Commentary Cellscience Reviews Vol 1 No.4 ISSN 1742-8130

The learning switch? CaMKII autophosphorylation and its role in learning vs. memory

Jennifer E. Mehren

Introduction

Over the last 20 years, calcium/calmodulin-dependent protein kinase II (CaMKII) has garnered much support as an important player in synaptic plasticity and behavior. CaMKII has been heavily studied because of its exceptionally high concentration in the mammalian forebrain, synaptic localization, multiple substrates, and ability to sense Ca²⁺ and act for prolonged periods following transient Ca²⁺ pulses (for review of CaMKII biochemistry, see Hudmon and Schulman, 2002). CaMKII autophosphorylation at residue Thr²⁸⁶ is crucial for long-term potentiation (LTP) in the hippocampus, a well-studied form of synaptic plasticity, which is thought to contribute to learning and memory (for reviews, see Lisman et al., 2002; Colbran and Brown, 2004). CaMKII has been dubbed a molecular memory "switch", the theory being that the autophosphorylated form of the enzyme "turns on" and maintains neuronal pathways representing memory traces (e.g. Miller and Kennedy, 1986; Lisman and Fallon, 1999). A recent study by Irvine et al. (2005) suggests that CaMKII autophosphorylation is not necessary for storage or recall of fear memory, but instead, it contributes to the early stages of acquisition of fear-conditioned memories.

Mouse genetic studies have shed much light on the role of CaMKII in synaptic plasticity and cognitive behavior (for review, see Elgersma et al., 2004). Knock-out mice deficient for α-CaMKII have defective LTP in hippocampal cells and defective spatial learning in the Morris water maze (Silva et al., 1992a; Silva et al., 1992b). Interestingly, these mice have decreased fear and low pain thresholds compared to wild type, which may have confounded behavioral results (Chen et al., 1994). A knock-in mouse expressing CaMKII-T286A, a form of CaMKII that is unable to autophosphorylate at Thr²⁸⁶, and therefore does not have prolonged activity after transient Ca²⁺ pulses, shows a lack of NMDAR-dependent LTP in hippocampal slices and deficient spatial learning in the Morris water maze (Giese et al., 1998). These results imply that the Ca²⁺-independent activity of CaMKII in particular is involved in cellular and behavioral plasticity. However, the role of CaMKII autophosphorylation in memory could not be assessed because the mutant mice were not able to learn the task to be remembered.

Fear and calcium-dependence

Giese and colleagues continued to study the T286A mutant mice, using different behavioral manipulations to tease apart the role of CaMKII autophosphorylation during memory acquisition (training) and memory recall (Irvine et al., 2005). They tested T286A mice in both cued and contextual fear conditioning, as well as passive avoidance, with variations (Irvine et al., 2005). Passive avoidance and contextual fear conditioning involve the amygdala and hippocampus, while the cued conditioning task is hippocampus-independent (Liang et al., 1982; Taubenfeld et al., 2001; Philips and LeDoux, 1992). In the passive avoidance task, mice were placed in a lit compartment, and allowed to step into an adjacent dark compartment where they received a mild foot-shock. After a single training trial, this procedure was repeated either immediately, 30 minutes, or 24 hours later. T286A mice showed no avoidance memory at any time point tested after a single training trial (Irvine at al., 2005). To find out whether the poor performance of the T286A mice was caused by an acquisition defect or a memory storage or retrieval defect, the investigators put the mutant mice through more intensive training, in which the passive avoidance task was repeated again until the mutants learned to stay in the lit compartment for more than 120 s. The mutants needed more training trials to reach this criterion compared with wild type mice, but after being trained in this manner, they displayed avoidance memory. T286A mutant mice also showed no cued (tone-shock association) fear memory or contextual fear memory after a single training trial, however wild type levels of memory could be obtained after 3 (cued) or 5 (contextual) training trials (Irvine et al., 2005). Using three different behavioral assays, Irvine et al. (2005) show that CaMKII autophosphorylation at Thr²⁸⁶ is needed for efficient training, but not for expression of fear memory.

Data obtained with T286A mice, suggesting that Ca²⁺-independent CaMKII activity contributes to learning, but is not necessary for memory, confirms fear conditioning results obtained in rats combining pharmacological inhibition of endogenous CaMKII with immunogold EM quantification of pCaMKII-Thr²⁸⁶ in lateral amygdala spines (Rodrigues et al., 2004). Rats were trained in cue conditioning with five tone-shock pairings, then their amydalas were immunostained and found to have significantly more pCaMKII-Thr²⁸⁶ compared with controls (Rodrigues et al., 2004). In another experiment in the same study, rats were trained in both cued and contextual memory after having their amygdalas infused with a CaMKII inhibitor, KN-62 either before training or before testing. Inhibition of CaMKII before training disrupted short- and long-term fear memories, while KN-62 infusion after training did not cause memory defects (Rodrigues et al., 2004). This is another piece of evidence that CaMKII activity is important for learning, but not necessarily for memory of cued and contextual fear tasks.

Constitutive activation of CaMKII and learning

If making CaMKII unable to autophosphorylate at Thr²⁸⁶ retards learning, how does the opposite molecular manipulation affect learning? Interestingly, and in accord with the work by Irvine et al. (2005), when Ca²⁺-independent CaMKII (T287D) is selectively overexpressed in Drosophila, better training is seen in courtship conditioning, while memory remains intact (Mehren and Griffith, 2004). However, transgenic mice overexpressing the constitutively active form of CaMKII (T286D) in a wild type background have defective spatial learning and memory, and defective cued and contextual memory (Bach et al., 1995; Mayford et al., 1996; Bejar et al., 2002). There is a dose-dependent effect on LTP, where mice with high levels of T286D have a deficit in 5 Hz-induced LTP, mice with low levels of T286D show an enhancement in this type of LTP (Bejar et al., 2002). The researchers conducted a gene expression analysis, and suggested that cellular and behavioral defects seen with high levels of T286D expression are possibly caused by compensatory changes in endogenous expression of genes involved in inhibitory neurotransmission (Bejar et al., 2002).

Conclusions

Ideally, we would be able to visualize autophosphorylation of CaMKII in learning pathways during tasks, and assess whether the Thr^286/7 phosphorylation persists through memory recall. Unfortunately, technology has not advanced this far yet, so we must interpret our genetic, cellular, and behavioral data the best we can. It is difficult to make direct reciprocal comparisons between the T286A knock-in mice and the overexpressing T286D mice. The CaMKII gene in the knock-in mice has been completely replaced with the T286A mutant sequence, and therefore expresses the mutated transgene through development and adulthood. The T286D transgene has been engineered to be regionally and temporally controlled, and is expressed in addition to endogenous CaMKII (Mayford et al., 1996; Bejar et al, 2002). αCaMKII and βCaMKII levels are not altered in T286A mutants (Giese et al., 1998), but it is plausible that with extended training, βCaMKII Ca²⁺-independent activity can compensate enough to produce learning in αCaMKII-T286A mutants. Subcellular distribution of βCaMKII should be investigated, as it has been shown that αCaMKII null mice have increased amounts of βCaMKII in the PSD (Elgersma et al., 2002). It would also be useful to conduct a microarray analysis of T286A mutants to see if compensatory expression changes occur in other endogenous learning-related enzymes such as PKC or PKA. Such compensatory changes might explain the ability of T286A mutants to learn with extended training.

It is important to note that CaMKII has different biochemical mechanisms of plasticity depending in which cells and behaviors it is acting. In rodent barrel cortex, overexpression of constitutively active CaMKII does not affect vibrissa deprivation plasticity, but the T286A mutation blocks this type of plasticity (Glazewski et al., 2000, 2001). In contrast, overexpression of constitutively active CaMKII in Drosophila led to an enhancement in training performance, while the T287A mutation did not have any effect (Mehren et al., 2004). Neither Ca²⁺-independent nor Ca²⁺-dependent CaMKII mutants affected memory of courtship conditioning (Mehren et al., 2004). Being able to measure learning performance separate from memory performance, and knowing which circuits and biochemical mechanisms are utilized in each is clearly important.

Irvine et al. do not mention whether T286A mice have altered pain thresholds, but all three tasks they look at involve foot shock. The criterion training where the T286A mutants require approximately three times the amount of foot shocks to produce avoidance memory suggests that these mutants could possibly have higher pain thresholds. The use of a reinforcer in lieu of foot shock in a hippocampal and amygdala-dependent assay would be a plus. Evidence continues to accumulate, but there is still much more work to be done to clarify the role of CaMKII autophosphorylation in learning and/or memory.

Acknowledgements: Special Thanks to Leslie C. Griffith for comments on the manuscript.

REFERENCES

Bach ME, Hawkins RD, Osman M, Kandel ER, Mayford M (1995) Impairment of spatial but not contextual memory in CaMKII mutant mice with a selective loss of hippocampal LTP in the range of the theta frequency. Cell 81:905-915.

Bejar R, Yasuda R, Krugers H, Hood K, Mayford M (2002) Transgenic calmodulin-dependent protein kinase II activation: dose-dependent effects on synaptic plasticity, learning, and memory. J Neurosci 22:5719-5726.

Chen C, Rainnie DG, Greene RW, Tonegawa S(1994). Abnormal fear response and aggressive behavior in mutant mice deficient for α-calcium-calmodulin kinase II. Science 266:291-294.

Colbran RJ, Brown AM (2004) Calcium/calmodulin-dependent protein kinase II and synaptic plasticity. Curr Opin Neurobiol 14:318-327.

Elgersma Y, Fedorov NB, Ikonen S, Choi ES, Elgersma M, Carvalho OM, Giese KP, and Silva AJ (2002) Inhibitory autophosphorylation of CaMKII controls PSD association, plasticity, and learning. Neuron 36:493-505.

Elgersma Y, Sweatt JD, Giese KP (2004) Mouse genetic approaches to investigating calcium/calmodulin-dependent protein kinase II function in plasticity and cognition. J. Neurosci 24:8410-8415.

Giese KP, Fedorov NB, Filipkowski RK, Silva AJ (1998) Autophosphorylation at Thr286 of the alpha calcium-calmodulin kinase II in LTP and learning. Science 279:870-873.

Glazewski S, Giese KP, Silva A, Fox K (2000) The role of alpha-CaMKII autophosphorylation in neocortical experience-dependent plasticity. Nat Neurosci 3:911-918.

Glazewski S, Bejar R, Mayfor M, Fox K (2001) The effect of autonomous alpha-CaMKII expression on sensory responses and experience-dependent plasticity in mouse barrel cortex. Neuropharmacology 41:771-778.

Hudmon A, Schulman H (2002) Neuronal Ca2+/calmodulin-dependent protein kinase II: the role of structure and autoregulation in cellular function. Annu Rev Biochem 71:473-510.

Irvine EE, Vernon J, Giese KP (2005) αCaMKII autophosphorylation contributes to rapid learning but is not necessary for memory. Nat Neurosci 8:411-412.

Liang KC, McGaugh JL, Martinez JL Jr, Jensen RA, Vasquez BJ, Messing RB (1982) Post-training amygdaloid lesions impair retention of an inhibitory avoidance response. Behav Brain Res 4:237-249.

Lisman JE and Fallon JR (1999) What maintains memories? Science 283:339-340.

Lisman J, Schulman H, Cline H (2002) The molecular basis of CaMKII function in synaptic and behavioural memory. Nat Rev Neurosci 3:175-190.

Mayford M, Bach ME, Huang YY, Wang L, Hawkins RD, Kandel ER (1996) Control of memory formation through regulated expression of a CaMKII transgene. Science 274:1678-1683.

Mehren JE, Griffith LC (2004) Calcium-independent calcium/calmodulin-dependent protein kinase II in the adult Drosophila CNS enhances the training of pheromonal cues. J. Neurosci 24:10584-10593.

Miller SG, Kennedy MB (1986) Regulation of brain type II Ca2+/calmodulin-dependent protein kinase by autophosphorylation: a Ca2+-triggered molecular switch. Cell 44:861-870.

Phillips RG, LeDoux JE (1992) Differential contribution of amygdala and hippocampus to cues and contextual fear conditioning. Behav Neurosci 106:274-285.

Rodrigues SM, Farb CR, Bauer EP, LeDoux JE, Schafe GE (2004) Pavlovian fear conditioning regulates Thr286 autophosphorylation of Ca2+/calmodulin-dependent protein kinase II at lateral amygdala synapses. J Neurosci 24:3281-3288.

Silva AJ, Paylor R, Wehner JM, Tonegawa S (1992a) Impaired spatial learning in alpha-calcium-calmodulin kinase II mutant mice [see comments]. Science 257:206-211.

Silva AJ, Stevens CF, Tonegawa S, Wang Y (1992b) Deficient hippocampal long-term potentiation in α-calcium-calmodulin kinase II mutant mice. Science 257:201-205.

Taubenfeld SM, Milekic MH, Monti B, Alberini CM (2001) The consolidation of new but not reactivated memory requires hippocampal C/EBPβ. Nat Neurosci 4:813-818.