Brain cells, or neurons, are the building blocks of the nervous system. They communicate with each other through synapses, where electrical signals are transmitted from one neuron to another. But how do neurons decide which synapses to form and how strong they should be? A new study by physicists and neuroscientists from the University of Chicago, Harvard and Yale provides a surprisingly simple answer.
Hebbian dynamics: A general principle of network organization
The researchers analyzed data from several model organisms, including fruit flies, roundworms, marine worms and mouse retina. They found that the distribution of synaptic strengths followed a similar pattern across all these species: a small number of synapses were much stronger than most, forming the backbone of the neural network. This pattern, known as a heavy-tailed distribution, is also seen in other types of networks, such as social interactions or the internet.
To explain this pattern, the researchers developed a mathematical model based on Hebbian dynamics, a concept proposed by Canadian psychologist Donald Hebb in 1949. Hebb’s rule states that “neurons that fire together, wire together”, meaning that the more two neurons activate together, the stronger their connection becomes. The model showed that Hebbian dynamics alone can produce heavy-tailed distributions of synaptic strengths, without relying on any biological features of the organisms.
Implications for understanding brain function and evolution
The study, published on January 17, 2024 in Nature Physics, suggests that Hebbian dynamics are a universal principle of network organization that applies across a wide range of organisms and potentially other types of networks as well. The researchers also showed that their model can predict how synaptic strengths change over time in response to different stimuli or learning tasks.
The findings have important implications for understanding how brain function and evolution are shaped by network properties. For example, the heavy-tailed distribution of synaptic strengths may allow neural networks to store more information and adapt more quickly to changing environments. The model also provides a simple way to compare neural connectivity across different species and identify common principles of brain organization.
“Our study shows that you can explain a lot of complex phenomena with very simple rules,” said Stephanie Palmer, PhD, Associate Professor of Physics and Organismal Biology and Anatomy at UChicago and senior author of the paper. “It’s exciting to see how general principles of networking and self-organization can account for the diversity and complexity of brain connectivity.”