Stanford Medicine Team Grows Human Brain Tissue Inside Mice Engineered to Lack a Cortex

For more than a decade, neuroscientists have grown tiny clusters of human brain tissue in laboratory dishes, coaxing human stem cells to self-organize into structures called cortical organoids that echo the earliest stages of the developing human cortex. These pinhead-sized assemblies have given researchers a rare window into human brain development, but they carry a fundamental limitation: isolated in a dish, they can mature only so far. Without a living body around them, they lack the blood supply, the electrical chatter and the connective wiring that shape a real brain.

Transplanting those organoids into animals offered a partial solution. In 2022, a Stanford University team led by neuroscientist Sergiu Pasca placed human cortical organoids into the brains of newborn rats. The human cells integrated and matured, eventually making up about one third of one side of the rat cortex. Yet the experiment ran into a wall. The rat's own cortex was already there, competing for space, and rat neurons mature faster than human ones, which capped how far the human tissue could develop.

Now Pasca's laboratory at Stanford Medicine has removed that obstacle. Rather than squeezing human tissue into an occupied brain, the team built a mouse that develops without almost any cortex at all. The result, described in a study published online September 16 in the journal Nature, is what the researchers and outside scientists describe as the most extensive integration of human brain cells into an animal nervous system achieved so far.

The work sits at the intersection of stem cell biology, developmental neuroscience and bioethics, and it is drawing attention from researchers who study conditions such as schizophrenia, epilepsy, cerebral palsy and profound autism, illnesses that originate in the human cortex but remain difficult to model in animals.

Key Facts

The Stanford Medicine team began by genetically engineering mice so that the precursor cells that would have formed the cerebral cortex did not survive. The researchers used a conditional knockout strategy, deleting the cohesion factor Esco2 in Emx1-expressing cells against an immunocompromised SCID background. Stanford Medicine reported on September 16 that the resulting animals, which the team calls apallial mice, lack almost their entire cerebral cortex along with the neighboring hippocampus, two structures that together account for roughly half of the brain's total volume. Magnetic resonance imaging and whole brain single-nucleus RNA sequencing confirmed that the mice had lost the dorsal and medial pallium and its derivatives, including the neocortex.

Into that vacant cavity the researchers transplanted human cortical organoids, pea-sized clusters grown from healthy donor stem cells that had first been transformed from human skin cells. Science News reported on September 16 that the transplantation took hold in 25 of 29 attempts. The mice were 5 to 17 days old at the time of injection, an age at which the brain's core wiring is already in place, and each animal received roughly 100,000 human cells.

The grafted tissue then expanded dramatically. Nature reported on September 16 that between two and three months after implantation the human tissue grew nearly fivefold and filled more than 90 percent of the vacant space, reaching up to 92 percent in some animals. Roughly 14 million mouse brain cells were replaced by about 4 million human ones. The human neurons did not stay confined to the skull: they sent projections deep into the rodents' spinal cords, and the transplanted tissue showed spontaneous electrical activity.

The grafts developed a range of cell types that recapitulate the cellular diversity of the developing human cerebral cortex, including large spindly cells similar to von Economo neurons, a type linked to social cognition in humans that had never before emerged in a laboratory dish and is thought to be especially susceptible to frontotemporal dementia. The tissue did not, however, acquire features of a mature cortex such as its characteristic layers.

Behavioral testing kept the result in perspective. Unsupervised machine learning using motion sequencing revealed distinct behavioral repertoires, with the transplanted animals positioned between intact mice and mice that lacked a cortex. The apallial mice showed deficits on tasks requiring the neocortex and hippocampus, including fine motor coordination and working memory, and the grafted mice recovered only part of that performance, with no sign of added intellect. The primary research paper, published in Nature on September 16, reported that hypoxia, or oxygen deprivation, triggered a cellular response in the human tissue and produced deficits in the CatWalk test of walking, underscoring the platform's potential for disease modelling and therapeutic discovery.

Analysis

The bigger picture here is that this experiment is less about building a smarter mouse than about building a better instrument for studying the human brain. For years, the central frustration of psychiatric and developmental neuroscience has been that human cortical tissue cannot be observed while it is alive, developing and wiring itself. Nature reported on September 16 that the study describes the most extensive integration of human brain cells into an animal so far, a claim that matters because it points to a system that can be perturbed, imaged and measured rather than inferred from postmortem samples.

The technical judgement behind the work is sharp and somewhat counterintuitive. The decisive move was not adding human cells but subtracting mouse ones. As Giorgia Quadrato of the University of Southern California said, the real innovation is removing the competition for space. Previous transplants into rats, including the team's own 2022 experiment, in which human organoids filled about one third of a cortical hemisphere, left human neurons developing in the context of an intact rodent cortex and its faster maturing connections. The apallial mouse removes both constraints at once, and the human tissue responded by expanding nearly fivefold and occupying more than 90 percent of the vacated cortex within three months.

That result should not be overstated. The grafts lacked the layered architecture that defines a mature human cortex, and they matured inside a mouse, surrounded by mouse blood vessels, mouse immune signals and mouse neural activity. Meanwhile, the behavioral data suggest that the additional human tissue did not confer any obvious cognitive advantage, an important safety signal rather than a footnote.

Ethicists and reviewers appear to agree on where the line sits. Madeline Lancaster of the University of Cambridge, who was not involved in the study, noted that the procedure's timing, days after the mouse pups' birth and past the point at which the brain's core wiring is in place, ensures that human cells cannot take over complex thinking. Insoo Hyun of the National University of Singapore, who led an outside expert panel, argued that what matters is not how much human tissue an animal contains but what it enables. The work underwent extensive oversight, including review by independent bioethics panels.

Why It Matters

The immediate payoff is disease modelling. The team used the mice to model cerebral palsy, a condition that affects 3 in 1,000 Americans. After the researchers induced oxygen deprivation, the human tissue showed cellular signs of injury and the animals developed problems with gait and limb coordination, a chain of events that is difficult to reproduce in any other system. Beyond cerebral palsy, Pasca said the methodology should speed research into schizophrenia, epilepsy and profound autism. The scale of those conditions is large: one in 20 American adults is troubled by a severe psychiatric illness, more than 1 in 100 adults suffers from schizophrenia, about 1 percent remain epileptic and 1 in every 218 American children meets the criteria for profound autism.

The platform also offers a new way to test drugs on human cortical circuitry inside a living nervous system. Pasca described the advance in blunt terms, saying the team now has a very powerful new system for understanding what makes the human brain unique and uniquely susceptible to disease. The discovery of von Economo like neurons, cells that had been seen only in autopsied human brains and never in laboratory culture, is an early demonstration of what that system can reveal.

Just as important is how the work was conducted. The mice are immunodeficient, the human tissue comes from consented donor stem cells, and independent bioethics panels reviewed the study before publication, a sign that the ethical framework is being built alongside the biology rather than after it.

Next Up

The Stanford team and others are likely to push the model toward more diseases and toward drug screening, using the xenocortical mice to watch how human cortical cells respond to injury, genetic mutations and candidate therapies. Whether the human tissue can be coaxed into forming the layered structure of a mature cortex, and whether it can be kept healthy for longer than a few months, are open questions that will shape how far the platform can travel.

For now, the study marks a clear step beyond the 2022 rat experiments, in which human organoids filled about a third of a cortical hemisphere. By removing the host cortex instead of competing with it, the researchers have turned a mouse brain into a temporary home for human tissue, and they have done so with oversight, with behavioral data and without any claim that the animals think differently. That combination, more than the sheer volume of human cells, is what makes the result notable.