Place cells—the hippocampal neurons discovered by Nobel laureate John O’Keefe, PhD—normally fire in crisp, rapid sequences that replay our recent experiences during rest, helping memories take hold. But in Alzheimer’s disease, that internal rehearsal appears to become undone. In a new study from University College London (UCL), researchers report that hippocampal replay still occurs in an Alzheimer’s mouse model, yet the sequences lose their structure, hinting at a subtle but profound breakdown in the machinery that stabilizes memories.
The work, published in Current Biology and titled “Disrupted hippocampal replay is associated with reduced offline map stabilization in an Alzheimer’s mouse model,” offers a mechanistic glimpse into how memory begins to fail long before neurons die or brain regions visibly degenerate. Instead of a missing process, the study suggests a malfunctioning one: the brain tries to consolidate memories, but the replay itself is scrambled.
“When we rest, our brains normally replay recent experiences—this is thought to be key to how memories are formed and maintained,” said co‑lead author Sarah Shipley, PhD, a senior research fellow in UCL’s cell and developmental biology department. “We found this replay process is disrupted in mice engineered to develop the amyloid plaques characteristic of Alzheimer’s, and this disruption is associated with how badly animals perform on memory tasks.”
To probe this breakdown, the team turned to the hippocampus, where place cells encode specific locations and fire in ordered sequences as an animal moves through space. During rest, those same sequences reactivate in compressed bursts—replay events thought to stabilize spatial maps and support long‑term memory. Using electrode arrays capable of tracking roughly 100 place cells simultaneously, the researchers monitored neural activity while mice navigated a radial arm maze.
In healthy mice, replay events during rest reinforced stable place‑cell representations. But in the Alzheimer’s model (the App NL‑G‑F knock‑in line), replay was fundamentally altered. The frequency of replay events remained normal, yet the internal structure was degraded: cell recruitment was disrupted, and co‑firing patterns within reactivation events were weakened. Place cells also became less stable over time, particularly after rest periods—precisely when replay should have strengthened them.
These neural disruptions had behavioral consequences. Affected mice performed worse in the maze, repeatedly revisiting arms they had already explored. “We’ve uncovered a breakdown in how the brain consolidates memories, visible at the level of individual neurons,” said co‑lead author Caswell Barry, PhD. “What’s striking is that replay events still occur—but they’ve lost their normal structure. It’s not that the brain stops trying to consolidate memories; the process itself has gone wrong.”
The team sees this mechanistic insight as a potential foothold for earlier diagnosis and more targeted therapies. “We hope our findings could help develop tests to detect Alzheimer’s early, before extensive damage has occurred, or lead to new treatments targeting this replay process,” Barry said. “We’re now investigating whether we can manipulate replay through the neurotransmitter acetylcholine, which is already targeted by drugs used to treat Alzheimer’s symptoms. By understanding the mechanism better, we hope to make such treatments more effective.”
