BDNF-dependent recycling facilitates TrkB translocation to postsynaptic density during LTP via a Rab11-dependent pathway

Abstract

Brain-derived neurotrophic factor (BDNF) plays an important role in the activity-dependent regulation of synaptic structure and function via tropomyosin related kinase B (TrkB) receptor activation. However, whether BDNF could regulate TrkB levels at synapse during long-term potentiation (LTP) is still unknown. We show in cultured rat hippocampal neurons that chemical LTP (cLTP) stimuli selectively promote endocytic recycling of BDNF-dependent full-length TrkB (TrkB-FL) receptors, but not isoform T1 (TrkB.T1) receptors, via a Rab11-dependent pathway. Moreover, neuronal-activity-enhanced TrkB-FL recycling could facilitate receptor translocation to postsynaptic density and enhance BDNF-induced extracellular signal-regulated kinase and phosphatidylinositol 3-kinase activation and rat hippocampal neuron survival. Finally, we found that cLTP could stimulate the switch of Rab11 from an inactive to an active form and that GTP-bound Rab11 could enhance the interaction between TrkB-FL and PSD-95. Therefore, the recycling endosome could serve as a reserve pool to supply TrkB-FL receptors for LTP maintenance. These findings provide a mechanistic link between Rab11-dependent endocytic recycling and TrkB modulation of synaptic plasticity.

Figure 1.

cLTP stimuli block BDNF-induced TrkB-FL receptor degradation. A, Schematic diagram of the cLTP induced by glycine. B, Example of continuous recordings from individual neurons immediately before (Basal) and 20 min after glycine (Glycine). C, Averaged responses from 6 cells (n = 80–120 events; 5 min per average; p < 0.01) are shown over a 60 min duration of recording. The amplitudes and frequencies of mEPSCs were compared. Inset, example mEPSC traces before and after glycine. D, Immunoblots of phospho-CREB1 and CREB1. Neurons were pretreated with glycine (200 μm) for 3 min and lysed 15 min after glycine was washed out. E, Left: Hippocampal neurons (9 DIV) were surfaced biotinylated and treated with BDNF (50 ng/ml) under the indicated conditions for 1 h. Surface-labeled receptors were detected by streptavidin pull-down followed by anti-TrkB immunoblotting. Right: Densitometric quantification of the results shown on the left. Data are shown as the mean ± SEM (n = 5, **p < 0.01 and ***p < 0.001 relative to the no BDNF control of each group, Student's t test). F, Left: Immunoblots of surface TrkB. Neurons (9 DIV) were treated with various agents as indicated, followed by biotinylation. Right: Densitometric quantification of the results shown on the left. Data are shown as the mean ± SEM (n = 3, *p < 0.05 and **p < 0.01 relative to the no ligand control, Student's t test) N.S. indicates not significant. G, Relative mRNA levels of TrkB-FL and TrkB.T1 in cultured hippocampal neurons (9 DIV) analyzed by real-time RT-PCR. Neurons were pretreated with glycine (200 μm) for 3 min, followed by BDNF (50 ng/ml) application for the indicated times. Data are shown as the mean ± SEM (n = 5, **p < 0.01 compared with baseline; Student's t test). H, Surface TrkB levels were detected by surface biotinylation as in F. Neurons (9 DIV) were pretreated with cycloheximide (CHX; 20 μg/ml) or rapamycin (RAP; 20 nm) to inhibit the protein translation.

Figure 2.

cLTP stimuli enhance the postendocytic recycling of TrkB-FL receptors, but not TrkB.T1 receptors. A, Left: TrkB recycling levels were measured by cleavable surface biotinylation (see Materials and Methods for details). Lane 1 refers to the total biotinylated TrkB receptors on neuronal surface; lane 2 shows a control for the efficiency of the stripping procedure; lane 3 shows the internalized biotinylated TrkB receptors before rewarming; lane 4 shows un-recycled (remaining) TrkB receptors after 30 min rewarming under basal condition; lane 5 shows un-recycled (remaining) TrkB receptors after 30 min rewarming upon cLTP stimuli. Right: Bar graphs representing means ± SEM (n = 5, ***p < 0.001, Student's t test). B, Left: Schematic of live cell ratiometric fluorescence-based recycling assay. Right: Live-cell ratiometric assay labeling of recycled TrkB under the indicated conditions (see Materials and Methods for details). The internalized TrkB is shown in green and the recycled TrkB is shown in red. Bottom: Enlarged images of framed regions, with the white arrows indicating the recycled TrkB. Scale bars, 20 μm. C, Quantitation of recycled TrkB levels in B as described in Materials and Methods. Graphs represent means ± SEM of five independent experiments (n ≥ 30 cells for each condition per experiment, ***p < 0.001, ^###p < 0.001, one-way ANOVA). D, Quantitation of recycled TrkB-FL levels treated with indicated drugs. Graphs represent means ± SEM of five independent experiments (n ≥ 30 cells for each condition per experiment, **p < 0.01, ***p < 0.001, one-way ANOVA).

Figure 3.

TrkB kinase domain is essential for cLTP-enhanced TrkB-FL recycling. A, Schematic presentation of FLAG-tagged TrkB-FL, TrkB.T1, and its mutants. ΔCT indicates that the CT region was deleted; ΔTK indicates that both the TK and CT regions were deleted; KD indicates TrkB kinase-dead mutant (K571A). B, Representative immunofluorescence images of recycled TrkB mutants in transfected hippocampal neurons (9 DIV) with or without glycine treatment. The internalized TrkB receptors are shown in green and the recycled TrkB is shown in red. Scale bars, 20 μm. C, Quantitative analysis of TrkB mutant recycling levels in hippocampal neurons expressing the indicated constructs. Graphs represent means ± SEM. of five independent experiments (n ≥ 30 cells for each condition per experiment, **p < 0.01, compared with their respective baselines, ^##p < 0.01 compared with basal TrkB-FL recycling level, Student's t test).

Figure 4.

Rab11 interacts and colocalizes with TrkB-FL. A, CO-IP was performed in lysates of HEK293 cells expressing FLAG-tagged TrkB-FL and Myc-tagged Rab11 mutants or Myc-Tagged Rab4. Cell lysates were immunoprecipitated with anti-Myc antibodies. B, FLAG-tagged TrkB mutants were coexpressed with Myc-tagged Rab11wt in HEK293 cells. Cell lysates was immunoprecipitated with anti-FLAG antibodies. C, Endogenous association of Rab11 with TrkB was detected in cultured hippocampal neurons (9 DIV) under the indicated conditions by CO-IP. A preimmune IgG (Rb) was used as a negative control. Each CO-IP experiment (A–C) was repeated three times and representative data are shown. D, Colocalization of internalized TrkB (derived by BDNF stimulation for 20 min, shown in green) with Rab11 (red) or Rab4 (red) in hippocampal neurons. Hippocampal neurons (9 DIV) transfected with FLAG tagged TrkB and MycRabs were starved overnight, fed with anti-FLAG antibodies (M1) for 30 min, treated with BDNF (50 ng/ml) for 20 min, fixed, and immunostained with anti-Myc antibody. Bottom: Higher magnification of vesicles. Scale bars, 20 μm. E, Colocalization quantification of internalized TrkB and Rabs. Data are shown as the mean ± SEM (n ≥ 30 cells for each condition per experiment, *p < 0.05, Student's t test).

Figure 5.

TrkB-FL and TrkB.T1 receptors are sorted into the Rab11- or Rab4-dependent recycling pathway, respectively. A, Representative immunofluorescence images of recycled TrkB-FL in neurons (9 DIV) expressing the indicated constructs. Scale bars, 20 μm. B, Quantitative analysis of TrkB-FL recycling in neurons expressing the indicated constructs. Graphs represent means ± SEM of three independent experiments (n ≥ 30 cells for each condition per experiment, ***p < 0.001 relative to their respective unstimulated group, ^#p < 0.05 relative to unstimulated TrkB-FL, Student's t test). C, Representative immunofluorescence images of recycled TrkB.T1 in neurons expressing the indicated constructs. Scale bars, 20 μm. D, Quantitative analysis of TrkB.T1 recycling in neurons expressing the indicated constructs. Graphs represent means ± SEM of three independent experiments (n ≥ 30 cells for each condition per experiment, ^#p < 0.05 relative to unstimulated TrkB.T1, Student's t test). E, Neurons (9 DIV) were transfected by the indicated siRNA and the TrkB recycling levels were measured by cleavable surface biotinylation 2 d after transfection. Lane 1 shows the internalized biotinylated TrkB receptors before rewarming; lane 2 shows un-recycled (remaining) TrkB receptors after 30 min rewarming under basal conditions; lane 3 shows un-recycled (remaining) TrkB receptors after 30 min rewarming upon cLTP stimuli. F, Quantitation of remaining TrkB levels in E. Bar graphs represent means ± SEM (n = 5, ***p < 0.001 relative to their respective basal control, ^#p < 0.05, Student's t test). G, Neurons transfected with the indicated siRNA (9 DIV) were stimulated with BDNF for 1 h, followed by surface biotinylation. H, I, Quantification of surface levels of TrkB-FL (D) and TrkB.T1 (E). Graphs represent means ± SEM (n = 3, **p < 0.01, ***p < 0.001 relative to their respective no BDNF treatment group, ^#p < 0.05, Student's t test).

Figure 6.

Increased TrkB-FL recycling upon cLTP stimuli enhances BDNF-induced sustained but not transient ERK and PI3K-AKT activation. A, Hippocampal neurons (7 DIV) were transfected with scrambled or siRab11 RNA as indicated. Two days later, the neurons were serum starved overnight and neurons with or without glycine (200 μm, 3 min) before stimulation were treated with BDNF (50 ng/ml) for 5 min and lysed. Total and phosphorylated protein levels were examined by immunoblotting. B–D, Quantitative analysis of TrkB (B), Akt (C), and Erk1/2 (D) activation. pTrk, pErk1/2, and pAkt levels were normalized to the phosphorylation levels of scramble group detected at the 5 min time point without glycine pretreatment. Graphs represent means ± SEM (n = 3, *p < 0.05 relative to their respective control group, Student's t test); N.S. indicates not significant. E, Hippocampal neurons (7 DIV) were transfected with scrambled or siRab11 RNA as indicated. Two days later, the neurons were serum starved overnight and neurons with or without glycine (200 μm, 3 min) before stimulation were treated with BDNF (50 ng/ml) for 60 min and lysed. Total and phosphorylated protein levels were examined by immunoblotting. F–H, Quantitative analysis of TrkB (F), Akt (G), and Erk1/2 (H) activation. pTrk, pErk1/2, and pAkt levels were normalized to the phosphorylation levels of scramble group detected at the 60 min time point without glycine pretreatment. Graphs represent means ± SEM (n = 3, **p < 0.01, ***p < 0.001 relative to their respective control group, ^#p < 0.05, Student's t test).

Figure 7.

cLTP-enhanced recycling facilitates TrkB-FL translocation into postsynaptic densities. A, Hippocampal neurons (18 DIV) were transfected with GFP and treated with various agents as indicated. Endogenous TrkB-FL stained by the Trk (C-14) antibody that recognizes the C terminus of Trk (catalog# sc-11; Santa Cruz Biotechnology) was labeled red. Arrowheads and arrows denote protrusions with and without TrkB in the spine head, respectively. Scale bars, 1 μm. B, Three types of TrkB-FL distribution in spines. Scale bars, 2 μm. C, Proportions of spines containing TrkB-FL in the indicated locations. Graphs represent means ± SEM (n = 116, 169, and 135 spines analyzed from 8, 8, and 10 neurons, respectively, *p < 0.05 relative to no BDNF control of each group, ^#p < 0.05, χ² test. D, Coimmunofluorescence labeling of recycled TrkB and PSD-95 in neurons (9 DIV) transfected with pcDNA3.1 or pCAGIG-GFP-Rab11S25N. Hippocampal neurons were stained for recycled TrkB (green) and PSD-95 (red) with (left) or without (right) glycine stimulation. Blue indicates neurons expressing pcAGIG-GFP-Rab11S25N. Colocalization is indicated by white arrows. Scale bars, 10 μm. E, Quantification analysis of the colocalization of recycled TrkB-FL or TrkB.T1 with PSD-95 in D. Graphs represent means ± SEM (n ≥ 30 cells for each condition per experiment, ***p < 0.001 relative to its baseline, ^#p < 0.05, Student's t test). F, Quantification analysis of the clusters of PSD95. Images were taken using a 100× objective and two-dimensional deconvolution software (MetaMorph) was used.

Figure 8.

Rab11 modulates the interaction between TrkB-FL and PSD-95. A, Myc-tagged Rab11 mutants or Rab11 siRNA were transfected with FLAG-TrkB-FL and GFP-PSD-95 in HEK293 cells. Anti-FLAG immunoprecipitates were analyzed by immunoblotting with PSD-95 antibody. B, FLAG-TrkB-FL, GFP-PSD-95, and Myc-Rab11WT constructs were cotransfected into HEK293 cells as indicated. Proteins were then incubated in the presence of nonhydrolyzable GTP or GDP and immunoprecipitated with FLAG antibody. Then samples were analyzed by immunoblotting with indicated antibodies. C, CO-IP of different Rab11 mutants with PSD-95 as indicated. Cell lysate was immunoprecipitated with anti-Myc antibody and analyzed by immunoblotting with anti-PSD-95 antibody. D, Hippocampal organotypic slices (P6 + 9 DIV, 400 μm) were stimulated with BDNF (50 ng/ml) for 1 h with or without glycine pretreatment and then slice lysate was immunoprecipitated with anti-TrkB antibody and analyzed by immunoblotting with anti-TrkB and PSD-95 antibodies. Nonimmune IgG was used as a negative control. E, Cultured hippocampal neurons were treated with glycine for 3 min and then washed out. One hour later, the neurons were homogenized and cytosol (Cyto) and membrane (Memb) fractions were extracted. The lysate was analyzed by immunoblotting with anti-Rab11 and Rab4 antibodies. Tubulin and calnexin were used as markers of cytosol and membrane protein, respectively. F, Immobilized GST-Sec15c (lanes 2 and 3) or control GST (lane 1) was incubated with cultured neurons extract undergoing glycine pretreatment or not. The beads were subsequently washed and bound proteins were analyzed by immunoblotting using anti-Rab11 antibody. GST fusion proteins were detected by Coomassie blue staining (bottom). All experiments were repeated a minimum of three times and representative data are shown.

Figure 9.

Neuronal-activity-dependent TrkB-FL recycling is involved in BDNF-induced neuronal survival. Hippocampal neurons (7 DIV) were transfected with the indicated siRNA. Two days later, the neurons were starved of B27 for 2 d. Drugs (50 ng/ml BDNF and/or 10 μm bicuculline) were applied to the cells as soon as B27 was removed. A, Immunoblotting analysis of cleaved caspase-3, Rab11, and tubulin. B, Summary of densitometric analysis of cleaved caspase-3. The levels of cleaved caspase-3 were normalized to scramble-basal group. Data are presented as the mean ± SEM (n = 3, **p < 0.01, ***p < 0.001 relative to the basal control of each group, ^#p < 0.05, one-way ANOVA). N.S. indicates not significant. C, Apoptosis was determined by TUNEL staining (green dots) 48 h after B27 deprivation and neurons were doubly stained with DAPI (blue dots). D, The percentage of cell death was quantified by dividing the number of apoptotic nuclei to a population of 2000 counted cells per condition. Graphs represent means ± SEM (n = 3, **p < 0.01, ***p < 0.001 relative to the basal control of each group, ^#p < 0.05, one-way ANOVA). N.S. indicates not significant.