Skip to content
Who's in the Video
Joseph LeDoux is a professor and a member of the Center for Neural Science and Department of Psychology at NYU. His work is focused on the brain mechanisms of emotion[…]

A conversation with the neuroscientist.


Question: What do we know about the amygdala?

Joseph LeDoux: Well the amygdala is a small region of the brain.  It’s in what’s called the temporal lobe of the brain.  If we divide the cerebral hemispheres of the four major lobes, occipital, temporal, parietal, frontal, it’s in the temporal lobe which is under the temporal bone, which is behind your ear.  

And the amygdala would be located at a point where a line that goes through your eye and through your ear intersect in your brain.  So imagine like someone has shot an arrow through your eye and another one through your ear, where those two come together would be roughly where it is.  And on both sides of the brain, of course, because most structures, in fact, all structures are bilateral in that sense.  

So its sometimes referred to as an almond or an almond-shaped structure, which is a bit of a misnomer because there’s one part of it that’s almond shaped and that caught the attention of early anatomists, and so they named this almond shaped structure the amygdala, but as the amygdala began to be defined more broadly, in includes a lot of structures other than that little almond-shaped piece.  

It’s most commonly associated with fear, although that’s not its only function, it’s its most heavily studied function.  And the reason for that is that it’s much easier to study fear than other emotions.  Things that are bad have more weight than things that are good in a sense.  You can put off eating, drinking, sex, whatever for an indefinite amount of time, but you have to respond to danger immediately.  You know, you don’t find great novels about happy subjects, it’s always about the bad stuff and it’s sometimes said that there’s this negative bias in emotion, which his that most research on emotion is about the negative stuff.  But the fact is that negative stuff turns out to be the most important for survival, in a sense, in terms of immediacy in having to respond to that.
 
Question:
To what other structures is the amygdala connected?

Joseph LeDoux:
So some of the key interconnections of the amygdala—and these connections actually define what it does in a sense, at least with respect to fear—the amygdala gets sensory information directly from the various sensory systems that process the external world.  So the visual system, the auditory system, olfactory, touch, pain, and so forth.  All of these kind of come together, or converge, in the amygdala.  And then the amygdala on the output side with all the systems involved in the emotional reactivity.  So, when you encounter sudden danger, you might freeze, your blood pressure and heart rate begin to rise, stress hormones are released, all of these things happen as a result of outputs of the amygdala.  

So the amygdala you can think of as this circle with one input coming in being the... or the input coming in being a sensory, flow of sensory information from the external world, then outputs being connections to systems involved in controlling  the responses.  But we have to expand those inputs, so it’s not just getting one sensory input, but all sensory inputs.  So each sensory system is coming in.  And it doesn’t stop there because in addition to getting information from sensory systems, it also gets information from higher-order systems, like the prefrontal cortex and higher-order association areas involved in various kinds of integrative activities in the cortex.  

Question:
What role does the amygdala play in learning adaptive behaviors?

Joseph LeDoux: What it does is a function of what connections it has.  So because it has connections with all of these sensory systems, it can take in information from the environment of a variety of types, and use that information.  And so if, you know, if a sound in the external world occurs right before something painful happens, you associate that sound with the painful event and then that sound will then later trigger a protective defense response.  But if the sound occurs just before food, when you’re hungry, then the sound will be associated with that kind of a positive or a repetitive event.  

And so, what the amygdala is doing is forming associations between random or neutral external stimuli and the kinds of reinforcing events that will stamp in those experiences in a stronger way.  So its creating these what’s called Pavlovian associations; you know, stimulus one plus stimulus two, if one of those is a biologically significant stimulus, then the other one will require some kind of biological significance itself whether it’s positive or negative.  

Now that information can then be used and this is sort of what I think you referred to in terms of adaptive behavior.  So that CS/US relationship or say, sound/food relationship can then be used to guide instrumental behavior, which is goal-directed behavior. So if in the past you obtained food at a certain location, the stimuli that are proximal to that location serve as conditions stimuli and you know, are reinforcing to you because they’re close to the actual goal and they take you towards the goal.  The same thing happens with reversive stimuli, it’s just that it works in the opposite way.

Question:
Can we think of the amygdala as the seat of what Freud called the unconscious?

Joseph LeDoux: I think it’s a distraction because... I mean, it’s true in the sense that the hippocampus is necessary to have a conscious recollection of some past event, and the amygdala participates in unconscious memory. But we shouldn’t really taint it with the Freudian concepts because that adds a lot of baggage.  

The amygdala is an unconscious processor because it’s just not connected with the conscious system.  It’s kind of like by default unconscious as opposed to being in the Freudian sense of unconscious something that was conscious, but was too anxiety-provoking and therefore shipped to the unconscious.  The amygdala gets direct sensory information and it learns and stores information on its own, and that information that’s stored then controls emotional responses.  The connectivity is hardwired, so one way to think about it is that a rat will respond to a cat without any learning by freezing, raising its blood pressure and heart rate and respiration and releasing stress hormones.  But it will also respond to a stimulus associated with a cat and have the same responses.  

So you don’t learn how to be afraid, your amygdala doesn’t have to learn what to do, it learns what to do it in response to.  So it learns what stimuli it should respond to.  So it’s learning and memory in that sense that we call an implicit kind of memory where you don’t have to have any conscious involvement.  

Whereas, the hippocampus is necessary to have a conscious memory.  So, yes, they do participate in conscious and unconscious memories, but not in the Freudian sense.  But there’s a whole other side of the amygdala’s role in memory, which is that when the amygdala is activated and all of those hormones and other things happened to get released, that provides information that feeds back to parts of the brain, like the hippocampus and allows them to store their memories in a much more efficient and strong way.  So we know that emotional memories are stored more vividly than other kinds of memories.  It used to be thought that they were more accurate, but in fact now we know that they are not more accurate, they’re just more vivid and strong in the personal sense.  But they can be highly inaccurate.  This is shown by studies of natural disasters and so forth, well not always so natural.  But like the Space Challenger Shuttle... or the shuttle, Space Shuttle Challenger explosion, a lot of people witnessed that and they were studied almost immediately by psychologists who made notes of exactly what their responses, what they were experiencing at the time and then a year later, they were surveyed again and the responses were completely different from what they remembered originally and then several years later it completely changed again.  

So what we remember is not necessarily what we experienced originally.  So the accuracy of those memories changes over time, but their strength in terms of your subjective feeling that it was a really powerful experience is there.  

Question:
Why study animal brains to learn more about our own brains?

Joseph LeDoux: I think that brain imaging is not a very good way to test subtle distinctions in a sense because, especially... well, if you look at what the... what you can learn from a brain image, it’s like trying to find out something about New York City by studying New York State.  In other words, it’s a very big representation and there’s a lot going on within those big blobs that you see on a brain image.  

And that’s why animal research is so important because we can go in and study individual cells and individual synapses on those cells, which is like, again, that’s a tiny part of New York City if you’re looking at the entire map of New York State.  And so you need to be able to get to that point of resolution  

Question:
Can we really apply what we learn from animal brains to our own?

Joseph LeDoux: Throughout the history of psychology there’s been a struggle between what can we learn about psychological states in humans and animals, what can we learn about psychological states by studying conscious states in people and unconscious states, and so forth.  There is always a struggle between animal/human, conscious/unconscious.  And the behaviorists got rid of that because they said we can’t study consciousness because it’s not a objective thing that you can measure.  So they said that what we should do is to just study observable behavior.  So they got rid of consciousness.  Then the cognitive revolution came along and brought the mind back to psychology, but it didn’t bring back the mind that the behaviorists got rid of, it brought back the mind as an information processing device.  

So when I got interested in emotion in the '70s, emotion was still being thought of in terms of subjective conscious experiences, whereas other aspects of psychology, like perception and memory, were thought of as information processing functions.  So you could learn a lot about the way the brain processes the redness of a sunset without actually knowing how it experiences the redness of the sunset.  

But in the study of emotion, that kind of leap had not been made, and so people were still thinking of emotion as the conscious feeling of, say, of being afraid, as opposed to the information processing that goes on when you detect danger.  So all animals have to be able to detect danger and respond to danger in order to stay alive, including humans.  And so my approach was to say, "Well, let’s forget about consciousness, it kind of gets in the way of studying what we want to study, which is how the brain processes information about emotional situations."  So I decided to study how the brain detects danger and response to danger independent of any conscious experience that the rat has or doesn’t have.  

So I remain completely agnostic about whether the rat experiences fear and instead focus totally on how rats and humans detect and respond to the danger.  Learn about new dangers and use that information in adaptive ways.  So that leaves a big hole in terms of what we can learn in rats, but it’s been an interesting time in research.  We’ve learned a lot about how the brain detects and responds to danger.  We’ve put together a pretty comprehensive understanding of that whole system.  

So we haven’t explained what the humanists want to know about when they talk about emotions, but we’ve learned a lot about what goes on in an emotional situation in terms of how the brain processes information.  

Question:
What are you currently researching?

Joseph LeDoux:
I think one of the more interesting things is our focus now on individual differences, you know, if you condition 10 rats or 20 rats to be afraid of the sound paired with a shock you find that some are very afraid and some are not very afraid and the other are kind of in the middle.  So the typical way of dealing with that is you average it altogether and you get the mean, and that’s what you study.  

The outliers are viewed as just kind of a nuisance which adds variance to the data, but now we’ve begun to study those previous nuisances to try to understand a little more about what’s really going on in terms of pathological fear, because almost all of the drugs that are developed to treat fear and anxiety are developed on that average animal, rather than the extremes. But what we really need to understand, I think, and the drugs to be much more effective and perhaps have fewer side effects if they were targeted for the animals with extreme fear.  

So we’re trying to come to the question of what causes animals to have this extreme fear.  What pushes them out to the ends of the distribution?  The basic idea is that, you know, one way to do this is to take animals that have the extreme fear and to start breeding them and create genetic lines that are fearful, but I think it’s also interesting to ask, given that it already exists in the population of rats, these extreme behaviors, what can we learn about, say the pathophysiology of extreme fear by studying those animals.  In other words, we don’t have to start breeding and creating genetic lines to get at what’s different because the difference is already there.

We can compare animals that are really afraid and those that are not afraid and look in their brains and see if there area any, for example, structural differences in the amygdala in terms of how the neurons, what their dendritic branches, what their axons are like.  What kinds of molecules are present in those neurons?  And to what extent?  So we can get a lot of information that might distinguish fearful and not so fearful rats that could provide important clues as to what pushes them out there towards the extremes.  

But that project is just beginning so we don’t have any answers, but I think it’s going to be an important project.
 
Question:
What are the biggest recent breakthroughs to have come out of your lab?
 

Joseph LeDoux: I can think of three things that we’ve done that have kind of been, I think – and I don’t say it to credit myself because a lot of the credit goes to the students and the post-docs that did this research.  In many cases they came up with these ideas. 

But one was the rediscovery of something called “reconsolidation,” which is when you retrieve a memory it becomes unstable and new information can be incorporated into the memory at that point.  It also means that when the memory is unstable, its restabilization process can be blocked.  And if you block that, the memory is weakened or dampened.  So that’s how we’ve been using reconsolidation to help people with traumatic memories and try and dampen their memories of those traumatic situations.  Not necessarily the conscious, cognitive memories, but the emotional component of the memory.  

So that’s one area that I think we’ve had a big impact in and Karim Nader really gets credit for triggering all of that research in my lab.  

Another area is a takeoff on that, which is a recent study that we published by Marie [...] is the lead author, showing that we could take what’s usually called exposure therapy, where—it’s also called extinction, where you give—you condition a rat to be afraid of something, then you give the tone over and over again and it stops being afraid.  And this is also used in the treatment of phobias and other kinds of anxiety problems by weakening the fear associations that trigger the emotion.  

But the problem is that the fear always bounces back at some point, say by stress.  So the patient has a fear of heights and then his or her mother dies and the fear of heights comes back even though it was successfully treated.  But we found a way in rats to prevent that bounceback of the fear.  We can more or less permanently dampen it.  And that has to do with a special procedure in which the extinction process is timed in a certain way.  It’s too complicated to explain in detail, but by simply timing certain presentations of the stimuli, extinction can be made to be more permanent.  

And a third area, which is kind of a technical advance.  Josh Johansen has recently published a study in my lab using optogenetics, which is a new way of altering brain activity by... what you do is you connect up a molecule that you’re interested in with a virus so it can be injected into the brain and alter genes in that brain area.  And what we’ve been injecting are molecules that are light-sensitive.  So you inject this virus into the brain and it puts a molecule in that part of the brain, but then when you later implant a small fiberoptic lamp in there, it’s very tiny, it’s just microscopic, but it goes right into the amygdala.  And when you flash light on the amygdala, it causes the neurons in the amygdala to be responsive to light and to depolarize.  So they have action potentials when the light shines on them.  

And what this allowed us to do was to test the basic idea that learning in the brain, or learning in this case in the amygdala, but in the brain in general, involves the depolarization of neurons, in other words the firing of action potentials, while the neuron is getting a meaningless stimulus.  So what we did was we presented the rat with the tone while we directly depolarized the lateral amygdala cells with the light.  And this created learning in the form of behavior when the rats later heard the tone.  

So this is a very important technical advance showing that a very basic principle of synaptic plasticity called Hebbian Learning after Donald Hebb, who said that learning involves the arrival of the weak synaptic input at the time when a strong synaptic input is activating the same cell.  So in our case, the weak synaptic input is the tone, and the strong synaptic input is the direct depolarization of the cells.  

This depolarization obviously substitutes for the electric shock that we would normally give in a conditioning experiment.  

So those are, I think, three things that we’ve done that have been important.  I don’t know if that – they’re the most important things in neuroscience...

Recorded on September 16, 2010
Interviewed by Max Miller


Related