Human brain organoids come of age—but still have room to grow

As the U.S. and Europe increasingly shift away from animal research, organoids have been touted as a promising alternative method for studying the brain.

The field has made significant progress since human brain organoids debuted in 2013. Researchers can now coax these clusters, fashioned from stem cells, to integrate a variety of cell types, including microglia, astrocytes and interneurons. They can even fuse brain-region-specific organoids into structures called assembloids that closely resemble circuits within the brain.

But in reality, these clumps of cells have “major shortcomings,” says Christiaan Huffels, a postdoctoral researcher who works with organoids in the lab of Elly Hol at University Medical Center Utrecht. These drawbacks include a lack of vascularization, which limits cell growth and survival and can stall the tissue’s maturation. “If we currently look at the electrophysiological profile of the neurons within the organoids that we use, I think you could compare it to premature or immature neuronal phenotype electrophysiologically,” Huffels says.

Organoids also lack the cytoarchitecture seen in the human brain, says Hongjun Song, professor of neurological sciences at the University of Pennsylvania. And there is the issue of variability: Each lab has its own protocol for growing organoids, so standardizing across labs, and even within the same lab, to get reproducible organoids is difficult, he says.

Neuroscientists are leveraging a variety of strategies to make more complex and biologically accurate organoids that are more reproducible and representative of a later stage in human development, Hong says. “The next frontier now is moving beyond general cell types, and see [if we can] study other biology. For example, now can we study circuitry formation? We’re not really quite there yet, but there are methodologies to study the next step.”

These improvements are vital for organoids to ever replace animal studies, says Corinna Ross, director of the Southwest National Primate Research Center. But that time has not yet arrived, she says. “To be blunt, garbage in is garbage out. And so if the organoids and the science isn’t ready, then the questions you’re asking won’t be answered.”

O

ne primary limitation for organoids to overcome is their lack of vasculature—although researchers are deploying various work-arounds.As an organoid balloons from a microscopic ball to a 5-millimeter spheroid, roughly the size of a pea, the cells deep within the structure lose access to the outside world. With no functional vasculature, outside molecules can diffuse only so far. As a result, the centermost cells fail to get sufficient nutrients and oxygen and eventually die off, leaving behind a necrotic core that can skew the development of the remaining cells.

A blood supply doesn’t just help cells survive; it also dictates normal development, says Alysson Muotri, professor of pediatrics and cellular and molecular medicine at the University of California, San Diego. “Vascularization is intrinsic to brain formation.”

Adding endothelial cells to a brain organoid or joining it with a vascular organoid can help deliver blood and nutrients to more cells within an organoid, says Hol, professor of glial biology. And applying or overexpressing angiogenic growth factors further boosts vascularization within the culture, studies show. These strategies endow the entire organoid with vessel-like structures that express junction proteins associated with the blood-brain barrier.

Neuroengineering also comes in handy. For example, researchers can grow brain organoids on scaffolds that mimic the vascular network or 3D printed vessels that distribute nutrients. The organoids grown with an engineered vasculature have improved proliferation and neural differentiation, studies show.

Slicing up an organoid after a few months of growth serves as another alternative: It gives the internal cells access to oxygen and nutrients, circumventing the need to vascularize. Air-liquid-interface organoids, in which the sliced-up structures’ top surface is exposed to air while keeping the bottom of the slice in contact with the culture media, can last for months, surviving for longer than whole organoids, and can develop functional axonal tracts. Using slices even lets the organoid’s cells proliferate and create an expanded cortical plate, work by Song shows.

These strategies may help protect brain organoids from hypoxia and necrosis but ultimately are not optimal ways to truly vascularize organoids, Hol says. “We do get endothelial cells in there. We sometimes do get something that looks like vasculature, but not one that is really functioning.”

And that limits just how healthily organoids can develop, Huffels says. Studies typically use 3- to 4-month-old organoids, which is roughly equivalent to the brain of a second-trimester fetus and before necrosis sets in. That provides a crucial timeframe to study development, but it limits researchers in how they can use organoids to study a more mature brain and processes such as neurodegeneration, Song says.

One way to speed up the maturation of brain organoids is by transplanting them into rodent brains, which can provide a supportive microenvironment—including a functional vascular system. Organoids grafted into the cortex of rodents integrate into the brain and take on more complex morphologies than their in-vitro counterparts, studies have shown.

Sending organoids to space also ages them faster. Organoids grown in orbit on the International Space Station for a month exhibit more genes associated with neuronal maturation compared with those grown on Earth, work from Muotri shows. And space-bound organoids also accumulate senescence features, a mark of aging, Muotri’s unpublished work has found. Those 30 days represent 10 years of aging on Earth, Muotri says. “We compress their developmental time in weeks.”

Tweaking the culture conditions, such as by keeping organoid density low in the culture plates and changing out the media more often, can improve the availability of oxygen and nutrients, minimize development of the necrotic core and preserve the organoid longer, a recent protocol suggests.

“To ask whether we can reach stages of maturation that are later in life, we need to grow these organoids for very long periods of time,” says Paola Arlotta, professor of stem cell and regenerative biology at Harvard University. Organoids age in lockstep with the human brain all the way up to 7 years, her work has found. These aged organoids acquire the appropriate mix of cell types and epigenetic features that show that they are indeed maturing similarly to a postnatal human brain, thanks in part to an activity-permissive medium (APM).

The organoids still have a necrotic core, but the signatures of hypoxia and apoptosis are attenuated in the APM culture conditions, Arlotta adds. Getting rid of the core completely “will most likely require the development of systems that allow media perfusion into the center of the organoids.”

T

he ultimate aim of all of these efforts is to create organoids with the correct cytoarchitecture: the right cell types in the right location connecting with the appropriate regions at the right time, Song says. And to that end, researchers have made advances in creating organoids with all six major brain cell types—neurons, astrocytes, microglia, oligodendrocytes, pericytes and endothelial cells—and organoids that resemble specific human brain regions, including telencephalic organoids that mimic the cellular composition, development and other features of the amygdala.One way researchers try to recreate the connections seen in a human brain is by developing region-specific organoids and fusing them together into assembloids, an approach pioneered by researchers at Stanford University. Last year, they generated somatosensory, dorsal spinal cord, thalamic and cerebral cortical organoids, recreating the connections in the pain pathway and showing that these four-part assembloids could respond to noxious stimuli. More recently, the team fused midbrain-hindbrain organoids, which contain serotonergic neurons, with human cortical organoids to study neuromodulatory systems.

But one challenge for these organoids is that the constituent cells may develop out of step with the brain. For example, radial glial cells produce neurons earlier than the mouse brain does in vivo, according to a new study. These differences are partly due to the organoid’s lack of the timely signals seen during development in the human brain, says Song, who was not involved in the work. During normal development, connections arise through a dynamic and reliable dance of morphogens, signals that are missing in organoid cultures.

“There’s lots of things communicating there that we are missing from our model,” Muotri says.

Promoting organoids to create their own morphogen-secreting centers or using platforms such as hydrogels aim to better simulate normal development. But these approaches often provide morphogens uniformly to the organoid, Song says. By contrast, microfluidic devices can provide signal gradients with spatiotemporal precision—and have enabled researchers to create a neural-tube-like structure, a study from 2024 shows.

But even with such precision, organoids can be highly variable.

Using the same protocol with the same cell line can still result in heterogeneous organoids, Muotri adds. “In the same way that [if] I give a recipe for a cake to two or three people, each one will present a different cake, recipes of organoids are in the same way: same protocol in different hands—even in my lab—huge variability.”

In the positive column, this variety of organoids reflects human variability, Hol says. But too much variability also hinders researchers, Song says, so the field needs to work on standardizing protocols to have reproducible organoids across labs—for example by using engineering approaches such as scaffolds to grow organoids of the exact same size, architecture and composition.

For some, including Muotri, organoids are a stepping stone toward eliminating animal research. “Full replacement should be the main goal,” he says.

But will organoids ever improve enough to be able to do so? “I think it all depends on the questions,” Song says. For studies examining tumor biology, for example, it is possible, he says. For others, such as those that look at function at the circuitry level, even if the field addresses all the problems, “we will need both.”

With reporting by Calli McMurray.