THURSDAY, Aug. 24 (HealthDay News) -- New research in mice points to a possible treatment for Alzheimer's disease, one that repairs brain cells so they can rid themselves of the amyloid beta proteins that are suspected of contributing to the disease.

By tinkering with an enzyme in the brains of mice afflicted with the rough equivalent of Alzheimer's in humans, scientists were able to improve the rodents' memories.

There's no way to know if the approach will work in humans, but researchers are hopeful. In essence, "we were able to restart -- or make more efficient -- the garbage disposal function of the cell," said study co-investigator Dr. Michael Shelanski, director of the Alzheimer's Research Center at Columbia University.

Despite intensive research, Alzheimer's disease remains very difficult to treat and affects about 4.5 million Americans, a number that is expected to soar in the coming decades as the population ages. Some drugs are available, but their effectiveness is limited, Shelanski said.

According to Shelanski, the new research expands on previous findings that suggested a shortage of an enzyme in the brain may be connected to Alzheimer's. The enzyme, known as Uch-L1, appears to be crucial to a cell's ability to get rid of malformed proteins and maintain memory.

"It has been widely assumed in Alzheimer's disease that a lot of the problem comes from the fact that proteins within the cell do not fold correctly, and therefore don't function correctly," Shelanski said. "The cells' garbage disposal tries to get rid of them."

Researchers engineered mice to suffer from a kind of rodent Alzheimer's disease. Then they boosted the level of the enzyme in the mice to see what would happen.

The treated mice were able to remember to avoid parts of a cage floor where they had earlier been exposed to a mild stimulus. The other mice forgot the places to avoid, the study found.

The study findings appear Thursday on the Web site of the journal Cell.

"We were able to greatly improve function in mice that had (the equivalent of) Alzheimer's disease for as long as a year," Shelanski said. That could reflect changes that would happen in a human who had Alzheimer's for many years, he added.

Will it work in humans? No one knows. One downside is that the treatment must be injected, not taken as a pill by mouth. And it might not be safe in people.

"This is hope, not proof, that even people with established Alzheimer's disease might be able to get some recovery from a therapy that works along these lines," Shelanski said.

Another Alzheimer's specialist called the study "interesting, exciting, and important."

The research points to "a previously understudied area of investigation that may be useful for developing drugs for Alzheimer's disease," said Dr. Sam Gandy, director of the Farber Institute for Neurosciences at Thomas Jefferson University in Philadelphia.

He added that the new research relies upon the most commonly accepted theory about what causes Alzheimer's, although it still remains to be proven.

Unfortunately, he said, it will take years to develop a drug for people.