Skip to main content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Proc Natl Acad Sci U S A. 2007 Feb 27; 104(9): 3579–3584.
Published online 2007 Feb 21. doi: 10.1073/pnas.0611698104
PMCID: PMC1805611
PMID: 17360685

Regulation of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor trafficking through PKA phosphorylation of the Glu receptor 1 subunit

Heng-Ye Man,* Yoko Sekine-Aizawa,* and Richard L. Huganir†‡
Author information Article notes Copyright and License information PMC Disclaimer

Abstract

α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors mediate the majority of excitatory synaptic transmission in the brain. Recent studies have shown that activation of PKA regulates the membrane trafficking of the AMPA receptor Glu receptor 1 (GluR1) subunit, but the role of direct phosphorylation of GluR1 in regulating receptor redistribution is not clear. Here we show that phosphorylation of the GluR1 subunit on serine 845 by PKA is required for PKA-induced increases in AMPA receptor cell-surface expression because it promotes receptor insertion and decreases receptor endocytosis. Furthermore, dephosphorylation of GluR1 serine 845 triggers NMDA-induced AMPA receptor internalization. These findings strongly suggest that dynamic changes in direct phosphorylation of GluR1 by PKA are crucial in the modulation of AMPA receptor trafficking and synaptic plasticity.

Keywords: glutamate receptors, long-term potentiation, synaptic plasticity

AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptors, which are responsible for the majority of excitatory synaptic transmissions in the mammalian central nervous system, undergo constant trafficking between the plasma membrane and cytosolic compartments. Alterations of these dynamic processes will redistribute AMPA receptors in neurons and might underlie changes in synaptic strength, such as long-term potentiation and long-term depression (1–5).

AMPA receptor trafficking is regulated by a variety of mechanisms (1–5). Activation of glutamatergic receptors, including NMDA receptors, AMPA receptors, and metabotropic Glu receptors (mGluR), triggers rapid AMPA receptor internalization (4), whereas selective activation of synaptic NMDA receptors facilitates AMPA receptor surface insertion (6, 23). Interactions of AMPA receptors with multiple proteins such as PICK1, NSF, and stargazin play a key role in AMPA receptor dynamics and its synaptic accumulation (4, 5).

The intracellular signaling pathways leading to AMPA receptor relocation are being actively examined and are still controversial. The involvement of different kinases and phosphatases in protein redistribution suggests that protein phosphorylation is a major cellular mechanism in regulating intracellular trafficking (7). AMPA receptors are members of a substrate of protein kinases that includes PKA, PKC, and Ca2+/calmodulin-dependent PK II (1). Phosphorylation of Glu receptor 1) (GluR1) subunits is closely correlated with AMPA receptor redistribution and expression of both long-term potentiation and long-term depression, indicating that PKA might regulate AMPA receptor trafficking via direct phosphorylation of GluR1 subunits (1, 8, 9). However, the specific role of GluR1 phosphorylation has not been clearly defined. Here we report that PKA phosphorylation of GluR1 at serine 845 (GluR1S845) increases AMPA receptor cell-surface expression as the result of a combination of increased receptor insertion and decreased internalization and that a dynamic dephosphorylation of this site is critical in NMDA-dependent AMPA receptor internalization.

Results

PKA Increases AMPA Receptor Cell-Surface Expression by Facilitating Receptor Plasma Membrane Insertion and Inhibiting Receptor Internalization.

To examine the role of PKA phosphorylation in the regulation of AMPA receptor trafficking in neurons, we used surface biotinylation techniques (9) (also see Methods). A 15-min treatment of cultured cortical neurons with the cAMP-elevating agent forskolin plus phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (forskolin treatment) increased AMPA receptor expression on the plasma membrane to 155.0 ± 13.2% of control without changing the total AMPA receptor protein amount (Fig. 1 A and B). The increase in surface AMPA receptor expression could be observed 5 min after forskolin application, and the PKA effect reached its plateau in 15 min (data not shown). AMPA receptors are undergoing continuous exocytosis and endocytosis, and a constant cell-surface-receptor level is maintained by balancing these two opposite trafficking processes. Modulation of either insertion or endocytosis will change the receptor abundance at the cell surface. The PKA activity-dependent increase of AMPA receptor cell-surface expression could have resulted from an increase in the rate of receptor membrane insertion, a decrease in the rate of receptor internalization, or both. By using a combination of surface biotinylation and surface stripping after receptor endocytosis, we found that forskolin treatment had little effect on constitutive AMPA receptor internalization (104.2 ± 8.3% of control) (Fig. 1 C and D). However, PKA activation reduced NMDA-induced receptor endocytosis (78.2 ± 5.5%; Fig. 1 E and F). Studies have indicated that constitutive and regulated (induced) receptor endocytosis might employ different cellular mechanisms. For instance, mutant dynamin blocks the induced endocytosis of AMPA receptors, although their constitutive trafficking is not affected (10). In addition, forskolin treatment affected only the NMDA-dependent AMPA receptor endocytosis, which suggests that PKA phosphorylation might not be critically involved in constitutive AMPA receptor endocytosis.

An external file that holds a picture, illustration, etc. Object name is zpq0070752920001.jpg

PKA activation increases AMPA receptor cell-surface expression in cultured cortical neurons. (A) Cell-surface biotinylation shows an increase in AMPA receptor surface expression (Surf) by using forskolin treatment (20 μM forskolin plus 50 μM 3-isobutyl-1-methylxanthine) (Forsk) compared with control (Con). This effect was abolished with PKA-specific inhibitor H89 (Forsk+H89). No changes were found in the total AMPA receptor protein amount (Total). (B) Densitometric quantitation of Western blots on AMPA receptor surface biotinylation. Data represent means ± SE (n = 4; ∗, P < 0.05 relative to control, t test). (C and D) Forskolin treatments show no effect on AMPA receptor constitutive internalization. Surface biotinylation at 4°C without stripping produced total surface receptors (T-surf), and stripping immediately after biotinylation showed low background (Strip and Con) (n = 3; P > 0.05, t test). (E and F) PKA inhibits NMDA-induced AMPA receptor internalization. Surface-biotinylated neurons were incubated at 37°C in the presence of 30 μM NMDA, with (NMDA+Forsk) or without (NMDA) forskolin treatment, for 15 min to induce receptor endocytosis (n = 4; ∗, P < 0.05, t test). (G and H) Surface biotinylation-based receptor insertion assays. Forskolin treatment decreased the remaining endocytosed receptor amount by using double strips (n = 2), indicating a facilitated receptor reinsertion. (I) AMPA receptor surface insertion by colorimetric assays. Surface AMPA receptors were first blocked with an anti-GluR1 N terminus antibody and a cold (nonconjugated) secondary antibody. After incubation at 37°C with (Forsk) or without (Con) forskolin treatment, newly inserted surface receptors were detected by using a second round of antibody labeling (n = 8; ∗, P < 0.05 compared with control, t test).

To investigate PKA's effects on AMPA receptor exocytosis, we first examined receptor reinsertion (recycling) into the plasma membrane after internalization. Surface-biotinylated and internalized receptors were allowed to recycle to the plasma membrane, with or without forskolin treatment. Biotinylated receptors on the cell surface were stripped after receptor internalization and stripped again after receptor recycling (see Methods). Because total constitutive receptor internalization is not altered by using forskolin, more labeled receptors remaining inside the cell (protected and nonstripped) under these conditions indicates a lower recycling rate. Following this protocol, we found that PKA treatment dramatically enhanced AMPA receptor recycling (Fig. 1 G and H). To investigate the surface insertion, rather than the reinsertion, of AMPA receptors, existing surface AMPA receptors in cortical neurons were blocked by using unlabeled antibodies, and colorimetric assays were performed after incubation at 37°C with or without forskolin treatment. As shown in Fig. 1I, PKA activation increased AMPA receptor membrane insertion by 64.8 ± 29.6%.

PKA Activity Specifically Increases GluR1 Subunit Cell-Surface Expression in Transfected HEK Cells.

Native AMPA receptors are heteromers composed of GluR1–GluR4 subunits. Studies have indicated that GluR distribution and trafficking are often regulated in a subunit-specific manner (11). To test whether PKA activation has distinct effects on different AMPA receptor subunits, we transiently transfected GluR1-GFP or GluR2-GFP AMPA receptor subunits in HEK cells. Either GluR1 or GluR2 alone is capable of surface expression, and either alone shows constitutive surface membrane insertion and internalization (11), indicating that both GluR1 and GluR2 subunits are undergoing constant trafficking. Using surface biotinylation, we found that a 15-min forskolin treatment increased cell-surface expression of GluR1-GFP (121.8 ± 5.2% of control) but not GluR2-GFP (95.4 ± 12.4% of control) (Fig. 2), indicating a specific PKA effect on GluR1.

An external file that holds a picture, illustration, etc. Object name is zpq0070752920002.jpg

PKA-dependent increase in AMPA receptor cell-surface expression is subunit-specific. (A) Surface biotinylation assays in transfected HEK cells. Forskolin treatment for 15 min increased the AMPA receptor subunit surface expression (Surf) in HEK cells expressing GluR1-GFP (Left) but not in those expressing GluR2-GFP (Right). (B) Quantitation of cell-surface expression of AMPA receptor subunits in HEK cells. Forskolin treatment (Forsk) increased GluR1-GFP surface expression (n = 3; ∗, P < 0.05, t test) (Left) but did not change the abundance of cell-surface GluR2-GFP (n = 3; P > 0.05) (Right). Con, control.

PKA Activation Increases GluR1 Surface Expression via GluR1S845 Phosphorylation.

There are many possible routes by which PKA might regulate AMPA receptor trafficking. For example, PKA might change the efficiency of trafficking machinery, regulate the association of AMPA receptors with cytosolic trafficking-related proteins, or alter vesicle or plasma membrane properties by using PKA downstream signaling. Given the specificity of PKA effects on GluR1 and the existence of a major PKA phosphorylation site at serine 845 in GluR1 C terminus (1), we explored whether GluR1S845 phosphorylation is involved in PKA-mediated AMPA receptor redistribution. In both cultured cortical neurons (Fig. 3A Left) and transfected HEK cells (Fig. 3A Right), 15 min of forskolin treatment greatly enhanced the phosphorylation level at GluR1S845, indicating that similar signaling mechanisms exist in cortical neurons and HEK cells. When GluR1-GFP and GluR1-GFP with a mutation at the PKA phosphorylation site (GluR1S845A-GFP) were transfected in HEK cells, surface staining showed that some transfected green cells lacked visible surface immunosignals (Fig. 3B). Forskolin application dramatically increased the surface-positive cells in cells expressing GluR1 (Fig. 3C Left) but not in cells expressing the GluR1S845A mutant (Fig. 3C Right). Consistently, biotinylation assays in HEK cells showed that the surface GluR1-GFP amount was increased by using PKA treatment, whereas no changes were found in HEK cells expressing GluR1S845A-GFP (124.2 ± 6.7% and 86.4 ± 11.4% of control, respectively; Fig. 3 D and E). Together, these data indicate that a direct phosphorylation at GluR1S845 is required to produce the PKA-dependent regulation of AMPA receptor cell-surface expression.

An external file that holds a picture, illustration, etc. Object name is zpq0070752920003.jpg

Phosphorylation of GluR1S845 is required in PKA activity-dependent regulation of AMPA receptor surface expression. (A) PKA phosphorylates GluR1S845 in both cortical neurons and GluR1-expressing HEK cells. Cultured cortical neurons (Left) or HEK cells transiently transfected with GluR1-GFP (Right) were treated with forskolin (Forsk) for 10 min, and GluR1S845 phosphorylation was examined by using anti-phospho-GluR1S845 (GluRI p-S845) antibodies. Compared with controls (Con), forskolin treatment greatly increased the phosphorylation level that was blocked by the PKA inhibitor H89 (Forsk+H89) (2 μM), and the PKC activator PMA (1 μM) had no effect. (B and C) PKA increases surface-GluR1 positive rate in transfected HEK cells. (B) In GluR1-GFP-expressing or GluR1S845A-GFP-expressing HEK cells, surface staining (red) revealed that a certain amount of cells had no visible surface labeling. (C) PKA treatment increased the percentage of surface-positive cells in the GluR1-expressing HEK population (Left) (n = 300 transfected cells in three experiments; ∗, P < 0.05 relative to control, t test) but not in cells expressing GluR1S845A (Right) (n = 300 transfected cells in three experiments). (D) Surface biotinylation assays showed no effect of forskolin treatment on GluR1S845A surface expression in transfected HEK cells. (E) Densitometric quantitation of surface biotinylation experiments in D. Forskolin treatment significantly increased the surface expression of GluR1-GFP (Left) (n = 5; ∗, P < 0.05, t test), but not GluR1S845A-GFP (Right) (n = 5; P > 0.05, t test).

PKA Activity Facilitates S845-Dependent GluR1 Cell-Surface Insertion.

To further investigate the role of the GluR1S845 phosphorylation site in PKA-induced AMPA receptor cell-surface insertion, we transfected cortical neurons with GluR1-GFP or GluR1S845A-GFP constructs that contained an α-bungarotoxin (Btx) binding site (BBS) in their extracellular N termini (BBS-GFP-GluR1). The BBS-tagged GluR1 targets to the plasma membrane in a manner indistinguishable from GluR1-GFP (Fig. 4A, Con) (12). Transfected neurons were first incubated with nonconjugated Btx at 17°C to block Btx binding sites on preexisting surface BBS-GFP-GluR1 (Fig. 4A, Block), and cells were then incubated at 37°C with or without forskolin treatment to induce GluR1 surface insertion. The newly inserted surface BBS-GFP-GluR1 receptors were labeled after the second round of incubation with rhodamine-conjugated Btx. The newly inserted BBS-GFP-GluR1 subunits formed clusters along dendrites and in the soma (Fig. 4B). Forskolin treatment facilitated the receptor insertion process. Quantitative analysis of the data for cluster intensity and cluster size both showed significant increases (Fig. 4 B and C). There was no increase in cluster number. In contrast, forskolin had no effect in BBS-GFP-GluR1S845A surface insertion, indicating that phosphorylation of GluR1S845 is required to produce PKA-dependent AMPAR cell-surface insertion.

An external file that holds a picture, illustration, etc. Object name is zpq0070752920004.jpg

Forskolin increases GluR1 cell-surface insertion rate in an S845-dependent manner in transfected cortical neurons. (A) Cortical neurons were transfected with GluR1 subunit (green) tagged with a BBS and GFP at its extracellular N terminus (BBS-GFP-GluR1). Rhodamine-Btx surface binding assay demonstrated that BBS-GFP-GluR1 was expressed on cell surface in clusters (Con). Immediately after incubation with free Btx, rhodamine-Btx labeling showed no signal, indicating a complete block of surface GluR1 BBS sites (Block). (B) Forskolin treatment facilitated GluR1 cell-surface insertion rate (BBS-GFP-GluR1) but had no effect on the mutant (GluR1S845A). (C) Quantitation of BBS-GFP-GluR1 plasma membrane insertion. Forskolin treatment increased both the cluster intensity (Left) and cluster size (Right) of the newly inserted surface BBS-GFP-GluR1 (n = 30; P < 0.05, t test).

Transient Dephosphorylation of GluR1S845 Associates with AMPA Receptor Internalization.

NMDA treatment causes fast and reversible dephosphorylation of GluR1S845 (9, 13, 14), and PKA activation inhibits NMDA-induced AMPA receptor internalization (Fig. 1 E and F), suggesting that dephosphorylation at GluR1S845 might be crucial in initiating NMDA-induced AMPA receptor endocytosis. If so, the dephosphorylation should occur on surface AMPA receptors and before AMPA receptor internalization. To explore this hypothesis, cultured neurons were incubated in hypertonic sucrose solution to prevent receptor internalization (10) during NMDA treatment, and surface AMPA receptors were isolated. As shown in Fig. 5 A and B, NMDA treatment caused dramatic dephosphorylation of surface GluR1S845, but sucrose alone had no effect. In addition, nearly complete GluR1S845 dephosphorylation was observed well before AMPA receptor internalization reached its plateau (Fig. 5C). We therefore reasoned that a rapid switch of GluR1S845 phosphorylation status might trigger the AMPA receptor endocytic process. If this were the case, the extent of NMDA-induced AMPA receptor internalization should correlate with the degree of surface GluR1 dephosphorylation. To test this hypothesis, neurons were first treated with forskolin for 5 min to elevate the level of GluR1S845 phosphorylation and were then treated with NMDA to induce receptor internalization. Consistent with our predictions, preincubation with forskolin increased NMDA-caused AMPA receptor internalization (135.1 ± 12.5% of NMDA control) (Fig. 5 D and E).

An external file that holds a picture, illustration, etc. Object name is zpq0070752920005.jpg

A quick switch of GluR1S845 from a phosphorylated to a nonphosphorylated state is crucial to NMDA-induced AMPA receptor internalization. (A and B) NMDA treatment induces cell-surface GluR1S845 dephosphorylation. Plasma membrane AMPA receptors were isolated by using surface biotinylation and probed by using anti-phospho-GluR1S845 (GluR1 P-S845) antibodies. When receptor internalization was blocked with hypertonic sucrose solution (Sucrose), NMDA still caused a dramatic dephosphorylation (NMDA+Sucrose) of GluR1 (n = 2). (C) GluR1 is dephosphorylated at a rate faster than AMPA receptors are internalized after NMDA application. At 10 min after NMDA treatment, a large amount of AMPA receptors remained on the surface, but almost all of the surface receptors were dephosphorylated. (D) PKA preactivation enhances NMDA-induced AMPA receptor internalization. Cells were surface-biotinylated, incubated with forskolin for 5 min, and treated with NMDA. Endocytosed receptors were collected after surface stripping. PKA pretreatment increased the NMDA-induced AMPA receptor internalization (Pre-forsk), an effect that was blocked by H89 (Pre-forsk+H89). (E) Quantification of D. NMDA-caused AMPAR internalization was increased significantly by preforskolin treatment (n = 3; ∗, P < 0.05, t test).

NMDA, but not AMPA, Fails to Induce GluR1S845A Internalization in Transfected Cortical Neurons.

If the dephosphorylation of the PKA site at GluR1S845 is crucial to GluR1 internalization, we wondered how the receptors respond to NMDA receptor activation when GluR1 is unable to be dephosphorylated. To explore this question, we transiently transfected cultured cortical neurons with GluR1-GFP and the GluR1 mutant GluR1S845A-GFP. Anti-GFP antibodies were used to label the surface receptor subunits at 4°C, and receptor internalization was induced at 37°C with or without NMDA treatment. The remaining surface receptor-bound antibodies were then stripped, and the internalized receptor–antibody complexes were visualized by using a fluorescence-conjugated secondary antibody. Surface staining of transfected neurons showed that GluR1-GFP and GluR1S845A-GFP had similarly clustered plasma membrane localization (Fig. 6A), and both GluR1-GFP and GluR1S845A-GFP showed very similar basal internalization rates (Fig. 6 B and D). However, although 30 μM NMDA treatment stimulated GluR1-GFP internalization in transfected neurons (0.58 ± 0.02 versus control 0.48 ± 0.04), internalization of GluR1S845A-GFP was not enhanced by NMDA (NMDA, 0.43 ± 0.03; control, 0.50 ± 0.04). In fact, in the majority of cells expressing GluR1S845A-GFP, the internalization was slightly inhibited by NMDA (Fig. 6 B and D). In contrast, 50 μM AMPA treatment induced GluR1S845A-GFP internalization (AMDA, 0.69 ± 0.03; control, 0.46 ± 0.03) to a level indistinguishable from that of GluR1-GFP (AMPA, 0.66 ± 0.04; control, 0.47 ± 0.04) (Fig. 6 E and F). Taken together, these data indicate that GluR1S845 dephosphorylation is required specifically for NMDA-dependent AMPA receptor internalization.

An external file that holds a picture, illustration, etc. Object name is zpq0070752920006.jpg

Preventing dynamic GluR1S845 dephosphorylation abolishes NMDA-mediated GluR1 internalization. (A) Surface expression of GluR1-GFP in cortical neurons. Cultured cortical neurons transfected with GluR1-GFP or GluR1S845A-GFP (green) were immunostained with anti-GFP antibodies (red) under nonpermeant conditions. (B) NMDA fails to induce GluR1S845-GFP internalization. Surface receptors were labeled with antibodies against GFP, and receptor endocytosis was induced by using NMDA treatment. The internalized receptors were visualized after surface acid stripping. Note that NMDA treatment enhanced internalization of wild-type GluR1-GFP but had no effect on GluR1S845A. (C) Stripping immediately after surface biotinylation showed complete removal of surface labeling. (D) Quantitation of the internalization assays in B. NMDA treatment significantly increased GluR1-GFP internalization (n = 22; ∗, P < 0.05, t test) but had no effect on the internalization of GluR1S845A-GFP (n = 24; P > 0.05, test). (E and F) In contrast, AMPA treatment (50 μM AMPA for 10 min) increased receptor internalization of both GluR1-GFP (n = 21; ∗, P < 0.05, t test) and GluR1S845A-GFP (n = 23; ∗, P < 0.05, t test).

Discussion

In this study, we found that activation of PKA induces a GluR1S845 phosphorylation-dependent increase in AMPA receptor surface expression, which results from a combination of an enhancement in AMPA receptor membrane insertion and an inhibition in AMPA receptor internalization. In addition, a rapid dephosphorylation of GluR1S845 is required for NMDA-induced AMPA receptor internalization. Consistent with our findings, one study demonstrated that activation of D1-type dopamine receptors promotes AMPA receptor insertion by using a PKA-dependent pathway in cultured nucleus accumbens neuron (15). PKA activity plays important roles in both long-term potentiation and long-term depression, which are also accompanied by changes in the PKA-dependent phosphorylation of AMPA receptor subunits (1, 4), indicating that GluR1S845 phosphorylation might be critical for the expression of various forms of synaptic plasticity. In the last several years, increasing amounts of data have indicated that the expression of long-term potentiation and long-term depression is produced through the increase or removal of synaptic AMPA receptors, respectively (1–5). To support the hypothesis that GluR1S845 phosphorylation is crucial in such receptor trafficking-mediated synaptic plasticity, AMPA receptor trafficking and GluR1S845 phosphorylation have been correlated (9). Our present findings provide a direct link between GluR1S845 phosphorylation and GluR1 trafficking, as well as between GluR1S845 dephosphorylation and NMDA-induced receptor internalization, and the findings are consistent with our recent in vivo data from mice lacking two phosphorylation sites in GluR1 (16). Furthermore, our previous work has shown that phosphorylation of GluR1S845 by using PKA increases receptor channel open probability without changing unitary conductance (17). This change might be an effect on AMPA receptor channel properties, or it might also be due to an increase in surface receptor numbers.

AMPA receptors delivered for membrane targeting can be newly synthesized protein or products of receptor recycling after endocytosis. These two distinct receptor insertion processes have not been differentially studied, and the cellular mechanisms for their regulation are largely unknown. By using colorimetric assays on endogenous AMPA receptor and immunocytochemistry on BBS-tagged GluR1 subunits, we found that PKA facilitates AMPA receptor insertion. This change could be accounted for solely by the observed enhancement in receptor reinsertion, or it could be the result of an increase in both insertion (new receptor exocytosis) and reinsertion (recycling). It is also interesting to note that, although GluR1 dephosphorylation is crucial in NMDA-induced AMPA receptor internalization, the unphosphorylatable GluR1S845 mutation failed to show enhanced basal endocytosis, indicating that the prerequisite for NMDA-induced AMPA receptor internalization is not an unphosphorylated or dephosphorylated steady state but a transient dephosphorylation of phospho-GluR1S845.

Two recent studies have suggested that GluR1S845 phosphorylation increases the extrasynaptic localization of GluR1 and that subsequent NMDA receptor-dependent signaling regulates access of GluR1 to the synapse (18, 19). In our studies we found that PKA activation increases the apparent synaptic localization of GluR1 in cultured neurons without additional treatment of the neurons. It is likely that basal NMDA receptor signaling due to spontaneous synaptic activity in our cultures was enough to permit the synaptic incorporation of extrasynaptic receptors.

How GluR1S845 phosphorylation regulates trafficking is not clear. Recent studies have shown that PKC phosphorylation of the GluR2 C terminus changes its interaction with cytosolic association partners and enhances AMPA receptor endocytosis (20). It is likely that rapid phosphorylation or dephosphorylation of GluR1S845 might similarly regulate receptor trafficking by modulating the association or dissociation of GluR1 with proteins that stabilize it in the plasma membrane or internal membranes. Phosphorylation of GluR1S845 might alter its association with AMPA receptor-interacting proteins, such as SAP-97, 4.1N, and PI3-kinase (21, 22), or with proteins involved in endocytosis or exocytosis, leading to changes in membrane trafficking. Future studies are needed to identify the critical GluR1-interacting proteins involved in this process.

Methods

Primary Cortical Neuron Cultures.

As described before (12), prenatal (embryonic day 18) rat cortical neurons were plated on 60-mm dishes (4 × 106 to 6 × 106), precoated with poly-lysine, and maintained in medium that was free of 2-amino-5-phosphonovaleric acid for 2–3 wk.

Biotinylation Assay of AMPA Receptor-Surface Expression.

High-density cultured cortical neurons (2- to 3-wk-old) were rinsed with aCSF (150 mM NaCl/3 mM KCl/2 mM CaCl2/1 mM MgCl2/10 mM Hepes/10 mM glucose, pH 7.4) and treated with 20 μM forskolin plus 50 μM 3-isobutyl-1-methylxanthine in aCSF at 37°C for 15 min or as otherwise indicated. Cells were then incubated at 10°C with 1 mg/ml sulfo-NHS-SS-biotin in aCSF for 30 min and lysed with RIPA buffer (0.15 mM NaCl/0.05 mM Tris·HCl, pH 7.2/1% Triton X-100/1% sodium deoxycholate/0.1% SDS) after three washes with aCSF. Biotinylated surface proteins were precipitated with immobilized streptavidin beads, and AMPA receptors were probed with anti-GluR C-terminal antibodies. The same procedures were followed for transfected HEK cells, except that the cells were washed with PBS and blots were probed with anti-GFP antibodies.

Biotinylation Assay of Receptor Internalization and Surface Reinsertion.

For AMPA receptor internalization assays, cortical neurons were surface-biotinylated as described above and treated with 30 μM NMDA with or without forskolin plus 3-isobutyl-1-methylxanthine for 15 min at 37°C to induce receptor endocytosis. The remaining surface-bound biotin was removed by using glutathione-stripping buffer (9). The internalized receptors that were protected from surface stripping were isolated by using streptavidin precipitation. For AMPA receptor reinsertion assays, which were performed after the first round of receptor internalization and surface stripping, neurons were transferred back to 37°C, with or without forskolin treatment, to allow the endocytosed receptors to recycle back to the cell surface. Cells were surface-stripped a second time, and the remaining intracellular biotin-bound receptors were isolated. Thus a greater amount of biotin-labeled receptors remaining inside the cell indicates slower surface reinsertion.

Immunocytochemisty of Receptor Surface Expression and Internalization.

Cultured cortical neurons or HEK cells were transfected by using GluR1-GFP or GluR1S845A-GFP with Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Two days after transfection, cells were fixed and the surface-expressed receptor subunits were labeled with anti-GFP primary antibodies and a Cy3-conjugated secondary antibody. Receptor internalization assays were performed essentially as described previously (10). In brief, cells were incubated live at 4°C with anti-GFP antibodies for 1 h and were then transferred to 37°C with or without 30 μM NMDA for 15 min to induce receptor internalization. The remaining surface-bound antibodies were stripped twice for 5 min each time by using acidic stripping buffer (0.5 M NaCl/0.2 M acetic acid) on ice, and cells were then fixed with 4% paraformaldehyde and permeabilized with 0.2% Triton X-100. The internalized receptors were then visualized after incubation by using Cy3-conjugated secondary antibodies. Immunofluorescence images were taken with a CCD camera with a ×63 objective, and the internalized receptor clusters were analyzed with a MetaMorph imaging system (Universal Imaging, Downingtown, PA).

Btx Binding Assays for GluR1 Cell-Surface Insertion in Transfected Cortical Neurons.

GluR1-GFP or GluR1S845A-GFP cDNA construct that contains a BBS in its N terminus (BBS-GluR1) was transfected by using the calcium phosphate method to DIV14 cortical neurons. The Btx binding experiments were performed 5 d after transfection as described (12). Briefly, the transfected neurons were first incubated with the nicotinic receptor antagonist tubocurarine (300 μM) in the culture medium at 37°C for 30 min to block binding of Btx to the endogenous nicotinic receptors and were then incubated with 10 μg/ml Btx at 17°C for 15 min to block the existing surface BBS-GluR1. After washing, cells were incubated with 1 μg/ml rhodamine-conjugated Btx at 37°C for 10 min to label the newly inserted receptors. The images were analyzed with MetaMorph analysis software (Universal Imaging), and the newly inserted surface GluR1 cluster intensities were normalized to the corresponding GFP signals.

Acknowledgments

We thank Lin Ding for providing cultured cortical neurons and Dr. Gareth Thomas and Julie Schafer for offering critical comments on the manuscript. This work was supported by The Howard Hughes Medical Institute. H.-Y.M. was supported by The Howard Hughes Medical Institute and the Canadian Institutes for Health Research.

Abbreviations

AMPAα-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
Btxα-bungarotoxin
BBSα-bungarotoxin binding site
GluR1glutamate receptor 1 subunit
GluR1S845GluR1 at serine 845.

Footnotes

Conflict of interest statement: Under a licensing agreement between Upstate Group, Inc., and The Johns Hopkins University, R.L.H. is entitled to a share of royalties received by the university on sales of products described in this article. R.L.H. is a paid consultant to Upstate Group, Inc. The terms of this arrangement are being managed by The Johns Hopkins University in accordance with its conflict of interest policies.

References

1. Song I, Huganir RL. Trends Neurosci. 2002;25:578–588. [PubMed] [Google Scholar]
2. Malinow R, Malenka RC. Annu Rev Neurosci. 2002;25:103–126. [PubMed] [Google Scholar]
3. Sheng M, Hyoung Lee S. Neurosci Res (NY) 2003;46:127–134. [PubMed] [Google Scholar]
4. Bredt DS, Nicoll RA. Neuron. 2003;40:361–379. [PubMed] [Google Scholar]
5. Collingridge GL, Isaac JT, Wang YT. Nat Rev Neurosci. 2004;5:952–962. [PubMed] [Google Scholar]
6. Lu W, Man H, Ju W, Trimble WS, MacDonald JF, Wang YT. Neuron. 2001;29:243–254. [PubMed] [Google Scholar]
7. Thomas GM, Huganir RL. Nat Rev Neurosci. 2004;5:173–183. [PubMed] [Google Scholar]
8. Esteban JA, Shi SH, Wilson C, Nuriya M, Huganir RL, Malinow R. Nat Neurosci. 2003;6:136–143. [PubMed] [Google Scholar]
9. Ehlers MD. Neuron. 2000;28:511–525. [PubMed] [Google Scholar]
10. Man HY, Lin JW, Ju WH, Ahmadian G, Liu L, Becker LE, Sheng M, Wang YT. Neuron. 2000;25:649–662. [PubMed] [Google Scholar]
11. Lee SH, Liu L, Wang YT, Sheng M. Neuron. 2002;36:661–674. [PubMed] [Google Scholar]
12. Sekine-Aizawa Y, Huganir RL. Proc Natl Acad Sci USA. 2004;101:17114–17119. [PMC free article] [PubMed] [Google Scholar]
13. Lee HK, Kameyama K, Huganir RL, Bear MF. Neuron. 1998;21:1151–1162. [PubMed] [Google Scholar]
14. Vanhoose AM, Clements JM, Winder DG. J Neurosci. 2006;26:1138–1145. [PMC free article] [PubMed] [Google Scholar]
15. Mangiavacchi S, Wolf ME. J Neurochem. 2004;88:1261–1271. [PubMed] [Google Scholar]
16. Lee HK, Takamiya K, Han JS, Man H, Kim CH, Rumbaugh G, Yu S, Ding L, He C, Petralia RS, et al. Cell. 2003;112:631–643. [PubMed] [Google Scholar]
17. Banke TG, Bowie D, Lee H, Huganir RL, Schousboe A, Traynelis SF. J Neurosci. 2000;20:89–102. [PMC free article] [PubMed] [Google Scholar]
18. Oh MC, Derkach VA, Guire ES, Soderling TR. J Biol Chem. 2006;281:752–758. [PubMed] [Google Scholar]
19. Gao C, Sun X, Wolf ME. J Neurochem. 2006;98:1664–1677. [PubMed] [Google Scholar]
20. Matsuda S, Mikawa S, Hirai H. J Neurochem. 1999;73:1765–1768. [PubMed] [Google Scholar]
21. Chung HJ, Xia J, Scannevin RH, Zhang X, Huganir RL. J Neurosci. 2000;20:7258–7267. [PMC free article] [PubMed] [Google Scholar]
22. Man HY, Wang Q, Lu WY, Ju W, Ahmadian G, Liu L, D'Souza S, Wong TP, Taghibiglou C, Lu J, et al. Neuron. 2003;38:611–624. [PubMed] [Google Scholar]
23. Liao D, Scannevin R, Huyanir RL. J Neurosci. 2001;21:6008–6017. [PMC free article] [PubMed] [Google Scholar]
Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
-