Once adult lab mice learn to associate a particular stimulus—a sound, a flash of light—with the pain of an electric shock, they don't easily forget it, even when researchers stop the shocks. But a new study in the December 23 issue of Science shows that the antidepressant Prozac (fluoxetine) gives mice the youthful brain plasticity they need to learn that a once-threatening stimulus is now benign. The research may help explain why a combination of therapy and antidepressants is more effective at treating depression, anxiety and post-traumatic stress disorder (PTSD) than either drugs or therapy alone. Antidepressants may prime the adult brain to rewire faulty circuits during therapy.

Nina Karpova, Eero Castrén and their colleagues at the University of Helsinki's Neuroscience Center created and extinguished fearful behaviors in mice. First, Castrén placed mice in a cage and repeatedly played a tone just before electrically shocking their feet. Soon the animals froze in fear whenever they heard the tone, at which point Castrén put them through "extinction training." He moved the mice to a different cage and played the same tone again. This time there was no electric shock.

Researchers have previously shown that young mice less than three weeks old quickly learn that the tone is no longer a herald of danger and stop freezing in fear. But adult mice are harder to put at ease. Even if the adults become less fearful during extinction training, their relaxation is not permanent—a week later the tone turns them into statues again.

In Castrén's study, adult mice that took fluoxetine while they went through extinction training behaved much like young mice—they lost their fear much faster than mice that were not taking the drug, and their anxiety did not return. In contrast, mice that were given fluoxetine but never went through extinction training remained anxious.

Castrén makes an analogy between these findings and the consensus that antidepressants in combination with therapy are almost always more effective than either antidepressants or therapy alone. Scientists know what most antidepressants do at the molecular level—they change the amounts of neurotransmitters in the spaces between neurons, for instance—but how these changes treat depression remains an open question. Research has not supported the idea that antidepressants treat depression simply by correcting chemical imbalances in the brain. More recently, researchers have hypothesized that depression kills neurons whereas antidepressants like Prozac encourage new neural growth in the brain. Castrén's study suggests Prozac returns regions of the brain to an immature state in which neurons make or break more connections with one another than is typical of the adult brain. In other words, Prozac increases brain plasticity.

Castrén looked for characteristic electrical and molecular signs of plasticity in the brains of mice that received fluoxetine and in those that did not. Specifically, Castrén looked in the amygdala at neural circuits responsible for fear responses. He found that fluoxetine increased levels of a cell-adhesion molecule associated with young neurons and decreased the levels of a transporter protein associated with adult neurons. He also found greater changes in membrane potential in neurons from the brains of mice that had learned to relax. These neurons were also better at synchronizing their communication through a process called long-term potentiation, which is crucial for learning and memory.

"We know that a combination of antidepressant treatment and cognitive behavioral therapy has better effects than either of these treatments alone, but the neurobiological basis is not known," Castrén says. "We show a possible mechanism is bringing the network into a more immature and plastic state."