By David Cyranoski

Zebrafish that stop swimming when left without company are showing promise as the first fish model of a human mood disorder.

In 2005, when neurobiologist Herwig Baier of the University of California, San Francisco, was screening thousands of zebrafish for vision problems, he found one that seemed a bit "off." If alone, especially after repeated periods of isolation, the fish would "freeze": just sit at the bottom of the tank (see video). If fish that swum normally were put in the tank, the relatively inactive fish became normal and swam around too.

Baier looked at the genetic mutations in the "frozen" fish and found one in the glucocorticoid receptor, a protein that is found in almost every cell and that senses cortisol--a hormone involved in the stress response. In the normal response to a stressful situation, the hypothalamus in the brain sends corticotropin-releasing hormone (CRH) to the pituitary gland, which releases adrenocorticotropic hormone (ACTH) to the adrenal gland. The adrenal gland in turn produces cortisol. Cortisol then effectively reduces levels of ACTH and CRH, completing the normal response that allows both humans and zebrafish to deal with stress.

In the frozen fish, however, Baier found that levels of all three hormones--CRH, ACTH and cortisol--were higher than normal. He guessed that the animals were unable to respond properly to chronic stress--a problem that is known to trigger anxiety or depression in humans. On the basis of that diagnosis, he started putting the antidepressant fluoxetine (originally marketed as Prozac) in their water. After four days, they started swimming around normally. Other antidepressants and anxiolytics--drugs used to treat anxiety--also worked as a pick-me-up, he says. "There's a long literature on chronic stress being related to depression, but the causal link is unknown," says Baier. "Now we might be able to simulate this in fish and study it."

Mutant marvel

Discussing his results at the Society for Neuroscience meeting in San Diego, Calif., last week, Baier says his mutants could represent the first fish model for a mood disorder and be a useful screening model for drugs.

"The fact that the key elements of stress and stress response are conserved in zebrafish is exciting because of the many experimental advantages of that model organism," says chemical geneticist Randall Peterson of the Massachusetts General Hospital, Harvard Medical School, Charlestown.

Studies this year support the idea that zebrafish can model complex behavior. Earlier this year, Peterson and his collaborator, neuroscientist Alex Schier of Harvard University in Cambridge, Mass., published the first two small-molecule screens based on zebrafish behavior. The studies focused on motor behavior, not mood disorders, but the team found that certain classes of psychotropic drugs, such as the antidepressants known as monoamine oxidase inhibitors, had recognizable, characteristic effects on behavior. "So, the idea of discovering new therapeutic approaches for anxiety or depression may not be so far-fetched," says Peterson.

If that proves true, the finding could accelerate drug-discovery programs. Chemical biology and drug discovery usually depend on screens of cells in lab dishes, for example. "But much of biology can't be easily reduced to an in vitro assay. This is particularly true for nervous-system disorders, where a complex, integrated nervous system is essential for research," says Peterson. Filling thousands of tiny wells on lab plates with zebrafish larvae and dousing them with candidate drug molecules offers the best of both worlds--high-throughput screening in a living system. "When we discover a new small molecule, we have the advantage of knowing that it works in vivo," he says.

Larval question

Baier plans to do the same sort of screening for new antidepressants and anxiolytics using his "frozen" fish. But Peterson does sound a note of caution. Baier's experiments used adult fish, whereas most zebrafish screens use larvae. Using adult for screening would be expensive and laborious. Baier plans to use larval stages but he admits it is not clear yet whether the larval zebrafish will react in the same way as the adult mutants.

The model will also have to convince the pharmaceutical industry that fish depression has significant similarities with the condition in people. This has not always been the case with mice, which as mammals should be a closer match to humans. "Considering the challenges of using rodents, including genetically engineered mice, to validly model human psychiatric diseases, zebrafish probably have some way to go before they are accepted as a translational model," said conference delegate Jeffrey Kogan, a behavioral neuroscientist working within the pharma industry who studies psychiatric disease.

It might, however, just take some time to sink in. In an e-mail to Nature a few days later, Kogan said that zebrafish might be a useful model organism for psychiatric disease after all. The huge numbers of fish that can be used in such studies would, for example, give the zebrafish mutant an advantage over the mouse. "I may have to reconsider my opinion," he says.