Network compensation of cyclic GMP-dependent protein kinase II knockout in the hippocampus by Ca2+-permeable AMPA receptors

Abstract

Gene knockout (KO) does not always result in phenotypic changes, possibly due to mechanisms of functional compensation. We have studied mice lacking cGMP-dependent kinase II (cGKII), which phosphorylates GluA1, a subunit of AMPA receptors (AMPARs), and promotes hippocampal long-term potentiation (LTP) through AMPAR trafficking. Acute cGKII inhibition significantly reduces LTP, whereas cGKII KO mice show no LTP impairment. Significantly, the closely related kinase, cGKI, does not compensate for cGKII KO. Here, we describe a previously unidentified pathway in the KO hippocampus that provides functional compensation for the LTP impairment observed when cGKII is acutely inhibited. We found that in cultured cGKII KO hippocampal neurons, cGKII-dependent phosphorylation of inositol 1,4,5-trisphosphate receptors was decreased, reducing cytoplasmic Ca(2+) signals. This led to a reduction of calcineurin activity, thereby stabilizing GluA1 phosphorylation and promoting synaptic expression of Ca(2+)-permeable AMPARs, which in turn induced a previously unidentified form of LTP as a compensatory response in the KO hippocampus. Calcineurin-dependent Ca(2+)-permeable AMPAR expression observed here is also used during activity-dependent homeostatic synaptic plasticity. Thus, a homeostatic mechanism used during activity reduction provides functional compensation for gene KO in the cGKII KO hippocampus.

Keywords: Ca2+-permeable AMPA receptors; LTP; calcineurin; gene knockout.

Fig. 1.

No impairment of hippocampal LTP in the cGKII KO. (A) Inhibition of cGKs significantly decreases LTP induced by TBS in WT slices, but not in the KO. Summary graph shows mean fEPSP slopes measured 60 min after LTP induction (n = 10 KO, 8 KO + Rp-8–pCPT–cGMS, 9 WT, and 7 WT + Rp-8–pCPT–cGMS animals; **P < 0.01, two-way ANOVA with Fisher’s LSD test). (B) Representative raw traces of fEPSPs in each condition shown in A. Solid lines represent baseline braces before TBS, and dotted lines indicate traces 120 min after TBS. (C) Input–output relation shows no difference in WT and KO slices (n = 14 KO and 14 WT animals).

Fig. 2.

Synaptic CPAR expression is required for hippocampal LTP in the KO. (A) Representative traces of mEPSC recordings in each condition (n, number of cells). (B) A CPAR-mediated increase in average mEPSC amplitude in the KO (*P < 0.05, ***P < 0.001, and ****P < 0.0001, one-way ANOVA with Fisher’s LSD test). (C) No difference in average mEPSC frequency in each condition. (D) A CPAR-mediated decrease in average decay time (peak to 10%) in the KO (****P < 0.0001, one-way ANOVA with Fisher’s LSD test). (E) Inhibition of CPAR blocks LTP in the KO but not in WT slices. Summary graph shows mean fEPSP slopes measured 60 min after LTP induction (n = 4 KO + naspm and 4 WT + naspm animals; ****P < 0.0001, two-way ANOVA with Fisher’s LSD test).

Fig. 3.

Synaptic elevation of GluA1 and reduction of GluA2/3 in hippocampal KO neurons. (A) Representative immunoblots and quantitative analysis of PSD from the hippocampus of WT and KO mice showing GluA1 levels are increased whereas GluA2/3 levels are reduced in the KO (n = 3 WT and 3 KO animals; *P < 0.05, **P < 0.01, and ****P < 0.0001, unpaired two-tailed Student’s t tests). (B) Representative immunoblots of surface biotinylation and a summary graph in the WT and KO neurons showing surface GluA1 levels are elevated, but surface GluA2/3 is decreased in the KO (n = 3 experiments; *P < 0.05 and ***P < 0.001, unpaired two-tailed Student’s t tests).

Fig. 4.

PKA and CaMKII-independent but NMDAR and PKC-dependent LTP in the KO. (A) Inhibition of PKA activity significantly reduces hippocampal LTP in WT but not in KO slices (n = 13 KO + KT5720, 9 WT, 10 KO, and 11 WT + KT5720 animals; ***P < 0.001, two-way ANOVA with Fisher’s LSD test). (B) NMDARs are required for LTP in both WT and KO hippocampus (n = 4 KO, 3 KO + APV, 4 WT, and 4 WT + APV animals; ***P < 0.001 and ****P < 0.0001, two-way ANOVA with Fisher’s LSD test). (C) Inhibition of CaMKII activity significantly decreases hippocampal LTP in WT but not in KO slices (n = 7 KO, 11 KO + KN-93, 6 WT, and 8 WT + KN-93 animals; *P < 0.05, two-way ANOVA with Fisher’s LSD test). (D) Inhibition of PKC significantly impairs LTP in both WT and KO hippocampus (n = 7 KO, 6 KO + chelerythrine, 6 WT, and 5 WT + chelerythrine animals; ***P < 0.001, two-way ANOVA with Fisher’s LSD test). All summary graphs show mean fEPSP slopes measured 60 min after LTP induction.

Fig. 5.

In vivo calcineurin activity is reduced in hippocampal KO neurons. Shown are representative images of CFP channel, FRET channel, and pseducolored emission ratio (Y/C) in each condition [blue (L), low emission ratio, and red (H), high emission ratio]. (Scale bar, 10 μm.) A summary graph shows a decrease in average of the emission ratio (Y/C) in the KO (n, number of cells; ***P < 0.001, unpaired two-tailed Student’s t tests).

Fig. 6.

Reduction of cGKII-dependent pIP3R(S1756) and Ca²⁺ signals in hippocampal KO neurons mediates synaptic CPAR expression. (A) Representative immunoblots and quantitative analysis of pIP3R(S1756) levels in WT and KO neurons in the presence of the PKA inhibitor showing that pIP3R(S1756) is reduced in the KO (n = 3 experiments; **P < 0.01, unpaired two-tailed Student’s t tests). (B) Normalized average of total Ca²⁺ signals in each condition reveals that reduced IP3R mediated Ca²⁺ signals in the KO (n, number of neurons; **P < 0.01, ***P < 0.001, and ****P < 0.0001, one-way ANOVA with Fisher’s LSD test). (C) Representative traces of mEPSC recordings in each condition (n, number of cells), average mEPSC amplitude, frequency, and decay time (peak to 10%) showing selective IP3R-dependent synaptic CPAR expression in the KO (**P < 0.01, one-way ANOVA with Fisher’s LSD test).