Picower Institute for Learning and Memory Inaugural Symposium: "The Future of the Brain," pt. 1
TONEGAWA: All right. Good morning, ladies and gentlemen. Welcome to the Inaugural Symposium of the Picower Institute For Learning Memory at MIT, the future of the brain. I am Susumu Tonegawa, director of the Picower Institute for Learning and Memory.
Today is the 1st of December. And the first is appropriate for such a day like today. This celebration and formal inauguration of the Picower Institute for Learning and Memory is the first page of a new chapter of the future of the brain research. Today marks our first milestone in the history of the Picower Institute, because this is the first time that all the members of the Picower Institute and the laboratories have been assembled under one roof. And isn't it magnificent roof we have?
The person I'm about to introduce has many firsts in her life and in her accomplishments. Too many to mention. But in my view, two stands out as particularly significant today. And this person is the first life scientist to lead MIT. And this person is also the first woman to hold this post of MIT president. And I'm of course speaking about Dr. Susan Hockfield, who is the 16th president of Massachusetts Institute of Technology.
Dr. Hockfield is a noted neuroscientist in her own right, but with her at its helm, MIT is poised to have many firsts in coming years, integrating life sciences with engineering and technology, which will inevitably foster the promise of many firsts, many discoveries. And these are the discoveries which will surely and dramatically enrich and improve the world we live in. So please welcome our president, Susan Hockfield.
HOCKFIELD: Good morning, everyone. Thank you, Susumu, for the introduction. I just have to say, I love this building. It's an enormous privilege and a really-- it's a great joy to join all of you this morning for the formal inauguration of the Picower Institute for Learning and Memory. I can't imagine a better way to inaugurate this new home and the first important new chapter than the distinguished symposium this morning on the future of the brain.
And what a spectacular symposium it promises to be. It's a roster of who's who in the global neuroscience community. We all know about Susumu's persuasive abilities and tenacious persistence. You can imagine how much he had to exercise those skills in assembling the stellar group that will be with us today. We have pioneers in brain research, the people I consider to be the real thought leaders in our field, who will help us celebrate the opening of this marvelous institute.
And we have a lot today to celebrate. We have a new place. We have a great vision. We have a remarkable community. And this all has been made possible through an extraordinary commitment by visionary donors. First a new place. Thanks to our architects, Charles Correa associates and Goody Clancy and associates. Thank you so much for providing this--
State of the art space for the difficult technical work of doing cutting edge neuroscience as well as together with compelling community space that catalyzes conversation and collaboration. It is a profoundly successful architectural phenomenon, and we really thank you.
Second, a great vision. The vision for the Picower is captured magnificently by the title of today's symposium, The Future of the Brain. The title is not The Future of Brain Research. The future of brain research would convey only what we will do, not the consequences of what we'll do. The consequence of the work that will be done at the Picower Institute for Learning and Memory will, I believe, change the future of the brain itself.
Let me explain a little bit. Through the Picower foundation, Barbara and Jeffrey Picower support many important programs, but their deepest passion has been for their programs in education and medical research. It's their vision, and it's a vision that's shared with the MIT Picower scientists, that here research that leads to a deeper understanding of learning and memory will translate into improvements in how we learn and remember. Those improvements one day will amplify the potential of education and will also reduce disease and age related memory loss. Simply said, neuroscience research at the Picower Institute will improve human lives by improving the human brain.
We enjoy a remarkable community here. The excellence of the Picower Institute today reflects the dynamism, the drive, and the insight of Susumu Tonegawa. Susumu is not just a brilliant researcher but also has proved to be a superb institution builder and a great talent spotter. He's assembled a cadre of scientists that week after week and paper after paper successfully tackles some of the most interesting and pressing problems in neuroscience today.
The incredible Picower team is joined in this magnificent building by a group of extraordinary colleagues in the McGovern Institute for Brain Research and the Department of Brain and Cognitive Sciences. Together more than 40 faculty. And I cannot imagine that anyone would fail to see that this is one of the very best communities of neuroscientists in the world.
The larger environment of the Picower amplifies its power. By any measure, this intersection of Vassar and Main Streets has to be one of the very most exciting and promising places for one of the most compelling intellectual domains of today, which is taking form at the convergence of the life sciences and engineering. Physically present right here, in an area that I often call the great circle, is an astounding array of activities at this convergence.
Of course, in this building, we have neuroscience. Across the street, in our new Ray and Maria Stata Center, it houses our departments of electrical engineering, computer science, and also our Department of Linguistics and Philosophy. Just behind that is a building that houses our division of biological engineering. Next to that, a building that holds our Department of Chemical Engineering. Right next to that is the Department of Biology. Next to that, our Center for Cancer Research.
Across the street, just diagonal from here, is a new building that's almost finished for the Broad Institute that will take the power of the human genome and turn it into new ways of discovering causes for diseases and their cures. Next to the Broad is the Whitehead Institute for Biomedical Research, which has been a fount of insights in molecular genetics for decades.
This is a great circle that brings together engineers and scientists, students and postdocs, to create the new world at this new intersection. I, of course, can't neglect to mention our neighbors, the almost 150 pharmaceutical, biotech, medical device companies that are within walking distance of our campus. And our neighbors in the other direction across the bridge, the Massachusetts General Hospital and three other of the leading academic hospitals in the nation. This is a powerful combination. This is a powerful place. It will, in many ways amplify the impact of the research that's done at the Picower.
Now, most importantly today, we celebrate visionary donors. The scientific power and accomplishment of the Picower scientists reflects the commitment of Barbara and Jeffrey Picower. Their generous gift established a permanent endowment to support Picower Institute faculty and their research as well as the construction of our magnificent new facilities. The unrestricted funding that they have provided sets these brilliant scientists free, now limited only by their imagination and their intellectual tenacity. It is a great and a long reaching gift. Thank you very much.
In addition to the generosity of the Picowers, the work of the Institute has been catalyzed and has been supported by other wonderful supporting organizations. The Sherman Fairchild Foundation really catalyzed these new neuroscience activities at MIT. The RIKEN Brain Science Institute has contributed importantly to our work. The NIMH and the HHMI has also provided important support, and we thank them all for the important funding for this endeavor in the past and in the years ahead.
Today we have technologies available to us that were simply unimaginable only a few years ago. And some of the best minds in the world are turning to the study of the brain. So I am profoundly optimistic about the future of brain science. My own belief is the work that will be done here at the Picower Institute will go a very long way toward alleviating the crippling diseases of the brain, toward enhancing education, and improving the quality of our lives.
In closing, let me just say again how inspired I am by the Picower Institute here at MIT. I believe I speak for all of us today when I thank, again, Barbara and Jeffrey Picower, the Picower's other supporters, Susumu Tonegawa, and his incredible colleagues for giving us a reason to expect revolutionary results from the Picower Institute in the years ahead. Welcome to MIT. Welcome to the symposium, and please, I hope you all enjoy the day. Thank you.
FLATOW: Welcome. My name is Ira Flatow. I Host Science Friday at NPR every Friday. I hope a few of you listen to us. Thank you, thank you.
I'd also like to thank all of you for coming today, and I'll just add my two cents to the beauty of this building. And I like how they built the skyboxes up there. You know, there's some people up there. And some Bob Uecker seats also, I think. We'll need you later, because we're going to be talking with the Nobel Prize winning panelists we have today. And let me tell you a bit about what's going to happen this morning.
We're going to have a panel discussion. I'll sort of be your tour guide for the day today. We have an unprecedented assemblage of Nobel Prize winners who, as Ed Sullivan used to say, right here on our stage will be assembled. And each one of them will present, make a short presentation, about where they see brain research going. They'll be sort of limited to the amount of time they can talk. But trying to get scientists to do anything is like herding cats, so we'll see how well that works out.
So each one of them is going to come up and make a short presentation. We'll have a break around 10:25, and we'll come back, we'll finish the presentations, and then we'll have questions from the audience. And I'm hoping everybody-- if you're up there, you can shout out a question. If you're down here, we'll have a couple of microphones to pass around so that you can ask the questions.
First up of our first Nobel Prize winners is you've already met Dr. Tonegawa. He is the winner of the 1987 Nobel Prize in Physiology or Medicine. He's also the director of the Picower Institute. So let's please welcome back Dr. Tonegawa.
TONEGAWA: Okay. The task assigned to me and my fellow Nobel laureates this morning, as Ira told you, is to predict about the future of brain and the brain research. And this is a very difficult thing to do. The French dramatist Eugene Ionesco said, we can only predict things after they have happened.
So, nevertheless, I urged my fellow speakers to venture into this difficult task. I myself had a good excuse to spend the summer-- actually, most of my time fulfilling another task. That is to introduce to you the Picower Institute. So I will give you a short account of what-- a biased account of what I think will happen in the brain, the brain research in the future at the end of my talk. But let me first give you a brief account of the mission of the Picower Institute and its work.
So Picower Institute is in the Department of Neuroscience Research Center at MIT. And its intellectual focus, as indicated here, is deciphering molecular cellular network and the brain systems mechanism for learning and memory and associated cognitive phenomena, such as attention, emotion, consciousness, et cetera in health and disease.
Now, here is a brief history of the Picower Institute. It started with a grant from the Sherman Fairchild Foundation at the small Center for Learning and Memory in 1994. And three years later, CLM was expanded on collaboration with RIKEN Brain Science Institute of Japan. And then the CLM was here further expanded in 2001 and renamed the Picower Center for Learning a memory with a large and generous gift from the Picower Foundation established by Barbara and Jeffrey Picower.
In 2003, we had a groundbreaking of the Brain Cognitive Science Project Building, this building here, of the homes for the Picower Center, McGovern Institute, and the Brain and Cognitive Science Department. This year earlier, this Picower Institute was renamed as a-- Picower Center was renamed as the Picower Institute for Learning and Memory, and the Picower building was being inaugurated.
Now, we put this learning and memory as part of the name of the Institute. So why do we study memory? Well, because memory is a mental groove and central to our mind. Now, for you in daily conversation, memory might be misty recollection of past times or a desperate effort to find lost keys. But actually, memory is much more than that. It is a mental groove to bind your life's experiences.
So imagine a life without memory capability. A life like that would be a life in dissolution, life with no connection to your past, present, and future, and it will be the life with no ties to people or to events occurring around you. Most tragically, the life without memory means you don't have sense of yourself and therefore, often the Alzheimer patient in advanced stage will ask you this question. Would you please remind me again who I am?
So for some of you, probably you will agree that memory and other cognitive function is based on the function of the brain. But some of you might be a little bit surprised if I tell you that this brain actually is a machine. And it is a machine. It's a machine evolutionarily assembled on this planet and the most complex machine. This diagram shows how the information flow in this machine. So the information is extracted from experiences and the environment through the sensory system shown here.
And some of this information is processed and stored in part of the brain, such as the neocortex and the other part, like the hippocampus and the basal ganglia. This stored information subsequently will be compared and integrated with the information which is arriving to the machine, to the person, to the animal, moment, moment, online information. And then the command signal will be generated, and that command will be executed through the motor system.
Now, one of the hallmark of the brain, particularly the human brain, is the enormous complexity and multi-layer organization. So at the bottom, you start with molecules, DNA proteins, and other variety of molecules with low molecular weight, which generate compose very specialized areas of a single cell such as synapse, dendrites, axons, soma, and so on.
And though the different parts of the cell compose a single independent neuron, which shows a particular function. And these single neurons interact with each other through synapses to form so-called brain ensemble, neuron ensembles. And these neuron ensembles of different kind will interact each other to form what is called functional brain systems such as the hippocampus.
And then a different brain system interact, such as cortex, amygdala, and hippocampus, to generate a function of the whole brain. And that is the basis of the cognition and behavior of whole live animals such as the Drosophila, mouse, monkey, and my postdoc [INAUDIBLE].
So you can imagine from this that if you really want to understand the mechanism underlying cognition, like memory, you need to identify the processes and the events occurring at all of these different level of complexity. And then further, to identify cause effect relationship of those events occurring at the different level of complexity. And then in order to do that, one need a variety of different experimental technology designed to detect these processes and events at a different level.
So at Picower Institute, we use a variety of different technologies. We use molecular biological, serobiological techniques like shown here. And also-- oops. I don't know how to [INAUDIBLE]. Okay. So then we use invertebrate and vertebrate genetics. We use electrophysiology of live cells at the brain slices or cultured neurons.
And we'll use also in vivo recording, so-called in vivo recording, of live animals by implanting electrode into the experimental animals. And then we also use very high tech imaging technology such as confocal microscopy, multi-photon microscopy, and so on. And then we also use non-invasive imaging technology like functional MRI. And finally, at the highest level, we also subject animal to the variety of behavioral studies.
Now, to carry out all this analysis is a daunting task. It is beyond expertise of one person or one laboratory. So at the Picower Institute, strategy is to assemble people trained in the different field with different expertise, but they are all interested in this large intellectual theme that is learning a memory and the brain plasticity. So here is a picture of the current Picower Institute faculty. Mark Bear, Ellie Nedivi, Carlos Lois, Yasunori Hayashi, myself, Morgan Sheng, Matt Wilson, Earl Miller, Troy Littleton, and Mriganka Sur.
The great thing about the Picower Institute faculty, as I said, their expertise is very complementary, and there are a lot of collaboration going on among different laboratory. At the same time, keeping the identity of each laboratory. Another feature of this faculty is that all of this faculty are relatively young, 30 and the 40's, except this guy.
We have also two faculty members coming next year in 2006. One is Li-Huei Tsai, currently at the Harvard Medical School. And she will give you a short presentation this afternoon. She is the expert in basic mechanism of neuronal migration as well as the codes and the mechanism for underlying Alzheimer's disease. And Wolfram Schultz is another person we expect to come next year, and he is currently the professor at Cambridge University. His expertise is reward signal and reward prediction signal and learning.
Now, I have this task of trying to give you a glimpse of some of the recent research accomplishment from the Picower Institute. But of course, I cannot do it justice on this subject, because my time is very limited. So I hope my colleagues will forgive me that the fact that I pick just a few of them. And even so, I can spend only two or three minutes for each lab and try to explain to you what they are interested in and what they have done recently.
The first is Morgan Sheng. Morgan Sheng is a great molecular neurobiologist and biochemist. And he has been-- at least one of the major goal of his laboratory is to decipher the molecular architecture of this very important unit called neuronal synapses, indicated in here.
So his research will address-- just like in order to build this magnificent suspension bridge or a motor engine, you have to identify the components, and then you have to know how these components interact with each other to form this wonderful product. And this is what they want to do for synapses. And so what is the molecular structure of underlying components, and how do they interact together to allow neuronal communication, what is called the synaptic transmission and the synaptic plasticity?
So here is the electromicrographic picture of a synapse, which comes with this round structure, which I hope you can see, which is the presynaptic vesicles containing neurotransmitter. In this case, glutamate. And then also the lower part comes with these glutamate receptors indicated by this blue [INAUDIBLE]. And the darker one being NMDA receptor and the lighter one AMPA receptors.
So much has been known conceptually about these receptors. But the detailed structure of them has been poorly known. And Sheng's laboratory have recently succeeded in visualizing AMPA receptors at very high resolution, high rate resolution using electromicroscopic technology. And interestingly, they found two forms of AMPA receptor. One is called the type I. The other is type II.
And they have shown that the transition from type I conformation to type II conformation correlated well with opening the receptor channel, which happens when the glutamate bind to these receptor at the synapse. This is a very important fundamental discoveries, which had a lot of implications for future drug discovery.
Now, my laboratory is interested in understanding the mechanism underlying memory but using rodent as the experimental animal. And we are studying this at both molecular and circuit levels. Earlier studies have shown that both in human, on this left, and also in animals, the part of the brain called the hippocampus played a crucial role in the most familiar form of memory called episodic memory. Memory of events, memory of episodes. However, little has been known about how each sub region of the hippocampus, such as [INAUDIBLE] and the CA1, and then the circuit connecting them would play a specific role in a specific aspect of episodic memory.
And our goal is to understand the function of a different part of the hippocampo sub region and the circuit. And our strategy, main strategy, is to use genetic engineering and produce mutant mouse strains in which a gene encoding at a specific receptor, in this case NMDA receptor that I referred to when we talked about Morgan's work, [INAUDIBLE] is deleted from the genome. [INAUDIBLE] receptor is not functional anymore, but only in a particular part of the hippocampus. Everywhere else the gene is normal. And [INAUDIBLE] animal to the behavioral studies and find out what kind of a deficit do they show.
Now, it turns out I can't tell you all the details. But if you knock out this one gene, one out of 30,000 gene, specifically in the CA1 circuitry in the hippocampus, then this animal have a severe problem in acquiring a memory, but even after repeated practice. but if you knock out the same gene, if you damage the same gene in this area called the CA3, then [INAUDIBLE] totally normal in acquiring memory. However, they exhibit severe problem in recording memory.
And as you know, the recording is as important as acquiring memory information, because if you don't record it, you can't use stored information. So they show a specific deficit during the record process. They can't record in the normal way. So what do I mean by normal recording process? In order to explain that to you, I want to ask you a simple question. Did you have a good dinner last night?
And many of you probably did have a good dinner. And you are, as I speak, you are reconnecting. You are remembering the dinner you had yesterday and the detail of that. With whom, where, and what you eat, and so on. In other words, you need only a fraction of relevant information to record rich content of event, in order to record the memory. So this process is called pattern completion. You can't see it up here, but pattern completion.
Now, it turns out that if you knock out the single protein in CA3 circuitry in the hippocampus, then you can acquire memory, but you can't recall the memory with limited cues. It turns out this is the capacity we normally have, but it is very sensitive to aging. So that as we get older, come to my age, we have often problem in recording the past event with a very limited cue. Everything get to the tip of the tongue. Although that doesn't mean that the memory is gone. It's there, but you need more cues to recall it. And it's the same thing in early Alzheimer patients. They exhibit this phenotype, this symptom.
Now, if we look at the same gene now in the dentate gyrus. This is called the dentate gyrus area. Then this animal also have no problem acquiring memory, generally. However, they have a different kind of deficit. They have a problem in distinguishing similar events, making a distinct memory for two similar events. For instance, if you translate into human, if you park the car yesterday in one of these lots and this morning park another place and then this afternoon when you go there and in the large parking lot, you will have a problem finding your car unless your activity in the dentate gyrus is normal. So that's the kind of thing we do in our lab.
And then I want to shift to Matt Wilson's lab. Matt Wilson's lab also are interested in use rodent as an experimental animal and study the hippocampus and some other area of the brain. But they study this at neuronal ensemble and the brain system rewards. We have a lot of collaboration with them. So they are also involved in the molecular studies, but their original expertise come from this great technology that he invented several years ago. That is while animal is exploring a space like that, [INAUDIBLE] and remembering the feature of this space, which will generate what is called spatial memory, memory of particular space environment.
As animal do that, map technology allows some monitoring of the change of the activity of a bunch of neuron points in the hippocampus as the animal form a memory. And this can be analyzed in such a way that you can detect what is called representation of the behavior memory at the level of activity of neuronal ensembles. And using this technology, for instance, this is an example of 10 cells here from top to bottom. This is the time. And when you see the color here, that means that that particular neuron, this is neuron one, neuron 10 here going horizontally. When you see this color, that means that neuron is firing and very active.
So animal is awake and exploring this circular track. And you can see that 10 neurons fires at a very specific pattern like this. Now, if you put the animal a few minutes later into a secluded area where they can take a rest and fall asleep, and during this REM sleep, they probably see dreams, as we do. And the amazing thing is, without any input from environment, totally separated, segregated from an environment, this same set of neurons fire in a pattern which is very similar to the pattern that they fired during awake exploration.
That means there is a replay of activity internally in the brain during the REM sleep. And since it is known that the replay will help to fortify the synapses involved in that neuronal ensemble, Matt and his colleagues propose this is a mechanism to stabilize memory, to consolidate memory. And so the message to the students who are going to take finals soon is that even if you want to cram, try to sleep the last three or four hours and have a dream before you go to that class.
All right, now I want to shift to Mark Bear. Mark Bear's lab's major interest is change of the cerebral cortex, particularly the visual cortex, by a sensory experience during development. Earlier, Mark made a great discovery on the mechanism underlying synaptic strengthening as well as weakening. And they have shown that one form of this synapse weakening due to the sensory stimulus called Long Term Depression, LTD, actually plays a crucial role, at least in part, for the [INAUDIBLE] effects of sensory deprivation, such as the visual deprivation, during the brain development.
Now, this is a very basic study, basic research, but you never know what basic research lead. So recent study has shown, study in their lab has shown, that LTD mechanism, this weakening of synapse, are leading to a discovery of molecular mechanism for fragile X mental retardation, which is the best known inheritable mental retardation and also known cause of autism. And amazingly, their recent study is suggesting that this malfunction can be correctable by drugs. And Mark is marshaling the resources to put this into the clinical trial. He hopes he can do that within a year or two.
And finally, I want to talk about Earl Miller's work, Earl Miller being the associate director of the Picower Institute. So Earl works also uses the electrophysiological technique applied to live animals, but he uses primate, monkeys, rather than rat. And the monkey has a certain advantage as an experimental animal, because they are closer to us. They have a more sophisticated cognitive function, such as decision making, some other, maybe consciousness, and so on.
So this is only one of the many, many accomplishment that Earl Miller's lab have made during the past several years. But what they are studying recently is how monkeys and we learn and follow rules, what is called rule learning. With a red traffic light, you don't cross a road. Green, you can go. That's rule learning.
Now, the dogma in the field has been that the evolutionarily more advanced area of the brain called the prefrontal cortex is the one which learns rule first. And then as you learn this and repeat this, then the more primitive part of the brain called the basal ganglia, shown on this side inside the brain, takes these familiar rules and form automatic habit. You don't have to think about when you cross a road here, you don't have to think about red, oh, should I? Oh yes, I can go. No, you do that automatically. So that's the way the rule learning usually occurs.
But their recent study has challenged the dogma in the field. They have shown that the basal ganglia learns a rule faster than the prefrontal cortex. But this is the studying activity following the activity of the neurons in these two areas of the brain. If you also follow the rule learning by behavioral experiment, they have shown that actually monkeys behavior follow the slower prefrontal cortex.
So the primitive basal ganglia may start learning the rules first, but the prefrontal cortex, the brain's so-called executive branch of the brain, and the final arbiter of behavior monitors the basal ganglia and more slowly but more judiciously arrive at the correct answer, correct behavior.
So that's all time I have, and hopefully you'll allow me to use another three minutes to speculate about the future. So I have one slide. And here it is. Oh, what happened here? Okay.
So as I said, future is difficult to predict. But I believe that in all of these works I have talked about is carried out using animal models. But I know that some work going on in the [INAUDIBLE] world, even at MIT, to try to understand how human acquire memory and so on. But currently, the major technology which is used to study human brain is MRI, Magnetic Resonance Imaging, and some other technology, which is called a non-invasive technology.
These technologies are very useful. It's generating, giving us a lot of information. But people are aware that there is a limitation in this technology in terms of temporal and spatial resolutions. So if we really want to understand the brain mechanism underlying human cognition, in the future, we need entirely new technology which is based on totally new principle, which we hope will be able to analyze what is going on in the brain even at sub-micron level, which is a level of the size of the single synapses.
Now, where does it come from? What kind of idea is there? I have no idea. And I don't think people have an idea either. But I would hope that a place like MIT is the ideal place to promote and explore this direction, because we have such a fantastic Department of Engineering and Department of Physics and so on. So we should have, we should promote the interaction between neuroscientist and engineer and physicist.
Of course, we have to continue to try to find a way to develop a new diagnostic and therapeutic method for psychiatric and neurodegenerative diseases. In addition to the traditional method of finding drugs by drug screening, as you know, that for some of these diseases, particularly for neurodegenerative diseases, stem cell research is going to become very important. And we can discuss this, the implication and promise of this technology, when we have a discussion later.
And I would also like to point out that although this is not, number three, this is not a real therapy for brain diseases, but this development of computer brain interface, a so-called neuronal prosthetic, is very important. Because this is a technology which will not cure the patient but will improve the quality of life of patients, sometimes very prolonged period, more than 10 years. Like for instance, Parkinson's disease. There have been this technology being applied and the patient can live by themselves. And I think a place like MIT also should promote technology like in this field.
And four and five are more philosophical things. And I would like to take that the neuroscience-- I will take this position. Neuroscience is the first natural science, branch of natural science, which has a promise to integrate natural science with humanity and the social sciences. Because after all, what we do, what people study in humanities and social studies are what is a reflection of what is happening in the brain. And also including engineering.
And of course, in order to do all this, there will be a serious issue of the establishment of new ethical guidelines, which we can also talk about later. So I will stop it here. And if I took them more time than assigned it to me, I apologize. Thank you very much.
FLATOW: I hope you were taking notes on this, because you will be tested later. There is so much to talk about later when we have our general discussion after the break. I'm hoping you will raise questions about some of the things that we're hearing about from the speakers this morning.
Our next speaker is Sydney Brenner. Dr. Brenner is a distinguished professor, research professor, at the Salk Institute. He is one of the past century's leading pioneers in genetics and molecular biology. Most recently, Dr. Brenner has been studying vertebrate gene and genome evolution, and his work in this area has resulted in new ways of analyzing gene sequences, which has developed a new understanding of the evolution of vertebrates.
I'm sure you're familiar with his work and his discoveries, among them being Dr. Brenner established the existence of messenger RNA. He demonstrated how the order of amino acids and proteins is determined. He also conducted the pioneering work in the roundworm, which has been used as a model organism and now widely used to study genetics. His research has garnered insight into aging, nerve cell function, controlled cell death, apoptosis. He's also the 2002 Nobel Prize winner in Physiology or Medicine. There's so much to talk about. We'll have hear what Dr. Brenner has chosen to speak about today. Dr. Sydney Brenner.
BRENNER: Well, thank you for the opportunity to talk about the future. In fact, I should be sitting down now, because my time is up.
But I'll speed up what I have to say so that I don't poach too much on Richard Axel's time. I think it's very hard to predict the future, because I think the future will be what we make it. If we know we're going to make it, then of course, we'll predict it. So let's see what we want to make of the future.
We talk about the future of the brain. My own brain has very little future. But I think the future of research into the brain is going to be extremely important. And the first thing I'd just like to emphasize, something that Susumu said, which is that we have become released from biological evolution. The moment primitive man gained memory, learned how to use tools, we entered a new phase of cultural revolution. Though we didn't have to wait millennia to gain mutations that would give us a furry skin to protect us from the cold. We simply went out and killed the bear and wore his skin.
So when you think about it, we are all the products of the brain. And as I think I've often said, quoting a very well known scientist called Woody Allen, that we haven't achieved more with the second most important organ in the male body than with the first organ. So the important thing to grasp is that it is also very important from this that we are the products of our brains and that all of our latter evolution since the last 100,000 years, perhaps, certainly in the last 25,000, has been the product of our brains.
But I want to just talk a little bit more rather on this rather broad area, on something which is a little detailed. And I'm sorry, but I'll have to mention the three letter word, DNA. There are enormously challenging questions left of a great intellectual challenge. And they are to try to gain understanding in what we've been doing for many years now, which is hard to gain an understanding of how this internal description within ourselves that is written in the DNA language of our genes gets transformed into fully functioning organisms. And I think that connection between genotype and phenotype, especially for complex animals, will remain the most challenging problem in biology.
And I don't think we can avoid it, because you realize that we can only ask three questions in the life sciences. We ask questions which is, how does it work, which is questions of function or physiology. We ask questions about how to get built, which is questions of development, construction. And then the fundamental question of all, which is how did it get that way? How did we actually, by changing these letters in the DNA, make things as complex as ourselves and so different from everything else? And everything is unique in that sense.
I've often said that I only have one wish for the future, which is to be allowed to come back, wherever I am, in 2053 on the 100th anniversary of the discovery of the double DNA helix. Just for one day, just to see what has happened. What the world is like, what science is like, and if they are listening somewhere, I just repeat the plea on this auspicious occasion.
I think we are being compelled. Biological sciences are being compelled, if not to act in that direction but certainly to pay lip service, to the practical applications of our work. That is fine, and I agree with that. But we should not let, if I can put it crudely, the pharmaceutical companies of the world direct our conceptual development. As I think it is still important, and especially for young people, that they should know that there are really deep, deep problems to be solved, deep intellectual problems to be solved.
I think such questions, for example, of can we compute behavior from a wiring diagram? I don't think we can do it yet. But certainly if we're going to gain an understanding of brains, we must accomplish that task. There are many questions like that in the brain sciences, and I think that looking at the audience upstairs who should be getting back to the labs to try and solve these problems, of course, they are the future of the science.
So I just would like to finish here and hope I've just given you a glimpse, I'm sure, just to open myself to many questions. But I didn't want to take much time from Richard Axel. Thank you very much.
FLATOW: Don't go to the lab just yet. It's good to see that you're all sticking around, because you'll have a chance to ask questions of our speakers later. Dr. Axel has been promoted heavily now, so he'd better have a heck of a lecture coming up. Our next Nobelist has enjoyed what you might call the sweet smell of success by studying smell itself. He's the co-winner of the 2004 Nobel Prize in Medicine or Physiology for clarifying how the olfactory system works.
Dr. Richard Axel is a university professor and professor of biochemistry and molecular biophysics. It's all up there. [INAUDIBLE] Pathology at the Columbia University College of Physicians and Surgeons. Dr. Axel's research continues to focus on olfactory perception. In particular, how the sense of smell is established during development, how it may change over time, and ultimately, how certain smells can elicit appropriate thoughts and behaviors. Please welcome Dr. Richard Axel.
AXEL: I too want to congratulate MIT, Susumu Tonegawa, the members of the Picower, for this marvelous institution, the opportunity to celebrate the opening of this edifice. Thank you for the opportunity to be here.
Now, as you might surmise, I'm going to talk to you today about one aspect of neuroscience that intensely interests me, perception. This is not a nose. It is portrayal by the surrealist Rene Magritte of his own brain's representation of the external world. It's a vignette that reveals a tension between image and reality, a sense of slippage that is a source of creativity that's persistent in art, brought to its culmination by the surrealists.
The problem as to how the brain represents the external world, of course, is not only a central theme in art but is at the very core of philosophy, psychology, and neuroscience. It is the very essence of the problem of perception. And this is what I want to talk with you about this morning.
Now, let's first consider the first problem in perception, and that is the recognition of sensory stimuli. There are many ways for organisms to probe the external world. Some smell it. Others listen to it. Many see it. Each species, therefore, lives in its own unique sensory world to which other species may be partially or even totally unaware. A whole series of specific devices alien to human perception have emerged. Bio sonar in bats. Infrared detectors in snakes. Sensitivity to magnetic fields in birds.
So what this means is that what an organism detects in its environment is only part of what is around it, and that part differs in different organisms. To me, this means the brain functions not by recording an exact and precise image of the world but by creating its own selective picture, a picture that's largely determined by what is important for the survival and reproduction of the species.
By inference, then, sensory impressions are apprehended through the lens of a particular perceiving brain, and the brain must therefore be endowed with an a priori potential to recognize the sensory world. Our perceptions are not direct recordings of the world around us. Rather, they are constructed internally according to a set of innate rules.
Colors, tones, taste, smells do not exist as such outside of the brain. Biological reality, I would argue, therefore reflects the particular representation of the external world that a brain is able to build and a brain builds with genes. We're therefore trapped in the representation of the world made possible by our genes and can't conceive of realizing it in any other way.
But how can genes provide insight into the astonishing problem, the problem of how the brain represents this world? The brain consists solely of a collection of excitable neurons. How is it that the rich array of mechanical, optical, chemical properties that define touch, hearing, vision, smell, taste can be represented by bits of electrical activity that can only vary in two parameters, time and space?
I'd like to illustrate this problem by discussing how the chemo sensory world is represented in the brain, how chemical structures can be represented in brain space. We'll move through this, I hope, rather quickly.
Let's consider the problem in the context of what Sydney has alluded to. That is, micro circuitry. The olfactory system is quite simple. You smell by virtue of an epithelium, a collection of sensory neurons that reside in the posterior recesses of your nose. The epithelium consists of a neuron, a sensory neuron, which sends a process out to the external world.
And it is on the very tips of this process that reside the receptors that bind odors and translate the energy of odor binding into alterations in the electrical properties of the membrane. And this is then transmitted down an axon into the brain. And this axon actually passes directly through the skull. So we have a direct connection in this system between the outside world and the brain.
We can divide the problem of how it is that we can smell as reductionists into two problems. First, then, how is it that we can recognize this universe of molecular structures that we call odors? Man, for instance, is thought to be able to recognize hundreds of thousands of discrete odors. I have no idea where this number comes from, but nonetheless, this is a serious problem. But a more complex problem is the problem of how we discriminate this vast array of odors. That is, how does the brain know what the nose is smelling?
Now, the first problem was solved by the efforts of Linda Buck in the laboratory about 15 years ago when Linda isolated the genes encoding the receptors that sit on the tips of these neurons that recognize odors. And what she observed is that unlike vision in which three photoreceptor molecules are capable of recognizing wavelengths of light that accommodate color vision, and only 30 receptors are required for taste, in humans, 900 receptors, perhaps 3% to 5% of our chromosome, is dedicated to the recognition of odors.
And this concept, this principle of a vast array of receptors, threads through all of eukaryotic species. So we recognize the vast array of molecular structures, the vast number of odors, by virtue of maintaining in our genomes a very large number of odorant receptor genes.
Now, these odorant receptor genes allow us to pose the second problem. That is, how does the brain know what the nose is smelling in molecular terms? We can now ask, then, how does the brain know which of the multiple receptors have been activated by a given odor? If I can come up with a cogent model for this, I have the beginnings of an understanding of perceptual discrimination. This was simplified for us by the observation that a given neuron in the nose makes only one receptor.
And so we could now reduce the problem of how does the brain know which of the multiple receptors were active by asking which-- how does the brain know which neurons have been activated? And by analogy with other sensory systems, we could then ask, does the brain know which neurons are activated by segregating the positions of those neurons in brain space?
Now, we've known for over a century that each of the sensory modalities, that is vision, touch, taste, smell, and hearing, indeed are segregated in cortical space, as you see here. But not only are these modalities segregated in cortical space. That the individual sensory regions of the brain are subdivided further, such that the position of electrical activity in a given region provides information about the position of a stimulus in space and about the quality of a stimulus.
If you allow me to return to the bat, for example, and you look at the auditory, the hearing centers of the bat brain, what you observe is that one important component of an auditory stimulus of a sound is, of course, frequency. And different frequencies activate neurons along an axis of the auditory center of the bat brain. So that the quality of sound is determined by the position of neural stimulation along this axis.
You also will notice, interestingly, that this frequency map in the brain is not proportionate. There is a huge area of the brain that is tuned to frequencies which surround about 60 kilohertz. And it's that particular frequency that is the frequency of the echo that the bat uses for its bio sonar ability. So the brain, this disproportion occurs to meet the unique evolutionary requirements of the individual organisms. Allows an organism to evolve within an ecologic niche to accommodate its particular needs. And those particular needs differ in different organisms.
Now, what about smell? Indeed, very quickly, a series of genetic studies initiated by Peter Mombaerts and Fan Wang in the laboratory told us that, indeed, there is a spatial map in olfactory centers in the brain. We could genetically alter mice so that neurons that make one of the 1,000 receptors were blue, and we could follow those axons as they course back into the brain. And we observe that all the neurons that make receptor A all converge on a fixed point in the brain, and that point differs from the neurons that make receptor B.
Moreover, this map is relatively constant in all individuals in a species. So there is a spatial map, and the identification of an anatomic map immediately suggests a model for odor discrimination, which says that different odors will activate different combinations of receptors, which in turn will activate different combinations of discrete loci in the brain. So that cherry might activate this array within the brain and lemon a second non-identical array. The quality of an odor in such a model, the quality of any percept in such a model, would be encoded by different spatial patterns of activity in the brain.
Is the model true? Can we look at the function of the brain in response to different olfactory precepts? And do we see that different odors elicit different patterns of activity, and can we relate these patterns of activity to different behaviors?
To do this, we switched organisms. This is a mammal. This is the fruit fly drosophila. The fruit fly drosophila. Not a terribly attractive visage, is it? Does afford some advantages. That is, it is an organism that is genetically facile that exhibits an array of olfactory driven behaviors that are mediated by a nervous system that is 100,000 times numerically simpler than that of a mammal.
And so indeed, we were able to address these issues of brain activity during perception by first observing that the fly brain remarkably, despite the 600 million years of evolution that separate flies from man, exhibit the very same principles of organization and function in the peripheral olfactory system despite the fact that a fly uses an antenna and we use a nose. Many receptors. One receptor per neuron. Neurons bearing a given receptor all project to a fixed point in the brain. This is called the antennal lobe. It's functionally equivalent to our olfactory bulb.
So we could in the fly, by using rather sensitive imaging technologies, multi-photon imaging technologies that Susumu alluded to, simply ask the question, can we identify unique patterns of neural activity that are the signature for different odors? And indeed, this is precisely what two fellows in the lab have done.
What we're looking at is the antennal lobe of a fly brain in response to the major component of apple versus the major component of banana. And we can ask, are they unique signatures? What happens when we give apple?
That's strange. Let's look at what hap-- strange. This is the pattern we see. This is a movie, but it's not playing. And when we give banana.
Suffice it to say that indeed different odors elicit very different patterns of neural activity. And every odor give us a very different pattern of neural activity. And these patterns are constant in all individuals in a species. So they form a signature for an odor.
Now, we can relate individual patterns to individual behaviors. One of the most robust behaviors that a fly exhibits is rather interesting and evolutionarily significant, and that is, when you hurt flies, when you stress flies, they exude a pheromone. And this pheromone tells other flies to avoid that environment. And they do so robustly.
One of the major components, in an experiment with Seymour Benzer and David Anderson, of such an olfactory avoidance cue we found is carbon dioxide. When we give carbon dioxide to a fly and image its brain, what we see is yet a different pattern of activity. So indeed, different odors elicit different patterns of activity. And we can demonstrate that this pattern of activity is responsible for avoidance behavior. For if we genetically silence this locus of the neurons that go into this locus or the neurons that come out of this locus, we can eliminate avoidance behavior.
So what I've shown you is that different odors indeed elicit different patterns of activity. And that different patterns of activity can be related to specific behaviors. I can, with accuracy, look down on this pattern and discern what the fly has encountered in nature.
But I've accomplished this, and now you have accomplished this with your eyes. How does the fly do it? How does the fly know that this pattern is apple and this pattern is banana? The fly brain does not have eyes to look down on this pattern. How then does the fly reconstruct this deconstructed image of an apple?
This is a simple form of an important and elusive problem. It's called the binding problem. How bits of electrical activity are bound into a meaningful percept. Inherent in the binding problem is yet another problem, the parsing problem. That is, if I mix apple and banana, I will generate a new pattern of five loci of activity. Yet I can distinguish apple and banana within this mix. How is it that an organism's brain can pass out of this set of discrete loci the individual units that constitute apple and banana?
The binding problem is one that is shared by all sensory systems. Elegant work by [INAUDIBLE], by Semir Zeki, have shown that to obtain knowledge of what it is seeing, the visual brain, for example, does not merely passively represent images that are reflected on the retina. Rather, the brain must actively deconstruct and then reconstruct the visual world. A visual image is deconstructed first in the retina and ultimately passes via parallel processing pathways that report distinct components of an image. And this is shown schematically here.
So color is represented in visual brain in a region known as V4, whereas cells in V5 are responsive to the movement of an image. And an adjoining area, V3, is responsive to form. V3 and V5 are independent of the color of the stimulus. And lesions in V4 allow one to see an image rather well with clarity but only in shades of gray.
This segregation immediately poses a problem in the reconstruction or binding of a perceptual image. How is a spatial map in the brain, such that I've described to you, read? How are bits of electrical activity integrated to allow for meaningful recognition of the quality and nature of a sensory image?
I've already argued that sensory input, the bottom up process, is incomplete. It results in an incomplete and selective image of physical reality that differs in different individuals. The image is completed by the brain by a top down process that brings experience, expectation, emotion to the binding process, to how you view an image. If this is true, then perception, as originally suggested by Richard Gregory, is only a hypothesis. A best guess that never truly approaches reality. Our assumptions are based in part on input and in part from the brain's stored record, a record that is nourished by sensory experience.
Let's consider the problem of binding that Salvador Dali poses for us in the slave market with the bust of Voltaire. Now, the visual information that is entering our eyes is the same for everyone in the room. But some of you will immediately recognize the bust of Voltaire, whereas others will only see two women, two nuns. We are binding the very same sensory stimuli in different ways. Once having seen this image, however, on second view, binding will be instantaneous. Experience shapes the way you integrate incoming sensory information.
Now, if perception really is just an assumption, a best guess, then you can get it wrong. This triangle, for example, provides a classic example of illusory binding. I don't know whether you can see it in this LED. Do you see a triangle?
This is an image composed simply of three Pac-Man. But this is not at all what you are focusing on. You see a triangle with equilateral sides that are quite clear. But these sides do not exist. Your brain is trying to use its preconceived notion of the visual world to make what is not. And this is an illusion.
Consider a second example. What you see is our President, Mr. Bush, along with our Vice President, Mr. Cheney, right? Wrong. What you see is our President, Mr. Bush, and his face is inserted into the face of Mr. Cheney. You are looking, unfortunately, at two Mr. Bush's.
Look at the nose, the nose, the mouth, the eyes. This is not Cheney at all. This is Bush. This is an illusion. You are seeing according-- you are taking the information that is going in and adding to it expectation and creating a top down, bottom up binding, which is wrong.
Finally, if perception is indeed a hypothesis, it can be challenged. And that is precisely what Rene Magritte did. Magritte defines common sense, the common sense of our brain, deliberately and with great success. In this painting, "Carte Blanche," it is confounding. It goes against everything the brain has ever seen, learned, and stored in its memory. We can have no preconceived notion here, because the brain cannot represent this bizarre scene. It is an act of the imagination, and it fascinates precisely because our brains cannot find a solution. It is a classic example of non-binding.
So let me very quickly return to the biology of the binding problem. How is this deconstructed olfactory map reconstructing? One disturbingly seductive model argues that the combinatorial of signals that we've observed in this part of the olfactory system as highly insular segregated neural signals might come together and report to a locus in higher olfactory centers, which would then provide us with a refined olfactory image, the notion of a jasmine cell in olfaction, analogous to a grandmother cell in vision.
We've tried to look at this in our laboratory and ask, are these spatial maps reflected in higher olfactory centers? So here you see the map in the antennal lobe. Neurons expressing a receptor go into a discrete insular locus, which is segregated from all other loci. How then do we bring together the combinatorial of activation? What we observe is that in high brain in the old olfactory cortex, there is also a map. It is invariant in all individuals, and different loci report to different loci in high olfactory centers.
But the character of the map is different. The map is disperse, it is interdigitated, and it is integrative. So it affords the opportunity for integration, suggesting the possibility that, indeed, this olfactory center is integrating, reading the combinatorial of spatial information again in a spatially defined way.
Now, let me summarize, then, by arguing that the elucidation of an olfactory map, both in the olfactory bulb, the antennal lobe, or the analogous visual centers, brings us closer to an understanding, but it still leaves us with a profound and basic problem in perception. For no matter how high in the fly brain we map a sensory circuit, we must still ask, who in the brain is looking down on this image, on this spatial pattern? Though we may look at these odor evoked images with our brains and recognize a spatial pattern is unique, and we can even readily associate them with particular stimuli, but the brain, I repeat, does not have eyes.
Who reads the map? How are the spatially defined bits of information in the visual system, the olfactory system, decoded to allow for the perception of an image? So we're left with an old problem. The problem of the ghost in the machine. And I leave this task to the Ghostbusters, the neuroscientists of the Picower Center, Earl Miller, Matt Wilson, and Susumu Tonegawa. Thank you.