Short Wave - Mapping The Entire Fruit Fly Brain

Episode Date: October 7, 2024

Fruit fly brains are smaller than a poppy seed, but that doesn't mean they aren't complex. For the first time, researchers have published a complete diagram of 50 million connections in an adult fruit flies brain. The journal Nature simultaneously published nine papers related to this new brain map. Until now, only a roundworm and a fruit fly larva had been mapped in this way.Read more of science correspondent Jon Hamilton's reporting here. Want to know more about the future of brain science? Email us at [e-mail] — we might cover it on a future episode! 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. Hey there, short wavers, or should I say my fly girls and fly guys, Regina Barbara here. And this episode is for you. And that's because we're going to talk about fly brains, specifically fruit fly brains. And here to explain why we should care about an insect brain that's smaller than a poppy seed is NPR's own fly brainiac, John Hamilton. Hey, Gina. How's it going? It's going well.

And let me tell you why we should care about fly brains. The reason is that... Okay, you already got me. I already care. This little tiny brain, scientists think, is going to be able to tell us a lot about how all brains work. I mean, even in a human brain. And that is especially true now because researchers have published the first complete wiring diagram of an adult fruit flies brain. Spoiler alert, this brain has 50 million connections. That seems like a lot for a very tiny brain.

Yeah, impressive, right, for this little creature that has only about $100. 140,000 neurons. For comparison, a human brain has roughly 86 billion neurons. Wow. Okay, so this fruit fly brain wiring diagram, it's a big deal, right? It is in the world of brain science. It even has this catchy name, flywire. And it was unveiled with a big splash in the journal Nature. They simultaneously published nine different papers related to this new brain map, which, by the way, is called a connectome. Okay, I like this term to connectome. Tell me where that word comes from. So a genome, you know, has all the genes in a cell or an organism. Yeah, I've been repeatedly told this because I continually forget biology facts. Yes.

So a connectome shows all the connections between neurons in a brain. Until now, the only complete connectomes were for a round worm and a fruit fly larva and a fruit fly larva kind of acts like a worm so it's not that interesting. adult fruit flies are way more interesting. True. I mean, they do stuff like flying and fighting. They search for mates. I mean, these insects even know how to flirt. I don't know how that's possible.

Like, how do fruit flies flirt? Male fruit flies actually sing these kind of love songs by extending one wing at a time and vibrating it. I heard about this from one of the scientists who was involved in the Kinecto project. Her name is Malamurthy. She's a professor at Princeton who has spent the past 12. 25 years studying what else fly brains. They're actually capable of pretty extraordinary behaviors. So one example, the male fly that produces his song, he chooses exactly what to sing

based on feedback he gets from the female fly. And he patterns every note in accordance with her behavior. Oh, my gosh. That's a clearly highly evolved species. Today on the show, what scientists are learning from a map showing. all the connections in a fruit flies brain. Plus, how at least one fly brain may achieve a kind of immortality. Immortality?

Focus, Gina. Okay. You're listening to Shortwave, the science podcast from NPR. Okay, so John, I think we all know that neurons are important in the brain. When did scientists get so interested in the connections between neurons? It's happened over the past 20 years or so. And the reason is that you can now identify and label every neuron in the brain. but that still doesn't tell you how a brain works.

So those neurons have to form networks and to communicate in order to make a memory or a decision. And that means you need lots and lots of connections. So around 2005, scientists started using this term connectome to describe the wiring in a brain. Then in 2010, a scientist named Sebastian Song gave a TED talk. It was called I am my connectome. And it went viral. It has been watched well over a million times. And Sebastian is another one of the scientists who played a critical role in creating the fruit fly connectome.

Okay. So how do he and other scientists make one of these like connectome maps? Like, I mean, do you have to find 50 million connections in that tiny brain? Gina, let's take a little audio voyage. My favorite kind. To the place that made a lot of this fruit fly connectome research possible. Okay. So welcome to the Janelia Research campus of the Howard Hughes Medical Institute. It sounds beautiful.

It is. It's in Ashburn, Virginia, about 30 miles from Washington, D.C. And right now we are in the main building. It's low and modern and kind of terraced into the hillside. Okay. Inside this building kind of looks like an office, except that all the pictures on the walls are of fruit flies. And right now we're being shown around by a senior scientist named Kenneth Hayworth. I'm thinking there's an awful lot of technologies for doing kind of comics now.

Kenneth takes us to a room that's about the size of a squash court. There's a machine in the center that looks like this oversized refrigerator plugged into a life support system. I think tubes, wires, lots of stainless steel. But it's actually running a whole wide central nervous system at the moment. So don't bang the machine. Then you wanted to bang on that machine. Absolutely not, Gina. My intrusive thoughts are much darker.

So this is it, Gina. the place where flybrain sausage gets made. That involves a process called ion milling. They shoot atoms like a sandblaster at a slice of brain to remove just a tiny bit of tissue. Then they send what's left into that refrigerator-looking thing, which, according to Kenneth, is actually a microscope. What you're looking at right here is the world's fastest scanning electron microscope.

It contains 91 beams. So it's essentially taking 91 images at the same time. That microscope, it runs day and night, and it still takes a couple of months to image the single brain of a fruit fly, right? But the result is millions of images that show every detail. Including all those connections between neurons, right? Well, sort of. The information is there in all those pictures. But in order to trace those connections, you actually have to assemble those two-dimensional images into one massive 3D image.

Okay. And that is what those scientists at Princeton did to create. the first complete connecto. Okay, using the images from that electron microscope, we just saw. No. Okay. That microscope didn't even exist when those scientists started their project back in 2008. Oh, wow.

Okay. So they used images from an earlier system at Janelia, and those images were made by a scientist named Davy Bach. So we're talking about more than 21 million electron microscope images of the brain of one individual female fruit fly. Wow. What the prison team did, and they were working with scientists at the University of Cambridge in the UK, was to figure out how to use AI, you know, artificial intelligence, to stitch together all those images and then recognize the connections between neurons. But the AI system made some mistakes. So you had to have hundreds of scientists from all around the world who agreed to be proof readers, essentially, and to find and fix all those mistakes. And then they were done.

Not quite. The researchers wanted to know what sort of connections they were looking at. So, for example, you have some connections that amplify messages, you have other ones that inhibit them. And that meant figuring out what chemical was being produced at each synapse, you know, the place where a message jumps from one neuron to the next. You can do that in the lab, but it's incredibly tedious. So a scientist at Jan Funca wondered whether a neural network, You know, artificial intelligence could learn to recognize the different types of synapses just by looking at them. And that seemed a bit impossible because humans can't do it. But it turned out we got lucky. And the neural network can actually do that and can do that with very high accuracy. We're in the neighborhood of like 90% accurate predictions. And that was quite a surprise to us. That's like a nerving.

And I should mention here that this project also identified thousands of different kinds of neurons in the fruit fly connectum. So, you know, it was a scientific tour to force. Okay. So now the scientists have this complete fruit fly connectome. What can they do with it? Like, where does this immortality come in? Well, one goal is to study how the brain processes information and how that information ultimately affects a creature's behavior, right? So obviously a fruit fly isn't doing anything like, you know, picking a retirement plan to invest in.

It's really hard. Right? For this brain, it is. But even fruit flies, they do make decisions. I talked about that with Srini Taraga at Janelia, and we talked about something called the startle response in a fruit fly. So this is when if you're trying to go and swat a fly, it jumps and flies away.

This particular reflex requires that they visually use their eyes to detect that someone is coming to swat them, make a decision, and then fire neurons that would rapidly allow the fly. to jump and then take off. I mean, I love that they're studying this. I mean, fruit flies are pretty fast, right? Like, they seem to get away. I usually get them because I'm so fast.

But, you know. They do. In that way, their brain is working faster than ours. And one thing Srinni is trying to understand is why this is possible. He's got a whole bunch of questions he hopes to answer, and he's going to do it by simulating an entire fruit fly brain. Ooh, okay. So he already has this.

virtual fly, he calls up on his computer. And it's a bit like a game, you know. He has this fly image and it crawls and jumps and reacts to things in the environment because it's controlled by circuits that simulate some of the things that are going on in an actual fly brain. Cool. So now that the entire fruit fly connectum is available, he's hoping to include all of the circuits in this virtual fly brain. Then scientists can use the simulation to get some idea of how a real fruit fly brain responds to different situations.

John, I really want to play this game. Like, is it available for me to play? Not yet. Shrini's working on it. So, okay, so this game, like, this completed, simulated brain is controlling a completely simulated fly. This is sounding very sci-fi. I'm super interested.

Yeah, totally. And remember that this particular simulation isn't some generic brain. It is based on all of the connections in one specific female fruit fly. So this, like, female fruit fly. fly is now like a cyberfly? Not completely, at least not yet, but that is part of the goal. And the sort of creepy aspect of all this is something that scientists are thinking about, as you might imagine.

So you remember Sebastian sung at Princeton, who did the viral TED talk about being his connectome? Here's what he had to say about your cyberfly. I joke that one fly died for this connectome, but that fly could actually live forever in virtual reality. So mind uploading might be. might be science fiction for humans, but it's starting to become mainstream science for flies.

There's got to be a movie in that. Got to be. Fruit Fly 2030. Let's do it. Yep. Thank you, John, for bringing us this incredible fun story. My pleasure.

This episode was produced by Rachel Carlson. It was edited by showrunner Rebecca Ramirez and John checked the facts. The audio engineer was Quacey Lee. Beth Donovan is our senior director and Colin Campbell is our senior vice president of podcasting strategy. I'm Regina Barber. Thank you for listening to Shortwave from NPR.