Institute News

Traffic lights for perception

Bonn, 26.11.2019. We see, we smell, we hear and we taste. We feel the sensation of touch on our skin. Every moment, our brain processes this data and transforms it into perception. This perception influences our behavior. The area of the brain dedicated to transforming sensory data into perception, and ultimately into behavior, is the neocortex. This structure, only a few millimeters thick and located at the outermost part of the brain, is almost exclusively found in mammals. Its complexity is astounding. Within every cubic millimeter, the neocortex contains hundreds of thousands of neurons, forming complex networks that are interconnected by up to a billion synapses – rendering the neocortex as one of the most complex structures known to Biology.

The neocortex: computational center of behavior

It’s not only the structure of the neocortex, which fascinates, but also its diverse functions. During the complex interplay of activity patterns between many neurons, the networks of the neocortex integrate sensory information, creating a representation of our environment. Relaying the results of these cortical computations to deeper regions of the brain is thereby an important aspect for generating behavior.

The neurons, which broadcast these results of cortical processing, have been known for decades. So-called pyramidal tract neurons receive information from thousands of neurons located throughout the neocortex, and transform these inputs into output signals.

A computer model of the neocortex unveils a new principle

Understanding the functionality of neuronal networks is a very challenging task. While it’s possible to experimentally measure how neurons react to sensory stimuli, the principles that underlie these activity patterns remain largely inaccessible by state-of-the-art techniques.

To overcome these present limitations, the Max Planck Research Group “In Silico Brain Sciences” at caesar combined measurements in the rodent brain with computer simulations. In the study, first activity patterns from all neurons that could potentially contribute to inputs that pyramidal tract neurons receive during sensory stimulation were measured. Subsequently, this empirical data was used to construct a realistic model of the neocortex and to constrain computer simulations of information flow through the model.

A traffic light for the neocortex

These simulations were not only able to reproduce measurements in the living animal, they also provided predictions of which neurons in the brain could regulate cortical output. Testing the predictions in the living animal, the scientists made a remarkable discovery. They identified special neurons, which act like a traffic light. Only when these cells become activated, the pyramidal tract neurons ares able to broadcast information out of the neocortex.

“The notion of a traffic light fits rather well”, says Dr. Marcel Oberlaender, lead investigator of the study. “We found that once sensory information reaches the neocortex, it first activates these ‘traffic light neurons’. Without activation of these cells, information could not leave the neocortex. One may say: these cells provide a ‘green light’ to the pyramidal tract neurons.”

The discovery of these “traffic light” neurons reveals a basic principle required for the transformation of sensory input into behavior. The study further provides a unique approach for combining experiments with simulations, opening new venues for investigating how functions are implemented within the brain.

The results of the study will be published on November 26, 2019, in the scientific journal “Neuron” (s2)

Original publikation:
Egger, Narayanan et al., Cortical Output Is Gated by Horizontally Projecting Neurons in the Deep Layers, Neuron (2019),

Wissenschaftlicher Kontakt:
Dr. Marcel Oberlaender
Forschungszentrum caesar
Max-Planck-Gruppe In Silico Brain Sciences
Ludwig-Erhard-Allee 2
53175 Bonn
Tel.: 0228 - 9656 380

Sebastian Scherrer
Forschungszentrum caesar
Ludwig-Erhard-Allee 2
53175 Bonn
Tel.: 0228 - 9656 139

For further information please contact: