Activation of glutamate receptors promotes Ca_V1.2 internalization and degradation. (A) Cortical neurons expressing YFP-HA-Ca_V1.2 channels were treated with 50 µM glutamate (Glu) for 10 min and stained with anti-HA antibodies without permeabilization. The top six panels show staining in neuronal cell bodies; to distinguish changes in staining intensity more easily, the insets show the same images using a rainbow color scheme. The bottom six panels show staining in dendrites. (B) Quantification of data from A (n = 30; mean ± SEM; *, P < 0.0001 by Student's t test). (C) Ratiometric measurement of intracellular Ca²⁺ in neurons incubated in the absence or presence of 50 µM glutamate for 10 min and depolarized with 67 mM KCl (n = 50; mean ± SEM). The panel on the left shows the mean Ca²⁺ responses, and the inset graphs show individual traces. (D) A 20-min recovery period after glutamate stimulation for 10 min did not lead to recovery of YFP-HA-Ca_V1.2 surface levels (n = 30; mean ± SEM; *, P < 0.0001 by Student's t test). (E) Cortical neurons were incubated with 50 µM glutamate for 15 min and analyzed by Western blotting with anti-Ca_V1.2 and anti-GAPDH antibodies (n = 3). Molecular mass is indicated in kilodaltons. (F) Cortical neurons were stimulated with glutamate for 10 min and then stained with anti-Ca_V1.2 and MAP2 antibodies to show changes in endogenous Ca_V1.2 in dendrites after stimulation. (G) Ratio of Ca_V1.2 to MAP2 fluorescence on neuronal dendrites treated as in F (n = 200 dendrites; mean ± SEM; *, P < 0.0001 by Student's t test). (H) Endogenous Ca_V1.2 surface levels 20 min after glutamate stimulation (n = 30; mean ± SEM; *, P < 0.0001 by Student's t test). Bars: (A, top) 20 µm; (A, bottom) 3 µm; (F) 5 µm.

NMDA receptors mediate Ca_V1.2 degradation in a Ca²⁺-dependent manner. (A) Ratiometric measurement of [Ca²⁺]_i in neurons treated with glutamate (Glu), 67 mM KCl, or 2 µg/ml ionomycin (n = 100; mean ± SEM). (B) Neurons were stimulated with 50 µM glutamate or 67 mM KCl for 10 min in the absence or presence of 2.5 mM EGTA. The graph shows the percent decrease in the ratio of Ca_V1.2 to MAP2 fluorescence on dendrites (n = 200 dendrites; mean ± SEM; *, P < 0.01 by one-way analysis of variance [ANOVA] with the Newman-Keuls multiple comparison test). (C) Change in the surface expression of Ca_V1.2 in dendrites of neurons treated with 50 µM glutamate for 10 min in the presence of 10 µM MK801 or 10 µM NBQX (n = 200 dendrites; mean ± SEM; *, P < 0.01 by one-way ANOVA). (D) Change in the surface expression of Ca_V1.2 in dendrites of neurons treated with 50 µM glutamate for 10 min in the presence of 300 µM DL-AP3 (n = 200 dendrites; mean ± SEM; NS by Student's t test).

Glutamate-induced Ca_V1.2 degradation is mediated by lysosomes. (A) Neurons were incubated with 100 µM chloroquine or 100 nM bafilomycin for 3 h or with 2 µM lactacystin for 1 h and then stimulated with 50 µM glutamate for 10 min. The graph shows the percent decrease in the ratio of Ca_V1.2 to MAP2 fluorescence on dendrites (n = 200 dendrites; mean ± SEM; *, P < 0.001 by one-way ANOVA). (B) Cortical neurons (3 d in vitro) were treated with or without 30 mM NH₄Cl for 3 h and then analyzed by immunocytochemistry with anti-Ca_V1.2 antibodies. Representative images are shown. (C) Cortical neurons were treated with or without 30 mM NH₄Cl for 30 min and then incubated with 50 µM glutamate for 15 min and subjected to immunoblot analysis with anti-Ca_V1.2 and anti-GAPDH antibodies. Molecular mass is indicated in kilodaltons. (D) Fluorescent images of neurons expressing Flag-Myc-Ca_V1.2 and LAMP-GFP. Cells were stained with anti-Myc and anti-GFP antibodies before and after treatment with 50 µM glutamate (Glu) for 15 min (n = 5 neurons and 23 dendrites). The arrows show colocalization of Ca_V1.2 and the lysosomal marker LAMP-GFP. Bars: (B) 30 µm; (D) 2.5 µm.

PIKfyve associates with Ca_V1.2. (A and B) Coimmunoprecipitation of full-length Myc-PIKfyve and of the Flag-tagged C terminus of Ca_V1.2 (Flag–Ca_V1.2 CT) coexpressed in HEK 293T cells with either anti-Flag (A) or anti-Myc (B) antibodies. (C) Coimmunoprecipitation of full-length Flag-Ca_V1.2 and Myc-PIKfyve in HEK 293T cells. Flag-Ca_V1.2 was immunoprecipitated with anti-Flag antibodies, and the immunoprecipitated material was analyzed using both Myc and Flag antibodies. (D) Schematic structure of deletion mutants of PIKfyve. (E and F) Immunoprecipitation (IP) of the deletion mutants of PIKfyve and of the C terminus of Ca_V1.2 expressed in HEK 293T cells. Cell lysates were subjected to immunoprecipitation with either anti-Myc (E) or anti-Flag (F) antibodies, and Western blots were analyzed to detect interacting proteins. (G) Schematic structure of deletion mutants of the C terminus of Ca_V1.2. (H) Coimmunoprecipitation of the Flag-Myc-Ca_V1.2 C terminus deletion mutants (CT-1 through CT-5) with full-length PIKfyve (HA-PIKfyve) expressed in HEK 293T cells using anti-Flag antibodies. Western blots were performed using both anti-Flag and anti-HA antibodies. (I) Coexpression of Ca_V1.2 CT-2 and CT-3 caused aggregation of GFP-PIKfyve in a structure close to the nucleus. Arrowheads show PIKfyve aggregates that are formed when PIKfyve is coexpressed with C-terminal fragments of Ca_V1.2. (J) Coimmunoprecipitation of endogenous Ca_V1.2 and PIKfyve from neurons before or after treatment with 50 µM glutamate for 5 min (n = 3). (A–C, E, F, H, and J) Molecular mass is indicated in kilodaltons. Ab, antibody; IB, immunoblot. Bar, 50 µm.

Activation of glutamate receptors increases PtdIns(3,5)P₂ levels. (A) Structure of myotubularin and of the GFP–2× GRAM reporter protein. (B) Fluorescent images expressing either GFP alone or the GFP–2× GRAM reporter before and after treatment with 50 µM glutamate (Glu) for 10 min. (C) Changes in perinuclear fluorescence observed in cells expressing GFP or GFP–2× GRAM after treatment with glutamate (n = 60 cells; 300 ROIs; mean ± SEM; *, P < 0.0001 by Student's t test). The areas targeted to quantify perinuclear localization are indicated in the illustration. The pink line indicates the edge of the cell, the dashed line indicates the nucleus, and the small yellow circles indicate perinuclear regions. (D) Change in perinuclear GFP–2× GRAM fluorescence before and after various types of stimulation, as indicated in Fig. 2 A (n = 60 cells; 300 ROIs; mean ± SEM; *, P < 0.001 by one-way ANOVA). (E) Cortical neurons stained with anti-PtdIns(3,5)P₂ antibodies before and after treatment with 50 µM glutamate for 10 min. (F) Quantification of the data in D showing an increase in PtdIns(3,5)P₂ levels in perinuclear regions (n = 100; mean ± SEM; *, P < 0.0001 by Student's t test). (G) Change in perinuclear GFP–2× GRAM fluorescence in neurons containing either shPIKfyve or shScrambled along with either a control vector or PIKfyve WTres or PIKfyve KEres after treatment with 50 µM glutamate (n = 60 cells; 300 ROIs; mean ± SEM; *, P < 0.01 by one-way ANOVA). Bars, 10 µm.

PIKfyve regulates glutamate-dependent Ca_V1.2 internalization and degradation. (A) Cortical neurons expressing YFP-HA-Ca_V1.2 with shPIKfyve or shScrambled were treated with 50 µM glutamate (Glu) for 10 min and stained with anti-HA antibodies without membrane permeabilization. The insets show the same images using a rainbow scheme. (B) Graph of the data in A illustrating the cell surface expression levels of Ca_V1.2 (n = 30; mean ± SEM; *, P < 0.0001 by Student's t test). (C) Dendrites from neurons stained with anti-Ca_V1.2 and anti-MAP2 antibodies and expressing either shPIKfyve or shScrambled, along with either a control vector or PIKfyve WTres or PIKfyve KEres. Neurons were stimulated with 50 µM glutamate for 10 min and then stained with anti-Ca_V1.2 or MAP2 antibodies. (D) Percent decrease of the ratio of Ca_V1.2 to MAP2 fluorescence on dendrites for the experiments in C (n = 100 dendrites; mean ± SEM; *, P < 0.01 by one-way ANOVA). (E) Dendrites from neurons expressing Flag-Myc-Ca_V1.2 or Flag-Myc–Ca_V1.2 ΔCT together with GFP, stimulated with glutamate, and immunostained with anti-Myc and anti-GFP antibodies. (F) The graph shows the percent decrease in the ratio of Ca_V1.2 to MAP2 fluorescence in dendrites in E (n = 100 dendrites; mean ± SEM; *, P < 0.005 by Student's t test). (G) Dendrites from neurons expressing Flag-Myc-Ca_V1.2 and LAMP-GFP in the presence or absence of shPIKfyve or shScrambled. Cells were stained with anti-Myc and anti-GFP antibodies before and after treatment with 50 µM glutamate for 15 min. (n = 5 neurons and 16–23 dendrites). The arrow shows Ca_V1.2 colocalization with the lysosomal marker LAMP-GFP. Bars: (A) 10 µm; (C, E, and G) 5 µm.

PIKfyve protects neurons against excitotoxic cell death. (A) Cortical neurons incubated with or without 10 µM MK801, 10 µM NBQX, 10 µM nimodipine (Nim), 0.5 µM agatoxin IVA (AgaIVA), 1 µM ω-conotoxin GVIA (ωCtx), or 10 µM SNX482 (SNX) and stimulated with 50 µM glutamate (Glu) for 30 min and then incubated for 9.5 h before analysis. Arrows indicate typical pyknotic cells. (B) Pyknotic nuclei quantified for each condition in A (n = 3 and >500 cells/sample; mean ± SEM; *, P < 0.001; and **, P < 0.05 by one-way ANOVA). (C) Cortical neurons were stimulated with 50 µM glutamate for 10 min and then incubated for 10 h in the absence of glutamate. Representative images of neurons expressing DsRed and either shPIKfyve or shScrambled along with the PIKfyve WTres, PIKfyve KEres, or the corresponding empty vector are shown. Cells were stained with the nuclear dye Hoechst. (D) Cell death was measured by counting pyknotic nuclei (n = 3 and >50 cells/sample; mean ± SEM; *, P < 0.001 by one-way ANOVA). (E) Cortical neurons were stimulated with 50 µM glutamate for 10 min in the presence or absence of 10 µM nimodipine and then incubated for 10 h. Cell death was measured by counting pyknotic nuclei (n = 3 and >50 cells/sample; mean ± SEM; *, P < 0.001 by one-way ANOVA). (F) Cortical neurons treated as in E. When indicated, NBQX was added at a 10-µM concentration (mean ± SEM; *, P < 0.001 by one-way ANOVA). Bars: (A) 30 µm; (C) 10 µm.

Model describing the process of Ca_V1.2 degradation in response to glutamate. Glutamate stimulation and high levels of intracellular Ca²⁺ lead to channel degradation in a PIKfyve-dependent manner, which protects the cell from Ca²⁺ toxicity. Stronger stimulation overrides this protective mechanism, and neurons undergo excitotoxicity.