Your brain begins as a single cell. After all, it houses an incredibly complex and powerful network of about 170 billion cells. How does it organize itself along the way? Neuroscientists at Cold Spring Harbor Laboratory have discovered a surprisingly simple answer that could have far-reaching implications for biology and artificial intelligence.
Stan Kerstiens, a postdoctoral fellow in Professor Anthony Zador’s lab, frames this question in terms of location. “A cell ‘sees’ only itself and its neighbors,” he explains. “But the fate of a cell is determined by where it is located. If a cell is in the wrong place, it will be the wrong one, and the brain will not develop correctly. Therefore, every cell must solve two questions: where am I and who do I need to become?”
In a study published in neuronKerstjens, Zador, and colleagues at Harvard University and ETH Zurich have proposed a new theory about how the brain is organized during development.
For a long time, researchers thought that cells exchanged location information primarily through chemical signaling. This works well when working with just a few cells, explains Carstiens. But the brain is not just a few cells. It’s billions of neurons, each of which needs to land in exactly the right place. Chemical signals can only travel a certain distance before attenuating. So how do cells deep in the developing brain automatically “know” their location?
The answer proposed by Kerstiens is very close. “Think about how the human population spreads across the country over generations,” he says. “Because descendants settle close to their parents, people who share ancestry end up living in neighborhoods, giving rise to large-scale geographic structures without long-distance communication. We argue that a similar principle is at work in the developing brain: Cells descended from the same ancestor tend to stay close to each other.”
To test this theory, Kerstjens and colleagues built what they called a “scalable location-based lineage-based model.” They started with theoretical calculations. They then tested their hypothesis on a large scale by examining individual and group gene expression in the developing mouse brain. Finally, they confirmed their results in zebrafish and showed that the model can be used for brains of different sizes.
Carstiens said the model supports the concept that chemical signaling works with lineage-based mechanisms to convey location information. Although his research focuses on the brain, the theory could be applied to many other types of developing tissues, including tumors. Similar to our own brain cells, self-replicating AI models that pass information from one generation to the next may also have implications.
Perhaps most importantly, showing how single cells grow into complex organs could help scientists solve fundamental mysteries of the mind.
The brain somehow makes us smarter. How was it possible to accumulate this ability not only during development, but also through a period of evolution? This is one piece of a larger puzzle. ”
Stan Kerstjens, Postdoctoral Researcher in the lab of Professor Anthony Zador
sauce:
Cold Spring Harbor Laboratory
Reference magazines:
Kerstiens, S. Others. (2026). A lineage-based model of scalable positional information in vertebrate brain development. neuron. DOI: 10.1016/j.neuron.2025.12.043. https://www.cell.com/neuron/abstract/S0896-6273(25)01000-1
