University of Tübingen: First comprehensive atlas of neuron types in the brain published
Neuroscientists have been researching the properties of neurons in the brain for more than 100 years. One of the central questions is to what extent the different types of nerve cells differ from one another and what influence this has on brain activity. The BRAIN Initiative Cell Census Network (BICCN), an international research collaboration in which researchers from the University of Tübingen are also involved, has now achieved a breakthrough: They have created a cell atlas that provides a unique overview of the various neuron types and their respective properties in the motor The cortex supplies the brain region that controls our movement processes – in the brains of mice, monkeys and humans. The atlas was published on Wednesday in a special edition of Naturereleased. The edition contains 17 papers, including an overview article that describes the entire atlas.
The aim of the BICCN research network was to describe the types of neurons in the motor cortex as comprehensively as possible. A catalog based on a “cell census” should be created. Similar to a population census, this should define the different types of neurons, including their specific characteristics and their distribution in the brain. With the help of new experimental techniques and data analysis methods, the scientists have now succeeded in collecting genetic information on more than a million cells. They recorded their spatial position, shape and electrical properties for some of the cells and determined their connections to other neurons in other areas of the brain. Since the researchers analyzed the cells of mice, marmoset monkeys and humans, they were even able to trace the evolutionary development of the various types of nerve cells. The result is a cell atlas that offers a previously unknown overview of the motor cortex and its development in the course of evolution.
Scientists from the University of Tübingen have contributed a study to this joint effort in which they characterize the different cell types in the motor cortex of the mouse on the basis of several data types. The work under the direction of Philipp Berens, professor at the Research Institute for Ophthalmology at the University of Tübingen and spokesman for the Cluster of Excellence “Machine Learning”, Prof. Andreas Tolias from Baylor College of Medicine in Houston, Texas (USA), and Prof. Rickard Sandberg from Karolinska Institutet in Stockholm (Sweden) provides one of the most complete descriptions to date of the variety of different neuron types in the mouse brain. It is part of the special edition of Nature and was previously published online.
Neuroscientists usually describe neurons in terms of three basic characteristics: their anatomy or what they look like under a microscope, their physiology or how they react to stimuli, and their transcriptome, which is the genetic information that is actually read in the cell. The team used a new, sophisticated experimental technique called “patch-seq” to compile a large database of anatomical, physiological and genetic information from cells in the mouse motor cortex.
The fact that the researchers were able to measure the three basic properties in over 1000 cells at the same time gave them a deep understanding of how the neurons in the motor cortex are related to one another. Using machine learning techniques, they brought together anatomical, physiological and genetic information and discovered relationships between neurons that were previously unknown. “The large genetic neuron families have different anatomical and physiological properties. But within each family, the neurons show a gradually changing anatomical and physiological diversity, ”explains Dr. Dmitry Kobak from the University of Tübingen, who is the lead author of the study together with Federico Scala from Baylor College of Medicine.
In analogy to the “tree of life”, which describes the relationships between the various species, the researchers came to the following conclusion: The neurons are subject to an order that consists of different, non-overlapping branches at the family level. Within each family, however, they show continuous changes in terms of their genetic, anatomical and physiological properties, so that the “tree of cell types” is more like a banana tree than an olive tree. With this, the scientists propose a completely new model for describing the diversity of neurons in the brain and their relationships to one another.
“The data from the new cell atlas of our research cooperation will be an invaluable resource for neuroscience,” says Philipp Berens. “By using our machine learning knowledge, we create the link between genetics and the physiology and anatomy of neurons. This can be crucial when it comes to understanding diseases that affect the brain at the cellular level. “