AI helps map the biology of spinal cord injuries in unprecedented detail

The spinal cord It is one of the most complex biological systems and, now, a scientific team, thanks to technologies cutting edge molecular mapping artificial intelligencehas managed to draw an open source atlas that offers a comprehensive understanding of the biology of spinal cord injuries.

The results are published in the journal Nature, in an article in which the researchers describe the difficult molecular processes that develop in each cell after a spinal cord injury, which opens the way to new, more effective and personalized therapies.

The work, carried out in rodents, not only identifies a specific set of neurons and genes that play a key role in recovery, but also proposes a gene therapy derived from the discoveries.

According to scientists at the Federal Polytechnic School of Lausanne (EPFL), it is ‘an important milestone’ that manages to map the cellular and molecular dynamics of paralysis in unprecedented detail, thanks to the open source project “Tabulae Paralytica”.

The human spinal cord is one of the most complex biological systems known to science: it is a mechanical, chemical and electrical arrangement of different types of cells that work in harmony to produce and regulate a multitude of neurological functions.

This cellular complexity amplifies the difficulties in effectively treating paralysis caused by spinal cord injuries. Until now, traditional imaging and mapping methods offered a generalized view of the cellular mechanisms of spinal cord injuries, but the lack of specificity blurred the different functions and reactions of cell types and hindered the development of specific treatments, explains a statement. of the EPFL.

“By offering an exceptionally detailed view of the cellular and molecular dynamics of spinal cord injury in mice, across space and time, the four cellular atlases that make up Tabulae Paralytica close a historic knowledge gap, paving the way for targeted treatments.” and better recovery”summarizes Grégoire Courtine.

The first treatment derived from this new understanding is a specific gene therapy, which takes advantage of a finding ‘crucial’: that a specific type of supporting cell called an astrocyte loses its ability to respond to injury in aging animals.

Another key result of the study is the identification of a specific subset of neurons, known as Vsx2 neurons, that are intrinsically equipped to promote recovery.

“Our previous studies had pointed in this direction, but with this new and refined understanding, we can say with certainty that Vsx2 neurons are largely responsible for the reorganization of neuronal circuits, which means that they are by far the largest population of neurons more interesting to repair spinal cord injuries”, says Jordan Squair.

Machine learning

To create the first comprehensive cellular map of spinal cord injury in rodent models, researchers used two innovative technologies.

The first, single-cell sequencing, examines the genetic makeup of each cell. Although it has been used for more than a decade, recent advances have allowed scientists to expand the process like never before, generating detailed reports of millions of spinal cord cells.

Second, spatial transcriptomics – a cutting-edge technology that shows where these cellular activities occur – expanded the map to the entire spinal cord, preserving the spatial context and relationships between different cell types.

The new data is so vast that new machine learning techniques had to be developed to take advantage of its complexity.

“We now have a detailed map that not only shows us which cells are involved, but also how they interact and change throughout the injury and recovery process,” Squair concludes.