Senior Science Editor
Researchers were able to reverse memory deficits in a mouse model of Alzheimer's disease (AD) by preventing cleavage of the protein tau by the enzyme caspase-2.
"There's a soluble aspect of tau that disrupts synaptic function, and we have identified a key player in why that occurs – the caspase-2 cleavage of tau," said Karen Ashe, a professor of neurology and neuroscience at the University of Minnesota, summarizing the results for Medical Device Daily.
"Reducing caspase-2 caused synapses to repair themselves," which suggests that targeting the process Ashe and her team have identified might offer advantages over current strategies, which are by and large aimed at "preventing further decline."
Amyloid beta plaques and tau tangles are two anatomical hallmarks of AD. But neither correlates particularly well with actual symptoms.
"The plaques don't correlate well at all," Ashe said. The pharma industry continues to learn that the hard way, with nothing to show, at least so far, for decades of strategies aimed at amyloid beta.
Tau pathology, which occurs later in the disease, correlates somewhat better. But in previous work, Ashe and her team had shown that tangles could be disentangled from cognitive deficits.
In work published in 2005, Ashe and her team showed that in a transgenic tau-driven model of AD, suppressing tau expression improved memory, even though tangles continued to accumulate. That work, Ashe said, demonstrated that "the tangles are not the cause of the disease."
For amyloid beta, the realization that plaques are not the cause of disease has shifted attention to targeting soluble oligomers.
In the Oct. 10, 2016, issue of Nature Medicine, Ashe and her team suggest that for tau, the key to disease lies in its cleavage by caspase-2.
Cleavage at a certain amino acid, they showed in a mouse model of tau-driven AD, led to tau's translocation into synapses, where it affected receptors and impaired synaptic transmission. Memories are physically stored as changes in synaptic strength, and mice developed memory deficits. Those deficits could be reversed by inhibiting caspase-2.
In their studies, Ashe and her team found that the cut form of tau and the full-length version co-operate in an as-yet-unknown way to produce neural damage and memory deficits. Mice expressing a version of tau that could not be cut by caspase-2 developed neither neurodegeneration nor memory deficits.
Figuring out the exact nature of the interaction, Ashe said, "is going to take more work...We need to understand whether the cleaved form is acting from inside the cell, or whether it might be getting released from the cell and acting to stimulate in another way."
What's already clear, though, is that cleaved tau is not working through tangles. The specific form studied by Ashe and her team, which is cut at amino acid 314 and called deltatau314, does not form tangles.
Ashe and her team plan to investigate "how delta-tau causes full-length tau to mislocalize to dendritic spines," she said. She also plans to work with a medicinal chemist to develop a caspase-2 inhibitor that crosses the blood-brain barrier.
Ashe said that caspase-2 is "potentially a really nice target." In knockout studies, mice lacking caspase-2 had a normal life expectancy of roughly 900 days, and physiological abnormalities did not become apparent until the animals were about 450 days old.
Clinical treatment with a caspase-2 inhibitor would not completely abolish caspase-2 activity. "In our studies, we did not need to lower the activity very much," she said. And "drug companies like proteases – they are relatively conventional targets."
Protease inhibitors are mainstays in antiviral treatments, and are being explored in many other indications as well.
Caspase-2 has also been implicated in Huntington's disease and in neurodegeneration in amyloid precursor protein (APP)-driven mouse models of AD, but the enzyme does not cut either mutant huntingtin or APP. In their paper, Ashe and her colleagues wrote that "an intriguing question is whether the cleavage of tau by caspase-2 mediates APP- and huntingtin-induced synaptic deficits."
Another advantage of going after caspase-2, she said, is that it targets a process that occurs late in AD.
Some form of amyloid beta, "probably not the plaques themselves," triggers abnormalities in tau, but tau pathology ultimately becomes unlinked from amyloid beta.
"We're not making any assumptions about what's triggering the problem," she said. "What we're trying to do is block tau from going into the dendritic spines, where it causes synaptic dysfunction."