Like the seeds of a forest, a few cells in embryos eventually sprout into an ecosystem of brain cells. Neurons get the most recognition for their computing power. But a host of other cells provides nutrition, clears the brain of waste, and wards off dangers, such as toxic protein buildup or inflammation.
This rich diversity underlies our ability to process information, transforming perception of the world and our internal dialogues into thoughts, emotions, memories, and decisions. Mimicking the brain could potentially lead to [energy-efficient computers](https://singularityhub.com/2025/10/21/unprecedented-artificial-neurons-are-part-biological-part-electrical-work-more-l…
Like the seeds of a forest, a few cells in embryos eventually sprout into an ecosystem of brain cells. Neurons get the most recognition for their computing power. But a host of other cells provides nutrition, clears the brain of waste, and wards off dangers, such as toxic protein buildup or inflammation.
This rich diversity underlies our ability to process information, transforming perception of the world and our internal dialogues into thoughts, emotions, memories, and decisions. Mimicking the brain could potentially lead to energy-efficient computers or AI systems. But we’re still decoding how the brain works.
One way to understand a machine is to first examine its parts. The landmark project BRAIN Initiative Cell Atlas Network (BICAN), launched in 2022, has parsed the brains of multiple species and compiled a census of adult brain cells with unprecedented resolution.
But brains are not computers. Their components aren’t engineered and glued on. They develop and interact cohesively over time.
Building on previous work, the BICAN consortium has now released results that peek inside the developing brain. By tracking genes and their expression in the cells of developing human and mouse brains, the researchers have built a dynamic picture of how the brain constructs itself.
This herculean effort could help scientists unravel the causes of neurodevelopmental disorders. In one study, led by Arnold Kriegstein at the University of California, San Francisco, scientists found brain stem cells that are potentially co-opted to form a deadly brain cancer in adulthood. Other studies shed light on imbalances between excitatory and inhibitory neurons—these ramp up or tone down brain activity, respectively—which could contribute to autism and schizophrenia.
“Many brain diseases begin during different stages of development, but until now we haven’t had a comprehensive roadmap for simply understanding healthy brain development,” said Kriegstein in a press release. “Our map highlights the genetic programs behind the growth of the human brain that go awry during specific forms of brain dysfunction.”
Shifting Landscape
Over a century ago, the first neuroscientists used brain cell shapes to categorize their identities. BICAN collaborators have a much larger arsenal of tools to map the brain’s cells.
A key technology called single-cell spatial transcriptomics detects which genes are turned on in cells at any given time. The results are then combined with the cells’ physical location in the brain. The result is a gene expression “heat map” that provides clues about a cell’s lineage and final identity. Like genealogical tracking, the technology traces the heritage of different types of brain cells and when they emerge while at the same time providing their physical address.
Like other organs, the brain grows from stem cells.
In early developmental stages, stem cells are nudged into different fates: Some turn into neurons, some turn into other cell types. So far, no single technology can “film” their journey. But BICAN’s new releases measuring gene expression through development offer a glimpse.
In one tour-de-force study, Kriegstein and team used a technique that maps gene variants to single cells during multiple stages of development. Many variants were previously linked to neurodevelopmental disorders, including autism, but their biological contribution remained mysterious.
The team gathered 38 donated human cortex samples—the outermost part of the brain—that spanned all three trimesters of pregnancy, after birth, and early adolescence.
They then grouped individual cells using gene expression data across samples. They found roughly 30 different types of cells that emerge during brain development, including excitatory and inhibitory neurons, supporting cells such as glia, and immune cells called microglia.
Some were linked to a single source. This curious cell type, dubbed tripotential intermediate progenitor cells, spawned an inhibitory neuron, star-shaped glia, and brain cells that wrap around neurons as protective sheathes of electrical insulation. The latter break down in neurological diseases like multiple sclerosis, resulting in fatigue, pain, and memory problems.
Many genes related to autism were turned on in immature neurons as they began their brain-wiring journey. Gene mutations, environmental influences, and other disruptions could interfere with their growth.
“These programs of gene expression became active when young neurons were still migrating throughout the growing brain and figuring out how to build connections with other neurons,” said study author Li Wang. “If something goes wrong at this stage, those maturing neurons might become confused about where to go or what to do.”
The mother cells also have a dark side. Scientists have long thought that glioblastoma, a fatal brain cancer, stems from multiple types of neural precursor cells. Because mother cells, marked by their distinctive gene expression profiles, develop into all three types of cells involved in the cancer, they’re essentially cancer stem cells that could be targeted for future treatments.
“By understanding the context in which one stem cell produces three cell types in the developing brain, we could be able to interrupt that growth when it reappears during cancer,” said Wang.
A Wealth of Data
Other BICAN studies also zeroed in on inhibitory neurons.
The authors of one hunted down a group of immature cells that shifted from making excitatory neurons to inhibitory ones during the middle of gestation, proving to be a balance between both forces. In another, in mice, researchers followed inhibitory neurons as they diversified and spread across the developing brain. More subtypes with unique gene expression profiles appeared in the cortex compared to deeper regions, which are more evolutionarily ancient.
Other studies investigated gene expression in neurodevelopment and how changes can lead to inflammation. Environmental influences such as social interactions played a role, especially in forming brain circuits tailored to gauging others’ behaviors. In developing mice, several genes related to social demands abruptly changed their activity during developmental milestones, including puberty.
Some cell types were shape-shifters. In mice, an immune challenge briefly changed microglia—the brain’s immune cells—back into a developmental-like state, suggesting these cells have the ability to turn back the clock.
The collection of studies only skims the surface of what BICAN’s database offers. Although the project mainly focused on the cortex, ongoing initiatives are detailing a cell atlas of the entire developing brain across dozens of timepoints and multiple species.
“Taken together, this collection from the BICAN turns the static portrait of cell types into a dynamic story of the developing brain,” wrote Emily Sylwestrak at the University of Oregon, who was not involved in the studies.