Hi, I’m Susan Lindquist. I’m at the Whitehead Institute at MIT and a member of the Howard Hughes Medical Institute. I’m here to tell you about protein folding as a powerful driver of evolutionary novelty. Last time, I talked to you about Hsp90 and the ways in which it can influence protein folding and the manifestation of genetic variation in very powerful ways. Today, I’m going to tell you about a very different way in which protein folding can influence the manifestation of genetic variation and lead to the appearance of all kinds of new traits. And that’s the prions. So, it’s an interesting story that starts in a place in New Guinea and will move on to Cambridge, Massachusetts. So, the protein folding problem, which drives all of this that I’m going to be talking to you about, is simply that proteins start out as long linear strings of amino acids and they have to fold into very complicated shapes like this in order to function, and they do that in a crazy environment. They do that in this really crowded environment of the cell where proteins are jostling around and bumping into each other all the time. The story on prions, as I mentioned, starts out in New Guinea, and there was a tribe where a very large number of people were dying from a very bizarre, horrible neurodegenerative disease, and Carleton Gajdusek found out that it was in fact due to an infectious agent. And Stan Prusiner — and many other investigators have contributed to this, I should say, but these guys were very, very, very important, pivotal in the field — found out that this disease was caused by protein folding. Now, the protein folding problem creating an infectious element was really quite an amazing thing, and it meant that proteins, when they change their folds, can have genetic manifestations… disease agents are usually agents that carry along DNA with them, so it was quite a revelation. But what I want to talk to you about is thinking about prions like this. Because the story of prions as disease-causing agents is so extraordinary and wonderful and amazing. It’s kind of dominated, and quite naturally so, the concept of prion biology. But what I think is that we have to think about prions as this creature, and it’s time to get rid of some of the baggage, because it’s my belief that prions actually are amazing protein-based genetic elements that can do all kinds of really wonderful things in biological systems. And I think also that we’re only at the very tip of the iceberg in revealing this. So, here’s some of the great things about prions that I want to tell you about. Prions form a mechanism for the inheritance of a protein-based trait. We’ve found more than 50 of them in yeast, we and other people have reported on these. 25 of them are as yet unpublished, but there’s just a lot of them. They cause dozens, a bewildering variety of all sorts of new traits. They allow organisms, the yeast organism, at least, that we’ve been looking at, to survive in fluctuating environments. They provide a fast route to the evolution of complexity and they’re reversible traits, so the organism can acquire these traits and they can lose them. They form a very sophisticated bet-hedging strategy for survival. And they form a system of biological memory, in which the change in protein folding self-perpetuates and that self-perpetuating change in protein folding changes the function of the protein, and that can both create new phenotypes and it actually serves as a kind of a biological memory, when you think about it, and there’s evidence now that this is really occurring at the ends of synapses in our brain and helping to maintain long-term synaptic connections. Also, we have found, in my lab, evidence that prions can enable life in complex biological systems and communities, creating different ways for organisms to relate to each other. So I’m going to try to give you an overview of this very complex and rich and wonderful subject. And I’m going to end with prions forming, really, the ultimate example of Lamarckian evolution, in which organisms can acquire a heritable new trait that they pass on from generation to generation by being exposed to a new environment. So, the prion story starts with these two strains of yeast, red versus white, and it turns out that these organisms have a very odd genetic behavior. Brian Cox did some wonderful work on this many years ago, and a lot of people thought, well, this is really kind of a strange thing, why is he working on this?… but his early work really laid the foundation for all of this, everything I’m going to be talking to you about today. What he found was the red strain was really very stable and you could streak it out and it would give rise to more red colonies, and every once in a while, they would turn white. And the white strain you could streak out and they were very stable, a very heritable trait, and every once in a while, they would turn red, and back and worth. So it was metastable inheritance. What Brian also found out is that when you mated the red cells to the white cells and then you sporulated and got out the genetic progeny, you would normally expect, if those traits were due to changes in the DNA, if the difference between the red and white cells was due to changes in the DNA, that some half of the progeny would now be red and half of the progeny would now be white, as those pieces of DNA reassorted in subsequent generations. And they weren’t. All the cells, all the progeny were white. So, an odd inheritance. Brian also realized that this change in color was due to a change in translation, the way in which messenger RNAs are decoded into proteins, that there was a translation termination defect, and the cells switched from red to white because of a change in ribosome readthrough of stop codons. And that was linked to this protein here, in this simple cartoon. It’s a protein that’s involved in translation termination, and it’s only this portion of the protein over here that’s actually required for translation termination activity. It’s a very important factor. It tells ribosomes to stop when they hit a stop codon. Attached to it are these two weird domains, called the N-terminal domain and the middle domain, that just have a very unusual type of amino acid composition. And it was discovered that those properties of inheritance really depended upon that… that is, those odd properties of inheritance depended upon that region of the protein. Now, Yury Chernoff comes into this story because he was searching for… what could govern this odd pattern of genetic behavior? It’s not like the genetics that we normally think about as being driven by changes in DNA. And he was working with Sue Liebman’s laboratory, and he did a screen for factors in the genome of yeast that could alter or influence this inheritance pattern. And what he came up with was Hsp104. But what the heck did that mean? Well, at that point, I got a phone call from Yury, because he knew that I had been working on Hsp104. My laboratory had shown that it saves cells from high-temperature death, and he was wondering if I knew any aspect of how it worked. And I did, I knew how Hsp104 worked, but nobody else did, and the reason for that was because we couldn’t get our paper published. Hsp104 did something unusual and it was very hard for people to accept what it did. So, what does Hsp104 do? And by the way, these are many, many, many people who contributed to this story of trying to figure out these different parts of this translation termination factor and which ones contributed to the genetic behavior. So, what does Hsp104 do? Hsp104 plays a major role in something called induced thermotolerance. I talked about this briefly in one of my earlier lectures. When organisms are exposed to mild heat temperatures, they make new proteins called heat shock proteins, and those proteins help to save them from the death caused by protein misfolding at high temperatures. Here’s an experiment in which the cells have been exposed to the same heat treatments, but in one case the cells have Hsp104 and in the other case the cells do not have Hsp104. It makes a big difference to their ability to survive, and the way it makes a difference in their ability to survive is that it takes apart protein aggregates. So, what are protein aggregates? These are well-folded proteins. You’ve all seen the effects of heat on these well-folded proteins. It causes proteins to aggregate. Now, Hsp104 I don’t want to say can unfry that egg, but just a little bit of that kind of aggregation occurs in cells in response to stresses, and that can kill them. Hsp104 saves them by disaggregating it. So, that led us to think, well maybe, how could it be involved in inheritance of this trait? Maybe the inheritance of that trait depended in some way upon protein aggregation phenomena? So we looked at the protein that had been tied to that trait, that translation termination factor, and asked whether it existed in a different conformational state in the red cells or the white cells. And sure enough, in the red cells the protein was soluble and functional, and in the white cells the protein was tied up in little aggregates. And you can see by the way… these little aggregates, they’re actually being inherited, they’re being passed from the mother cell into the daughter cell. Now, about the time when we were in the midst of these experiments, and we hadn’t yet published anything and we hadn’t yet been able to publish even our story on Hsp104, although we did wind up getting a really great paper and the suggestions of the reviewers, in the long run, really did make a better paper… while we were in the midst of this, along came a paper by Reed Wickner, which was a very, very clever interpretation of some genetic experiments on another factor that was inherited in a very bizarre way, inherited and had the same kind of genetic properties that the PSI element had, that that red/white element had. This was called URE3, and he suggested that the way in which this was inherited was that it was due to some kind of a prion-like phenomenon, a self-perpetuating change in protein function. Not knowing whether it was an aggregate, or whether it was a change in the activity of the protein, or what was going on, but he also suggested that this might apply to… this prion-like mechanism might apply to that inheritance of that red/white trait. So, as I say, it could have been anything, it could have been an enzyme that catalyzed its own modification, it could have been lots of different things, but because we knew that it was controlled by Hsp104, which we knew was involved in protein aggregation, this paper really made a tremendous amount of sense. And so we decided to look at this in even greater detail, and we found that, in fact, the aggregation state of that protein, Sup35, that translation termination factor, was not just a generalized aggregate, but a very special kind of aggregate, a self-templating amyloid aggregate. So, on the left there you see the protein fibers of Sup35 that we found, after a great deal of effort. Aggregated proteins are a lot of work to work with, but we eventually looked at the aggregates under the electron microscope and, lo and behold, they’re not just a mess of aggregates, they were very highly organized amyloid filaments. And they were organized in a very interesting way, because the central spine of the filament was that region of the protein that had previously been shown to be essential for the inheritance of the white trait, and the functional part of the protein, that’s normally functioning in translation, is stuck out on the outside, so when this protein gets assembled into this amyloid fiber, it can no longer get to the ribosome and is no longer functioning. And we were then able to actually monitor the assembly kinetics of this protein, and we found that the protein initially assembles in a test tube very, very inefficiently and very, very slowly. It took hours and hours and hours for the protein to assemble… but, the key thing that I think explains the ability of this protein to serve as an element of genetic inheritance is that if you take a very small amount of this preassembled protein and add it to the start of a new assembly reaction, this one right here, what you see is that… boom… the preassembled fibers have the capacity to completely, very, very rapidly convert all the other protein to that same state. So, now you can begin to see how this could explain the inheritance of this trait, because if you have a soluble protein that’s functional and the cells are red, and an insoluble protein that’s not functional, so that the cells are white. You mate those cells together and then you sporulate and segregate the genomes, well, the protein has served as a template to change the protein in the other cell type, and so that when the cells sporulate, all of the progeny are going to be white, as Brian Cox had shown. And it doesn’t segregate with the DNA, it doesn’t matter which cells get which chromosome, because the trait is not based upon the inheritance of a DNA change, it’s based upon the inheritance of a protein with an altered conformational state. And how Hsp104 figures in this is that Hsp104 controls that amyloid state. I told you that it saves cells from heat shock by disaggregating proteins; it also can disaggregate those amyloids. So… and in fact it can cut them into little bits and pieces. So, here are fibers of that protein, amyloid fibers, very tough, difficult to dissolve biochemically — you have to go through incredible lengths to get them to dissolve — but here’s what Hsp104 does. It chops them into little tiny pieces, and that allows them to be inherited, and to form the template that goes into the daughter cell and allows the daughter cell’s proteins to change. So, the whole thing, putting that together, then, looks like this. You have red cells that are carrying a particular gene, and when ribosomes do what they’re supposed to do, the translation termination factor is in its soluble state, it tells the ribosomes to stop when the ribosomes see the stop codon. But that protein can assemble into this self-perpetuating amyloid, and when it does there’s not very much of the translation termination factor around, and so quite a few of the ribosomes wind up reading through that stop codon, ignoring it. That changes the cells from having a red pigment to having a white pigment. Now, the cells turn white, and the next key, the inheritance of this, is that the self-templating amyloid protein is actually cut by Hsp104 to allow its orderly segregation into daughter cells for inheritance. And in fact with Helen Saibil, we overexpressed this protein to the point where it made massive assemblies, amyloid assemblies in the cell, and used some very, very sophisticated imaging techniques, like EM tomography, this was all done by Helen by the way, to suggest that, just like polytene chromosomes — larger assemblies of polytene chromosomes gave us our first view of the organization of DNA in chromosomes — we think these larger assemblies are giving us the first view of the way in which they, actually a rather sophisticated mitotic apparatus that breaks these fibers apart, is working. But in any case, Hsp104 is key to it, but it’s not the only key to it. There are multiple other proteins that are involved in helping these amyloid fibers to be partitioned to daughter cells in a very orderly way, guaranteeing the inheritance of the trait that’s caused when this protein changes its conformational state. So this provides a completely coherent biochemical explanation for this phenomenon. Is it really responsible, is it the only thing that’s going on? Well, we did a lot of things. We had mutations in the prion that didn’t assemble very well, and they could not inherit the trait. And we had ones that caused it to assemble more and they were more likely to inherit the trait, we did all sorts of things, but let me just tell you about one line of evidence that was particularly fun. We figured that, well, if this really was responsible for that new trait, this protein assembly process, then we should be able to use that knowledge to create a new prion, all of our own. And so what we did was to take the Sup35 translation termination factor and in the genome we deleted the prion element from that strain, because we didn’t want it to interfere with our new prions that we were going to be making. And then we took glucocorticoid receptor, which is a steroid hormone receptor from rat, and asked whether or not we could monitor its activity in a yeast cell, and see a change in its activity state when we fused it to that domain which we now call the prion domain of this protein, which is responsible for this bistable state. And sure enough… we had a reporter in there that, when the glucocorticoid receptor was working, it would turn the cells the blue. When the glucocorticoid receptor was assembled into a self-perpetuating aggregated template, the cells became white, because it was no longer working and, moreover, they passed that white state on to their progeny in a very, very stable way. So, they could either pass on the blue state or they could pass on the white state. So, the Weissman lab did another wonderful experiment. They took the cell walls off of yeast cells and they did a protein-only transformation, assembling the protein into those fibers that I told you that we made earlier, they made them under a couple of different conditions, and they used those fibers, they stuck them into the yeast cells, and they were able to show that the cells turned from red to white in a heritable way. So, that protein-only transformation really was another nail establishing the coffin, establishing that this genetic trait was due to the inheritance of a protein with an altered conformation. Now, here we have this prion protein. As I mentioned, it’s a translation termination factor, a really important factor in the cell that determines when ribosomes will interpret stop codons properly, and it’s conserved in all eukaryotes, and here we have this domain stuck onto the end of it that has been conserved, by the way, for 800 million years of evolution, and its regulation by Hsp104 has been conserved for 800 million years of evolution. And so the question becomes, why? Why in heaven’s name would cells allow this translation termination factor to suddenly be sucked out of solution so that ribosomes aren’t performing properly? And it occurred to us that this had to have some meaning, because otherwise, yeast can very, very rapidly evolve, and they could have acquired mutations in that prion domain that would have still allowed the protein to have its translation termination function, but not allowed it to be sucked up into these aggregates. So, one thing that occurred to us was that the readthrough of ribosomes might be occurring not just on that one messenger RNA that Brian Cox had first discovered, but actually it might be happening on messenger RNAs that were coming from all over the genome, and in that case we might expect to see some really new and interesting traits if we started growing strains in different conditions, not on rich media, in the laboratory. So, here’s an example of a really wonderful trait that’s caused by the appearance of this same prion. You can see that the colony morphology of these cells has changed completely. Cells that normally look like this and create colonies like this created very, very different types of colonies that adhere to each other, the cells adhere to each other, they stick to each other, they have different abilities to grow in different environments. And here’s an example of the fact that these prions actually spontaneously appeared every once in a while. And we tried growing cells under all sorts of different conditions, kind of like the story I told you about Hsp90 a little while ago, and we found that in different strains we got completely different traits. And that makes sense when you realize that the regions that are downstream of stop codons are not normally under much selective pressure, they’re free to vary quite a bit, and so changes in those downstream sequences might be expected to accumulate over the course of evolution, and then we the cell switches into the prion state by, perhaps, sheer happenstance, that will cause the creation of many new phenotypes. So, you get lots of different traits in lots of different strains, and the traits of different strains were completely dependent upon the genotype of that strain. So, what we think is that this prion allows cells to sample genetic variation, kind of at a genome-wide scale, and it allows them to acquire some really complicated traits that it would be very hard for them to acquire by individual mutation for example, of a stop codon, that would cause one individual messenger RNA to be readthrough. Rather, multiple messages being readthrough at the same time could create quite complicated traits. So, if this mechanism has really existed, this prion has existed, in order to create evolutionary novelty and the ability of the cells to sometimes be able to exploit new environments, then evolutionary biologists and we ourselves realized that if this was how this prion was serving such a purpose in evolution, then switching should increase with stress. That is, it should be tuned such that under stressful environments the cells would be more likely to switch into the prion state, because they would be more likely to need novel new phenotypes under stress. And so we asked whether or not it does, and the answer to that was yes. And then the question of course becomes how, but for us there’s a pretty simple, logical explanation for that, and that is because of how stress influences the protein folding and protein homeostasis pathways of the cell. So, stress increases the likelihood that proteins will misfold, making it more likely that those prion domains will now suddenly acquire that aggregated amyloid state and perpetuate that state to their daughter cells. Stress also increases the rate at which cells lose prions, because the stresses cause induction of chaperone proteins, and induction of protein degradation mechanisms, and all sorts of other things that influence protein homeostasis, so these prions, which normally appear only relatively rarely, would, just because of the very nature of protein homeostasis and the way in which stress influences protein homeostasis, be more likely to appear and disappear under conditions when the cells might need new phenotypes. So, here’s the way we think it’s working: cells switch into the prion state just every once in a while, and in fact with PSI, the prion that I’ve just been talking to you about, that translation termination factor, that happens about one in a million cells. And generally one would expect that when the translation termination activity isn’t working well and ribosomes are reading through stop codons, it’s generally not going to create a good trait, and so that one in a million cell might die — not a big loss to the population. But you can imagine that, if the environment changes, and that’s what we found, is that in different environments, prions could sometimes give cells an ability to survive conditions that they otherwise could not possibly survive, then the cells would now be able to live in an environment where they otherwise wouldn’t be able to live. They would proliferate, the non-prion cells might disappear, and this allows this genome, the prion would allow this genome to survive under conditions when those cells would otherwise not be able to survive, because the formation of that prion has allowed all kinds of new genetic variation to be exploited, some of which is beneficial. Now, the cool thing is that because this is all tied to protein homeostasis in the cell, under conditions when the environments have changed where cells are not doin’ so well, and protein homeostasis is not going so well, they are more likely to switch. And of course the return is also going to be influenced by stress. So, the protein homeostasis… if cells change… the prion may be really great here, but if the environment changes, under these new circumstances, the prion might not be so good, but under those stressful conditions, they’d be more likely to try out the loss of the prion, because chaperones have been upregulated and they’ve destabilized the whole system. So, we decided to look for new prions. How many new prions might there be? We surveyed the yeast genome and we found that there were about 100 proteins that had domains on them that looked a lot like the Sup35 domain, and so we decided, because we know so much about Sup35 and how it behaves and looks when it changes into a prion state, we borrowed that prion domain of Sup35, we took it away, and we added the prion domain of these other potential new prions. And sure enough, when we tested them out, many, many different protein domains had the ability to switch cells from red to white in a very heritable way. Once the cells switched from red to white, they could be struck out and maintain that characteristic for generation after generation, and in every case that depended upon the formation of prion amyloids, and, in fact, it depended upon Hsp104. So, we also went back, not just to the prion domain that we were testing, but we went back to the endogenous protein and asked whether those proteins could switch states. We couldn’t handle all of them, but we looked at several of them and they created some really interesting beneficial new traits, and intriguingly many of those proteins are RNA binding proteins or DNA binding proteins, so they sit in the middle of regulatory networks in such a way that they’re really primed to change the way information is being decoded and really create some really complex, novel, new traits. Here’s just one example. So, this is the prion known as MOT3, and that’s the prion- cell and that’s the prion+ cells growing in rich media. These are a variety of cells that have the prion. And now we wash away that media… sorry, these cells from that media, and we look to see if any of the cells remain, and in fact the prion, in this case, has allowed many of these cell types to acquire a new invasive growth phenotype. It also has create a capacity, in some of the strains, to flocculate, that is, to group together, which really changes their growth properties. It really makes them, in many ways, function as a community of yeast rather than as individuals. And you can see that in different strains, we’re getting different phenotypes. So, again, a variety of different phenotypes, and MOT3, by the way, is a transcription repressor, so when it goes into the prion state it can alter the expression of lots of different genes. So, this had us excited and we were very, very interested in it, and a lot of other people thought it was very interesting too, but there were also a lot of skeptics. And with good reason, because after all Saccharomyces cerevisiae is the best understood organism on the planet, so the question arises, well, if there’s so many of these things, why weren’t these discovered before? And I have an answer to one aspect of that question, because I took the Cold Spring Harbor yeast course many years ago. It was a wonderful, wonderful course and I learned so much, and it empowered my research in so many ways, but there was one thing that was very interesting about that course. They told us that whenever we found a new phenotype, a new trait in a yeast cell, you should cross it back to the original and then look for it to segregate two-to-two, so that you had some clean system in which to investigate. And things that didn’t segregate two-to-two were complex and would be too difficult to deconstruct, and you shouldn’t work on them. Since we’ve been talking about these prions, I’ve actually had an awful lot of people come up to me and say, “I had these weird traits in yeast that were segregating that way too and my advisor made me give it up.” So, I think this is kind of a wonderful story where we really get into habits of doing science in particular types of ways, and we really need to remember that it’s sometimes time to break away from those. Anyway, the other criticism was, well, so far these things had only been found in laboratory strains and maybe it was just an artifact of laboratory strains, so we went to the same broad group of strains that we’d worked with with our Hsp90 investigations, strains collected from all over the world with all sorts of different properties and lots of different ecological niches, and we asked whether or not any of them had prions. And we found that a lot of them did. So, here’s an example of a prion that exists in a wild strain of yeast, it’s a wine strain, and this is the cells growing in grape must media, and it turned out that we were able to create isogenic cells in which we caused the cells to lose the prion. It didn’t really make any difference, really, to their growth on normal media, rich media that we use in the laboratory, but when you look at their ability to grow in grape must medium, that growth was really dependent upon the prion. So the prion was creating, in this case, a very valuable trait for that strain. And we have found, in fact, that prions very frequently create interesting new traits that can be beneficial. In fact, in the strains that we’ve looked at so far, about 25% of these traits allow the cells to grow under conditions where they otherwise just simply could not grow. So, again, because they normally appear quite rarely and you only lose a small percentage of the cells if they’re not beneficial, but when they do appear they allow cells to grow under conditions where they couldn’t otherwise, we think this is a really plausible bet-hedging strategy for the acquisition of genetic diversity in the organism, creating lots of new phenotypes, which allows it to exploit fluctuating environments and grow in a variety of different situations. And I just showed you one phenotype from PSI; we’ve seen many phenotypes from many other prions as well, and in fact, although we haven’t identified many of the prions that exist in those strains, the genetic behavior tells us that about 236 out of the 700 wild yeast strains we looked at do carry prions. So, the final part of this story that I want to tell you about concerns the ways in which a prion can influence the dynamics of the growth of these microbes in communities. Now, of course, almost all experimental investigations in the laboratory work with organisms in pure culture, so, yeast growing or bacteria growing, and that’s natural that we started out doing experiments that way, because otherwise we have everything all mixed up and it’s just impossible to do the kinds of experimental investigation that we needed to do over the last century to understand biological systems. But organisms never, virtually never, grow under those circumstances in the wild. In the wild, they grow in mixed communities. So, what I’m going to tell you about is just the first case, I think, of a prion influencing the dynamics of community, organization, and organismal communication to create more diverse and more robust communities. It’s the only we’ve really looked at so far and my guess is that that’s just, again, the tip of the iceberg. So, yeast cells have a very particular type of metabolism. They’re very, very fastidious, and that’s the reason why we love them. When we give them glucose, sugars like from grapes, they take those sugars and they convert them into ethanol extremely efficiently. They don’t use the ethanol. They only make the sugars into ethanol; only when all the sugar is exhausted do they start turning around and using the ethanol. So they will not grow on another carbon source if there’s any glucose present — very, very fastidious. But we found a prion that cells can acquire that allows them to bypass this system. They don’t grow quite as well in pure glucose when they have that prion, and that’s probably why this was not really found and looked at before, but in mixed carbon sources where there’s other carbon sources around, and a little bit of glucose too, those cells can now grow when they otherwise would not be able to. So, how did we investigate this story? Well, we took advantage of a glucose mimetic. It’s a compound called glucosamine, that looks just like glucose, probably to most of you, and it does to the yeast cells — it has this little amino group here — but what it does is the yeast cells take this glucosamine up, they think they’re in glucose, and the shut off the pathways for utilization of all other carbon sources. And so here you see an example of this. We’ve got cells growing without the prion. Growing in pure glycerol they’re just fine, but if you try to get them to grow in glycerol with just a little bit of this glucosamine in there, they think, “Oh no, I do not want to use any other carbon source, I only want to use glucose, and I won’t use that other carbon source until all the glucose is gone,” they can’t metabolize it, so they’re just stuck. This allowed us to search for strains that might bypass that metabolic problem, and allow the cells to grow in the presence of other carbon sources when there was a little bit of glucose present. And as you can see, we’ve got strains here that do that very, very well. This strain is genetically identical to this strain and we’ve tested that in many, many ways. Moreover, these cells, once they acquire that ability to grow on that other carbon source, can be processed and passaged for hundreds of generation, back on pure glucose, but they remember. They retain that trait, it’s a biological memory, and they retain that trait for many, many generations. That allows them, next time they’re put on that diverse carbon source, to be able to grow on them. So, it really changes their metabolic potential in a really quite strong way. Now, the cool thing about this in terms of functioning in a community… when you think about it, carbon source utilization strategies might change quite a bit depending on whether you have other neighbors that are competing with you for those carbon sources. And this is something we found actually by accident. Jessica Brown was working on strains, looking at what causes prion switching, and she suddenly found a colony around which there were all kinds of strains that had switched into the prion state, and that colony in the middle was a bacteria. So, in this experiment we tested that more rigorously. Do bacteria have the capacity to secrete a factor that will cause the yeast cells to switch their metabolism in a heritable way by the induction of this prion? So, in this experiment, what we’ve done is we’ve plated… left these wells empty and we’ve plated the gar- yeast cells here, the GAR+ yeast cells — GAR is for glucose-associated repression, I should have said that, that’s the name of that prion — and then we’ve plated gar- cells here. So, these cells, remember, are genetically identical; the only difference between them is that these cells have switched into a prion state that is inherited in a non-Mendelian fashion, because it’s based upon protein-based inheritance. And the cells are being plated on glycerol with, again, just a little bit of glucosamine. And you can see, by the way, that the cells do switch spontaneously, every once in a while, into this prion state, which allows them to grow, but they really only switch that way every once in a while. This is a dilution of strains across this plate. Now, the next experiment is done exactly the same way, except we’re going to plate bacteria that seem to have the capacity to induce this prion here. And what you can see is that the presence of the bacteria on that plate… a substance has diffused down the plate and influenced these yeast cells here to now switch their metabolic state and be able to grow under this condition when they otherwise wouldn’t, but it’s a diffusible compound and it doesn’t get far enough to get all the way down to this row, here. We established by lots of experiments that I won’t take the time to show you, but we could use conditioned media that bacteria had grown in, and get rid of the bacteria, and we could use the conditioned media to induce this heritable trait. So, what happens? The cells usually are using glucose and they make lots of ethanol and only some ATP, but when they’re in the presence of glucose and other carbon sources, this prion now allows them to make less ethanol, but lots of ATP, and what does that do? It causes both organisms to flourish. So, it turns out that when the yeast cells switch, as I mentioned, it allows them to grow on different kinds of carbon sources, even when glucose is present. It also, for reason we don’t quite understand, creats increased longevity in those cells — they can live in culture for much longer periods of time — and the bacteria get a big advantage out of it, because the yeast cells are making less ethanol and that means that the bacteria can grow up to much higher concentrations, because the bacteria don’t like ethanol — it poisons them. Now, it turns out that this is a very highly conserved property of yeast and bacterial strains from all over the world. We’ve found lots and lots of yeast strains are capable of this kind of a switch, and many, many different bacteria are capable of inducing it. But does it matter in the real world? Well, I can tell you one way in which it clearly matters in the real world, and that is it spoils wine fermentations. So, the yeast are making less ethanol, they’re making other kinds of metabolic byproducts, and it really makes lousy wine, and in fact it turns out the bacteria that contaminate spoiled wine preparations, that winologists have been studying for a long period of time, turn out to be super-inducers. They make lots of this compound that switches the yeast cells into this new prion state. So, the yeast get tremendous advantages out of it, the bacteria get tremendous advantages out of it, it’s to the detriment of man, but to the benefit of both those organisms. And again, in terms of the conservation of this, across tens of millions of years of yeast evolution, these strains have really exactly… what seems to be exactly the same mechanisms. Bruxellensis, which has diverged from Saccharomyces… these strains don’t have it, they have a different kind of metabolic profile… bruxellensis is another yeast that has that same kind of fastidious lifestyle in terms of carbon utilization that Saccharomyces has, and it also has that GAR prion-like state. Even pombe, which diverged hundreds of millions of years ago, although the mechanism isn’t quite the same, has the capacity to switch into an epigenetic state that changes its metabolism heritably, so vast amounts of evolutionary distances… these abilities to switch carbon source in a heritable way, induced by environmental factors, has been conserved. So, here we have really what I consider the ultimate example of Lamarckian evolution. We have chemical communication between a prokaryote and a eukaryote that transforms metabolism in a heritable way, and the simple exposure of the yeast cells to that chemical compound causes them to change their biology in a way that’s heritable for hundreds of generations. So, again, when considering Jean-Baptiste Lamarck, I think it’s time to give him back his dignity. There are mechanisms by which changes in the environment can produce new traits that can be passed on to progeny. So, once again, I want to end my lecture by acknowledging the amazing and wonderful people in my laboratory who have done this work over the years. I’ve spoken about things that each of them have done and if you’re interested in this work, I would really urge you to take a look at some of the papers. I think they’re pretty cool. Thank you.