In the 1960s, two University of Pennsylvania psychology graduate students discovered that when dogs received electrical shocks that they could not control, they later showed signs of anxiety and depression, but when dogs could end the shocks by pressing a lever, they didn't. What's more, the dogs that received the uncontrollable shocks in the first experiment didn't even try to escape shocks in a later experiment, even though all they needed to do was jump a low barrier.

The two researchers—Martin E.P. Seligman, PhD, and Steven F. Maier, PhD—termed their discovery "learned helplessness," and their findings are now a staple of introductory psychology textbooks. Seligman went on to further explore the finding, while Maier went in a different direction, retraining as a neuroscientist and studying the effects of stress on the immune system.

But 30 years after the experiment, Maier found himself thinking about that work and wondering if he could find a neural circuit for learned helplessness. With help from students and colleagues at the University of Colorado, where he's a psychology and neuroscience professor, Maier succeeded—and his findings suggest that the dogs from that early experiment were not, in fact, learning helplessness. They were failing to learn control.

"The default position of the brain is to assume that stress is not controllable," he said.

Ancient structures respond

To begin their search for the brain basis of learned helplessness, Maier and his colleagues had to identify a part of the brain that facilitates activation in the amygdala, which plays a major role in fear and anxiety responses, but that inhibits activation in the dorsal periaqueductal gray matter, which triggers fight or flight responses. A review of the literature turned up the dorsal raphe nucleus (DRN) as a likely candidate, since that cluster of neurons in the brain stem releases serotonin into the forebrain and limbic systems as well as the neighboring periaqueductal gray.

To explore the DRN's role in learned helplessness, Maier and his colleagues ran an experiment where they exposed rats to either controllable or uncontrollable tail shocks. The researchers measured the adult rats' DRN serotonin levels throughout the experiment and found that all of the animals' levels spiked when they were first exposed to the shock. But as soon as the rats learned to control the shock by pressing the levers, their serotonin levels dropped.

After the procedure, the researchers placed an unfamiliar juvenile rat in the cage of the rats that had been through the uncontrollable or controllable shock procedures. Usually, adult rats will sniff a juvenile rat, and that's what the animals that experienced controllable shock did, but the rats that had been through the uncontrollable shock procedure cowered in their cages and did not explore the newcomer. Their DRN activation also spiked and stayed high throughout the social stress test, while the other rats' DRNs stayed calm.

"It was the release of serotonin that was responsible for these behavioral effects," concluded Maier.

The mystery, however, was not solved. Past research shows the DRN, which resides in the ancient brain stem, is not smart enough to know whether stress is controllable or not—it just responds to stress in general. Some other part of the brain, said Maier, must be giving the DRN its instructions.

'Cool it, brain stem'

That area, according to research by Maier and his colleagues, appears to be the ventromedial prefrontal cortex (vMPC), a part of the mammalian brain's frontal lobe. In a series of studies published last year, Maier and his colleagues found that when they deactivated the vMPC while animals received controllable shocks, the DRN stayed active, and the rats later showed the signs of anxiety and depression you'd expect only if they had not been in control of the situation. Also, when the researchers activated the vMPC in rats receiving uncontrollable shocks, DRN activation dropped off, and the animals did not show later effects of learned helplessness.

"This is the illusion of control at the level of neurochemistry," Maier said.

Taken together, the findings suggest that in the face of any stressor, the DRN activates the body's ancient stress responses, but if that stressor turns out to be controllable, the vMPC steps in and calms the DRN's response. "It's like the forebrain is saying, 'Cool it, brain stem, we have the situation under control,'" Maier said.

Looking back on his early research, Maier now realizes that the dogs in his seminal study were not learning helplessness, they were just staying in their natural state. Only with training and input from the vMPC, which evolved later than the DRN, do animals learn to relax when a situation is under control.