Kamran Dibba, an anesthesiologist at the University of Michigan, and his colleagues have found that during sleep, some neurons not only replay the recent past but also anticipate future experiences.
To dynamically track the spatial tuning of neurons offline, Mahboudi others We used a novel Bayesian learning approach based on spike-triggered average decoded positions in population recordings from freely moving rats.
“Certain neurons fire in response to certain stimuli,” Dr. Dibba said.
“Neurons in the visual cortex fire when presented with an appropriate visual stimulus, and the neurons we study show location preference.”
In their study, Dr. Dibba and his co-authors aimed to study the process by which these specialized neurons generate representations of the world after new experiences.
Specifically, the researchers tracked sharp ripples, patterns of neural activity that are known to play a role in consolidating new memories and, more recently, have also been shown to tag which parts of a new experience will be stored as a memory.
“In this paper, for the first time, we observe individual neurons stabilizing spatial representations during rest periods,” said Rice University neuroscientist Dr. Caleb Kemele.
“We imagined that some neurons might change their representation, mirroring the experience we've all had of waking up with a new understanding of a problem.”
“But to prove this, we needed to trace how individual neurons achieve spatial tuning – the process by which the brain learns to navigate new routes and environments.”
The researchers trained rats to run back and forth on a raised track with liquid rewards at each end, and observed how individual neurons in the animals' hippocampus “spiked” in the process.
By calculating the average spike rate over multiple round trips, the researchers were able to estimate a neuron's place field – the area of the environment that a particular neuron is most “interested” in.
“The key point here is that place fields are inferred using the animal's behavior,” Dr Kemele said.
I’ve been thinking for a long time about how we can assess neuronal preferences outside the labyrinth, such as during sleep,” Dr. Dibba added.”
“We addressed this challenge by relating the activity of individual neurons to the activity of all the other neurons.”
The scientists also developed a statistical machine learning approach that uses other neurons they examined to infer where the animals were in their dreams.
The researchers then used the dreamed locations to estimate the spatial tuning process of each neuron in the dataset.
“The ability to track neuronal preferences in the absence of stimulation was a significant advance for us,” Dr. Dibba said.”
This method confirmed that the spatial representation formed during the experience of a novel environment remained stable in most neurons throughout several hours of sleep following the experience.
But as the author predicted, there was more to the story.”
“What I liked most about this study, and why I found it so exciting, was that it showed that stabilizing memories of experiences isn’t the only thing these neurons do during sleep. It turns out some of them are doing other things after all,” Dr. Kemmele said.”
“We can see these other changes that occur during sleep, and then when we put the animals back into the environment, we can see that these changes actually reflect something that the animals learned while they were asleep.”
“It’s as if the animal is exposed to that space a second time while they’re sleeping.”
This is important because it provides a direct look at the neuroplasticity that occurs during sleep.
“It appears that brain plasticity and rewiring require very fast timescales,” Dr. Dibba said.”
This study paper In the journal Nature.
_____
K. Mabudi others2024. Recalibration of hippocampal representations during sleep. Nature 629, 630-638; doi: 10.1038/s41586-024-07397-x
Source: www.sci.news
