Do long-term memories punch holes in the brain?

How the brain encodes and stores memories is one of the enduring mysteries of neuroscience. Memories are thought to be encoded by the strengthening of synaptic connections, and many researchers believe that they are retained by proteins at the synapses. And yet, while memories can persist over our entire lifetimes, these synaptic proteins are continuously being destroyed and replaced, over a time-frame of hours or days.

Several years ago, neuroscientist Roger Tsien of the University of California, San Diego, proposed that long-term memories are stored in patterns of holes created within a lattice-like structure called the perineuronal net. Sakina Palida, a graduate student in Tsien’s lab, presented evidence for such a ‘punch-card’ mechanism at the annual meeting of the Society for Neuroscience in Chicago earlier this month.

The perineuronal net is a specialization of the extracellular matrix, consisting of fibrous proteins and carbohydrates that form a coat that surrounds nerve cell bodies and fibers, and plays a critical role in the formation, stabilization, and maintenance of synaptic connections.

Palida and her colleagues fused the genes encoding a small perineuronal net ‘linker’ protein to that of a yellow fluorescent protein, enabling them to monitor the dynamics of the protein in real time. When this fusion gene was introduced into nerve cells growing in Petri dishes, the cells synthesized a fluorescent linker protein that they secreted and then integrated into the cell membrane from the outside, where it remained stable for up to 180 days.

The researchers then treated cells with brain-derived neurotrophic factor (BDNF), a growth factor that can strengthens synapses and make them larger, and found that this resulted in a breakdown of extracellular matrix proteins, leading to a remodelling of the perineuronal net in the area immediately surrounding the strengthened synapses.

“The perineuronal net is highly stable, but is locally degraded when synapses are strengthened,” says Palida. “The strengthened synapses produce a protein that chews up the matrix to form this hole structure, which remains very stable over time, regardless of what synaptic proteins come and go through it.”

Another set of experiments showed that mice lacking the enzymes needed to degrade the perinuronal net have deficient long-term, but not short-term, memories, adding weight to the idea that the perineuronal net contributes to the storage of infomration over long periods of time.

Other research implicates the perineuronal net in Alzheimer’s Disease and other dementias, but so far the results are conflicting, and so its exact role is unclear. Some post-mortem studies show that certain components of the perineuronal net are lost in the brains of patients with Alzheimer’s; in others, however, the net remains unaffected, and in some it even appears to protect neurons against degeneration.

A better understanding of how the perineuronal net contributes to long-term memories could therefore open up new avenues for treating these devastating diseases.

“Our results suggest that the perineuronal net is an ideal substrate for encoding and storing memories in the brain over time,” says Palida, “and we’re currently making transgenic animals expressing this [fluorescent linker] protein, which will allow us to monitor the dynamics of the perineuronal net simultaneously with synaptic dynamics.”


Originally posted on Brain Decoder, October 25th, 2015.

Header image: Confocal micrographs showing the perineuronal net (stained green) in three different regions of the mouse brain. (Palida, S. F., et al.)