Short Wave - Brain Cells In A Dish Play Pong And Other Brain Adventures

Episode Date: October 21, 2022

The world of brain research had two incredible developments last week. Researchers have taught a dish of brain cells to play the video game Pong to help develop more intelligent AI. Separately, scientists transplanted human brain organoids into a living animal with the hope of using them as models of human disease. Jon Hamilton talks with host Aaron Scott about this research and its implications.See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.NPR Privacy Policy

You're listening to Shortwave from NPR. Halloween season is upon us, shortwavers, and you know what that means. Brains! Human brains that can live on their own or bring monsters to life. Oh, my perfectest brain. Her brain kept alive by experimental science. By amazing. But I digress. We are a science show, and this is the stuff of science fiction.

I mean, you can't really bond a human brain to another animal, and certainly it's not like their brains out there in a petri dish somewhere, sitting around playing video games. Okay, so, Aaron, about that. Oh, hi, John Hamilton, MPR brain expert. Yeah, hi. I should just say it's not science fiction anymore. Go on. I mean, just last week, I,

I was looking at these two pretty amazing studies about how clusters of living brain cells are doing things that totally sound like sci-fi. Oh, so Halloween dreams do come true? Apparently, I mean, do you want to start with a study about the thing that sounds super creepy or the one that involves a video game from the 1970s? I want to save the creepy for last, so let's start with the 70s video game. I might have grown up in the OG Nintendo generation, but I did have cool older neighbors and they still had their Atari, and this, it sounds like Pong.

Correct. And just FYI, I am one of those extremely cool older neighbors who used to play Pong. Of course you are. Because, I mean, why would you get an actual ping pong table when you and your friends could sit in front of a tiny Minochrome monitor and make an illuminated dot go back and forth? Right, right, and thus begin a new generation of couch potatoes. But back to my point, the giant leap forward for science is that today, nearly 50 years after Pong change the world, a dish full of brain cells has learned to play it. It's all part of an experiment to understand how a network of brain cells is able to learn. That is a big deal because a living

brain has a kind of intelligence that a computer usually doesn't have. Here's an example. Let's say you're a tea drinker. You're at someone else's house and they send you to the kitchen to make a cup. What do you do? I rummaged through drawers looking for tea bags. I find a kettle or I pot, put water in it, put it on the stove, bada bing, bada boom, I got a cup of tea. Right. You kind of figure it out. And Brett Kagan, he's the chief scientific officer at a company called cortical labs in Melbourne, Australia.

Kagan told me that figuring stuff out is what living brains are really good at. You might have never been to someone else's house, but with a bit of rummaging and searching, you can probably make a decent cup of tea as long as I've got the ingredients. Simple for a person, really tough for a computer, even a powerful one. Computers really struggle. when they need to take something that they've learned in one context and generalize it to another. So cortical labs has been trying to understand how living brain cells acquire this kind of intelligence.

By teaching them to play pong. Aaron, by giving them the tools to learn to play pong. And also on the show today, I want to tell you about a study that could lead to new treatments for some brain disorders. But it also raises some questions about how far scientists should go in replicating aspects of the human brain. in another animal like a rat. I'm Aaron Scott, and you're listening to Shortwave, the Daily Science Podcast from NPR. Okay, John, so before we hand the joystick over to the brain cells,

can we just talk about the fact that there are actually jars or, you know, at least petri dishes of brain cells in labs doing things? Yeah, I mean, this is pretty new. So over the past couple of decades, scientists have gotten really good at using stem cells You know those cells that can turn into anything. They get them from animals or people, and they use them to grow neurons in the lab. They even know how to coax these neurons into becoming specific types of brain cells. And as the cells grow, they do what brain cells do, which is they make connections.

So eventually, you have not just a dish full of cells, but cells that are wired together into a network. Okay, okay. So let us continue on de Pong. Right. So Brett Kagan says the Pong experiment. was a way for the company to answer a question about a network of brain cells. So remember, we were talking about making a cup of tea easy for you, hard for a computer. Kagan had this idea that the cells might get smarter if they got some sort of feedback

that let them know when they screwed up and when they did something right. If we do allow the cells to know the outcome of their actions, will they actually be able to change in some sort of goal-directed way? So to find out the scientists used a system they created called, wait for it. Dish brain. Dish brain. Great. I mean, I know, I know, I know. It uses a layer of living neurons that are grown on a special silicon chip at the bottom of this thumb-sized dish filled with nutrients.

This chip is connected to a computer. That means the system can detect the electrical signals produced by the neurons. It can also deliver electrical signals to them. So in the experiment, the computer generated a game of Pong. then it started sending signals to the cells that told them where the ball was. Kagan says at the same time, the computer began monitoring signals sent by the cells. And what we did is we took that information and we allowed it to influence this Pong game that they were playing so they could move the paddle around. First, they needed to learn to interpret the signals telling them where the ball was, you know, kind of like learning to use their eyes. Then they needed to learn what signals to send to the computer to move the paddle in front of the ball. Oh, and they needed a reason to do all that stuff.

You know, motivation. Right. So Kagan says the system provided them what you might call rewards and punishment in the form of electrical stimulation. If they hit the ball, we gave them something predictable. So very, very simple, predictable stimulus was the same every time. When they missed it, they got something that was totally unpredictable, white noise, but different white noise every time. And it worked. The brain cells never really got that good.

But they got a lot better. And it's worth noting that when the system used human brain cells instead of mouse cells, they seemed to play even better. Kagan told me it's important to remember that all of these pong playing networks contained fewer cells than the brain of a cockroach. If you could see a cockroach playing a game of pong and it was able to hit the ball twice as often as it was missing it, you would be pretty impressed with that cockroach. I should probably mention here that the results were published in, the journal Neuron. Yeah, John, back to that fundamental question of why. I mean, most parents want

children to play fewer video games, and yet here scientists are teaching games to brain cells. Is this somehow about using biology to help computers learn? Yeah, it's about using biology to help artificial intelligence become more intelligence. It's also about seeing whether you can combine a biological element into an intelligent device, right? I talked to a guy named Stephen M. Potter. He's an adjunct associate professor at Georgia Tech. He told me that the future we're talking about here is probably still a long way off. The idea of a computer that has some living components is exciting and it's starting to become reality.

However, the kinds of learning that these things can accomplish is quite rudimentary right now. Right. So Pong. But even so, Potter says the Pong playing system is a great tool for doing research. This is sort of a semi-living animal model that one can use to study all sorts of mechanisms in the nervous system. Semi-living animal model. Yeah, that one sounds like a failed Halloween costume. Aaron, if you think that sounds like a failed Halloween costume, you are really going to love the next study. It involves brain organoids.

Now that one sounds like a real Halloween costume, or they like brains with legs? Tentacles. Oh, no. Just kidding. Just kidding. A brain organoid is a cluster of brain cells that usually grows to about the size and shape of a small p. They're typically grown in something called a bioreactor, which keeps them constantly moving in this kind of nutrient soup. And in many ways, each organoid mimics the early development of a human brain, which means they could offer up a compelling way to study brain disorders if researchers could overcome a few hurdles. I talked to Dr. Sergio Pashka of Stanford University, and he says these clumps of human brain cells are a bit of a disappointment because they really haven't yet revealed much about complicated disorders. No matter how long we've kept them in a dish, they still do not become as complex as human neurons would be in an actual human brain.

The brain develops by having the cells do something, you know, control an arm or beat a heart or blink an eye. And when a cluster of cells is just in a dish or a bioreactor, there's only so much they can do. So Pashka's lab began looking for ways to make organoids more sophisticated. The team decided to take the clusters out of a dish and transplant them into the brains of newborn rats. Pashka says it worked. We discovered that the graph grows over the span of a few months about nine times in volume. And in the end, it covers roughly about a third. of a rat's hemisphere.

Okay, this is too much for my little brain to process, John. What do the human cells do? Do they make the rats super smart? They do not make the rats super smart. In fact, you can't really tell that these rats have transplanted cells in them. They act normally. Okay. But what's interesting is that these cells, this clump of human cells growing inside

the rat's brain, it eventually, it starts acting like a part of the rat's own. brain. Nothing has been removed. The rat tissue is just pushed aside, but now you also have a group of human cells that are integrating into the circuitry. This is blowing my mind, integrating into the circuit. Wild. Right? Bashka's team placed each organoid in an area of the rat brain that processes sensory information. So, you know, the signals coming from the eyes or ears or nose or, you know, with rats, so whiskers. After a few months, he says, the human cells seemed to be reacting to whatever the rat was sensing. When you stimulate the whiskers of the rat,

the majority of human neurons are engaged in an electrical activity that follows that stimulation. Another experiment they did showed that the human cells could even influence a rat's behavior. So the team decided to study this rare genetic syndrome that affects the brain, the heart. It can also cause a form of autism. It's called Timothy syndrome, and it's really sad. kids who have it usually die before they get to age three. The team compared organoids made from the stem cells of healthy people with organoids made from the stem cells of patients with the syndrome.

Pashka says that while growing in the dish, all of the clusters looked exactly the same. But once we transplanted and we looked 250 days later, we discovered that while control cells grew dramatically, patient cells failed to do so. These results appear in the journal Nature, and Paola Arlata of Harvard University says they suggest a new way to study how psychiatric disorders affect the circuits in a human brain. But Arlotta says as brain organoids become more like actual human brains, scientists are going to have to consider the ethical implications of their research. We need to be able to watch it, consider it, discuss it, and stop it if we think one day that we are at a point where we shouldn't progress. I think we're far, far away from that point right now. Other scientists agree with her.

I mean, even the most advanced brain organoids are still really pretty basic. I talked to In Suhoun, a bioethicist at the Museum of Science in Boston. He says this sort of research often gets a negative reaction from non-scientists. There is a tendency for people to assume that when you transfer the biomaterials from one species into another, you transfer the essence of that animal to the other. Hune says in this case, the human brain cells are simply helping rats act like rats. John, I'm just going to hope we're not back here in a year talking about how the rats are rising up. Thank you for this look at some truly provocative research.

Always a pleasure, Aaron. Pleasant dreams. This episode was produced by Margaret Serino, edited by our supervising senior editor, Giselle Grayson, and fact-checked by Britt Hansen, our audio engineer was Jobi Tenseco. Brendan Crump is our podcast coordinator. Beth Donovan is our senior director of programming, and Anya Grunman is our senior vice president of programming. I'm Aaron Scott.

Thanks as always for listening to Shortwave from NPR.