captioning sponsored by rose communications from our studios in new york city, this is charlie rose. >> rose: tonight we continue our journey to the most exciting frontier of science, the brain, last month, we examined how the brain interprets information from the five senses, focusing on the visual system. tonight we'll show you how the brain uses that information to interact with the outside world. our subject is the acting brain. taken together, the parts of the central nervous system devoteed to movement are known as the motor system. the motor system allows to us plan, coordinate every action needed to survive in the physical world. all of the movements we make from the breathing of our heart to the hitting of a tennis ball are controlled by the brain and nervous system. as the legendary biologist charles sharng on the once said "to move things is all that mankind can do and for this task the soul excue tonight is a muscle whether it be a whispering of the syllable or felling of a forest. how does the brain do it? how does the brain translate subjective intentions into physical actions? what happens in the brain when we learn a new skill? why are some of us graceful and others clumsy sfwhoo? why does practice make perfect? much like the visual system, the motor system is astounding in its complexity, it controls over 650 muscles, giving rise to an immense repertoire of movement and actions. coordinating these muscles is a tremendous charge with the motor system carried out mostly without conscious instruction, reflexes, for instance, allow us to respond immediately and unconsciously to changes in our environment. other essential functions such as breathing are also performed automatically and unconsciously. but most of our physical abilities must be learned through practice during infancy and childhood we learn to crawl, walk and use language. by the time we reach adulthood these difficult tasks have become effortless. in, most of what the motor system does is taken for granted until it is interrupted by injury or disease. tonight we'll examine two devastating motor illnesses, stroke and a.l.s., also known as lou gehrig's disease. both of these diseases are tragically common stroke is the leading cause of death in the united states. but for those who survive, the brain'splastyty shows remarkable recovery. despite decades of research, a.l.s. is 1200% fatal. joining me this evening is a remarkable group of scientists who have devoted their careers and their lives to understanding hui the brain controls movement. they are thomas jess is, he studies circuits that forms the basis for the entire motor system. he is a howard hughes medical investigator. daniel wolpert, his research uses mechanical models that mimic human behavior. he heads the wolpert lab at cambridge university in england. john krakauer, he is interested in how the brains learns new skills and how it can regain function even after stroke or injury. he is an associate professor of neurology and neuroscience at columbia university. robert brown, he is a geneticist physician and all around expert on lou gehrig's disease. he teaches and practices at the university of massachusetts. and once again my co-host is dr. eric kandel, he is, as you know, a nobel laureate, a professor at columbia university and also an investigator at the howard hughes medical research center. i'm pleased to have one more time a chance to talk about this extraordinary thing, this brain. we've gone from the... we've gone from the general, to visual now to movement. what are the themes we'll look at tonight? >> well, charlie, you outlined them so beautifully in your introduction. every behavior is mediated through the motor systems, from the simplest to the most complex. all sensory perception, visual perception reaches its completion through the actions of motor systems. in fact, we can think of the motor systems in some ways as being the mirror image of the sensory system. the sensory systems create a ski ma, an internal representation in our brain of the outside world. the motor system uses that internal representation in actions. and like the sensory systems, the motor system is localized to particular regions and it has three important components, a high arkky, if you will, to decide, to make an action, to pick up a glass of water, to mobilize the muscles to actually make that movement and then report back that that movement has been carried out successfully. motor actions vary tremendously from the simplest kind of action to marvelous pirouettes that ballet dancers can do or high jumping pole vaulting extremely skilled performances. some of these are inborn and as you indicate misdemeanor are learned. and the flexibility, the plasticity of the motor system is extraordinary. throughout our lives we can continue to modify our behavior. we play tennis next to one another. i see your tennis game getting better month to month. even at my age. (laughs) >> rose: so it's not getting better at my age, is it? >> even at my age it has a chance of getting better and we only see, as you indicated, that this automatic behavior we take for granted because we do so much of it automatically, we only see that it is important to us when something goes wrong. we can see this, for example, in the sad case of lou gehrig. can we have the video on lou gehrig, please? spectacular baseball player who was hitting about .350 for most of his career. an extraordinary career running over 12 years. never missed a game. game after game throughout the season. all of a sudden he began to find a weakening in his stroke. he still saw the ball as clearly as he ever did, but he couldn't get his bat around him powerfully. and you can see the dramatic decline in his batting average going from .350 to .150 in the period of a year and a half. and ultimately, for the first time in over a thousand games, he had to be benched. that was because he developed a.l.s., a disease we're going to consider today. >> rose: just make this point about science. science builds on what you have learned before. here is sir charles sherrington, this book "the integrative action of the nervous system, published in 1960. this is also based on some lectures he did at yale. it's a remarkable story of how science works. jup >> you're absolutely right. there are occasional people-- we call them giants-- who not only make extraordinary contributions sherrington worked out the simplest reflex pathways in the final cord, stimulated the motor cortex, saw how it moved particular limbs but alsoed that insight to see how the whole brain works. the sbi graveive action of the nervous system means sensory information is come in, it's processed to give rise to a variety of movements. this not only applies to motor systems, it applies to every aspect of nervous system function. there is xi station of appropriate movements, inhibition of inappropriate movements. this is a brilliant peres yent set of insight wes owe to sherrington. we're still living in a sherringtonian world today. >> rose: we talk about and touch on all these things we have been talking about. eric and i have assembled a group of people i think you will find extraordinary in terms of their insight and how they can demonstrate what we are talking about. because we're talking about motor functions, what you can see. you will also discover as you listen to them that there is a remarkable thing that is coming out of this, insight into how the brain functions in its highest level. so we begin with understanding the significance of the motor system. >> well, i think we have to ask a very fundamental question, perhaps the most fundamental question we can ever ask and i think it's remiss that eric hasn't asked it. why do animals have brains? it's a fundamental question because there are many species who don't have brains. that so that's a fundamental question we should be taught on our first day of school. and if you think about it, it's obvious. we have one for one reason and one reason only, and that's to produce adaptable and complex movement. there's no other reason to have evolved a brain. so the only way we can affect the outside world is through contractions of muscles. so think about communications, speech, gestures, sign language, they're all mediated by contracting muscles. so we need to remember that things like sensory processing, perceptual system, memory and cognitive processs are all important but they can only be important to drive action or suppress future actions. there's no point laying down memories of childhood or perceiving the color of a rose if it doesn't lead you to do something different with your motor system later in life. so if you think from an evolutionary point of view, there's no point having the thinking processs if they can't be expressed through action. so i'm a really movement chauvinist. i though understand the brain we have to understand movement, which is the final output. >> rose: to understand the brain we have to understand movement. >> we can't look at memory or perception in isolation from action. and we can say if you don't believe in this argue. there's many species who live very happy lives in our planet, do very well socially but they don't need to move. so the tree is a nice example. it doesn't require complex movement, it hasn't evolved a brain. but the clinching for those who don't believe this is this animal here. this is the humble sea squirt. it's a rudimentary animal, it has a brain, a spinal cord and it swims in its juvenile life. and at some point it implants itself on a rock and never leaves the rock again. and the first thing in implanting on that rock is to digest its own brain and nervous system for food. once it doesn't need to move, it doesn't need that brain anymore. so i think it's really the brain is there for movement. >> rose: and is this the reason we've never had a robot that can be as graceful as a six-year-old walking down a road? >> that's a very good question. so we can ask how well are we going understanding how the brain controls movement and the answer is i think we've done pretty well but we've got a long way to go. and you can begin by asking how well can you build machine which is can do the things humans do. how well can we build machines that decide what chess piece to move where. if we pit gary kasparov against i.b.m. blue, occasionally blue will win. and if it played any one here it will win every time. (laughter) but we need to think about now build magazine which has got the manual dexterity of a five-year-old. pit a five-year-old's dexterity to manipulate a chess piece against the best robot, there's no competition. the five-year-old wins easily. and so we can say why is that the first problem of deciding what to move so easy and the control problem so hard. one reason is the five-year-old child will tell you. you need to solve what piece can move where. look at all possible movements a president end of the game and choose what makes you win. the algorithm is very simple. the problem is moves. but with fast computers and approximations we get close to the answer. when it comes to being dexterous it's very unyear problems you have to solve to be dexterous and there are real problems in sensing the world and acting upon chit have a lot of complexity. but i'd love to show you a video. >> rose: please do. >> i'd like to show you a video of what's cutting edge in robotics and cutting edge in human performance just to give you a feel of how close we're getting. so if we could roll the video of the robot. what this video shows is the end of a three-year project by my colleagues in germany teaching a robot to pick up a glass and pick up a jug of water and pour water into a glass. as you can see, it does the task but it's not doing it anywhere as fluidly or speedily as a human would do it. you would regard this as poor performance. that is very challenging task. and if if you want it to do something different, there's another three year project. there's no generalization from one task to another. so this is a fundamentally very difficult problem. let's care in now to what we regard, perhaps, as the cutting edge human performance. so you see a small child, a nine-year-old, winning the world record for cup stacking. now, cup stacking is a popular support. it involves taking cups and stacking them and unstacking them in a particular sequence as fast as you can. >> rose: this is great! (laughs) >> we've got a very long way to go before we get anywhere near to building these machines or understanding how that child learns to do that task. >> rose: let's look further at the anatomy. talk about how about how it works. >> well, the mysteries that daniel has just outlined for you have been puzzling neuroscientists for over 100 years or so and charles sherrington was really the first of the modern generations of neural scientists who tried to understand the mechanisms that really linked the way into the way out. so in the brain, in the nervous system, there are many ways in. the individual senses that we've talked about and previous discussions. sense of vision, a sense of touch, the sense of smell. but sherrington and people of his time realized that despite all of the many ways in, there's only way out, and that's the motor system. and sherrington was the first to articulate this idea that the motor system is the final common pathway of all of the sensory world that impinges on the brain and the way that that information is processed to produce coordinated movement. and so in the 19th century, there had been a series of experiments trying to understand and deconstruct what movement really is. even the simplist of movements to move one's hand or to move one's thumb rishz several components or several processing events. the first is actually to plan the movement. one has to be able to control the nature of the movement before there's any sign of muscle contraction. the second is the execution of movement and the thirdd is to achieve some reporting of the consequences of that movement. and sherrington and his colleagues really outlined the idea that these different functions, planning, execution, and reporting, are assigned to different regions of the brain. and the first experiments that changed the way we think about the field were some electrical stimulation experiments. so with the crude tools available, people start to stimulate different regions of the surface of the brain and look for the consequences in terms of movement. so some areas where you're stimulated gave no overt signs of mooucht. but there are other areas, hot spots, if you like, which we now know is the motor cortex and the premotor cortex and they're shown here in this colored region where very low intensity stimulation would illicit movement. and more than that, that was a precise register or map between the place in which the stimulating probe was located and the type of movement. and so it turns out in the remarkable way that somehow the brain achieves a map of the body surface and so stimulation in one place will produce the movement of the thumb. stimulation of an adjacent reason will produce a movement of the wrist and elbow and shoulder and yet another region will produce movements of the leg. so in this way people began to realize that there was this precise register between the surface of the brain and the mull groups, the 600 or so you mentioned that have to be activated in precise pattern to produce a coordinated movement. >> rose: this is really a beautiful example, perhaps the best example we have in the brain of functional localization. that different aspects of behavior, different aspects of sensory perception are localized. the detailed map of the movements of the body on the surface of the brain. >> sherrington began to reason that if the cortex is the site of initiation or planning of movement yet the muscles are out in the periphery, there has to be an anatomical conduit or link between the site of stimulation and the site of muscle contraction and what sherrington in a systematic way did is to show that the information that is being initiated in these cortical regions has always of that information regardless of the type of movement, has to be funneled down the central nervous system, essentially to the spinal cord. and sherrington began to examine the nature of an organization of circuits that existed within the spinal cord. and very soon he realized and articulated within that within the spinal court... so this is now a cross section, a cut slab through the spinal cord, that the key intermediary between all of this cortical information and the actual contraction of the muscle was this group of neurons shown here which are spinal motor neurons. so of the tens of thousands of neurons, classes of neurons that existed in the brain, only one of those classes, the motor neuron, actually sends a process out of the central nervous system, out of the brain and spinal, to communicate with the periphery. so all of these dexterous tasks that daniel showed you are really dependent on the activity of motor neurons. and if we have 600 muscle groups we now know that in order to accommodate the combination of muscle activity, we need 600 motor neuron subtypes. so in a nutshell, the problem that the brain has to solve is how to initiate movement exactly when you want to move, how to control exactly which part of the body you want to move. and all of that information has to be funneled down into the spinal court and activate just the right set of motor neurons in the right combination to produce coordinated movement. >> rose: is there a dramatic difference between a reflex action and a conscious action? >> yes. so this is the other great insight that sherrington had and it comes to the third component of movement. there's no point in moving or acting if the brain and the central nervous system and the body doesn't get some reporting of the consequences of action. so sherrington spent much of his life not only working on the way the cortex interacts with the spinal cord but the way this information is fed back through recording of the state of muscle contraction. so any coordinated movement, the cup stacking that we saw, requires the fact that the brain received online information about the efficacy with which those motor tasks are performed. and that feedback information comes in through sets of sensory neurons that are acting as if you like a strain gauge, monitoring the intensity and the timing of mull contraction. and feeding that in the simplest reflex circuit back directly to form a single synaptic connection with the motor neuron. so this is an involuntary stretch reflex circuit, if you like. so perhaps we could first of all demonstrate this mono synaptic reflex in action. >> rose: all right. (laughs) >> so i'm going to try to fool tom's nervous system. in particular his low spinal cord, into thinking that this knee has been flexed and what you will see is that independently of any of tom's efforts of the contrary, the nervous system will respond by attempting to extend the knee back out and i will fool it by using a reflex hammer and using a very standard reflex. tom you just want to relax? what you can see is that a short percussion here, he kicks the leg back up. this involves three elements. it involves the final cord perceiving the fact there's been a stretch. this has a sensory feedback essentially in the spine cord, there's an integration of that sensory information in the spinal cord to the motor output which allows essentially a motor signal let's correct the problem let's extend the knee. (laughter) this would new york his sleep. so here is the spinal cord, here are the muscles, the knee, this pathway we've activated the sensory input into the spinal cord thenedly one synapse we've transformed sensory input to motor out put. normally sensory input comes in from vision and is transformed through a series of pathways to a motor action. here is the simplest version of this sensory motor transformation that exists in this central nervous system and sherrington, again, through these very simple devices would have many of the ideas about the way that circuits in the spinal cord integrate this information. >> this is a beautiful example. because one of the great challenges is to understand the logic of neural circuitry. here in 1900 he worked out the logic for the simplist neural circuit we know. >> if one looks through the history of neural science, many of the core principles we now accept as part of the canon of the field first came through studying this sensory and motor circuit. first of all, the existence of reflex circuits, second the nature of synaptic communication between one cell and another. these were all worked out by essentially studying the simple circuit. >> one general thing you can get from this. two things. one is in terms of the architecture of the circuit it's recapitulated up the hierarchy, it just get mrs. complicated. so off seinetory input, processing that goes on and then an output. so when you talk about reflex and voluntary movement, you can say voluntary movements have to require circuits based on this architecture but are more complicated. another interesting thing is that this circuit isn't there for that reflex you saw. the brain has been clever enough to use that building block for voluntary behavior. so the same spinal circuits that you can actually isolate with a reflex hammer are also incorporated into useful behaviors. >> rose: all right. let's talk a bit about disease and what happens to a sensory neuron when we have an awful thing happen to a person, whether it's a stroke or a.l.s. >> i think one of the questions is in fact if we start with this very figure, what happens if one interrupts various components of a circuitry. if, for example, there's a viral infection that eliminates the function of the sensory feedback, what are the consequences for movement? i think we have a very good video which will illustrate some of these features. so here is a gentleman who has, in fact, had such a viral infection whose motor think and powers are perfectly normal but whose ability to know where his limbs and joints are in space is severely impaired. thus when he wishes to take a sip from a cup of coffee, it's a herculean labor to get the fingers to grip the handle and still requires both hands to get the cup up to his lips. and that's without any loss of power whatever. this illustrates dramatically how critical it is to have that sensory input. guide, modulate, a servomechanism the control of final motor output. >> in this case, the motor neurons are still present, still connected with muscle. the only thing missing is the ability of that motor system to receive sensory feedback. >> so this will show the same gentleman being asked simply to stand, something any one of cus do. and what one can see... and this is a real-time photo. this is a laborious, slow act which essentially he cannot carry out despite having full power in his legs. so as you can see, he falls forward then rolls until he achieves an upright position. he's now out of the chair but a long way from standing. and in itself is an achievement, hence the thumb's up. but severely disabled because of loss of sensory function. >> when you have a desired behavior and you have to generate the behavior, the flip side is very important to learning and that's prediction. so control is about what i want to do, how do i generate the command. but the other side you have to learn is how to predict. that's saying given the commands i send out can i predict what will happen. and we know the brain has an internal stimulator. when i send a command out and get feedback from my arm, that's governed by physics. the physics of my body and sensory receptors. but we know within the brain there's a neural stimulator this which says i can see the command going out, let me anticipate or predict what's going to happen. so as i'm moving around the world doing things, i've got something in my brain simulating what's happening and trying to anticipate it. why would you want one given it's really going to happen? there's several good reasons. one is you can sit here and simulate movements to decide what the best one is. a really important one is one of the problems with mutual is when i send a command out and get the feedback there's a quart over a second delay between sending out and getting the feedback. there's a very long time if you want to play a fast tennis stroke. so one of the ideas is rather than working on reality. you can work on your internal stimulation which is possible. and we have a demo we'd like to show you to prove to you prediction. and what we've brought along is one of the heaviest books you can find in the library. (laughter) this is written by tom and eric. it's the standard textbook in the field. this is called waiter task. a waiter brings youing? a tray and off choice of when you lift it off the tray or the waiter does. so i want you to support this book. and now he's supporting the book. he's doing that having to contract his muscles. normally we wait. we like to have naive subjects but john's naive enough. what i'm going to do is remove the book and ask him to keep his hand still. watch what happens to his hand. effectively it's impossible to keep his hand still, to anticipate the timing and therefore relax the muscles. but if i ask john to remove the book with his other hand, his brain will predict the consequences and we can do it without any motion at all. this is something you can try at home. it's impossible to keep your hand still when someone else moves it because the brain is anticipating the actions of one part. and this mechanism explains why we can't tickle ourselves. so we know our actions affect the way we perceive sensory feedback and that's because of this predictive mechanism. i predict what ice going to happen and therefore i'm not surprised by it is. (laughter) >> rose: two questions. the import of all this or in terms of what insight in terms of understanding function. >> one insight is now we predict what we do, we're beginning to predict what other people do. that's a very exciting arena. not only do i have to predict my actions but your actions and john has some nice data on this. >> right. so in other words to become skilled, it's not enough to know how your body's going to respond to commands but you need to be able to predict what the player or your opponent is going to do ahead of time. >> rose: exactly! now we're talking about our favorite support, tennis. >> basketball! (laughter) >> well, i don't know a lot about that but i'll manage. this is a very interesting study that came out initially where they were interested in seeing whether it would be possible for an observer to know whether a basketball player at the free throw line will hit or miss the basket. >> rose: right. >> and they looked at professional basketball players watching this, coaches, and regular spectators. and what they found was that professional basketball players to a high degree of accuracy can actually tell you whether the blal go in or not before it ever leaves the shooter's hands. >> this is the observer. the observer can predict. they are basically running, as daniel said, a sort of simulation of what you can imagine is themselves doing it and therefore... and the way they do it and what was done in this study is that it looks as though before the ball ever leaves the hand here there is differing degrees of activation in the observer's motor cothaex go to the same very muscles they themselves would use to perform this task. and that seems to be a readout or simulation of what's happening when they're watching the player and that's why they can predict if the ball will go in or not. because a sense they've simulated doing it themselves. >> rose: here's what i need know. what people can do that? >> the basketball players themselves can do it. even coachers and observers can't do it. >> rose: a highly skill basketball player can watch another basketball player and predict before the ball leave his hands if it goes into the basket? >> and your question is crucial. which is in order to get these predictive circuits working, you have to have done motor practice yourself. >> i want to take you back to the case that bob describe odd motor performance under conditions in which sensory feedback is lost. the two videos that we saw. what we saw is the normal response to the loss of sensory information. but there are a few remarkable individuals who, faced with the same viral loss of sensory input have somehow found ways of co-opting other sensory systems to compensate for the loss of the sense of touch and the sense of limb position. and so the general idea that we'll be talking about is that normally the loss of limb position and muscle contraction leads to this dramatic miscoordination. but there's an individual in england, in waterman, who was incapacitated for the first two or three years but by a supreme act of will somehow learned to use his visual input to compensate for the loss of sensory input and recovered over the course of these two or three years through intense efforts to the point that he can now move, walk down the street. and this is a remarkable recovery for someone who was essentially bedridden for the first two or three years following this viral infection. so if we look at the video here, we can see he's walking, his step is somewhat abnormal but if you look, he's looking at his feet the entire time. now, he's going to collide with another and despite that he waves him on but keeps his eyes on his feet. and if you deprive waterman of vision, then his motor performance reduces and is degraded to something that we saw earlier. so this is a remarkable case of plasticity in a sense, to take the five or six senses that come in and then in an alternating way to change the sense that is being used to feed into the motor system. >> what can use that plasticity in overcoming the consequence of stroke? >> no so in stroke... i'll just give you an example. this this is a brain image and this white area shows you where a stroke has happened. this tissue has died because of loss of oxygen and blood. this person... this is on the left side of their body, they'll have a right-sided paralysis. so the big challenge for people after stroke is everybody gets better to some degree. the brain has an intrinsic plasticity that allows it to get better. what is that plasticity, how can we enhance it and how can prepredict how good a job is going to do? and interestingly enough, we've been able to show that early after stroke we can identify brain networks that are going to tell you what the potential for the brain is to recover in future and somehow what we think is happening is that... walk you through this. this is the side view of the brain. this is the motor cortex in red and this is the part way down to the spinal cord tom told us about. and a stroke like the one i showed you in that image is where you interrupt that descending pathway, where that crosses. so now there's no long a pathway down to the spinal cord and the personal loses their ability to move. but you also see in this lighter color other descending pathways coming from other areas of the brain. so that what we think is happening is over time that brain network i showed you that seems to predict people's recovery is somehow training people to unmask and strengthen these weaker connections down to the spinal cord that over time can substitute for this one we've lost and now they're just as intense as the original pathway was here at the beginning and through this learning process you've now found a new route down to the spinal cord to get function back. >> this is sort of the parallel in the motor system, what you saw for visual compensation. you're using another pathway to take over for damage to the initial pathway. and one of the really wonderful things that john and other people of his generation has shown is that unlike the early teaching in which physical therapy is introduced rather late after stroke, these people now initiate very early. you may want to elaborate. >> i think going back to your question about children being better learned than adults. what's interesting is those heightened plastic conditions that you see in childhood seem to be recapitulated briefly after injury to the brain. >> rose: now why briefly? >> well, we don't really know. what seems to happen is that genes that promote growth factors get turned on and then a little later genes that inhibit those growth factors get turned on. i think the idea would be that you don't want this plasticity to go on forever uncontrolled. so enough brief window that's a little like plasticity you see in childhood in early development and the key thing is to try and intervene, as eric said, in this window of heightened plasticity with still stimulation and with behavior to get more recovery than we used to. >> rose: the critical question about the critical time is how long is the time? >> all right. so in animal models it looks like it's about two to three weeks. >> rose: oh! >> we think, based on our work in humans, that that window may be as long as three months, although my feeling is we should really be getting in in the first month. >> rose: that's... that's... >> what is really remarkable about this is a complete reversal on what we were taught when we were in medical school in which one did not intervene with physical rehabilitation until much later in the disease. now these people intervening very early, taking advantage of this critical period which is therapeutically so beneficial. >> rose: let me go to a.l.s. and talk about a.l.s. and what we're learning there and where we are and what we know. >> the best way to introduce that is to turn again to the figure that we've seen previously of from tom jessell, making the point that the final common output for all movement and, indeed, any form of human exsuppression this spinal motor neuron which goes from the spinal cord down to the muscle. i want to digress for just a moment to comment a bit on this particular cell. this is an extraordinary entity, it's among the largest of cells in the body. it is so large that you can almost see it with the naked eye. it's a tenth of a millimeter. but what's extraordinary is, despite the enormous size of the cell body, the process that goes down to muscle itself is truly extraordinary. so that, for example, if this room were a motor neuron, say, in your low back, sending a process out to a muscle down here, that muscle would be about 150 miles away. >> rose: wow! (laughs) >> and that's valid. and so the point is that somehow this sell has to sustain this process, keep it alive, basically, to function normally. but here's the extraordinary point. this is a cell that essentially has almost no capacity to repair. it has a limited capacity to sprout. it has no capacity to divide and replace itself. so the extraordinary question is really why it survives as long as it does. so in motor neuron disease, the fundmental problem is this that this cell dies. you've heard the in stroke and spinal cord injury and other diseases one can recruit other parts of the nervous system to supplement or make up for dysfunction and produce pseudomotor output which is almost normal and we've seen examples of that here. but the problem, is when this final common pathway is shot, there is no compensation. there is nothing else. and so sadly when people have motor neuron disease and this cell type dies, typically they in turn die in three to five years. >> rose: they can't breathe or... >> exactly. so all muscle then becomes deer in separated, unplugd from motor nerves, breathing fails and they die. >> rose: do we know what causes that neuron to die? >> so we don't. and, in fact, i'll comment on that because that's a burning question in terms of the way we understand this problem clinically. we suspect there are many factors that contribute. certainly about 10% of cases we know that there are mutant genes that make sick or toxic proteins that actually actively kill a neuron. but we also know that behavior probably plays a role. there's some data that soccer players in italy, for example, have a higher-than-expected incident of a.l.s. we know head trauma may be a factor. we know that the environment may also influences, of course, the disease. and there may be just some role for what we call bad luck. >> this is such an important principle because what bob is showing here for amy trow if i can lateral sclerosis is almost the general principle in the nervous system. most diseases-- parkinson's-- they have multiple causes. so one doesn't think of simply a single factor like genes. that's one way of producing the disease but there are other ways of producing it also. >> yeah. that's exactly right. so one of the ways that scholars over the last several years have tried to understand this is to look at that 10% of cases where we know a miscreant gene and a sick protein have caused the disease. and so basically one can take, for example, a family. and here we have a pedigree with individuals in light and dark coloring, those who the light color having the disease, those with dark don't. and one can essentially profile d.n.a. and use profiles of d.n.a. to screen the families and essentially in that manner find the disease-causing genes. and i'll just tell you that it's... >> rose: and once you do that? >> well, the question is once one has a disease-causing gene, can one use that both to understand the disease and ultimately, of course, to find therapy. and the answer is that thus far there have been huge insights, enormous insights into the complexity of how a motor neuron dies when there's a mutant gene. but the ability to use that information therapeutically is still not yet realized. >> this has been true time and time again. that even when we know the gene, even when we know what is called the mechanism of path though genesis, how that gene does its damage, that mutant gene, we're stale long way if treatment. it takes a long time and they will discuss various therapeutic approaches that are possible in order to ameliorate the disease. it's a very difficult challenge. >> rose: any therapies in terms of stem cell or any other area of... >> we think that there are a variety of therapies that are very promising. i'll give you one or two examples. so typically in most of the a.l.s. genes that have been found so far, the concept is that the mutant gene makes a toxic protein and that wreaks havoc in many ways up and down the motor nerve. so clearly what one would like to do is turn off the expression of the toxic protein from the sick gene. turning off the poison at the top. stopping the dominos from falling before the dominos fall over. so that's a technique calledar nay silences. but the question of whether or not one can use stem cells to think closely about the biology of the disease is of enormous urgency and importance. let me give you company prime minister pl. it was reported approximately one year ago by colleagues here in new york and boston that one could take a skin biopsy from an a.l.s. patient, from there that prepare skin cells which could be transformed into a variety of molecular interventions into stem cells from which one could differentiate both motor neurons and some of the supporting cells from motor neurons, so for the first time one had in a petri dish a representative motor neuron from a living a.l.s. patient in a genetic con thaex reproduced the patient itself. >> one of the fascinating things about this area is that little tom jessell here has been a pioneer in studying how motor neurons developed. in fact, how the spinal cord develops. so he has figured out all the reagents, all the chemicals that are necessary and are used normally in order to get a motor neuron to grow. and that's been tremendously helpful in allowing you to do that with stem cells. >> yes, to come back to that example that bob pointed out. this arrow from stem cell to motor neuron is easy to draw on paper. (laughter) you can see that. (laughter) >> rose: it's not a slam dunk. >> but in a sense, a stem cell has the capacity by definition to give rise to every cell in the body-- to a heart cell, to a liver cell, to a kidney cell, to a hair cell. so the trick in a culture dish is how to coax those stem cells not to undergo embark on these thousands of different pathways but to find the recipe in which you can convert a significant number of these cells to one class of cell, a neuron. and then not just to a neuron, now a motor neuron. and so remarkably it's now possible to do that by adding just a couple of small molecule chem kls to that starting stem cell mix. you can get 50% of those cells to become motor neurons and the other cells become the cells that are sitting in the environment of the motor neuron. and in the context of a.l.s., i think that's had a dramatic impact because it means now for the first time we can... the field can essentially grow billions of motor neurons derived from an individual patient and then apply the strength of modern molecular biology and sell biology to understand the causal path though genesis. we know there are a dozen different genes that will give rise to what we call a.l.s. when they're mutated. are these really the same disease rohr they diseases that simply affect the same sell type? the motor neuron. so now i think the field can begin to study in a rational way the origins of the disease and in principle to think about replacing dying or dead motor neurons with other motor neurons. now, while that may be a promising approach, for example, in diabetes ngor parkinson's disease, if we come back to the circuit issues of 600 muscle groups each have to be enervated by their own set of motor neurons which have to receive the same input from cortex, it begins to illustrate the daunting challenges that face a cell replacement strategy. but perhaps through having patient-specific diseaseed motor neurons growing in a tissue culture dish, you can then begin to combine 24 this with chemical biology, with the pharmaceutical industry to screen for compounds that slow the degeneration of these motor neurons in a dish and ask which of them compounds will actually be effective in animal models. >> so you've pointed out two uses for stem cells. one is to study mechanism of the disease and therefore to develop drugs that inhibit that mechanism. and the other which you said is actually replace and your argument is this is unlikely to be very successful. could you give examples in which it might be successful? >> yes, i'll mention that them briefly. so a disease like type one diabetes which is in very large part the loss of insulin production is a much more attractive candidate for stem cell therapy, for turning stem cells into the insulin-producing cells and then reintroducing them back into the body. for a disease like a.l.s. where the organization of the sir scut so important, it's not impossible but it's difficult to think of how in an adult animal or an adult individual patient who's lost 50% of their motor neurons how you reintroduce those motor neurons. >> rose: what i'd like to do is something i think daniel may have touched on but i want to make sure i understand it. this idea that if we understand how the acting brain works and motor functions, it will be a key to unlocking? >> well, other... >> rose: other higher forms of brain sfungs? >> i think people may disagree around the table but i think we can't study things like perception and how that affects movement because effectively the way we use information is very important. and we don't do perception in isolation from action. so i think we can probably study perception and we'll all all about perception. but unless we can link that to the action we won't have the whole picture. so in the end we have to understand how the information is used and therefore once we solve motor control we will have by definition have had to solve all the other things. >> what has also emerged, and we preferred to this earlier in looking at the hand of a basketball player shooting a basket is that even when we don't engage in movements, our motor systems are stimulating the movement. so there's a lot more movement going on in our brain that is visible to the outside world. >> the point being made on that is that thought is basically movement planning without the movement. from an evolutionary standpoint, you could imagine if we understand motor planning and stimulation without the movement it's very likely those planning processs will co-opt it for higher level thought. >> rose: the hierarchy idea. explain that to me one more time. >> well, in order to carry out a movement, if i want to shake dan's hand, i have to decide that i want to do that. >> rose: decide, plan you want to do it. >> then you have to plan which muscles are involved then i need to recruit the motor system and spinal cord in order to carry out that movement. so there's a hierarchy, very much like we saw in the visual system. there are more complex processes steps. and what we also saw, each of those steps can be selectively interfered with by different diseases and each of these can be strengthened by learning. >> rose: i've got thousands of questions but i have limited amount of time. let me... what's the most... one question you most want to answer? >> for me, i would love to know what the algorithm or computer code the brain uses to generate skilled movement. so if we can write down the equation which is really explained how we learn and how we control the movements of our bodies, that would be wonderful for three reasons. one, it would be intellectual curiosity. just to know that would be wonderful. >> rose: (laughs) yes. >> but the only reason is it has two benefits. one is a minor one. if we could build machine which is could do human-like tasks that would be wonderful to put that into a technology domain. but also i think uerstanding basic function, we understand disease. i mean, there's lots of examples where if i understand the basic mechanisms we can understand what goes wrong with disease but also for rehabilitation. so if we really understood how normal people learn it would inform us for the sorts of ways we could improve learning and the sort of patient which is john has talked about already. >> i would real like to know what... good drugs, brain stimulation techniques and training protocols. could we make people after brain injury and spinal cord injury really much, much better than we can now? i mean, is that window that i talked about exploitable to the extent that we can get huge beneficial effects of rehabilitation? i really don't know how far we can go. >> 500 million years of evolution have been devoted to building a wiring diagram that produces action with a sophistication that we've begun to discuss. i'd be interested in understanding the way in which the assembly of motor circuits really explains the biomechanical demands that those circuits need to engage in in everyday life. to what extent are all our motor functions really prewired through genetic perhaps. to what extent does experience and environment build on that pre-wiring and if we understood that in the context of the motor system i think we'd have a new insight into the way that wiring diagram controls all aspects of brain function. and i think that's the ghel probably is attainable in the next two or three decades. really to have that this molecular... this combination of behavior and molecular programming of circuits and wiring diagrams. >> so tom has taught us over the years how a stem cell assumes an identity as a motor neuron and with that identity establishes a biochemical and a morphological identity that says i am a ngor motor neuron extensor process and functions to go to and activate contraction of muscle. what i'd like to know is how those developmental phenomena are recapitulated in an adult. so as we age and as the motor neuron is exposed to a variety of age-dependent stressors, the system nonetheless is able to reactivate some of the early developmental kinds of processes that establish the identity in the first place to sustain normal function. because i think it's through understanding these age-dependent responses to injury in the motor neuron that will finally come up with therapies for disorders like a.l.s. >> rose: sum up what we just saw but also tell us what we have to look forward to in our next episode. >> well, so far we've spoken about individual perception, individual actions. in the next episode, we're going to speak about group interactions, social behavior. we're immensely social beings. we bond with one another, we like certain people. i enjoy being with charlie rose. these are very important aspects of human interaction. but we not only bond together but we also form groups that are aggressive to anyone from the outside. what is the nature of bonding? what is the nature of aggression? can we understand how that's represented in the brain? can we see examples of social behavior, of bonding and of aggression in simple animals or is this something that's exclusive to higher primates, to human beings. to what degree is social behavior determined by genes? to what degree is it determined by learning, by later experience? are there diseases that selectively affect the social action of human beings? these are all topics we're going to consider in the next program. >> rose: i can't wait. see you then, episode four of the charlie rose brain series. captioning sponsored by rose communications captioned by media access group at wgbh access.wgbh.org tavis: good evening. from los angeles, i am tavis smiley. first of tonight, the first