Reimagining axons with the pearls-on-a-string model

In a provocative new study, scientists challenge a fundamental tenet in neuroscience about the shape of axons — the long, thin filaments radiating from nerve cells that transmit electrical signals from cell to cell – and propose a new model for understanding how information is transmitted in the brain. The study, led by Shigeki Watanabe of Johns Hopkins School of Medicine, was partly conducted in the Marine Biological Laboratory (MBL) Neurobiology course and appears this week in Nature Neuroscience.

For more than 70 years, scientists have depicted axons as ultrathin cables, varying in diameter along their length but roughly cylindrical in shape. Electrical signals (action potentials) were thought to travel through them at constant velocity, like cars speeding through a tunnel. (This concept and velocity equation came from the historic study of action potentials in the squid giant axon by Alan Hodgkin and Andrew Huxley in the 1940s-50s, partly at MBL.)

However, Watanabe and team demonstrate that axons actually have a “pearls-on-a-string” morphology at the nanoscale level – lengths of cable interspersed with bulges they call “nano-pearls” (or nonsynaptic boutons). The speed of the action potential isn’t constant, they assert, but modulated by changes in the size of the nano-pearls, which in turn are caused by mechanical changes in the axon’s membrane and cytoskeleton as the action potential travels through.

We can think of this like cars traveling on a highway. If you have a four-lane highway through a tunnel, the cars travel normally. But if the highway has four lanes, then narrows to one lane, then goes back to four lanes, one lane – that’s how axons actually look. And you’d think the flow of traffic wouldn’t be too great.”

Shigeki Watanabe of Johns Hopkins School of Medicine

“But what’s interesting is the size of the pearls-on-a-string can change, at certain locations,” Shigeki continues. “We’ve shown you can modulate the nano-pearls’ size by changing factors in the local area, such as cholesterol in the plasma membrane. That, in turn, modulates the speed of the action potential. So, axons are highly flexible in that sense.”

Flash-freezing before imaging led to discovery

This axon ultrastructure that Watanabe and team describe is far below the diffraction limit of light microscopy, with the axon tract being about 60 nm in diameter with repeated nano-pearls about 200 nm in diameter. (The observations were made in unmyelinated axons in a mouse nervous system.)

“The reason people have missed this axon morphology before, and that we were able to observe this, is because we are looking at cryo-preserved tissues under an electron microscope,” Watanabe said. “Usually, people use chemicals to process samples for electron microscopy and then dehydrate these tissues, which is like making a grape into a raisin. But when you cryo-preserve, it’s like you’re making a frozen grape. You can preserve the actual shape.”

Understanding neurodegenerative disease

This discovery has implications for understanding neurodegenerative disease, Watanabe said. Alzheimer’s disease, for example, is associated with misregulation of cholesterol in the brain. Watanabe’s study shows the size of nano-pearls is modified by cholesterol moving onto or out of the neuronal plasma membrane, which in turn regulates the conduction velocity of action potentials. If this mechanism is impaired, it may eventually lead to axonal death.

“It will be interesting in the future to look at the mutations that lead to neurodegeneration, what axon morphology looks like in those neurons, and whether axon plasticity is still present,” he said.

Since 2015, Watanabe has been on the faculty of the MBL Neurobiology course, where part of the research was conducted. First author Jacqueline Griswold and co-authors Siyi Ma and Renee Pepper are MBL Neurobiology course alumni.

Part of this work was supported by an MBL Whitman Fellowship to Watanabe.