Human-Derived Brain Organoids Create Functional Networks in Mice

For over a decade, Stanford University neuroscientist Sergiu Pasca has looked for strategies to study human brain functions and its disorders. In Pasca’s eyes, using human induced pluripotent stem cells to create human-derived brain organoids—three-dimensional, self-organized tissues grown in a dish—and transplanting them into rodent hosts is an opportunity to achieve this goal.

In a study published in Nature, his team described a new mouse model into which they transplanted human-derived neural organoids and showed how these structures integrate into the mouse brain.¹ The grafts not only developed into distinct human brain cells that show incipient signs of local organization, but they also extensively connected with the host’s nervous system to be functionally active and regulate the animal’s behavior.

“Understanding the brain and [its] disorders will require getting access to the cells and the circuits of the human brain,” Pasca said. “The model that we developed is one step forward in that direction, and it allows us to actually study human cells in much more physiological conditions. In doing so, [this] will hopefully allow us to ask questions about disease and therapeutic development.”

Previous work by Pasca and his team showed how human stem-cell-derived organoids transplanted into the developing brains of rats formed mature brain cells that connected with parts of the rodent cortex to influence behavior.^2,3 Still, Pasca noted, competition for connectivity with rat brain cells, as well as space constraints within the cranial cavity, posed limitations on how well those organoids integrated with the brain host. “It's still very difficult. You're adding human cells to what the rat already has, and rat cells will connect with each other much more efficiently, so human cells are essentially left behind,” he explained.

In the new study, Pasca’s team devised a strategy to buy some time for the neural organoids to develop and better connect to the host’s brain by providing them with more space. To achieve this, the researchers genetically engineered a mouse strain to prevent the formation of the cerebral cortex and hippocampus at early stages of development. Pasca noted that previous attempts to remove large portions of the rodent brain were very difficult, with animals showing severe motor and cognitive deficits and poor survival rates. To improve the new model’s odds, Pasca’s team spent years optimizing the rearing conditions of these mice, which allowed the animals to grow well and look much more like regular mice. Yet, “when you do an MRI, you can see that half of the volume of the brain is gone, and it's filled with liquid,” said Pasca. Quantification of the cell types through single-cell profiling revealed that 98 percent of the cortex and hippocampus were absent in these mice.

Image showing top-down view of mice brains. — By genetically engineering mice to prevent the development of a large portion of their brains, researchers were able to improve human-derived brain organoid implantation into the mouse brain. Left, middle, and right brains correspond to control, genetically engineered mice lacking brain portions, and graft mouse recipients, respectively.
Pasca lab, Stanford University. (CC BY 4.0)

Transplantation of the human-derived cortical cell organoids into the mice’s brains led to substantial graft growth that formed many human cortical cell subtypes, some of which have been challenging to obtain in previous in vitro and in vivo preparations. This was the case for the von Economo neurons, a set of cells primarily affected in brain disorders such as frontotemporal dementia.⁴ As these cells are one of the first to die in patients with this form of dementia, the fact that they were present in the mouse model may provide a way to better understand their role in that disorder, Pasca noted.

Using calcium imaging and electrophysiological recording, Pasca’s team also showed that these human brain cells connected with the host’s nervous system, exhibiting patterns of cell activity similar to those seen in neural circuits.

While the mice lacking large portions of their brains did not show severe behavioral deficits, they did show working memory impairments, which the researchers reversed with the brain organoid transplants. “[This] means that human cells are probably contributing somehow, even if indirectly, to the functioning of the animal,” Pasca said.

Finally, to probe the model’s ability to replicate disease phenotypes, the researchers exposed control and human brain organoid mouse recipients to lower oxygen levels to induce a hypoxia injury in the brain cells. Exposure to the hypoxic conditions led to the infiltration of microglia and astroglia into the graft, a response frequently seen after brain injuries, as well as to the expression of a cellular marker induced by low-oxygen stress. These findings suggest the model’s potential use to study perinatal hypoxia and associated conditions such as cerebral palsy.

While these results are informative for a milder type of injury, assessing how these animals would respond to more severe injuries would be an interesting next step, said Ruslan Rust, a neuroscientist at the University of Southern California who was not involved in the study. “[The study] is very fascinating,” added Rust. “It can be a really powerful [way] to study diseases in a more humanized model. If we treat human cells in a living organism, maybe that's more predictive of how successful something will be in a clinical trial later on,” he said.

Despite the model’s promising applications, Pasca explained it still has limitations, including the absence of GABAergic cell development within the graft, which could be necessary to create a more complex model, and the lack of cortical layering. “That's why the system is not an all-in-one, universal system that replaces everything,” he said. “It's just a very useful tool for asking specific questions and specific disease questions.”

References