UC Berkeley researchers develop first-ever atlas of brain cells

photo of Helen Bateup
Dirk Hockemeyer/Courtesy

Related Posts

For the first time ever, a large number of researchers — including several from UC Berkeley — have unveiled an “atlas” of brain cells, the outcome of five years of work culminating in the release of 17 studies Wednesday.

The studies specifically focused on the mammalian primary motor cortex, the part of the brain responsible for controlling movement, according to campus statistics chair and paper co-author Sandrine Dudoit.

This research is especially noteworthy because it studied mouse, marmosets and human brains, according to Helen Bateup, a campus neurobiology associate professor and a co-author of the project’s flagship synthesis paper.

The flagship paper completes a “cross-species analysis,” which allows the scientists to standardize cell types and taxonomy for numerous mammals, Bateup added.

“The marker genes identified for the newly discovered cell types will allow researchers to target those particular cell types using mouse models or other systems to understand their functions in both health and disease,” Bateup said in an email. 

Analyzing this data, however, was not without its challenges. Specifically, the researchers faced the problem of “clustering,” in which algorithms give scientists distinctive “clusters” of data collected — in this case, cells.

In many cases, it is unclear as to whether or not certain clusters were actually distinct from other clusters, which was an issue the researchers had to overcome.

“Clustering is a challenging problem: No matter what the data or clustering algorithm, we get clusters,” Dudoit said in an email. “Whether these clusters are ‘real’ is another story.”

According to Dudoit, to solve their difficulties, she and the rest of the statistical team, including campus associate statistics professor Elizabeth Purdom, developed numerous tools, including a learning framework allowing researchers to find “consensus clustering” and a procedure to merge clusters.

In addition to Dudoit and Purdom, other campus-affiliated researchers played a key role in the project. Campus associate cellular and developmental biology professor Dirk Hockemeyer and John Ngai, a former campus professor who now serves as the director of the National Institutes of Health BRAIN Initiative, the initiative responsible for coordinating this project, were involved in the project as well.

According to Bateup, Ngai coordinated campus researchers’ efforts, while Hockemeyer and his lab used CRISPR-Cas9 to see and alter the cell types. Bateup’s lab, in turn, helped characterize these mice, finding these cell types’ location in the brain and where their projections go.

The researchers hope the work will help eventually create a comprehensive atlas of the entire brain and potentially have medical benefits.

“Detailed information about the types of cells that make up the brain and their properties lays the foundations for discovering new ways to treat, cure, and even prevent brain disorder,” Dudoit said in the email.

Lukas Carbone is a general assignment reporter. Contact him at [email protected].