|Home | About | Journals | Submit | Contact Us | Français|
The group I metabotropic glutamate receptor 5 (mGluR5) can modulate addiction, pain and neuronal cell death. Expression of some mGluRs, such as group II and III mGluRs, has been reported in microglia and may affect their activation. However, the expression and role of mGluR5 in microglia is unclear. Using immunocytochemistry and Western blot, we demonstrate that mGluR5 protein is expressed in primary microglial cultures. Activation of mGluR5 using the selective agonist (RS)-2-chloro-5-hydroxyphenylglycine (CHPG) significantly reduces microglial activation in response to lipopolysaccharide (LPS), as indicated by a reduction in nitric oxide, reactive oxygen species and TNFα production. Microglial-induced neurotoxicity is also markedly reduced by CHPG treatment. The anti-inflammatory effects of CHPG are not observed in microglial cultures from mGluR5 knockout mice and are blocked by selective mGluR5 antagonists, suggesting these actions are mediated by the mGluR5 receptor. Anti-inflammatory actions of mGluR5 activation are attenuated by PLC and PKC inhibitors, as well as by calcium chelators, suggesting that mGluR5 activation in microglia involves the Gαq-protein signal transduction pathway. These data indicate that microglial mGluR5 may represent a novel target for modulating neuroinflammation, an important component of both acute and chronic neurodegenerative disorders.
Metabotropic glutamate receptors (mGluRs) are G-protein coupled receptors that modulate glutamatergic activity. Within the central nervous system, these receptors have been studied largely in neurons, although they are also found in glia. Classified into three groups based upon their signal transduction pathways and pharmacological profiles, activation of several mGluR subtypes affect neuronal viability both in vitro and after central nervous system injury. Within group I receptors, mGluR1 activation exacerbates neuronal death, particularly necrotic cell death (Allen et al. 2000). In contrast, activation of mGluR5 can inhibit caspase-dependent neuronal apoptosis in cell culture models (Movsesyan et al. 2004). Although mGluR5 antagonists have also been reported to be neuroprotective, such effects appear to be unrelated to actions at the mGluR5 receptor (Lea et al. 2005).
In neurons and astrocytes, mGluR5 receptors are coupled to Gαq-proteins and activation induces phospholipase C (PLC), leading to an increase in inositol triphosphate (IP3) and intracellular calcium, as well as activation of protein kinase C (PKC). Alternatively, mGluR5 is coupled to a Gαs-protein in T-lympocytes; mGluR5 activation in these cells induces cAMP and reduces lymphocyte proliferation (Pacheco et al. 2004).
mGluRs have been identified on microglia, but their roles in neuroinflammation have received limited attention. Taylor and colleagues have reported modulatory effects by group II and III receptors in opposite directions in vitro (Taylor et al. 2002; Taylor et al. 2003), but such data are hard to reconcile with in vivo studies showing virtually identical neuroprotective profiles after activation of these receptors (Allen et al. 1999; Bond et al. 1999; Faden et al. 1997; Folbergrova et al. 2005; Movsesyan and Faden 2006; Vernon et al. 2007). Although the presence of mGluR5 mRNA has been observed in microglia (Biber et al. 1999), neither protein expression nor its role in microglial activation and neuroinflammation has been evaluated.
We show that mGluR5, but not mGluR1, is substantially expressed in microglia and negatively regulates the release of microglial associated inflammatory factors and related neurotoxicity. These effects are mediated in part by G-protein signal transduction mechanisms, including activation of the PLC - PKC signal transduction system.
Microglial cells were obtained from postnatal day 2 Sprague Dawley rat pups, day 2 mGluR5 knockout and wild-type mouse pups, or day 2 C57Bl6 mouse pups and cultured as described (Byrnes et al. 2006). Briefly, the whole brain was carefully dissected and homogenized in L15 media (Gibco, Carlsbad, CA). Mixed glial cultures were incubated for 8 – 10 days at 37°C with 5% CO2 in Dulbecco’s Modified Eagle Media (Gibco) with 10% Fetal Calf Serum (Hyclone, Logan, UT), 1% L-glutamine (Gibco), 1% Sodium Pyruvate (Gibco), and 1% Pen/Strep (Fisher, Pittsburgh, PA). After the initial incubation, the cells were shaken for 1 hour at 100 rpm and 37°C. Detached microglia were collected and replated as purified cultures with greater than 96% purity.
CHPG ((RS)-2-chloro-5-hydroxyphenylglycine; Tocris, Ellisville, MO) was applied to cells for 1 hr prior to lipopolysaccharide (LPS; 100ng/ml) stimulation. The mGluR5 antagonist MTEP (3-[(2-methyl-1,3-thiazol-4-yl)ethynyl]pyridine; 100μM) was a gift from Merck Research Laboratories (Rahway, NJ), and was administered 30 min prior to CHPG administration. Forskolin (50μM, Tocris) and dibutyryl cAMP (200μM, Biomol, Plymouth Meeting, PA) were administered 1 hr prior to LPS stimulation. SQ22356, RpcAMPs (Calbiochem, San Diego, CA), R0318220, U-73122 (10μM; BioMol, Plymouth Meeting, PA), and BAPTA-AM (1.5μM; Molecular Probes, Carlsbad, CA) were administered 30 minutes prior to CHPG administration. All chemicals were prepared and stored according to the manufacturer’s guidelines.
At 24 hours post-purification, microglia in 24-well plates on glass cover-slips were stimulated with LPS (100 ng/ml) or were untreated (control) for an additional 24 hours. Cells were then fixed in 4% paraformaldehyde and subjected to standard immunohistochemistry, performed as described (Byrnes et al. 2006) using antibodies against mGluR1α and mGluR5 (Chemicon, Billerica, MA), OX42 (Serotec, Raleigh, NC), ED1 (Serotec), and Galectin-3 (Abcam, Cambridge, MA). Confocal fluorescence microscopy imaging was performed using Zeiss LSM 510 Meta confocal laser scanning microscope.
At various time points after application of mGluR agonists/antagonists, proliferation of microglia in 96-well plates was assessed using the MTS assay (MTS tetrazolium compound; Cell Titer 96® Aqueous One Solution, Promega, Madison, WI); cell death was assessed with the lactate dehydrogenase (LDH) release assay (CytoTox 96® nonradioactive cytotoxicity assay kit, Promega, Madison, WI). Both were performed according to the manufacturer’s protocols.
Nitric oxide (NO) production was assayed using the Griess Reagent Assay (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions.
TNFα protein secreted into the media was assessed using the TNFα ELISA kit according to the manufacturer’s protocol (Endogen, Rockford, IL).
Intracellular reactive oxygen species (ROS) production was assessed at 24 hours after stimulation by measuring the oxidation of 5 (and 6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate-acetyl ester (CM-H2DCFDA; Molecular Probes, Eugene, OR). Media from microglia plated into 96-well plates was aspirated and replaced with warmed 1X PBS. CM-H2DCFDA (10 μM) was added to microglia and incubated for 45 minutes. Fluorescence was measured using excitation and emission wavelengths of 490 and 535 nm, respectively.
Rat primary cortical neuronal cultures were derived from E18 rat cortices (Taconic, Germantown, NY) as previously described (Mukhin et al. 1998). At 24 hours after stimulation, microglia in transwell inserts were washed in media and inserted into 24-well plates containing neurons (neurons were at day 5 in vitro). Twenty-four hours later, microglia were removed and the LDH release assay was used to assess neuronal cell death.
Primary microglia were cultured in 96-well plates and incubated overnight with 0.625 μCi/well myo-[3H]inositol (NEN, Boston, MA) to label the cell membrane phosphoinositides and perform the assay, as described previously (Surin et al. 2007). Briefly, cells were washed in Locke’s buffer and incubated with or without 100μM CHPG for 1 hr at 37 °C in Locke’s buffer containing 20 mM LiCl to block inositol phosphate degradation. Inositol phosphates were then extracted with 0.1 M HCl for 10 min. The separation of [3H]inositol phosphates was performed by ion-exchange chromatography on AG 1-X8 resin (200–400 mesh; Bio-Rad Laboratories, Hercules, CA). Total [3H]inositol phosphates were eluted from the columns with 0.5 ml of 0.1 M formic acid/1 M ammonium formate. The collected samples were mixed with Safety-Solve cocktail (RPI, Mount Prospect, IL) and measured by scintillation counting.
Quantitative data are presented as mean +/- SEM. Data were analyzed using Student’s t-test or one-way ANOVA, where appropriate. All statistical tests were performed using the GraphPad Prism Program, Version 3.02 for Windows (GraphPad Software, Inc. San Diego, CA). A p value < 0.05 was considered statistically significant.
Western blot analysis demonstrated that mGluR5 protein is highly expressed in rat whole brain microglial cultures whereas expression of the other group I mGluR, mGluR1, is barely detectable in these cells (Fig. 1A). Rat cortical neuron (RCN) samples, which constitutively express all mGluR receptors, were run alongside as positive controls for the antibodies.
Immunocytochemistry was also used to confirm mGluR5 expression in microglial cultures (Fig. 1B). Immunolabeling for mGluR1α revealed no receptor expression on microglia (Fig. 1B). Strong mGluR5 immunolabeling was detected on the surface of primary microglia. mGluR5 labeling is found particularly on the cell periphery, consistent with membrane localization. Further, to confirm microglial expression, double-labeling was performed with common markers of microglia and mGluR5. mGluR5 was expressed on cells that were also positive for OX42 and ED1 (Fig. 1C). Primary antibodies were excluded from the incubation mix for immunocytochemistry to serve as negative controls and showed no labelling (data not shown).
In order to determine if microglia respond to mGluR5 stimulation, microglia were cultured in 96-well plates and mGluR5 agonists/antagonists were added alone or in combination to determine their effect on microglial activation. Pre-treatment with CHPG was found to reduce expression of several independent markers of microglial activation, including ROS production, NO production, proliferation and TNFα production.
CHPG’s effects were determined to be dose dependent, with maximal effects at 100μM, as measured by both ROS and NO production (p<0.05; Fig. 2A, B). CHPG addition without LPS stimulation had no effect on microglial ROS or NO production (Fig. 2A, B).
Addition of LPS (100ng/ml) to microglia resulted in an increase in proliferation over 24 hours, which was blocked by 1 hour pre-treatment with CHPG (100μM; p<0.05) (Fig. 3A). NO production was also significantly increased by 15 and 24 hours after LPS stimulation (Fig. 3B). Pre-treatment with CHPG significantly attenuated the LPS induced increase in NO production by 24 hours (p<0.05; Fig. 3B). TNFα production was up-regulated by LPS stimulation within 30 minutes and remained at increased levels through 24 hours (p<0.001; Fig. 3C). CHPG pre-treatment significantly attenuated TNFα production from 30 minutes through 8 hours, delaying the increase in TNFα observed after LPS stimulation.
Galectin-3, a marker of microglial activation (Byrnes et al. 2006), was also expressed when microglia were stimulated with LPS (Fig. 4). Pre-treatment with CHPG resulted in a loss of Galectin-3 staining, similar to control cultures (Fig. 4).
Addition of activated microglia to neuronal cultures is known to induce neuronal cell death. In order to determine if mGluR5 plays a role in microglial-induced neurotoxicity, agonists/antagonists were added to microglia prior to co-culture with neurons. Stimulation of microglia with LPS prior to co-culture significantly decreased neuron viability, as measured by neuronal LDH release (p<0.05; Fig. 5).
Pre-treatment of microglia with CHPG prior to LPS addition resulted in a significant increase in neuronal viability (p<0.05; Fig. 5). In order to eliminate the possibility of a direct effect of CHPG on neurons in this study, microglia were washed prior to addition to neurons. Neither LPS nor CHPG had any direct effect on neurons when added without microglia (data not shown). Addition of the selective mGluR5 antagonist MTEP (100 μM) prior to CHPG pre-treatment reversed the protective effect of CHPG, reducing neuronal viability (p<0.05; Fig. 5), suggesting that CHPG acts through the mGluR5 receptor.
To confirm that CHPG is acting through the mGluR5 receptor, microglia from mGluR5 knockout (-/-) and wild-type (+/+) mice were obtained and stimulated with LPS. LPS stimulation increased microglial proliferation, NO production and ROS production in both knockout and wild-type cells (Fig. 6). However, one hour pre-treatment of microglia with CHPG attenuated the effects of LPS in wild-type microglia (Fig. 6A, C, E) but not those from mGluR5 knockouts (Fig. 6B, D, F), thus demonstrating that CHPG acts through mGluR5 receptor activation.
In T-lymphocytes, activation of mGluR5 receptors induces activation of adenylyl cyclase, cAMP production and PKA induction. To determine if this pathway is utilized in microglia, cAMP was measured in microglia following CHPG application. CHPG significantly increased cAMP production by 24 hours (Fig. 7A). To further investigate the potential role of cAMP in microglial activation, cAMP inducers, forskolin (50μM) and dibutyryl cAMP (dbcAMP; 200μM), were applied to microglia prior to LPS stimulation; each caused suppression of microglial activation as measured by NO production, similar to that of CHPG (p<0.05; Fig. 7B).
In order to determine if this pathway was required for CHPG’s activities in microglia, cells were pre-treated with the adenylyl cyclase inhibitor SQ22356 (0.1 – 500 μM) or the PKA inhibitor RpcAMPs (10 – 500 μM) prior to addition of CHPG. Pre-treatment with either inhibitor failed to induce any significant reduction in CHPG’s effects on microglial activation, as measured by ROS production (Fig. 7C, D), suggesting that cAMP production is not necessary for CHPG’s activity in microglial cells.
In neurons and astrocytes, mGluR5 stimulation results in Gαq activation, triggering PLC phosphorylation, hydrolysis of phosphatidyl inositol (PI), PKC activation and calcium release. In order to determine if these signal transduction pathways were involved in CHPG signalling in microglia, hydrolysis of PI was assessed. CHPG treatment resulted in a significant increase PI hydrolysis (p<0.01; Fig. 8A), suggesting the activation of Gαq. To further assess the role of this pathway, cells were pre-treated with inhibitors of PLC (U-73122) and PKC (R0318220) prior to CHPG and LPS addition. Using microglial activation outcomes such as ROS production, NO production and proliferation, U-73122 and R0318220 were found to block the effects of CHPG, returning NO, ROS and proliferation to LPS-induced levels (Fig. 8B, C, D). The calcium chelator (BAPTA-AM) also reversed the effects of CHPG (Fig. 8C). These inhibitors did not affect the activity of LPS or of microglia alone when introduced without CHPG pre-treatment (data not shown), demonstrating that their activities were specific for the CHPG signal transduction pathway.
Previous studies have reported the presence of mGluRs on immune cells such as microglia, macrophages and lymphocytes (Geurts et al. 2003; Pacheco et al. 2004; Rezzani et al. 2003; Taylor et al. 2005). Biber et al (Biber et al. 1999) demonstrated mGluR5 mRNA in microglia, but did not examine the protein or functional implications. The current study shows that in a purified rat microglia culture system, microglia express mGluR5 constitutively (Fig. 1), similar to T lymphocytes (Pacheco et al. 2004), while the other group I mGluR, mGluR1, is barely expressed.
Although mGluR1 and mGluR5 receptors share certain common signal transduction mechanisms, they have remarkably different profiles in CNS injury. Agonists of mGluR1 exacerbate necrotic cell death (Allen et al. 2000), whereas selective antagonists are neuroprotective in vitro and in vivo (Faden et al. 2001; Fei et al. 2006; Szydlowska et al. 2007). In contrast, activation of mGluR5 inhibits caspase dependent neuronal apoptosis in cell culture models (Movsesyan et al. 2004). Although mGluR5 antagonists have been reported to be neuroprotective, it has been convincingly established using mGluR5 knockout technology that such effects are unrelated to actions at the mGluR5 receptor and instead are mediated by direct actions on the NMDA receptor (Lea et al. 2005).
In the current work, we show that mGluR5 stimulation inhibits microglial activation in vitro. Markers of microglial activation, including proliferation, NO, TNFα and ROS production and Galectin-3 expression, as well as microglial-induced neurotoxicity were significantly attenuated by pre-treatment with the mGluR5 agonist CHPG. To verify that these effects were mediated by the mGluR5 receptor, rather than a non-receptor related mechanism, we demonstrated that CHPG’s effects were lost in microglia from mGluR5 knockout mice and markedly attenuated by a selective mGluR5 receptor antagonist.
NO, TNFα and ROS each have been shown to contribute to neuronal damage and death in vitro and in vivo (He et al. 2003; Iravani et al. 2002; Jeohn et al. 2000; Juurlink and Paterson 1998; Lewen et al. 2000; Munch et al. 2003; Taylor et al. 2005; Zou and Crews 2005). For example, ROS induces lipid peroxidation, which results in cell membrane damage and cell death, as well as glutamate release to contribute to further excitotoxic cell death (Barger et al. 2007). NO is also associated with neuronal death in vitro and in vivo, through the production of the toxic metabolite peroxynitrite or via direct action on lipid membranes of the cell or mitochondria (Gibbons and Dragunow 2006; Mander and Brown 2005). TNFα can cause cell death directly by binding to neuronal TNF receptors linked to death domains that activate caspase-dependent apoptosis (Zhao et al. 2001). This cytokine can also induce glutamate release and enhance excitotoxicity (Zou and Crews 2005). In addition, TNFα can induce additional release of ROS, by inducing NADPH oxidase activity (Li et al. 2005). In our hands, LPS-stimulated microglia induced neuronal cell death, which was attenuated by CHPG treatment; it is unclear which of these factors, if not all, are responsible for the observed neurotoxicity. Because early suppression of TNFα is protective, whereas knockout of the receptor exacerbates damage after central nervous system injury, it has been suggested that this cytokine contributes to early pathophysiological changes but may facilitate subsequent plasticity (Bethea et al. 1999; Cheng et al. 1994; Kim et al. 2001; Knoblach and Faden 1998; Knoblach et al. 1999; Nakajima and Kohsaka 2004). In this regard, it should be noted that CHPG treatment of microglial cells markedly suppressed early TNFα release yet had modest to no effect on later cytokine levels. Finally, Galectin-3, a marker of microglial activation, was also up-regulated after LPS stimulation but reduced by CHPG treatment. Galectin-3 is a carbohydrate binding protein that is up-regulated following injury and plays a role in phagocytosis and perpetuation of the inflammatory response (Byrnes et al. 2006).
In lymphocytes, mGluR5 is coupled to a Gαs-protein, with activation resulting in cAMP production (Pacheco et al. 2004). Our data indicate that treatment of microglia with CHPG induces cAMP production, and that other cAMP inducers, forskolin and dbcAMP, have similar effects on microglial activity. However, blocking the production of cAMP does not mitigate the effects of CHPG, suggesting that cAMP is not a critical component in the CHPG signal transduction pathway in microglia.
In neurons and astrocytes, mGluR5-mediated signal transduction involves activation of PLC and PKC, increased inositol triphosphate (IP3), and the release of intracellular calcium. Inhibitors of PLC (U-73122) and PKC (R0318220), as well as calcium chelation (BAPTA-AM), blocked CHPG’s effects, suggesting that CHPG is operating through the PLC/PKC/calcium pathway. In the PLC/PKC pathway, activation of PLC leads to DAG synthesis and IP3 mediated release of Ca2+ (Pin and Duvoisin 1995). Both DAG and Ca2+ are capable of activating PKC, which has a multitude of actions within the cell, including activation of transcription factors and alteration of potassium channel activity (Endoh 2004; Mannaioni et al. 2001), which may alter microglial responses. Precisely how PKC affects microglial activation in response to CHPG remains to be elucidated. It should be noted that although Ro318220 is a PKC inhibitor, it also inhibits the activity of mitogen-activated protein kinase phosphatase-1 (MKP-1) (Guo et al. 1998). The latter enzyme is induced by TNFα, and operates to inactivate mitogen-activated protein kinases and JNK (Keyse 1995). As these actions can serve to modulate inflammation, they may reflect another potential mechanism of CHPG action within microglia.
Previous studies have demonstrated that PKC is also involved in the regulation of COX2 in activated microglia (Akundi et al. 2005), and that inhibitors of PKCα/βII reduce microglial activation (Nikodemova et al. 2007). In our study, the pan-PKC inhibitor Ro318220 at the concentrations used (1.5μM) did not have an effect on microglial production of NO or on microglial proliferation, but did reverse the effect of CHPG, suggesting that different PKC isoforms may be involved in these two systems. Work has suggested that mGluR5 involves phosphorylation of the PKCε isoform (Lee and Ro 2007; Olive et al. 2005). However, the PKC isoforms that are involved in microglial activation or in mGluR5 signalling have not been fully elucidated.
CHPG provides protection in a rat model of focal brain ischemia (Bao et al. 2001). However, the authors did not evaluate potential mechanism of the neuroprotection, nor examine effects on microglia or inflammation. Our present findings suggest that neuroprotective effects of CHPG may reflect, in part, an mGluR5 receptor mediated reduction in the post-injury inflammatory response. As such, the present observations suggest a novel strategy for reducing neuroinflammation, which may have therapeutic potential in a number of neurological disorders in addition to stroke. Preliminary observations from our lab suggest that CHPG can limit microglial activation after traumatic brain or spinal cord injury in rodents. Moreover, as mGluR5 stimulation also exerts anti-apoptotic effects on neurons (Movsesyan et al. 2004), such treatment may have multi-potential neuroprotective actions. As both neuroinflammation and caspase-dependent neuronal apoptosis have been implicated in many acute and chronic neurodegenerative disorders (Block and Hong 2005; Tansey et al. 2007; Yakovlev and Faden 2001), mGluR5 agonist therapy may have broad therapeutic relevance.
The authors would like to thank Dr. Ahdeah Pajoohesh-Ganji for her technical assistance and Dr. Jarda Wroblewski for instruction and use of the PI hydrolysis assay. The work was supported by NIH NINDS R01 grant 5R01NS037313-08.