Rap1-induced p38 Mitogen-activated Protein Kinase Activation Facilitates AMPA Receptor Trafficking via the GDI·Rab5 Complex

Recent evidence has emphasized the importance of p38 mitogen-activated protein kinase (MAPK) in the induction of metabotropic glutamate receptor (mGluR)-dependent long term depression (LTD) at hippocampal CA3-CA1 synapses. However, the cascade responsible of mGluR to activate p38 MAPK and the signaling pathway immediately downstream from it to induce synaptic depression is poorly understood. Here, we show that transient activation of group I mGluR with the selective agonist (S)-3,5-dihydroxyphenylglycine (DHPG) activates p38 MAPK through G protein βγ-subunit, small GTPase Rap1, and MAPK kinase 3/6 (MKK3/6), thus resulting in mGluR5-dependent LTD. Furthermore, our data clearly show that an accelerating AMPA receptor endocytosis by stimulating the formation of guanyl nucleotide dissociation inhibitor-Rab5 complex is a potential downstream processing of p38 MAPK activation to mediate DHPG-LTD. These results suggest an important role for Rap1-MKK3/6-p38 MAPK pathway in the induction of mGluR-dependent LTD by directly coupling to receptor trafficking machineries to facilitate the loss of synaptic AMPA receptors.

Long term depression (LTD) 1 is a persistent activity-dependent decrease of synaptic efficacy that together with the converse process, long term potentiation, has been considered to be crucial for information storage in the brain (1,2). In the hippocampus, LTD is divided into three categories: homosynaptic, heterosynaptic, and associative LTD (3). The bestcharacterized form of homosynaptic LTD is induced in the CA1 region of the hippocampus by prolonged low frequency synaptic stimulation via a NMDA receptor-dependent rise in postsynaptic [Ca 2ϩ ] i and the activation of serine/threonine protein phosphatases (4). Recent work has shown that mechanistically distinct type of LTD can be induced in the CA1 region by other types of synaptic stimulation or brief pharmacological treatments. For example, a prolonged period of paired-pulse stimulation or a direct application of the selective group I mGluR agonists, such as DHPG, can induce a robust mGluR-dependent form of LTD that is independent of NMDA receptor activation (5). In contrast to the mechanisms of NMDA receptor-dependent LTD, which are fairly well established, the mechanisms of induction and the site of expression of mGluR-dependent LTD are still a matter of some considerable debate. Current studies have reported that the induction of mGluR-dependent LTD does not require extracellular Ca 2ϩ (6), Ca 2ϩ release from intracellular stores (7), activation of Ca 2ϩ /calmodulin-dependent protein kinase II (8), protein kinase A or protein kinase C (7,9), or serine/ threonine protein phosphatases (9). However, this form of LTD requires activation of G q -type G proteins (10), a local translation of dendritic mRNA (5,11), a long-lasting loss of postsynaptic AMPA receptors (12,13), and activation of protein tyrosine phosphatases (14). In addition, more recent studies suggest that p38 MAPK signaling also serves as a signal mediator in the induction of mGluR-dependent LTD (15,16). However, it is still unclear how mGluR activation leads to activate p38 MAPK and the signaling downstream of p38 MAPK to mediate mGluR-dependent LTD induction.
In an attempt to address these questions, in this study we have investigated the mechanisms involved in the induction of DHPG-LTD in hippocampal CA1 neurons. We confirm the previously described the role for p38 MAPK in the induction of DHPG-LTD. We show that by releasing G␤␥-subunits, the mGluR5 may activate Rap1 to mediate DHPG-induced p38 MAPK activation and an accelerating loss of surface AMPA receptor by stimulating the formation of GDI⅐Rab5 complex is a potential downstream processing of p38 MAPK activation to mediate DHPG-LTD.

EXPERIMENTAL PROCEDURES
Electrophysiology-Hippocampal slices (400 m) were prepared from 21-28-day-old young male Sprague-Dawley rats using standard procedures (17,18), allowed to recover for a minimum of 1 h, and then transferred to a submersion-type recording chamber continually perfused with 30 -32°C oxygenated ACSF solution (in mM: 117 NaCl, 4.7 KCl, 2.5 CaCl 2 , 1.2 MgCl 2 , 25 NaHCO 3 , 1.2 NaH 2 PO 4 , 11 glucose). Area CA3 was surgically removed after sectioning. Extracellular and whole cell recordings were carried out with Axoclamp-2B or Axopatch 200B amplifier (Axon Instruments, Union City, CA). The responses were low pass-filtered at 2 kHz, digitally sampled at 5-10 kHz, and analyzed using pCLAMP software (Version 8.0, Axon Instruments). In whole cell recordings, cells were held at Ϫ70 mV while series and input resistances were monitored throughout each experiment. The evoked postsynaptic responses were induced in CA1 stratum radiatum by stim-ulation (0.02-ms duration) of Schaffer collateral/commissural afferents at 0.033 Hz with a bipolar stainless steel-stimulating electrode. Field excitatory postsynaptic potentials (fEPSPs) were recorded with a glass pipette filled with 1 M NaCl (2-3-megohm resistance), and the initial slope was measured. Whole cell recording of EPSCs was made from CA1 pyramidal cells, which were identified under a differential interference contrast microscope. Patch pipettes (3-5 megohms) filled with the following internal solution were used (in mM: 110 potassium gluconate, 30 KCl, 10 HEPES, 1 MgCl 2 , 0.5 EGTA, 4 Na 2 ATP, 0.3 Na 3 GTP, 7 phosphocreatine, 5 lidocaine N-ethyl bromide quaternary (QX-314), pH 7.3, 290 -295 mosmol). The amplitude of evoked EPSCs was measured. To ensure stability of the whole cell recordings, electrical stimulation was initiated before the cell was patched. We waited for ϳ5 min in the cell-attached configuration before break-in to wash off any residual internal solution spilled from the approaching pipette.
Application of Drugs and Proteins-When experiments were performed in the presence of inhibitors, slices were preincubated in the inhibitors for a minimum of 1 h before and then transferred to the recording chamber where they were continually perfused with a solution containing the inhibitors. Drugs were diluted from stock solutions just before application. Stock solutions of PD98059, SP600125, GGTI-298, and FTI-277 were dissolved in Me 2 SO and stored at Ϫ20°C. The final Me 2 SO concentration in the superfusate was Ͻ0.1%. At this concentration, Me 2 SO had no effect on fEPSPs (17,18). LY341495 and LY367385 were prepared by first dissolving them into an equimolar amount of NaOH as a concentrated stock solution and then diluting to their final concentration in ACSF. Purified bovine brain G␤␥ was dissolved at 50 g/ml in 50 mM HEPES, 1 mM dithiothreitol, 1 mM EDTA, and 0.1% (1.25 M) Lubrol, pH 7.6, and stored at Ϫ80°C. To inactivate the protein, the stock solution was boiled at 95°C for 10 min. Boiled or nontreated G␤␥ was diluted with the internal solution added with 0.05% CHAPS for the final G␤␥ concentration to 100 nM. DHPG, MPEP, LY341495, LY367385, PD98059, SB203580 hydrochloride, and SP600125 were purchased from Tocris Cookson (Bristol, United Kingdom). Guanosine-5Ј-O-(2-thiodiphosphate) (GDP␤S) was from Sigma. GGTI-298, FTI-277, and bovine brain G␤␥ were obtained from Calbiochem.
Western Blotting-For each experimental group, homogenates from at least three slices were pooled. The microdissected subregions were lysed in ice-cold Tris-HCl buffer solution, pH 7.4, containing a mixture of protein phosphatase and proteinase inhibitors (50 mM Tris-HCl, 100 mM NaCl, 15 mM sodium pyrophosphate, 50 mM sodium fluoride, 1 mM sodium orthovanadate, 5 mM EGTA, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 M microcystin-LR, 1 M okadaic acid, 0.5% Triton X-100, 2 mM benzamidine, 60 g/ml aprotinin, 60 g/ml leupeptin) to avoid dephosphorylation and degradation of proteins and ground with a pellet pestle (Kontes Glassware, Vineland, NJ). Samples were sonicated and spun down at 15,000 ϫ g at 4°C for 10 min. The supernatant was then assayed for total protein concentration using Bio-Rad Bradford protein assay kit (Hercules, CA). Each sample was separated in 10% SDS-PAGE gel for p38 MAP kinase and MKK3/6 detection and in 12.5% gel for Rap1 detection. Following the transfer on polyvinylidene difluoride membranes, blots were blocked in buffer solution containing 5% milk and 0.1% Tween 20 in phosphate-buffered saline (124 mM NaCl, 4 mM KCl, 10 mM Na 2 HPO 4 , 10 mM KH 2 PO 4 , pH 7.2) for 1 h and then blotted for 2 h at room temperature with antibodies that recognize phosphorylated p38 MAPK (1:1000, New England BioLabs, Beverly, MA) or phosphorylated MKK3/6 (1:1000, Cell Signaling Technology). It was then probed with horseradish peroxidase-conjugated secondary antibody for 1 h and developed using the ECL immunoblotting detection system. The immunoblots using phosphorylation site-specific antibodies were subsequently stripped and reprobed with an antibody that recognizes p38 MAPK (1:500) or MKK3/6 (1:500). For immunoprecipitation of Rab5⅐GDI complex, homogenates (ϳ500-g protein) was incubated with anti-Rab5 antibody (1:50, BD Transduction Laboratories) for 4 h at 4°C. The antibody-protein complexes were then pelleted with protein A-Sepharose beads. Proteins eluted from the beads were subjected to 10% SDS-PAGE gel and immunoblotting for anti-GDI antibody (1:500, Synaptic Systems). Immunoblots were analyzed by densitometry.
Measurement of Rap1 and Ras Activity-The activity of Rap1 and Ras was determined using Rap1 and Ras activation assay kit (Upstate Biotechnology, Lake Placid, NY), respectively, following manufacturer's instructions. Rap1-GTP and Ras-GTP from various treated lysates were pulled down using the glutathione S-transferase fusion protein corresponding to human Rap1-binding domain of RalGDS or human Rasbinding domain of Raf-1 bound to agarose for 1 h at 4°C. The beads were washed with ice-cold lysis buffer and resuspended in SDS sample buffer. Western blotting (see above) was performed with anti-Rap1 or anti-Ras antibody (Upstate Biotechnology) to detect the presence of Rap-1-GTP and Ras-GTP.
Biochemical Measurements of Surface-expressed AMPA Receptors-Biotinylation experiments were performed as described previous (18). The microdissected subregions were then incubated with ACSF containing 1 mg/ml Sulfo-NHS-LC-Biotin (Pierce) for 30 min on ice. Unreacted biotinylation reagent was washed once with ice-cold ACSF and quenched by two successive 20-min washes in ACSF containing 100 mM glycine followed by two washes in ice-cold Tris-HCl buffer solution. The microdissected subregions were lysed and ground with a pellet pestle. Samples were sonicated and spun down at 14,000 ϫ g at 4°C for 15 min. 30-g lysates of the resulting supernatant were removed to measure total GluR1, and 90-g lysates of the supernatant were incubated with 100 l of 50% neutravidin-agarose (Pierce) for 3 h at 4°C to measure the isolated biotinylated proteins. After the neutravidin-agarose was washed five times with homogenate buffer, bound proteins were eluted with SDS sample buffer by boiling for 5 min. Total protein and isolated biotinylated proteins were analyzed by quantitative immunoblotting with polyclonal anti-GluR1 C terminus (1:1000, Upstate Biotechnology).
Data Analysis-All of the values reported are means Ϯ S.E. The significance of the difference between the mean was calculated by a paired or unpaired Student's t test as appropriate. Probability values of p Ͻ 0.05 were considered to represent significant differences.

DHPG-induced an mGluR5-dependent LTD-
Using field potential recordings, we first tested for the existence of an mGluRdependent form of LTD in the CA1 region of hippocampal slices. As reported previously (19,20), application of DHPG (50 M) for 5 min resulted in a substantial LTD of fEPSPs (DHPG-LTD). The synaptic responses were depressed to 63.5 Ϯ 5.1% (n ϭ 8; p Ͻ 0.05) measured at 40 min after washout of DHPG (Fig. 1A). This LTD was completely blocked by bath application of the broad-spectrum mGluR antagonist LY341495 (10 M; 94.9 Ϯ 5.3%, n ϭ 5; p Ͻ 0.05) (21), suggesting that it is mediated by mGluR (Fig. 1B). To further investigate which subtype of group I mGluR underlies this LTD, we next examined the effect of specific antagonists for mGluR1 and mGluR5 on this LTD induction. Application of the highly selective mGluR5 antagonist MPEP (10 M) (22) completely blocked the DHPG-LTD (96.3 Ϯ 3.8%, n ϭ 5; p Ͻ 0.05, Fig. 1C), whereas the highly selective mGluR1 antagonist LY367385 (100 M) (23) had no effect (58.9 Ϯ 4.5%, n ϭ 5; p Ͼ 0.05, Fig. 1D). These results, which are similar to previous findings in juvenile rat hippocampal slices (19,20), confirm that activation of mGluR5 is required for the induction of DHPG-LTD.
To further determine whether postsynaptic G protein activation mediated DHPG-LTD, we performed whole cell recordings from CA1 pyramidal cells using a potassium gluconate-based pipette solution where GTP was replaced with 1 mM GDP␤S to inhibit the activation of postsynaptic G proteins. Whereas DHPG induced a robust LTD of EPSCs in control recordings with GTP-containing solution (62.3 Ϯ 4.9%, n ϭ 7), only a small initial depression and no persistent depression were observed in recordings where GDP␤S was used in the pipette solution (106.9 Ϯ 5.4%, n ϭ 7; p Ͻ 0.05) (Fig. 3, A and B). As previously reported, after rupture with pipette solution containing GDP␤S, EPSCs gradually grew during the first 5-15 min and were stable for the remainder of the experiments. The initial "run-up" effects of GDP␤S might in part be attributed to the effects on GDP␤S on AMPA receptor endocytosis (28,29). Although these results indicate that DHPG-LTD is dependent on the activation of postsynaptic G q , no role has been found for the common G q ␣-stimulated signals (i.e. phospholipase C and protein kinase C) following mGluR5 activation in this form of LTD (30). Consequently, another mechanism-unrelated G␣ q must underlie the mGluR-dependent LTD. Because G␤␥ released by G protein-coupled receptor activation has important role in the G protein-mediated signaling and there is growing evidence demonstrating that several G protein-coupled receptors induce p38 MAPK activation through G␤␥ (31-33), we next investigated the role of G␤␥ in the induction of DHPG-LTD. To address it, we loaded G␤␥ (100 nM) directly into the cells. After establishing the whole cell configurations, EPSCs began to decrease and this "run-down" stabilized after ϳ20 min at ϳ40% reduction (Fig. 3C1). This manipulation completely blocked the induction of DHPG-LTD (102.9 Ϯ 4.6%, n ϭ 6; p Ͻ 0.05), whereas interleaved cells recorded from a similar period of time with the heat-inactivated G␤␥ exhibited stable LTD (63.5 Ϯ
Rap1 Mediates DHPG-induced p38 MAPK Activation-Having confirmed the role of G␤␥ and p38 MAPK in DHPG-LTD, we next identified the signaling components involved in G␤␥mediated p38 MAPK activation. Among many potential intracellular factors involved in p38 MAPK activation, we investigated the potential role of two small GTPases, Rap1 and Ras. In several experimental models, the family of small GTPase plays a key role in the activation of MAPK. Recently, it has been reported that Rap1, a Ras-like small GTPase, contributes to the induction of activity-induced LTD via phosphorylation of p38 MAPK, whereas Ras mediates the signaling activated by Ca 2ϩ /calmodulin-dependent protein kinase II that is responsible for long term potentiation (34,35). To determine whether Rap1 plays any role in mediating DHPG-LTD, we examined the induction of DHPG-LTD in the presence of membrane-permeable geranylgeranyltransferase inhibitor, GGTI-298, which blocks Rap1 activation (IC 50 ϭ 3 M) by preventing its recruitment to plasma membrane (36). As shown in Fig. 4A, GGTI-298 (10 M) abolished the ability of DHPG to induce LTD (97.8 Ϯ 3.6%, n ϭ 6; p Ͻ 0.05). In contrast, pretreatment of the slices with Ras farnesyltransferase inhibitor FTI-277 (10 M) (IC 50 ϭ 0.1 M) (37) failed to affect the induction of DHPG-LTD (66.3 Ϯ 4.5%, n ϭ 6, p Ͼ 0.05, Fig. 4B). Similarly, GGTI-298 (10 M) largely blocked the increase in p38 MAPK phosphorylation following the DHPG application (Fig. 4C). In contrast, FTI-277 (10 M) had no effect on the DHPG-induced increase in p38 MAPK phosphorylation. To directly demonstrate that Rap1 participates in the signaling pathway responsible for DHPGinduced activation of p38 MAPK, we investigated the activation of Rap1 by means of a pull-down assay using the glutathione S-transferase-RalRBD fusion protein. Application of DHPG (50 M) for 5 min induced a large increase in the amount of Rap1-GTP (62.8 Ϯ 5.7%, n ϭ 5, p Ͻ 0.05, Fig. 4D). This DHPG-induced Rap1 activation was blocked by the MPEP (10 M) but not by the LY367385 (100 M). In contrast to Rap1, we found that the level of GTP-Ras was not changed following DHPG application (105.3 Ϯ 4.9%, n ϭ 4; p Ͼ 0.05) (Fig. 4E).
These results indicate that p38 MAPK acts as a downstream target of Rap1 to induce DHPG-LTD.

DHPG-induced Activation of p38 MAPK by G␤␥ and Rap1
Occurs at the level of MKK3/6 -In vivo, p38 MAPK can be activated by upstream MKK3 and MKK6 (38). Thus, any changes in MKK3/6 phosphorylation should result in changes in p38 MAPK activity. As shown in Fig. 4F, phospho-MKK3/6 was readily detected in untreated control slices and levels were increased to 136.3 Ϯ 6.5% (n ϭ 6) of control following DHPG application. This DHPG-induced increase in phospho-MKK3/6 was fully prevented by pretreatment of the slices with GGTI-298 (10 M) but not by FTI-277 (10 M). Thus, MKK3/6 seems to be an immediate upstream signaling component of p38 MAPK to underlie DHPG-LTD.
Involvement of p38 MAPK-stimulated AMPA Receptor Endocytotic Trafficking in the DHPG-LTD-What is the mechanism that couples an activation of p38 MAPK to synaptic depression? It has recently been demonstrated that DHPG-LTD appears to require loss of AMPA receptors from the postsynaptic membrane surface (12,13) and p38 MAPK has been identified to have the ability to accelerate the endocytosis trafficking by stimulating the activity of GDI in extracting Rab5 from endosomal membranes and forming a GDI⅐Rab5 complex (39). This complex is required for ligand sequestration into clathrincoated pits. The potential involvement of p38 MAPK in DHPGinduced loss of surface AMPA receptors was first investigated. In agreement with previous reports (12,13), application of DHPG (50 M) caused a significant reduction in surface expression of AMPA receptors (GluR1: 65.6 Ϯ 6.9% of control slices, n ϭ 5; p Ͻ 0.05) (Fig. 5A). This DHPG-induced loss of surface AMPA receptors was completely blocked by SB203580 (1 M) and GGTI-298 (10 M).
We finally assessed whether p38 MAPK contributes to regulate of endocytic trafficking of surface AMPA receptors by stimulating the formation of GDI⅐Rab5 complex. Following application of DHPG, tissue lysates were immunoprecipitated with anti-Rab5 antibody and immunoblotted with an antibody specific for GDI to measure GDI association with Rab5. Fig. 5B demonstrates that DHPG induced an increase in GDI associa- p38 MAPK Mediates DHPG-LTD tion with Rab5 by 34.4 Ϯ 4.9% (n ϭ 5; p Ͻ 0.05). This DHPGinduced increase in the formation of the GDI⅐Rab5 complex was inhibited by SB203580 (1 M). These observations demonstrate that DHPG accelerates a rapid loss of surface AMPA receptors, presumably via p38 MAP kinase activation to stimulate the formation of GDI⅐Rab5 complex.

DISCUSSION
Our results provide a plausible mechanism of the induction of DHPG-LTD in hippocampal CA1 neurons (Fig. 6). Activation of mGluR5 by DHPG releases G␤␥-dimers, which in turn promotes the exchange of GDP with GTP of Rap1 and activates a sequential kinase cascades that includes MKK3/6 and p38 MAPK. Activated p38 MAPK enhances the activity of GDI in extracting Rab5 from endosomal membranes leading to a reduced expression of postsynaptic AMPA receptor and eventually resulting in the reduction in synaptic strength observed with LTD.
The finding that DHPG-LTD was completely eliminated in mice lacking G␣ q strongly suggests that G q plays a crucial role in the induction of DHPG-LTD (10). However, recent work has shown that neither the activation of protein kinase C nor inositol trisphosphate-mediated Ca 2ϩ release from intracellular stores, a core signaling pathway to the G␣ q activation, is required for DHPG-LTD. This implies that other signaling process downstream of G q activation may underlie such LTD. Because in many cases it is the G␤␥-dimers rather than G␣ that transmit the signals through G protein-coupled receptors (40), it is likely that the signal from the mGluR to activate p38 MAPK is mediated by G␤␥. This prediction is supported by our finding that intracellular loading of G␤␥ mimicked and occluded the DHPG effect. Similarly, in human embryonic kidney 293 cells G q/11 -coupled m1 muscarinic receptor was found to stimulate p38 MAPK by G␤␥ (31). What is the mechanism that couples G␤␥ to the activation of p38 MAPK? Because the structure of p38 MAPK does not contain the specific G␤␥-binding domain (41), it is unlikely that G␤␥ activates p38 MAPK through a direct G␤␥/p38 MAPK association. In this case, additional downstream effectors of G␤␥ are also required. G␤␥ can activate MAPK via Ras in other cell types. For example, it has been shown that G␤␥ could recruit phosphatidylinositol 3-kinase to the plasma membrane, enhancing the activity of Src-like kinase, which in turn leads to the activation of the Shc-Grb2-Sos-Ras pathway, resulting in increased MAPK activity (42,43). In this study, our results suggest that the activation of Rap1 and MKK3/6 is essential for DHPG-induced p38 MAPK activation. Active Rap enhances the levels of active p38 MAPK, consistent with recent findings indicating p38 MAPK as a downstream target of Rap (43) and able to control some forms of LTD (15). Although how G␤␥ increases in Rap1 activity remains to be elucidated, these results imply a model where Rap1 can specifically serve as the downstream effector of G␤␥ to stimulate MKK3/6 and p38 MAPK activation.
It has recently reported that the expression of DHPG-LTD is associated with a loss of AMPA receptors from postsynaptic membrane surface (12) and that disruption of endocytic processes by injecting D15, which inhibits the association of dynamin with Quantification of Ras-GTP immunoreactivity over control base-line levels revealed that DHPG has no effect on Ras activation. F, representative immunoblot showing DHPG-induced increase in phospho-MKK3/6 levels was significantly reduced in GGTI-298treated slices but not by FTI-277-treated ones. Group data showing the normalization of phospho-MKK3/6 to the nonphosphorylated form was determined in each group of six experiments.

p38 MAPK Mediates DHPG-LTD
amphiphysin, prevents the induction of DHPG-LTD (13). We have confirmed these results and find that DHPG-induced internalization of AMPA receptors requires the activation of Rap1 and p38 MAPK. Indeed, evidence that active Rap1 enhances levels of active p38 MAPK, which in turn leads to the removal of synaptic AMPA receptors, resulting in occluding the induction of LTD, has been provided more recently in a study on cultured hippocampal slices (34). To date, the locus of expression of DHPG-LTD is highly controversial with evidence for both presynaptic (6,29,44,45) and postsynaptic changes (12,13). Here, we provide evidence that interventions that attenuated the DHPG-induced loss of postsynaptic surface AMPA receptors also prevented the expression of DHPG-LTD, supporting a postsynaptic locus for the induction of DHPG-LTD.
How might p38 MAPK activation lead to the removal of surface AMPA receptors? A recent study in nonneuronal cells has indicated that p38 MAPK-dependent response to stresses such as hydrogen peroxide and UV irradiation stimulates the formation of the GDI⅐Rab5 complex and accelerates endocytosis, whereas inhibition of the p38 MAPK signaling has the opposite effects (39). In addition, the GDI⅐Rab5 complex has been shown to be an important component of the machinery controlling clathrin-coated endocytotic vesicle formation (46). Here we find that the cytosolic association of GDI and Rab5 was significantly enhanced by DHPG and that such an effect was prevented by blocking p38 MAPK signaling. These results imply that stimulated formation of GDI⅐Rab5 complexes are probably one of the steps mediating the signal initiated by activated p38 MAPK and eventually leading to the loss of surface AMPA receptors. The mechanism by which p38 MAPK regulates GDI in the cytosolic cycle of Rab5 and regulates AMPA receptor trafficking remains to be determined. At excitatory synapses, the loss of AMPA receptors can occur either by diminished recycling or by enhanced endocytosis (47). According to this view, it will be important for future studies to identify which of these mechanisms underlie the long-lasting reduction of surface AMPA receptors during the expression of DHPG-LTD.
Previous work from ours and other laboratories has demonstrated that a rapid dendritic protein synthesis is required for the induction of DHPG-LTD and depotentiation (11,18,48). Although in situ hybridization data indicate that various mRNAs and polyribosomes are present in the dendrites of mature neurons in vivo (49,50), the identities of the dendritic protein required for DHPG-LTD are not known. Because the expression of DHPG-LTD is correlated with a rapid loss of AMPA receptors from postsynaptic membrane surface, one possibility is that the functional consequence of dendritic protein synthesis is involved in promoting AMPA receptor endocytosis in response to application of DHPG. Further work involving the use of functional knock-outs of candidate proteins or proteomic approach is needed to assess the potential roles of candidate proteins in the expression of DHPG-LTD.
In conclusion, we have provided further evidence for a functional contribution of p38 MAPK in the induction of mGluRdependent LTD. The p38 MAPK cascade is a major signaling pathway activated by Rap1 and also has been demonstrated to be an important mediator of AMPA receptor trafficking during synaptic plasticity. It remains to be determined whether such p38 MAPK signaling pathway is also responsible for the induction of other forms of LTD. showing that 30 min following DHPG surface GluR1 levels were significantly reduced, which can be prevented by SB203580 or GGTI-298 pretreatment. B, representative immunoblot showing that DHPG treatment increased the amount of GDI associated with Rab5. Pretreatment with SB203580 (1 M) significantly reduced the formation of the GDI⅐Rab5 complex. Densitometric quantitation from five separate experiments is summarized in the histogram plot.