SELF-INFLICTED WOUNDS: By deleting a single gene, researchers at Duke University Medical Center stumbled onto a possible mouse model for OCD.
© JING LU, JEFF WELCH & GUOPING FENG

A research team studying brain signals in mice accidentally stumbled upon what could be an important discovery that could lead to understanding and successfully treating obsessive-compulsive disorder (OCD).

The finding identifies a new potential target for treating the psychological syndrome, which affects some 2.2 million Americans and is characterized by symptoms including anxiety and excessive behavior such as repeated hand washing and pulling out one's own hair.

Researchers at Duke University Medical Center made the discovery after deleting a gene in mice while studying neuronal communication in the striatum, a structure in the midbrain that plays a role in information processing, decision making and movement. They had set up 24-hour video surveillance of the critters in their cages after the animals developed skin lesions on their heads and necks four to six months after their birth.

"These mice stay by themselves and are grooming themselves all the time," says Guoping Feng, an assistant professor of neurobiology at Duke and co-author of a report on the findings published in Nature. He says the mice also show telltale signs of anxiety, hewing to the sides of their cages and staying out of both bright and open spaces.

"We were not specifically looking for OCD … the phenotype itself is by accident," notes Feng. But, the serendipitous discovery shows "how synaptic dysfunction can lead to abnormal function."

Feng and his team had wanted to better understand the internal communication among neurons in the striatum and the transfer of information between cells in the striatum and the cerebral cortex (the outer layer and central processing unit of the brain). The scientists targeted the gene SAPAP3 for deletion. SAPAP3 codes for a protein that affects the building blocks of synapses (the communication portals between neurons). Specifically, the protein works by recruiting receptors for the neurotransmitter glutamate, a chemical messenger involved in learning and memory, to migrate into the synapse for more efficient communication between cells.

The current treatment for OCD, which only works in about 50 percent of the cases, includes giving patients certain selective serotonin reuptake inhibitors (SSRIs) such as Prozac, which improve the transmission of the neurotransmitter serotonin in synapses. Feng and his colleagues administered the drug for six days to the perpetually self-grooming animals, during which time their obsessive behavior attenuated and their anxiety abated. "It's very interesting that serotonin has some effect on these mice," he says, noting that the neurotransmitter may also modulate the activity of glutamate in neighboring synapses.

Similarly, when seven-day-old transgenic mice received injections of a vector containing the missing SAPAP3 gene directly into their striatum, they did not develop the OCD symptoms. Taken together, "this is the first direct evidence that cortico-striatal communication is critical to this OCD-like behavior," Feng says.

Apparently, blocking this protein's formation in the striatum is enough for the OCD symptoms to be manifest. Feng points out, however, that he and his team cannot be sure that this is the sole model for studying the syndrome; no cause is known for the disorder, but it is believed to result from multiple genetic flaws or disruptions.

"Of course, we cannot say whether the mice have obsessive thoughts, but in my opinion this does not detract in a major way from the model," says Ann Graybiel, a professor of neuroscience at the Massachusetts Institute of Technology. "The field urgently needs models of complex genetic disorders. The obsessive compulsive-spectrum disorders have been particularly difficult to study because of the complexity and variety of symptoms and syndromes and apparent high incidence of their co-morbidity."

Steven Hyman, a neurobiologist at Harvard Medical School, concurs that the model may be incomplete, but notes that this is a solid first step. "Despite these reservations," he writes in an editorial accompanying the report in Nature, this study "gives us a compelling clue that the compulsive behavior associated with this condition is due to a synaptic abnormality in these neural loops."