FEAR CONDITIONING can be brought about by pairing a sound and a mild electric shock to a rat's foot (top). After several such pairings, the rat becomes fearful from hearing the sound alone (bottom).

Jennifer T. was absolutely certain during the trial: This man was her rapist. She had studied his features when the crime happened, and had later identified him both in pictures and in person. Based on her testimony, Ronald C. was sentenced to life in 1986. Eleven years later, though, a DNA test revealed that another man, named Bobby P., was the real attacker.

The reversal made headlines in June this year in several major papers, including the New York Times. Most asked the same thing: How could Jennifer T.s memory have fooled her so completely? Now, a new study from New York University published in the August 17 issue of Nature may provide an answer.

The study looked at so-called long-term memories in rats. Most memories are in fact short-lived and easily forgotten. It is only through a process called "consolidation" that some make the transition from short-term to long-term storage. And this process requires not only structural changes in nerve cells but the synthesis of new proteins in these cells. Scientists long believed that after this consolidation, memories were pretty stable and not easily disturbed. As it turns out, they were wrong.

The new work reveals that at least one form of long-term memory--that of fearful experiences--seems to become fragile when reactivated. Joseph LeDoux and his colleagues conditioned rats much in the same way Pavlov trained his famous dog--while the animals heard a tone, they received a mild electric shock to the foot. The scientists found that after only one training session, the tone itself was enough to make the rats freeze with fear; they remembered and anticipated the unpleasant shocks.

Once the animals were so conditioned, the researchers played the tone again. Now, though, at the very moment in which a rat remembered the shock, the scientists injected an antibiotic that inhibits the synthesis of new proteins into its amygdala--a part of the brain long known to store lasting memories of fearful experiences. And when the tone sounded again a day later, most rats treated with the antibiotic did not react to it anymore; they seemed to have lost their memory of the shock.

Of interest, this fear amnesia occurred only when the infusion and the memory reactivation happened at the same time. If the rats received their dose of inhibitor six hours after they had remembered the shock, or when the researchers did not activate the memory at all, their reaction to the tone remained unaffected. Also, the scientists could erase the memory equally well one day or 14 days after the initial training; the power of the antibiotic seemed to have nothing to do with how old the memory was.

Although it isn't entirely clear what these results mean, the researchers speculate that each time a fearful memory is revived, it again enters a labile state, and protein synthesis is necessary to "reconsolidate" it. In the absence of new protein production, it may be lost permanently.

That said, it is conceivable that while a memory is in the "limbo" that LeDoux's work documents, new impressions might mix in with it before it settles again, creating what is known as false memory syndrome. Perhaps while Jennifer T. recalled her ordeal, the memory of it became vulnerable. Looking at photos of Ronald C. then incorporated him into that memory, making her believe that he was the man who attacked her.

Scientists must still discover whether memories other than fearful ones undergo a similar kind of reprocessing when they are retrieved. Whatever the outcome, though, the results may be clinically useful in helping people with post-traumatic stress disorder shake their haunting memories once and for all.


EMOTION, MEMORY AND THE BRAIN, by Joseph E. LeDoux. Scientific American, June 1994. (This article is available for purchase at Scientific American's Online Archive.)