Picture a network of rubber bands stretched between brain cells. When we're young, these bands are elastic: when we learn something new, they stretch easily and change shape. But with age, too much glue gets into the network, and it hardens. The connections between neurons can no longer rearrange themselves, and memory starts to fail. For a long time, scientists thought the neurons themselves were to blame, but it turned out that other cells play the leading role: oligodendrocytes.
Oligodendrocytes are helper cells that wrap nerve fibers in insulation so signals can travel faster. They also release proteins that fill the space between cells, the extracellular matrix. One of these proteins, HAPLN2, works like glue: it binds the matrix together at the spots where a nerve impulse jumps from one segment of a fiber to another. In old mice and elderly humans, there's too much of this glue, especially in the brain region responsible for memory.
This discovery changes how we look at brain aging. It used to seem that oligodendrocytes simply suffer from old age, but now it's clear: they actively make memory worse by releasing too much glue. If we can create a drug that lowers HAPLN2 levels, we might be able to restore flexibility to the aging brain and halt memory decline. No such drugs exist today, but now we know which direction to look.
The nodes of Ranvier, where HAPLN2 accumulates, are gaps in the myelin sheath about 1 micrometer long. That is exactly where the nerve impulse "leaps," speeding up tens of times over. In older people, there are fewer such stretches, and the signal slows down.
In Greg Egan's novel Permutation City, characters control their own biochemistry to change their personalities. Manipulating the extracellular matrix, as in this work, is a first step toward influencing memory through the molecular environment of neurons rather than through the neurons themselves.
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.