 Hello, and welcome to the OIST podcast, bringing you the latest in science and tech from the Okinawa Institute of Science and Technology Graduate University. My name is Andrew Scott. In this episode, we are journeying to the smallest reaches of biology as we explore how individual cells are able to communicate with each other with an amazing degree of precision. Joining us on this journey is Dr. Thomas Kornberg, Professor of Biochemistry and Biophysics in the Cardiovascular Research Institute at the University of California, San Francisco. Dr. Kornberg's work focuses on cell-to-cell signalling and how these processes give rise to the creation of organs and body structure. Working primarily with Drosophila fruit flies, his discoveries have a surprising amount of applicability to other creatures too, including humans, with potentially huge effects. I chatted with Dr. Kornberg after his presidential lecture during his visit to OIST. Dr. Thomas Kornberg, thank you very much for speaking with us and welcome to the podcast. Imagine you are at a cocktail party and you have to introduce yourself and what you do to someone you've just met. How would you do that? So the questions that we focus on in the lab are how cells talk to each other. And this is the very basis of multicellularity. And basically, everyone at a cocktail party understands that there are specialized cells in the body called neurons, and neurons have this fantastic ability to reach out over very long distances to contact other cells and to communicate directly with those cells even at distances of meters. And we've all grown up with the idea that that feature of neuronal signalling is unique to neurons. I mean, neurons are uniquely endowed cells that do amazing things. The idea has always been that all the other cells do it differently. They communicate with, by other mechanisms. But what we've learned in our studies over the last couple decades is in fact that neuronal type signalling of reaching out over long distance, making contact, and communicating at short distance in a very highly specific and targeted way is in fact not unique to neurons. In fact, every cell type that we've looked at does that. And it's only with the new techniques that have emerged over the last decades, genetic techniques and histological techniques, new microscopes, new kinds of fluorescent probes that we've been able to see these things. They've always been there, but the ability to see them has been beyond our grasp. But with the new genomics, new microscopes, and new proteins, kinds of fluorescent proteins that are being invented daily, we can now begin to see these other types of structures which have long been invisible. But the basic take home message is that all cells, we can see, communicate at short distance and long by reaching out and touching. You touched upon like there are various methods by which cells can connect to one another, neurons perhaps being the ones that people are most familiar with. Can you tell me a little bit about the things that you're studying? We came up with a name for these, which I thought was important to do, given that we assumed that they were doing something which was unique. Looking back on it 20 years later, we realize that probably what they're doing is not unique. But these kinds of structures do exist, and there are many cells. So we called them, based upon their appearance when we first saw them, cytonemes, nemes being fingers or thread, and cytobene, they clearly had cytoplasm in them. Well, we didn't, and we knew that they were an organelle generally known as phillipodia. It's been observed for many decades, 100 years. But what we were suggesting or what our data indicated was these were phillipodia of a special type that were involved in signaling, as opposed to other types of phillipodia, which when we looked in the cell biology textbooks said, what was the role of phillipodia? Well, the role of phillipodia was to involve in cell movement and force generation. So we said, OK, so these are phillipodia, which are doing something else. They're involved not in force generation, not in cell movement. So force generation meaning like the mechanisms that allow cells to move around. Correct. They basically reach out and pull a cell from one place to another. And they do that. Looking back on it now 20 years later, I have to think, are there phillipodia which are not involved in signaling? And I don't know that there are not. So we call them cytonemes. They've now been, these kinds of structures have been observed in many other contexts at this point, in research in other labs over the past 20 years. And people give them various names. I think they're basically all doing the same thing, which is allowing cells to reach out and scout out, investigate their environment. And when they make appropriate contact, and what we've shown is when they do make appropriate contact, they make chemical synapses, organized structures that allow exchange of information. And that's basically how cells know where they are relative to their far and near neighbors. So you mentioned briefly that this has quite a long history, like people are being observing the structures, not knowing quite what they were for about 100 years or so. Can you give me a little summary of how this history began? I'm thinking of like the Ethel Brown Hydra experiment. So the experiment that you're referring to was an experiment done by a graduate student at Columbia University on the first decade of the 20th century. And she was looking at an organism called Hydra, which is still worked on today. And she was doing the kind of experimental embryology, which was available to people in those days. They didn't have genetics. They didn't have proteins. They didn't have the kind of microscopes that we have today. But what they could do is very skillfully take out little pieces of the animal and transplant it and put it in other places and ask, what happens? What they observed is that if they transplanted a little piece of Hydra from one place to another, that occasionally they could see that not only did the expected structure appear, that is, if they took it from a place in the animal, it was going to make a tentacle, that when they transplanted it to another animal, they saw a tentacle appear. That's fine. Those cells remembered, even after this gross insult of dissection and transplantation, but occasionally they saw that the surrounding tissue into which that material had been placed also were tentacles. In other words, the host tissue, which would normally never grow a tentacle, did because it was influenced by these new neighbors that it had, suggesting that cells have the ability to instruct their neighbors. Now, that basic idea then was beautifully and powerfully developed by this German group of Speymann, who showed that he could basically reproduce that phenomenon in amphibian embryos and there show that he could create basically whole new animals growing out from these areas around transplanted tissue. He was able to identify specific regions of an amphibian embryo that had this capacity to induce the neighbors to create new structures. The important issue there was it wasn't just creating a monster, it wasn't creating abnormal structures. It was creating normal structures, but in an abnormal place. That's a normal developmental procedure processes were ongoing, but it's going on in the wrong place as a function of instructions received from these specialized cells, which were called organizing centers or signaling centers. This idea has a long history and really is the foundation for how we think about how organs and tissues are organized. That is what I like to say is development is not a democracy. All cells are not equal. There are some specialized cells, which we call centers or organizers, which instruct their neighbors in a developmental field what to do. We take that as the conceptual basis for organogenesis and defining morphology. The question that we address is how these substances, which are instructive, how they work and how they distribute over space and time. Thinking about the cytonyms connecting one cell to another, and you mentioned that they're exchanging information. What kind of information have you observed them swapping hand? What form does that take? It was assumed back to the days of Speymann in the 1920s that there were substances, chemical substances, which would induce these responses. It was assumed, and that went on for more than 50 years, that these substances would be small molecules, small organic molecules on the order of size of ATP, a couple hundred molecular weight, and they would be free to diffuse in and out of cells so they could move across the tissue by just unimpeded by cell membranes. Wonderful theories of how they might work were developed for such small molecule, small diffusible molecules. But then, one of the triumphs of Drosophila developmental genetics was the discovery these inducers are not small molecules, they are actually proteins. Proteins do not easily diffuse in and out of cells. Proteins have to move, have to be exported from cells, they have to be taken up by cells, and so these distributions of these informational packets which are encoded within these signaling proteins have to be created. And what was now established is that there are cells, groups of cells that make these signaling proteins. Those are the signaling centers. Those are the signaling centers that were discovered way back in the 1920s. They make these instructional inducers, these proteins. And then what we know is that those proteins can actually move long distances. And so what we've shown is, by characterizing these cytonyms, is that these are the structures, the conduits that move these proteins over long distances. So because we're dealing with an audio medium here, I'd like to kind of take the listeners on this sort of journey. So if we were to shrink ourselves down and put ourselves down to the cellular level inside a drosophila, and we were to observe these structures, what would we see? A textbook drawing, image of a cell has a nucleus inside, and it has cytoplasm surrounding that nucleus, and then a cell membrane. And that membrane forms an even, smooth barrier that separates the internal contents of the cell, the cytoplasm and the nucleus, from the external environment. That's the image of a cell that we have. But in fact, cell membranes do not have smooth external surfaces. And so what we know today, now that we have much better and more sensitive ways of looking, so that we can create fluorescent proteins that embed in the membrane and light up the membrane. And if we, under the appropriate conditions, when we look at the distribution of those fluorescent proteins, they're in the plasma membrane, but the plasma membrane is not a flat surface. It's populated by many protrusions and extensions. And what we observe is that the cells are actually quite hairy, if you will. But those hairs are not constant. They extend out. They act. And it's a very, very active process. So the cell seems to be exploring the world around it in a very active way by extending out these little fingers of membrane inside the cytoplasm, and that they're actively scouting out, we think, what's around the cell. So these fingers themselves are the cytonemes? And some of those are certainly cytonemes. I won't say that all of them are. And one of the, we know so little really about the cell biology or the biology of these structures that the mantra that we have in the lab, don't make rules. Don't assume, don't circumscribe what we see. Look at what we see, describe it, and hopefully unencumbered by the prejudice of dogma. So you mentioned earlier when you lectured that these cytonemes, these finger-like projections coming from cells are not localized. They're not limited to the cells in their immediate vicinity, but they can actually stretch an incredibly long way. How long relative to maybe cell size are we talking about? How long of a distance can these transfers take place on? Well, again, I'm not going to make a rule. Okay. So one flippant answer to that is that we, as developmental biologists, can define what we call as a developmental field. By that, what we mean is that we know where these signaling centers are, and we know how far their influence extends. That's our field. Okay. And in different contexts, that field can vary enormously. In the context that we study, we know that these substances can move and the cytonemes can extend on the distances on the order of 100, 125 microns. So to put that into context, the diameter of an individual cell in our system is around two to two and a half microns. So we're talking 40 to 50 cells. In other contexts where they've been observed and characterized such as the growing limb bud of a mouse in a mouse embryo, they can extend for hundreds of microns. So if we scaled that up to say this, if our cell was the size of a car, we'd be talking like hundreds of cars lengths. So that's quite an achievement for them to be able to do this. It is. And remember that they are extremely active. They extend out and then they retract. That's what we see. So it's a very active dynamic process. But after all, so is development. Cells are growing, shapes are changing. And if this is the means by which cells understand where they are in an organism, in a tissue, in an organ, it's changing shape in real time. That's what you need. That's what you have to have. So you were with fruit flies exploring this particular phenomena. What does their activity look like for the development of a fruit fly? So in fairness, we've focused on a particular organ at a particular time in development. We've not characterized all of development. We've not characterized all tissues in the fly. The fly has many. But what I can say is that the processes that we've studied and for which we can now make some pretty definitive conclusions about what these structures are doing. They are involved in the exchange and the transfer and the transport of the signal proteins in space and time. That biology is conserved. It's conserved at different times in Drosophila development. It's conserved between different tissues in Drosophila. And it's conserved across virtually every animal that we know. So many, many years ago, when I was a student, these signaling proteins had not yet been identified. We really didn't have a clue what embedded this kind of information. And the founder of this institute, Sydney Brenner, who was himself involved in the elucidation of the genetic code and the understanding of how the information and genetic code was elaborated and used. And so was intimately involved in coming to understand that the genetic code is universal in bacteria, in plants, in animals. And it's that universality that's really given birth to the back technology revolution because although many people doubted it, in fact, but because of the universality of the genetic code, human proteins can be produced in bacteria such as insulin and growth hormone and that protein is the same protein that it would be produced in a human cell because the genetic code and these basic biochemistry is universal. But if we go back 40 years, or our understanding of how these developmental processes worked, we understood at that point that the genetic code was universal. We had no idea about how different organs took on different shapes and functions and how different cell types appeared in one place or another. And so Sydney Brenner stated the dilemma, the conundrum, in this way. He said, you know what, what really worried him was that the genetic code would be the last elegant solution in nature. That is, every animal, every organ, every tissue of different times would sort of piece together and cobble together different ways of doing something to create something. And now 40 years later what we know is that there is a beautiful, elegant solution to this kind of developmental patterning and programming. That is, we now know the vocabulary of signaling proteins that's responsible for regulating growth, patterning, and morphogenesis. And that vocabulary is incredibly small. It consists of proteins which have these wonderful names, hedgehog, bone morphogenetic protein, wingless or wint, fibroblast growth factor, epidermal growth factor, notch. These come from phenotypes mostly in flies. But what we know now is that same vocabulary of signaling proteins is used not just in the fly wing or the organs that we study, not just in marvel development at the stage we study, but it's also that those same proteins are responsible for organizing embryogenesis, tissue types in an embryo all the way to the adult. And it's not limited just to the fly. Virtually every animal that's been examined, it's that same vocabulary. So now that we know the vocabulary, now we have to understand the grammar and the language and the syntax. And so that's what most of us are doing. What sparked your personal interest to pursue this field? Was there an inciting incident or was it something that kind of built up over time? It built up over time and the questions changed. So when I got into developmental biology, I would argue that the developmental biology at that time was a matter of semantics. And it consisted basically of observation and descriptions of what was being observed. There was very little functional studies that could be done. And so people who were studying Drosophila or fly development described what they saw and they created a language to describe them. People studying chick development created their own language. Mouse development created their own language. Every animal had its own language. And at that time, as I said, we didn't know if there was any correlation or any elegance to development. It was appropriate to do that. But now 40 years later, we know that the basic processes are all the same. And in fact, there needs to be a common language to describe these developmental processes and all these different animals. So at the time I got into the field, it was sort of pretty dirty. And there wasn't much that we could extract other than describing what was going on. But what attracted me to a particular question in Drosophila was a discovery by a couple Spanish developmental geneticists, Antonio Garcia-Balito, Enes Morata and Pedro Rapal, and a British developmental biologist, Peter Lawrence. And what they showed was that the fly wing was basically subdivided into two parts. It was a binary decision. And there was a gene called engrailed that was needed in one part and not the other. And if you lacked engrailed function, those two parts didn't remain separate, but they got intermingled in the cells for overgrew. So that attracted me because it was a very clear mechanism for establishing the difference between these two parts. And so my initial interest was in understanding how that boundary between these two cell populations was established. So for many years, we studied the genetics to understand how that gene worked. And so the juxtaposition of these two parts of the wing is called a border, a compartment border, and these two parts were called compartments. And the phenotype, what happened if that border was abnormal, is that the cells no longer respected that border and overdrew. So in a sense, it was a model for cancer. Did I think that that would have any relevance whatsoever to human cancer? OK, so these genetic studies allowed us to definitively show the requirement for this gene in only one compartment of the wing. The question arose as we then determined that this gene encoded a protein, which is a transcription factor, which regulates the expression of other genes. Then we wanted to know, OK, what is it regulating to carry out this role? So among the genes that we identified that it regulates was a gene which we cloned and characterized called hedgehog. And we showed that hedgehog expression is regulated by Engrail and hedgehog carries out this amazing role of moving from the cells where it's made to regulate the cells across the compartment border. And in the absence of or with abnormal hedgehog function, cells didn't know what to do and they overdrew. It's consistent with this tumor model, cancer model. Again, did I think that would have any relevance to human tumors? I didn't think so. But others have subsequently shown that hedgehog is not simply a doosophila protein, but there are vertebrate homologs of hedgehog, which are called sonic hedgehog and Indian hedgehog. And that that pathway that we and others helped delineate of how hedgehog works is functioning not only in doosophila, but in most human cells. And a barren hedgehog signaling is responsible for the most common form of human cancer. There are drugs now in the clinic directed against the hedgehog pathway to treat patients that have defects in hedgehog signaling. So there is a direct parallel from very basic studies of fly wing development all the way to human disease. It's really quite amazing that that genetic structure has been preserved through such a massive phylogical jump. Exactly. But it's not true just of hedgehog. It's true of BMP. It's true of wingless, wind. It's true of FGF, it's EGF. So these are the fundamental systems which are responsible for defining space and pattern in animals. So then having established that the important function of these signaling proteins is to move in space, to move from one cell to another, then the next question became, okay, how do they do it? And that's where informed by the biology that we and others had established, we were prepared to interpret or to take seriously a chance observation when we saw these structures which we now call cytonyms that were present in an oriented distribution in these tissues. And that's within what we've been studying for the past 20 years. So from the point that you are now, and if you'll indulge a little bit of speculation, is there a pathway that we can follow from your observations to a presumed treatment for tumors in the future? Absolutely. Work that I described in the seminar, we've shown that the tumor types that require interaction with stromal cells, and there are many such types of tumors. That is, tumors are composed not just of the tumor cells, but also the normal cells around them, and they talk to each other and cooperate with each other. And that cooperation is responsible for neoplasia, the overgrowth of tumors, as well probably as leading to metastasis. So it's very important that cell-cell communication between tumor cells and stromal cells. So our work in the fly on model systems of human tumors has taught us that these tumor stromal cell communication is cytonym-mediated. And if we block cytonym-mediated communication between tumor and stromal cells, tumors don't grow. So we literally can cure lethal flies of lethal tumors by inhibiting the ability of the tumors to communicate by cytonyms with these stromal cells. So do I think, in my wildest imagination, that this then could identify for us important new targets for cancer therapy in humans? Absolutely. It's a wonderful opportunity. This could be very big indeed. Outside of your scientific work, you're also quite noted as a very capable cellist. I just wanted to ask if there's any kind of overlap perhaps in the way you approach science to the way you play music or when you're conducting research, is there any particular kind of music that plays in your head? All those are interesting questions. I'm afraid I don't have a good answer for you. If I rephrase the question a bit differently to say, is there any value to musical training for a scientist? I would say absolutely. I think there's a tremendous value for musical training actually for any vocation. And I'll illustrate that in the following way. So when I graduated from undergraduate school, I had not done the typical course. That is, I did take a course in chemistry. I did take a course in physics and biology. But I opted out of every laboratory course. So I had never set foot in a laboratory. And I did that because at the same time I was an undergraduate, I was also a student in the Julliard Music Conservatory. And I explained to my professors, I didn't have time to spend hours and hours every afternoon in the lab. And they were all very accommodating. But then when I finished undergraduate, I did walk into a laboratory prior to my entering graduate school. What I discovered is that in contrast to other beginning students, when I set up an experiment, I didn't necessarily expect it to work. If it didn't work, I set it up again and again and again until it did. And after all, that's the training you get as a musician. You don't expect to be able to play a concerto the first time you sit down and read the music. It's a long process to develop the ability to play. And you get constant feedback, constant instruction down to minutia of how to sit, how to hold your hands, what to do and every note and so forth. So you become accustomed to being directed and to working for the long term to get something to actually work and something to develop that is good. And students in other disciplines don't get that. They don't get a personal attention to detail or the idea of working through difficulties to actually achieve something. So when I walked into the laboratory and what I observed is I would set up an experiment again and again and again until it got to work. Some of my classmates or the others were entering the lab if the experiment didn't work, they were out. They came back the next day or several days later because they got frustrated. So I think the training one gets as a musician in terms of focus, detail, working for the long term, taking direction is the best education one can get for any field. And that's what I said to my children when they were growing up when they said although they played instruments and they took music lessons they said we are not going to become professional musicians so we are not going to be that serious but I said that's fine. But the training that you will get will do you well, will be good for whatever profession you do decide to take. Is there a question that you've always wanted to be asked but haven't and if so what's the answer? Or is a question you wish somebody else would ask of science? I'll give an answer which is sort of oblique to that which is to say that how lucky and fortunate I am to be able to devote my career to asking questions that society would support that. I can walk into my lab and ask any question I want. Nobody is going to tell me what I should do. And what a treat, what a rare luxury that is to have that freedom and of course the satisfaction and the joy of getting unexpected answers and to understanding something in a way that you never thought about it before. So I feel very fortunate to be able to have been able to do that. Science is wonderful but it is not universal and so I'm fortunate to be here living in a time when I could take advantage of that. Thomas Kornberg, thank you very much. Thanks for listening to this podcast. It was recorded and edited by me, Andrew Scott. Special thanks to Dr. Thomas Kornberg. If you enjoyed this episode, subscribe to get more as soon as we release them. And we always love to read your reviews so why not let the world know what you think of the show? You can also find us on Facebook and Twitter or send us an email to media.oist.jp Thanks again for listening. See you next time.