Login to Your Account

Researchers alter memories: Bad to good and vice versa


By Anette Breindl
Science Editor

"Recalling a memory is not like playing a tape recorder," Susumu Tonegawa told reporters. "It is a creative process."

Normally, that creativity comes from the organism – man or mouse – doing the remembering. But Tonegawa, who is at the Massachusetts Institute of Technology, and his colleagues have managed to be creative with mouse memories, by altering the association between so-called episodic memories and the emotions that accompanied them.

By artificially reactivating a negative memory trace using optogenetics, and pairing it with a positive experience, the authors were able to pair the content of an old memory trace with a new emotion – essentially turning bad memories into good ones, and vice versa.

Senior author Tonegawa, first author Roger Redondo, and their colleagues published their work in the Aug. 28, 2014, issue of Nature.

In previous experiments, the authors had managed to give an emotional value to a memory that had originally been neutral. The new work extends that by changing the emotional tone of a memory, making fearful memories happy and vice versa.

The approach works because the factual content of a memory – for example, having been in a certain place before – is stored separately from its emotional content.

"In our day-to-day lives," Tonegawa explained, "we encounter a variety of events and episodes that give positive or negative impact to our emotions. . . . Emotions are intimately associated with past events."

For example, if a person is mugged in an alley, the memories of the alley will be aversive, and the alley will be a place where that person will not wish to return.

Using footshocks rather than mouse muggings, Tonegawa and his team essentially demonstrated that in such a case, the fear of the alleyway is stored in a separate brain region from factual memories of the alleyway, such as where it is located, what it looks like and the memory of having been there.

The authors developed a method for artificially activating the parts of the memory stored in either the hippocampus or the amygdala, and pairing that activation with a new positive or negative experience.

Pairing activation of a previously negative hippocampal memory with a pleasurable experience led mice, over time, to remember the memory as being pleasurable. "They can link up with either fear or pleasure," first author Redondo told reporters.

The amygdala memory neurons, on the other hand, "did not switch their response," Redondo said. "We could not get them to switch the type of response they were eliciting. And this suggests that they are dedicated to producing one type of emotion or another."

"We strongly believe that the amygdala comes with two types of cells that are prewired during the development of the organism," he said. "We don't have markers of these two cells at the moment," but work to identify such markers is ongoing.

Overall, Redondo said, the results led the team to develop a model where "in the brain, there exist competing neuronal circuits" for memory storage.

The hippocampus stores episodic memories, "but does not store the valence associated with an experience." That emotional valence is stored by two types of amygdala cell populations – one involved in fear and one in positive memories.

Overall, he said, "competition mediated by plasticity of the connection of these circuits dictates the overall emotional value and the direction of valence" of a given memory.

And that competition can be manipulated.

"With these new optogenetic tools that we . . . and others have developed . . . now we can go inside the brain and manipulate this circuit and change the way the mouse reacts to a memory being brought back to mind without any drugs – there is not a single drug used in this study," Redondo said. "And this memory manipulation occurs without the mouse ever being brought back to the original place where the memory was formed initially."

Given that the precise activation of memories in the mice took a mix of gene therapy and implanted devices, for now the work is firmly rooted in basic science, and practical applications of the findings in humans are "very much [in] the future," Tonegawa said.

"But now we know where the critical connection is between the hippocampus and the amygdala. So in the future – and it is, I emphasize, in the future – the critical site where this valence switch occurs could be the targets for drugs or some type of therapies" such as deep or surface electrical stimulation.

"It may be a lengthy process," Tonegawa said. But "technology develops so fast these days, and I'm quite optimistic that eventually, something like this will be done."