Making Memories With Fruit Flies
IRA FLATOW, host:
This is SCIENCE FRIDAY from NPR News. I'm Ira Flatow.
Ms. SARAH PALIN (Former Republican Governor, Alaska): Sometimes, these dollars, they go to projects having little or nothing to do with the public good. Think, like, fruit fly research in Paris, France. I kid you not.
FLATOW: You might remember that from last year's election, Sarah Palin mocking fruit fly research. Well, fruit flies, as you may know, are widely used in science, and neuroscience specifically, to study everything from brain tumors to addiction. And that's because, compared to people, fruit flies have a much simpler brain. It's easy to map out and understand how the parts work together.
Tweaking just a few neurons can cause a fly to alter its behavior in revealing ways, and scientists are reporting in the journal Cell this week that they did just that. By manipulating a handful of fruit fly brain cells, they were able to create false memories in the flies. Why would they want to do that?
Here to talk about it is Gero Miesenbock. He's professor of physiology at the University of Oxford in England. Thanks for talking with us today.
Professor GERO MIESENBOCK (Physiology, Oxford University): It's a pleasure.
FLATOW: Why did you do this to fruit flies?
Prof. MIESENBOCK: We want to understand sort of the fundamental problem that every animal that moves about faces. Life for a fly and for you and for me is a string of choices, and to make these choices properly, we have to evaluate our options and choose the best course of actions, and we get it wrong, learn from our mistakes.
This is a very, very important function of the brain. Some argue it's the most important thing that the brain does, but very little is known about how the brain does precisely that.
FLATOW: You mean learning from our mistakes.
Prof. MIESENBOCK: Exactly.
FLATOW: And so what did you do with the fruit flies to learn about the mistakes they would make?
Prof. MIESENBOCK: Well, we tried to identify the cells that tell the fly's brain that the animal has done something wrong and then store that information in memory so that it can be used when a similar situation recurs in the future and hopefully help avoid committing the same mistake again.
FLATOW: So you created, basically, a false memory in the fruit fly's brain, that it had done something wrong.
Prof. MIESENBOCK: Exactly.
FLATOW: And how did you do that?
Prof. MIESENBOCK: We developed several years ago a method that allows us to remotely control groups of neurons that are genetically defined in the brains of animals with light. And in this case, we have used this method to search for the neurons that would send so-called error signals, signals that tell the fly that something has gone wrong.
The logic of the experiment was that if we managed to send the appropriate error signal while the fly was doing something, in this case smelling an odor, the flies would avoid that odor if we had hit the right neurons and turned them on.
FLATOW: So how did you hit the right neurons? How were you able to modify the right neurons?
Prof. MIESENBOCK: We relied on genetics to target different sets of neurons that produce the transmitter dopamine, which is probably familiar to many of the listeners as the key - a key ingredient in the development of Parkinson's Disease and also a key reward-signaling mechanism in humans.
In flies, it also has something to do with reward and punishment, but its polarity is flipped. So in contrast to vertebrates, dopamine is a punishing signal in flies. And so we narrowed our search to the dopamine-producing neurons in the fly's brain - there's about 300 of them, 300 out of a total of 100,000 brain cells - and made different groups of these 300 neurons light-sensitive. And we observed that remote-controlling some groups led to learned behavior, was able to instruct memories, in other words, whereas remote-controlling different groups was not able to instruct memories. And then by elimination, we narrowed the responsible set of cells down to just a dozen.
FLATOW: So you found a dozen cells. You genetically engineered these cells so that they would respond to laser light.
Prof. MIESENBOCK: Yes.
FLATOW: And then you shone the light on the fruit flies.
Prof. MIESENBOCK: Yes.
FLATOW: And so they released the dopamine, thinking that they had done something bad.
Prof. MIESENBOCK: Exactly.
FLATOW: And so they avoided - by shining the light on it, you created this memory, so they avoided something that wasn't even there. They thought they avoided it.
Prof. MIESENBOCK: No, they avoided the odor. So the odor was the sensory cue that was paired with the error signal. So when the odor was presented again, next time without the light, the fly indicated through its avoidance that it had learned that negative association.
FLATOW: Wow. And you just had to use 12 cells to do this?
Prof. MIESENBOCK: Yes. It's likely that the actual number of cells that's responsible for this instructive effect is a subset of those 12 cells. There was a study published in the same issue of Cell from a different group, a professor called Waddell at the University of Massachusetts, that studied the motivational use of memory in flies. And their study also zeroed in on just the same 12 cells. So it seems that there is functional subdivision within that cluster of 12 neurons.
FLATOW: Could you find other neurons to make the fruit flies do other things by just shining the light on them?
Prof. MIESENBOCK: Well, obviously, all neurons do different kinds of things in the brain. So you can - by rendering different types of neurons light-sensitive, you can elicit different forms for behavior. For example, last year, my lab published a study in which we had rendered cells that express the principal sex-determination factor in the brain light-sensitive. There's about 2,000 of these neurons, and when we turned these neurons on in male flies, they started to exhibit the typical male courtship behavior. They stuck out one wing and started to sing a song that normally female flies find irresistible.
Now the big surprise came when we did the same manipulation in female flies. Females obviously never go through male-specific courtship. But lo and behold, we were able to turn that circuit on artificially and make female flies behave as if they were males. So there's a latent capacity for male-like behavior wired into the female brains.
FLATOW: Did you shine this in their eyes? Where did you have to shine the laser beam?
Prof. MIESENBOCK: You shine it onto the entire animal. So the light is intense enough to penetrate the cuticle, the chitinous exoskeleton of the fly, and to activate whatever brain neurons are genetically sensitized to respond to the beam of light.
FLATOW: Can you do this in higher forms of animals?
Prof. MIESENBOCK: Yes, this entire field, which sort of originated with our first study in 2005 and the methodological papers in the years before, has really exploded in the past several years. It now even has a name. It's called optogenetics, and there's many, many different kinds of laboratory animals that are being studied using optogenetic techniques, from nematodes to flies to fish to mice and even to primates.
FLATOW: So it's like a having a remote control for your TV. When you want the animal to do something, you just shine the light on them.
Prof. MIESENBOCK: That's essentially, yes, what it is. It's an ability to control the function of targeted brain circuits with exquisite precision without destroying the normal function of the brain.
FLATOW: And how high, you know - you know what I'm fishing for here, is the human connection.
Prof. MIESENBOCK: I think the human connection is quite far off. There's so many ethical, regulatory and technical hurdles that have to be overcome. Don't forget that one of the key elements in what we do is a genetic modification. It's genetic engineering.
So in the human case, that would mean gene therapy, that you would have to be infected with a virus that then delivers a light-sensitive protein to some nerve cells in the brain. And as we all know, there's significant regulatory barriers at the moment to gene therapy in humans.
FLATOW: But you could do it in lower animals.
Prof. MIESENBOCK: Yes, you can. And it has been done.
FLATOW: Yeah, and you could do it in, I don't know, just pick an animal - a dog, a cat, something like that?
Prof. MIESENBOCK: Yes, I think that there's no barrier to - there's no technical barrier to doing that in any kind of animal.
FLATOW: And what would you like to study most in these animals, to do this?
Prof. MIESENBOCK: For me, I mean, I think one of the hallmarks of intelligence is the ability to adapt to a changing environment and not to make the same - to keep making the same mistake many, many times.
Prof. MIESENBOCK: The - how such intelligent, adaptive behavior emerges from physical interactions of components, nerve cells that are by themselves unintelligent, I think is a big mystery, and that's a problem that I'd like to study.
FLATOW: Could you work on addiction or overcoming addiction if you could find out which of the cells are involved?
Prof. MIESENBOCK: Of course. I mean, what makes addictive substances so powerful is that they play just with our endogenous mechanisms of reward and punishment, right? So by manipulating these mechanisms and finding out more about how they work and how they are normally controlled, that, of course, has major implications for many more applied fields of science.
FLATOW: So you say this is a whole new field that's opening now.
Prof. MIESENBOCK: I do think so, yes.
FLATOW: Optogenetics. We should keep an ear, an eye out for that?
Prof. MIESENBOCK: Yeah.
FLATOW: Yeah. And well, we - so now that you've done this fruit fly experiment, and you're, like, known as the king of the fruit flies, where do you go from here?
Prof. MIESENBOCK: We would really like to understand how those 12 cells are controlled, right? I mean, what we have found is essentially the messenger. We have found the channel through which something in the fly's brain tells some other part in the fly's brain that the contents of memory should be updated, that some new information should be written into memory. But what we do not know yet is the author of the message that's being sent by those 12 cells. That's our next challenge, to try and understand what the signals are that lie upstream of these 12 dopamine-producing neurons and control their activity.
FLATOW: Can you actually trace it back upstream?
Prof. MIESENBOCK: We're working on it. So I - optimistically, I would say yes. I hope we are able to trace those signals back upstream. And what that should then reveal is a fundamental learning algorithm. There's a huge literature in diverse fields such as psychology, engineering, machine learning, that deals with so-called neuromorphic circuits, so circuits constructed from neuron-like components that calculate such error signals. But all of these circuits are, at the moment, just speculation. There's no biological example that has been solved, and that's quite ironic, considering that biology is the inspiration for all this sort of artificial adaptive intelligence.
FLATOW: Well, we wish you good luck.
Prof. MIESENBOCK: Thank you.
FLATOW: Dr. Miesenbock, thank you for taking time to be with us.
Prof. MIESENBOCK: My pleasure.
FLATOW: Gero Miesenbock is a professor of physiology at the University of Oxford in England. We're going to take a short break and switch gears and come back and talk a little lunacy here, talk about the moon and all these - you've been hearing a lot of stuff that's been going on with space probes and whatever on the moon. We'll wrap it all up. Stay with us. We'll be right back.
(Soundbite of music)
FLATOW: I'm Ira Flatow. This is SCIENCE FRIDAY from NPR News.
NPR transcripts are created on a rush deadline by a contractor for NPR, and accuracy and availability may vary. This text may not be in its final form and may be updated or revised in the future. Please be aware that the authoritative record of NPR's programming is the audio.