Synaptic ERK2 Phosphorylates and Regulates Metabotropic Glutamate Receptor 1 In Vitro and in Neurons

Abstract

A synaptic pool of extracellular signal-regulated kinases (ERK) controls synaptic transmission, although little is known about its underlying signaling mechanisms. Here, we found that synaptic ERK2 directly binds to postsynaptic metabotropic glutamate receptor 1a (mGluR1a). This binding is direct and the ERK-binding site is located in the intracellular C-terminus (CT) of mGluR1a. Parallel with this binding, ERK2 phosphorylates mGluR1a at a cluster of serine residues in the distal part of mGluR1a-CT. In rat cerebellar neurons, ERK2 interacts with mGluR1a at synaptic sites, and active ERK constitutively phosphorylates mGluR1a under normal conditions. This basal phosphorylation is critical for maintaining adequate surface expression of mGluR1a. ERK is also essential for controlling mGluR1a signaling in triggering distinct postreceptor signaling transduction pathways. In summary, we have demonstrated that mGluR1a is a sufficient substrate of ERK2. ERK that interacts with and phosphorylates mGluR1a is involved in the regulation of the trafficking and signaling of mGluR1.

Keywords: Cerebellum; G protein-coupled receptor; IP3; MAPK; Phosphorylation; Src; mGluR.

Fig. 1

ERK2-mediated phosphorylation of mGluR1a. a GST-fusion proteins containing distinct intracellular domains of rat mGluR1a. b A representative autoradiograph illustrating the ERK2-induced phosphorylation of GST-mGluR1a-CT2 and Elk-1 but not GST and other mGluR1a fragments. A gel staining with Coomassie Brilliant Blue (CBB) was present below to show protein loading. Note that the CT2 segment was markedly phosphorylated by ERK2. c A representative autoradiograph above a CBB staining illustrating phosphorylation of GST-mGluR1a-CT2 by active but not inactive ERK2. d A representative autoradiograph above a CBB staining illustrating phosphorylation of GST-mGluR1a-CT2 in the presence but not absence of ATP. e CIP-mediated dephosphorylation of ERK2-phosphorylated GST-mGluR1a-CT2. All phosphorylation reactions were carried out at 30 °C for 30 min (b–e) followed by dephosphorylation reactions (e). The reactions were then subjected to gel electrophoresis followed by autoradiography. Solid and open arrows indicate phosphorylated GST-mGluR1a-CT2 (pGST-mGluR1a-CT2) and Elk-1, respectively. Red stars in CBB staining images (b–d) indicate the CT2 and Elk-1 bands that were phosphorylated by ERK2. Note that these bands smeared and shifted upward as a result of slower migration due to phosphorylation

Fig. 2

Phosphorylation of mGluR1a-CT2 by ERK2. a Amino acid sequence analysis of mGluR1a-CT2(P1001-L1199). Eight ERK2 phosphorylation motifs (T/SP) are highlighted in red. Potential phosphorylation sites include T1064 (1), S1098 (2), T1143 (3), S1147 (4), T1151 (5), S1154 (6), S1169 (7), and T1178 (8). b A representative phosphoamino acid analysis showing GST-mGluR1a-CT2 phosphorylation at serine (pSer), but not threonine (pThr) and tyrosine (pTyr), residues. c, d Phosphorylation of mGluR1a-CT2 at pS/TP (c) and pTP (d) motifs by active ERK2. e Five mutants (M1–5) derived from mGluR1a-CT2(P1001-L1199). Site-directed mutants include mutations of T1064/S1098 to alanines (M1), T1064/S1098/T1143/S1147 to alanines (M2), T1064/S1098/T1143/S1147/T1151/S1154 to alanines (M3), four serines (S1098/S1147/S1154/S1169) to alanines (M4), and four threonines (T1064/T1143/T1151/T1178) to alanines (M5). e Phosphorylation of mGluR1a-CT2 mutants and WT at pS/TP by active ERK2. Note that no signal was detected in the M4 mutant. g Phosphorylation of GST-mGluR1a-CT2 and Elk-1 at PXpSP by active ERK2. Phosphorylation of CT2 and Elk-1 was detected by immunoblot (IB) with a phosphomotif antibody against pS/TP (c, f), pTP (d), or PXpSP (g)

Fig. 3

ERK2 binding to mGluR1a. a In vitro binding assays with immobilized GST-fusion proteins containing distinct intracellular domains of mGluR1a and purified ERK2. Inactive ERK2 proteins were used in these assays. The inactive state of ERK2 was confirmed by the lack of phosphorylation signals from ERK2 proteins when tested in immunoblots (IB) with a phospho-specific antibody against pERK1/2. b In vitro binding assays with immobilized mGluR1a CT1 fragments and inactive or active ERK2. c GST-fusion proteins containing different fragments of mGluR1a CT1. d, e In vitro binding assays with immobilized GST-fusion mGluR1a-CT1a–c proteins and inactive (d) or active (e) ERK2. Note that CT1a but not CT1b and CT1c precipitated ERK2 and pERK2. ERK2 and pERK2 proteins bound to GST-fusion proteins were visualized with immunoblots using the specific antibodies against ERK2 and pERK1/2, respectively

Fig. 4

Interactions of ERK2 with mGluR1a in rat cerebellar neurons. a, b Immunoblot detection of mGluR1a (a) and mGluR5 (b) proteins in synaptic membranes extracted from the cerebellum and striatum. Note that a smaller amount of cerebellar proteins (2.5 µg) than that of striatal proteins (25 µg) was loaded (a). Major dimer bands around 250 kDa and weak monomer bands (130–140 kDa) were seen for mGluR1a in the cerebellum and striatum, while mGluR5 bands (mainly dimers) were seen in the striatum but not cerebellum. c, d Immunoblot detection of ERK1/2 (c) and pERK1/2 (d) proteins in whole-cell homogenates and synaptic plasma membranes of cerebellar neurons (12.5 µg per lane). Note that ERK2 and pERK2 were expressed at a higher level than respective ERK1 and pERK1 in synaptic locations. e Coimmunoprecipitation (IP) assays of ERK1/2 and mGluR1a with the anti-mGluR1a antibody (Ab). Note that ERK2 clearly existed in mGluR1a precipitates (lane 5). Lanes 3 and 4 showed no specific bands due to the lack of a precipitating antibody (L3) and the use of an irrelevant IgG (L4). f Reverse coimmunoprecipitation assays of ERK1/2 and mGluR1a with the anti-ERK1/2 antibody. g Coimmunoprecipitation of pERK1/2 and mGluR1a. Synaptic proteins solubilized from enriched synaptic plasma membranes of the rat cerebellum were used in above IP assays. Precipitated proteins were visualized by immunoblots (IB) with indicated antibodies

Fig. 5

ERK-mediated phosphorylation of mGluR1a in rat cerebellar neurons. a–c Phosphorylation of mGluR1a at ERK-preferred motifs, including S/TP (a), TP (b), and PXSP (c). Cerebellar mGluR1a was immunopurified by the anti-mGluR1a antibody. Phosphorylation signals at three phosphomotifs (pS/TP, pTP, and PXpSP) in immunopurified mGluR1a were detected by immunoblots (IB) with indicated antibodies. d Dephosphorylation of mGluR1a PXSP phosphorylation by λ-protein phosphatase (λ-PP). Cerebellar slices were incubated with λ-PP (200–400 units) for 1 h at 30 °C. Note that PXSP phosphorylation was dephosphorylated by λ-PP. e Representative immunoblots illustrating effects of U0126 on phosphorylation of mGluR1a at PXSP and TP motifs and on pERK1/2-mGluR1a and ERK1/2-mGluR1a interactions in cerebellar neurons. f, g Quantifications of effects of U0126 on mGluR1a PXSP (f) and TP (g) phosphorylation. h Quantifications of effects of U0126 on pERK2-mGluR1a and ERK2-mGluR1a interactions. Cerebellar slices were incubated with U0126 (0.5 or 5 µM) for 30 min at 30 °C (e–h). Slices were collected for mGluR1a immunoprecipitation (IP). Phosphorylation levels of mGluR1a at PXSP and TP and proteins bound to immunopurified mGluR1a (pERK1/2 and ERK1/2) were visualized by immunoblots. Values in graphs were measured as ratios of PXpSP to mGluR1a (f), pTP to mGluR1a (g), pERK2 to mGluR1a (h), and ERK2 to mGluR1a (h). All values were analyzed by one-way ANOVA: PXpSP (f), F(2, 7) = 5.05, p < 0.05; pTP (g), F(2, 7) = 0.26, p > 0.05; pERK2 (h), F(2, 7) = 37.44, p < 0.05; and ERK2 (h), F(2, 7) = 14.67, p < 0.05. Data are presented as means ± SEM (n = 3–4 per group). *p < 0.05 versus vehicle

Fig. 6

ERK-mediated surface expression of mGluR1a in rat cerebellar neurons. a A representative mGluR1a immunoblot from rat viable cerebellar slices treated with BS³ or vehicle control (Con). b Quantification of intracellular mGluR1a dimers in BS³-treated and control slices. c, d A representative immunoblot of mGluR5 (c) or α-actinin (d) from rat viable cerebellar slices treated with BS³ or vehicle control. e Effects of U0126 on surface and intracellular expression of mGluR1a. Note that U0126 significantly reduced surface expression of mGluR1a. f Effects of U0126 on the surface-to-intracellular ratio (S:I ratio) of mGluR1a expression. g Effects of U0126 on α-actinin expression. Representative immunoblots are shown to the left of the quantified data (e, g). Cerebellar slices were incubated with vehicle (Veh) or U0126 (5 µM) for 30 min at 30 °C (e–g). After drugs were washed off, slices were used for BS³ crosslinking assays. All values were analyzed by Student’s t test: BS³-unlinked dimers (b), t = 4.45, p < 0.05; surface mGluR1a (e), t = 2.48, p < 0.05; intracellular mGluR1a (e), t = 2.19, p > 0.05; S:I ratio (f), t = 4.09, p < 0.05; and α-actinin (g), t = 0.68, p > 0.05. Data are presented as means ± SEM (n = 4–6 per group). *p < 0.05 versus vehicle

Fig. 7

ERK regulates the mGluR1-mediated IP₃ production in rat cerebellar neurons. a A rapid elevation of cytosolic IP₃ levels following application of DHPG (100 µM). b Effects of the mGluR1 antagonist 3-MATIDA on the DHPG-stimulated IP₃ formation. Note that 3-MATIDA completely blocked IP₃ responses to DHPG. c Effects of U0126 on the DHPG-stimulated IP₃ production. Note that U0126 at 5 µM significantly reduced IP₃ responses to DHPG. d Tat-fusion mGluR1a interaction-dead (Tat-mGluR1a-id) and sequence-scrambled control (Tat-mGluR1a-con) peptides. The underlined letters (LNIFRRKK) represent core hydrophobic and basic residues in a D domain-like sequence of mGluR1a-CT. e, f Effects of Tat-fusion peptides on the ERK2-mGluR1a association (e) and DHPG-stimulated IP₃ production (f). Note that Tat-mGluR1a-id but not Tat-mGluR1a-con peptides reduced the ERK2-mGluR1a association and IP₃ responses to DHPG. Experiments were carried out in rat cerebellar slices. 3-MATIDA (10 µM) or U0126 (0.5 or 5 µM) was applied 30 min before and during 20–25 s incubation of DHPG (100 µM) (b, c). Tat-fusion peptides (5 µM in e and 2 or 5 µM in f) were added to cerebellar slices for 30 min. Slices were then used for coimmunoprecipitation detection of the ERK2-mGluR1a association (e) or IP₃ measurements (f). Values in the panel a were analyzed by Student’s t test (t = 3.21, p < 0.05). Values in panels b, c, e, and f were analyzed by one-way ANOVA: 3-MATIDA + DHPG (b), F(3, 16) = 35.05, p < 0.05; U0126 + DHPG (c), F(5, 26) = 19.10, p < 0.05; ERK2-mGluR1a interactions (e), F(2, 15) = 7.50, p < 0.05; and Tat-peptides + DHPG (f), F(9, 29) = 9.55, p < 0.05. Data are presented as means ± SEM (n = 4–7 per group). *p < 0.05 versus vehicle (a, e) or vehicle + vehicle (b, c, and f). +p < 0.05 versus vehicle + DHPG (b, c, and f)

Fig. 8

Effects of inhibition of ERK by U0126 on the DHPG-stimulated Src Y416 phosphorylation in rat cerebellar neurons. a Effects of DHPG (100 µM, 15 min) on Src Y416 phosphorylation. b Effects of the mGluR1 antagonist 3-MATIDA on the DHPG-stimulated Y416 phosphorylation. Note that 3-MATIDA completely blocked the Y416 phosphorylation induced by DHPG. c Effects of U0126 (0.5 and 5 µM, 30 min) on basal Y416 phosphorylation. d Effects of U0126 on the DHPG-stimulated Y416 phosphorylation. Note that U0126 at 5 µM substantially blocked the Y416 phosphorylation induced by DHPG. 3-MATIDA (10 µM) or U0126 (0.5 or 5 µM) was applied 30 min before and during vehicle (Veh) or DHPG incubation (100 µM, 15 min) (b, d). Experiments were carried out on rat cerebellar slices. Values in the panel a were analyzed by Student’s t test (t = 3.08, p < 0.05). Values in panels b–d were analyzed by one-way ANOVA: 3-MATIDA (b), F(3, 16) = 10.02, p < 0.05; U0126 (c), F(2, 9) = 0.46, p > 0.05; and U0126 + DHPG (d), F(3, 16) = 11.22, p < 0.05. Data are presented as means ± SEM (n = 4–5 per group). *p < 0.05 versus vehicle (a) or vehicle + vehicle (b, d). +p < 0.05 versus vehicle + DHPG (b, d)