Today I met Daphna Shohamy. Will I remember who she is if I run into her later for dinner? If I see her, I probably won’t relive her morning talk but memories of CNS likely (hopefully!) will come to mind to help me remember who she is. This is how Shohamy likes to explain memory reactivation, a process that many scientists believe is critical to not only long-term memory formation but also for everyday decision-making.
New work, presented this morning at a session chaired by Lila Davachi (NYU) at the CNS annual meeting in New York, is elucidating how memory reactivation contributes to memory persistence and integration into our everyday lives. Shohamy of Columbia University focused on how that process affects decision-making, while Ken Paller of Northwestern University, talked about the influence of sleep.
Memories to use
“To guide decisions, memory must be useful,” Shohamy said. The memories need to be adaptive and relevant but the challenge is that we don’t often know in the moment which memories we’ll need to use later on.
Take, for example, going to Starbucks or your regular coffee shop: You encode tons of memories about which drinks you like, the organization of the shop, the baristas’ personalities, etc. And you will then draw from your memories of those experiences if you go to a new coffee shop, Shohamy explained. So how does your brain on a daily basis prioritize which memories to reactivate to help guide decisions in future (unpredictable) scenarios, like in a new coffee shop?
A growing body of work suggests that reward plays a huge role in this process. Several studies have used tasks during which people form associations, learn about rewards, and then make decisions that require integration of this information. For example, a photo is paired with a colorful shape and participants learn that the shape leads to reward. The participants then come to favor the associated photo through its intial link with reward.
This automatic and unconscious process places values on how the brain decides which memories to reactivate, Shohamy said. “Reactivation allows rewarding events to change value of memories.”
The integration of information happens, she said, before decisions are faced. And it happens via connectivity between the hippocampus and ventromedial prefrontal cortex (vmPFC), based on work correlating brain activity (recorded with fMRI) with the decision-making behavior.
To help understand which memories are strengthened by reward, another experiment looked at participants navigating a maze, where some mazes led to reward and others not. The researchers found that with the mazes that offered reward, people most remembered the objects geographically close to the rewards in the maze but only after 24 hours. They need that time period for the reactivation to be successful, Shohamy said: “Reward retroactively enhances memories for sequences of events but only after reactivation.”
In general, the memory reactivation work suggests we create adaptive models of world; these models are biased by what’s important and relevant for us, Shohamy said. Future work is focusing on the circuit dynamics that support the integration of information and the role of dopamine in memory consolidation.
Memories to sleep on
Your sleep determines what you know and who you are. -Ken Paller, CNS 2016
You’ll forget most of what you do today. Key to what you remember is what you rehearse, and much of that rehearsal happens in your sleep. That’s how Ken Paller kicked off his talk about the role of sleep in memory reactivation.
Sleep, he said, is far from trivial in our lives: Your sleep influences what you know through your memories and therefore determines who you are.
So the question becomes: Can we use sleep not only to better understand how memories get stored but also to target what we want to remember?
The answers may very well lie in the hippocampus, which affects acquisition, storage, and retrieval of memories, Paller said, which we know, in part, through work done in patients with amnesia. His and others’ research has shown a way to do “targeted memory reactivation” through sensory cues, like odors or sounds, played in sleep after pairing the cues with tasks while awake.
The work represents a new way of thinking about sleep, overturning previous dogma in the field that sensory signals are blocked in sleep. While the signals are not as strong in sleep as when awake, scientists now know they are not blocked, Paller explained. It also builds off work done on memory while awake and focuses on slow wave sleep rather than REM.
“Entering slow wave sleep is not like shutting down a computer,” Paller said. High neuronal activity in slow wave sleep can enable brain interactions important for memory consolidation.
Among many findings, researchers have found better memory retention after sleep compared to no sleep, better retention the more slow wave sleep someone has after learning, and that sensory cues during sleep can provoke memory reactivation. For example, researchers had people remember the placement of pictures while listening to various sounds (like a “meow” for a picture of a cat) and then played the sounds while participants took a nap. They found that there was more forgetting for the uncued photos after the nap than for those that were cued with the sounds played while the participants slept.
But it’s not like in the Dexter cartoon, where playing a record during sleep will suddenly imbue the listener with new knowledge, of say a foreign language. Pre-sleep accuracy of the material being learned matters. Memory reactivation during sleep, Paller said, is like when you are awake except that when awake, you can incorporate new information to update your thinking. In either case, it’s the hippocampal interactions with distributed neocortical networks that changes how the information is stored.
Paller toured the audience through a variety of studies using sensory cues for target memory reactivation, including for fear conditioning with odor cues and even targeted reactivation to help reduce implicit bias (also see the post on Phelps’ work on this). Importantly, he said, the work with sensory cues in sleep has not yet indicated any deleterious effects on people’s sleep.
He said he sees a great future in more work on memory reactivation during sleep, including finding ways to reactivate the memories we really want to store and ways to help rehabilitate people with motor or language deficits. He also speculated that perhaps treatment of psychiatric disorders, as well as reinforcing good habits or countering bad ones (e.g. smoking, unhealthy eating), could be supplemented by targeted memory reactivation in sleep.
-Lisa M.P. Munoz