Human Cortex Grown Inside Mice
Stanford University researchers removed most of the cortex from developing mice and let human brain organoid cells fill the gap. The resulting animals did better than cortex-free mice on several tests, yet they still trailed normal mice, and the human tissue never formed the layered structure of a real cortex.
Key Takeaways
- Stanford researchers engineered mice lacking most of their cortex, then implanted human cortical organoids that supplied 92% of the cells in the rebuilt region.
- The human cells formed all major cortical neuron types and long-range connections reaching the spinal cord, but not the distinct layers of a normal cortex.
- On behavior, body weight, maze and motor tests, the mice landed between normal and cortex-free animals, and they showed no gain in associative memory.
Scores Between Two Extremes
The team filmed the animals in an open area and used a machine learning classifier to group similar behavior patterns. Normal mice and cortex-free mice formed two separate clusters. Mice with human cortical tissue formed a third, distinct from both.
Body weight followed the same pattern. Mice without a cortex weighed much less than normal mice, and the transplanted animals fell in between.
In a simple maze, cortex-free mice performed at the level of random chance. Transplanted mice beat chance but still lagged behind normal mice. Tests of fine motor coordination gave another in-between result.
One test showed no benefit at all. On forming associative memories, the transplanted mice did no better than mice with no cortex.
Making Room in a Mouse Brain
The work, published on Wednesday in Nature (DOI: 10.1038/s41586-026-11032-2), comes from the lab of Stanford neuroscientist Sergiu Pasca. The team made a bold choice: it removed the cortex, the region that handles decision-making, memory and other complex functions.
The researchers found a gene that is active in almost all cortical cells. They used it to trigger the deletion of another gene that cells need to separate chromosomes when they divide. Most cells that would have formed the mature cortex died, and brain volume fell by half. Stanford says the mice also lacked most of the hippocampus, a structure important for memory.
Remarkably, the mice could survive this. To help them, the team removed most other pups from each litter so the cortex-free mice got enough milk. The animals stayed with their mothers longer and then received very high-calorie food. With those steps, nearly all of them survived.
The mice were also immunocompromised to prevent an immune attack on human cells, a weakness that poses less risk in a sterile animal facility.
Next, the researchers implanted human cortical organoids into the space where the mouse cortex had failed to develop. Stanford calls the resulting animals “xenocortical” mice.
What the Human Cells Built
A little over 85% of implanted animals took up the graft. In those mice, human cells made up 92% of the cells in the cortex.
The human tissue produced all the major types of neurons known in the cortex. Stanford also reports a rare cell type called von Economo neurons, which are considered vulnerable in some forms of dementia.
Long-distance wiring appeared as well. Processes from human cells reached as far as the spinal cord, and the neurons fired in synchronized bursts, a sign of some coordination.
Large-scale structure did not follow. Cell types that normally occupy separate cortical layers tended to cluster near each other, but the layers themselves never formed.
Pasca has pushed back on labels such as “humanized mice” or “mini-brains.” In his view, the animals keep a mouse nervous system that simply holds a larger volume of integrated human cortical tissue.
Why Organoids Need a Body
Organoids are small, three-dimensional patches of tissue grown from stem cells. They produce many of the cell types, and some of the structures, found in real organs. That makes them a better model than flat layers of loose cells for diseases driven by interactions among specialized cells.
Their limits are clear, though. An organoid has no circulatory system linking it to a liver that could process the chemicals it produces. No immune cells travel through it.
Brain organoids face an extra hurdle. Real brain regions constantly trade signals with neighbors and with distant areas. A dish offers none of that, which weakens studies of diseases that disrupt communication between regions.
Scientists have implanted human neural stem cells into the brains of other species, where they join the host's nervous system. Those cells, however, work alongside healthy host neurons, which makes their own contribution hard to isolate. In 2022, Pasca's group placed human cortical organoids into newborn rats, where the tissue matured far better than in a dish.
Clearing out host cells brings its own trouble. Mice need their brain cells, and human neurons mature much more slowly than mouse neurons. With only about 21 days of gestation, a mouse may not wait for human cells to wire up, and without mouse tissue nothing supplies the signals that organize them.
Open Questions Before Disease Studies
The team has not yet published the careful anatomy and development work needed to show which structures the human cells form. It also has not measured variation between individual animals. Until then, nobody can link the modest improvements to specific functions.
The model does show one human trait. The transplanted cells responded to brief periods of low oxygen the way normal human cells do. That is still far from proving the system can stand in for a disease such as ALS.
The disorganized connections could turn out to be a lasting flaw. Only deeper study of what these cells do will reveal whether such brains can ever model human neural processing.
A Push Toward Human-Based Models
Regulators want research tools that behave more like people. In 2025, the U.S. Food and Drug Administration unveiled a plan to reduce and eventually replace many animal tests in drug development, naming lab-grown organoids among the preferred methods. Xenocortical mice fall between the two approaches. They still rely on animals, but they aim to show how human tissue behaves inside a living body.
Pasca's group frames the work around neurodevelopmental and psychiatric conditions that begin before birth. In Stanford Medicine's account of the study, the researchers describe the model as a way to test how disease-linked changes to human brain circuits play out in an intact nervous system. Studies like this also keep attention on the ethics of placing human neural tissue in animals, a question that grows sharper as grafts become larger and better connected.
Written by Vytautas Valinskas