Single-cell exploration of the mouse brain reveals new cell type

The 100 billion cells in the human brain have a vast array of specializations — distinct shapes, types, functions, and connectivities — all of which contribute to the cellular diversity underlying our complex behavior as humans. Yet scientists don’t know all the cell types in the brain. 

“In order to really appreciate how the functions of the brain are actually arising, we need to know comprehensively what kinds of cells are there in this complex organ,” said Evan Macosko, an assistant professor and attending psychiatrist at Massachusetts General Hospital and an institute member at the Stanley Center for Psychiatric Research of the Broad Institute of MIT and Harvard.

Now, a study by Macosko, first authors Velina Kozareva and Caroline Martin also of the Stanley Center (now at MIT), Wade Regehr of Harvard Medical School, and colleagues is bringing scientists one step closer to a more thorough understanding of the cellular make-up of the brain.

The study is part of an ambitious project launched in 2017 by the The BRAIN Initiative’s Cell Census Network (BICCN) to create an atlas of all brain cells in the mouse, the most widely used model system in neuroscience, using a technique that analyzes the genetic activity in individual cells called single nucleus RNA sequencing (snRNA-seq). The first installment of the Macosko team’s effort, published today in Nature, reports a thorough characterization of all cell types in the mouse cerebellum — a region of the brain that controls movement through balance and coordination of the limbs and had not been explored in detail by previous single-cell studies. The study is one of 17 papers released by the BICCN in Nature, including one published this June by Paola Arlotta, Aviv Regev, and colleagues that teased apart the molecular signals and timing driving development in the mouse cerebral cortex.

The findings reveal a new cell type and different ways of categorizing known cell types. The study also shows far more complexity in the brain than previously thought, suggesting that much more about the brain’s myriad of cell types remains to be discovered. 

“We're going to learn a lot about how the brain works just by having this map,” said Macosko.

How scientists define cells

Over 130 years ago, Santiago Ramón y Cajal was the first scientist to see individual brain cells, staining them black and observing shapes that inspired intricate drawings still displayed in museums today. Because he distinguished cells by their shape, or morphology, rather than genetic information, Cajal missed key differences that would later become clear with more advanced technology such as electron microscopy and single cell RNA sequencing, which allows for systematic, high-throughput investigation of gene expression. 

Today, scientists use gene expression, which affects cell function, to distinguish cell types, though neuroscientists debate precise definitions of brain cell types. Some populations of cells also exist on a continuum, rather than having a discrete set of properties, which contributes to the brain’s complexity. The new study found some new examples of cells on a continuum. For example, brain cells called unipolar brush cells, which were previously categorized as two discrete populations, exhibit gradients of both gene expression and electrical properties. 

“One of the biggest challenges that we're appreciating as we sample more and more of the brain is that cells that we thought were very discrete and different from each other are just two ends of a continuum,” said Macosko. 

A new kind of cell

When Cajal was staining cells in the outer layer of the cerebellum, he noticed cells, later known as molecular layer interneurons (MLIs), that looked like one of two other brain cell types: stellate cells, which are small and star-shaped, and basket cells, which have long branching extensions that give the cells their name. For decades afterward, scientists studied MLIs and found that they had different electrical properties, and that some of the cells appeared to have properties of both basket and stellate cells. 

Using their high-throughput snRNA-seq approach, Macosko and his team found that MLIs were actually two distinct types of cells, distinguished not by their morphology but rather by their gene expression and subcellular structure; one population had connections between cells called gap junctions and the other did not. The team also discovered that the two kinds of MLIs were present in the human brain. 

“There's a lot more complexity in the brain than we thought,” said Macosko. “This was completely hidden from Cajal because he couldn’t see this distinction by just the morphology.” 

As Macosko and colleagues continue to profile regions across the mouse brain, their atlas will illuminate cellular diversity and set the stage for a human brain cell atlas, which Macosko says could dramatically improve scientific understanding of human health and disease. He adds that because he and his colleagues discovered a new cell type that makes up as much as one-third of cells in the outer layer of the cerebellum, new discoveries in other, less-studied regions of the brain are only just beginning.

Macosko’s group has already begun examining other brain regions in mice, including those in the pallidum, which are involved in mood, anxiety, and other behaviors. “We are finding that 90 percent of the cell populations in these regions are completely unknown, or are barely known by neuroscientists.”

“Hundreds of cell types have not been studied,” he said. “I'm quite confident of that.” 

Watch a webinar describing BICCN efforts to understand the cellular composition of the mammalian brain, and in particular the primary motor cortex, on Wednesday, October 27 at 1 pm EST.

This work was supported by the Stanley Center for Psychiatric Research and the National Institute of Mental Health at the National Institutes of Health.