A DOZEN tiny, creamy balls are suspended in a dish of clear, pink liquid. Seen with the naked eye, they are amorphous blobs. But under a powerful microscope, and with some clever staining, their internal complexity is revealed: intricate whorls and layers of red, blue and green.
These are human brain cells, complete with branching outgrowths that have connected with one other, sparking electrical impulses. This is the stuff that thoughts are made of. And yet, these collections of cells were made in a laboratory – in this case, in the lab of at the University of Cambridge.
Advertisement
The structures, known as brain organoids or sometimes “mini-brains”, hold immense promise for helping us understand the brain. They have already produced fresh insights into how this most mysterious organ functions, how it differs in people with autism and how it goes awry in conditions such as dementia and motor neurone disease. They have even been made to grow primitive eyes.
To truly fulfill the potential of mini-brains, however, neuroscientists want to make them bigger and more complex. Some are attempting to grow them with blood vessels. Others are fusing two organoids, each mimicking a different part of the brain. Should they succeed, their lab-grown brains could model development and disease in the real thing in greater detail than ever before, paving the way to new insights and treatments.
But as researchers seek to make mini-brains genuinely worthy of the name, they move ever closer to a crucial question: at what point will their creations approach sentience?
The key to developing organoids was the discovery of stem cells, which have greater ability to multiply and be manipulated in a dish than ordinary bodily cells. The field leapt forward with the development of “induced pluripotent stem cells”, in 2006, when Japanese researchers showed that it is possible to take a sample of skin cells from an adult and genetically reprogram them into stem cells resembling those of an embryo just a few days after conception. With the right chemical cues, the cells could then be coaxed to mature into all manner of other tissue types, including brain cells or neurons.
The downside of growing neurons on a flat surface in a dish is that it is an artificial setting. In real life, cells exist in three dimensions, jostling up against each other and in constant communication. The key to creating a three-dimensional ball of cells is to inject stem cells into a small blob of gel, suspended in liquid and constantly jiggled to stop it from sinking to the floor of the dish. Lancaster was among the first to do this with human stem cells that were then nudged into becoming neural stem cells. She found that the cells multiplied and developed into different tissue types, until they formed a ball – or a cerebral organoid.
Growing brain-like tissue
In 2013, Lancaster described how, after two months, her organoids had developed lobes of brain-like tissue containing dozens of different types of mature neurons and immature neural stem cells in separate layers, organised around fluid-filled cavities. “The beauty of organoids is that they generate this tissue architecture,” says Lancaster.
The promise was clear: organoids offered a unique opportunity for neuroscientists to bypass the practical and ethical issues with using human embryonic tissue samples to study the brain. Many other labs rushed to join the field, experimenting with different methods and coaxing their organoids into replicating different parts of the brain.
The balls of cells follow similar patterns to embryonic development, in terms of which cell layers form, which genes are active and which proteins are made. “One of the fascinating things is that there’s a self-patterning process,” says at the Salk Institute for Biological Studies in La Jolla, California. “You can just put [stem] cells in a dish under certain conditions and they’ll form into structures that are quite remarkable.”
Equally startling is that the mature neurons fire electrical signals, and different neurons sometimes show patterns of synchronous activity known as brainwaves, detectable by placing mini-brains on a grid of recording microelectrodes. They are , according to at the University of California, San Diego, who has done some of the most attention-grabbing work on mini-brains.
Some groups have seen a further change in the organoids’ behaviour around the nine-month mark. It is as if the cells register that if they were in a real-life brain, they would be born around that time, says at Stanford University in California. “They start to resemble cells of the postnatal brain. They change their profile, the proteins they express.” For instance, in newborn babies’ brains, there is a change in the shape of a molecule called the NMDA receptor, involved in chemical signalling between neurons. This is also seen in Paşca’s mini-brains .
This isn’t to suggest that mini-brains are the same as the brain of a real fetus or baby, not least because they cannot be made to grow beyond a few millimetres in diameter. But they have already produced some intriguing results.
Lancaster’s main goal is to understand brain development and evolution. The two questions are linked because evolutionary change usually happens by modifying existing embryonic development pathways. “It’s such a fundamental question that human beings have been wondering for millennia,” says Lancaster. “What makes humans special as a species?” By comparing human mini-brains with those made from cells from chimpanzees and gorillas, last year her group found a genetic change that helps explain how human brains got so big. In the uterus, our neural stem cells spend longer at a stage where they are multiplying and so .
Other groups are creating mini-brains that mimic various medical conditions, made using skin cells taken from affected people. If genes play a major role in the condition, the organoid should share some features with the brains of those affected. Take motor neuron disease, which causes progressive paralysis, and frontotemporal dementia, an unusual form of dementia that tends to strike relatively early, in people’s 50s. These two conditions may seem unrelated, but, in fact, they are often caused by mutations in the same small set of genes.
Those mutations usually diminish brain cells’ ability to cope with damage to proteins and DNA that is linked to rising levels of a protein called poly(GA). And sure enough, the neurons of organoids made from cells taken from people with these conditions also showed these same features, , according to work published last year by at the University of Cambridge. That means he can watch how organoids develop the features of a disease over time. “You can see what’s going wrong,” he says.
Organoids are also being developed to help us understand autism, where people can have difficulties with communication and social skills and may have repetitive behaviours. Several groups using organoids made from cells from autistic people have uncovered hints about a key difference in their brains.
at Yale University and her colleagues looked at the balance between “excitatory” brain cells, whose raised activity triggers the firing of downstream neurons, and their rarer “inhibitory” counterparts, which block firing in subsequent neurons. They found that . It still isn’t clear how this relates to features of autism, but Vaccarino suggests that the greater number of inhibitory neurons could affect how the fetal brain wires up. “They regulate early circuitry,” she says.
Vaccarino adds that her findings show that organoids can shed light on genetic brain conditions even when the mutations responsible are unknown. The families in her study have some members with profound autism and some without, but the genetic variants responsible – which seem to be different for each family – are still undetermined. Yet the elevated numbers of inhibitory neurons in all the organoids made from cells from autistic people could be traced to higher levels of a molecule called FOXG1. Presumably, many different autism genes act through FOXG1, which is now a prime target for further investigation.
The power of organoids
These are important insights, demonstrating the power of organoids to study the brain in ways that would otherwise be impossible. But organoids are still an imperfect model of the human brain. The most obvious difference is that, for the moment, their growth stalls after a few months, when they are just a few millimetres wide.
The reason brain organoids don’t keep growing is because they lack a blood supply to deliver oxygen and nutrients. Unlike natural tissue, organoids must rely on whatever seeps in passively from the nutrient liquid in their dish. And once they reach a certain size, that isn’t enough. They stop growing and cells at their centre start to die. Vascularisation isn’t only important for size, though. The growth of blood vessels also regulates neural development, cell migration and circuit formation during early brain development. Their absence, then, means organoids aren’t as realistic as they could be, and that limits how much we can learn, says Lancaster. “I’m not sure that they’re really what I would call healthy brain tissue.” The problem is that it has proven extremely tricky to vascularise any lab-grown tissue, never mind brain tissue. Not that it has stopped people trying.
Gage’s approach is to transplant a small, immature organoid into the brain of a mouse, which lets the mouse’s blood vessels invade the human mini-brain. His team found that the human brain cells accepted ingrowing blood vessels and . The big drawback, as Gage acknowledges, is that you no longer have a completely human mini-brain.
At least two groups have managed to get human blood vessels growing in human mini-brains, although the work is still in its early stages. One group did this by first creating blood vessel organoids from immature stem cells, and then fusing these with brain organoids – the result was vascularised mini-brains that lasted for at least 50 days. Another group has added to some of the neurons a gene normally active in developing blood vessel cells, and made it switch on when the organoids were 18 days old. In effect, this made brain cells turn into blood vessel cells, and they self-organised into a network of blood vessels.
The resulting brain organoids seemed to have more mature neurons, although they were no bigger than ordinary organoids, says In-Hyun Park at Yale School of Medicine, who led the work. His team is now trying to encourage the structures to grow a lymphatic system, a network of tubes similar to capillaries that take away waste from tissues.
Muotri and his colleagues take a different approach. They grow mini-brains around fibres made from a material that dissolves if the dish is exposed to certain chemicals. After several weeks, the fibres are destroyed, leaving hollow channels inside the organoid. From that point, the organoid is kept in a chamber that gently pumps the nutrient fluid through the channels. Immature blood vessel stem cells are introduced and they stick to the walls, flatten out and become the walls of blood-vessel-like tubes, says , also at the University of California, San Diego. The work hasn’t been published yet, but she says that perfusing the organoids with the nutrient liquid changes them, leading to more mature synapses, the junctions between neurons.
Others are exploring another route to greater complexity: to make several organoids that mimic different parts of the brain, then connect them to each other, making a larger structure known as an assembloid. Paşca, for example, has joined up a cortical organoid to a spinal cord organoid, which connects to a bundle of muscle fibres in a dish, in order to simulate the circuitry that controls movement in living animals. When the cortical organoid is stimulated, the .
Migrating neurons
Assembloids are already shedding light on another key feature of fetal brain development, the migration of neurons. “It’s almost a rule: neurons do not rest in the place they were born,” says Paşca. “That migration is not just a movement. Very often, it involves maturation that’s only triggered by migration.” When Paşca’s group merged a cortical organoid with an organoid representing a part of the brain called the subpallium, it triggered neurons from the latter to migrate into the former.
The other major deficit of mini-brains, as things stand, is that they lack sensory input from the environment, which is necessary for brain circuits to develop normally. Muotri has taken the first steps in trying to provide that for his mini-brains, by wirelessly connecting them to small, crab-like robots that move around the floor. “The stage where we are now is closing the feedback loop,” he says.
The organoids sit on top of an array of electrodes that can both detect electrical impulses in the brain cells and feed information back. “The robot is communicating what it is seeing to the organoid and the organoid is reacting to that,” says Muotri. While there are , the group hasn’t yet published the results of how the feedback loop affects the mini-brains.
It remains to be seen which of these methods will be the best for making bigger and better mini-brains. But assuming that one or more of them will succeed, at some point, researchers will potentially confront the opposite problem: how to make sure their creations don’t become too lifelike. That is because they have to consider the possibility that their brain organoids could become sentient, and therefore capable of feeling pain.
Sure, they are nowhere near as complex as the human brain – but neither is the brain of a mouse, and there are strict laws on what kind of experiments may be done on mice, says Lancaster. The development of brainwaves in organoids, as shown by Muotri, don’t amount to consciousness, but Muotri predicts that organoids will inevitably reach some kind of mouse-like level of sentience. “We’re gonna get there,” he says. “And then we’ll have to deal with the consequences.”
How would we know if organoids do approach the consciousness levels of a mouse? And if any do, would we be obliged to destroy them – or to keep them alive for as long as possible? Addressing such questions is going to be hard because we don’t understand the scientific basis of consciousness. Ironically, mini-brains could help with that. But at this stage, no one knows quite what they will be able to teach us, says Gage. “We are just at the beginning.”
New Scientist audio You can now listen to many articles – look for the headphones icon in our app newscientist.com/app
