Acute BDNF treatment upregulates GluR1-SAP97 and GluR2-GRIP1 interactions: implications for sustained AMPA receptor expression

Abstract

Brain-derived neurotrophic factor (BDNF) plays several prominent roles in synaptic plasticity and in learning and memory formation. Reduced BDNF levels and altered BDNF signaling have been reported in several brain diseases and behavioral disorders, which also exhibit reduced levels of AMPAr subunits. BDNF treatment acutely regulates AMPA receptor expression and function, including synaptic AMPAr subunit trafficking, and implicates several well defined signaling molecules that are required to elicit long term potentiation and depression (LTP and LTD, respectively). Long term encoding of synaptic events, as in long term memory formation, requires AMPAr stabilization and maintenance. However, factors regulating AMPAr stabilization in neuronal cell membranes and synaptic sites are not well characterized. In this study, we examine the effects of acute BDNF treatment on levels of AMPAr-associated scaffolding proteins and on AMPAr subunit-scaffolding protein interactions. We also examine the effects of BDNF-dependent enhanced interactions between AMPAr subunits with their specific scaffolding proteins on the accumulation of both types of proteins. Our results show that acute BDNF treatment upregulates the interactions between AMPAr subunits (GluR1 and GluR2) with their scaffold proteins SAP97 and GRIP1, respectively, leading to prolonged increased accumulation of both categories of proteins, albeit with distinct mechanisms for GluR1 and GluR2. Our findings reveal a new role for BDNF in the long term maintenance of AMPA receptor subunits and associated scaffolding proteins at synapses and further support the role of BDNF as a key regulator of synaptic consolidation. These results have potential implications for recent findings implicating BDNF and AMPAr subunits in various brain diseases and behavioral disorders.

Figure 1. Prolonged BDNF treatment of cultured neocortical neurons enhances GluR1- and SAP97-like immunoreactivities and increased AMPAr subunit expression on neuronal cell membrane.

Cultured neurons are treated with BDNF (50 ng/ml), or not (control) for 4–5 days. At the end of treatment, neurons are lysed and probed with anti-GluR1 and anti-GluR2 using total cell lysates (A) and cytoplasmic and membrane fractions (B). Results were obtained from 8 experiments (A, B) *: p<0.05; **: p<0.005. C) Similarly treated cultures, (BDNF (50 ng/ml), or not (control) for 4–5 days, were stained with the indicated antibodies: Upper panels, anti-N-terminal domain of GluR1; middle panels, anti-C-terminal domain of GluR1; lower panels, anti-SAP97.

Figure 2. Acute BDNF treatment increases GluR1-SAP97 and GluR2-GRIP1 interactions in neurons and enhances GluR1 and GluR2 expression on neuronal cell surface.

DIV 14 cultured neurons acutely-treated with BDNF (1 h, 50 ng/ml), or not, are lysed and processed for co-immunoprecipitation using anti-SAP97 (A) and anti-GRIP1 (B). Immunoblots are probed with anti-GluR1 and anti-GluR2 (upper blots in A and B, respectively). Re-probing the same blots with anti-SAP97 and anti-GRIP1 indicates equal loading (lower blots in A and B, respectively). C) Acute BDNF treatment (1 h, 50 ng/ml) does not increase total GluR1 and GluR2 protein levels in similarly treated cultures. D) Similarly treated cultures, as in (C), are treated with biotinylation reagent and biotinylated total surface proteins are analyzed with anti-GluR1 and anti-GluR2 antibodies. Results are representative of 4 experiments; *: p<0.01; Student’s t-test.

Figure 3. GluR1 knock-down in cultured cortical neurons.

A) GluR1-specific dsRNA reduces GluR1-like but not GluR2-like immunostaining in transduced neurons. B) Immunoblots show specific decline of GluR1 immunoreactivity in cultured neurons transduced with GluR1-specifc dsRNA; GluR2/3 levels are slightly reduced but reduction is not statistically significant. SAP97 protein levels show a highly significant decline following GluR1-specific dsRNA transduction. C) Quantification of 6 experiments carried out as in B. *p<0.05, **p<0.01; Student’s t-test.

Figure 4. Over-expression of GluR1 and GluR2 C-terminal decoy peptides eliminates BDNF-mediated upregulation of AMPAr subunits and their scaffolding proteins.

DIV 10 cultured cortical neurons are infected with a control Sindbis virus to express EGFP alone or with viruses encoding EGFP-R1 and EGFP-R2 and treated with BDNF for 1 day or not. Protein samples are extracted and analyzed by immunoblotting as indicated in “Materials and Methods”. Over-expression of the decoy peptides, corresponding to the C-terminal domains of GluR1 and GluR2, results in specific elimination of the up-regulatory effects of BDNF on the corresponding AMPAr subunit and its interacting partner PDZ protein(s).

Figure 5. Effects of prolonged BDNF treatment on heterologous proteins expressed in HEKTrkB cells.

A) BDNF treatment leads to TrkB phosphorylation (p-TrkB) in HEKTrkB cells. Cells are treated with BDNF (50 ng/ml) for 1 h after which culture medium is replaced with fresh medium. Cells are incubated for the indicated times before being lysed and used for analysis with TrkB and p-TrkB antibodies. B) HEKTrkB cells are transfected overnight with pCMV-EGFP (31). Cells were re-suspended and split into two equal numbers before being re-plated and treated with BDNF (50 ng/mL, daily for 3 days). Similar levels of EGFP fluorescence is indicative of similar expression levels in BDNF-treated or untreated HEKTrkB cells. C) Cells are treated as in (B); protein samples are collected from cells treated with BDNF for 1, 2, or 3 days or not and analyzed by immunoblotting for EGFP, GluR2 and GRIP1 protein contents. The results show that chronic BDNF treatment in HEKTrkB cells does not increase the expression of any of the heterologous proteins. D) Cells are transfected as in (C); total RNA is extracted from cells treated with BDNF or vehicle (control) for 1, 2, or 3 days and analyzed by RT-PCR. The results indicate that similar levels of exogenously-expressed mRNA (GluR2 and GRIP1) are present in BDNF-treated and untreated HEKTrkB cells. Amplification of endogenously-expressed Actin mRNA serves as an internal control.

Figure 6. BDNF treatment of HEKTrkB cells increases GluR2-GRIP1 interaction and upregulates their total protein levels.

A) Upper panel, immunoprecipitation is done with anti-N-terminal GluR2 antibody and immunoblotting with anti-GRIP1; lower panel, immunoprecipitation is done with anti-GRIP1 and immunoblotting with anti-C-terminal GluR2/3 antibody. The results indicate increased interaction between GluR2 and GRIP1 following 1 h BDNF treatment (50 ng/mL) of HEKTrkB cells cotransfected with GluR2 and GRIP1 expression vectors. B) HEKTrkB cells are transfected with WT GluR2 and GRIP1. One day after transfection, cells are treated with BDNF or not for 1, 2, or 3 days. Protein samples are analyzed for the total protein content of GluR2 and GRIP1 as well as for phospho-GluR2 (p-GluR2-Ser880). C) Western blotting of proteins samples from cultured neurons treated with BDNF or not were run in parallel with samples extracted from HEKTrkB cells that were transfected with wild type (WT) or deletion mutant of GluR2 lacking the last 5, 10, or 21 aminoacids (Δ5, Δ10, and Δ21, respectively). The C-terminal antibody against GluR2/3 only labeled GluR2 WT. The N-terminal anti-GluR2 labeled all constructs. D) HEKTrkB cells were transfected with GluR2Δ5 together with GRIP1. The results indicate gradual accumulation of WT GluR2 but not GluR2Δ5 in samples treated with BDNF.