Host: Benjamin Thompson
Welcome back to the Nature Podcast. This week: integrating human brain organoids into the brains of rats, and the exoskeleton boots that learn as you walk. I’m Benjamin Thompson.
[Jingle]
Host: Benjamin Thompson
A few weeks ago on the podcast feed, you might have heard one of our long read shows, where I read a feature article by Kendall Powell, which looked at efforts from labs around the world to transplant human cells into animal brains to understand more about things like brain development and disease progression. This week in Nature, one of those groups has a new paper out, demonstrating that human neurons transplanted into rats can integrate into the brain and even influence behaviour. Anand Jagatia takes up the story.
Interviewer: Anand Jagatia
In 2008, a team of scientists showed that you could take some human skin cells, reprogram them into stem cells, and get them to develop into complex 3D brain tissue – a small mass of brain cells known as a neural organoid.
Interviewee: Sergiu Pasca
Neural organoids are clumps of cells that are derived in a laboratory dish. They're not miniaturised versions of the entire brain. They are supposed to model aspects of the anatomy or function of the nervous system.
Interviewer: Anand Jagatia
This is Sergiu Pasca from Stanford University in the US. As Sergiu says, a neural organoid isn’t a tiny brain. The human brain is immensely complex, made of billions of neurons that form countless connections. But even so, since the 2000s, organoids have become a powerful tool for studying neural development and the evolution of the brain, as well as certain neuropsychiatric conditions.
Interviewee: Sergiu Pasca
We need models to study these conditions that maintain some of the genetic or genomic context. So, models built from stem cells derived from patients are important to start tackling or start asking questions about the biology of these neuropsychiatric conditions.
Interviewer: Anand Jagatia
In their efforts to better understand these conditions, researchers like Sergiu have connected multiple brain organoids together into circuits called ‘assembloids’. But an organoid in a dish will only develop so far, so if you want to mature it further, you have to transplant it into the brain of another animal, like a rat. This has been done successfully in adult rats, but, again, there are obstacles.
Interviewee: Sergiu Pasca
So, there are a number of limitations of transplanting into the adult rat. And of course, one of them is that the rat brain is already very well formed, so it is much more challenging for the organoid to integrate, just simply because the rat brain is, so to speak, less interested in forming new connections.
Interviewer: Anand Jagatia
But now, in a paper in Nature this week, Sergiu and his colleagues have overcome some of these issues using a novel approach – transplanting human brain organoids into newborn rats.
Interviewee: Sergiu Pasca
We place the organoids through surgery, right into the part of the cortex that responds to whiskers in the rat. Then we discovered that over the next few months, blood vessels will grow within the human graft, and that's how the graft survives. It can grow up to 9-10 times in size. And essentially, in the end, you obtain a unit of human cortex that sits onto one side on one hemisphere of the rat and covers roughly about one third of the rat’s hemisphere.
Interviewer: Anand Jagatia
This unit of human cortex didn't look exactly like it would in a human brain in terms of its structure and cell types. But the neurons in it grew and developed much more than organoids usually do in vitro. And the team showed just how much the human brain neurons could assimilate into the rat cortex.
Interviewee: Sergiu Pasca
For instance, we discovered that you could record the activity of human neurons inside the rat brain. And if you were to move the whiskers of the rat, you could actually see activity of these neurons following that stimulation. That tells us that some of these human neurons have connected to the pathway of the rat that responds to whisker stimulation.
Interviewer: Anand Jagatia
So, these human neurons were able to receive inputs, but were they also able to generate output? That is, could activity in the human brain cells influence the rat’s behaviour?
Interviewee: Sergiu Pasca
So, to stimulate human neurons, we leverage a technology that involves putting inside the human neurons a protein that is sensitive to light. So, when you deliver blue light to these neurons, neurons become electrically active. We train the rats to associate stimulation of human neurons with delivery of a reward. And we found that, indeed, if you do training for a couple of weeks, rats will learn that and you can stimulate human neurons and trigger in the rat reward-seeking behaviour. So, in this case, the rat will seek to drink water. That tells us that human neurons are well integrated into the circuitry, but I think it speaks more to the potential of this approach moving forward.
Interviewer: Anand Jagatia
Sergiu thinks this system could be used to more accurately model disorders affecting the brain, as well as to test drugs for those conditions. And by studying how these organoids mature in vivo, it might one day be possible to replicate that development more realistically for in vitro models. But of course, research like this raises a whole host of ethical questions concerning animal welfare, human donor consent, and the consequences of putting human brain tissue into another species.
Interviewee: Sergiu Pasca
We've been very carefully monitoring the animals to make sure that there's like no suffering. For instance, that the animals would not develop any seizures. Another ethical aspect is whether we're going to see any augmentation or alterations of the behaviour or physiology of the host. One question is whether you would expect that the rats will perform much better in certain tasks, right, for instance, in a memory task, and we did not find that that was the case. We did not find any changes in the behaviour of the rat.
Interviewer: Anand Jagatia
In these experiments, the rats with human neurons growing inside them didn’t behave any differently to controls. But what would the ethical implications be if they did? If they had better cognitive abilities, would that make them more human? And there are bigger questions around whether organoids can have moral status or even consciousness. These issues will become more pressing as the field develops. But as far as this study goes, Sergiu points out that the fraction of human neurons is small – about 1 or 2 million out of 30 million rat brain cells. And there are differences between the human and rat cells that Sergiu thinks put a limit on how integrated they can become.
Interviewee: Sergiu Pasca
It takes more than a week for the rat brain to develop. But it takes more than 20 weeks for humans to make all the neurons in the cerebral cortex. And so, by the time human neurons just start to form in vivo, the rat brain is already quite advanced. So, there's always this disconnect, so to speak, between the two species, and I think that that poses some natural barriers to how advanced this integration can actually be.
Interviewer: Anand Jagatia
Navigating the ethics of all of this will require active conversations between scientists, ethicists and the public to determine how this field should – and should not – progress. But Sergiu adds that that we also have to think about the ethical consequences of not performing this kind of work.
Interviewee: Sergiu Pasca
I mean, the biology of the human brain is still a mystery to a large extent, and so is the biology of neuropsychiatric disorders. And it's becoming clear that moving forward, we're going to need human models to also tackle these diseases. And certainly, the more human the models are, the more concerned we are about, ethical aspects that surround this work. But together, I think we have a moral imperative just to support and to study these disorders that are those that are causing so much suffering.
Host: Benjamin Thompson
That was Sergiu Pasca from Stanford University in the US. Look out for a link to his paper, and an associated News and Views article, in the show notes. Coming up, we’ll be hearing about some exoskeleton boots that could help people walk more efficiently. Now, though, with a slightly different feel to usual, it’s time for the Research Highlights, read this week by Noah Baker.
[Jingle]
Noah Baker
Ever wondered what gives jazz its unique sound? That groovy, foot-tapping, head-bobbing je ne sais quoi. Well, there’s a name for it: swing. And it’s all about timing – subtle deviations from the beat to create a mood. And now, a team of researchers are trying to quantify just how much musicians need to deviate for that thing to become swing. The team digitally manipulated four jazz recordings, shifting the piano soloists’ timing with respect to the rhythm section. In some versions, there was a delay of 30 milliseconds in a selection of the soloists’ speeds. In other versions, all of their beats were delayed. The team then asked a group of jazz musicians to listen to both versions and asked them: did it swing? By a significant margin, participants rated the versions where only some of the beats deviated to have more swing. They reported these having a pleasant friction between the performers, although they couldn't pinpoint the nature of the difference. You can read more in Communications Physics.
[Jingle]
Noah Baker
China is the world's largest producer of cement, steel and other building materials – the production of which emits huge amounts of carbon dioxide. But a new analysis is suggesting that powering these heavy industries with hydrogen could be a cost-effective way to reduce China's carbon emissions – and its contribution to climate change. China aims to reach net-zero carbon emissions by 2060, but decreasing carbon output in heavy industry is challenging. Clean hydrogen, however, provides an opportunity. This is hydrogen that's created using renewable energy or decarbonised fossil fuels, and it yields only water when it's burned. So, a team of researchers in the US has modelled how clean hydrogen could be used in China, and its cost effectiveness. And they got some interesting results. For example, they found that by 2060, clean hydrogen could supply 29% of the energy demand for steelmaking. The analysis also showed that by turning to hydrogen, China could avoid spending nearly US$2 trillion between 2020 and 2060 on other clean-energy solutions. What's more, the study also indicated that clean hydrogen could help other countries shrink their carbon footprints from heavy industry. Read more in Nature Energy.
[Jingle]
Host: Benjamin Thompson
Next up, reporter Dan Fox has been investigating a new exoskeleton.
Interviewer: Dan Fox
Exoskeletons are devices designed to augment and enhance mobility, and they are fast becoming a reality, thanks to an explosion in research in recent years.
Interviewee: Patrick Slade
An exoskeleton is really beneficial because it allows us to improve certain aspects of our mobility.
Interviewer: Dan Fox
That's Patrick Slade from Stanford University.
Interviewee: Patrick Slade
For example, if we want to walk with less effort, or walk more quickly, or overcome, perhaps, a mobility disorder which limits our movements.
Interviewer: Dan Fox
But now, Patrick and his team have created a new exoskeleton, which they hope will overcome one particular challenge.
Interviewee: Patrick Slade
It's actually quite difficult to make walking easier, or to reduce knee pain, or to do these things that we care about because when we walk around every day, we walk in short bouts of walking, we change speeds. We're not walking on a treadmill at a fixed speed like we do in the lab.
Interviewer: Dan Fox
Patrick's new exoskeleton is smart and able to learn and adapt as you walk.
Interviewee: Patrick Slade
So, what we've built is an ankle exoskeleton, so this is with a motorised unit, sort of like a motorised boot, and it applies assistance as you're walking. So, as you walk, it can apply assistance with push off, and this can both make walking easier, so require less effort, and increase walking speed.
Interviewer: Dan Fox
Exoskeletons like this have been created before. But what sets this one apart is its ability to adapt. You see, to provide the most benefit, exoskeletons need to be carefully personalised to each user.
Interviewee: Patrick Slade
In the past, we've done this in the lab using a big machine which measures your breathing, your oxygen and your carbon dioxide to understand your energy expenditure. And it's this huge lab-based equipment, and it takes hours to run these experiments, so it's not really feasible to do on devices that you might purchase and wear. And so, one really cool part of this project is we figured out how to actually personalise assistance for each person as they're walking in everyday life.
Interviewer: Dan Fox
Now, Patrick's device doesn't change its shape or mould to the user's leg, but it does specifically tailor how it provides assistance to suit every user. And importantly, it does this on the fly, as it's being used – no big lab calibration or bulky equipment required. Instead, when a user first puts on the device, the exoskeleton immediately starts to monitor the way they move.
Interviewee: Patrick Slade
So, as they're walking in short bouts, starting, stopping and also changing speeds, the device is able to collect data based on how they're moving and figure out what's the best way to assist each person.
Interviewer: Dan Fox
There are various ways that the exoskeleton can tweak its support, from the timing of a boost to a step to the torque or angle of the assistance. But to work out which intervention may be most helpful, it runs a model trained on lab data which predicts metabolic benefits. This then allows the exoskeleton to try different forms of assistance, ranking and comparing until it's optimised for the best results.
Interviewee: Patrick Slade
And this doubles the benefits compared to a generic non-personalised controller.
Interviewer: Dan Fox
Because the system is designed to work for any user without extra programming, Patrick and his team wanted to really put it through its paces, so they sent subjects out onto their university campus wearing the exoskeleton, and prompted them to walk in a variety of different ways.
Walk as if you’re walking across the street.
Interviewer: Dan Fox
These are some of the prompts from the experiment.
Walk as if you're walking home after a really bad day.
Interviewer: Dan Fox
They found that with this real-world optimisation, participants walked around 9% faster and used 17% less energy while wearing the exoskeleton boots compared to normal shoes. Patrick says this optimisation technique could have an impact across robotics.
Interviewee: Patrick Slade
The really exciting science takeaway is that we've developed a method for robots to optimise as humans wear them during everyday life. And so, this idea and these approaches can be broadly applied to many different kinds of robotic devices beyond just exoskeletons where you have a human and a robot working together. And so, this could be an exciting chance to see developments in, for example, human-robot worker collaboration or other aspects of sort of smart home systems.
Interviewer: Dan Fox
And he hopes that further research will allow the ankle exoskeleton to help those who need it most.
Interviewee: Patrick Slade
We've never seen a device that you can put on and take outside and move how you normally do – starting, stopping and changing speeds – and actually see mobility benefits. And so, that's really exciting because in the lab, these devices have been very promising and there's been decades of research, and now we're seeing that these can be translated to perhaps commercial products that you can buy and wear and improve your mobility and be able to do things that you otherwise couldn't do. We're excited and starting to study how we can apply these same ideas and devices to older adults, for example, to help make walking easier and increase speed of walking, and how specific populations who have other challenges beyond just walking effort and walking speed, for example, knee pain, we can try to improve other metrics using the same sorts of personalisation approaches in the real world.
Host: Benjamin Thompson
That was Patrick Slade from Stanford University in the US. Look out for a link to the paper in the show notes, and to a video featuring Patrick's new design, which is up on our YouTube channel. That's all for the show. But before I go, just a quick bit of info. Next week, Nature has a special issue focusing on racism in science, and we're doing the same here on the Nature Podcast. So, we'll have something a little different for you – a special mini-series investigating racism in health. Look out for that wherever you get your podcasts. As always, don't forget you can keep in touch with us over on Twitter – we’re @NaturePodcast – or you can send us an email to podcast@nature.com. I'm Benjamin Thompson. Thanks for listening.
