A small piece of human brain tissue was placed into a newborn mouse. What happened over the next few months was far more extensive than a simple transplant. The tissue grew through the space where the mouse’s cerebral cortex would normally have developed. It formed different types of human cortical neurons.
Some of those neurons sent long projections through the mouse brain and into its spinal cord. By three months, human-derived tissue made up 91.9% of the measured cortical tissue volume in the transplanted animals. The experiment comes from Stanford researchers led by neuroscientist Sergiu Pașca. Their findings were published in Nature on September 16, 2026.
The researchers call the approach xenocortication , a term for developing human cortical tissue inside an animal brain. The result sounds almost like science fiction. The biology is more precise than that. These were not ordinary mice receiving a human brain transplant. The animals had first been genetically engineered to lose most of the brain regions that normally form the neocortex and hippocampus. That created an unusual amount of space for the human tissue to develop.
Stanford did not just transplant human tissue; it changed the space around it
The clever part of the experiment happened before the human organoids went in. A human brain organoid grown in a laboratory is not a miniature adult brain. It is a three-dimensional collection of developing human neural cells that can organize into structures resembling parts of the cerebral cortex. Researchers have used these organoids for years, but they have a major limitation. They develop outside the body.
The Stanford team took a different route. They created what the study calls apallial mice . Through a genetic manipulation, the researchers depleted most glutamatergic neurons normally produced by the dorsal and medial pallium, the embryonic structures that give rise to much of the neocortex and hippocampus. MRI showed that the resulting mice had about half the total brain tissue of normal controls.
Newborn mice were then given human cortical organoids. The researchers typically transplanted more than one organoid into the enlarged cortical cavity. The organoids were already developing human neural cells when they were implanted.
The graft survived in 86.2% of 29 mice used to assess transplantation success. In another group followed with MRI, the graft volume increased about 4.7 times between two and three months after transplantation. By three months, it occupied almost all of the available cortical tissue volume measured in the study.
That is the detail that separates this work from a conventional organoid transplant. The researchers created an environment where the human cortex had room to become a substantial part of the animal’s developing brain.
The human neurons wired into the mouse, and some rarely seen cells appeared
Growth alone would not have been enough. The researchers wanted to know whether the human tissue could actually become part of the animal’s nervous system. Their experiments found evidence that it could.
Human neurons formed organized axonal pathways. Some projections traveled through deeper parts of the mouse brain. Others reached the cervical spinal cord. Tracing experiments also showed mouse neurons sending inputs toward the human graft. The two nervous systems were therefore not simply sitting beside one another. They were exchanging connections.
Electrical recordings added another layer. The human-derived cortex showed spontaneous activity across the graft. Calcium imaging revealed coordinated patterns that resembled developing neural circuits. The tissue was behaving as an active network rather than as a collection of disconnected human cells.
The researchers also used oxygen deprivation as a proof-of-concept injury experiment. The xenocortical mice were exposed to five hours of low oxygen. Their human-derived cortical tissue showed substantial damage, and the animals developed problems with gait and balance. Normal mice and the cortex-depleted mice did not show the same response.
That does not mean the model recreates cerebral palsy. It cannot establish that from one experiment. What it does show is that researchers can expose developing human cortical tissue to an injury inside a living animal and then measure both cellular damage and changes in behavior. That opens a route for studying how human neurons respond to conditions that are difficult to reproduce in a dish.
A mouse containing a large amount of human cortical tissue is not the same thing as a mouse with a human brain. The researchers tested the animals’ behavior rather than assuming what the human tissue might have done to them. Across several tests, xenocortical mice generally retained locomotion and performed broadly like normal mice.
The study did find selective differences, including changes in limb coordination and the organization of spontaneous behavior. That leaves the central ethical issue where it belongs: with the possibility that future versions of these models could become increasingly complex.
