Neuroscientists around the world race towards their ultimate goal: a three-dimensional map of the brain. Harvard-Professor Jeff Lichtman is not just one of them, but a godfather of connectomics. An interview with someone who dedicated his life to a piece of cortex, a cube just one millimeter on a side.
[ZEISS Stories] Prof. Lichtman, in order to understand how nerve cells communicate with each other, you and your colleagues cut mouse brain into slices a thousand times thinner than a human hair. You scan them with electron microscopes, assemble them digitally, color each cell in order to visualize the network, spend months and months mapping them, just to realize one bold vision. A three-dimensional map of the brain. How did all that start?
[Jeff Lichtman] When I was a graduate student I worked on a very unimportant part of the nervous system of a rat and the way it controls saliva secretion in its mouth. I remember trying to explain to my parents what I was working on, but even then I wondered why would anyone care about this? I really wasn’t sure about it myself. But after a couple of years I discovered that in this region, baby neurons are much more interconnected than later due to massive pruning of synapses. It was exactly the opposite of what I would have imagined. My intuition was that adults should have this very complicated wiring diagram and babies should have a simple one. And frankly, that neural circuit reorganization is all I’ve been studying ever since.
[ZEISS Stories] What was the implication of this discovery?
[Jeff Lichtman] Maybe it tells us what learning is all about. As a baby we come into the world, wired up for every possibility and through experience we prune away only but a small set of those wires. Finally we become well trained in a sub-set of things with very strong connections. But once we go down one path, it gets harder to go down others because we eliminated connections. For example, it’s easy to learn a language as a child, but not so easy as an adult. The wiring diagram then is somehow the tell-tale result, the physical instantiation (to use a fancy word) of the experiences you have had. The fundamental question then is how experiences, that is, learned information, turns into wiring.
[ZEISS Stories] The National Geographic says: You’re trying to understand how the brain really works by understanding its wires. What is it that you do and do not understand?
[Jeff Lichtman] I should start by saying that I have banned the word understanding from my lab. We do not use it. That’s because we don’t believe that what we’re doing necessarily will provide us with understanding, as crazy as that sounds. People often assume that the purpose of all science is to understand something. But there is a kind of science with the purpose to describe something, to provide information about something you know nothing about. To get wiring diagrams of the brain is such science.
[ZEISS Stories] Isn’t that the first step in understanding – describing something, getting the data?
[Jeff Lichtman] One could think so. But there is some irony here. The more information you have, the harder it is to “understand” it in the traditional sense. For example, let’s say you want to understand New York City. One way to do that would be to map out all the streets, the buildings and the rooms, the location of every person, of every car, of every animal and so on. Would that complete description help you to understand it? Probably not. In fact, it would be harder than ever. Not only would the sheer mass of information be overwhelming. But New York City is an organic complicated mess where millions of actions happen simultaneously. To understand the whole thing would be impossible.
Now think of the brain as the same thing but magnified. It’s not a relatively flat structure like NYC but a volume in which information is flowing through nerve cells and blood supplies, where all sorts of information are coming in from the outside and getting processed. For the first time ever we will have data about all these streams of information in one piece of brain tissue. But anyone who thinks that this will suddenly end in an “Aha”-moment is I think fooling himself.
[ZEISS Stories] Then what are you striving for if not to understand the brain?
[Jeff Lichtman] A complete description. A description in which I suspect there will be many surprises. Ones that will tell us that what we thought we understood, in fact was a misunderstanding. The brain is the most complicated thing on the planet and we’re clarifying how far we are from actually making sense of it. That to me is a great thing about connectomics, our term for mapping the brain. It confronts you with a more accurate rendering of the world inside your head that challenges accepted dogma. It doesn’t replace one dogma with another, but it does say that many of the dogmatic statements neuroscientists believe are not exactly right.
[ZEISS Stories] Sounds like you are heading into trouble with fellow colleagues.
[Jeff Lichtman] Let’s put it this way: The scientific world view has to get to the point where there is appreciation for the idea that detailed description has value per se.
[ZEISS Stories] Till then, what’s the biggest challenge in publishing the map?
[Jeff Lichtman] If you write such describing papers, it’s not enough to say “here’s the data”. People would not know what they are supposed to do with it. You have to say “here’s the data and look at these things that you can see in this data that are incompatible with what you knew before.” Forcing people to start thinking about the brain in a different way is one challenge. Another is the digital challenge of sharing the data. To store the scans of just one cubic millimeter of a mouse brain you need about 2000 terabytes, or 2 petabytes of data capacity. That’s as much as 15,300 movies! That’s like publishing a book with one trillion pages. How would people read it? And how do you mine data that size? It’s an area of really interesting active questions.
[ZEISS Stories] You’re used to finding things out and discovering them. You had your first researching microscope as a third grader, looking into things from nearby pond water and then much later watching the development of neuromuscular junctions. This kind of work is more circumscribed. On the other hand, there seems to be a drawback with this project – mapping out the brain – that it will not be achieved in your lifetime.
[Jeff Lichtman] I’m resigned to the fact that the mapping of a whole human brain – requiring exabytes or even zettabytes of data – will perhapsnot happen in my lifetime. But I’m also resigned to the fact that it might never happen. It’s simply not where people would go. There’s an amount of data you need before you got the picture, but you don’t need everything. For example, if you are interested in categorizing beaches, you can find some with very fine sand, others with rougher sand or pebbles. But there is a certain number of beaches you can go to after which it is very surprising perhaps impossible to find one that does not fir into the categories you already have. Then you’re done. You don’t have to go to all the other beaches in the world because chances are you got 99.9999 percent categorized. It’s the same here. We should continue reconstructing brain tissue until it gets boring. And by boring I mean the variations, the ways things differ from one place to another, stop being surprising.
[ZEISS Stories] How much of a brain do you have to map to get “the picture”, as you said?
[Jeff Lichtman] I hope that long before actually mapping a whole brain we can have a sense of what we would get doing the whole thing. I don’t say we will understand why every wire is where it is, but we will come to a point where looking at another piece will not provide any new insights. And it might even be possible that a cubic millimeter might reach that plateau.
[ZEISS Stories] But all our brains are different. How could one cubic millimeter tell us enough about them?
[Jeff Lichtman] Even though every brain is different, they are all based on the same fundamental method of taking experiences and turning it into some kind of wiring diagram. Once you crack that code, it would just be variations of that theme. And my suspicion is, if you cracked it for visual information, that you stored as visual memory, you would also have cracked it for auditory memory, taste and scent etc. Once the information gets beyond your sense organs like the eyes or ears it’s just electrical signals running very much the same way, no matter what brain you’re in. So, all the encoding is probably based on the same principles.
[ZEISS Stories] What is it that drives you to achieve this once unattainable goal?
[Jeff Lichtman] It’s the thing that I thought about deeply when I was young beginning graduate student. It may have etched away from my brain all the other things I could have thought about (by pruning). So I’m just on this one path and it seems to me like the most important thing in the world is to figure out how each of us gets a behavioral repertoire and massive core of information that’s unique based entirely on our personal life experiences. That just seems really interesting to me!