A Jekyll-and-Hyde Enzyme's Role In Alzheimer's

In their article in the December 8, 2005, issue of Neuron, Li-Huei Tsai and colleagues said their findings in mice could offer a model to study how normal neural mechanisms can go awry to contribute to the brain degeneration of AD.

The researchers' studies concentrated on a key brain regulatory enzyme called Cdk5. Elevated levels of Cdk5 had been found in the brains of AD patients, as well as elevated levels of a regulatory protein called p25 that is known to hyperactivate Cdk5. Researchers had found that the hyperactivating p25 is produced when the normal regulatory protein for Cdk5, called p35, is snipped into the shorter p25 by another enzyme.

In their experiments, the researchers used mice in which they could switch on p25 at will in the brain's learning and memory center, the hippocampus. In these mice, they found that switching on p25 for only two weeks actually enhanced learning and memory compared to normal mice. The two-week p25 mice could learn more quickly to associate a tone or environment with a mild electrical shock and to find a platform submerged in a pool of murky water.

In contrast, found the researchers, mice in which p25 had been switched on for six weeks showed impaired learning and memory in these tasks. Physiological studies showed that these mice showed significant brain atrophy and loss of hippocampal neurons. The researchers also showed that the two-week pulse of p25 did not cause neurodegeneration and had long-lasting effects on enhancing memory.

Studies of electrophysiological activity in the brains of the mice showed that the two-week p25 animals showed enhancement in the mechanism called "long-term potentiation" (LTP), which is central to learning and memory processes. In contrast, the six-week p25 animals showed a decrease in LTP.

Also, microscopy studies revealed that the two-week p25 animals showed higher connectivity among their neurons than did the six-week p25 animals.

The researchers concluded that their findings "indicate that transient p25 production facilitated learning, whereas the ability to form new memories was impaired by prolonged p25 expression." Also, they wrote that their data suggest that Cdk5 activity plays a role in the adult brain in forming new connections among neurons.

"It is possible that p25 is produced to compensate for the loss of Cdk5 activity during aging," they wrote. "In this scenario, chronic exposure to AD risk factors would further increase p25 levels to a critical concentration that ultimately contributes to neuronal loss.

"Taken together these data suggest that p25 production in vivo is not detrimental per se but can lead to neuronal cell death when p25 levels are chronically high," concluded Tsai and colleagues.

"In summary, it is intriguing that several studies suggest that during the pathogenesis of AD, which manifests over several years, compensatory mechanisms that initially enhance neuroplasticity eventually become maladaptive when chronically activated," they wrote. "A similar scenario can be envisioned for p25. Thus, the present study provides evidence that p25 generation might be a compensatory phenomenon to enhance neuroplasticity. These mice may also serve as a model whereby a factor that promotes plasticity can eventually contribute to neurodegeneration," they wrote.

The researchers include Andre Fischer and Farahnaz Sananbenesi of Harvard Medical School; Petti T. Pang and Bai Lu of the National Institute of Mental Health; and Li-Huei Tsai of Harvard Medical School and Howard Hughes Medical Institute. A.F. is a fellow of the Humboldt society, F.S. is a fellow of the German research foundation (DFG). L.-H.T. is an investigator of the Howard Hughes Medical Institute. This work is partially supported by NIH grant (NS051874) to L.-H.T and by funds from the NIH intramural research program to B.L.

Fischer at al.: "Opposing roles of transient and prolonged expression of p25 in synaptic plasticity and hippocampus-dependent memory" Publishing in Neuron, Vol. 48, 825-838, December 8, 2005, DOI 10.1016/j.neuron.2005.10.033, www.neuron.org