Human cortical organoids fill a missing cortical niche in mice

Transplanting human neural organoids into rodents can provide vascularisation, sensory input and months of in-vivo development, but a small graft still has to compete with an intact host cortex. Kaganovsky, Kelley, Gschwind and colleagues change that starting condition. They first deplete much of the mouse's glutamatergic cortex and hippocampus, then use the available developmental niche to support human stem-cell-derived cortical organoids.

The central advance

The engineered immunodeficient host—described as an apallial mouse—creates space for human cortical tissue without removing the systems needed to sustain the animal. Four organoids were engrafted into neonatal mice in the optimized procedure. The grafts expanded, connected and occupied most of the combined cortical volume in animals analysed at three months.

This is more than a larger xenograft. The human tissue generated several cortical cell classes, including deep-layer extratelencephalic projection neurons, and sent projections into the mouse nervous system. Graft-wide calcium imaging and electrophysiology identified coordinated activity with features of developing neural circuits.

What the evidence shows

Across 29 mice and three human iPSC lines, the authors report an 86.2% graft success rate. Graft volume increased about 4.7-fold between two and three months. In a subset of seven animals from two lines, human tissue accounted for an average 91.9% of the combined cortical tissue volume at three months.

The mice retained broad locomotor ability, while showing selective changes in limb coordination and behaviour. These results support functional interaction between graft and host; they do not demonstrate human-like cognition or a human brain inside a mouse.

Why it matters for model builders

The platform moves organoid transplantation toward a whole-system perturbation model. A large, distributed graft may make it easier to connect cell-level phenotypes with circuit activity, injury responses and behaviour. It may also provide room for human tissue to develop projection patterns that are difficult to study in an isolated dish.

The host design is the enabling technology. By reducing competition from mouse cortical neurons, it changes the biological question from whether a graft can survive at one site to how human cortical tissue develops when given access to much of the relevant niche.

The necessary caution

This is a chimeric developmental model with a deliberately abnormal host. The mouse cortex is genetically depleted, the human tissue remains immature, and species-mismatched cues shape every stage of integration. Behavioural differences cannot be assigned to a single human circuit without further causal work. Ethical review must also develop alongside any increase in graft scale, integration or functional complexity.

HCMX take

The paper's value is not the sensational image of a “humanized” cortex. It is a technically controlled way to ask how human cortical tissue behaves when space, blood supply and long-range connections are no longer the main constraints. Used carefully, that could expose phenotypes that conventional organoids and small grafts cannot reach.