACC-3′, respectively. The TaqMan fluorogenic probe was 5′-6FAM-TGCTCCTCACCCACACCGTCAGCTAMRA-3′. During PCR amplification, 5′ nuclease activity of AmpliTaq Gold DNA polymerase cleaves the TaqMan probe, separating the 5′ reporter dye from 3′ quencher dye, resulting in increased fluorescence of the reporter. Reaction conditions for PCR were as follows: 40 cycles of amplification by denaturing to 95°C for 15 sec and extending at 60°C for 1 min. The threshold cycle, which correlates inversely with the target mRNA levels, was measured as the cycle number at which the reporter fluorescent emission increases above a threshold level. The TNF mRNA levels were corrected for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) RNA using a VIC probe according to the protocol of the manufacturer.

Western blot analysis. Western blots were performed for the analysis of ERK (p44 and p42), JNK, and p38 activation using a MAP kinase antibody kit according to the protocol of the manufacturer. In brief, the cells were washed with PBS, lysed by adding SDS sample buffer, and sonicated. After heating to 95°C for 5 min, the protein samples were separated by SDS-PAGE and blotted onto polyvinylidene difluoride membranes. The membranes were blocked with blocking buffer containing 3% skim milk for 3 hr at room temperature and incubated with primary antibody with gentle agitation overnight at 4°C. After washing, the membranes were incubated for 1 hr at room temperature with horseradish peroxidase-conjugated secondary antibody and horseradish peroxidase-conjugated anti-biotin antibody to detect biotinylated protein markers. The membranes were then washed and incubated with Lumi GLO, and the proteins were detected by exposure to x-ray film.

Viability assay. The viability of neurons was assessed by a 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay on the basis of the reduction of the tetrazolium salt MTT by the mitochondrial dehydrogenase in living cells.

Statistical analysis. Student's t test was used in all determinations of statistical significance.

Results

The roles of ERK, JNK, and p38 MAP kinase in TNF mRNA expression and the production and release of TNF in ATP-stimulated microglia

ATP rapidly activates ERK, JNK, and p38 MAP kinase in microglia. To investigate possible roles of ERK, JNK, and p38 in ATP-induced TNF release, we examined the inhibitory effects of 10 μm 1,4-diamino-2,3-dicyano-1,4-bis[2-amino-phenylthio]-butadiene (U0126) [which inhibits MEK (mitogen-activated protein kinase kinase)], 30 μm anthra[1,9-cd]pyrazol-6(2H)-one (SP600125) (which inhibits JNK), and 15 μm 4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)IH-imidazole (SB203580) (which inhibits p38) on the release of TNF from ATP-stimulated microglia. All three inhibitors proved to be capable of reducing the amounts of TNF released by ATP-treated microglia very significantly, suggesting that TNF release is dependent on the activities of all three MAP kinases (Fig. 1A). To confirm the supposed selectivity of these particular inhibitors, we conducted experiments with the two additional inhibitory compounds 2-(2′-amino-3′-methoxyphenyl)-oxanaphthalen-4-one (PD98059) (which inhibits MEK) and 4-(4-fluorophenyl)-2-(4-hydroxyphenyl)-5-(4-pyridyl)IH-imidazole (SB202190) (which inhibits p38) and found that 25 μm PD98059 and 10 μm SB202190 were also capable of potently suppressing the ATP-induced release of TNF (by 57 and 60%, respectively). Next we examined the effects of three of these inhibitors (U0126, SP600125, and SB203580) on the intracellular accumulation of TNF in microglia during a 2 hr period of stimulation with ATP to determine whether all three MAP kinases are involved in the regulation of the de novo production of TNF. Once again, all three inhibitors markedly reduced the intracellular levels of TNF in ATP-stimulated microglia (Fig. 1B). These data suggest that ERK, JNK, and p38 MAP kinase are all involved in both the production and the release of TNF by ATP-treated microglia.

Figure 1.
Figure 1.

Effects of U0126 (U), SP600125 (SP), and SB203580 (SB) on ATP-induced TNF release, intracellular TNF production, and mRNA expression in microglia. The cells were treated with 10 μm U0126, 30 μm SP600125, or 15 μm SB203580 for 15 min and stimulated with 1 mm ATP for 3 hr (A), 2 hr (B), and 1 hr (C). The released TNF (A) and the intracellular TNF contents (B) were measured by ELISA. Values are expressed as mean ± SEM of percentage of release compared with ATP alone from three independent experiments. Values of 100% for the release and intracellular production of TNF in ATP-stimulated microglia were 178.8 ± 22.2 and 407.0 ± 45.6 pg/106 cells, respectively. C, The expression of TNF mRNA was quantified by real-time RT-PCR. Values are shown as the ratio of TNF mRNA versus GAPDH mRNA. Data are expressed as mean ± SEM of ratio of expression compared with ATP or BzATP alone from three independent experiments. **p < 0.01; ***p < 0.001, significantly different from the control (Con) (t test).

To determine whether all three kinases also regulate TNF transcription, we used real-time RT-PCR analysis to examine the effects of the three inhibitors on ATP-induced TNF mRNA expression. As shown in Figure 1C, ATP treatment led to a marked increase in TNF mRNA levels; moreover, the extent of this increase was greatly reduced in the presence of either U0126 or SP600125 but did not appear to be affected by SB203580. Thus, although all three kinases appear to be critical for ATP-induced TNF production, they appear to have different modes of action, with ERK and JNK acting to regulate transcription of the TNF gene and p38 only acting post-transcriptionally to interfere with the production or release of TNF.

p38 mediates the transport of TNF mRNA from the nucleus to the cytoplasm

A recent report indicates that the nucleocytoplasmic transport of TNF mRNA in lipopolysaccharide (LPS)-stimulated macrophages is regulated by ERK-dependent signals (Dumitru et al., 2000). We therefore attempted to clarify the mechanism by which p38 mediates post-transcriptional regulation of TNF production by examining the localization of TNF mRNA in the nucleus and the cytoplasm of microglia. We used Western blots of the marker proteins octamer-binding protein 1 (Oct-1) and heat shock protein 90 (Hsp-90) to confirm the efficacy of the technique used to separate the nuclear and cytoplasmic fractions (Fig. 2B). Our results indicated that ATP treatment led to significantly increased TNF mRNA levels in the cytoplasm as well as in the nucleus; presumably, this meant that the TNF mRNA had been exported from the nucleus to the cytoplasm so that new TNF protein molecules could then be synthesized on ribosomes (Fig. 2A). Interestingly, exposure of ATP-treated microglia to SB203580 did not appear to have any effect on the significantly increased levels of TNF mRNA that we observed in nuclei, but it did appear to prevent any comparable increase in cytoplasmic TNF mRNA levels from becoming evident (Fig. 2A). These data could be indicating that p38 has a role in regulating the nucleocytoplasmic transport of TNF mRNA in ATP-stimulated microglia, although we cannot rule out other possible explanations, including, for example, differences in the half-life of TNF mRNA in the nucleus and the cytoplasm.

Figure 2.
Figure 2.

Distribution of TNF mRNA induced by ATP in the nucleus and cytoplasm of microglia treated with SB203580 (SB). A, The cells were treated with 15 μm SB203580 for 15 min and stimulated with 1 mm ATP for 1 hr. Nuclear and cytoplasmic fractions of the cells were separated by NE-PER kit, and RNA was extracted from each fraction. Values are shown as the ratio of TNF mRNA levels versus GAPDH mRNA levels. Data are expressed as mean ± SEM of ratio of expression compared with ATP alone from three independent experiments. *p < 0.05 significantly different from the control (Con) (t test). B, Nuclear and cytoplasmic extracts were separated by SDS-PAGE and probed with antibodies to Oct-1 [nuclear protein (n)] and Hsp-90 [cytoplasmic protein (c)] to determine the efficiency of nucleocytoplasmic separation.

P2X7 receptors are coupled to JNK and p38 activation but not to ERK activation

One reason for suspecting that ATP-induced TNF release is mediated by P2X7 receptors is that the optimal concentration of ATP that induces TNF release is approximately equivalent to the concentration needed to activate P2X7 (i.e., of the order of 1 mm). Moreover, although we reported previously that BzATP (a P2X7 agonist) is able to mimic the effects of ATP in provoking TNF release (Hide et al., 2000), we now find that α,β-methylene ATP (which is known to be a P2X receptor agonist and not a P2X7 agonist) has no such effect (data not shown). We were able to confirm the involvement of P2X7 receptors by showing that brilliant blue G (BBG), a P2X7 receptor blocker, suppresses ATP-as well as BzATP-induced TNF release (Fig. 3A). We then tried to determine whether P2X7 receptors are able to activate MAP kinases by examining the effects of BBG on the ATP-induced activation of ERK, JNK, and p38. The results in Figure 3C indicate that BBG is capable of selectively inhibiting the ATP-induced activation of JNK and p38 but appears to have little if any effect on the activation of ERK. Similar results were obtained using BzATP-stimulated cells (Fig. 3B,D). These results indicate that the activation of P2X7 receptors does lead to JNK and p38 activation, and that ERK activation occurs by some other, as yet unidentified, route.

Figure 3.
Figure 3.

Effects of BBG on the release of TNF and the activation of ERK, JNK, and p38 MAP kinase in ATP- or BzATP-stimulated microglia. The cells were treated with BBG (0.1 or 1 μm) for 5 min and stimulated with 1 mm ATP (A, C) or 100 μm BzATP (B, D) for 3 hr (A, B) or 10 min (C, D). A, B, The release of TNF was measured by ELISA. Values are expressed as mean ± SEM of percentage of release compared with ATP or BzATP alone from six independent experiments. Values for 100% of release of TNF were 183.8 ± 46.3 and 350.7 ± 152.2 pg/106 cells in ATP- or BzATP-stimulated microglia, respectively. *p < 0.05; ***p < 0.001, significantly different from the control (Con) (t test). C, D, The phosphorylated (active) and total ERK, JNK, and p38 were detected by Western blotting using antibodies that recognize phosphorylated and both phosphorylated and nonphosphorylated enzymes, respectively. The levels of each total MAPK were confirmed to be identical for each lane. Similar results were obtained in at least three independent experiments.

The P2X7 receptor-mediated activation of JNK and p38 is protein tyrosine kinase dependent

Despite the fact that the P2X7 receptor can and does function as a cation channel, activation of JNK and p38 appears to be independent of extracellular Ca2+ (Hide et al., 2000) and may therefore be dependent on a mechanism that is independent of Ca2+ influx. We therefore decided to investigate a possible involvement of PTK (protein tyrosine kinase) in this pathway by studying the effects of a general PTK inhibitor, genistein, on both TNF release and the activation of each of the MAP kinases that are induced by ATP. We found that genistein significantly inhibited the ATP-induced activation of JNK and p38 (but not that of ERK) (Fig. 4C) as well as the ATP-induced release of TNF, and that it did all of these things in a concentration-dependent manner (Fig. 4A). We obtained similar results with a second PTK inhibitor known as tyrphostin A25 (data not shown). We then found that the src-family PTK inhibitor PP2 also appeared to suppress ATP-induced TNF release, and that its negative control counterpart PP3 had no comparable effect (Fig. 4B). PP2 also appeared to inhibit the BzATP-induced activation of JNK and p38 but not that of ERK (Fig. 4D). Together, these findings suggest that a PTK, almost certainly a src-family PTK, can act downstream of the P2X7 receptor to activate both JNK and p38; this could then lead to the synthesis and subsequent transport of TNF mRNA.

Figure 4.
Figure 4.

Effects of genistein and PP2 on the release of TNF and the activation of ERK, JNK, and p38 MAP kinase induced by ATP. Microglia were treated with genistein, a nonselective PTK inhibitor (A, C), with PP2, an src-family-selective PTK inhibitor, or with PP3, an inactive analog of PP2 (B, D); stimulated with 1 mm ATP or 100 μm BzATP; and then measured for the release of TNF after 3 hr (A, B) and MAP kinase activation after 10 min (C, D). Values are expressed as mean ± SEM of percentage of release compared with ATP alone from three independent experiments. Values of 100% for the release of TNF were 136.0 ± 37.0 (A) and 133.6 ± 65.1 (B) pg/106 cells. *p < 0.05; **p < 0.01, significantly different from the control (Con) (t test). C, D, The phosphorylated (Phospho) (active) and total ERK, JNK, and p38 were detected by Western blotting using antibodies that recognize either phosphorylated or both phosphorylated and unphosphorylated enzymes, respectively. The (total) levels of each MAP kinase were confirmed to be identical for each lane. Similar results were obtained in at least three independent experiments. Geni, Genistein.

P2X7 receptor-stimulated microglia rescue the neurons from glutamate-induced cell death by releasing TNF

Possible interactions between microglia, the TNF they produce, and neuronal tissue are still not well understood. To determine whether P2X7 receptor-stimulated microglia and the TNF they release are harmful or protective to surrounding neurons, we developed a neuron-microglia coculture system in which primary cultured neurons and microglia are held in separate compartments that nonetheless allow them to exchange freely diffusible factors. A preliminary experiment then showed that even a brief (10 min) exposure of neurons to 100 μm glutamate followed by culture for 24 hr led to ∼50% of them dying (Fig. 5A). There was no significant difference in glutamate neurotoxicity between cultures of neurons alone and neuron-microglia cocultures (Fig. 5A,B), although the neurons in cocultures that had been exposed to BzATP for 24 hr before the addition of glutamate appeared to have been given a significant level of protection against glutamate-induced cell death (Fig. 5B). No comparable protection was evident in neurons that were cultured on their own with BzATP (Fig. 5A), nor was there any evidence of a reduction in their viability (Fig. 5A,B). Interestingly, we also found that TNF-α protease inhibitor (TAPI), a well known inhibitor of TNF-α-converting enzyme (TACE), strongly inhibited the BzATP-induced release of TNF (Fig. 5D), and that both TACE and anti-TNF antibody significantly suppressed the protective effect that is afforded to neurons by BzATP-treated microglia (Fig. 5C). Furthermore, this protective effect was abolished by BBG, confirming the involvement of P2X7 receptor (Fig. 5C). Similar results were obtained using ATP-stimulated microglia (data not shown). Together, our results strongly indicate that the neuroprotective effect of P2X7 receptor-stimulated microglia stems very directly from their ability to release TNF. Whether other substances are involved remains to be seen.

Figure 5.
Figure 5.

Effects of BzATP on the viability of primary cortical neurons in “neuron alone” cultures or neuron-microglia cocultures. Primary cultures of rat cortical neurons alone (A) or cocultures of neurons with microglia (B) were treated with 100 μm BzATP for 24 hr and then stimulated with 100 μm glutamate for 10 min. After 24 hr of incubation, neuronal cell viability was determined by MTT assay. Values are expressed as mean ± SEM of percentage of viability of control cells from three independent experiments. **p < 0.01, significantly different from glutamate alone (t test). n.s., Not significant. C, Effects of TAPI, TACE inhibitor, anti-TNF, and BBG on BzATP-induced neuroprotection. Cortical neurons cocultured with microglia were treated with 50 μm TAPI, 10 μg/ml anti-TNF, and 1 μm BBG for 5 min before BzATP application. After 24 hr of coculture, the neurons were stimulated with 100 μm glutamate for 10 min. After 24 hr of additional incubation, neuronal cell viability was determined by MTT assay. Values are expressed as mean ± SEM of percentage of viability of control cells from three independent experiments. **p < 0.01, significantly different from BzATP-glutamate application (t test). D, Effects of TAPI on BzATP-induced TNF release from rat microglia. ***p < 0.001.

Discussion

In the present study, we have shown that three MAP kinase family members (ERK, JNK, and p38) all make important contributions to the ATP-induced production of TNF in microglia, and that their precise individual roles may differ. Thus, ERK and JNK appear to regulate TNF mRNA synthesis, whereas the effects of p38 seem to be limited to the post-transcriptional level and probably involve the nucleocytoplasmic transport of TNF mRNA. Other findings led us to suggest that activation of JNK and p38 by the P2X7 receptor is likely to be mediated by a src-family PTK. This study also revealed the biological significance of the TNF released from P2X7 receptor-activated microglia.

The precise mode of action of ERK and JNK in microglial TNF production is not yet known, but it may be that they exert their effects on the promoter domain of the TNF gene by activating one of the nuclear factor-κB, nuclear factor of activated T cells, or activator protein-1 transcription factors. As the p38-mediated post-transcriptional regulation of TNF production, it is known, for example, that p38 regulates the expression of several RNA-binding proteins, which then help to stabilize TNF mRNA by interacting with the adenine uracil-rich element (ARE) region in the 3′ untranslated region of TNF mRNA (Kontoyiannis et al., 2001; Mahtani et al., 2001). Other proteins such as eukaryotic polypeptide chain initiation factor-4E, eukaryotic polypeptide chain elongation factor-1, and eukaryotic polypeptide chain releasing factor-1 also appear to contribute to translational efficiency, whereas elF-4E, a translation start factor, responds to a regulatory process downstream of p38 MAP kinase (Lee et al., 2000). Interestingly, Dumitru et al. (2000) have suggested that the transport efficacy of TNF mRNA is regulated through the ARE via ERK, although our findings suggest that p38-mediated post-transcriptional regulation of TNF production regulates the nucleocytoplasmic transport of TNF mRNA in ATP-activated microglia. Another recent paper suggested that MAP kinase-activated protein kinase 2 (MAPKAPK2), which is activated in response to direct phosphorylation by p38, also appears to be involved in the post-transcriptional regulation of TNF biosynthesis in LPS-stimulated macrophages, but it is not yet known whether MAPKAPK2 participates directly in the regulatory process (Kotlyarov et al., 1999).

Recent reports indicate that P2X7 receptors are implicated in diverse cell functions, such as apoptosis, transcription, microvesicle shedding, and synaptic transmission (Humphreys et al., 2000; MacKenzie et al., 2001; Armstrong et al., 2002), and that they have a role in the release of cytokines such as IL-1β and IL-18 (Sanz and Di Virgilio, 2000; Mehta et al., 2001). Activation of P2X7 receptors may also be involved in triggering the release of TNF from rat microglia, given that: (1) a relatively high concentration of ATP (1 mm) is needed to ensure that TNF is released, (2) the P2X7 agonist BzATP is much more effective as an inducer of TNF release than ATP (Hide et al., 2000), and (3) the P2X7 antagonist BBG is a powerful inhibitor of ATP-BzATP-induced TNF release (Fig. 3A,B). We have also shown that BBG inhibits the activation of JNK and p38 but not ERK, thereby indicating that P2X7 receptors may well be involved in the activation of JNK and p38. The P2X7 receptor signaling pathways that lead up to the activation of MAP kinases still have to be properly characterized. Our data suggest that a PTK may be involved in activating JNK and p38 but not ERK. Although we do not know which PTK is involved, it may be an src-family tyrosine kinase such as Lyn or Lck, primarily because the src-family PTK-selective inhibitor PP2, but not its negative control PP3, strongly suppresses JNK and p38 activation as well as ATP-induced TNF release. As suggested recently by Denlinger et al. (2001), the Src homology 3 (SH3) domains of src-family kinases may well interact with the SH3 domain-binding motif in the P2X7 C-terminal domain. It has also been reported that AMPA receptors, which also act as glutamate ion channel receptors, are capable of promoting glial cell line-derived neurotrophic factor release by a mechanism that is: (1) dependent on the Lyn tyrosine kinase and (2) independent of ion flux (Hayashi et al., 1999). ATP-induced p38 and JNK activation does not appear to be associated with Ca2+ influx (Hide et al., 2000; our unpublished observation), so it is conceivable that any src-family PTK family that plays a part in the transduction of the P2X7 receptor does so because of its C-terminal domain rather than because of its ability to function as an ion channel.

P2X7 receptors are believed to be involved in the activation of ERK in several cell types (Panenka et al., 2001; Budagian et al., 2003; Gendron et al., 2003). In the present study, however, activation of ERK by ATP (or even by the more powerful BzATP) was not sensitive to the P2X7 antagonist BBG. Several lines of evidence point to the involvement of other P2 receptors in the activation of ERK. It has been demonstrated, for instance, that in rat primary astrocytes, ATP activates ERK via a P2Y-PKCδ-phospholipase D pathway (Neary et al., 1999), whereas the ATP released during stretch is more likely to activate ERK through P2X2 and P2Y1 (Neary et al., 2003). ADP and UTP also activate ERK possibly via P2Y12 and P2Y2 receptors expressed on microglia. Together, these findings lend support to the view that ERK activation may involve several non-P2X7 P2 receptors. However, a recent study in which P2X7 C-terminal and N-terminal mutants were expressed in human embryonic kidney 293 cells established that the N terminal of P2X7 has an essential role in ERK activation and that the C terminal is required for Ca2+ entry (Amstrup and Novak, 2003). One possible explanation for the obvious discrepancy between these results and our findings could be that the P2X7 N-terminal-associated activation signals are insensitive to BBG; unfortunately, we do not yet know whether this is the case. Nevertheless, ERK activation is clearly important for TNF-producing cells, given that it plays a key role in both the production of TNF and its subsequent release and that it does so by modulating the activity of TACE, a protein that converts 26 kDa membrane-bound pro-TNF into 17 kDa soluble TNF at the cell surface (Bezzi et al., 2001).

It is still not entirely clear whether microglia protect or harm neurons or whether TNF is beneficial or toxic (Arnett et al., 2001; Combs et al., 2001; Fontaine et al., 2002). Thus, for example, TNF appears to enhance injury, as shown by the fact that the injection of neutralizing TNF antibody into lesion sites significantly reduces experimental ischemic and traumatic injury (Barone et al., 1997; Meistrell et al., 1997). There is also recent evidence that indicates that TNF can provide protection to neurons because it is able to encourage the expression of anti-apoptotic and anti-oxidative proteins. Moreover, evidence from experiments in TNF-deficient mice indicates that although TNF has a deleterious effect during the acute response that occurs in a traumatized brain, it also has a key part to play in both the long-term behavioral recovery and the histological repair of the tissues (Scherbel et al., 1999). It is of course possible that the effects of TNF depend just as much on the postinjury time course as on its expression levels. Recent reports indicate that the dual actions of TNF are mediated via different TNF receptors, with the p55 TNF receptor 1 (TNFR1) eliciting neurotoxic effects and the p75 TNF receptor 2 (TNFR2) eliciting neuroprotection (Yang et al., 2002). Interestingly, these two receptors have been shown to have similarly specific roles in oligodendrocytes, with TNFR1 being implicated in demyelination and TNFR2 in remyelination (Arnett et al., 2001).

Given that one of our primary aims in this study was to better understand the biological significance of the TNF that is released from ATP- or BzATP-activated microglia, we decided to construct a primary neuron-microglia coculture system that could be used to screen for possible effects. Our results clearly demonstrate that BzATP-stimulated microglia provide neurons with effective protection against glutamate-induced cell death. This protective effect appears to be mediated by a soluble factor or factors released from P2X7-activated microglia. One such neuroprotective factor is almost certain to be TNF, because microglia-mediated neuroprotection was suppressed in the presence of either TACE inhibitor or anti-TNF antibody. Given that ATP is likely to be released by cells that have been damaged by trauma, inflammation, or ischemia, and that low ATP concentrations can lead to chemotaxis of microglia (Honda et al., 2001), it seems not unreasonable to hypothesize that ATP can act as an emergency messenger that recruits microglia to a damaged brain area, at which point a pathway that involves ATP-activated P2X7 receptors takes over and provokes the microglia to secrete neuroprotective factors such as TNF.

Here, we provided new information about the mechanisms that underlie the production of TNF via P2X7 receptors and MAP kinases in rat microglia and have clearly established a neuroprotective role for the TNF that the ATP-activated microglia may go on to release. A better understanding of the mechanisms by which microglia are transformed into cells that can protect neurons, including the mechanism that leads to P2X7 receptor activation, will undoubtedly help in the development of more rational approaches to the entire spectrum of neural diseases.

Footnotes

  • This work was supported by Grants-in-Aid 11670089 and 15590229 from the Ministry of Education, Science, and Culture of Japan and partly supported by the Promotion of Fundamental Studies in Health Science of the Organization for Pharmaceutical Safety and Research. We thank Dr. Peter Thatam (University College London, London, UK), Dr. Donald MacPhee (Radiation Effects Research Foundation, Hiroshima, Japan), and Masaya Tanaka for helpful comments; Yuko Kawai and Akiyo Matsubara for technical assistance; and the Research Center for Molecular Medicine (Hiroshima University School of Medicine) for the use of its facilities.

  • Correspondence should be addressed to Izumi Hide, Department of Pharmacology, Graduate School of Biomedical Sciences, Hiroshima University, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8551, Japan. E-mail: ihide{at}hiroshima-u.ac.jp.

  • Copyright © 2004 Society for Neuroscience 0270-6474/04/240001-07$15.00/0

References

  1. ↵
    Amstrup J, Novak I (2003) P2X7 receptor activates ERK1/ERK2 independent of Ca2+ influx. Biochem J 374: 51-61.
    OpenUrlCrossRefPubMed
  2. ↵
    Armstrong JN, Brust TB, Lewis RG, MacVicar BA (2002) Activation of presynaptic P2X7-like receptors depresses mossy fiber-CA3 synaptic transmission through p38 mitogen-activated protein kinase. J Neurosci 22: 5938-5945.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    Arnett HA, Mason J, Marino M, Suzuki K, Matsushima GK, Ting JP (2001) TNFα promotes proliferation of oligodendrocyte progenitors and remyelination. Nat Neurosci 4: 1116-1122.
    OpenUrlCrossRefPubMed
  4. ↵
    Barone FC, Arvin B, White RF, Miller A, Webb CL, Willette RN, Lysko PG, Feuerstein GZ (1997) Tumor necrosis factor-alpha. A mediator of focal ischemic brain injury. Stroke 28: 1233-1244.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Bezzi P, Domercq M, Brambilla L, Galli R, Schols D, De Clercq E, Vescovi A, Bagetta G, Kollias G, Meldolesi J, Volterra A (2001) CXCR4-activated astrocyte glutamate release via TNFα: amplification by microglia triggers neurotoxicity. Nat Neurosci 4: 702-710.
    OpenUrlCrossRefPubMed
  6. ↵
    Budagian V, Bulanova E, Brovko L, Orinska Z, Fayad R, Paus R, Bulfone-Paus S (2003) Signaling through P2X7 receptor in human T cells involves p56lck, MAP kinases, and transcription factors AP-1 and NF-kappa B. J Biol Chem 278: 1549-1560.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Combs CK, Karlo JC, Kao SC, Landreth GE (2001) β-Amyloid stimulation of microglia and monocytes results in TNFα-dependent expression of inducible nitric oxide synthase and neuronal apoptosis. J Neurosci 21: 1179-1188.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Denlinger LC, Fisette PL, Sommer JA, Watters JJ, Prabhu U, Dubyak GR, Proctor RA, Bertics PJ (2001) Cutting edge: the nucleotide receptor P2X7 contains multiple protein- and lipid-interaction motifs including a potential binding site for bacterial lipopolysaccharide. J Immunol 167: 1871-1876.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    Dumitru CD, Ceci JD, Tsatsanis C, Kontoyiannis D, Stamatakis K, Lin JH, Patriotis C, Jenkins NA, Copeland NG, Kollias G, Tsichlis PN (2000) TNF-α induction by LPS is regulated posttranscriptionally via a Tpl2/ERK-dependent pathway. Cell 22: 1071-1083.
    OpenUrl
  10. ↵
    Ferrari D, Chiozzi P, Falzoni S, Hanau S, Di Virgilio F (1997) Purinergic modulation of interleukin-1β release from microglial cells stimulated with bacterial endotoxin. J Exp Med 185: 579-582.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    Fontaine V, Mohand-Said S, Hanoteau N, Fuchs C, Pfizenmaier K, Eisel U (2002) Neurodegenerative and neuroprotective effects of tumor necrosis factor (TNF) in retinal ischemia: opposite roles of TNF receptor 1 and TNF receptor 2. J Neurosci 22: 1-7.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Gendron FP, Neary JT, Theiss PM, Sun GY, Gonzalez FA, Weisman GA (2003) Mechanisms of P2X7 receptor-mediated ERK1/2 phosphorylation in human astrocytoma cells. Am J Physiol Cell Physiol 284: C571-C581.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Hayashi T, Umemori H, Mishina M, Yamamoto T (1999) The AMPA receptor interacts with and signals through the protein tyrosine kinase Lyn. Nature 397: 72-76.
    OpenUrlCrossRefPubMed
  14. ↵
    Hide I, Tanaka M, Inoue A, Nakajima K, Kohsaka S, Inoue K, Nakata Y (2000) Extracellular ATP triggers tumor necrosis factor-alpha release from rat microglia. J Neurochem 75: 965-972.
    OpenUrlCrossRefPubMed
  15. ↵
    Honda S, Sasaki Y, Ohsawa K, Imai Y, Nakamura Y, Inoue K, Kohsaka S (2001) Extracellular ATP or ADP induce chemotaxis of cultured microglia through Gi/o-coupled P2Y receptors. J Neurosci 21: 1975-1982.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    Humphreys BD, Rice J, Kertesy SB, Dubyak GR (2000) Stress-activated protein kinase/JNK activation and apoptotic induction by the macrophage P2X7 nucleotide receptor. J Biol Chem 275: 26792-26798.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Inoue K (2002) Microglial activation by purines and pyrimidines. Glia 40: 156-163.
    OpenUrlCrossRefPubMed
  18. ↵
    Inoue K, Nakajima K, Morimoto T, Kikuchi Y, Koizumi S, Illes P, Kohsaka S (1998) ATP stimulation of Ca2+-dependent plasminogen release from cultured microglia. Br J Pharmacol 123: 1304-1310.
    OpenUrlCrossRefPubMed
  19. ↵
    James G, Butt AM (2002) P2Y and P2X purinoceptor mediated Ca2+ signalling in glial cell pathology in the central nervous system. Eur J Pharmacol 447: 247-260.
    OpenUrlCrossRefPubMed
  20. ↵
    Kontoyiannis D, Kotlyarov A, Carballo E, Alexopoulou L, Blackshear PJ, Gaestel M, Davis R, Flavell R, Kollias G (2001) Interleukin-10 targets p38 MAPK to modulate ARE-dependent TNF mRNA translation and limit intestinal pathology. EMBO J 20: 3760-3770.
    OpenUrlAbstract
  21. ↵
    Kotlyarov A, Neininger A, Schubert C, Eckert R, Birchmeier C, Volk HD, Gaestel M (1999) MAPKAP kinase 2 is essential for LPS-induced TNF-α biosynthesis. Nat Cell Biol 1: 94-97.
    OpenUrlCrossRefPubMed
  22. ↵
    Lee YB, Schrader JW, Kim SU (2000) p38 map kinase regulates TNF-α production in human astrocytes and microglia by multiple mechanisms. Cytokine 12: 874-880.
    OpenUrlCrossRefPubMed
  23. ↵
    MacKenzie A, Wilson HL, Kiss-Toth E, Dower SK, North RA, Surprenant A (2001) Rapid secretion of interleukin-1β by microvesicle shedding. Immunity 15: 825-835.
    OpenUrlCrossRefPubMed
  24. ↵
    Mahtani KR, Brook M, Dean JL, Sully G, Saklatvala J, Clark AR (2001) Mitogen-activated protein kinase p38 controls the expression and posttranslational modification of tristetraprolin, a regulator of tumor necrosis factor α mRNA stability. Mol Cell Biol 21: 6461-6469.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Mehta VB, Hart J, Wewers MD (2001) ATP-stimulated release of interleukin (IL)-1β and IL-18 requires priming by lipopolysaccharide and is independent of caspase-1 cleavage. J Biol Chem 276: 3820-3826.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    Meistrell III ME, Botchkina GI, Wang H, Di Santo E, Cockroft KM, Bloom O, Vishnubhakat JM, Ghezzi P, Tracey KJ (1997) Tumor necrosis factor is a brain damaging cytokine in cerebral ischemia. Shock 8: 341-348.
    OpenUrlPubMed
  27. ↵
    Nakajima K, Hamanoue M, Shimoto M, Takei N, Kosaka S (1989) Characterization of microglia isolated from a primary culture of embryonic rat brain by a simplified method. Biomed Res 10: 411-423.
    OpenUrl
  28. ↵
    Nakajima K, Shimojo M, Hamanoue M, Ishiura S, Sugita H, Kohsaka S (1992) Identification of elastase as a secretory protease from cultured rat microglia. J Neurochem 58: 1401-1408.
    OpenUrlCrossRefPubMed
  29. ↵
    Neary JT, Kang Y, Bu Y, Yu E, Akong K, Peters CM (1999) Mitogenic signaling by ATP/P2Y purinergic receptors in astrocytes: involvement of a calcium-independent protein kinase C, extracellular signal-regulated protein kinase pathway distinct from the phosphatidylinositol-specific phospholipase C/calcium pathway. J Neurosci 19: 4211-4220.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    Neary JT, Kang Y, Willoughby KA, Ellis EF (2003) Activation of extracellular signal-regulated kinase by stretch-induced injury in astrocytes involves extracellular ATP and P2 purinergic receptors. J Neurosci 23: 2348-2356.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    Norenberg W, Langosch JM, Gebucke-Haerter PJ, Illes P (1994) Characterization and possible function of adenosine 5′-triphosphate receptors in activated rat microglia. Br J Pharmacol 111: 942-950.
    OpenUrlCrossRefPubMed
  32. ↵
    North RA, Surprenant A (2000) Pharmacology of cloned P2X receptors. Annu Rev Pharmacol Toxicol 40: 563-580.
    OpenUrlCrossRefPubMed
  33. ↵
    Panenka W, Jijon H, Herx LM, Armstrong JN, Feighan D, Wei T, Yong VW, Ransohoff RM, MacVicar BA (2001) P2X7-like receptor activation in astrocytes increases chemokine monocyte chemoattractant protein-1 expression via mitogen-activated protein kinase. J Neurosci 15: 7135-7142.
  34. ↵
    Sak K, Webb TE (2002) A retrospective of recombinant P2Y receptor subtypes and their pharmacology. Arch Biochem Biophys 397: 131-136.
    OpenUrlCrossRefPubMed
  35. ↵
    Sanz JM, Di Virgilio F (2000) Kinetics and mechanism of ATP-dependent IL-1β release from microglial cells. J Immunol 164: 4893-4898.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Scherbel U, Raghupathi R, Nakamura M, Saatman KE, Trojanowski JQ, Neugebauer E, Marino MW, McIntosh TK (1999) Differential acute and chronic responses of tumor necrosis factor-deficient mice to experimental brain injury. Proc Natl Acad Sci USA 96: 8721-8726.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    Shigemoto-Mogami Y, Koizumi S, Tsuda M, Ohsawa K, Kohsaka S, Inoue K (2001) Mechanisms underlying extracellular ATP-evoked interleukin-6 release in mouse microglial cell line, MG-5. J Neurochem 78: 1339-1349.
    OpenUrlCrossRefPubMed
  38. ↵
    Yang L, Lindholm K, Konishi Y, Li R, Shen Y (2002) Target depletion of distinct tumor necrosis factor receptor subtypes reveals hippocampal neuron death and survival through different signal transduction pathways. J Neurosci 22: 3025-3032.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Production and Release of Neuroprotective Tumor Necrosis Factor by P2X7 Receptor-Activated Microglia
Tomohisa Suzuki, Izumi Hide, Katsutoshi Ido, Shinichi Kohsaka, Kazuhide Inoue, Yoshihiro Nakata
Journal of Neuroscience 7 January 2004, 24 (1) 1-7; DOI: 10.1523/JNEUROSCI.3792-03.2004
del.icio.us logo Digg logo Reddit logo Twitter logo Facebook logo Google logo Mendeley logo

Jump to section

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.