20 July 2007. Memories are not always worth keeping. People who suffer from post-traumatic stress disorder, for example, wrestle to suppress thoughts and feelings that take a heavy emotional toll. Unfortunately, it is not always possible to purge bad memories, but as scientists begin to understand the biology behind memory extinction, there is a glimmer of hope that they might find a therapeutic approach that helps. In this week’s Nature Neuroscience online, researchers led by Li-Huei Tsai at MIT reveal one aspect of memory extinction that may be amenable to manipulation. They report that in mice, contextual fear extinction is impaired by Cdk5, a kinase best known for its role in the developing nervous system. Cdk5 has also been implicated in the pathology of Alzheimer disease and other neurodegenerative disorders (see ARF related news story).
Memory extinction is often measured in mice using a fear conditioning paradigm, where animals learn to associate a condition, such as a specific cage or environment, with a mild though unpleasant shock. Extinction comes when the animal is repeatedly re-exposed to the same environment but without the stimulus. The animal typically freezes in expectation of the shock, but as the mouse learns that no shock is coming, this reaction subsides.
To test the role of Cdk5 in fear extinction, first author Farahnaz Sananbenesi and colleagues injected butyrolactone, or roscovitine, inhibitors of the kinase into mouse hippocampi at various times during a 6-day memory extinction experiment. Animals treated with the inhibitors showed significantly enhanced fear extinction than control mice. A single injection of 50 ng of butyrolactone was sufficient to reduce the percentage of freezing from about 55 to 25 one day after the initial fear conditioning. In contrast, control animals typically took 4 days of trials to achieve this level of memory extinction. The timing of the inhibition seems crucial. When given 15 minutes prior or 3 hours after the trial, the Cdk5 inhibitors had no effect on freezing; instead they had to be given immediately after the extinction trials to work, “suggesting that Cdk5 activity affects the consolidation of extinction in our procedure,” write the authors.
Because inhibitors can often have pleiotropic effects, Sananbenesi and colleagues used a more direct approach to test the role of Cdk5 in extinction. They tested transgenic mice harboring an inducible Cdk5 activator, CK-p25, in the same experimental paradigm. Activation of Cdk5 would be expected to prevent extinction, and this is just what the researchers found. When p25 was constitutively active, animals showed absolutely no fear extinction whatsoever, but when p25 expression was turned off, extinction returned to normal levels. “Thus, increased Cdk5 activity prevented extinction, which can be reversed on downregulation of Cdk5 activity,” they write.
Cdk5 has been linked to positive processes such as learning and memory, and also to the Aβ and tau pathology associated with AD. How might it also prevent fear extinction? It seems that this may be linked to a complex signal transduction pathway involving the small GTPase Rac-1, which regulates the localization of Cdk5 and its coactivator p35, and p21-activated kinase-1 (PAK-1). The researchers found that during the extinction process there was a redeployment of Cdk5, p35, and Rac-1 from the membrane fraction to the cytosol. They found they could facilitate this redistribution by administering a Rac-1 inhibitor, NSC23760, which also promoted memory extinction. Since Cdk5/p35 is known to inhibit PAK-1 activity in a Rac-1-dependent manner, the authors questioned what role the p21-activated kinase plays in the process. They found that a dominant-negative form of PAK-1 prevents memory extinction when introduced into mouse hippocampus via viral vector.
“In summary, our data indicate that during extinction membrane depletion of Cdk5 activity and dissociation of p35 from PAK-1 in the cytosol removes the inhibitory tone on PAK-1 activity,” write the authors. They also note that other regulatory molecules, such as phosphatases, most likely regulate PAK-1 during memory extinction as well.
How these findings related to AD, particularly the fear and anxiety experienced by patients in the late stages of disease, is unclear. However, given that Cdk5 may contribute to tau and Aβ pathology and that lower PAK-1 levels have been linked to synaptic loss in AD (see ARF related news story), the relationship is worth watching.—Tom Fagan.
Sananbenesi F, Fischer A, Wang X, Schrick C, Neve R, Radulovic J, Tsai L-H. A hippocampal Cdk5 pathway regulates extinction of contextual fear. Nature Neuroscience. 2007, July 15. Advanced online publication. Abstract