Regulation of N-Methyl-d-aspartic Acid (NMDA) Receptors by Metabotropic Glutamate Receptor 7*♦

Background: The metabotropic glutamate receptors (mGluRs) are potential novel targets for mental disorders. Results: Activation of mGluR7 significantly reduced NMDAR-mediated currents and NMDAR surface expression via an actin-dependent mechanism. Conclusion: mGluR7, by affecting the cofilin/actin signaling, regulates NMDAR trafficking and function. Significance: It provides a potential mechanism for understanding the role of mGluR7 in mental health and disorders. Emerging evidence suggests that metabotropic glutamate receptors (mGluRs) are potential novel targets for brain disorders associated with the dysfunction of prefrontal cortex (PFC), a region critical for cognitive and emotional processes. Because N-methyl-d-aspartic acid receptor (NMDAR) dysregulation has been strongly associated with the pathophysiology of mental illnesses, we examined the possibility that mGluRs might be involved in modulating PFC functions by targeting postsynaptic NMDARs. We found that application of prototypical group III mGluR agonists significantly reduced NMDAR-mediated synaptic and ionic currents in PFC pyramidal neurons, which was mediated by mGluR7 localized at postsynaptic neurons and involved the β-arrestin/ERK signaling pathway. The mGluR7 modulation of NMDAR currents was prevented by agents perturbing actin dynamics and by the inhibitor of cofilin, a major actin-depolymerizing factor. Consistently, biochemical and immunocytochemical results demonstrated that mGluR7 activation increased cofilin activity and F-actin depolymerization via an ERK-dependent mechanism. Furthermore, mGluR7 reduced the association of NMDARs with the scaffolding protein PSD-95 and the surface level of NMDARs in an actin-dependent manner. These data suggest that mGluR7, by affecting the cofilin/actin signaling, regulates NMDAR trafficking and function. Because ablation of mGluR7 leads to a variety of behavioral symptoms related to PFC dysfunction, such as impaired working memory and reduced anxiety and depression, our results provide a potential mechanism for understanding the role of mGluR7 in mental health and disorders.

Previous studies have shown that mGluR7 suppresses voltage-gated Ca 2ϩ channels and inhibits synaptic transmission via a presynaptic mechanism depending on PKC and the scaffolding protein PICK1 (16,17). In the hippocampal mossy fiber pathway, mGluR7 acts as a switch to control the bidirectional plasticity of forward inhibition via compartmentalized Ca 2ϩ channel regulation (18,19). In neocortical regions, mGluR7 mRNA is intensely expressed, whereas the mRNA for other members of group III mGluRs is barely detectable (20,21). Electron microscopy studies have revealed that mGluR7, in addition to the presynaptic presence, is also postsynaptically distributed in various regions (22)(23)(24). Because of the strong effect of mGluR7 on presynaptic glutamate release, the function of mGluR7 at postsynaptic neurons has not been demonstrated in brain slice recordings.
A potential target of postsynaptic mGluR7 that is critical for cognition and emotion is the NMDA glutamate receptor because NMDAR dysregulation has been strongly associated with the pathophysiology of schizophrenia and other mental illnesses (25,26). Because ablation of mGluR7 leads to a variety of behavioral symptoms related to the abnormality of PFC, a brain region critical for cognitive and emotional processes (27)(28)(29), we examined the impact of mGluR7 on NMDA receptors in PFC pyramidal neurons. We demonstrate that mGluR7 inhibits NMDAR-mediated currents through a mechanism involving actin-based trafficking of NMDA receptors. It provides a potential molecular and cellular mechanism underlying the involvement of mGluR7 in regulating PFC functions.

EXPERIMENTAL PROCEDURES
Electrophysiological Recordings in Slices-All experiments were carried out with the approval of the State University of New York at Buffalo Animal Care Committee. The voltage clamp recording technique was used to measure NMDAR-EPSC or AMPAR-EPSC in PFC pyramidal neurons as described previously (30,31). Briefly, PFC brain slices (300 m) from juvenile (3-5 weeks postnatal) Sprague-Dawley rat were used. Cells were visualized with a 40ϫ water immersion lens and illuminated with near infrared (IR) light, and the image was detected with an IR-sensitive charge-coupled device (CCD) camera. Tight seals (2-10 gigaohms) from visualized pyramidal neurons were obtained by the "blow and seal" technique. A MultiClamp 700A amplifier was used for these recordings. The access resistances ranged from 13 to 18 megaohms and were compensated 50 -70%. A bipolar stimulating electrode (FHC, Inc., Bowdoinham, ME) was positioned ϳ100 m from the neuron under recording. Evoked currents were generated with 50-s pulses (delivered every 30 s) from a stimulation isolation unit controlled by an S48 pulse generator (Astro-Med, Inc., West Warwick, RI). For NMDAR-EPSC recording, stimuli were given 3 s after neurons (clamped at Ϫ70 mV) were depolarized to ϩ60 mV to fully relieve the voltage-dependent Mg 2ϩ block of NMDAR channels.
Whole-cell Recordings in Dissociated or Cultured Neurons-Acutely dissociated PFC pyramidal neurons from 4-week-old rats or PFC primary cultures from 17-18-day rat embryos were prepared using procedures described previously (30). Recordings of whole-cell NMDA-elicited currents employed similar techniques as described previously (30,32). Recordings were obtained with an Axon Instruments 200B patch clamp amplifier that was controlled and monitored with an IBM PC running pCLAMP with a DigiData 1320 series interface. Electrode resistances were typically 2-4 megaohms in the bath. After seal rupture, series resistance (4 -10 megaohms) was compensated (70 -90%) and periodically monitored. The cell membrane potential was held at Ϫ60 mV. NMDA (100 M) was applied for 2 s every 30 s to minimize desensitization-induced decrease of current amplitude.
Data analyses were performed with Clampfit (Axon instruments) and KaleidaGraph (Albeck Software, Reading, PA). ANOVA tests were performed to compare the differential degrees of current modulation between groups subjected to different treatments.
Co-immunoprecipitation-After treatment, PFC slices were homogenized in 1 ml of radioimmune precipitation lysis buffer. The lysates were ultracentrifuged at 100,000 ϫ g at 4°C for 30 min. The supernatant fractions were collected and incubated with anti-NR1 (Millipore, 2 g) at 4°C overnight followed by incubation with protein A/G plus agarose (Santa Cruz Biotechnology) for 2 h. Immunoprecipitates were then subjected to 7.5% SDS-PAGE, transferred to nitrocellulose membranes, and blotted with anti-actin (Santa Cruz Biotechnology, 1:2000) or anti-PSD95 (Affinity BioReagents, 1:2000).
Immunocytochemical Staining-Neurons grown on coverslips were fixed in 4% paraformaldehyde in PBS for 20 min at room temperature and then washed three times with PBS. Neurons were then permeabilized with 0.1% Triton X-100 in PBS for 5 min followed by a 1-h incubation with 5% bovine serum albumin (BSA) to block nonspecific staining. For the staining of surface NR1, neurons were fixed but not permeabilized. Neurons were incubated with the primary antibody (anti-MAP2, 1:500, Chemicon; or anti-NR1 extracellular domain, 1:100, Chemicon) at 4°C overnight. After washing, neurons were incubated with Alexa Fluor 488 (green)-conjugated secondary antibody (Molecular Probes, 1:500) for 2 h at room temperature. For the staining of F-actin, neurons were incubated with Alexa Fluor 568 (red)-conjugated phalloidin (1 unit/ml, Invitrogen) at room temperature for 15 min. After washing in PBS for three times, coverslips were mounted on slides with VECTASHIELD mounting media (Vector Laboratories, Burlingame, CA).
Fluorescent images were obtained using a 100ϫ objective with a cooled charge-coupled device camera mounted on a Nikon microscope. All specimens were imaged under identical conditions and analyzed using identical parameters with the ImageJ software. The surface NR1 clusters or F-actin clusters were measured. To define dendritic clusters, a single threshold was chosen manually so that clusters corresponded to puncta of at least 2-fold greater intensity than the diffuse fluorescence on the dendritic shaft. On each coverslip, the cluster density of 4 -6 neurons (3-4 dendritic segments of 50 m length per neuron) was measured. Three to four independent experiments for each of the treatments were performed. Quantitative anal-mGluR7 Regulation of NMDA Receptors yses were conducted blindly (without knowledge of experimental treatment).
Biochemical Measurement of Surface NMDA Receptors-The surface NMDA receptors were detected as described previously (31,33). Briefly, after treatment, cortical cultures were incubated with artificial cerebrospinal fluid containing 1 mg/ml Sulfo-NHS-LC-Biotin (Pierce Chemical Co.) for 20 min on ice. The cultures were then rinsed three times in TBS to quench the biotin reaction followed by homogenization in 0.5 ml of modified radioimmune precipitation buffer. Lysates were centrifuged at 16,000 ϫ g at 4°C. Supernatant was collected and incubated with NeutrAvidin-agarose (Pierce) for 2 h at 4°C. Bound proteins were resuspended in SDS sample buffer and boiled. Quantitative Western blots were performed on both total and biotinylated (surface) proteins using anti-NR1 (1:1000; Upstate Biotech Millipore).

Activation of mGluR7 Significantly Decreases NMDAR Currents in PFC Pyramidal
Neurons-To examine the potential impact of mGluRs on excitatory synaptic transmission in PFC pyramidal neurons, we examined the effect of L-AP4, a prototypical agonist for group III mGluRs (34,35), on NMDAR-EPSC and AMPAR-EPSC in PFC slices. As shown in Fig. 1A, bath application of L-AP4 (200 M) caused a strong reversible reduction of the amplitude of NMDAR-EPSC (43.1 Ϯ 6.8%, n ϭ 18). A marked reduction of AMPAR-EPSC by L-AP4 (Fig. 1B, 58.3 Ϯ 9.3%, n ϭ 14) was also observed. To distinguish the presynaptic versus postsynaptic mechanism underlying these effects of L-AP4, we measured the paired pulse ratio of NMDAR-EPSC and AMPAR-EPSC (evoked by paired pulses with a 100-or 40-ms interval, respectively), a readout sensitive to presynaptic glutamate release. As shown in Fig. 1C, L-AP4 did not alter the paired pulse ratio of NMDAR-EPSC (4.9 Ϯ 0.6%, n ϭ 8), but significantly increased the paired pulse ratio of AMPAR-EPSC (37.2 Ϯ 4.2%, n ϭ 8). This suggests that L-AP4 regulates NMDAR-EPSC mainly via a postsynaptic mechanism, whereas regulating AMPAR-EPSC through a presynaptic action.
To verify the postsynaptic effect of L-AP4 on NMDARs, we measured the ionic current elicited by exogenous NMDA application (100 M) in acutely dissociated PFC pyramidal neurons. As shown in Fig. 1D, bath application of L-AP4 (200 M) reversibly reduced NMDAR currents (24.3 Ϯ 3.6%, n ϭ 12). In contrast, L-AP4 application did not have any significant effect on AMPA (100 M)-elicited currents (2.6 Ϯ 1.3%, n ϭ 10). To further characterize the effect of L-AP4 on NMDAR currents, we carried out the dose-response experiments. As shown in NMDAR functional properties are determined by subunit composition. The primary NMDARs in mature cortical synapses, which are composed of NR1/NR2A or NR1/NR2B, differ in deactivation kinetics and subcellular distribution. To determine which subpopulation(s) of NMDARs is targeted by L-AP4, we applied the selective inhibitor of NR2B subunit, ifenprodil.
To confirm that the effect of L-AP4 on NMDAR currents was mediated through group III mGluRs, we applied selective receptor agonists and antagonists. As shown in Fig. 2A Using cell lines that express specific mGluR subtypes, it has been determined that neither L-AP4 nor L-SOP possesses activity at group I or II mGluRs at concentrations below 1000 M, whereas L-AP4 has EC 50 of 160 -500 M to mGluR7 and 0.4 -1.2 M to other members in group III (mGluR4,6,8) (34,35). Similarly, L-SOP has EC 50 of ϳ200 M to mGluR7 and 2-5 M to other members in group III (34,35). Because of the pharmacological profile of mGluRs, previous physiological studies on the role of group III mGluRs, especially mGluR7, in synaptic transmission have commonly used high doses (500 -1000 M) of L-AP4 (19, 36 -38). Our dose-response studies suggest that mGluR7 is the receptor mediating the effect of L-AP4 on NMDAR currents. To confirm this, we carried out a cellular knockdown experiment with mGluR7 siRNA transfection in PFC cultures. As shown in Fig. 2C, the reducing effect of L-AP4 on NMDAR currents was abolished in mGluR7 siRNA-transfected neurons (control: 22.3 Ϯ 3.1%, n ϭ 8; mGluR7 siRNA: 5.5 Ϯ 1.1%, n ϭ 6). Taken together, these results suggest that activation of mGluR7 suppresses NMDAR currents via a postsynaptic mechanism in PFC pyramidal neurons.
␤-Arrestin2/ERK Signaling Pathway Is Involved in mGluR7 Regulation of NMDAR Currents-We then investigated the signaling mechanisms underlying mGluR7 reduction of NMDAR currents. Group III mGluRs are thought to be primarily coupled to G i/o proteins to inhibit cAMP formation and PKA activation (39). Thus, we first tested the involvement of PKA. As shown in Fig. 3, A and C, the L-AP4 reduction of NMDAR currents in dissociated PFC pyramidal neurons was not significantly changed by either PKA activator cpt-cAMP (21.6 Ϯ 3.8%, n ϭ 8) or PKA inhibitor PKI14 -22 (22.2 Ϯ 3.5%, n ϭ 8). Other than the PKA signaling pathway, previous studies have shown that L-AP4 can activate PI3K and ERK pathways in cultured cerebellar granule cells (21). Thus, we tested their potential involvement. As shown in Fig. 3, B and C, the L-AP4 reduction of NMDAR currents was largely blocked by the ERK kinase inhibitor PD98059 (50 M, 8.6 Ϯ 2.5%, n ϭ 10) or U0126 (20 M, 7.1 Ϯ 2.1%, n ϭ 10), but not by the PI3K inhibitor wortmannin (0.5 M, 17.2 Ϯ 3.3%, n ϭ 8) or LY294002 (100 M, 17.8 Ϯ 3.5%, n ϭ 8). These results suggest that ERK signaling is specifically involved in mGluR7 regulation of NMDAR currents in PFC pyramidal neurons.
We further tested whether group III mGluRs could lead to the activation of ERK (indicated by Thr 202 /Tyr 204 phospho-ERK) in PFC. As shown in Fig. 3D Next, we tested what signaling pathway underlies the activation of ERK by L-AP4. It has been shown that some G proteins form a signaling complex with the multifunctional adaptor and transducer molecule, ␤-arrestin1/2, which recruits and activates components of the MAP kinase cascades (40,41). Thus, we examined the potential involvement of ␤-arrestins in L-AP4 activation of ERK and L-AP4 suppression of NMDAR currents. The siRNA against ␤-arrestin1 or ␤-arrestin2, which can spe-
mGluR7 Reduction of NMDAR Currents Depends on Mechanism Involving Cofilin-regulated Actin Cytoskeleton-We next investigated the potential mechanisms underlying mGluR7 regulation of NMDAR currents. After NMDARs leave the endoplasmic reticulum, they are transported first along microtubules on dendritic shafts and then along actin filaments in dendritic spines before being delivered to the synaptic membrane. Our previous studies have shown that neuromodulators could affect NMDAR trafficking via mechanisms depending on the integrity of these cytoskeletal systems (30,32). Thus, we examined whether L-AP4 reduction of NMDAR currents is through an actin-or microtubule-dependent mechanism. As shown in Fig. 5A, pretreatment with the actin-depolymerizing agent latrunculin B (5 M, 20 min) prevented the effect of L-AP4 on NMDAR currents (6.9 Ϯ 1.6%; n ϭ 8, Fig. 5C). Dialysis with the actin stabilizer phalloidin (2 M) also blocked the effect of L-AP4 (8.6 Ϯ 1.3%, n ϭ 9, Fig. 5C). In contrast, the effect of L-AP4 on NMDAR currents was not altered by the microtubule depolymerizer nocodazole (50 M, Fig. 5B, 20.8 Ϯ 3.8%, n ϭ 6, Fig. 5C) or the microtubule stabilizer taxol (10 M, Fig. 5B, 21.9 Ϯ 3.9%, n ϭ 6, Fig. 5C). It suggests that mGluR7 reduces NMDAR currents by interfering with the trafficking of NMDA receptors along actin filaments.
The dynamics of actin assembly are regulated by several key proteins, one of which is cofilin (43), an important member of the actin-depolymerizing factor (ADF) family. The activity of cofilin is mainly regulated by phosphorylation at a single site (Ser 3 ). Dephosphorylation at this site greatly increases the actin-depolymerizing activity of cofilin (44). Thus, we tested whether mGluR7 might regulate actin-based NMDAR trafficking by interfering with cofilin-mediated actin depolymerization.
First, we examined whether mGluR7 could affect the phosphorylation and activity of cofilin. As shown in Fig. 6, A and B, with an antibody specifically recognizing Ser3 phospho-cofilin, L-AP4 treatment (200 M, 10 min) significantly decreased the level of Ser 3 -phosphorylated cofilin in cultured PFC neurons (63 Ϯ 11% of control, n ϭ 8), which was prevented by pretreatment with the ERK kinase inhibitor PD98059 (105 Ϯ 19% of control, n ϭ 6). This suggests that mGluR7 activation increases the actin-depolymerizing activity of cofilin via an ERK-dependent mechanism. Consistently, the ERK signaling pathway has been shown to mediate cofilin dephosphorylation and activation induced by Src or Ras in non-neuronal cells (45,46).
Next, we tested whether cofilin was involved in L-AP4 regulation of NMDAR currents. Because cofilin is inactivated by phosphorylation at Ser 3 and reactivated by dephosphorylation, two peptides consisting of 1-16 residues of cofilin with or without Ser 3 -phosphorylated (47,48) were used. The phospho-cofilin peptide serves as an inhibitor of endogenous cofilin because it binds to cofilin phosphatases and thus prevents the dephosphorylation (activation) of endogenous cofilin, whereas    Fig. 6D). These results suggest that mGluR7 may reduce NMDAR currents by increasing cofilin activity and the ensuing actin depolymerization.
Activation of mGluR7 Decreases Surface Level of NMDARs through Actin-dependent Mechanism-Next, we would like to know how mGluR7-induced actin depolymerization leads to the suppression of NMDAR currents. Previous studies have suggested that the integrity of actin cytoskeleton is important for stabilizing and/or promoting surface NMDAR expression (30,49), so we tested the involvement of NMDAR trafficking in mGluR7 regulation of NMDAR currents. We first examined whether mGluR7 affects the association of NMDARs with actin cytoskeleton. As shown in Fig. 8, L-AP4 treatment caused a significant decrease of the binding between NR1 and actin (53.0 Ϯ 8.8%, n ϭ 8), and this effect was blocked by pretreatment (15 min prior to L-AP4 application) with the mGluRIII antagonist CPPG (8.0 Ϯ 1.3%, n ϭ 6) or the membrane-permeable actin stabilizer phalloidin oleate (5.0 Ϯ 1.1%, n ϭ 6). Because the clathrin-dependent   internalization of NMDARs is prevented by the postsynaptic scaffolding protein PSD-95 (50, 51), we further tested whether mGluR7 affects the association of NR1 and PSD-95 in an actin-dependent manner. As shown in Fig. 8, L-AP4 treatment significantly decreased the binding between NR1 and PSD-95 (37.0 Ϯ 5.6%, n ϭ 8), which was also prevented by CPPG (6.0 Ϯ 1.1%, n ϭ 6) or phalloidin oleate (2.0 Ϯ 0.5%, n ϭ 6). These data suggest that mGluR7-induced actin depolymerization results in less NR1 binding to actin assembly, which may contribute to the dissociation of NMDARs from the scaffold protein complex.
The decreased binding between NR1 and PSD-95 in response to L-AP4 could result in reduced surface expression of NMDARs. To test this, we performed immunochemical staining of surface NR1 using an antibody recognizing the extracellular domain of NR1 in cultured PFC neurons under the nonpermeabilized condition. As shown in Fig To further complement the staining results, we also performed surface biotinylation experiments (31,33) to measure the level of surface NR1 in PFC slices. Surface proteins were labeled with sulfo-NHS-LC-biotin, and then biotinylated surface proteins were separated from nonlabeled intracellular proteins by reaction with NeutrAvidin beads. Surface and total proteins were subjected to electrophoresis and probed with an antibody against the NR1 subunit. As shown in Fig. 9, B and C, L-AP4 treatment significantly decreased the level of surface NR1 (28.7 Ϯ 2.1% decrease, n ϭ 8), which was blocked by pretreatment with phalloidin oleate (5.5 Ϯ 0.6% decrease, n ϭ 8). These results suggest that mGluR7 activation decreases the level of surface NMDARs through an actin-de-

DISCUSSION
In this study, we have revealed the regulation of NMDA receptors by mGluR7 signaling (Fig. 10). The mGluR subtypes, which have heterogeneous distribution and differential signal transduction mechanisms, are thought to serve as a modulator to fine-tune synaptic efficacy in distinct manners (34, 52) All three groups of mGluRs have been linked to long-term synaptic plasticity such as long-term potentiation/long-term depression (53,54). Previous slice recordings have shown that group III mGluRs could act presynaptically to reduce glutamatergic and GABAergic transmission at different synapses (36,(55)(56)(57). However, mGluR4 has been found to be expressed postsynaptically as well as presynaptically in hippocampus by anatomic studies (58). Functional studies also suggest that postsynaptic mGluR4 might mediate the reducing effect of a low concentration of L-AP4 (20 M) on the inward current evoked by exogenous NMDA application in nucleus accumbens slices (59). The pre-and postsynaptic presence of mGluR7 (22)(23)(24) suggests that mGluR7 may have postsynaptic functions, which are masked by presynaptic actions when examining in brain slices.
Because the effects of mGluR agonists on NMDAR-EPSC in slices could be mediated by either presynaptic or postsynaptic actions, we mainly used freshly isolated neurons from slices, which have the following advantages. First, it is a pure postsynaptic system, so we are able to reveal the postsynaptic action of mGluRs. Second, NMDAR-mediated current can be measured accurately with fast application of ligand (NMDA) to the cell under recording. Third, the intracellular mechanism underlying mGluR regulation of NMDARs can be investigated directly without the indirect effects from neighboring neurons at the circuit. Using combined electrophysiological, biochemical, and immunocytochemical approaches, we have revealed an inhibitory effect on NMDAR trafficking and function by group III mGluRs in PFC pyramidal neurons. Pharmacological studies and cellular knockdown experiments indicate that mGluR7 is the receptor mediating this postsynaptic effect.   Emerging evidence suggests that NMDA receptors undergo regulated transport to and from the cell surface and lateral diffusion at synaptic and extrasynaptic sites in the plasma membrane (60). NMDARs are tethered to actin cytoskeleton via scaffolding and adaptor proteins such as ␣-actinin and PSD-95 (61,62). Actin depolymerization reduces NMDA channel activity (63) and the number of synaptic NMDAR clusters (49). Moreover, actin depolymerization plays a key role in triggering long-term depression of NMDA synaptic responses in hippocampus (64). Here we provide multiple lines of evidence showing that mGluR7, via activation of ERK, increases the actin-depolymerizing activity of cofilin, leading to the loss of actin filaments at synapses. Consequently, the association of NMDARs with actin and PSD-95 is reduced, which causes the loss of surface NMDAR clusters and the depression of NMDAR currents. This mGluR7-mediated regulation of NMDA responses is different from group I mGluR-induced long-term depression, which is due to the increased AMPAR endocytosis rate (65,66) mediated by the dendritic synthesis of Arc/Arg3.1, an activity-regulated cytoskeleton-associated protein (67).
Unlike the presynaptically located mGluRs, which inhibit transmitter release and therefore affect the response of many postsynaptic receptor subtypes, the postsynaptic modulation of NMDARs by mGluR7 offers a way for more specific regulation. Given the key role of NMDARs in cognition and emotion (25,68), our results provide a potential mechanism for the involvement of mGluR7 in a variety of PFC-controlled mental functions, which was revealed from mGluR7 deficient mice (9,10,14).