What if aging memories could be made young again? Scientists at Switzerland's EPFL have shown that it may be possible — by resetting the molecular clock of the exact brain cells that store our memories.
In a study published in the journal Neuron in February 2026, a team led by Johannes Gräff at EPFL's Brain Mind Institute demonstrated that 'partial reprogramming' of specific neurons called engrams — the sparse groups of cells that form the physical trace of a memory — can restore learning and memory in aged mice.
The approach uses a gene therapy cocktail known as 'OSK' (Oct4, Sox2, and Klf4), three genes that previous research has shown can reset age-related molecular changes in cells. But rather than applying this broadly across the entire brain, the EPFL team did something elegant: they targeted only the engram neurons that were active during learning.
The results were striking. In aged mice, the OSK treatment not only recovered recent memories stored in the hippocampus but also restored long-term memories in the prefrontal cortex. The treated neurons showed physical signs of rejuvenation, including restored nuclear structures and improved firing patterns that typically degrade with age.
Perhaps most remarkably, the approach also worked in mouse models of Alzheimer's disease, improving spatial learning strategies and restoring long-term memory that had previously been impaired.
'Memories are thought to rely on these sparse groups of neurons called engrams,' explained the research team. 'In aged brains and Alzheimer's models, engrams malfunction and memory recall suffers. We asked: could rejuvenating these specific neurons recover memory after decline has already begun?'
The answer, resoundingly, is yes.
The key innovation is precision. Rather than attempting to rejuvenate the entire brain — which carries risks of uncontrolled cell growth — this approach specifically targets only the neurons involved in storing particular memories. It's like replacing worn-out components in a machine rather than rebuilding the whole thing.
The implications are profound. Alzheimer's disease affects more than 55 million people worldwide, with that number expected to nearly triple by 2050. Current treatments can slow decline but cannot restore lost cognitive function. This research suggests that memory loss may not be the one-way street it was long assumed to be.
The team is now working to refine the approach and explore its potential for human applications. While gene therapy in the human brain remains complex, the fundamental principle — that aging neurons can be molecularly rejuvenated to restore their function — opens an entirely new avenue for treating age-related cognitive decline and neurodegenerative disease.
As the researchers put it: 'This suggests a new paradigm. Memory decline isn't necessarily permanent. The hardware is still there — it just needs to be rebooted.'