Laboratory rats are different from you and me, as F. Scott Fitzgerald might have said, were he a neuroscientist. Rats make good models for Pavlovian fear conditioning, but trying to derive an understanding of human behavior from them is complicated by our messy cognitive and social environment, said Elizabeth Phelps, Ph.D., a professor of psychology and neuroscience at New York University. She spoke at the sixth annual Amygdala Conference, held at the Uniformed Services University of the Health Sciences in Bethesda, Md., in April.

In her lab, Phelps tries to understand the acquisition, storage, and expression of fear in humans by extending information gathered from the study of neural systems in rats.

Basic fear-conditioning experiments on humans are not that different from those for rats. The rodents learn to associate an electric shock with some harmless stimulus, such as a sound, so that when they hear the sound later, they react as if shocked.

Working with nonrodent species, college undergraduates, for example, Phelps presents a blue square accompanied by a mild shock. Later, the students exhibit a physiological fear response when shown the blue square in the absence of the shock.

So far, so good. But humans' higher cognitive abilities appear to both hurt and help them, depending on circumstances. For instance, simply telling the students about the shock associated with the blue square is enough to generate a fear response. That may explain why people don't have to be physically hurt to develop stress disorders.

However, humans may have some advantages when it comes to getting rid of fears.

The primary mechanism for doing so is extinction. Repeated exposure to a learned fear stimulus (like the sound for the rats and the blue square for the students) with no aversive consequences (the shock) results in a diminished fear response in both species, said Phelps.

"This is the ‘getting back on the horse’ technique," she said. "Actually, the fear doesn't go away, but there is new learning that says the stimulus is safe. There are two different, competing memories, the safe and the not safe."

The amygdala is involved in the acquisition of both fear memory and the extinction of fear. Imaging studies show an increased response in the amygdala during the acquisition of fear and a decreased response during extinction.

Evolution has, however, conferred some advantages on humans. For one thing, they can think about what's going on and regulate their emotions.

"We can reevaluate the stimulus associated with negative affect into neutral or positive terms," she said. This is the basis of cognitive-behavioral therapy. "When using thoughts to control emotions, you are co-opting the methods of extinction."

Fear, like other memories, can return, which has drawn attention to the process of memory retrieval. The old view was that remembering something was like pulling a folder out of a file cabinet, flipping through it, and returning it to its place, unchanged. The current view (termed reconsolidation) is that a memory is retrieved, updated, and then returned to storage somewhat modified.

Rats go through the same process, which can be controlled by injecting a protein-synthesis inhibitor directly into the amygdala, which interrupts consolidation of fear memories. The technique cannot be applied in humans, of course, and attempts to reproduce the effect with oral propranolol have been unsuccessful, she said.

However, humans may be able to update a memory with new information about the memory or the stimulus that includes safety information. Phelps and her colleagues have shown that extinction training during reconsolidation can reduce fear.

How big is the window for this safety learning? Phelps reported on experiments showing that safety learning that takes places 10 minutes after memory retrieval can produce an effect that lasts up to one year. Training immediately after retrieval or six hours later, however, had no effect. But that's where the research stands at the moment.

"Only three papers have been published on the matter of timing," said Phelps. "The basic science is so new that translating this research to the clinic is premature."

Human responses to both fear and extinction are also complicated by our rich social environment and our symbolic means of communication, leading to a flexibility in humans that makes research more difficult than it is in rats.

"How do you reactivate the right memories?" asked Phelps. "Just knowing something occurred can drive fear memories. You can get rid of the extinction memory too, so you can reverse progress. Anyway, it's the natural process of the brain to rework memories. It happens constantly."