doi: 10.1523/JNEUROSCI.4538-13.2014.
Phosphorylation of Ser1166 on GluN2B by PKA is critical to synaptic NMDA receptor function and Ca2+ signaling in spines
Affiliations
- PMID:24431445
- PMCID: PMC3891964
- DOI: 10.1523/JNEUROSCI.4538-13.2014
Item in Clipboard
Phosphorylation of Ser1166 on GluN2B by PKA is critical to synaptic NMDA receptor function and Ca2+ signaling in spines
Jessica A Murphy et al. J Neurosci..
Display options
Format
Display options
Format
Abstract
The NMDA-type glutamate receptor (NMDAR) is essential for synaptogenesis, synaptic plasticity, and higher cognitive function. Emerging evidence indicates that NMDAR Ca(2+) permeability is under the control of cAMP/protein kinase A (PKA) signaling. Whereas the functional impact of PKA on NMDAR-dependent Ca(2+) signaling is well established, the molecular target remains unknown. Here we identify serine residue 1166 (Ser1166) in the carboxy-terminal tail of the NMDAR subunit GluN2B to be a direct molecular and functional target of PKA phosphorylation critical to NMDAR-dependent Ca(2+) permeation and Ca(2+) signaling in spines. Activation of β-adrenergic and D1/D5-dopamine receptors induces Ser1166 phosphorylation. Loss of this single phosphorylation site abolishes PKA-dependent potentiation of NMDAR Ca(2+) permeation, synaptic currents, and Ca(2+) rises in dendritic spines. We further show that adverse experience in the form of forced swim, but not exposure to fox urine, elicits striking phosphorylation of Ser1166 in vivo, indicating differential impact of different forms of stress. Our data identify a novel molecular and functional target of PKA essential to NMDAR-mediated Ca(2+) signaling at synapses and regulated by the emotional response to stress.
Figures
Ser1166 in GluN2B is a direct target of PKA phosphorylation critical to PKA potentiation.a, HEK293 cells expressing WT or mutant GluN2B subunits were immunoprecipitated and incubated with PKA-C in the presence of γ-32P-ATP. Top, Autoradiogram depicting incorporation of γ-32P into GluN2B. Bottom, Western blot probed with an antibody to GluN2B. Right, Summary data. Values for mutant receptors were normalized to values for WT receptors.b, PKA potentiation of NMDA-evoked currents is abolished in HEK293 cells expressing GluN1 with GluN2B(S1166A); bar above sample records represents duration of NMDA+glycine application (N). Left, Sample traces at 0 (1) and 7 (2) min. Right, Summary time course data of the peak NMDAR current, defined as the initial maximum current. Potentiation is defined as the ratio of the peak current recorded from the same cell at 7 min versus 0 min. Values for WT and mutant receptors in the presence (PKA-C) and absence (Control) of PKA-C were normalized to values for WT receptors in the absence of PKA-C (control). WT+PKA-C,n = 9 cells; S1166A+PKA-C,n = 11 cells; WT control,n = 7 cells; S1166A control,n = 6 cells from 4 to 9 independent experiments.c, PKA potentiation of NMDA-evoked currents is abolished in hippocampal neurons expressing GluN2B(S1166A). Left, Representative traces. Right, Summary time course data as inb. WT+PKA-C,n = 5 cells; S1166A+PKA-C,n = 8 cells; WT control,n = 5 cells; S1166A control,n = 5 cells from 3 or 4 independent experiments.d, Far left, Representative traces for individual GluN1/GluN2B(S1166A) (left), WT (center), and S1166D (right) channels. Low gating activity of S1166A relative to that of WT is restored with the phosphomimetic mutation, S1166D. Single-channel activity was recorded with cell-attached pipettes as inward currents through one S1166A, WT, or S1166D channel expressed in a HEK293 cell. Far right, Summary of single-channel parameters for S1166A (gray,n = 14), WT (black,n = 19), and S1166D (white,n = 16). The apparent differences in Po and closed time values for either the S1166A or S1166D mutant versus WT were not significant, consistent with variability in the basal phosphorylation state of WT GluN1/2B receptors. *p < 0.05. **p < 0.01. ****p < 0.0001. Here and in all figures, error bars indicate SEM.
Ser1166 is critical to modulation of NMDAR Ca2+ permeability and NMDA-evoked Ca2+ rises by PKA. Representative traces (a) and current–voltage relationship (b) for NMDA-elicited currents recorded from HEK293 cells expressing GluN1 with WT and mutant GluN2B(S1166A) in the absence and presence of H-89. H-89 induced a negative shift in reversal potential in cells expressing WT (ΔErev = −6.79 mV), but not GluN2B(S1166A) (ΔErev = 0.46 mV).c, Summary data depicting reversal potential (Erev); WT,n = 6 cells; S1166A,n = 6 cells from 3 independent experiments. *p = 0.0207; n.s., Not significant.d–f, PKA potentiation of NMDA-evoked rises in Ca2+ is diminished in HEK293 cells expressing GluN1 with GluN2B(S1166A).d, Representative images of basal and NMDA-evoked Ca2+ rises before (control), during (H-89), and after washout of H-89 (recovery) for cells expressing GluN1 with WT or S1166A receptors.e, Time course data of cells ind.f, Left, Summary data for NMDA-evoked Ca2+ rises. Right, Summary data expressed as percentage inhibition by H-89; WT,n = 7 cells; S1166A,n = 5 cells from 5 independent experiments. *p = 0.0165. **p < 0.01. Scale bar, 30 μm.
Ser1166 of GluN2B is essential to β-adrenergic-induced Ca2+ rises in spines.a, Representative 2PLSM image of a dendritic spine from a CA1 hippocampal neuron filled with the red fluorophore AlexaFluor-594 and the green Ca2+-sensitive indicator Fluo-5F. The yellow arrowhead indicates the location of a 500 μs pulse of 720 nm laser light used to focally uncage glutamate near the spine head. White dashed line indicates the region used for line scan analysis of fluorescence. Scale bar, 1 μm.b, Red and green fluorescence in the spine head (sp) and neighboring dendrite (dend) measured in a line scan over the region indicated by the dashed line ina. Inset scale bars, 100 ms and 5% ΔG/Gsat. Yellow arrowhead indicates time of uncaging. The increase in green fluorescence indicates increased intracellular Ca2+, and the corresponding spine Ca2+ transient (white trace) is below.c, Average trace of uncaging-evoked ΔCa2+spine produced by single uncaging stimuli in neurons expressing GluN2B(WT) (left) or GluN2B(S1166A) (right) under control conditions (black) or after isoproterenol incubation (red). Mean (solid lines) and mean ± SEM (shaded regions) are shown.d, Average trace of uEPSCs produced by single uncaging stimuli in neurons expressing GluN2B(WT) (left) or GluN2B(S1166A) (right) under control conditions (black) or after isoproterenol incubation (red).e, Population data for peak ΔCa2+spine and (f) uEPSC elicited by single uncaging stimuli in neurons expressing GluN2B(WT) or GluN2B(S1166A) under control conditions (Ctrl) or after isoproterenol incubation. Red bars represent the mean of distribution ± SEM. *p < 0.001.
Physiologically relevant stimuli that activate cAMP/PKA signaling promote Ser1166 phosphorylation.a, PKA-C promotes phosphorylation of GluN2B at Ser1166 after immunoprecipitation from brain lysates in the presence or absence of phosphatase (PP) inhibitors as indicated (+/−). Top, GluN2B immunocomplexes were incubated with no enzyme (control), PKA-C, or alkaline phosphatase (AP). Membranes were probed with antibodies to pSer1166-GluN2B in the absence or presence of phosphopeptide and total GluN2B to detect IP levels. Middle, Input samples (lysates). Bottom, Summary data showing pSer1166-GluN2B normalized to total GluN2B and then to control (no PKA-C);n = 3. *p = 0.0138 for control versus PKA-C.b,c, β-Adrenergic and D1/D5 dopamine receptor agonists promote phosphorylation of Ser1166 in forebrain slices.b, Top, Slices treated with vehicle or isoproterenol in the absence or presence of the β-adrenergic antagonist propranolol. Membranes were probed with antibodies to pSer1166-GluN2B, pSer897-GluN1, and pSer845-GluA1 to assess phosphorylation or to GluN2B, GluN1, or GluA1 to determine total subunit levels. Bottom, Quantification of phosphorylation levels of Ser1166, Ser897, and Ser845 normalized to total IP levels of GluN2B, GluN1, and GluA1, respectively, and then normalized to vehicle.c, Slices were treated with vehicle, the D1/D5 dopamine receptor agonist SKF81297, or with SKF81297 in the presence of the D1/D5 dopamine receptor antagonist SCH23390. GluN2B and GluA1 were immunoprecipitated and processed for immunoblotting as above. Images are from the same blot and exposure, but non-relevant lanes were removed. Bottom, Summary data as inb. *p < 0.05. ***p < 0.001. ****p < 0.0001.
Phosphorylation of GluN2B at Ser1166 is induced by forced swim, but not exposure to fox urine,in vivo.a, Representative Western blots (top) and summary data (bottom) for hippocampal lysates from rats exposed to fox urine for 5 min. Western blots were probed with antibodies to pSer1166-GluN2B, pSer897-GluN1, and pSer845-GluA1 and then with antibodies to total GluN2B, GluN1, and GluA1. Exposure to fox urine increased phosphorylation of GluA1 at Ser845, with little or no effect on phosphorylation levels of GluN2B at Ser1166 or GluN1 at Ser897.n = 4 animals per treatment group, **p = 0.0030.b, Representative Western blots (top) and summary data (bottom) for hippocampal lysates from rats exposed to a single episode of forced swim for 5 min. Western blots were probed with antibodies as above for exposure to fox urine. Forced swim induced striking phosphorylation of GluN2B at Ser1166, GluN1 at Ser897, and GluA1 at Ser845. Values for phosphorylated protein were normalized to the corresponding value for total protein and then to the corresponding value for control animals.n = 5 animals per treatment group. ***p = 0.0006 (S1166 control vs FS). ***p = 0.0002 (S845 control vs FS). ****p < 0.0001.
Forced swim (FS)-induced phosphorylation of Ser1166 is PKA-dependent.a, Schematic of experimental paradigm for injections and FS.b,c, Top, Representative Western blots. Bottom, Summary data for lysates of hippocampal CA1 from rats injected with Rp-8-Br-cAMPS or saline 1 h before a single episode of FS for 5 min. Western blots were probed with antibodies to pSer1166-GluN2B, pSer897-GluN1, pSer845-GluA1, and total GluN2B, GluN1, and GluA1, respectively. Rp-8-Br-cAMPS markedly reduced FS-induced phosphorylation of GluN2B at Ser1166 and GluA1 at Ser845, with little or no effect on pSer897-GluN1. Values for phosphorylated protein were normalized to the corresponding value for total protein and then to the corresponding value for control animals.n = 5 animals per treatment group. *p < 0.05. ***p < 0.001.
References
- Bloodgood BL, Sabatini BL. NMDA receptor-mediated calcium transients in dendritic spines. Biology of the NMDA receptor. In: Van Dongen AM, editor. Frontiers in neuroscience. Boca Raton, FL: CRC; 2009. - PubMed
Publication types
MeSH terms
Substances
Grants and funding
LinkOut - more resources
Full Text Sources
Other Literature Sources
Miscellaneous