Brain Protein Linked to Obsessive-Compulsive Disorder, in Mice

DURHAM, N.C., Aug. 22 -- The mystery of obsessive-compulsive disorder may be at least partly solved, with a discovery that implicates a structural protein missing in the brains of certain excessively groomed mice.

Mice lacking the protein exhibit excessive grooming to the point of causing skin damage and increased anxiety, characteristics reminiscent of those in human OCD, Guoping Feng, Ph.D., of Duke, and colleagues, reported in the Aug. 23 issue of Nature.

What's more, treatment with fluoxetine (Prozac) alleviates the symptoms, as it does in about half of human OCD patients, Dr. Feng and colleagues found.

The finding was serendipitous, the researchers said, because they were simply trying to understand the function of brain proteins known as SAPAP, including SAPAP3, which plays a role in the glutamate chemical messenger system.

As part of the studies, they generated genetic knock-out mice lacking the gene for SAPAP3.

The resulting mice appeared normal until they were between four and six months old, when they started developing lesions on their heads, necks and snouts.

Videotapes of the animals showed they were grooming themselves significantly more than wild-type mice even during times when they would normally be asleep (P<0.05, P<0.01, and P<0.001, depending on the time of day).

In a series of psychological tests, the researchers showed that the knockout mice were also significantly more anxious and timid than their wild-type cousins.

For instance, Dr. Feng and colleagues said, anxious mice tend to stick close to the walls of a circular open cage, while less timid animals spend more time exploring the open center.

The knockout mice, over a half-hour test, averaged less than 10% of their time in the center of the test apparatus, while wild-type mice spent between 20% and 35% of the time in that region. The differences were significant at P<0.05 and P<0.01, depending on the time during the test.

"We were surprised by the magnitude of this phenomenon," Dr. Feng said. "The parallels with OCD were pretty striking."

Treating the animals with fluoxetine for six days -- once a day at 5 mg/kg -- significantly reduced the excessive grooming in the knockout mice (at P<0.01) but had no effect on wild-type mice.

To nail down the link, the researchers created a lentivirus that carried either the gene for green fluorescent protein (GFP) or a fusion of the GFP gene and that for SAPAP3.

When the viruses were injected into the striata of knockout mice, those getting the fusion gene and beginning to express SAPAP3 saw the excessive grooming markedly reduced, from about 70 bouts per hour in the control mice to fewer than 30. The difference was significant at P<0.01.

"Since this is the first study to directly link OCD-like behaviors to abnormalities in the glutamate system in a specific brain circuit, it may lead to new targets for drug development," Dr. Feng said.

The research "sharpens our focus" on the frontal-striatal-thalamic circuits as possible sources of OCD, commented Steven Hyman, M.D., of Harvard, a former director of the National Institute of Mental Health.

But in an accompanying viewpoint article, Dr. Hyman cautioned that the main element of OCD that causes distress to humans is unwanted intrusive thoughts, which "cannot be mimicked in mice, at least not in any obvious way."

He added that the researchers may have hit on something "but we must retain a healthy caution."