Neuroscience: Neuronal networks in dishes


A simple model of the brain consisting of four circuits of rat neurons reveals how the brain strikes the right balance between two processing strategies

By varying the degree of physical coupling between four modules made up of circuits of rat neurons, AIMR researchers have mimicked the activity pattern that occurs in actual brains.
By varying the degree of physical coupling between four modules made up of circuits of rat neurons, AIMR researchers have mimicked the activity pattern that occurs in actual brains.

© 2019 Hideaki Yamamoto

By precisely engineering the interface between living cells and a glass coverslip, AIMR researchers have constructed small circuits of rat neurons in a dish1. This allowed them to mimic the way neurons are connected in the brain, revealing important insights into the mysterious field of brain dynamics.

The human brain is made up of specialized regions that process signals from different sources. For example, one region of the brain processes the colors our eyes see, while another region processes language.

But in addition to this segregation of activities, processed signals from the different brain regions need to be integrated. Evidence of this integration is found in that fact that the brain recognizes the words red and blue more quickly if they are written as red and blue than if they are written as red and blue, explains Hideaki Yamamoto of the AIMR at Tohoku University.

“Generally speaking, integration is when different neuronal groups activate coherently, whereas segregation is a state where the neuronal groups activate independently,” says Yamamoto. “These two states of neural activation are well balanced in the brain and underpin its ability to carry out complex computation.”

But it has not been clear how the brain coordinates these very different activities. “The neural basis of information processing in the brain is one of the biggest challenges in modern natural science,” notes Yamamoto.

Now, Yamamoto, working with researchers at Tohoku University, the University of Barcelona, Tohoku Fukushi University, Waseda University and Yamagata University, has used a bottom-up approach to explore this problem. Specifically, they created a very simple model of a brain by linking four modules made up of circuits of rat neurons. The researchers then investigated how they integrate when they varied the degree of physical coupling between them (see image).

The team discovered that segregation and integration coexist only when there is almost no coupling between the four modules. If the coupling is any stronger, then integration predominates, whereas the modules operate independently of each other if it is weaker.

“By introducing modular organization to in vitro neuronal networks, we succeeded in suppressing excessive coherence between the networks and realized activity patterns that more closely resemble those observed in actual brains,” says Ayumi Hirano-Iwata of the AIMR, who led the group. “While the observed dynamics are still far from the complexity in brains, we were able to gain insight into the fundamental mechanisms that shape brain dynamics.”

In addition to demonstrating the advantages bestowed by the modular organization of the brain, the findings illustrate the power of using biophysical approaches to explore collective phenomena in complex systems.


  1. Yamamoto, H., Moriya, S., Ide, K., Hayakawa, T., Akima, H., Sato, S., Kubota, S., Tanii, T., Niwano, M., Teller, S. et al. Impact of modular organization on dynamical richness in cortical networks. Science Advances 4, eaau4914 (2018). | article

This research highlight has been approved by the author of the original article and all empirical data contained within has been provided by said author.