How a fly flies
-
0:01 - 0:04I grew up watching Star Trek. I love Star Trek.
-
0:04 - 0:09Star Trek made me want to see alien creatures,
-
0:09 - 0:11creatures from a far-distant world.
-
0:11 - 0:14But basically, I figured out that I could find
-
0:14 - 0:17those alien creatures right on Earth.
-
0:17 - 0:19And what I do is I study insects.
-
0:19 - 0:23I'm obsessed with insects, particularly insect flight.
-
0:23 - 0:26I think the evolution of insect flight is perhaps
-
0:26 - 0:28one of the most important events in the history of life.
-
0:28 - 0:31Without insects, there'd be no flowering plants.
-
0:31 - 0:33Without flowering plants, there would be no
-
0:33 - 0:36clever, fruit-eating primates giving TED Talks.
-
0:36 - 0:38(Laughter)
-
0:38 - 0:40Now,
-
0:40 - 0:43David and Hidehiko and Ketaki
-
0:43 - 0:46gave a very compelling story about
-
0:46 - 0:49the similarities between fruit flies and humans,
-
0:49 - 0:51and there are many similarities,
-
0:51 - 0:54and so you might think that if humans are similar to fruit flies,
-
0:54 - 0:58the favorite behavior of a fruit fly might be this, for example --
-
0:58 - 1:00(Laughter)
-
1:00 - 1:03but in my talk, I don't want to emphasize on the similarities
-
1:03 - 1:06between humans and fruit flies, but rather the differences,
-
1:06 - 1:11and focus on the behaviors that I think fruit flies excel at doing.
-
1:11 - 1:14And so I want to show you a high-speed video sequence
-
1:14 - 1:18of a fly shot at 7,000 frames per second in infrared lighting,
-
1:18 - 1:22and to the right, off-screen, is an electronic looming predator
-
1:22 - 1:24that is going to go at the fly.
-
1:24 - 1:26The fly is going to sense this predator.
-
1:26 - 1:28It is going to extend its legs out.
-
1:28 - 1:30It's going to sashay away
-
1:30 - 1:32to live to fly another day.
-
1:32 - 1:35Now I have carefully cropped this sequence
-
1:35 - 1:38to be exactly the duration of a human eye blink,
-
1:38 - 1:41so in the time that it would take you to blink your eye,
-
1:41 - 1:44the fly has seen this looming predator,
-
1:44 - 1:50estimated its position, initiated a motor pattern to fly it away,
-
1:50 - 1:55beating its wings at 220 times a second as it does so.
-
1:55 - 1:57I think this is a fascinating behavior
-
1:57 - 2:00that shows how fast the fly's brain can process information.
-
2:00 - 2:03Now, flight -- what does it take to fly?
-
2:03 - 2:06Well, in order to fly, just as in a human aircraft,
-
2:06 - 2:09you need wings that can generate sufficient aerodynamic forces,
-
2:09 - 2:12you need an engine sufficient to generate the power required for flight,
-
2:12 - 2:14and you need a controller,
-
2:14 - 2:17and in the first human aircraft, the controller was basically
-
2:17 - 2:21the brain of Orville and Wilbur sitting in the cockpit.
-
2:21 - 2:24Now, how does this compare to a fly?
-
2:24 - 2:27Well, I spent a lot of my early career trying to figure out
-
2:27 - 2:31how insect wings generate enough force to keep the flies in the air.
-
2:31 - 2:33And you might have heard how engineers proved
-
2:33 - 2:36that bumblebees couldn't fly.
-
2:36 - 2:38Well, the problem was in thinking that the insect wings
-
2:38 - 2:41function in the way that aircraft wings work. But they don't.
-
2:41 - 2:44And we tackle this problem by building giant,
-
2:44 - 2:48dynamically scaled model robot insects
-
2:48 - 2:51that would flap in giant pools of mineral oil
-
2:51 - 2:53where we could study the aerodynamic forces.
-
2:53 - 2:55And it turns out that the insects flap their wings
-
2:55 - 2:58in a very clever way, at a very high angle of attack
-
2:58 - 3:01that creates a structure at the leading edge of the wing,
-
3:01 - 3:04a little tornado-like structure called a leading edge vortex,
-
3:04 - 3:07and it's that vortex that actually enables the wings
-
3:07 - 3:11to make enough force for the animal to stay in the air.
-
3:11 - 3:13But the thing that's actually most -- so, what's fascinating
-
3:13 - 3:16is not so much that the wing has some interesting morphology.
-
3:16 - 3:20What's clever is the way the fly flaps it,
-
3:20 - 3:23which of course ultimately is controlled by the nervous system,
-
3:23 - 3:26and this is what enables flies to perform
-
3:26 - 3:28these remarkable aerial maneuvers.
-
3:28 - 3:30Now, what about the engine?
-
3:30 - 3:33The engine of the fly is absolutely fascinating.
-
3:33 - 3:35They have two types of flight muscle:
-
3:35 - 3:38so-called power muscle, which is stretch-activated,
-
3:38 - 3:42which means that it activates itself and does not need to be controlled
-
3:42 - 3:45on a contraction-by-contraction basis by the nervous system.
-
3:45 - 3:49It's specialized to generate the enormous power required for flight,
-
3:49 - 3:52and it fills the middle portion of the fly,
-
3:52 - 3:53so when a fly hits your windshield,
-
3:53 - 3:55it's basically the power muscle that you're looking at.
-
3:55 - 3:58But attached to the base of the wing
-
3:58 - 4:00is a set of little, tiny control muscles
-
4:00 - 4:04that are not very powerful at all, but they're very fast,
-
4:04 - 4:07and they're able to reconfigure the hinge of the wing
-
4:07 - 4:09on a stroke-by-stroke basis,
-
4:09 - 4:12and this is what enables the fly to change its wing
-
4:12 - 4:15and generate the changes in aerodynamic forces
-
4:15 - 4:17which change its flight trajectory.
-
4:17 - 4:21And of course, the role of the nervous system is to control all this.
-
4:21 - 4:22So let's look at the controller.
-
4:22 - 4:25Now flies excel in the sorts of sensors
-
4:25 - 4:27that they carry to this problem.
-
4:27 - 4:31They have antennae that sense odors and detect wind detection.
-
4:31 - 4:33They have a sophisticated eye which is
-
4:33 - 4:35the fastest visual system on the planet.
-
4:35 - 4:38They have another set of eyes on the top of their head.
-
4:38 - 4:40We have no idea what they do.
-
4:40 - 4:43They have sensors on their wing.
-
4:43 - 4:46Their wing is covered with sensors, including sensors
-
4:46 - 4:48that sense deformation of the wing.
-
4:48 - 4:50They can even taste with their wings.
-
4:50 - 4:53One of the most sophisticated sensors a fly has
-
4:53 - 4:55is a structure called the halteres.
-
4:55 - 4:57The halteres are actually gyroscopes.
-
4:57 - 5:01These devices beat back and forth about 200 hertz during flight,
-
5:01 - 5:04and the animal can use them to sense its body rotation
-
5:04 - 5:08and initiate very, very fast corrective maneuvers.
-
5:08 - 5:10But all of this sensory information has to be processed
-
5:10 - 5:14by a brain, and yes, indeed, flies have a brain,
-
5:14 - 5:17a brain of about 100,000 neurons.
-
5:17 - 5:19Now several people at this conference
-
5:19 - 5:24have already suggested that fruit flies could serve neuroscience
-
5:24 - 5:27because they're a simple model of brain function.
-
5:27 - 5:29And the basic punchline of my talk is,
-
5:29 - 5:32I'd like to turn that over on its head.
-
5:32 - 5:35I don't think they're a simple model of anything.
-
5:35 - 5:37And I think that flies are a great model.
-
5:37 - 5:40They're a great model for flies.
-
5:40 - 5:42(Laughter)
-
5:42 - 5:45And let's explore this notion of simplicity.
-
5:45 - 5:48So I think, unfortunately, a lot of neuroscientists,
-
5:48 - 5:49we're all somewhat narcissistic.
-
5:49 - 5:53When we think of brain, we of course imagine our own brain.
-
5:53 - 5:55But remember that this kind of brain,
-
5:55 - 5:56which is much, much smaller
-
5:56 - 5:59— instead of 100 billion neurons, it has 100,000 neurons —
-
5:59 - 6:02but this is the most common form of brain on the planet
-
6:02 - 6:05and has been for 400 million years.
-
6:05 - 6:07And is it fair to say that it's simple?
-
6:07 - 6:09Well, it's simple in the sense that it has fewer neurons,
-
6:09 - 6:11but is that a fair metric?
-
6:11 - 6:13And I would propose it's not a fair metric.
-
6:13 - 6:16So let's sort of think about this. I think we have to compare --
-
6:16 - 6:18(Laughter) —
-
6:18 - 6:23we have to compare the size of the brain
-
6:23 - 6:25with what the brain can do.
-
6:25 - 6:28So I propose we have a Trump number,
-
6:28 - 6:31and the Trump number is the ratio of this man's
-
6:31 - 6:35behavioral repertoire to the number of neurons in his brain.
-
6:35 - 6:37We'll calculate the Trump number for the fruit fly.
-
6:37 - 6:40Now, how many people here think the Trump number
-
6:40 - 6:42is higher for the fruit fly?
-
6:42 - 6:45(Applause)
-
6:45 - 6:48It's a very smart, smart audience.
-
6:48 - 6:52Yes, the inequality goes in this direction, or I would posit it.
-
6:52 - 6:54Now I realize that it is a little bit absurd
-
6:54 - 6:58to compare the behavioral repertoire of a human to a fly.
-
6:58 - 7:02But let's take another animal just as an example. Here's a mouse.
-
7:02 - 7:06A mouse has about 1,000 times as many neurons as a fly.
-
7:06 - 7:08I used to study mice. When I studied mice,
-
7:08 - 7:11I used to talk really slowly.
-
7:11 - 7:13And then something happened when I started to work on flies.
-
7:13 - 7:16(Laughter)
-
7:16 - 7:19And I think if you compare the natural history of flies and mice,
-
7:19 - 7:23it's really comparable. They have to forage for food.
-
7:23 - 7:25They have to engage in courtship.
-
7:25 - 7:29They have sex. They hide from predators.
-
7:29 - 7:31They do a lot of the similar things.
-
7:31 - 7:32But I would argue that flies do more.
-
7:32 - 7:36So for example, I'm going to show you a sequence,
-
7:36 - 7:40and I have to say, some of my funding comes from the military,
-
7:40 - 7:42so I'm showing this classified sequence
-
7:42 - 7:46and you cannot discuss it outside of this room. Okay?
-
7:46 - 7:48So I want you to look at the payload
-
7:48 - 7:51at the tail of the fruit fly.
-
7:51 - 7:53Watch it very closely,
-
7:53 - 7:57and you'll see why my six-year-old son
-
7:57 - 8:02now wants to be a neuroscientist.
-
8:02 - 8:03Wait for it.
-
8:03 - 8:05Pshhew.
-
8:05 - 8:08So at least you'll admit that if fruit flies are not as clever as mice,
-
8:08 - 8:13they're at least as clever as pigeons. (Laughter)
-
8:13 - 8:17Now, I want to get across that it's not just a matter of numbers
-
8:17 - 8:19but also the challenge for a fly to compute
-
8:19 - 8:22everything its brain has to compute with such tiny neurons.
-
8:22 - 8:25So this is a beautiful image of a visual interneuron from a mouse
-
8:25 - 8:28that came from Jeff Lichtman's lab,
-
8:28 - 8:31and you can see the wonderful images of brains
-
8:31 - 8:34that he showed in his talk.
-
8:34 - 8:37But up in the corner, in the right corner, you'll see,
-
8:37 - 8:41at the same scale, a visual interneuron from a fly.
-
8:41 - 8:43And I'll expand this up.
-
8:43 - 8:45And it's a beautifully complex neuron.
-
8:45 - 8:48It's just very, very tiny, and there's lots of biophysical challenges
-
8:48 - 8:52with trying to compute information with tiny, tiny neurons.
-
8:52 - 8:56How small can neurons get? Well, look at this interesting insect.
-
8:56 - 8:58It looks sort of like a fly. It has wings, it has eyes,
-
8:58 - 9:01it has antennae, its legs, complicated life history,
-
9:01 - 9:04it's a parasite, it has to fly around and find caterpillars
-
9:04 - 9:05to parasatize,
-
9:05 - 9:09but not only is its brain the size of a salt grain,
-
9:09 - 9:11which is comparable for a fruit fly,
-
9:11 - 9:14it is the size of a salt grain.
-
9:14 - 9:18So here's some other organisms at the similar scale.
-
9:18 - 9:22This animal is the size of a paramecium and an amoeba,
-
9:22 - 9:26and it has a brain of 7,000 neurons that's so small --
-
9:26 - 9:28you know these things called cell bodies you've been hearing about,
-
9:28 - 9:30where the nucleus of the neuron is?
-
9:30 - 9:33This animal gets rid of them because they take up too much space.
-
9:33 - 9:36So this is a session on frontiers in neuroscience.
-
9:36 - 9:41I would posit that one frontier in neuroscience is to figure out how the brain of that thing works.
-
9:41 - 9:47But let's think about this. How can you make a small number of neurons do a lot?
-
9:47 - 9:49And I think, from an engineering perspective,
-
9:49 - 9:51you think of multiplexing.
-
9:51 - 9:54You can take a hardware and have that hardware
-
9:54 - 9:55do different things at different times,
-
9:55 - 9:58or have different parts of the hardware doing different things.
-
9:58 - 10:02And these are the two concepts I'd like to explore.
-
10:02 - 10:03And they're not concepts that I've come up with,
-
10:03 - 10:08but concepts that have been proposed by others in the past.
-
10:08 - 10:11And one idea comes from lessons from chewing crabs.
-
10:11 - 10:13And I don't mean chewing the crabs.
-
10:13 - 10:16I grew up in Baltimore, and I chew crabs very, very well.
-
10:16 - 10:19But I'm talking about the crabs actually doing the chewing.
-
10:19 - 10:21Crab chewing is actually really fascinating.
-
10:21 - 10:24Crabs have this complicated structure under their carapace
-
10:24 - 10:26called the gastric mill
-
10:26 - 10:28that grinds their food in a variety of different ways.
-
10:28 - 10:33And here's an endoscopic movie of this structure.
-
10:33 - 10:36The amazing thing about this is that it's controlled
-
10:36 - 10:39by a really tiny set of neurons, about two dozen neurons
-
10:39 - 10:44that can produce a vast variety of different motor patterns,
-
10:44 - 10:49and the reason it can do this is that this little tiny ganglion
-
10:49 - 10:53in the crab is actually inundated by many, many neuromodulators.
-
10:53 - 10:55You heard about neuromodulators earlier.
-
10:55 - 10:57There are more neuromodulators
-
10:57 - 11:03that alter, that innervate this structure than actually neurons in the structure,
-
11:03 - 11:07and they're able to generate a complicated set of patterns.
-
11:07 - 11:10And this is the work by Eve Marder and her many colleagues
-
11:10 - 11:13who've been studying this fascinating system
-
11:13 - 11:15that show how a smaller cluster of neurons
-
11:15 - 11:17can do many, many, many things
-
11:17 - 11:22because of neuromodulation that can take place on a moment-by-moment basis.
-
11:22 - 11:24So this is basically multiplexing in time.
-
11:24 - 11:27Imagine a network of neurons with one neuromodulator.
-
11:27 - 11:30You select one set of cells to perform one sort of behavior,
-
11:30 - 11:33another neuromodulator, another set of cells,
-
11:33 - 11:35a different pattern, and you can imagine
-
11:35 - 11:39you could extrapolate to a very, very complicated system.
-
11:39 - 11:41Is there any evidence that flies do this?
-
11:41 - 11:44Well, for many years in my laboratory and other laboratories around the world,
-
11:44 - 11:47we've been studying fly behaviors in little flight simulators.
-
11:47 - 11:48You can tether a fly to a little stick.
-
11:48 - 11:51You can measure the aerodynamic forces it's creating.
-
11:51 - 11:53You can let the fly play a little video game
-
11:53 - 11:57by letting it fly around in a visual display.
-
11:57 - 12:00So let me show you a little tiny sequence of this.
-
12:00 - 12:01Here's a fly
-
12:01 - 12:04and a large infrared view of the fly in the flight simulator,
-
12:04 - 12:06and this is a game the flies love to play.
-
12:06 - 12:09You allow them to steer towards the little stripe,
-
12:09 - 12:11and they'll just steer towards that stripe forever.
-
12:11 - 12:15It's part of their visual guidance system.
-
12:15 - 12:17But very, very recently, it's been possible
-
12:17 - 12:22to modify these sorts of behavioral arenas for physiologies.
-
12:22 - 12:25So this is the preparation that one of my former post-docs,
-
12:25 - 12:27Gaby Maimon, who's now at Rockefeller, developed,
-
12:27 - 12:29and it's basically a flight simulator
-
12:29 - 12:32but under conditions where you actually can stick an electrode
-
12:32 - 12:34in the brain of the fly and record
-
12:34 - 12:38from a genetically identified neuron in the fly's brain.
-
12:38 - 12:40And this is what one of these experiments looks like.
-
12:40 - 12:43It was a sequence taken from another post-doc in the lab,
-
12:43 - 12:44Bettina Schnell.
-
12:44 - 12:48The green trace at the bottom is the membrane potential
-
12:48 - 12:50of a neuron in the fly's brain,
-
12:50 - 12:53and you'll see the fly start to fly, and the fly is actually
-
12:53 - 12:56controlling the rotation of that visual pattern itself
-
12:56 - 12:58by its own wing motion,
-
12:58 - 13:00and you can see this visual interneuron
-
13:00 - 13:04respond to the pattern of wing motion as the fly flies.
-
13:04 - 13:06So for the first time we've actually been able to record
-
13:06 - 13:09from neurons in the fly's brain while the fly
-
13:09 - 13:13is performing sophisticated behaviors such as flight.
-
13:13 - 13:15And one of the lessons we've been learning
-
13:15 - 13:18is that the physiology of cells that we've been studying
-
13:18 - 13:20for many years in quiescent flies
-
13:20 - 13:23is not the same as the physiology of those cells
-
13:23 - 13:25when the flies actually engage in active behaviors
-
13:25 - 13:28like flying and walking and so forth.
-
13:28 - 13:31And why is the physiology different?
-
13:31 - 13:33Well it turns out it's these neuromodulators,
-
13:33 - 13:37just like the neuromodulators in that little tiny ganglion in the crabs.
-
13:37 - 13:39So here's a picture of the octopamine system.
-
13:39 - 13:41Octopamine is a neuromodulator
-
13:41 - 13:45that seems to play an important role in flight and other behaviors.
-
13:45 - 13:48But this is just one of many neuromodulators
-
13:48 - 13:49that's in the fly's brain.
-
13:49 - 13:52So I really think that, as we learn more,
-
13:52 - 13:54it's going to turn out that the whole fly brain
-
13:54 - 13:57is just like a large version of this stomatogastric ganglion,
-
13:57 - 14:02and that's one of the reasons why it can do so much with so few neurons.
-
14:02 - 14:04Now, another idea, another way of multiplexing
-
14:04 - 14:06is multiplexing in space,
-
14:06 - 14:08having different parts of a neuron
-
14:08 - 14:10do different things at the same time.
-
14:10 - 14:12So here's two sort of canonical neurons
-
14:12 - 14:14from a vertebrate and an invertebrate,
-
14:14 - 14:17a human pyramidal neuron from Ramon y Cajal,
-
14:17 - 14:21and another cell to the right, a non-spiking interneuron,
-
14:21 - 14:25and this is the work of Alan Watson and Malcolm Burrows many years ago,
-
14:25 - 14:29and Malcolm Burrows came up with a pretty interesting idea
-
14:29 - 14:31based on the fact that this neuron from a locust
-
14:31 - 14:33does not fire action potentials.
-
14:33 - 14:35It's a non-spiking cell.
-
14:35 - 14:38So a typical cell, like the neurons in our brain,
-
14:38 - 14:41has a region called the dendrites that receives input,
-
14:41 - 14:43and that input sums together
-
14:43 - 14:46and will produce action potentials
-
14:46 - 14:48that run down the axon and then activate
-
14:48 - 14:50all the output regions of the neuron.
-
14:50 - 14:53But non-spiking neurons are actually quite complicated
-
14:53 - 14:56because they can have input synapses and output synapses
-
14:56 - 15:00all interdigitated, and there's no single action potential
-
15:00 - 15:03that drives all the outputs at the same time.
-
15:03 - 15:07So there's a possibility that you have computational compartments
-
15:07 - 15:11that allow the different parts of the neuron
-
15:11 - 15:13to do different things at the same time.
-
15:13 - 15:18So these basic concepts of multitasking in time
-
15:18 - 15:20and multitasking in space,
-
15:20 - 15:23I think these are things that are true in our brains as well,
-
15:23 - 15:26but I think the insects are the true masters of this.
-
15:26 - 15:29So I hope you think of insects a little bit differently next time,
-
15:29 - 15:32and as I say up here, please think before you swat.
-
15:32 - 15:35(Applause)
- Title:
- How a fly flies
- Speaker:
- Michael Dickinson
- Description:
-
An insect's ability to fly is perhaps one of the greatest feats of evolution. Michael Dickinson looks at how a common housefly takes flight with such delicate wings, thanks to a clever flapping motion and flight muscles that are both powerful and nimble. But the secret ingredient: the incredible fly brain. (Filmed at TEDxCaltech.)
- Video Language:
- English
- Team:
- closed TED
- Project:
- TEDTalks
- Duration:
- 15:55
Thu-Huong Ha edited English subtitles for How a fly flies | ||
Thu-Huong Ha approved English subtitles for How a fly flies | ||
Thu-Huong Ha edited English subtitles for How a fly flies | ||
Morton Bast accepted English subtitles for How a fly flies | ||
Morton Bast edited English subtitles for How a fly flies | ||
Joseph Geni added a translation |