Researchers identify brain region that deciphers between new and old memories

Have you ever walked down the street and seen someone you weren’t sure if you knew or if they just looked familiar? Whether that man was a high school classmate or just someone that looked like him is all up to a specific part of the brain that creates and processes memory, found researchers from Johns Hopkins University.

In a new study published in the journal Neuron, researchers identified the part of the hippocampus that deciphers a memory from an entirely new experience. The hippocampus, part of the brain’s limbic system, is necessary to form episodic memory, or memory of an event. Shrinkage of this area is severe in patients with advanced Alzheimer’s disease; forgetfulness is one of the first symptoms of the disease.

The theory is that the brain stores different parts of experiences— such as where and when it occurs— in different parts of the memory. For example, in thinking about what you had for breakfast this morning, one set of brain cells stores the sight of the bowl, another the smell, another the taste, another the time of day, what you may have been reading at the time, and so forth.

If someone asks what you had for breakfast,  you may remember what you were reading, and the theory goes that the cells reactivate to remind you what you were eating, what it smelled like and tasted like.

“The connections between [brain cells] are strengthened during the actual experience,” James J. Knierim, a professor of neuroscience at the Zanvyl Krieger Mind/Brain Institute at the Johns Hopkins University, told

However, you may remember one part of the memory, but not another part. For whatever reason, the connections of the memory that you remember were strengthened, or perhaps overridden by later memories.

“This is one of the critical problems the brain has to solve,” Knierim said. “A lot of our experiences have a lot of overlapping components.”

The  process of recognizing the now unfamiliar face is called pattern completion, where the brain collects the rest of the cell activity in order to retrieve the full memory. If you do not know that person, the brain must create a new set of cells to encode the memory of this new person. This process, pattern separation, helps you avoid repeating the same accidental false recognition again.

“The brain has to make a decision, given ambiguous input: Is this the person I knew before and I’m recalling or is it someone who looks like that person?” Knierim said.

Researchers identified that the CA3 region of the hippocampus is responsible for these memory decisions. Previously, the theory was that the CA3 region acted as the judge of whether something was the same or different. The study found that the mechanisms are more nuanced; that different parts of the CA3 region are biased toward making opposing decisions and they pass on these different decisions to other brain areas.

“It seems to be that some parts of the brain want to be biased to say, ‘This is the same person,’ and, ‘This is a different person,’” Knierim said, adding that it’s still unclear why this happens.

Researchers recorded the activity of brain cells in the hippocampi of rats in a test environment. Over 10 days, the rats built mental maps of their surroundings. The team then altered the space so items were mismatched from their original locations and found that the “pattern separating” part of the rats’ CA3 created a new memory of the altered environment. But, the “pattern completing” part of CA3 tended to retrieve a similar activity pattern that encoded the original memory. The CA3 was acting toward both decisions.

“The experience was analogous to you one day going into your office and someone rearranged the environment. Your brain has to make the same kind of determination: Is it really my office and altered, or did I open the wrong door in the wrong place?” Knierim said.

Their finding may help understand how memory is affected in diseases like Alzheimer’s.

“This basic science will allow us to understand why would a certain therapy work better, a certain drug work better,” Knierim said. “It’s something that doesn’t have a direct implication right now… but will allow us eventually go come to those kind of treatments.”

One issue with Alzheimer’s research is that by the time memory symptoms occur, the brain damage is already done. A big push now is to figure out whose brain is already being affected, before symptoms start, Knierim said.  To do this, researchers could use imaging scanners to measure brain activity in non-diagnosed patients and look at the hippocampus activity.

If researchers observe low activity during recognition tests, they could start therapies in the hopes of preventing brain damage— and stop Alzheimer’s from developing.

“Maybe the kind of work we’re doing will lead clinicians to a good idea of how to test for subtle signs of hippocampus damage, before the disease progresses,” Knierim said. “There’s still no treatment, but that might be the root necessary to treat disease before it causes all that damage.”