CaMKII T286 phosphorylation has distinct essential functions in three forms of long-term plasticity

Abstract

The Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) mediates long-term potentiation or depression (LTP or LTD) after distinct stimuli of hippocampal NMDA-type glutamate receptors (NMDARs). NMDAR-dependent LTD prevails in juvenile mice, but a mechanistically different form of LTD can be readily induced in adults by instead stimulating metabotropic glutamate receptors (mGluRs). However, the role that CaMKII plays in the mGluR-dependent form of LTD is not clear. Here we show that mGluR-dependent LTD also requires CaMKII and its T286 autophosphorylation (pT286), which induces Ca²⁺-independent autonomous kinase activity. In addition, we compared the role of pT286 among three forms of long-term plasticity (NMDAR-dependent LTP and LTD, and mGluR-dependent LTD) using simultaneous live imaging of endogenous CaMKII together with synaptic marker proteins. We determined that after LTP stimuli, pT286 autophosphorylation accelerated CaMKII movement to excitatory synapses. After NMDAR-LTD stimuli, pT286 was strictly required for any movement to inhibitory synapses. Similar to NMDAR-LTD, we found the mGluR-LTD stimuli did not induce CaMKII movement to excitatory synapses. However, in contrast to NMDAR-LTD, we demonstrate that the mGluR-LTD did not involve CaMKII movement to inhibitory synapses and did not require additional T305/306 autophosphorylation. Thus, despite its prominent role in LTP, we conclude that CaMKII T286 autophosphorylation is also required for both major forms of hippocampal LTD, albeit with differential requirements for the heterosynaptic communication of excitatory signals to inhibitory synapses.

Keywords: Ca(2+)/calmodulin-dependent protein kinase II (CaMKII); N-methyl-D-aspartate receptor (NMDA receptor, NMDAR); hippocampus; long-term depression (LTD); long-term potentiation (LTP); metabotropic glutamate receptor (mGluR); neurons; synaptic plasticity.

Figure 1

mGluR-LTD requires the CaMKIIα isoform and its autophosphorylation at T286 but not T305/306.A, example traces and time course of synaptic response, measured by excitatory postsynaptic potential (EPSP) slope of the CA3-CA1 Schaffer collateral pathway, in wildtype hippocampal slices before and after chemical mGluR stimulation with DHPG (10 μM for 10 min). B, example traces and time course of synaptic response in CaMKIIα KO slices before and after the same DHPG treatment as in A show abolished mGluR-LTD. C, example traces and time course of synaptic response in T286A slices before and after DHPG treatment. D, example traces and time course of synaptic response in T305/306AV slices before and after stimulation with DHPG (10 μM for 10 min). E, quantification of the change in synaptic response (measured by EPSP slope) after DHPG stimulation in wildtype, CaMKIIα KO, T286A, and T305/306AV slices. Both CaMKII KO and T286A animals demonstrated severe impairments in DHPG-induced mGluR LTD when compared with wildtype animals while T305/306AV slices did not show any LTD deficit, indicating CaMKII and its T286 autophosphorylation are required for mGluR LTD (one-way ANOVA, Tukey’s post hoc test versus WT, ∗∗∗p < 0.001, n = 6, 6, 7, 5 slices).

Figure 2

T286A mutation does not affect the LTP-induced movement of overexpressed GFP-CaMKII.A, representative images of rat hippocampal neurons (DIV 14–17) expressing intrabodies for detection of PSD-95 (red) and gephyrin (blue) to label excitatory and inhibitory synapses, respectively, and overexpressing either CaMKII wildtype or T286A (green) before and 1 min following chemical NMDAR-LTP stimulation (cLTP; 100 μM glutamate/10 μM glycine, 1 min). The scale bar represents 10 μm. B, quantification of wildtype or T286A CaMKII at excitatory (red) and inhibitory (blue) synapses before and 1 min post cLTP stimulation. Both overexpressed wildtype and T286A CaMKII translocated to excitatory, but not inhibitory, synapses following cLTP treatment (paired t test, wildtype: ∗∗∗∗p < 0.0001, T286A: ∗∗∗p = 0.0003, n = 21, 12 neurons). C, full time course of wildtype (gray) and T286A CaMKII (orange) movement to excitatory and inhibitory synapses following cLTP stimulation indicating similar translocation dynamics for both constructs.

Figure 3

LTP-induced movement of endogenous CaMKII to excitatory synapses is accelerated by T286 autophosphorylation.A, representative images of hippocampal neurons from CaMKII T286A mutant mice (DIV 14–17) expressing intrabodies against endogenous PSD-95 (red) and gephyrin (blue), to label excitatory and inhibitory synapses, respectively, and CaMKII (green) before, 1 min, and 2 min after cLTP treatment. The scale bar represents 10 μm. B, quantification of CaMKII T286A at excitatory (red) and inhibitory (blue) synapses before, 1 min, and 2 min post cLTP stimulation. Similar as described in wildtype neurons, endogenous CaMKII in hippocampal neurons from T286A mutant mice moved to excitatory but not inhibitory synapses in response to cLTP stimulation. However, the endogenous T286A mutant moved more slowly to excitatory synapses, as significant synaptic enrichment was seen only at 2 min but not at 1 min after cLTP stimuli. These results indicate T286 autophosphorylation accelerates LTP-induced CaMKII synaptic targeting (one-way ANOVA, Tukey’s post hoc test versus pre, ∗∗p = 0.0054, n = 13 neurons). C, full time course of T286A CaMKII movement to excitatory and inhibitory synapses following cLTP stimulation. For comparison, the previously described movement of wildtype CaMKII is illustrated in gray (11).

Figure 4

CaMKII movement to inhibitory synapses in response to NMDAR-LTD stimuli requires T286 autophosphorylation.A, representative images of hippocampal neurons from CaMKII T286A mutant mice (DIV 14–17) expressing intrabodies against endogenous PSD-95 (red) and gephyrin (blue), to label excitatory and inhibitory synapses, respectively, and CaMKII (green) before and 5 min after chemical NMDAR-LTD stimuli (cLTD; 30 μM NMDA/10 μM CNQX/10 μM glycine, 1 min). The scale bar represents 10 μm. B, quantification of T286A CaMKII at excitatory (red) and inhibitory (blue) synapses prior to and 5 min post cLTD stimulation. In response to cLTD, endogenous T286A mutant CaMKII did not move to either excitatory or inhibitory synapses (in contrast to CaMKII wildtype, which has been described to move to inhibitory but not excitatory synapses in response to LTD). C, minimal movement of the T286A mutant to excitatory synapses (red) and lack of movement to inhibitory synapses (blue) following cLTD is further illustrated in a full time course up to 10 min after stimulation. For comparison, the previously described movement of wildtype CaMKII is illustrated in gray (11).

Figure 5

mGluR stimulation does not promote CaMKII movement to synapses.A, representative images of rat hippocampal neurons (DIV 14–17) expressing intrabodies targeting endogenous CaMKII (green), as well as PSD-95 (red) and gephyrin (blue) to label excitatory and inhibitory synapses, respectively, before and 0, 5, 10, and 20 min after stimulation with DHPG (100 μM for 10 min). The scale bar represents 10 μM. The left panel shows a full time course quantification of endogenous wildtype CaMKII at excitatory (red) and inhibitory (blue) synapses prior to and 0, 5, 10, and 20 min post DHPG stimulation. No CaMKII movement to either excitatory or inhibitory synapses was found in response to DHPG stimulation (in contrast to NMDAR-LTD stimuli, where wildtype CaMKII has been described to move to inhibitory but not excitatory synapses). B, for comparison with mutant mice, hippocampal neurons of wildtype mice were subjected to the same experiments as the rat neurons shown in A, with essentially the same results. C, representative images of hippocampal neurons from CaMKII T305/306AV mutant mice (DIV 14–17) expressing intrabodies targeting endogenous CaMKII (green), as well as PSD-95 (red) and gephyrin (blue) to label excitatory and inhibitory synapses, respectively, before and 0, 5, 10, and 20 min after stimulation with DHPG (100 μM for 10 min). The scale bar represents 10 μM. The left panel shows a full time course quantification of endogenous T305/6AV CaMKII at excitatory (red) and inhibitory (blue) synapses prior to and 0, 5, 10, and 20 min post DHPG stimulation. Again, no movement of CaMKII to either excitatory or inhibitory synapses was found in response to DHPG stimulation.