Developmental restoration of LTP deficits in heterozygous CaMKIIα KO mice

Abstract

The Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) is a major mediator of long-term potentiation (LTP) and depression (LTD), two opposing forms of synaptic plasticity underlying learning, memory and cognition. The heterozygous CaMKIIα isoform KO (CaMKIIα^+/-) mice have a schizophrenia-related phenotype, including impaired working memory. Here, we examined synaptic strength and plasticity in two brain areas implicated in working memory, hippocampus CA1 and medial prefrontal cortex (mPFC). Young CaMKIIα^+/- mice (postnatal days 12-16; corresponding to a developmental stage well before schizophrenia manifestation in humans) showed impaired hippocampal CA1 LTP. However, this LTP impairment normalized over development and was no longer detected in older CaMKIIα^+/- mice (postnatal weeks 9-11; corresponding to young adults). By contrast, the CaMKIIα^+/- mice failed to show the developmental increase of basal synaptic transmission in the CA1 seen in wild-type (WT) mice, resulting in impaired basal synaptic transmission in the older CaMKIIα^+/- mice. Other electrophysiological parameters were normal, including mPFC basal transmission, LTP, and paired-pulse facilitation, as well as CA1 LTD, depotentiation, and paired-pulse facilitation at either age tested. Hippocampal CaMKIIα levels were ∼60% of WT in both the older CaMKIIα^+/- mice and in the younger WT mice, resulting in ∼30% of adult WT expression in the younger CaMKIIα^+/- mice; levels in frontal cortex were the same as in hippocampus. Thus, in young mice, ∼30% of adult CaMKIIα expression is sufficient for normal LTD and depotentiation, while normal LTP requires higher levels, with ∼60% of CaMKIIα expression sufficient for normal LTP in adult mice.

Keywords: CaMKII; LTD; LTP; schizophrenia; synaptic transmission.

Fig. 1.

Hippocampal CA1 LTP but not depotentiation or long-term depression (LTD) is impaired in the young p12–16 CaMKIIα^+/− mice. Field excitatory postsynaptic potentials (fEPSPs) were recorded in the CA1 dendritic layer in response to CA1 Schaffer collateral stimulation. Long-term potentiation (LTP) was induced by high frequency stimulation (HFS; two 1 strains of 100-Hz stimulation separated by 20 s). LTD and depotentiation were induced by low -frequency stimulation (LFS; 900 pulses at 1 Hz). A: LTP and depotentiation plots and quantification from young p12–16 mice. Quantifications are from time points A, B, and C as indicated, representing a 20-min baseline, and last 10 min of LTP and depotentiation respectively. Right: example traces (average of 3 responses, representing 1 min) pre (black)- or post (grey)-HFS stimulation from time points A and B. B: depotentiation plot and quantification after renormalization to the last 10 min of LTP. Quantifications are from time points A and B as indicated, representing the renormalized last 10 min of LTP and the last 10 min of depotentiation. Right: example traces (average of 3 responses, representing 1 min) pre (black)- or post (grey)-LFS stimulation from time points A and B. LTP traces are the same as time point B traces in A. C: LTD plot and quantification from young postnatal day (p) 12–16 mice. Quantifications are from time points A and B as indicated, representing a 20-min baseline and last 10 min of LTD, respectively. Right: example traces (average of 3 responses, representing 1 min) pre (black)- or post (grey)-LFS from time points A and B. WT, wild type. **P < 0.01, ***P < 0.001 significantly different from prestimulus value indicating LTP, depotentiation, or LTD or as indicated between groups, one-way ANOVA followed by Tukey's honestly significant difference (HSD); NS, not significant.

Fig. 2.

Hippocampal CA1 LTP and depotentiation are normal in the older 9–11 wk CaMKIIα^+/− mice. fEPSPs were recorded in the CA1 dendritic layer in response to CA1 Schaffer collateral stimulation. LTP was induced by HFS (two 1 strains of 100-Hz stimulation separated by 20 s). Depotentiation was induced by LFS (900 pulses at 1 Hz). A: LTP plot and quantification from 9–11 wk mice. Quantifications are from time points A and B as indicated, representing 20 min baseline, and last 10 min of LTP, respectively. Right: example traces (average of 3 responses, representing 1 min) pre (black)- or post (grey)-HFS from time points A and B. B: Depotentiation plot and quantification after renormalization to the last 10 min of LTP. Quantifications are from time points A and B as indicated, representing the renormalized last 10 min of LTP and the last ten min of depotentiation. Right: example traces (average of 3 responses, representing 1 min) pre (black)- or post (grey)-LFS from time points A and B. LTP traces are the same as time point B traces in A. ***P < 0.001, significantly different from pre stimulus value indicating LTP or depotentiation; NS, not significant, one-way ANOVA followed by Tukey's HSD.

Fig. 3.

Hippocampal CA1 paired-pulse ratio is normal in both young and older CaMKIIα^+/− mice. Paired pulses (50-ms interpulse interval) were conducted at the beginning (after slice stabilization) and end of each experiment. A: quantification of paired-pulse ratio (first pulse/second pulse) from p12–16 young mice recorded at the beginning and end of each experiment. Right: example paired-pulse traces (average of 3 responses, representing 1 min) (pulse 1 black, pulse 2 grey) recorded at the beginning and end of each experiment. B: quantification of paired-pulse ratio (first pulse/second pulse) from 9–11 wk mice. Right: example traces (average of 3 responses, representing 1 min) (pulse 1 black, pulse 2 grey) recorded at the beginning and end of each experiment. C: percent amplitude of facilitation and decay during the first 50 pulses of 100-Hz stimuli used to induce LTP from young p12–16 mice. D: percent amplitude of facilitation and decay during the first 50 pulses of 100-Hz stimuli used to induce LTP from 9–11 wk mice. NS, not significant, one-way ANOVA for A and B, and for C, not significant at any pulse number by nonpaired, two-tailed t-test.

Fig. 4.

Additive effects of age and genotype on CaMKIIα expression in the hippocampus results in 4 times lower expression in the young CaMKIIα^+/− mice. Hippocampal homogenates were made from p12–16 and 9–11 wk WT and CaMKIIα^+/− mice (N = 3 for each age and genotype) and compared with SDS-PAGE and Western blot analysis for CaMKIIα and β, normalized to β-actin loading control. Densitometry is quantified as immunodetection value (IDV) normalized to values from 9–11 wk WT mice. A: representative blots. B: quantification of CaMKIIα expression. C: quantification of CaMKIIβ expression. *P < 0.05, **P < 0.01, ***P < 0.001, significantly different from 9–11 wk WT mice or between indicated groups; NS, not significant, two-way ANOVA (age × genotype) followed by Tukey's HSD; #main effect of age and different by t-test between respective ages but not significant by Tukey's HSD multiple comparison.

Fig. 5.

mPFC LTP and paired-pulse facilitation is normal in 9–11 wk CaMKIIα^+/− mice. fEPSPs were recorded in mPFC layer 5 in response to layer 2–3 stimulation. LTP was induced by HFS (two 1 strains of 100-Hz stimulation separated by 20 s). Paired pulses (50-ms interpulse interval) were conducted at the beginning (after slice stabilization) and end of each experiment. A: LTP plots and quantification from young p12–16 mice. Quantifications are from time points A and B as indicated, representing 15 min baseline, and last 10 min of LTP respectively. Right: example traces (average of 3 responses, representing 1 min) pre (black)- or post (grey)-HFS stimulation from time points A and B. B: quantification of paired-pulse ratio (first pulse/second pulse). Right: example traces (average of 3 responses, representing 1 min) (pulse 1, black, pulse 2 grey) recorded at the beginning and end of each experiment. **P < 0.01, ***P < 0.001 significantly different from prestimulus value indicating LTP, one-way ANOVA followed by Tukey's HSD; NS, not significant.

Fig. 6.

Frontal cortex CaMKIIα expression does not differ from hippocampus in 9–11 wk mice. Hippocampal and frontal cortex homogenates were made from 9–11 wk WT and CaMKIIα^+/− mice (N = 3 hippocampus and cortex were taken from same mouse) and compared with SDS-PAGE and Western blot analysis for CaMKIIα and β, normalized to β-actin loading control. Densitometry is quantified as immunodetection value (IDV) and normalized to values from WT hippocampus. A: representative blots. B: quantification of CaMKIIα expression. C: quantification of CaMKIIβ expression. *P < 0.05, **P < 0.01, two-way ANOVA (brain area × genotype) followed by Tukey's HSD; NS, not significant.

Fig. 7.

CaMKIIα^+/− mice show a selective deficit in synaptic efficacy at the hippocampal CA1 synapse that develops with age. Input/output (I/O) curves were generated by increasing stimulus strength by 5 μA until responses leveled off or in hippocampus showed signs of population spiking. Regression analysis of the slope (mV/ms) of response vs. the corresponding stimulus intensity (μA) are compared; A–C include best fit lines for each genotype (left) and representative traces (right). A: I/O plot from hippocampus CA1 in p12–16 mice. B: I/O plot from hippocampus CA1 in 9–11 wk mice. C: I/O plot for mPFC layer 5 from 9–11 wk mice. ***P < 0.001, different curves are needed for each genotype; NS, one curve fits both genotypes, sum of squares F-test. For clarity, best-fit lines are shown for each genotype rather than one fit in A and C.