Transgenic calmodulin-dependent protein kinase II activation: dose-dependent effects on synaptic plasticity, learning, and memory

Abstract

Genetic disruption of calmodulin-dependent protein kinase II (CaMKII) function alters hippocampal synaptic plasticity and memory in mice. We used transgenic mice carrying a tetracycline-regulated, calcium-independent form of CaMKII (CaMKII-Asp286) to investigate the role of CaMKII activation on synaptic plasticity and behavior. Mice expressing low levels of a CaMKII-Asp286 transgene have facilitated low-frequency (5 Hz)-induced long-term potentiation (LTP), whereas mice with high levels of transgene expression have a deficit in this form of plasticity. Behavioral impairments on fear-conditioned memory and visible water maze correlate with the level of CaMKII-Asp286 expression. Mice with high levels of CaMKII-Asp286 have reversible, compensatory changes in the expression of genes associated with inhibitory neurotransmission. These results demonstrate that in the hippocampus, CaMKII activation facilitates the induction of low-frequency LTP, but at high levels of expression, compensatory mechanisms act to inhibit the induction of this form of LTP. The most severe behavioral impairments are associated with activation of this compensatory mechanism.

Fig. 1.

Transgene regulation. A, Expression of transgenic CaMKII–Asp286 mRNA is shown by in situ hybridization in these half-coronal forebrain sections taken from B13-DEV(high) mice. Mice treated with a low-dose doxycycline diet for >3 weeks show no expression (On DOX). Transgene mRNA expression is readily apparent within 2 d and is present at high levels at 14 d after mice are switched to a Dox-free diet. Brain slices used in this figure were prepared in parallel, probed with the same radiolabeled oligonucleotide, and exposed to film for the same amount of time. B, Mice raised with their transgene suppressed (B13) do not express as much CaMKII–Asp286 mRNA as adults compared with mice that have expressed the transgene throughout development (B13-DEV). The average level of expression in B13-DEV mice has been normalized to 100%. For hippocampal expression, p < 0.03; for striatal expression, p < 0.001.

Fig. 2.

Behavioral effects of CaMKII–Asp286 transgene expression. A, Timeline showing regulation of the CaMKII–Asp286 transgene. B13 mice (gray line) had their transgene off throughout development. In B13-DEV mice (black line), the transgene was expressed during development and was first suppressed at 6 weeks of age. At ∼11 weeks of age, both groups of mice had their transgene induced for 21 d before behavioral testing. B, The latency for mice to reach the visible platform is shown as a function of training trial. Across all trials, B13-DEV(high) mice have longer latencies than B13(low) mice (p < 0.02), which have longer latencies than control mice (p < 0.001).C, Degree of freezing in response to training, context, and auditory cues after tone–shock pairings. Despite having similar degrees of baseline freezing (p > 0.05 by ANOVA), B13-DEV(high) mice freeze less to context and cues than do B13(low) mice (p < 0.015) and control mice (p < 0.001).

Fig. 3.

Correlations between CaMKII–Asp286 gene dose and behavioral performance. A, Contextual fear-conditioned freezing 24 hr after training is plotted as a function of CaMKII–Asp286 mRNA levels in the hippocampus. TheR² value for all transgenic mice is 0.683 (p < 0.001). B, The average latency for mice to reach the visible water-maze platform on day 2 of training is plotted as a function of hippocampal CaMKII–Asp286 mRNA levels. The R²value for all transgenic mice is 0.527 (p < 0.001). Regardless of transgene exposure during development, animals with a greater transgene dose as adults were more likely to have severe deficits on these tasks. One unit of expression is defined as the amount of mRNA expressed by mice carrying only the tetO-linked CaMKII–Asp286 transgene (no tTA transgene). The control group includes six wild-type mice and three mice carrying only the tetO-linked CaMKII–Asp286 transgene.

Fig. 4.

Reversible behavioral deficits in B13-DEV(high) mice. A, Timeline showing the CaMKII–Asp286 transgene activation pattern for the experiment shown in B.B, Contextual freezing is impaired in B13-DEV(high) mice 24 hr after training with the transgene on and at 6 weeks after transgene suppression (measured during the baseline period immediately before retraining). Twenty-four hours after retraining with the transgene off, B13-DEV(high) mice show normal freezing to context compared with controls. When the transgene is induced after normal learning, B13-DEV(high) mice have a complete lack of freezing to the context learned previously. **p < 0.001.C, B13-DEV(high) mice fail to locate a visible platform in the water maze, whereas controls acquire this task readily. After transgene suppression, B13-DEV(high) mice learn to locate the visible platform (performance on last two trials, wild-type vs B13 mice;p = 0.14; NS).

Fig. 5.

CaMKII–Asp286 expression alters theta-frequency LTP. Hippocampal LTP induced by 5 Hz stimulation in B13(low) and B13-DEV(high) groups during CaMKII–Asp286 expression is shown. Hippocampal slices from B13(low) and B13-DEV(high) groups were stimulated at 0.02 Hz while EPSPs were recorded from CA1 dendritic fields. After 20 min of baseline, slices were stimulated at 5 Hz for 30 sec, followed by 1 hr of 0.02 Hz recording. Note how B13(low) and B13-DEV(high) mutants have opposing responses to 5 Hz stimulation. B13-DEV(high) mutants have an LTP deficit, whereas B13(low) mutants trend toward enhanced LTP at this stimulation frequency. Representative EPSPs for each group of animals tested are shown before and after the 5 Hz stimulation. Calibration, 10 msec, 2 mV.

Fig. 6.

Striatal GAD65 and tachykinin expression correlates with CaMKII–Asp286 expression. A, Striatal expression of CaMKII–Asp286 and GAD65 co-vary, with anR² value for all transgenic mice of 0.722 (p < 0.001). B, Striatal expression of CaMKII–Asp286 and Tac co-vary, with anR² value for all transgenic mice of 0.865 (p < 0.001). Expression levels were assayed by Northern blot of total striatal RNA and normalized for levels of tubulin mRNA. One unit of expression is defined as the amount of mRNA expressed by mice carrying only the tetO-linked CaMKII–Asp286 transgene (no tTA transgene). The control group includes four wild-type mice and two mice carrying only the tetO-linked CaMKII–Asp286 transgene. C, Northern blots for GAD65 and Tac mRNA. Whole-forebrain RNA from B13-DEV(high) mutant mice with their transgene induced (High ON) and suppressed (High OFF) as well as from B13(low) mutants with their transgene induced (Low ON) is probed for GAD65. At this exposure, significant expression is seen only in the B13-DEV(high) mice tested with their transgene expressed (High ON).