Skip secondary menu

A Rollercoaster of Seizure-Like Activity May Damage the Alzheimer's Brain


For release: Tuesday, November 27, 2007

Although seizures are not a common symptom of Alzheimer's disease (AD), the brains of people with AD could be humming with seizure-like activity, interrupted by quiet rebound periods that do more harm than good.

In a new study published in Neuron*, researchers observed this pattern of brain activity in a mouse model of AD.  The study's lead investigator, Lennart Mucke, M.D., Director of the Gladstone Institute of Neurological Disease at the University of California in San Francisco, now plans to determine whether the same activity contributes to memory loss in people with AD.   If so, it's possible that anti-seizure medications might slow the disease, for which there is currently no cure.

Dr. Mucke's research, which was funded in part by the National Institute of Neurological Disorders and Stroke (NINDS), also helps resolve doubts about amyloid-beta – the protein fragment that accumulates in the brains of people with AD.  Amyloid-beta is widely believed to be the causative agent in AD, but even experts cautiously refer to this idea as the "amyloid-beta hypothesis," partly because they're unsure how the accumulation of amyloid-beta leads to memory deficits.  The new study not only provides a possible explanation, it also shows that certain changes in the brains of mice with AD are correlated with the level of amyloid-beta the mice produce.

Amyloid-beta is a snippet of a larger, normal protein called the amyloid precursor protein (APP).  Although the reasons for amyloid-beta accumulation are complex and poorly understood in most cases, a rare, early-onset form of AD has been linked to genetic mutations in APP that render it more susceptible to the action of protein-cutting enzymes.  The mice studied by Dr. Mucke were genetically engineered to carry these mutations in human APP.

The same mouse model has been used in countless previous studies, too.  Researchers have examined the mouse's hippocampus – a brain region involved in memory and known to shrink in AD – and found signs that it's inefficient at encoding new memories.  They've also tested the ability of the mice to learn mazes, which the mice do poorly.  Dr. Mucke and his collaborators tried something new: using electrodes to record electrical activity over the surface of the whole brain, called electroencephalography (EEG), as the mice roamed around their cages.

The EEGs revealed "epileptiform" activity – recurrent, abnormal spikes in brain activity reminiscent of those seen in epilepsy.  Occasionally, the spikes intensified and spread across the brain, creating the barcode-like pattern of brain activity that's characteristic of a seizure.  For a short time after the seizure, brain activity fell below normal. 

"We saw very frequent fluctuations between over-excitation and over-inhibition," says Dr. Mucke.  During these highs and lows of brain activity, the mice never developed convulsions or other obvious signs of a seizure; they typically sat grooming themselves or just paused briefly in their exploration of the cage.

But Dr. Mucke and his team did see changes in granule cells – a type of cell in the hippocampus – consistent with a see-sawing of excitation and inhibition.  As in mouse models of epilepsy, the granule cells of APP transgenic mice turned on proteins that are "hallmarks of over-excitation," says Dr. Mucke.  (The increased protein levels paralleled the amount of amyloid-beta produced by different strains of APP transgenic mice.)  But unlike in chronic epilepsy, there was a rampant growth of nerve endings known to inhibit the activity of granule cells.

Dr. Mucke thinks the growth of these inhibitory connections is a unique repair response to seizures that ends up backfiring.

"In epilepsy, the inhibitory system fails, but here it becomes over-active," he says.  "It probably prevents the seizures from escalating into convulsive seizures, but it may also interfere with normal excitatory processes required for memory formation and learning."

Epileptiform activity has been observed in as many as 10-20 percent of AD cases.  Dr. Mucke and his collaborators are investigating the possibility that it's more common, but has gone unnoticed because no one has had reason to look.  They're also beginning to test anti-seizure medications in the APP transgenic mice.  Since many of these medications are already approved by the Food and Drug Administration for use against epilepsy, they could be fast-tracked through clinical trials for AD.

*Palop J et al. "Aberrant Excitatory Neuronal Activity and Compensatory Remodeling of Inhibitory Hippocampal Circuits in Mouse Models of Alzheimer's Disease."  Neuron, September 6, 2007, Vol. 55, pp. 697-711.

-By Daniel Stimson, Ph.D.

Last Modified December 5, 2007