A full-fledged Alzheimer's mouse.

John D. Gearhart's latest litter of mice is just a few weeks old, but already he has high expectations for them. He hopes the mice will provide a model of Alzheimer's disease, a neurodegenerative disorder. Other investigators pursing such a model, which would help test potential therapies, have had only limited success so far.

Gearhart, of the Johns Hopkins University School of Medicine in Baltimore, and his colleagues believe they may have had better luck because of a twist they've added to the common strategy for making an Alzheimer's mouse.

To construct a mouse model of an illness, researchers add to the animal's genome a gene they believe may be responsible for the disease. In the case of Alzheimer's, attention has focused on a gene that in humans directs the synthesis of a molecule called amyloid precursor protein, or APP. The brains of people who had Alzheimer's often display dense plaques of a protein called beta-amyloid, a fragment of APP. Many investigators therefore believe that extra or abnormal APP production may cause the disorder.

The APP gene, however, is huge, much bigger than genes researchers can easily put into the mouse genome. As a result, notes Gearhart, most

researchers have taken a shortcut, inserting only severely abbreviated versions of the APP gene that contain little more than the minimum amount of DNA necessary to make the protein.

But last year, Gearhart and his colleagues successfully added to the mouse genome the full version of the human APP gene and the DNA that normally surrounds it, an amount of genetic material about 60 times greater than researchers have added in previous attempts. They have now extended their work, creating new strains of mice with full-sized APP genes that include mutations found in some of the inherited forms of Alzheimer's disease.

"I think it's the only paradigm that can mimic the human system," says Gearhart, who must wait to see if the mice develop Alzheimer's as they age.